commit 8b81781d2da17944e61530b5fcdc3c564bdbe6e2 Author: ModelHub XC Date: Tue Aug 25 11:51:19 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: eaddario/Hammer2.1-7b-GGUF Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..7428ec4 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,38 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +*.gguf filter=lfs diff=lfs merge=lfs -text +*.logits filter=lfs diff=lfs merge=lfs -text +*.dat filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..86d94a1 --- /dev/null +++ b/.gitignore @@ -0,0 +1,584 @@ +# https://github.com/github/gitignore + +# ------------- +# General Linux +# ------------- +.directory +.fuse_hidden* +.nfs* +.Trash-* +*~ + +# ----------- +# General OSX +# ----------- +._* +.apdisk +.AppleDB +.AppleDesktop +.AppleDouble +.com.apple.timemachine.donotpresent +.DocumentRevisions-V100 +.DS_Store +.fseventsd +.LSOverride +.Spotlight-V100 +.TemporaryItems +.Trashes +.VolumeIcon.icns +Network Trash Folder +Temporary Items + +# --------------- +# General Windows +# --------------- +[Dd]esktop.ini +*.cab +*.lnk +*.msi +*.msix +*.msm +*.msp +*.stackdump +$RECYCLE.BIN/ +ehthumbs_vista.db +ehthumbs.db +Thumbs.db +Thumbs.db:encryptable + +# ----------------- +# General JetBrains +# ----------------- +.idea_modules/ +.idea/ +*.iml +*.ipr +*.iws + +# --------------- +# General VS Code +# --------------- +!.vscode/*.code-snippets +!.vscode/extensions.json +!.vscode/launch.json +!.vscode/settings.json +!.vscode/tasks.json +.history/ +.vscode/ +*.vsix + +# --------------------- +# General Visual Studio +# --------------------- +__pycache__/ +_NCrunch_* +_pkginfo.txt +_Pvt_Extensions +_ReSharper*/ +_TeamCity* +_UpgradeReport_Files/ +!?*.[Cc]ache/ +!.vscode/extensions.json +!**/[Pp]ackages/build/ +.*crunch*.local.xml +.builds +.cr/personal +.fake/ +.ionide/ +.localhistory/ +.mfractor/ +.ntvs_analysis.dat +.paket/paket.exe +.sass-cache/ +.vs/ +.vscode/* +.vshistory/ +[Aa][Rr][Mm]/ +[Aa][Rr][Mm]64/ +[Bb]in/ +[Bb]uild[Ll]og.* +[Dd]ebug/ +[Dd]ebugPS/ +[Dd]ebugPublic/ +[Ee]xpress/ +[Ll]og/ +[Ll]ogs/ +[Oo]bj/ +[Rr]elease/ +[Rr]eleasePS/ +[Rr]eleases/ +[Tt]est[Rr]esult*/ +[Ww][Ii][Nn]32/ +*_h.h +*_i.c +*_p.c +*_wpftmp.csproj +*- [Bb]ackup ([0-9]).rdl +*- [Bb]ackup ([0-9][0-9]).rdl +*- [Bb]ackup.rdl +*.[Cc]ache +*.[Pp]ublish.xml +*.[Rr]e[Ss]harper +*.appx +*.appxbundle +*.appxupload +*.aps +*.azurePubxml +*.bim_*.settings +*.bim.layout +*.binlog +*.btm.cs +*.btp.cs +*.build.csdef +*.cachefile +*.code-workspace +*.coverage +*.coveragexml +*.dbmdl +*.dbproj.schemaview +*.dotCover +*.DotSettings.user +*.dsp +*.dsw +*.e2e +*.GhostDoc.xml +*.gpState +*.ilk +*.iobj +*.ipdb +*.jfm +*.jmconfig +*.ldf +*.mdf +*.meta +*.mm.* +*.ncb +*.ndf +*.nuget.props +*.nuget.targets +*.nupkg +*.nvuser +*.obj +*.odx.cs +*.opendb +*.opensdf +*.opt +*.pch +*.pdb +*.pfx +*.pgc +*.pgd +*.pidb +*.plg +*.psess +*.publishproj +*.publishsettings +*.pubxml +*.pyc +*.rdl.data +*.rptproj.bak +*.rptproj.rsuser +*.rsp +*.rsuser +*.sap +*.sbr +*.scc +*.sdf +*.sln.docstates +*.sln.iml +*.snupkg +*.suo +*.svclog +*.tlb +*.tlh +*.tli +*.tlog +*.tmp +*.tmp_proj +*.tss +*.user +*.userosscache +*.userprefs +*.vbp +*.vbw +*.VC.db +*.VC.VC.opendb +*.VisualState.xml +*.vsp +*.vspscc +*.vspx +*.vssscc +*.xsd.cs +**/[Pp]ackages/* +**/*.DesktopClient/GeneratedArtifacts +**/*.DesktopClient/ModelManifest.xml +**/*.HTMLClient/GeneratedArtifacts +**/*.Server/GeneratedArtifacts +**/*.Server/ModelManifest.xml +~$* +$tf/ +AppPackages/ +artifacts/ +ASALocalRun/ +AutoTest.Net/ +Backup*/ +BenchmarkDotNet.Artifacts/ +bld/ +BundleArtifacts/ +ClientBin/ +coverage*.info +coverage*.json +coverage*.xml +csx/ +dlldata.c +DocProject/buildhelp/ +DocProject/Help/*.hhc +DocProject/Help/*.hhk +DocProject/Help/*.hhp +DocProject/Help/*.HxC +DocProject/Help/*.HxT +DocProject/Help/html +DocProject/Help/Html2 +ecf/ +FakesAssemblies/ +FodyWeavers.xsd +Generated_Code/ +healthchecksdb +ipch/ +MigrationBackup/ +mono_crash.* +nCrunchTemp_* +node_modules/ +nunit-*.xml +OpenCover/ +orleans.codegen.cs +Package.StoreAssociation.xml +paket-files/ +project.fragment.lock.json +project.lock.json +publish/ +PublishScripts/ +rcf/ +ScaffoldingReadMe.txt +ServiceFabricBackup/ +StyleCopReport.xml +TestResult.xml +UpgradeLog*.htm +UpgradeLog*.XML +x64/ +x86/ + +# ---------------------- +# General Archived Files +# ---------------------- +*.7z +*.bz2 +*.bzip +*.bzip2 +*.deb +*.dmg +*.egg +*.gem +*.gz +*.gzip +*.iso +*.jar +*.lzma +*.rar +*.rpm +*.tar +*.tgz +*.txz +*.tzst +*.xar +*.xpi +*.xz +*.zip +*.zst + +# ----- +# C/C++ +# ----- +.tmp_versions/ +*.a +*.app +*.cmd +*.d +*.dll +*.dSYM/ +*.dylib +*.elf +*.exe +*.exp +*.gch +*.hex +*.i*86 +*.idb +*.ko +*.la +*.lai +*.lib +*.lo +*.map +*.mod* +*.o +*.out +*.slo +*.so +*.so.* +*.su +*.x86_64 +dkms.conf +Mkfile.old +Module.symvers +modules.order + +# ---- +# CUDA +# ---- +*.cubin +*.fatbin +*.gpu +*.i +*.ii +*.ptx + +# -- +# Go +# -- +.env +*.exe~ +*.test +go.work +go.work.sum + +# ---- +# Java +# ---- +*.class +*.ctxt +*.ear +*.hprof +*.nar +*.tar.gz +*.war +hs_err_pid* +replay_pid* + +# ----- +# Julia +# ----- +*.jl.*.cov +*.jl.cov +*.jl.mem +deps/build.log +deps/deps.jl +deps/downloads/ +deps/src/ +deps/usr/ +docs/build/ +docs/site/ +Manifest.toml + +# ------------------------------ +# JavaScript / Node / TypeScript +# ------------------------------ +.cache +.cache/ +.docusaurus +.dynamodb/ +.env.development.local +.env.local +.env.production.local +.env.test.local +.eslintcache +.fusebox/ +.grunt +.lock-wscript +.next +.node_repl_history +.npm +.nuxt +.nyc_output +.parcel-cache +.pnp.* +.pnpm-debug.log* +.rpt2_cache/ +.rts2_cache_cjs/ +.rts2_cache_es/ +.rts2_cache_umd/ +.serverless/ +.stylelintcache +.temp +.tern-port +.vscode-test +.vuepress/dist +.yarn-integrity +.yarn/build-state.yml +.yarn/cache +.yarn/install-state.gz +.yarn/unplugged +*.lcov +*.pid +*.pid.lock +*.seed +*.tsbuildinfo +bower_components +dist +jspm_packages/ +lerna-debug.log* +lib-cov +logs +npm-debug.log* +out +pids +report.[0-9]*.[0-9]*.[0-9]*.[0-9]*.json +web_modules/ +yarn-debug.log* +yarn-error.log* + +# ------ +# Python +# ------ +__pypackages__/ +.coverage +.coverage.* +.dmypy.json +.eggs/ +.hypothesis/ +.installed.cfg +.ipynb_checkpoints +.mypy_cache/ +.nox/ +.pdm-build/ +.pdm-python +.pdm.toml +.pybuilder/ +.pyre/ +.pytest_cache/ +.Python +.python-version +.pytype/ +.ropeproject +.scrapy +.spyderproject +.spyproject +.tox/ +.venv +.webassets-cache +*.cover +*.egg-info/ +*.manifest +*.mo +*.pot +*.py,cover +*.py[cod] +*.sage.py +*.spec +*$py.class +/site +build/ +celerybeat-schedule +celerybeat.pid +cover/ +coverage.xml +cython_debug/ +db.sqlite3 +db.sqlite3-journal +develop-eggs/ +dist/ +dmypy.json +docs/_build/ +downloads/ +eggs/ +env.bak/ +env/ +htmlcov/ +instance/ +ipython_config.py +lib/ +lib64/ +local_settings.py +MANIFEST +nosetests.xml +parts/ +pip-delete-this-directory.txt +pip-log.txt +profile_default/ +sdist/ +share/python-wheels/ +target/ +var/ +venv.bak/ +venv/ +wheels/ + +# ---- +# Rust +# ---- +**/*.rs.bk +Cargo.lock +debug/ + +# ----- +# Scala +# ----- + +# ----- +# CMake +# ----- +_deps +cmake_install.cmake +CMakeCache.txt +CMakeFiles +CMakeLists.txt.user +CMakeScripts +CMakeUserPresets.json +compile_commands.json +CTestTestfile.cmake +install_manifest.txt +Makefile +Testing + +# ------ +# Gradle +# ------ +!gradle-wrapper.jar +!gradle-wrapper.properties +!src/**/build/ +.classpath +.gradle +.gradletasknamecache +.project +**/build/ +gradle-app.setting + +# ----- +# Maven +# ----- +.mvn/ +buildNumber.properties +dependency-reduced-pom.xml +pom.xml.next +pom.xml.releaseBackup +pom.xml.tag +pom.xml.versionsBackup +release.properties + +# --------- +# Terraform +# --------- +.terraform.tfstate.lock.info +.terraform/ +.terraformrc +*_override.tf +*_override.tf.json +*.tfstate +*.tfstate.* +*.tfvars +*.tfvars.json +crash.*.log +crash.log +override.tf +override.tf.json +terraform.rc \ No newline at end of file diff --git a/Hammer2.1-7b-F16.gguf b/Hammer2.1-7b-F16.gguf new file mode 100644 index 0000000..6589aaf --- /dev/null +++ b/Hammer2.1-7b-F16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d414b3f9607335467c199f2378d25d8f19bcb465376412aca4c2742f5a90452d +size 15232123808 diff --git a/Hammer2.1-7b-IQ3_M.gguf b/Hammer2.1-7b-IQ3_M.gguf new file mode 100644 index 0000000..c01f906 --- /dev/null +++ b/Hammer2.1-7b-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f463f9b3bec8377f80cea123711fcdab55613f4f0ff352091e9280ba048dbe4b +size 3483159680 diff --git a/Hammer2.1-7b-IQ3_S.gguf b/Hammer2.1-7b-IQ3_S.gguf new file mode 100644 index 0000000..8525962 --- /dev/null +++ b/Hammer2.1-7b-IQ3_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:97cb3efe4747109275b82568ad6b4751b035befa08dc22d1e8e80a5dc3ea6476 +size 3254777856 diff --git a/Hammer2.1-7b-IQ4_NL.gguf b/Hammer2.1-7b-IQ4_NL.gguf new file mode 100644 index 0000000..7cd1d94 --- /dev/null +++ b/Hammer2.1-7b-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:88a7050e28853a0249d05c75f37747628867822c182d302da4c101eddb3ea95e +size 4174122624 diff --git a/Hammer2.1-7b-Q3_K_L.gguf b/Hammer2.1-7b-Q3_K_L.gguf new file mode 100644 index 0000000..a5e34ce --- /dev/null +++ b/Hammer2.1-7b-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:033d0484dc19b37492f9b96022d9777702b5809a749127c2a7bd1e4f8c5cb536 +size 3535142016 diff --git a/Hammer2.1-7b-Q3_K_M.gguf b/Hammer2.1-7b-Q3_K_M.gguf new file mode 100644 index 0000000..ef56abe --- /dev/null +++ b/Hammer2.1-7b-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:676d89f69276fe589b30c458f509aac0f50e81590ff4fbcfa4c729d30bdc147f +size 3367955584 diff --git a/Hammer2.1-7b-Q3_K_S.gguf b/Hammer2.1-7b-Q3_K_S.gguf new file mode 100644 index 0000000..a7e4ca9 --- /dev/null +++ b/Hammer2.1-7b-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73cf6d45841ed732d1b000b878ba00c1043f671c76a959a66d3002f038d458c +size 3135327168 diff --git a/Hammer2.1-7b-Q4_K_M.gguf b/Hammer2.1-7b-Q4_K_M.gguf new file mode 100644 index 0000000..2c89816 --- /dev/null +++ b/Hammer2.1-7b-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cd698cb286d02e407ef41b2304f4b7fd1f7d45eece1780120016e5fbc41e8020 +size 4178136704 diff --git a/Hammer2.1-7b-Q4_K_S.gguf b/Hammer2.1-7b-Q4_K_S.gguf new file mode 100644 index 0000000..c069eaf --- /dev/null +++ b/Hammer2.1-7b-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d01de17d7bf4309e0e5aad479c6b2d69e56ba9ca9814c16d1318dd814308090c +size 4056108672 diff --git a/Hammer2.1-7b-Q5_K_M.gguf b/Hammer2.1-7b-Q5_K_M.gguf new file mode 100644 index 0000000..8f4634d --- /dev/null +++ b/Hammer2.1-7b-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb13f56264f87a5e1bf15d82cb7bc32218b83995fe562da4cce6f659bc18684f +size 5093254208 diff --git a/Hammer2.1-7b-Q5_K_S.gguf b/Hammer2.1-7b-Q5_K_S.gguf new file mode 100644 index 0000000..bd63758 --- /dev/null +++ b/Hammer2.1-7b-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d7940c008124ba606a6c60381c78f24c92937e3a547d61e1aa7332295cba0ea7 +size 4963599424 diff --git a/Hammer2.1-7b-Q6_K.gguf b/Hammer2.1-7b-Q6_K.gguf new file mode 100644 index 0000000..5d75aec --- /dev/null +++ b/Hammer2.1-7b-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a566f0f780e41ae072ba7f40cf2d660c12ad0fc8fcd54de0ea9d50c90d913f72 +size 6226368096 diff --git a/Hammer2.1-7b-Q8_0.gguf b/Hammer2.1-7b-Q8_0.gguf new file mode 100644 index 0000000..44d1289 --- /dev/null +++ b/Hammer2.1-7b-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac8f52cffb377844263f785b4a18409cb7e7dcac1882952d802df2b5af851854 +size 7265396096 diff --git a/README.md b/README.md new file mode 100644 index 0000000..86f0732 --- /dev/null +++ b/README.md @@ -0,0 +1,143 @@ +--- +base_model: +- MadeAgents/Hammer2.1-7b +datasets: +- eaddario/imatrix-calibration +language: +- en +license: +- cc-by-nc-4.0 +pipeline_tag: text-generation +tags: +- gguf +- quant +- experimental +--- + +# Experimental layer-wise quantization of MadeAgents/Hammer2.1-7b + +Using [LLaMA C++]() release [b5170]() for quantization. + +Original model: [MadeAgents/Hammer2.1-7b](https://huggingface.co/MadeAgents/Hammer2.1-7b) + +From the original model creators: + +> Hammer refers to a series of lightweight Large Action Models. Currently, we are releasing Hammer 2.1 models ([0.5B](https://huggingface.co/MadeAgents/Hammer2.1-0.5b), [1.5B](https://huggingface.co/MadeAgents/Hammer2.1-1.5b), [3B](https://huggingface.co/MadeAgents/Hammer2.1-3b), and [7B](https://huggingface.co/MadeAgents/Hammer2.1-7b)) with strong function calling capability. These models are based on the Qwen 2.5 coder series and utilize [function masking techniques](https://arxiv.org/abs/2410.04587) and other advanced technologies. Hammer 2.1 series bring significant enhancements, while still maintaining the basic functionality of Hammer 2.0's Single-Turn interaction and further strengthening other capabilities. +> +> The Hammer 2.1 models, fine-tuned from the Qwen 2.5 coder series, inherit Hammer 2.0's advantages and are enhanced as follows: +> - Multi-Step Function Calling: The assistant can perform multiple internal function calls to handle a single user request, actively planning and gathering information to fulfill complex tasks. +> - Multi-Turn Function Calling: Enables continuous and context-aware interactions over multiple exchanges, with each turn potentially containing multiple steps, for a more natural conversation experience. +> - Enhanced Irrelevant Information Inspection: Better at identifying when provided functions are irrelevant to a user query, by providing a non-function call response. + +# PLEASE READ THIS BEFORE USING THESE EXPERIMENTAL VERSIONS! + +An area of personal interest is finding ways to optimize the inference performance of LLMs when deployed in resource-constrained environments like commodity hardware, desktops, laptops, mobiles, edge devices, etc. There are many approaches to accomplish this, including architecture simplification and knowledge distillation, but my focus has been primarily on quantization and pruning. + +The method used to produce these experimental versions is covered in [Squeezing Tensor Bits: the quest for smaller LLMs](), but at a high level it involves using custom versions of `llama-imatrix` and `llama-quantize` to identify the influential tensors, and quantize the most important layers to higher bit precision and the less important to lower bits. This process was partly inspired by Dumitru's et al [Layer-Wise Quantization: A Pragmatic and Effective Method for Quantizing LLMs Beyond Integer Bit-Levels](). + +There’re two pull requests ([imatrix]() & [quantize]()) to merge these changes back into the core llama.cpp project. This may or may not ever happen so, until then, the modified versions will be available on [GitHub](). + +For testing and comparison I'd normally use models produced by [Unsloth]() ([Daniel and Michael Han]() do some really advanced level stuff!) and [Bartowski]() (see credits below), but they don't provide GGUF versions of this model, so all tests and comparisons are done against naive quantizations obtained by simply running `llama-quantize` with no further optimization. + +All experimental versions were generated using an appropriate imatrix created from calibration datasets available at [eaddario/imatrix-calibration](). At its core, an Importance Matrix (imatrix) is a table or, more broadly, a structured representation that scores the relative importance of different features or parameters in a machine learning model. It essentially quantifies the "impact" each feature has on a specific outcome, prediction, or relationship being modeled, and it helps to counterbalance the negative effects of quantization and pruning. + +The process to generate these models is roughly as follows: + +1. Convert the the original model's tensors to [GGUF]() F16* +2. Estimate the Perplexity score for the F16 model (baseline) using the [wikitext-2-raw-v1]() dataset, and save the [logits]() +3. Generate an [imatrix]() from selected calibration datasets +4. Determine tensor and layer Importance Score contribution using a modified version of `llama-imatrix` +5. Select an appropiate quant level for each tensor using a modified version of `llama-quantize` +6. Calculate Perplexity, KL Divergence, ARC (Easy+Challenge), HellaSwag, MMLU, Truthful QA and WinoGrande scores for each quantized model +7. Keep versions with the best scores +8. Repeat until all desired quants are created. I find that quantizations below Q3/IQ3 are not fit for my purposes and therefore do not usually generate them, but happy to provide other quants on request. + +*[BF16]() would be preferred, but Apple's GPUs don't support it yet, and therefore any operations are executed in the CPU, making it unacceptably slow. This is expected to change in the near term but until then, if you are using Apple kit avoid using any models tagged BF16 + +# Models + +### Sizes (in GB) +| Model | Naive | Repo | Shrinkage | +| ------------------------------------------------- | ----: | ---: | --------: | +| [Hammer2.1-7b-IQ3_M](./Hammer2.1-7b-IQ3_M.gguf) | 3.57 | 3.48 | 2.5% | +| [Hammer2.1-7b-IQ3_S](./Hammer2.1-7b-IQ3_S.gguf) | 3.50 | 3.25 | 7.1% | +| [Hammer2.1-7b-IQ4_NL](./Hammer2.1-7b-IQ4_NL.gguf) | 4.44 | 4.17 | 6.1% | +| [Hammer2.1-7b-Q3_K_L](./Hammer2.1-7b-Q3_K_L.gguf) | 4.09 | 3.54 | 13.4% | +| [Hammer2.1-7b-Q3_K_M](./Hammer2.1-7b-Q3_K_M.gguf) | 3.81 | 3.37 | 11.5% | +| [Hammer2.1-7b-Q3_K_S](./Hammer2.1-7b-Q3_K_S.gguf) | 3.49 | 3.14 | 10.0% | +| [Hammer2.1-7b-Q4_K_M](./Hammer2.1-7b-Q4_K_M.gguf) | 4.68 | 4.18 | 10.7% | +| [Hammer2.1-7b-Q4_K_S](./Hammer2.1-7b-Q4_K_S.gguf) | 4.46 | 4.06 | 9.0% | +| [Hammer2.1-7b-Q5_K_M](./Hammer2.1-7b-Q5_K_M.gguf) | 5.44 | 5.09 | 6.4% | +| [Hammer2.1-7b-Q5_K_S](./Hammer2.1-7b-Q5_K_S.gguf) | 5.31 | 4.96 | 6.6% | +| [Hammer2.1-7b-Q6_K](./Hammer2.1-7b-Q6_K.gguf) | 6.25 | 6.23 | 0.3% | +| [Hammer2.1-7b-Q8_0](./Hammer2.1-7b-Q8_0.gguf) | 8.10 | 7.27 | 10.2% | + +### Perplexity and KL Divergence scores +| Model | μPPL | 𝜌PPL | μKLD | RMS Δp | +| ------------------------------------------------- | ------------------: | -----: | -----------------: | -----------: | +| [Hammer2.1-7b-IQ3_M](./Hammer2.1-7b-IQ3_M.gguf) | 9.827809 ±0.071236 | 98.97% | 0.059970 ±0.000226 | 6.574 ±0.032 | +| [Hammer2.1-7b-IQ3_S](./Hammer2.1-7b-IQ3_S.gguf) | 9.942560 ±0.071243 | 98.64% | 0.079935 ±0.000289 | 7.445 ±0.038 | +| [Hammer2.1-7b-IQ4_NL](./Hammer2.1-7b-IQ4_NL.gguf) | 9.585928 ±0.068060 | 99.51% | 0.026627 ±0.000108 | 4.346 ±0.026 | +| [Hammer2.1-7b-Q3_K_L](./Hammer2.1-7b-Q3_K_L.gguf) | 10.010842 ±0.071669 | 98.55% | 0.080230 ±0.000317 | 7.556 ±0.040 | +| [Hammer2.1-7b-Q3_K_M](./Hammer2.1-7b-Q3_K_M.gguf) | 10.109660 ±0.072549 | 98.44% | 0.086444 ±0.000342 | 7.845 ±0.041 | +| [Hammer2.1-7b-Q3_K_S](./Hammer2.1-7b-Q3_K_S.gguf) | 10.224307 ±0.073520 | 98.16% | 0.101951 ±0.000397 | 8.487 ±0.044 | +| [Hammer2.1-7b-Q4_K_M](./Hammer2.1-7b-Q4_K_M.gguf) | 9.513981 ±0.067589 | 99.56% | 0.023792 ±0.000107 | 4.057 ±0.026 | +| Hammer2.1-7b-Q4_K_M (naive) | 9.469517 ±0.067164 | 99.70% | 0.016770 ±0.000081 | 3.364 ±0.022 | +| [Hammer2.1-7b-Q4_K_S](./Hammer2.1-7b-Q4_K_S.gguf) | 9.529467 ±0.067693 | 99.53% | 0.025429 ±0.000116 | 4.191 ±0.026 | +| [Hammer2.1-7b-Q5_K_M](./Hammer2.1-7b-Q5_K_M.gguf) | 9.411472 ±0.066924 | 99.88% | 0.005881 ±0.000035 | 2.028 ±0.015 | +| [Hammer2.1-7b-Q5_K_S](./Hammer2.1-7b-Q5_K_S.gguf) | 9.410802 ±0.066883 | 99.88% | 0.006252 ±0.000036 | 2.085 ±0.015 | +| [Hammer2.1-7b-Q6_K](./Hammer2.1-7b-Q6_K.gguf) | 9.384561 ±0.066615 | 99.96% | 0.001779 ±0.000018 | 1.101 ±0.012 | +| [Hammer2.1-7b-Q8_0](./Hammer2.1-7b-Q8_0.gguf) | 9.379380 ±0.066571 | 99.98% | 0.000692 ±0.000012 | 0.706 ±0.010 | +| [Hammer2.1-7b-F16](./Hammer2.1-7b-F16.gguf) | 9.366577 ±0.066397 | 100% | N/A | N/A | + +### ARC, HellaSwag, MMLU, Truthful QA and WinoGrande scores +Scores generated using [llama-perplexity]() with 750 tasks per test, and a context size of 768 tokens. + +For the test data used in the generation of these scores, follow the appropiate links: [HellaSwag](), [ARC, MMLU, Truthful QA]() and [WinoGrande]() + +| Model | ARC | HellaSwag | MMLU | Truthful QA | WinoGrande | Avg Score | +| ------------------------------------------------- | --------------: | --------: | --------------: | --------------: | --------------: | --------: | +| [Hammer2.1-7b-IQ3_M](./Hammer2.1-7b-IQ3_M.gguf) | 56.5333 ±1.8113 | 71.46 | 38.0000 ±1.7736 | 29.8667 ±1.6723 | 70.6667 ±1.6636 | 53.31 | +| [Hammer2.1-7b-IQ3_S](./Hammer2.1-7b-IQ3_S.gguf) | 55.8667 ±1.8143 | 70.26 | 38.6667 ±1.7794 | 30.9333 ±1.6889 | 70.4000 ±1.6680 | 53.23 | +| [Hammer2.1-7b-IQ4_NL](./Hammer2.1-7b-IQ4_NL.gguf) | 57.3333 ±1.8072 | 71.20 | 38.5333 ±1.7783 | 30.1333 ±1.6766 | 69.8667 ±1.6766 | 53.41 | +| [Hammer2.1-7b-Q3_K_L](./Hammer2.1-7b-Q3_K_L.gguf) | 57.2000 ±1.8079 | 69.86 | 36.6667 ±1.7608 | 29.6000 ±1.6680 | 70.9333 ±1.6591 | 52.85 | +| [Hammer2.1-7b-Q3_K_M](./Hammer2.1-7b-Q3_K_M.gguf) | 55.7333 ±1.8149 | 69.87 | 37.2000 ±1.7661 | 29.8667 ±1.6723 | 70.4000 ±1.6680 | 52.61 | +| [Hammer2.1-7b-Q3_K_S](./Hammer2.1-7b-Q3_K_S.gguf) | 56.5333 ±1.8113 | 68.80 | 37.3333 ±1.7674 | 29.0667 ±1.6591 | 69.3333 ±1.6849 | 52.21 | +| [Hammer2.1-7b-Q4_K_M](./Hammer2.1-7b-Q4_K_M.gguf) | 57.8667 ±1.8042 | 71.20 | 38.0000 ±1.7736 | 29.8667 ±1.6723 | 69.8667 ±1.6766 | 53.36 | +| Hammer2.1-7b-Q4_K_M (naive) | 57.8313 ±1.8080 | 73.47 | 35.0667 ±1.7436 | 33.9683 ±2.6727 | 70.8000 ±1.6614 | 54.23 | +| [Hammer2.1-7b-Q4_K_S](./Hammer2.1-7b-Q4_K_S.gguf) | 57.8667 ±1.8042 | 71.07 | 37.8667 ±1.7724 | 29.6000 ±1.6680 | 69.6000 ±1.6807 | 53.20 | +| [Hammer2.1-7b-Q5_K_M](./Hammer2.1-7b-Q5_K_M.gguf) | 58.0000 ±1.8034 | 71.33 | 38.6667 ±1.7794 | 30.2667 ±1.6787 | 70.6667 ±1.6636 | 53.79 | +| [Hammer2.1-7b-Q5_K_S](./Hammer2.1-7b-Q5_K_S.gguf) | 58.4000 ±1.8010 | 71.47 | 38.8000 ±1.7805 | 30.5333 ±1.6828 | 70.6667 ±1.6636 | 53.97 | +| [Hammer2.1-7b-Q6_K](./Hammer2.1-7b-Q6_K.gguf) | 58.8000 ±1.7984 | 72.00 | 38.9333 ±1.7816 | 29.7333 ±1.6702 | 70.4000 ±1.6680 | 53.97 | +| [Hammer2.1-7b-Q8_0](./Hammer2.1-7b-Q8_0.gguf) | 58.5333 ±1.8002 | 72.00 | 38.9333 ±1.7816 | 30.1333 ±1.6766 | 70.2667 ±1.6702 | 53.97 | +| [Hammer2.1-7b-F16](./Hammer2.1-7b-F16.gguf) | 58.9333 ±1.7976 | 72.13 | 39.0667 ±1.7827 | 30.4000 ±1.6807 | 70.5333 ±1.6658 | 54.21 | + +### Tokens per Second - Benchmarks +Scores generated using [llama-bench](https://github.com/ggml-org/llama.cpp/tree/master/examples/llama-bench). Naive Q4_K_M quantization included for comparison. + +| model | size | params | backend | threads | test | t/s | +| ------------------------------------------------- | -------: | -----: | ---------- | ------: | ------------: | ------------: | +| [Hammer2.1-7b-Q4_K_M](./Hammer2.1-7b-Q4_K_M.gguf) | 3.89 GiB | 7.61 B | Metal,BLAS | 6 | pp512 | 336.11 ± 0.60 | +| [Hammer2.1-7b-Q4_K_M](./Hammer2.1-7b-Q4_K_M.gguf) | 3.89 GiB | 7.61 B | Metal,BLAS | 6 | tg128 | 29.32 ± 0.14 | +| [Hammer2.1-7b-Q4_K_M](./Hammer2.1-7b-Q4_K_M.gguf) | 3.89 GiB | 7.61 B | Metal,BLAS | 6 | pp1024+tg1024 | 48.00 ± 0.16 | +| Hammer2.1-7b-Q4_K_M (naive) | 4.35 GiB | 7.61 B | Metal,BLAS | 6 | pp512 | 355.08 ± 0.19 | +| Hammer2.1-7b-Q4_K_M (naive) | 4.35 GiB | 7.61 B | Metal,BLAS | 6 | tg128 | 28.21 ± 0.00 | +| Hammer2.1-7b-Q4_K_M (naive) | 4.35 GiB | 7.61 B | Metal,BLAS | 6 | pp1024+tg1024 | 45.92 ± 1.18 | + +# Metrics used +**[Perplexity]():** one of the key metrics used in NLP evaluation. It measures the quality of a language model by evaluating how well it predicts the next token given a particular sequence of words. A PPL of **1** indicates an exact match between predicted and actual, whereas values greater than one indicate a degree of "surprise" the generated token differs from the expected. + +**[Kullback–Leibler (KL) Divergence]():** a statistical measure of how much a probability distribution differs from another. When quantizing models (or altering the original tensors in any way for that matter), the closest we can preserve the weights' probability distribution to the original model the better, thus the closest to **0** the better. + +**[AI2 Reasoning Challenge (ARC)]():** a benchmark to evaluate the ability of AI models to answer complex science questions that require logical reasoning beyond pattern matching. + +**[HellaSwag]():** the Harder Endings, Longer contexts, and Low-shot Activities for Situations With Adversarial Generations (bit of a mouthful!) is a benchmark designed to test commonsense natural language inference. It requires the model to predict the most likely ending of a sentence. + +**[MMLU]():** the Massive Multitask Language Understanding evaluates LLMs’ general knowledge and problem-solving abilities across 57 subjects, including elementary mathematics, US history, computer science, and law. + +**[Truthful QA]():** evaluates how well LLMs generate truthful responses to questions. It identifies whether AI models can avoid generating false or misleading information, particularly in areas where human knowledge is prone to misconceptions. + +**[Winogrande]():** based on the [Winograd Schema Challenge](), is a natural language understanding task requiring models to resolve ambiguities in sentences involving pronoun references. + +## Credits +A big **Thank You!** to [Colin Kealty]() for the many contributions and for being one of the best sources of high quality quantized models available in Hugginface, and a really big ***Thank You!*** to [Georgi Gerganov]() for his amazing work with **llama.cpp** and the **ggml/gguf** libraries. diff --git a/imatrix/imatrix-Hammer2.1-7b-micro.dat b/imatrix/imatrix-Hammer2.1-7b-micro.dat new file mode 100644 index 0000000..b3a3574 --- /dev/null +++ b/imatrix/imatrix-Hammer2.1-7b-micro.dat @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d33304f929548e097be5afdc04c00ab91a729ec3cfbb30639fae5555f56dd65 +size 4551051 diff --git a/imatrix/imatrix-Hammer2.1-7b-small.dat b/imatrix/imatrix-Hammer2.1-7b-small.dat new file mode 100644 index 0000000..a8cbe98 --- /dev/null +++ b/imatrix/imatrix-Hammer2.1-7b-small.dat @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7866bac797b7678785e273994874f59fbf2a331dfc0e0e5fd0ec7fcecea754ff +size 4551051 diff --git a/imatrix/imatrix-Hammer2.1-7b-tiny.dat b/imatrix/imatrix-Hammer2.1-7b-tiny.dat new file mode 100644 index 0000000..65317f8 --- /dev/null +++ b/imatrix/imatrix-Hammer2.1-7b-tiny.dat @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e2b2c0fa25b7286cea883300aa904f099ffd60e8630da388bc492c8ed6be0a1 +size 4551050 diff --git a/logits/Hammer2.1-7b-F16.logits b/logits/Hammer2.1-7b-F16.logits new file mode 100644 index 0000000..f242f52 --- /dev/null +++ b/logits/Hammer2.1-7b-F16.logits @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c2428621962d4b5693e38f2e0a470cce82b7b65283f9e0b753d5aaabfe4e6aa +size 45194911608 diff --git a/scores/Hammer2.1-7b-F16.arc b/scores/Hammer2.1-7b-F16.arc new file mode 100644 index 0000000..21735f2 --- /dev/null +++ b/scores/Hammer2.1-7b-F16.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 29 key-value pairs and 339 tensors from ./Hammer2.1-7b-F16.gguf (version GGUF V3 (latest)) + +Final result: 58.9333 +/- 1.7976 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 6229.12 ms +llama_perf_context_print: prompt eval time = 102811.05 ms / 35972 tokens ( 2.86 ms per token, 349.88 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 103951.36 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-F16.hsw b/scores/Hammer2.1-7b-F16.hsw new file mode 100644 index 0000000..75c2320 --- /dev/null +++ b/scores/Hammer2.1-7b-F16.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 29 key-value pairs and 339 tensors from ./Hammer2.1-7b-F16.gguf (version GGUF V3 (latest)) + +750 72.13333333% [68.8181%, 75.2230%] + + +llama_perf_context_print: load time = 488.78 ms +llama_perf_context_print: prompt eval time = 371768.51 ms / 126038 tokens ( 2.95 ms per token, 339.02 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 377711.49 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-F16.md b/scores/Hammer2.1-7b-F16.md new file mode 100644 index 0000000..5813a63 --- /dev/null +++ b/scores/Hammer2.1-7b-F16.md @@ -0,0 +1,1025 @@ +# Hammer2.1-7b-F16.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 32 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 29 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | UINT32 | 1 | general.file_type | 1 | +| 19 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 20 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 21 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 22 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 23 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 24 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 25 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 26 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 27 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 28 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 29 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 30 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 31 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 32 | UINT32 | 1 | general.quantization_version | 2 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-F16.gguf - GGUF Internal File Dump](#hammer21-7b-f16gguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | token_embd.weight | 0x5ab3a0 | 0x40cc5c00 | +| 1 | blk.0.attn_norm.weight | 0x41270fa0 | 0x3800 | +| 2 | blk.0.ffn_down.weight | 0x412747a0 | 0x8180000 | +| 3 | blk.0.ffn_gate.weight | 0x493f47a0 | 0x8180000 | +| 4 | blk.0.ffn_up.weight | 0x515747a0 | 0x8180000 | +| 5 | blk.0.ffn_norm.weight | 0x596f47a0 | 0x3800 | +| 6 | blk.0.attn_k.bias | 0x596f7fa0 | 0x800 | +| 7 | blk.0.attn_k.weight | 0x596f87a0 | 0x380000 | +| 8 | blk.0.attn_output.weight | 0x59a787a0 | 0x1880000 | +| 9 | blk.0.attn_q.bias | 0x5b2f87a0 | 0x3800 | +| 10 | blk.0.attn_q.weight | 0x5b2fbfa0 | 0x1880000 | +| 11 | blk.0.attn_v.bias | 0x5cb7bfa0 | 0x800 | +| 12 | blk.0.attn_v.weight | 0x5cb7c7a0 | 0x380000 | +| 13 | blk.1.attn_norm.weight | 0x5cefc7a0 | 0x3800 | +| 14 | blk.1.ffn_down.weight | 0x5cefffa0 | 0x8180000 | +| 15 | blk.1.ffn_gate.weight | 0x6507ffa0 | 0x8180000 | +| 16 | blk.1.ffn_up.weight | 0x6d1fffa0 | 0x8180000 | +| 17 | blk.1.ffn_norm.weight | 0x7537ffa0 | 0x3800 | +| 18 | blk.1.attn_k.bias | 0x753837a0 | 0x800 | +| 19 | blk.1.attn_k.weight | 0x75383fa0 | 0x380000 | +| 20 | blk.1.attn_output.weight | 0x75703fa0 | 0x1880000 | +| 21 | blk.1.attn_q.bias | 0x76f83fa0 | 0x3800 | +| 22 | blk.1.attn_q.weight | 0x76f877a0 | 0x1880000 | +| 23 | blk.1.attn_v.bias | 0x788077a0 | 0x800 | +| 24 | blk.1.attn_v.weight | 0x78807fa0 | 0x380000 | +| 25 | blk.2.attn_norm.weight | 0x78b87fa0 | 0x3800 | +| 26 | blk.2.ffn_down.weight | 0x78b8b7a0 | 0x8180000 | +| 27 | blk.2.ffn_gate.weight | 0x80d0b7a0 | 0x8180000 | +| 28 | blk.2.ffn_up.weight | 0x88e8b7a0 | 0x8180000 | +| 29 | blk.2.ffn_norm.weight | 0x9100b7a0 | 0x3800 | +| 30 | blk.2.attn_k.bias | 0x9100efa0 | 0x800 | +| 31 | blk.2.attn_k.weight | 0x9100f7a0 | 0x380000 | +| 32 | blk.2.attn_output.weight | 0x9138f7a0 | 0x1880000 | +| 33 | blk.2.attn_q.bias | 0x92c0f7a0 | 0x3800 | +| 34 | blk.2.attn_q.weight | 0x92c12fa0 | 0x1880000 | +| 35 | blk.2.attn_v.bias | 0x94492fa0 | 0x800 | +| 36 | blk.2.attn_v.weight | 0x944937a0 | 0x380000 | +| 37 | blk.3.attn_norm.weight | 0x948137a0 | 0x3800 | +| 38 | blk.3.ffn_down.weight | 0x94816fa0 | 0x8180000 | +| 39 | blk.3.ffn_gate.weight | 0x9c996fa0 | 0x8180000 | +| 40 | blk.3.ffn_up.weight | 0xa4b16fa0 | 0x8180000 | +| 41 | blk.3.ffn_norm.weight | 0xacc96fa0 | 0x3800 | +| 42 | blk.3.attn_k.bias | 0xacc9a7a0 | 0x800 | +| 43 | blk.3.attn_k.weight | 0xacc9afa0 | 0x380000 | +| 44 | blk.3.attn_output.weight | 0xad01afa0 | 0x1880000 | +| 45 | blk.3.attn_q.bias | 0xae89afa0 | 0x3800 | +| 46 | blk.3.attn_q.weight | 0xae89e7a0 | 0x1880000 | +| 47 | blk.3.attn_v.bias | 0xb011e7a0 | 0x800 | +| 48 | blk.3.attn_v.weight | 0xb011efa0 | 0x380000 | +| 49 | blk.4.attn_norm.weight | 0xb049efa0 | 0x3800 | +| 50 | blk.4.ffn_down.weight | 0xb04a27a0 | 0x8180000 | +| 51 | blk.4.ffn_gate.weight | 0xb86227a0 | 0x8180000 | +| 52 | blk.4.ffn_up.weight | 0xc07a27a0 | 0x8180000 | +| 53 | blk.4.ffn_norm.weight | 0xc89227a0 | 0x3800 | +| 54 | blk.4.attn_k.bias | 0xc8925fa0 | 0x800 | +| 55 | blk.4.attn_k.weight | 0xc89267a0 | 0x380000 | +| 56 | blk.4.attn_output.weight | 0xc8ca67a0 | 0x1880000 | +| 57 | blk.4.attn_q.bias | 0xca5267a0 | 0x3800 | +| 58 | blk.4.attn_q.weight | 0xca529fa0 | 0x1880000 | +| 59 | blk.4.attn_v.bias | 0xcbda9fa0 | 0x800 | +| 60 | blk.4.attn_v.weight | 0xcbdaa7a0 | 0x380000 | +| 61 | blk.5.attn_norm.weight | 0xcc12a7a0 | 0x3800 | +| 62 | blk.5.ffn_down.weight | 0xcc12dfa0 | 0x8180000 | +| 63 | blk.5.ffn_gate.weight | 0xd42adfa0 | 0x8180000 | +| 64 | blk.5.ffn_up.weight | 0xdc42dfa0 | 0x8180000 | +| 65 | blk.5.ffn_norm.weight | 0xe45adfa0 | 0x3800 | +| 66 | blk.5.attn_k.bias | 0xe45b17a0 | 0x800 | +| 67 | blk.5.attn_k.weight | 0xe45b1fa0 | 0x380000 | +| 68 | blk.5.attn_output.weight | 0xe4931fa0 | 0x1880000 | +| 69 | blk.5.attn_q.bias | 0xe61b1fa0 | 0x3800 | +| 70 | blk.5.attn_q.weight | 0xe61b57a0 | 0x1880000 | +| 71 | blk.5.attn_v.bias | 0xe7a357a0 | 0x800 | +| 72 | blk.5.attn_v.weight | 0xe7a35fa0 | 0x380000 | +| 73 | blk.6.attn_norm.weight | 0xe7db5fa0 | 0x3800 | +| 74 | blk.6.ffn_down.weight | 0xe7db97a0 | 0x8180000 | +| 75 | blk.6.ffn_gate.weight | 0xeff397a0 | 0x8180000 | +| 76 | blk.6.ffn_up.weight | 0xf80b97a0 | 0x8180000 | +| 77 | blk.6.ffn_norm.weight | 0x1002397a0 | 0x3800 | +| 78 | blk.6.attn_k.bias | 0x10023cfa0 | 0x800 | +| 79 | blk.6.attn_k.weight | 0x10023d7a0 | 0x380000 | +| 80 | blk.6.attn_output.weight | 0x1005bd7a0 | 0x1880000 | +| 81 | blk.6.attn_q.bias | 0x101e3d7a0 | 0x3800 | +| 82 | blk.6.attn_q.weight | 0x101e40fa0 | 0x1880000 | +| 83 | blk.6.attn_v.bias | 0x1036c0fa0 | 0x800 | +| 84 | blk.6.attn_v.weight | 0x1036c17a0 | 0x380000 | +| 85 | blk.7.attn_norm.weight | 0x103a417a0 | 0x3800 | +| 86 | blk.7.ffn_down.weight | 0x103a44fa0 | 0x8180000 | +| 87 | blk.7.ffn_gate.weight | 0x10bbc4fa0 | 0x8180000 | +| 88 | blk.7.ffn_up.weight | 0x113d44fa0 | 0x8180000 | +| 89 | blk.7.ffn_norm.weight | 0x11bec4fa0 | 0x3800 | +| 90 | blk.7.attn_k.bias | 0x11bec87a0 | 0x800 | +| 91 | blk.7.attn_k.weight | 0x11bec8fa0 | 0x380000 | +| 92 | blk.7.attn_output.weight | 0x11c248fa0 | 0x1880000 | +| 93 | blk.7.attn_q.bias | 0x11dac8fa0 | 0x3800 | +| 94 | blk.7.attn_q.weight | 0x11dacc7a0 | 0x1880000 | +| 95 | blk.7.attn_v.bias | 0x11f34c7a0 | 0x800 | +| 96 | blk.7.attn_v.weight | 0x11f34cfa0 | 0x380000 | +| 97 | blk.8.attn_k.bias | 0x11f6ccfa0 | 0x800 | +| 98 | blk.8.attn_k.weight | 0x11f6cd7a0 | 0x380000 | +| 99 | blk.8.attn_output.weight | 0x11fa4d7a0 | 0x1880000 | +| 100 | blk.8.attn_q.bias | 0x1212cd7a0 | 0x3800 | +| 101 | blk.8.attn_q.weight | 0x1212d0fa0 | 0x1880000 | +| 102 | blk.8.attn_v.bias | 0x122b50fa0 | 0x800 | +| 103 | blk.8.attn_v.weight | 0x122b517a0 | 0x380000 | +| 104 | blk.10.attn_norm.weight | 0x122ed17a0 | 0x3800 | +| 105 | blk.10.ffn_down.weight | 0x122ed4fa0 | 0x8180000 | +| 106 | blk.10.ffn_gate.weight | 0x12b054fa0 | 0x8180000 | +| 107 | blk.10.ffn_up.weight | 0x1331d4fa0 | 0x8180000 | +| 108 | blk.10.ffn_norm.weight | 0x13b354fa0 | 0x3800 | +| 109 | blk.10.attn_k.bias | 0x13b3587a0 | 0x800 | +| 110 | blk.10.attn_k.weight | 0x13b358fa0 | 0x380000 | +| 111 | blk.10.attn_output.weight | 0x13b6d8fa0 | 0x1880000 | +| 112 | blk.10.attn_q.bias | 0x13cf58fa0 | 0x3800 | +| 113 | blk.10.attn_q.weight | 0x13cf5c7a0 | 0x1880000 | +| 114 | blk.10.attn_v.bias | 0x13e7dc7a0 | 0x800 | +| 115 | blk.10.attn_v.weight | 0x13e7dcfa0 | 0x380000 | +| 116 | blk.11.attn_norm.weight | 0x13eb5cfa0 | 0x3800 | +| 117 | blk.11.ffn_down.weight | 0x13eb607a0 | 0x8180000 | +| 118 | blk.11.ffn_gate.weight | 0x146ce07a0 | 0x8180000 | +| 119 | blk.11.ffn_up.weight | 0x14ee607a0 | 0x8180000 | +| 120 | blk.11.ffn_norm.weight | 0x156fe07a0 | 0x3800 | +| 121 | blk.11.attn_k.bias | 0x156fe3fa0 | 0x800 | +| 122 | blk.11.attn_k.weight | 0x156fe47a0 | 0x380000 | +| 123 | blk.11.attn_output.weight | 0x1573647a0 | 0x1880000 | +| 124 | blk.11.attn_q.bias | 0x158be47a0 | 0x3800 | +| 125 | blk.11.attn_q.weight | 0x158be7fa0 | 0x1880000 | +| 126 | blk.11.attn_v.bias | 0x15a467fa0 | 0x800 | +| 127 | blk.11.attn_v.weight | 0x15a4687a0 | 0x380000 | +| 128 | blk.12.attn_norm.weight | 0x15a7e87a0 | 0x3800 | +| 129 | blk.12.ffn_down.weight | 0x15a7ebfa0 | 0x8180000 | +| 130 | blk.12.ffn_gate.weight | 0x16296bfa0 | 0x8180000 | +| 131 | blk.12.ffn_up.weight | 0x16aaebfa0 | 0x8180000 | +| 132 | blk.12.ffn_norm.weight | 0x172c6bfa0 | 0x3800 | +| 133 | blk.12.attn_k.bias | 0x172c6f7a0 | 0x800 | +| 134 | blk.12.attn_k.weight | 0x172c6ffa0 | 0x380000 | +| 135 | blk.12.attn_output.weight | 0x172feffa0 | 0x1880000 | +| 136 | blk.12.attn_q.bias | 0x17486ffa0 | 0x3800 | +| 137 | blk.12.attn_q.weight | 0x1748737a0 | 0x1880000 | +| 138 | blk.12.attn_v.bias | 0x1760f37a0 | 0x800 | +| 139 | blk.12.attn_v.weight | 0x1760f3fa0 | 0x380000 | +| 140 | blk.13.attn_norm.weight | 0x176473fa0 | 0x3800 | +| 141 | blk.13.ffn_down.weight | 0x1764777a0 | 0x8180000 | +| 142 | blk.13.ffn_gate.weight | 0x17e5f77a0 | 0x8180000 | +| 143 | blk.13.ffn_up.weight | 0x1867777a0 | 0x8180000 | +| 144 | blk.13.ffn_norm.weight | 0x18e8f77a0 | 0x3800 | +| 145 | blk.13.attn_k.bias | 0x18e8fafa0 | 0x800 | +| 146 | blk.13.attn_k.weight | 0x18e8fb7a0 | 0x380000 | +| 147 | blk.13.attn_output.weight | 0x18ec7b7a0 | 0x1880000 | +| 148 | blk.13.attn_q.bias | 0x1904fb7a0 | 0x3800 | +| 149 | blk.13.attn_q.weight | 0x1904fefa0 | 0x1880000 | +| 150 | blk.13.attn_v.bias | 0x191d7efa0 | 0x800 | +| 151 | blk.13.attn_v.weight | 0x191d7f7a0 | 0x380000 | +| 152 | blk.14.attn_norm.weight | 0x1920ff7a0 | 0x3800 | +| 153 | blk.14.ffn_down.weight | 0x192102fa0 | 0x8180000 | +| 154 | blk.14.ffn_gate.weight | 0x19a282fa0 | 0x8180000 | +| 155 | blk.14.ffn_up.weight | 0x1a2402fa0 | 0x8180000 | +| 156 | blk.14.ffn_norm.weight | 0x1aa582fa0 | 0x3800 | +| 157 | blk.14.attn_k.bias | 0x1aa5867a0 | 0x800 | +| 158 | blk.14.attn_k.weight | 0x1aa586fa0 | 0x380000 | +| 159 | blk.14.attn_output.weight | 0x1aa906fa0 | 0x1880000 | +| 160 | blk.14.attn_q.bias | 0x1ac186fa0 | 0x3800 | +| 161 | blk.14.attn_q.weight | 0x1ac18a7a0 | 0x1880000 | +| 162 | blk.14.attn_v.bias | 0x1ada0a7a0 | 0x800 | +| 163 | blk.14.attn_v.weight | 0x1ada0afa0 | 0x380000 | +| 164 | blk.15.attn_norm.weight | 0x1add8afa0 | 0x3800 | +| 165 | blk.15.ffn_down.weight | 0x1add8e7a0 | 0x8180000 | +| 166 | blk.15.ffn_gate.weight | 0x1b5f0e7a0 | 0x8180000 | +| 167 | blk.15.ffn_up.weight | 0x1be08e7a0 | 0x8180000 | +| 168 | blk.15.ffn_norm.weight | 0x1c620e7a0 | 0x3800 | +| 169 | blk.15.attn_k.bias | 0x1c6211fa0 | 0x800 | +| 170 | blk.15.attn_k.weight | 0x1c62127a0 | 0x380000 | +| 171 | blk.15.attn_output.weight | 0x1c65927a0 | 0x1880000 | +| 172 | blk.15.attn_q.bias | 0x1c7e127a0 | 0x3800 | +| 173 | blk.15.attn_q.weight | 0x1c7e15fa0 | 0x1880000 | +| 174 | blk.15.attn_v.bias | 0x1c9695fa0 | 0x800 | +| 175 | blk.15.attn_v.weight | 0x1c96967a0 | 0x380000 | +| 176 | blk.16.attn_norm.weight | 0x1c9a167a0 | 0x3800 | +| 177 | blk.16.ffn_down.weight | 0x1c9a19fa0 | 0x8180000 | +| 178 | blk.16.ffn_gate.weight | 0x1d1b99fa0 | 0x8180000 | +| 179 | blk.16.ffn_up.weight | 0x1d9d19fa0 | 0x8180000 | +| 180 | blk.16.ffn_norm.weight | 0x1e1e99fa0 | 0x3800 | +| 181 | blk.16.attn_k.bias | 0x1e1e9d7a0 | 0x800 | +| 182 | blk.16.attn_k.weight | 0x1e1e9dfa0 | 0x380000 | +| 183 | blk.16.attn_output.weight | 0x1e221dfa0 | 0x1880000 | +| 184 | blk.16.attn_q.bias | 0x1e3a9dfa0 | 0x3800 | +| 185 | blk.16.attn_q.weight | 0x1e3aa17a0 | 0x1880000 | +| 186 | blk.16.attn_v.bias | 0x1e53217a0 | 0x800 | +| 187 | blk.16.attn_v.weight | 0x1e5321fa0 | 0x380000 | +| 188 | blk.17.attn_norm.weight | 0x1e56a1fa0 | 0x3800 | +| 189 | blk.17.ffn_down.weight | 0x1e56a57a0 | 0x8180000 | +| 190 | blk.17.ffn_gate.weight | 0x1ed8257a0 | 0x8180000 | +| 191 | blk.17.ffn_up.weight | 0x1f59a57a0 | 0x8180000 | +| 192 | blk.17.ffn_norm.weight | 0x1fdb257a0 | 0x3800 | +| 193 | blk.17.attn_k.bias | 0x1fdb28fa0 | 0x800 | +| 194 | blk.17.attn_k.weight | 0x1fdb297a0 | 0x380000 | +| 195 | blk.17.attn_output.weight | 0x1fdea97a0 | 0x1880000 | +| 196 | blk.17.attn_q.bias | 0x1ff7297a0 | 0x3800 | +| 197 | blk.17.attn_q.weight | 0x1ff72cfa0 | 0x1880000 | +| 198 | blk.17.attn_v.bias | 0x200facfa0 | 0x800 | +| 199 | blk.17.attn_v.weight | 0x200fad7a0 | 0x380000 | +| 200 | blk.18.ffn_gate.weight | 0x20132d7a0 | 0x8180000 | +| 201 | blk.18.ffn_up.weight | 0x2094ad7a0 | 0x8180000 | +| 202 | blk.18.attn_k.bias | 0x21162d7a0 | 0x800 | +| 203 | blk.18.attn_k.weight | 0x21162dfa0 | 0x380000 | +| 204 | blk.18.attn_output.weight | 0x2119adfa0 | 0x1880000 | +| 205 | blk.18.attn_q.bias | 0x21322dfa0 | 0x3800 | +| 206 | blk.18.attn_q.weight | 0x2132317a0 | 0x1880000 | +| 207 | blk.18.attn_v.bias | 0x214ab17a0 | 0x800 | +| 208 | blk.18.attn_v.weight | 0x214ab1fa0 | 0x380000 | +| 209 | blk.8.attn_norm.weight | 0x214e31fa0 | 0x3800 | +| 210 | blk.8.ffn_down.weight | 0x214e357a0 | 0x8180000 | +| 211 | blk.8.ffn_gate.weight | 0x21cfb57a0 | 0x8180000 | +| 212 | blk.8.ffn_up.weight | 0x2251357a0 | 0x8180000 | +| 213 | blk.8.ffn_norm.weight | 0x22d2b57a0 | 0x3800 | +| 214 | blk.9.attn_norm.weight | 0x22d2b8fa0 | 0x3800 | +| 215 | blk.9.ffn_down.weight | 0x22d2bc7a0 | 0x8180000 | +| 216 | blk.9.ffn_gate.weight | 0x23543c7a0 | 0x8180000 | +| 217 | blk.9.ffn_up.weight | 0x23d5bc7a0 | 0x8180000 | +| 218 | blk.9.ffn_norm.weight | 0x24573c7a0 | 0x3800 | +| 219 | blk.9.attn_k.bias | 0x24573ffa0 | 0x800 | +| 220 | blk.9.attn_k.weight | 0x2457407a0 | 0x380000 | +| 221 | blk.9.attn_output.weight | 0x245ac07a0 | 0x1880000 | +| 222 | blk.9.attn_q.bias | 0x2473407a0 | 0x3800 | +| 223 | blk.9.attn_q.weight | 0x247343fa0 | 0x1880000 | +| 224 | blk.9.attn_v.bias | 0x248bc3fa0 | 0x800 | +| 225 | blk.9.attn_v.weight | 0x248bc47a0 | 0x380000 | +| 226 | blk.18.attn_norm.weight | 0x248f447a0 | 0x3800 | +| 227 | blk.18.ffn_down.weight | 0x248f47fa0 | 0x8180000 | +| 228 | blk.18.ffn_norm.weight | 0x2510c7fa0 | 0x3800 | +| 229 | blk.19.attn_norm.weight | 0x2510cb7a0 | 0x3800 | +| 230 | blk.19.ffn_down.weight | 0x2510cefa0 | 0x8180000 | +| 231 | blk.19.ffn_gate.weight | 0x25924efa0 | 0x8180000 | +| 232 | blk.19.ffn_up.weight | 0x2613cefa0 | 0x8180000 | +| 233 | blk.19.ffn_norm.weight | 0x26954efa0 | 0x3800 | +| 234 | blk.19.attn_k.bias | 0x2695527a0 | 0x800 | +| 235 | blk.19.attn_k.weight | 0x269552fa0 | 0x380000 | +| 236 | blk.19.attn_output.weight | 0x2698d2fa0 | 0x1880000 | +| 237 | blk.19.attn_q.bias | 0x26b152fa0 | 0x3800 | +| 238 | blk.19.attn_q.weight | 0x26b1567a0 | 0x1880000 | +| 239 | blk.19.attn_v.bias | 0x26c9d67a0 | 0x800 | +| 240 | blk.19.attn_v.weight | 0x26c9d6fa0 | 0x380000 | +| 241 | blk.20.attn_norm.weight | 0x26cd56fa0 | 0x3800 | +| 242 | blk.20.ffn_down.weight | 0x26cd5a7a0 | 0x8180000 | +| 243 | blk.20.ffn_gate.weight | 0x274eda7a0 | 0x8180000 | +| 244 | blk.20.ffn_up.weight | 0x27d05a7a0 | 0x8180000 | +| 245 | blk.20.ffn_norm.weight | 0x2851da7a0 | 0x3800 | +| 246 | blk.20.attn_k.bias | 0x2851ddfa0 | 0x800 | +| 247 | blk.20.attn_k.weight | 0x2851de7a0 | 0x380000 | +| 248 | blk.20.attn_output.weight | 0x28555e7a0 | 0x1880000 | +| 249 | blk.20.attn_q.bias | 0x286dde7a0 | 0x3800 | +| 250 | blk.20.attn_q.weight | 0x286de1fa0 | 0x1880000 | +| 251 | blk.20.attn_v.bias | 0x288661fa0 | 0x800 | +| 252 | blk.20.attn_v.weight | 0x2886627a0 | 0x380000 | +| 253 | blk.21.attn_norm.weight | 0x2889e27a0 | 0x3800 | +| 254 | blk.21.ffn_down.weight | 0x2889e5fa0 | 0x8180000 | +| 255 | blk.21.ffn_gate.weight | 0x290b65fa0 | 0x8180000 | +| 256 | blk.21.ffn_up.weight | 0x298ce5fa0 | 0x8180000 | +| 257 | blk.21.ffn_norm.weight | 0x2a0e65fa0 | 0x3800 | +| 258 | blk.21.attn_k.bias | 0x2a0e697a0 | 0x800 | +| 259 | blk.21.attn_k.weight | 0x2a0e69fa0 | 0x380000 | +| 260 | blk.21.attn_output.weight | 0x2a11e9fa0 | 0x1880000 | +| 261 | blk.21.attn_q.bias | 0x2a2a69fa0 | 0x3800 | +| 262 | blk.21.attn_q.weight | 0x2a2a6d7a0 | 0x1880000 | +| 263 | blk.21.attn_v.bias | 0x2a42ed7a0 | 0x800 | +| 264 | blk.21.attn_v.weight | 0x2a42edfa0 | 0x380000 | +| 265 | blk.22.attn_norm.weight | 0x2a466dfa0 | 0x3800 | +| 266 | blk.22.ffn_down.weight | 0x2a46717a0 | 0x8180000 | +| 267 | blk.22.ffn_gate.weight | 0x2ac7f17a0 | 0x8180000 | +| 268 | blk.22.ffn_up.weight | 0x2b49717a0 | 0x8180000 | +| 269 | blk.22.ffn_norm.weight | 0x2bcaf17a0 | 0x3800 | +| 270 | blk.22.attn_k.bias | 0x2bcaf4fa0 | 0x800 | +| 271 | blk.22.attn_k.weight | 0x2bcaf57a0 | 0x380000 | +| 272 | blk.22.attn_output.weight | 0x2bce757a0 | 0x1880000 | +| 273 | blk.22.attn_q.bias | 0x2be6f57a0 | 0x3800 | +| 274 | blk.22.attn_q.weight | 0x2be6f8fa0 | 0x1880000 | +| 275 | blk.22.attn_v.bias | 0x2bff78fa0 | 0x800 | +| 276 | blk.22.attn_v.weight | 0x2bff797a0 | 0x380000 | +| 277 | blk.23.attn_norm.weight | 0x2c02f97a0 | 0x3800 | +| 278 | blk.23.ffn_down.weight | 0x2c02fcfa0 | 0x8180000 | +| 279 | blk.23.ffn_gate.weight | 0x2c847cfa0 | 0x8180000 | +| 280 | blk.23.ffn_up.weight | 0x2d05fcfa0 | 0x8180000 | +| 281 | blk.23.ffn_norm.weight | 0x2d877cfa0 | 0x3800 | +| 282 | blk.23.attn_k.bias | 0x2d87807a0 | 0x800 | +| 283 | blk.23.attn_k.weight | 0x2d8780fa0 | 0x380000 | +| 284 | blk.23.attn_output.weight | 0x2d8b00fa0 | 0x1880000 | +| 285 | blk.23.attn_q.bias | 0x2da380fa0 | 0x3800 | +| 286 | blk.23.attn_q.weight | 0x2da3847a0 | 0x1880000 | +| 287 | blk.23.attn_v.bias | 0x2dbc047a0 | 0x800 | +| 288 | blk.23.attn_v.weight | 0x2dbc04fa0 | 0x380000 | +| 289 | blk.24.attn_norm.weight | 0x2dbf84fa0 | 0x3800 | +| 290 | blk.24.ffn_down.weight | 0x2dbf887a0 | 0x8180000 | +| 291 | blk.24.ffn_gate.weight | 0x2e41087a0 | 0x8180000 | +| 292 | blk.24.ffn_up.weight | 0x2ec2887a0 | 0x8180000 | +| 293 | blk.24.ffn_norm.weight | 0x2f44087a0 | 0x3800 | +| 294 | blk.24.attn_k.bias | 0x2f440bfa0 | 0x800 | +| 295 | blk.24.attn_k.weight | 0x2f440c7a0 | 0x380000 | +| 296 | blk.24.attn_output.weight | 0x2f478c7a0 | 0x1880000 | +| 297 | blk.24.attn_q.bias | 0x2f600c7a0 | 0x3800 | +| 298 | blk.24.attn_q.weight | 0x2f600ffa0 | 0x1880000 | +| 299 | blk.24.attn_v.bias | 0x2f788ffa0 | 0x800 | +| 300 | blk.24.attn_v.weight | 0x2f78907a0 | 0x380000 | +| 301 | blk.25.attn_norm.weight | 0x2f7c107a0 | 0x3800 | +| 302 | blk.25.ffn_down.weight | 0x2f7c13fa0 | 0x8180000 | +| 303 | blk.25.ffn_gate.weight | 0x2ffd93fa0 | 0x8180000 | +| 304 | blk.25.ffn_up.weight | 0x307f13fa0 | 0x8180000 | +| 305 | blk.25.ffn_norm.weight | 0x310093fa0 | 0x3800 | +| 306 | blk.25.attn_k.bias | 0x3100977a0 | 0x800 | +| 307 | blk.25.attn_k.weight | 0x310097fa0 | 0x380000 | +| 308 | blk.25.attn_output.weight | 0x310417fa0 | 0x1880000 | +| 309 | blk.25.attn_q.bias | 0x311c97fa0 | 0x3800 | +| 310 | blk.25.attn_q.weight | 0x311c9b7a0 | 0x1880000 | +| 311 | blk.25.attn_v.bias | 0x31351b7a0 | 0x800 | +| 312 | blk.25.attn_v.weight | 0x31351bfa0 | 0x380000 | +| 313 | blk.26.attn_norm.weight | 0x31389bfa0 | 0x3800 | +| 314 | blk.26.ffn_down.weight | 0x31389f7a0 | 0x8180000 | +| 315 | blk.26.ffn_gate.weight | 0x31ba1f7a0 | 0x8180000 | +| 316 | blk.26.ffn_up.weight | 0x323b9f7a0 | 0x8180000 | +| 317 | blk.26.ffn_norm.weight | 0x32bd1f7a0 | 0x3800 | +| 318 | blk.26.attn_k.bias | 0x32bd22fa0 | 0x800 | +| 319 | blk.26.attn_k.weight | 0x32bd237a0 | 0x380000 | +| 320 | blk.26.attn_output.weight | 0x32c0a37a0 | 0x1880000 | +| 321 | blk.26.attn_q.bias | 0x32d9237a0 | 0x3800 | +| 322 | blk.26.attn_q.weight | 0x32d926fa0 | 0x1880000 | +| 323 | blk.26.attn_v.bias | 0x32f1a6fa0 | 0x800 | +| 324 | blk.26.attn_v.weight | 0x32f1a77a0 | 0x380000 | +| 325 | blk.27.attn_norm.weight | 0x32f5277a0 | 0x3800 | +| 326 | blk.27.ffn_down.weight | 0x32f52afa0 | 0x8180000 | +| 327 | blk.27.ffn_gate.weight | 0x3376aafa0 | 0x8180000 | +| 328 | blk.27.ffn_up.weight | 0x33f82afa0 | 0x8180000 | +| 329 | blk.27.ffn_norm.weight | 0x3479aafa0 | 0x3800 | +| 330 | blk.27.attn_k.bias | 0x3479ae7a0 | 0x800 | +| 331 | blk.27.attn_k.weight | 0x3479aefa0 | 0x380000 | +| 332 | blk.27.attn_output.weight | 0x347d2efa0 | 0x1880000 | +| 333 | blk.27.attn_q.bias | 0x3495aefa0 | 0x3800 | +| 334 | blk.27.attn_q.weight | 0x3495b27a0 | 0x1880000 | +| 335 | blk.27.attn_v.bias | 0x34ae327a0 | 0x800 | +| 336 | blk.27.attn_v.weight | 0x34ae32fa0 | 0x380000 | +| 337 | output_norm.weight | 0x34b1b2fa0 | 0x3800 | +| 338 | output.weight | 0x34b1b67a0 | 0x40cc5c00 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | F16 | +| 337 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | F16 | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 1 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 3 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 4 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 5 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 7 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 8 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 9 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 11 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 12 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 13 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 15 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 16 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 17 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 19 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 20 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 21 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 23 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 24 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 25 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 27 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 28 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 29 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 31 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 32 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 33 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 35 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 36 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 37 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 39 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 40 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 41 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 43 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 44 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 45 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 47 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 48 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 49 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 51 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 52 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 53 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 55 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 56 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 57 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 59 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 60 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 61 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 63 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 64 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 65 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 67 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 68 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 69 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 71 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 72 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 73 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 75 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 76 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 77 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 79 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 80 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 81 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 83 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 84 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 85 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 87 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 88 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 89 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 91 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 92 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 93 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 95 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 96 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 97 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 98 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 99 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 100 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 101 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 102 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 103 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 209 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 211 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 212 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 213 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 104 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 105 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 106 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 107 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 108 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 109 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 110 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 111 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 112 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 113 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 114 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 115 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 116 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 117 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 118 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 119 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 120 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 121 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 122 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 123 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 124 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 125 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 126 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 127 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 128 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 129 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 130 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 131 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 132 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 133 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 134 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 135 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 136 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 137 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 138 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 139 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 140 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 141 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 142 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 143 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 144 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 145 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 146 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 147 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 148 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 149 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 150 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 151 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 152 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 153 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 154 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 155 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 156 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 157 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 158 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 159 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 160 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 161 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 162 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 163 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 164 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 165 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 166 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 167 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 168 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 169 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 170 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 171 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 172 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 173 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 174 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 175 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 176 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 177 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 178 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 179 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 180 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 181 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 182 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 183 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 184 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 185 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 186 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 187 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 188 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 189 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 190 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 191 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 192 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 193 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 194 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 195 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 196 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 197 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 198 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 199 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 200 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 201 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 202 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 203 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 204 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 205 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 207 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 226 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 228 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 214 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 215 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 216 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 217 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 218 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 219 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 221 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 222 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 223 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 224 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 225 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 229 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 231 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 232 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 233 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 235 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 236 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 237 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 239 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 240 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 241 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 243 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 244 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 245 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 247 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 248 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 249 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 251 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 252 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 253 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 255 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 256 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 257 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 259 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 260 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 261 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 263 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 264 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 265 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 267 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 268 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 269 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 271 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 272 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 273 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 275 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 276 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 277 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 279 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 280 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 281 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 283 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 284 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 285 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 287 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 288 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 289 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 291 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 292 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 293 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 295 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 296 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 297 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 299 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 300 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 301 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 303 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 304 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 305 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 307 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 308 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 309 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 311 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 312 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 313 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 315 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 316 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 317 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 319 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 320 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 321 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 323 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 324 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 325 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | F16 | +| 327 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 328 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | F16 | +| 329 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 331 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 332 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 333 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | F16 | +| 335 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 336 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-F16.mmlu b/scores/Hammer2.1-7b-F16.mmlu new file mode 100644 index 0000000..c638bf2 --- /dev/null +++ b/scores/Hammer2.1-7b-F16.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 29 key-value pairs and 339 tensors from ./Hammer2.1-7b-F16.gguf (version GGUF V3 (latest)) + +Final result: 39.0667 +/- 1.7827 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 527.95 ms +llama_perf_context_print: prompt eval time = 194228.90 ms / 67719 tokens ( 2.87 ms per token, 348.66 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 196042.71 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-F16.tqa b/scores/Hammer2.1-7b-F16.tqa new file mode 100644 index 0000000..4e69259 --- /dev/null +++ b/scores/Hammer2.1-7b-F16.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 29 key-value pairs and 339 tensors from ./Hammer2.1-7b-F16.gguf (version GGUF V3 (latest)) + +Final result: 30.4000 +/- 1.6807 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 508.17 ms +llama_perf_context_print: prompt eval time = 146171.42 ms / 49696 tokens ( 2.94 ms per token, 339.98 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 148488.63 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-F16.wng b/scores/Hammer2.1-7b-F16.wng new file mode 100644 index 0000000..045f303 --- /dev/null +++ b/scores/Hammer2.1-7b-F16.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 29 key-value pairs and 339 tensors from ./Hammer2.1-7b-F16.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.5333 +/- 1.6658 + +llama_perf_context_print: load time = 536.90 ms +llama_perf_context_print: prompt eval time = 61935.07 ms / 21448 tokens ( 2.89 ms per token, 346.30 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 62681.37 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-IQ3_M.md b/scores/Hammer2.1-7b-IQ3_M.md new file mode 100644 index 0000000..0ee537b --- /dev/null +++ b/scores/Hammer2.1-7b-IQ3_M.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-IQ3_M.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 27 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-IQ3\_M.gguf - GGUF Internal File Dump](#hammer21-7b-iq3_mgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0xdebe7c4 | +| 1 | output_norm.weight | 0xe469ca0 | 0x3800 | +| 2 | token_embd.weight | 0xe46d4a0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x1c32bc80 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x1c32c480 | 0xab800 | +| 5 | blk.0.attn_norm.weight | 0x1c3d7c80 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x1c3db480 | 0x6e4000 | +| 7 | blk.0.attn_q.bias | 0x1cabf480 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x1cac2c80 | 0x4b0800 | +| 9 | blk.0.attn_v.bias | 0x1cf73480 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x1cf73c80 | 0xc0800 | +| 11 | blk.0.ffn_down.weight | 0x1d034480 | 0x246c000 | +| 12 | blk.0.ffn_gate.weight | 0x1f4a0480 | 0x18c9800 | +| 13 | blk.0.ffn_norm.weight | 0x20d69c80 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x20d6d480 | 0x18c9800 | +| 15 | blk.1.attn_k.bias | 0x22636c80 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x22637480 | 0xab800 | +| 17 | blk.1.attn_norm.weight | 0x226e2c80 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x226e6480 | 0x6e4000 | +| 19 | blk.1.attn_q.bias | 0x22dca480 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x22dcdc80 | 0x4b0800 | +| 21 | blk.1.attn_v.bias | 0x2327e480 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x2327ec80 | 0xc0800 | +| 23 | blk.1.ffn_down.weight | 0x2333f480 | 0x246c000 | +| 24 | blk.1.ffn_gate.weight | 0x257ab480 | 0x1bd2800 | +| 25 | blk.1.ffn_norm.weight | 0x2737dc80 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x27381480 | 0x1bd2800 | +| 27 | blk.2.attn_k.bias | 0x28f53c80 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x28f54480 | 0xab800 | +| 29 | blk.2.attn_norm.weight | 0x28fffc80 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x29003480 | 0x6e4000 | +| 31 | blk.2.attn_q.bias | 0x296e7480 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x296eac80 | 0x4b0800 | +| 33 | blk.2.attn_v.bias | 0x29b9b480 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x29b9bc80 | 0xc0800 | +| 35 | blk.2.ffn_down.weight | 0x29c5c480 | 0x246c000 | +| 36 | blk.2.ffn_gate.weight | 0x2c0c8480 | 0x1bd2800 | +| 37 | blk.2.ffn_norm.weight | 0x2dc9ac80 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x2dc9e480 | 0x1bd2800 | +| 39 | blk.3.attn_k.bias | 0x2f870c80 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x2f871480 | 0xab800 | +| 41 | blk.3.attn_norm.weight | 0x2f91cc80 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x2f920480 | 0x6e4000 | +| 43 | blk.3.attn_q.bias | 0x30004480 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x30007c80 | 0x4b0800 | +| 45 | blk.3.attn_v.bias | 0x304b8480 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x304b8c80 | 0xc0800 | +| 47 | blk.3.ffn_down.weight | 0x30579480 | 0x246c000 | +| 48 | blk.3.ffn_gate.weight | 0x329e5480 | 0x1bd2800 | +| 49 | blk.3.ffn_norm.weight | 0x345b7c80 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x345bb480 | 0x1bd2800 | +| 51 | blk.4.attn_k.bias | 0x3618dc80 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x3618e480 | 0xab800 | +| 53 | blk.4.attn_norm.weight | 0x36239c80 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x3623d480 | 0x6e4000 | +| 55 | blk.4.attn_q.bias | 0x36921480 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x36924c80 | 0x4b0800 | +| 57 | blk.4.attn_v.bias | 0x36dd5480 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x36dd5c80 | 0xc0800 | +| 59 | blk.4.ffn_down.weight | 0x36e96480 | 0x246c000 | +| 60 | blk.4.ffn_gate.weight | 0x39302480 | 0x1bd2800 | +| 61 | blk.4.ffn_norm.weight | 0x3aed4c80 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x3aed8480 | 0x1bd2800 | +| 63 | blk.5.attn_k.bias | 0x3caaac80 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x3caab480 | 0xab800 | +| 65 | blk.5.attn_norm.weight | 0x3cb56c80 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x3cb5a480 | 0x6e4000 | +| 67 | blk.5.attn_q.bias | 0x3d23e480 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x3d241c80 | 0x4b0800 | +| 69 | blk.5.attn_v.bias | 0x3d6f2480 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x3d6f2c80 | 0xc0800 | +| 71 | blk.5.ffn_down.weight | 0x3d7b3480 | 0x246c000 | +| 72 | blk.5.ffn_gate.weight | 0x3fc1f480 | 0x1bd2800 | +| 73 | blk.5.ffn_norm.weight | 0x417f1c80 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x417f5480 | 0x1bd2800 | +| 75 | blk.6.attn_k.bias | 0x433c7c80 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x433c8480 | 0xab800 | +| 77 | blk.6.attn_norm.weight | 0x43473c80 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x43477480 | 0x6e4000 | +| 79 | blk.6.attn_q.bias | 0x43b5b480 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x43b5ec80 | 0x4b0800 | +| 81 | blk.6.attn_v.bias | 0x4400f480 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x4400fc80 | 0xc0800 | +| 83 | blk.6.ffn_down.weight | 0x440d0480 | 0x246c000 | +| 84 | blk.6.ffn_gate.weight | 0x4653c480 | 0x18c9800 | +| 85 | blk.6.ffn_norm.weight | 0x47e05c80 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x47e09480 | 0x18c9800 | +| 87 | blk.7.attn_k.bias | 0x496d2c80 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x496d3480 | 0xab800 | +| 89 | blk.7.attn_norm.weight | 0x4977ec80 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x49782480 | 0x6e4000 | +| 91 | blk.7.attn_q.bias | 0x49e66480 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x49e69c80 | 0x4b0800 | +| 93 | blk.7.attn_v.bias | 0x4a31a480 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x4a31ac80 | 0xc0800 | +| 95 | blk.7.ffn_down.weight | 0x4a3db480 | 0x246c000 | +| 96 | blk.7.ffn_gate.weight | 0x4c847480 | 0x18c9800 | +| 97 | blk.7.ffn_norm.weight | 0x4e110c80 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x4e114480 | 0x18c9800 | +| 99 | blk.8.attn_k.bias | 0x4f9ddc80 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x4f9de480 | 0xab800 | +| 101 | blk.8.attn_norm.weight | 0x4fa89c80 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x4fa8d480 | 0x6e4000 | +| 103 | blk.8.attn_q.bias | 0x50171480 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x50174c80 | 0x4b0800 | +| 105 | blk.8.attn_v.bias | 0x50625480 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x50625c80 | 0xc0800 | +| 107 | blk.8.ffn_down.weight | 0x506e6480 | 0x246c000 | +| 108 | blk.8.ffn_gate.weight | 0x52b52480 | 0x18c9800 | +| 109 | blk.8.ffn_norm.weight | 0x5441bc80 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x5441f480 | 0x18c9800 | +| 111 | blk.9.attn_k.bias | 0x55ce8c80 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x55ce9480 | 0xab800 | +| 113 | blk.9.attn_norm.weight | 0x55d94c80 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x55d98480 | 0x6e4000 | +| 115 | blk.9.attn_q.bias | 0x5647c480 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x5647fc80 | 0x4b0800 | +| 117 | blk.9.attn_v.bias | 0x56930480 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x56930c80 | 0xc0800 | +| 119 | blk.9.ffn_down.weight | 0x569f1480 | 0x246c000 | +| 120 | blk.9.ffn_gate.weight | 0x58e5d480 | 0x1bd2800 | +| 121 | blk.9.ffn_norm.weight | 0x5aa2fc80 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x5aa33480 | 0x1bd2800 | +| 123 | blk.10.attn_k.bias | 0x5c605c80 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x5c606480 | 0xab800 | +| 125 | blk.10.attn_norm.weight | 0x5c6b1c80 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x5c6b5480 | 0x6e4000 | +| 127 | blk.10.attn_q.bias | 0x5cd99480 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x5cd9cc80 | 0x4b0800 | +| 129 | blk.10.attn_v.bias | 0x5d24d480 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x5d24dc80 | 0xc0800 | +| 131 | blk.10.ffn_down.weight | 0x5d30e480 | 0x246c000 | +| 132 | blk.10.ffn_gate.weight | 0x5f77a480 | 0x18c9800 | +| 133 | blk.10.ffn_norm.weight | 0x61043c80 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x61047480 | 0x18c9800 | +| 135 | blk.11.attn_k.bias | 0x62910c80 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x62911480 | 0xab800 | +| 137 | blk.11.attn_norm.weight | 0x629bcc80 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x629c0480 | 0x6e4000 | +| 139 | blk.11.attn_q.bias | 0x630a4480 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x630a7c80 | 0x4b0800 | +| 141 | blk.11.attn_v.bias | 0x63558480 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x63558c80 | 0xc0800 | +| 143 | blk.11.ffn_down.weight | 0x63619480 | 0x246c000 | +| 144 | blk.11.ffn_gate.weight | 0x65a85480 | 0x18c9800 | +| 145 | blk.11.ffn_norm.weight | 0x6734ec80 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x67352480 | 0x18c9800 | +| 147 | blk.12.attn_k.bias | 0x68c1bc80 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x68c1c480 | 0xab800 | +| 149 | blk.12.attn_norm.weight | 0x68cc7c80 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x68ccb480 | 0x6e4000 | +| 151 | blk.12.attn_q.bias | 0x693af480 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x693b2c80 | 0x4b0800 | +| 153 | blk.12.attn_v.bias | 0x69863480 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x69863c80 | 0xc0800 | +| 155 | blk.12.ffn_down.weight | 0x69924480 | 0x246c000 | +| 156 | blk.12.ffn_gate.weight | 0x6bd90480 | 0x18c9800 | +| 157 | blk.12.ffn_norm.weight | 0x6d659c80 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x6d65d480 | 0x18c9800 | +| 159 | blk.13.attn_k.bias | 0x6ef26c80 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x6ef27480 | 0xab800 | +| 161 | blk.13.attn_norm.weight | 0x6efd2c80 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x6efd6480 | 0x6e4000 | +| 163 | blk.13.attn_q.bias | 0x6f6ba480 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x6f6bdc80 | 0x4b0800 | +| 165 | blk.13.attn_v.bias | 0x6fb6e480 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x6fb6ec80 | 0xc0800 | +| 167 | blk.13.ffn_down.weight | 0x6fc2f480 | 0x246c000 | +| 168 | blk.13.ffn_gate.weight | 0x7209b480 | 0x18c9800 | +| 169 | blk.13.ffn_norm.weight | 0x73964c80 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x73968480 | 0x18c9800 | +| 171 | blk.14.attn_k.bias | 0x75231c80 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x75232480 | 0xc0800 | +| 173 | blk.14.attn_norm.weight | 0x752f2c80 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x752f6480 | 0x6e4000 | +| 175 | blk.14.attn_q.bias | 0x759da480 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x759ddc80 | 0x543800 | +| 177 | blk.14.attn_v.bias | 0x75f21480 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x75f21c80 | 0xfc000 | +| 179 | blk.14.ffn_down.weight | 0x7601dc80 | 0x246c000 | +| 180 | blk.14.ffn_gate.weight | 0x78489c80 | 0x18c9800 | +| 181 | blk.14.ffn_norm.weight | 0x79d53480 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x79d56c80 | 0x18c9800 | +| 183 | blk.15.attn_k.bias | 0x7b620480 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x7b620c80 | 0xc0800 | +| 185 | blk.15.attn_norm.weight | 0x7b6e1480 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x7b6e4c80 | 0x6e4000 | +| 187 | blk.15.attn_q.bias | 0x7bdc8c80 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x7bdcc480 | 0x543800 | +| 189 | blk.15.attn_v.bias | 0x7c30fc80 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x7c310480 | 0xfc000 | +| 191 | blk.15.ffn_down.weight | 0x7c40c480 | 0x246c000 | +| 192 | blk.15.ffn_gate.weight | 0x7e878480 | 0x18c9800 | +| 193 | blk.15.ffn_norm.weight | 0x80141c80 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x80145480 | 0x18c9800 | +| 195 | blk.16.attn_k.bias | 0x81a0ec80 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x81a0f480 | 0xc0800 | +| 197 | blk.16.attn_norm.weight | 0x81acfc80 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x81ad3480 | 0x6e4000 | +| 199 | blk.16.attn_q.bias | 0x821b7480 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x821bac80 | 0x543800 | +| 201 | blk.16.attn_v.bias | 0x826fe480 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x826fec80 | 0xfc000 | +| 203 | blk.16.ffn_down.weight | 0x827fac80 | 0x246c000 | +| 204 | blk.16.ffn_gate.weight | 0x84c66c80 | 0x18c9800 | +| 205 | blk.16.ffn_norm.weight | 0x86530480 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x86533c80 | 0x18c9800 | +| 207 | blk.17.attn_k.bias | 0x87dfd480 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x87dfdc80 | 0xc0800 | +| 209 | blk.17.attn_norm.weight | 0x87ebe480 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x87ec1c80 | 0x6e4000 | +| 211 | blk.17.attn_q.bias | 0x885a5c80 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x885a9480 | 0x543800 | +| 213 | blk.17.attn_v.bias | 0x88aecc80 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x88aed480 | 0xfc000 | +| 215 | blk.17.ffn_down.weight | 0x88be9480 | 0x246c000 | +| 216 | blk.17.ffn_gate.weight | 0x8b055480 | 0x18c9800 | +| 217 | blk.17.ffn_norm.weight | 0x8c91ec80 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x8c922480 | 0x18c9800 | +| 219 | blk.18.attn_k.bias | 0x8e1ebc80 | 0x800 | +| 220 | blk.18.attn_k.weight | 0x8e1ec480 | 0xc0800 | +| 221 | blk.18.attn_norm.weight | 0x8e2acc80 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0x8e2b0480 | 0x6e4000 | +| 223 | blk.18.attn_q.bias | 0x8e994480 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0x8e997c80 | 0x543800 | +| 225 | blk.18.attn_v.bias | 0x8eedb480 | 0x800 | +| 226 | blk.18.attn_v.weight | 0x8eedbc80 | 0xfc000 | +| 227 | blk.18.ffn_down.weight | 0x8efd7c80 | 0x246c000 | +| 228 | blk.18.ffn_gate.weight | 0x91443c80 | 0x18c9800 | +| 229 | blk.18.ffn_norm.weight | 0x92d0d480 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x92d10c80 | 0x18c9800 | +| 231 | blk.19.attn_k.bias | 0x945da480 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x945dac80 | 0xc0800 | +| 233 | blk.19.attn_norm.weight | 0x9469b480 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x9469ec80 | 0x6e4000 | +| 235 | blk.19.attn_q.bias | 0x94d82c80 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x94d86480 | 0x543800 | +| 237 | blk.19.attn_v.bias | 0x952c9c80 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x952ca480 | 0xfc000 | +| 239 | blk.19.ffn_down.weight | 0x953c6480 | 0x246c000 | +| 240 | blk.19.ffn_gate.weight | 0x97832480 | 0x18c9800 | +| 241 | blk.19.ffn_norm.weight | 0x990fbc80 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x990ff480 | 0x18c9800 | +| 243 | blk.20.attn_k.bias | 0x9a9c8c80 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x9a9c9480 | 0xc0800 | +| 245 | blk.20.attn_norm.weight | 0x9aa89c80 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x9aa8d480 | 0x6e4000 | +| 247 | blk.20.attn_q.bias | 0x9b171480 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x9b174c80 | 0x543800 | +| 249 | blk.20.attn_v.bias | 0x9b6b8480 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x9b6b8c80 | 0xfc000 | +| 251 | blk.20.ffn_down.weight | 0x9b7b4c80 | 0x246c000 | +| 252 | blk.20.ffn_gate.weight | 0x9dc20c80 | 0x1bd2800 | +| 253 | blk.20.ffn_norm.weight | 0x9f7f3480 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x9f7f6c80 | 0x1bd2800 | +| 255 | blk.21.attn_k.bias | 0xa13c9480 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xa13c9c80 | 0xc0800 | +| 257 | blk.21.attn_norm.weight | 0xa148a480 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xa148dc80 | 0x6e4000 | +| 259 | blk.21.attn_q.bias | 0xa1b71c80 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xa1b75480 | 0x543800 | +| 261 | blk.21.attn_v.bias | 0xa20b8c80 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xa20b9480 | 0xfc000 | +| 263 | blk.21.ffn_down.weight | 0xa21b5480 | 0x246c000 | +| 264 | blk.21.ffn_gate.weight | 0xa4621480 | 0x1bd2800 | +| 265 | blk.21.ffn_norm.weight | 0xa61f3c80 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xa61f7480 | 0x1bd2800 | +| 267 | blk.22.attn_k.bias | 0xa7dc9c80 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xa7dca480 | 0xc0800 | +| 269 | blk.22.attn_norm.weight | 0xa7e8ac80 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xa7e8e480 | 0x6e4000 | +| 271 | blk.22.attn_q.bias | 0xa8572480 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xa8575c80 | 0x543800 | +| 273 | blk.22.attn_v.bias | 0xa8ab9480 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xa8ab9c80 | 0xfc000 | +| 275 | blk.22.ffn_down.weight | 0xa8bb5c80 | 0x246c000 | +| 276 | blk.22.ffn_gate.weight | 0xab021c80 | 0x1bd2800 | +| 277 | blk.22.ffn_norm.weight | 0xacbf4480 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xacbf7c80 | 0x1bd2800 | +| 279 | blk.23.attn_k.bias | 0xae7ca480 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xae7cac80 | 0xc0800 | +| 281 | blk.23.attn_norm.weight | 0xae88b480 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xae88ec80 | 0x6e4000 | +| 283 | blk.23.attn_q.bias | 0xaef72c80 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xaef76480 | 0x543800 | +| 285 | blk.23.attn_v.bias | 0xaf4b9c80 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xaf4ba480 | 0xfc000 | +| 287 | blk.23.ffn_down.weight | 0xaf5b6480 | 0x246c000 | +| 288 | blk.23.ffn_gate.weight | 0xb1a22480 | 0x1bd2800 | +| 289 | blk.23.ffn_norm.weight | 0xb35f4c80 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xb35f8480 | 0x1bd2800 | +| 291 | blk.24.attn_k.bias | 0xb51cac80 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xb51cb480 | 0xc0800 | +| 293 | blk.24.attn_norm.weight | 0xb528bc80 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xb528f480 | 0x6e4000 | +| 295 | blk.24.attn_q.bias | 0xb5973480 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xb5976c80 | 0x543800 | +| 297 | blk.24.attn_v.bias | 0xb5eba480 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xb5ebac80 | 0xfc000 | +| 299 | blk.24.ffn_down.weight | 0xb5fb6c80 | 0x246c000 | +| 300 | blk.24.ffn_gate.weight | 0xb8422c80 | 0x1bd2800 | +| 301 | blk.24.ffn_norm.weight | 0xb9ff5480 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xb9ff8c80 | 0x1bd2800 | +| 303 | blk.25.attn_k.bias | 0xbbbcb480 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xbbbcbc80 | 0xc0800 | +| 305 | blk.25.attn_norm.weight | 0xbbc8c480 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xbbc8fc80 | 0x6e4000 | +| 307 | blk.25.attn_q.bias | 0xbc373c80 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xbc377480 | 0x543800 | +| 309 | blk.25.attn_v.bias | 0xbc8bac80 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xbc8bb480 | 0xfc000 | +| 311 | blk.25.ffn_down.weight | 0xbc9b7480 | 0x246c000 | +| 312 | blk.25.ffn_gate.weight | 0xbee23480 | 0x1bd2800 | +| 313 | blk.25.ffn_norm.weight | 0xc09f5c80 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xc09f9480 | 0x1bd2800 | +| 315 | blk.26.attn_k.bias | 0xc25cbc80 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xc25cc480 | 0xc0800 | +| 317 | blk.26.attn_norm.weight | 0xc268cc80 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xc2690480 | 0x6e4000 | +| 319 | blk.26.attn_q.bias | 0xc2d74480 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xc2d77c80 | 0x543800 | +| 321 | blk.26.attn_v.bias | 0xc32bb480 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xc32bbc80 | 0xfc000 | +| 323 | blk.26.ffn_down.weight | 0xc33b7c80 | 0x246c000 | +| 324 | blk.26.ffn_gate.weight | 0xc5823c80 | 0x1bd2800 | +| 325 | blk.26.ffn_norm.weight | 0xc73f6480 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xc73f9c80 | 0x1bd2800 | +| 327 | blk.27.attn_k.bias | 0xc8fcc480 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xc8fccc80 | 0xc0800 | +| 329 | blk.27.attn_norm.weight | 0xc908d480 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xc9090c80 | 0x6e4000 | +| 331 | blk.27.attn_q.bias | 0xc9774c80 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xc9778480 | 0x543800 | +| 333 | blk.27.attn_v.bias | 0xc9cbbc80 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xc9cbc480 | 0xfc000 | +| 335 | blk.27.ffn_down.weight | 0xc9db8480 | 0x246c000 | +| 336 | blk.27.ffn_gate.weight | 0xcc224480 | 0x1bd2800 | +| 337 | blk.27.ffn_norm.weight | 0xcddf6c80 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xcddfa480 | 0x1bd2800 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:------| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | IQ3_S | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | IQ3_S | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-IQ3_S.md b/scores/Hammer2.1-7b-IQ3_S.md new file mode 100644 index 0000000..c934bad --- /dev/null +++ b/scores/Hammer2.1-7b-IQ3_S.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-IQ3_S.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 26 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-IQ3\_S.gguf - GGUF Internal File Dump](#hammer21-7b-iq3_sgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0xc671d9c | +| 1 | output_norm.weight | 0xcc1d260 | 0x3800 | +| 2 | token_embd.weight | 0xcc20a60 | 0xc671d9c | +| 3 | blk.0.attn_k.bias | 0x19292800 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x19293000 | 0xab800 | +| 5 | blk.0.attn_norm.weight | 0x1933e800 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x19342000 | 0x543800 | +| 7 | blk.0.attn_q.bias | 0x19885800 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x19889000 | 0x4b0800 | +| 9 | blk.0.attn_v.bias | 0x19d39800 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x19d3a000 | 0xc0800 | +| 11 | blk.0.ffn_down.weight | 0x19dfa800 | 0x1bd2800 | +| 12 | blk.0.ffn_gate.weight | 0x1b9cd000 | 0x18c9800 | +| 13 | blk.0.ffn_norm.weight | 0x1d296800 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x1d29a000 | 0x18c9800 | +| 15 | blk.1.attn_k.bias | 0x1eb63800 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x1eb64000 | 0xab800 | +| 17 | blk.1.attn_norm.weight | 0x1ec0f800 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x1ec13000 | 0x543800 | +| 19 | blk.1.attn_q.bias | 0x1f156800 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x1f15a000 | 0x4b0800 | +| 21 | blk.1.attn_v.bias | 0x1f60a800 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x1f60b000 | 0xc0800 | +| 23 | blk.1.ffn_down.weight | 0x1f6cb800 | 0x1bd2800 | +| 24 | blk.1.ffn_gate.weight | 0x2129e000 | 0x1bd2800 | +| 25 | blk.1.ffn_norm.weight | 0x22e70800 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x22e74000 | 0x1bd2800 | +| 27 | blk.2.attn_k.bias | 0x24a46800 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x24a47000 | 0xab800 | +| 29 | blk.2.attn_norm.weight | 0x24af2800 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x24af6000 | 0x543800 | +| 31 | blk.2.attn_q.bias | 0x25039800 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x2503d000 | 0x4b0800 | +| 33 | blk.2.attn_v.bias | 0x254ed800 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x254ee000 | 0xc0800 | +| 35 | blk.2.ffn_down.weight | 0x255ae800 | 0x1bd2800 | +| 36 | blk.2.ffn_gate.weight | 0x27181000 | 0x1bd2800 | +| 37 | blk.2.ffn_norm.weight | 0x28d53800 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x28d57000 | 0x1bd2800 | +| 39 | blk.3.attn_k.bias | 0x2a929800 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x2a92a000 | 0xab800 | +| 41 | blk.3.attn_norm.weight | 0x2a9d5800 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x2a9d9000 | 0x543800 | +| 43 | blk.3.attn_q.bias | 0x2af1c800 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x2af20000 | 0x4b0800 | +| 45 | blk.3.attn_v.bias | 0x2b3d0800 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x2b3d1000 | 0xc0800 | +| 47 | blk.3.ffn_down.weight | 0x2b491800 | 0x246c000 | +| 48 | blk.3.ffn_gate.weight | 0x2d8fd800 | 0x1bd2800 | +| 49 | blk.3.ffn_norm.weight | 0x2f4d0000 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x2f4d3800 | 0x1bd2800 | +| 51 | blk.4.attn_k.bias | 0x310a6000 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x310a6800 | 0xab800 | +| 53 | blk.4.attn_norm.weight | 0x31152000 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x31155800 | 0x543800 | +| 55 | blk.4.attn_q.bias | 0x31699000 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x3169c800 | 0x4b0800 | +| 57 | blk.4.attn_v.bias | 0x31b4d000 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x31b4d800 | 0xc0800 | +| 59 | blk.4.ffn_down.weight | 0x31c0e000 | 0x1bd2800 | +| 60 | blk.4.ffn_gate.weight | 0x337e0800 | 0x1bd2800 | +| 61 | blk.4.ffn_norm.weight | 0x353b3000 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x353b6800 | 0x1bd2800 | +| 63 | blk.5.attn_k.bias | 0x36f89000 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x36f89800 | 0xab800 | +| 65 | blk.5.attn_norm.weight | 0x37035000 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x37038800 | 0x543800 | +| 67 | blk.5.attn_q.bias | 0x3757c000 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x3757f800 | 0x4b0800 | +| 69 | blk.5.attn_v.bias | 0x37a30000 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x37a30800 | 0xc0800 | +| 71 | blk.5.ffn_down.weight | 0x37af1000 | 0x1bd2800 | +| 72 | blk.5.ffn_gate.weight | 0x396c3800 | 0x1bd2800 | +| 73 | blk.5.ffn_norm.weight | 0x3b296000 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x3b299800 | 0x1bd2800 | +| 75 | blk.6.attn_k.bias | 0x3ce6c000 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x3ce6c800 | 0xab800 | +| 77 | blk.6.attn_norm.weight | 0x3cf18000 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x3cf1b800 | 0x543800 | +| 79 | blk.6.attn_q.bias | 0x3d45f000 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x3d462800 | 0x4b0800 | +| 81 | blk.6.attn_v.bias | 0x3d913000 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x3d913800 | 0xc0800 | +| 83 | blk.6.ffn_down.weight | 0x3d9d4000 | 0x1bd2800 | +| 84 | blk.6.ffn_gate.weight | 0x3f5a6800 | 0x18c9800 | +| 85 | blk.6.ffn_norm.weight | 0x40e70000 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x40e73800 | 0x18c9800 | +| 87 | blk.7.attn_k.bias | 0x4273d000 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x4273d800 | 0xab800 | +| 89 | blk.7.attn_norm.weight | 0x427e9000 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x427ec800 | 0x543800 | +| 91 | blk.7.attn_q.bias | 0x42d30000 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x42d33800 | 0x4b0800 | +| 93 | blk.7.attn_v.bias | 0x431e4000 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x431e4800 | 0xc0800 | +| 95 | blk.7.ffn_down.weight | 0x432a5000 | 0x246c000 | +| 96 | blk.7.ffn_gate.weight | 0x45711000 | 0x18c9800 | +| 97 | blk.7.ffn_norm.weight | 0x46fda800 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x46fde000 | 0x18c9800 | +| 99 | blk.8.attn_k.bias | 0x488a7800 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x488a8000 | 0xab800 | +| 101 | blk.8.attn_norm.weight | 0x48953800 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x48957000 | 0x543800 | +| 103 | blk.8.attn_q.bias | 0x48e9a800 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x48e9e000 | 0x4b0800 | +| 105 | blk.8.attn_v.bias | 0x4934e800 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x4934f000 | 0xc0800 | +| 107 | blk.8.ffn_down.weight | 0x4940f800 | 0x1bd2800 | +| 108 | blk.8.ffn_gate.weight | 0x4afe2000 | 0x18c9800 | +| 109 | blk.8.ffn_norm.weight | 0x4c8ab800 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x4c8af000 | 0x18c9800 | +| 111 | blk.9.attn_k.bias | 0x4e178800 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x4e179000 | 0xab800 | +| 113 | blk.9.attn_norm.weight | 0x4e224800 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x4e228000 | 0x543800 | +| 115 | blk.9.attn_q.bias | 0x4e76b800 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x4e76f000 | 0x4b0800 | +| 117 | blk.9.attn_v.bias | 0x4ec1f800 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x4ec20000 | 0xc0800 | +| 119 | blk.9.ffn_down.weight | 0x4ece0800 | 0x246c000 | +| 120 | blk.9.ffn_gate.weight | 0x5114c800 | 0x1bd2800 | +| 121 | blk.9.ffn_norm.weight | 0x52d1f000 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x52d22800 | 0x1bd2800 | +| 123 | blk.10.attn_k.bias | 0x548f5000 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x548f5800 | 0xab800 | +| 125 | blk.10.attn_norm.weight | 0x549a1000 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x549a4800 | 0x543800 | +| 127 | blk.10.attn_q.bias | 0x54ee8000 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x54eeb800 | 0x4b0800 | +| 129 | blk.10.attn_v.bias | 0x5539c000 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x5539c800 | 0xc0800 | +| 131 | blk.10.ffn_down.weight | 0x5545d000 | 0x246c000 | +| 132 | blk.10.ffn_gate.weight | 0x578c9000 | 0x18c9800 | +| 133 | blk.10.ffn_norm.weight | 0x59192800 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x59196000 | 0x18c9800 | +| 135 | blk.11.attn_k.bias | 0x5aa5f800 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x5aa60000 | 0xab800 | +| 137 | blk.11.attn_norm.weight | 0x5ab0b800 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x5ab0f000 | 0x543800 | +| 139 | blk.11.attn_q.bias | 0x5b052800 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x5b056000 | 0x4b0800 | +| 141 | blk.11.attn_v.bias | 0x5b506800 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x5b507000 | 0xc0800 | +| 143 | blk.11.ffn_down.weight | 0x5b5c7800 | 0x1bd2800 | +| 144 | blk.11.ffn_gate.weight | 0x5d19a000 | 0x18c9800 | +| 145 | blk.11.ffn_norm.weight | 0x5ea63800 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x5ea67000 | 0x18c9800 | +| 147 | blk.12.attn_k.bias | 0x60330800 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x60331000 | 0xab800 | +| 149 | blk.12.attn_norm.weight | 0x603dc800 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x603e0000 | 0x543800 | +| 151 | blk.12.attn_q.bias | 0x60923800 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x60927000 | 0x4b0800 | +| 153 | blk.12.attn_v.bias | 0x60dd7800 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x60dd8000 | 0xc0800 | +| 155 | blk.12.ffn_down.weight | 0x60e98800 | 0x1bd2800 | +| 156 | blk.12.ffn_gate.weight | 0x62a6b000 | 0x18c9800 | +| 157 | blk.12.ffn_norm.weight | 0x64334800 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x64338000 | 0x18c9800 | +| 159 | blk.13.attn_k.bias | 0x65c01800 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x65c02000 | 0xab800 | +| 161 | blk.13.attn_norm.weight | 0x65cad800 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x65cb1000 | 0x543800 | +| 163 | blk.13.attn_q.bias | 0x661f4800 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x661f8000 | 0x4b0800 | +| 165 | blk.13.attn_v.bias | 0x666a8800 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x666a9000 | 0xc0800 | +| 167 | blk.13.ffn_down.weight | 0x66769800 | 0x1bd2800 | +| 168 | blk.13.ffn_gate.weight | 0x6833c000 | 0x18c9800 | +| 169 | blk.13.ffn_norm.weight | 0x69c05800 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x69c09000 | 0x18c9800 | +| 171 | blk.14.attn_k.bias | 0x6b4d2800 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x6b4d3000 | 0xc0800 | +| 173 | blk.14.attn_norm.weight | 0x6b593800 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x6b597000 | 0x543800 | +| 175 | blk.14.attn_q.bias | 0x6bada800 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x6bade000 | 0x543800 | +| 177 | blk.14.attn_v.bias | 0x6c021800 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x6c022000 | 0xc0800 | +| 179 | blk.14.ffn_down.weight | 0x6c0e2800 | 0x1bd2800 | +| 180 | blk.14.ffn_gate.weight | 0x6dcb5000 | 0x18c9800 | +| 181 | blk.14.ffn_norm.weight | 0x6f57e800 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x6f582000 | 0x18c9800 | +| 183 | blk.15.attn_k.bias | 0x70e4b800 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x70e4c000 | 0xc0800 | +| 185 | blk.15.attn_norm.weight | 0x70f0c800 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x70f10000 | 0x543800 | +| 187 | blk.15.attn_q.bias | 0x71453800 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x71457000 | 0x543800 | +| 189 | blk.15.attn_v.bias | 0x7199a800 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x7199b000 | 0xc0800 | +| 191 | blk.15.ffn_down.weight | 0x71a5b800 | 0x1bd2800 | +| 192 | blk.15.ffn_gate.weight | 0x7362e000 | 0x18c9800 | +| 193 | blk.15.ffn_norm.weight | 0x74ef7800 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x74efb000 | 0x18c9800 | +| 195 | blk.16.attn_k.bias | 0x767c4800 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x767c5000 | 0xc0800 | +| 197 | blk.16.attn_norm.weight | 0x76885800 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x76889000 | 0x543800 | +| 199 | blk.16.attn_q.bias | 0x76dcc800 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x76dd0000 | 0x543800 | +| 201 | blk.16.attn_v.bias | 0x77313800 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x77314000 | 0xc0800 | +| 203 | blk.16.ffn_down.weight | 0x773d4800 | 0x1bd2800 | +| 204 | blk.16.ffn_gate.weight | 0x78fa7000 | 0x18c9800 | +| 205 | blk.16.ffn_norm.weight | 0x7a870800 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x7a874000 | 0x18c9800 | +| 207 | blk.17.attn_k.bias | 0x7c13d800 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x7c13e000 | 0xc0800 | +| 209 | blk.17.attn_norm.weight | 0x7c1fe800 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x7c202000 | 0x543800 | +| 211 | blk.17.attn_q.bias | 0x7c745800 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x7c749000 | 0x543800 | +| 213 | blk.17.attn_v.bias | 0x7cc8c800 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x7cc8d000 | 0xc0800 | +| 215 | blk.17.ffn_down.weight | 0x7cd4d800 | 0x1bd2800 | +| 216 | blk.17.ffn_gate.weight | 0x7e920000 | 0x18c9800 | +| 217 | blk.17.ffn_norm.weight | 0x801e9800 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x801ed000 | 0x18c9800 | +| 219 | blk.18.attn_k.bias | 0x81ab6800 | 0x800 | +| 220 | blk.18.attn_k.weight | 0x81ab7000 | 0xc0800 | +| 221 | blk.18.attn_norm.weight | 0x81b77800 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0x81b7b000 | 0x543800 | +| 223 | blk.18.attn_q.bias | 0x820be800 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0x820c2000 | 0x543800 | +| 225 | blk.18.attn_v.bias | 0x82605800 | 0x800 | +| 226 | blk.18.attn_v.weight | 0x82606000 | 0xc0800 | +| 227 | blk.18.ffn_down.weight | 0x826c6800 | 0x246c000 | +| 228 | blk.18.ffn_gate.weight | 0x84b32800 | 0x18c9800 | +| 229 | blk.18.ffn_norm.weight | 0x863fc000 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x863ff800 | 0x18c9800 | +| 231 | blk.19.attn_k.bias | 0x87cc9000 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x87cc9800 | 0xc0800 | +| 233 | blk.19.attn_norm.weight | 0x87d8a000 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x87d8d800 | 0x543800 | +| 235 | blk.19.attn_q.bias | 0x882d1000 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x882d4800 | 0x543800 | +| 237 | blk.19.attn_v.bias | 0x88818000 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x88818800 | 0xc0800 | +| 239 | blk.19.ffn_down.weight | 0x888d9000 | 0x246c000 | +| 240 | blk.19.ffn_gate.weight | 0x8ad45000 | 0x18c9800 | +| 241 | blk.19.ffn_norm.weight | 0x8c60e800 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x8c612000 | 0x18c9800 | +| 243 | blk.20.attn_k.bias | 0x8dedb800 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x8dedc000 | 0xc0800 | +| 245 | blk.20.attn_norm.weight | 0x8df9c800 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x8dfa0000 | 0x543800 | +| 247 | blk.20.attn_q.bias | 0x8e4e3800 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x8e4e7000 | 0x543800 | +| 249 | blk.20.attn_v.bias | 0x8ea2a800 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x8ea2b000 | 0xc0800 | +| 251 | blk.20.ffn_down.weight | 0x8eaeb800 | 0x246c000 | +| 252 | blk.20.ffn_gate.weight | 0x90f57800 | 0x1bd2800 | +| 253 | blk.20.ffn_norm.weight | 0x92b2a000 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x92b2d800 | 0x1bd2800 | +| 255 | blk.21.attn_k.bias | 0x94700000 | 0x800 | +| 256 | blk.21.attn_k.weight | 0x94700800 | 0xc0800 | +| 257 | blk.21.attn_norm.weight | 0x947c1000 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0x947c4800 | 0x543800 | +| 259 | blk.21.attn_q.bias | 0x94d08000 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0x94d0b800 | 0x543800 | +| 261 | blk.21.attn_v.bias | 0x9524f000 | 0x800 | +| 262 | blk.21.attn_v.weight | 0x9524f800 | 0xc0800 | +| 263 | blk.21.ffn_down.weight | 0x95310000 | 0x246c000 | +| 264 | blk.21.ffn_gate.weight | 0x9777c000 | 0x1bd2800 | +| 265 | blk.21.ffn_norm.weight | 0x9934e800 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0x99352000 | 0x1bd2800 | +| 267 | blk.22.attn_k.bias | 0x9af24800 | 0x800 | +| 268 | blk.22.attn_k.weight | 0x9af25000 | 0xc0800 | +| 269 | blk.22.attn_norm.weight | 0x9afe5800 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0x9afe9000 | 0x543800 | +| 271 | blk.22.attn_q.bias | 0x9b52c800 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0x9b530000 | 0x543800 | +| 273 | blk.22.attn_v.bias | 0x9ba73800 | 0x800 | +| 274 | blk.22.attn_v.weight | 0x9ba74000 | 0xc0800 | +| 275 | blk.22.ffn_down.weight | 0x9bb34800 | 0x246c000 | +| 276 | blk.22.ffn_gate.weight | 0x9dfa0800 | 0x1bd2800 | +| 277 | blk.22.ffn_norm.weight | 0x9fb73000 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0x9fb76800 | 0x1bd2800 | +| 279 | blk.23.attn_k.bias | 0xa1749000 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xa1749800 | 0xc0800 | +| 281 | blk.23.attn_norm.weight | 0xa180a000 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xa180d800 | 0x543800 | +| 283 | blk.23.attn_q.bias | 0xa1d51000 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xa1d54800 | 0x543800 | +| 285 | blk.23.attn_v.bias | 0xa2298000 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xa2298800 | 0xc0800 | +| 287 | blk.23.ffn_down.weight | 0xa2359000 | 0x246c000 | +| 288 | blk.23.ffn_gate.weight | 0xa47c5000 | 0x1bd2800 | +| 289 | blk.23.ffn_norm.weight | 0xa6397800 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xa639b000 | 0x1bd2800 | +| 291 | blk.24.attn_k.bias | 0xa7f6d800 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xa7f6e000 | 0xc0800 | +| 293 | blk.24.attn_norm.weight | 0xa802e800 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xa8032000 | 0x543800 | +| 295 | blk.24.attn_q.bias | 0xa8575800 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xa8579000 | 0x543800 | +| 297 | blk.24.attn_v.bias | 0xa8abc800 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xa8abd000 | 0xc0800 | +| 299 | blk.24.ffn_down.weight | 0xa8b7d800 | 0x246c000 | +| 300 | blk.24.ffn_gate.weight | 0xaafe9800 | 0x1bd2800 | +| 301 | blk.24.ffn_norm.weight | 0xacbbc000 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xacbbf800 | 0x1bd2800 | +| 303 | blk.25.attn_k.bias | 0xae792000 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xae792800 | 0xc0800 | +| 305 | blk.25.attn_norm.weight | 0xae853000 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xae856800 | 0x543800 | +| 307 | blk.25.attn_q.bias | 0xaed9a000 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xaed9d800 | 0x543800 | +| 309 | blk.25.attn_v.bias | 0xaf2e1000 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xaf2e1800 | 0xc0800 | +| 311 | blk.25.ffn_down.weight | 0xaf3a2000 | 0x246c000 | +| 312 | blk.25.ffn_gate.weight | 0xb180e000 | 0x1bd2800 | +| 313 | blk.25.ffn_norm.weight | 0xb33e0800 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xb33e4000 | 0x1bd2800 | +| 315 | blk.26.attn_k.bias | 0xb4fb6800 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xb4fb7000 | 0xc0800 | +| 317 | blk.26.attn_norm.weight | 0xb5077800 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xb507b000 | 0x543800 | +| 319 | blk.26.attn_q.bias | 0xb55be800 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xb55c2000 | 0x543800 | +| 321 | blk.26.attn_v.bias | 0xb5b05800 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xb5b06000 | 0xc0800 | +| 323 | blk.26.ffn_down.weight | 0xb5bc6800 | 0x246c000 | +| 324 | blk.26.ffn_gate.weight | 0xb8032800 | 0x1bd2800 | +| 325 | blk.26.ffn_norm.weight | 0xb9c05000 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xb9c08800 | 0x1bd2800 | +| 327 | blk.27.attn_k.bias | 0xbb7db000 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xbb7db800 | 0xc0800 | +| 329 | blk.27.attn_norm.weight | 0xbb89c000 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xbb89f800 | 0x543800 | +| 331 | blk.27.attn_q.bias | 0xbbde3000 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xbbde6800 | 0x543800 | +| 333 | blk.27.attn_v.bias | 0xbc32a000 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xbc32a800 | 0xc0800 | +| 335 | blk.27.ffn_down.weight | 0xbc3eb000 | 0x246c000 | +| 336 | blk.27.ffn_gate.weight | 0xbe857000 | 0x1bd2800 | +| 337 | blk.27.ffn_norm.weight | 0xc0429800 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xc042d000 | 0x1bd2800 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:--------| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | IQ3_XXS | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | IQ3_XXS | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:--------| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_XXS | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_XXS | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ3_S | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:--------| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_XXS | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | IQ4_NL | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-IQ4_NL.md b/scores/Hammer2.1-7b-IQ4_NL.md new file mode 100644 index 0000000..6ea20bc --- /dev/null +++ b/scores/Hammer2.1-7b-IQ4_NL.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-IQ4_NL.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 25 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-IQ4\_NL.gguf - GGUF Internal File Dump](#hammer21-7b-iq4_nlgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x123979e0 | +| 1 | output_norm.weight | 0x12942ea0 | 0x3800 | +| 2 | token_embd.weight | 0x129466a0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x20804e80 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x20805680 | 0xc0800 | +| 5 | blk.0.attn_norm.weight | 0x208c5e80 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x208c9680 | 0x6e4000 | +| 7 | blk.0.attn_q.bias | 0x20fad680 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x20fb0e80 | 0x543800 | +| 9 | blk.0.attn_v.bias | 0x214f4680 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x214f4e80 | 0xee000 | +| 11 | blk.0.ffn_down.weight | 0x215e2e80 | 0x2c84000 | +| 12 | blk.0.ffn_gate.weight | 0x24266e80 | 0x1bd2800 | +| 13 | blk.0.ffn_norm.weight | 0x25e39680 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x25e3ce80 | 0x1bd2800 | +| 15 | blk.1.attn_k.bias | 0x27a0f680 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x27a0fe80 | 0xc0800 | +| 17 | blk.1.attn_norm.weight | 0x27ad0680 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x27ad3e80 | 0x6e4000 | +| 19 | blk.1.attn_q.bias | 0x281b7e80 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x281bb680 | 0x543800 | +| 21 | blk.1.attn_v.bias | 0x286fee80 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x286ff680 | 0xee000 | +| 23 | blk.1.ffn_down.weight | 0x287ed680 | 0x2c84000 | +| 24 | blk.1.ffn_gate.weight | 0x2b471680 | 0x246c000 | +| 25 | blk.1.ffn_norm.weight | 0x2d8dd680 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x2d8e0e80 | 0x246c000 | +| 27 | blk.2.attn_k.bias | 0x2fd4ce80 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x2fd4d680 | 0xc0800 | +| 29 | blk.2.attn_norm.weight | 0x2fe0de80 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x2fe11680 | 0x6e4000 | +| 31 | blk.2.attn_q.bias | 0x304f5680 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x304f8e80 | 0x543800 | +| 33 | blk.2.attn_v.bias | 0x30a3c680 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x30a3ce80 | 0xee000 | +| 35 | blk.2.ffn_down.weight | 0x30b2ae80 | 0x2c84000 | +| 36 | blk.2.ffn_gate.weight | 0x337aee80 | 0x246c000 | +| 37 | blk.2.ffn_norm.weight | 0x35c1ae80 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x35c1e680 | 0x246c000 | +| 39 | blk.3.attn_k.bias | 0x3808a680 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x3808ae80 | 0xc0800 | +| 41 | blk.3.attn_norm.weight | 0x3814b680 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x3814ee80 | 0x6e4000 | +| 43 | blk.3.attn_q.bias | 0x38832e80 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x38836680 | 0x543800 | +| 45 | blk.3.attn_v.bias | 0x38d79e80 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x38d7a680 | 0xee000 | +| 47 | blk.3.ffn_down.weight | 0x38e68680 | 0x2c84000 | +| 48 | blk.3.ffn_gate.weight | 0x3baec680 | 0x246c000 | +| 49 | blk.3.ffn_norm.weight | 0x3df58680 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x3df5be80 | 0x246c000 | +| 51 | blk.4.attn_k.bias | 0x403c7e80 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x403c8680 | 0xc0800 | +| 53 | blk.4.attn_norm.weight | 0x40488e80 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x4048c680 | 0x6e4000 | +| 55 | blk.4.attn_q.bias | 0x40b70680 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x40b73e80 | 0x543800 | +| 57 | blk.4.attn_v.bias | 0x410b7680 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x410b7e80 | 0xee000 | +| 59 | blk.4.ffn_down.weight | 0x411a5e80 | 0x2c84000 | +| 60 | blk.4.ffn_gate.weight | 0x43e29e80 | 0x246c000 | +| 61 | blk.4.ffn_norm.weight | 0x46295e80 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x46299680 | 0x246c000 | +| 63 | blk.5.attn_k.bias | 0x48705680 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x48705e80 | 0xc0800 | +| 65 | blk.5.attn_norm.weight | 0x487c6680 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x487c9e80 | 0x6e4000 | +| 67 | blk.5.attn_q.bias | 0x48eade80 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x48eb1680 | 0x543800 | +| 69 | blk.5.attn_v.bias | 0x493f4e80 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x493f5680 | 0xee000 | +| 71 | blk.5.ffn_down.weight | 0x494e3680 | 0x2c84000 | +| 72 | blk.5.ffn_gate.weight | 0x4c167680 | 0x246c000 | +| 73 | blk.5.ffn_norm.weight | 0x4e5d3680 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x4e5d6e80 | 0x246c000 | +| 75 | blk.6.attn_k.bias | 0x50a42e80 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x50a43680 | 0xc0800 | +| 77 | blk.6.attn_norm.weight | 0x50b03e80 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x50b07680 | 0x6e4000 | +| 79 | blk.6.attn_q.bias | 0x511eb680 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x511eee80 | 0x543800 | +| 81 | blk.6.attn_v.bias | 0x51732680 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x51732e80 | 0xee000 | +| 83 | blk.6.ffn_down.weight | 0x51820e80 | 0x2c84000 | +| 84 | blk.6.ffn_gate.weight | 0x544a4e80 | 0x1bd2800 | +| 85 | blk.6.ffn_norm.weight | 0x56077680 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x5607ae80 | 0x1bd2800 | +| 87 | blk.7.attn_k.bias | 0x57c4d680 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x57c4de80 | 0xc0800 | +| 89 | blk.7.attn_norm.weight | 0x57d0e680 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x57d11e80 | 0x6e4000 | +| 91 | blk.7.attn_q.bias | 0x583f5e80 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x583f9680 | 0x543800 | +| 93 | blk.7.attn_v.bias | 0x5893ce80 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x5893d680 | 0xee000 | +| 95 | blk.7.ffn_down.weight | 0x58a2b680 | 0x2c84000 | +| 96 | blk.7.ffn_gate.weight | 0x5b6af680 | 0x1bd2800 | +| 97 | blk.7.ffn_norm.weight | 0x5d281e80 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x5d285680 | 0x1bd2800 | +| 99 | blk.8.attn_k.bias | 0x5ee57e80 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x5ee58680 | 0xc0800 | +| 101 | blk.8.attn_norm.weight | 0x5ef18e80 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x5ef1c680 | 0x6e4000 | +| 103 | blk.8.attn_q.bias | 0x5f600680 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x5f603e80 | 0x543800 | +| 105 | blk.8.attn_v.bias | 0x5fb47680 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x5fb47e80 | 0xee000 | +| 107 | blk.8.ffn_down.weight | 0x5fc35e80 | 0x2c84000 | +| 108 | blk.8.ffn_gate.weight | 0x628b9e80 | 0x1bd2800 | +| 109 | blk.8.ffn_norm.weight | 0x6448c680 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x6448fe80 | 0x1bd2800 | +| 111 | blk.9.attn_k.bias | 0x66062680 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x66062e80 | 0xc0800 | +| 113 | blk.9.attn_norm.weight | 0x66123680 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x66126e80 | 0x6e4000 | +| 115 | blk.9.attn_q.bias | 0x6680ae80 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x6680e680 | 0x543800 | +| 117 | blk.9.attn_v.bias | 0x66d51e80 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x66d52680 | 0xee000 | +| 119 | blk.9.ffn_down.weight | 0x66e40680 | 0x2c84000 | +| 120 | blk.9.ffn_gate.weight | 0x69ac4680 | 0x246c000 | +| 121 | blk.9.ffn_norm.weight | 0x6bf30680 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x6bf33e80 | 0x246c000 | +| 123 | blk.10.attn_k.bias | 0x6e39fe80 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x6e3a0680 | 0xc0800 | +| 125 | blk.10.attn_norm.weight | 0x6e460e80 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x6e464680 | 0x6e4000 | +| 127 | blk.10.attn_q.bias | 0x6eb48680 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x6eb4be80 | 0x543800 | +| 129 | blk.10.attn_v.bias | 0x6f08f680 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x6f08fe80 | 0xee000 | +| 131 | blk.10.ffn_down.weight | 0x6f17de80 | 0x2c84000 | +| 132 | blk.10.ffn_gate.weight | 0x71e01e80 | 0x1bd2800 | +| 133 | blk.10.ffn_norm.weight | 0x739d4680 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x739d7e80 | 0x1bd2800 | +| 135 | blk.11.attn_k.bias | 0x755aa680 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x755aae80 | 0xc0800 | +| 137 | blk.11.attn_norm.weight | 0x7566b680 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x7566ee80 | 0x6e4000 | +| 139 | blk.11.attn_q.bias | 0x75d52e80 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x75d56680 | 0x543800 | +| 141 | blk.11.attn_v.bias | 0x76299e80 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x7629a680 | 0xee000 | +| 143 | blk.11.ffn_down.weight | 0x76388680 | 0x2c84000 | +| 144 | blk.11.ffn_gate.weight | 0x7900c680 | 0x1bd2800 | +| 145 | blk.11.ffn_norm.weight | 0x7abdee80 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x7abe2680 | 0x1bd2800 | +| 147 | blk.12.attn_k.bias | 0x7c7b4e80 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x7c7b5680 | 0xc0800 | +| 149 | blk.12.attn_norm.weight | 0x7c875e80 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x7c879680 | 0x6e4000 | +| 151 | blk.12.attn_q.bias | 0x7cf5d680 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x7cf60e80 | 0x543800 | +| 153 | blk.12.attn_v.bias | 0x7d4a4680 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x7d4a4e80 | 0xee000 | +| 155 | blk.12.ffn_down.weight | 0x7d592e80 | 0x2c84000 | +| 156 | blk.12.ffn_gate.weight | 0x80216e80 | 0x1bd2800 | +| 157 | blk.12.ffn_norm.weight | 0x81de9680 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x81dece80 | 0x1bd2800 | +| 159 | blk.13.attn_k.bias | 0x839bf680 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x839bfe80 | 0xc0800 | +| 161 | blk.13.attn_norm.weight | 0x83a80680 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x83a83e80 | 0x6e4000 | +| 163 | blk.13.attn_q.bias | 0x84167e80 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x8416b680 | 0x543800 | +| 165 | blk.13.attn_v.bias | 0x846aee80 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x846af680 | 0xee000 | +| 167 | blk.13.ffn_down.weight | 0x8479d680 | 0x2c84000 | +| 168 | blk.13.ffn_gate.weight | 0x87421680 | 0x1bd2800 | +| 169 | blk.13.ffn_norm.weight | 0x88ff3e80 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x88ff7680 | 0x1bd2800 | +| 171 | blk.14.attn_k.bias | 0x8abc9e80 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x8abca680 | 0xfc000 | +| 173 | blk.14.attn_norm.weight | 0x8acc6680 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x8acc9e80 | 0x6e4000 | +| 175 | blk.14.attn_q.bias | 0x8b3ade80 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x8b3b1680 | 0x6e4000 | +| 177 | blk.14.attn_v.bias | 0x8ba95680 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x8ba95e80 | 0xfc000 | +| 179 | blk.14.ffn_down.weight | 0x8bb91e80 | 0x2c84000 | +| 180 | blk.14.ffn_gate.weight | 0x8e815e80 | 0x1bd2800 | +| 181 | blk.14.ffn_norm.weight | 0x903e8680 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x903ebe80 | 0x1bd2800 | +| 183 | blk.15.attn_k.bias | 0x91fbe680 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x91fbee80 | 0xfc000 | +| 185 | blk.15.attn_norm.weight | 0x920bae80 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x920be680 | 0x6e4000 | +| 187 | blk.15.attn_q.bias | 0x927a2680 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x927a5e80 | 0x6e4000 | +| 189 | blk.15.attn_v.bias | 0x92e89e80 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x92e8a680 | 0xfc000 | +| 191 | blk.15.ffn_down.weight | 0x92f86680 | 0x2c84000 | +| 192 | blk.15.ffn_gate.weight | 0x95c0a680 | 0x1bd2800 | +| 193 | blk.15.ffn_norm.weight | 0x977dce80 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x977e0680 | 0x1bd2800 | +| 195 | blk.16.attn_k.bias | 0x993b2e80 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x993b3680 | 0xfc000 | +| 197 | blk.16.attn_norm.weight | 0x994af680 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x994b2e80 | 0x6e4000 | +| 199 | blk.16.attn_q.bias | 0x99b96e80 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x99b9a680 | 0x6e4000 | +| 201 | blk.16.attn_v.bias | 0x9a27e680 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x9a27ee80 | 0xfc000 | +| 203 | blk.16.ffn_down.weight | 0x9a37ae80 | 0x2c84000 | +| 204 | blk.16.ffn_gate.weight | 0x9cffee80 | 0x1bd2800 | +| 205 | blk.16.ffn_norm.weight | 0x9ebd1680 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x9ebd4e80 | 0x1bd2800 | +| 207 | blk.17.attn_k.bias | 0xa07a7680 | 0x800 | +| 208 | blk.17.attn_k.weight | 0xa07a7e80 | 0xfc000 | +| 209 | blk.17.attn_norm.weight | 0xa08a3e80 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0xa08a7680 | 0x6e4000 | +| 211 | blk.17.attn_q.bias | 0xa0f8b680 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0xa0f8ee80 | 0x6e4000 | +| 213 | blk.17.attn_v.bias | 0xa1672e80 | 0x800 | +| 214 | blk.17.attn_v.weight | 0xa1673680 | 0xfc000 | +| 215 | blk.17.ffn_down.weight | 0xa176f680 | 0x2c84000 | +| 216 | blk.17.ffn_gate.weight | 0xa43f3680 | 0x1bd2800 | +| 217 | blk.17.ffn_norm.weight | 0xa5fc5e80 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0xa5fc9680 | 0x1bd2800 | +| 219 | blk.18.attn_k.bias | 0xa7b9be80 | 0x800 | +| 220 | blk.18.attn_k.weight | 0xa7b9c680 | 0xfc000 | +| 221 | blk.18.attn_norm.weight | 0xa7c98680 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0xa7c9be80 | 0x6e4000 | +| 223 | blk.18.attn_q.bias | 0xa837fe80 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0xa8383680 | 0x6e4000 | +| 225 | blk.18.attn_v.bias | 0xa8a67680 | 0x800 | +| 226 | blk.18.attn_v.weight | 0xa8a67e80 | 0xfc000 | +| 227 | blk.18.ffn_down.weight | 0xa8b63e80 | 0x2c84000 | +| 228 | blk.18.ffn_gate.weight | 0xab7e7e80 | 0x1bd2800 | +| 229 | blk.18.ffn_norm.weight | 0xad3ba680 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0xad3bde80 | 0x1bd2800 | +| 231 | blk.19.attn_k.bias | 0xaef90680 | 0x800 | +| 232 | blk.19.attn_k.weight | 0xaef90e80 | 0xfc000 | +| 233 | blk.19.attn_norm.weight | 0xaf08ce80 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0xaf090680 | 0x6e4000 | +| 235 | blk.19.attn_q.bias | 0xaf774680 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0xaf777e80 | 0x6e4000 | +| 237 | blk.19.attn_v.bias | 0xafe5be80 | 0x800 | +| 238 | blk.19.attn_v.weight | 0xafe5c680 | 0xfc000 | +| 239 | blk.19.ffn_down.weight | 0xaff58680 | 0x2c84000 | +| 240 | blk.19.ffn_gate.weight | 0xb2bdc680 | 0x1bd2800 | +| 241 | blk.19.ffn_norm.weight | 0xb47aee80 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0xb47b2680 | 0x1bd2800 | +| 243 | blk.20.attn_k.bias | 0xb6384e80 | 0x800 | +| 244 | blk.20.attn_k.weight | 0xb6385680 | 0xfc000 | +| 245 | blk.20.attn_norm.weight | 0xb6481680 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0xb6484e80 | 0x6e4000 | +| 247 | blk.20.attn_q.bias | 0xb6b68e80 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0xb6b6c680 | 0x6e4000 | +| 249 | blk.20.attn_v.bias | 0xb7250680 | 0x800 | +| 250 | blk.20.attn_v.weight | 0xb7250e80 | 0xfc000 | +| 251 | blk.20.ffn_down.weight | 0xb734ce80 | 0x2c84000 | +| 252 | blk.20.ffn_gate.weight | 0xb9fd0e80 | 0x246c000 | +| 253 | blk.20.ffn_norm.weight | 0xbc43ce80 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0xbc440680 | 0x246c000 | +| 255 | blk.21.attn_k.bias | 0xbe8ac680 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xbe8ace80 | 0xfc000 | +| 257 | blk.21.attn_norm.weight | 0xbe9a8e80 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xbe9ac680 | 0x6e4000 | +| 259 | blk.21.attn_q.bias | 0xbf090680 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xbf093e80 | 0x6e4000 | +| 261 | blk.21.attn_v.bias | 0xbf777e80 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xbf778680 | 0xfc000 | +| 263 | blk.21.ffn_down.weight | 0xbf874680 | 0x2c84000 | +| 264 | blk.21.ffn_gate.weight | 0xc24f8680 | 0x246c000 | +| 265 | blk.21.ffn_norm.weight | 0xc4964680 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xc4967e80 | 0x246c000 | +| 267 | blk.22.attn_k.bias | 0xc6dd3e80 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xc6dd4680 | 0xfc000 | +| 269 | blk.22.attn_norm.weight | 0xc6ed0680 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xc6ed3e80 | 0x6e4000 | +| 271 | blk.22.attn_q.bias | 0xc75b7e80 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xc75bb680 | 0x6e4000 | +| 273 | blk.22.attn_v.bias | 0xc7c9f680 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xc7c9fe80 | 0xfc000 | +| 275 | blk.22.ffn_down.weight | 0xc7d9be80 | 0x2c84000 | +| 276 | blk.22.ffn_gate.weight | 0xcaa1fe80 | 0x246c000 | +| 277 | blk.22.ffn_norm.weight | 0xcce8be80 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xcce8f680 | 0x246c000 | +| 279 | blk.23.attn_k.bias | 0xcf2fb680 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xcf2fbe80 | 0xfc000 | +| 281 | blk.23.attn_norm.weight | 0xcf3f7e80 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xcf3fb680 | 0x6e4000 | +| 283 | blk.23.attn_q.bias | 0xcfadf680 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xcfae2e80 | 0x6e4000 | +| 285 | blk.23.attn_v.bias | 0xd01c6e80 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xd01c7680 | 0xfc000 | +| 287 | blk.23.ffn_down.weight | 0xd02c3680 | 0x2c84000 | +| 288 | blk.23.ffn_gate.weight | 0xd2f47680 | 0x246c000 | +| 289 | blk.23.ffn_norm.weight | 0xd53b3680 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xd53b6e80 | 0x246c000 | +| 291 | blk.24.attn_k.bias | 0xd7822e80 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xd7823680 | 0xfc000 | +| 293 | blk.24.attn_norm.weight | 0xd791f680 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xd7922e80 | 0x6e4000 | +| 295 | blk.24.attn_q.bias | 0xd8006e80 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xd800a680 | 0x6e4000 | +| 297 | blk.24.attn_v.bias | 0xd86ee680 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xd86eee80 | 0xfc000 | +| 299 | blk.24.ffn_down.weight | 0xd87eae80 | 0x2c84000 | +| 300 | blk.24.ffn_gate.weight | 0xdb46ee80 | 0x246c000 | +| 301 | blk.24.ffn_norm.weight | 0xdd8dae80 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xdd8de680 | 0x246c000 | +| 303 | blk.25.attn_k.bias | 0xdfd4a680 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xdfd4ae80 | 0xfc000 | +| 305 | blk.25.attn_norm.weight | 0xdfe46e80 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xdfe4a680 | 0x6e4000 | +| 307 | blk.25.attn_q.bias | 0xe052e680 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xe0531e80 | 0x6e4000 | +| 309 | blk.25.attn_v.bias | 0xe0c15e80 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xe0c16680 | 0xfc000 | +| 311 | blk.25.ffn_down.weight | 0xe0d12680 | 0x2c84000 | +| 312 | blk.25.ffn_gate.weight | 0xe3996680 | 0x246c000 | +| 313 | blk.25.ffn_norm.weight | 0xe5e02680 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xe5e05e80 | 0x246c000 | +| 315 | blk.26.attn_k.bias | 0xe8271e80 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xe8272680 | 0xfc000 | +| 317 | blk.26.attn_norm.weight | 0xe836e680 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xe8371e80 | 0x6e4000 | +| 319 | blk.26.attn_q.bias | 0xe8a55e80 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xe8a59680 | 0x6e4000 | +| 321 | blk.26.attn_v.bias | 0xe913d680 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xe913de80 | 0xfc000 | +| 323 | blk.26.ffn_down.weight | 0xe9239e80 | 0x2c84000 | +| 324 | blk.26.ffn_gate.weight | 0xebebde80 | 0x246c000 | +| 325 | blk.26.ffn_norm.weight | 0xee329e80 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xee32d680 | 0x246c000 | +| 327 | blk.27.attn_k.bias | 0xf0799680 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xf0799e80 | 0xfc000 | +| 329 | blk.27.attn_norm.weight | 0xf0895e80 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xf0899680 | 0x6e4000 | +| 331 | blk.27.attn_q.bias | 0xf0f7d680 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xf0f80e80 | 0x6e4000 | +| 333 | blk.27.attn_v.bias | 0xf1664e80 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xf1665680 | 0xfc000 | +| 335 | blk.27.ffn_down.weight | 0xf1761680 | 0x2c84000 | +| 336 | blk.27.ffn_gate.weight | 0xf43e5680 | 0x246c000 | +| 337 | blk.27.ffn_norm.weight | 0xf6851680 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xf6854e80 | 0x246c000 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-------| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | IQ4_NL | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | IQ3_S | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-------| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ3_S | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ3_S | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_XS | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ3_S | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-------| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | IQ4_NL | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | IQ4_NL | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | IQ4_NL | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q3_K_L.md b/scores/Hammer2.1-7b-Q3_K_L.md new file mode 100644 index 0000000..420af24 --- /dev/null +++ b/scores/Hammer2.1-7b-Q3_K_L.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q3_K_L.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 13 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q3\_K\_L.gguf - GGUF Internal File Dump](#hammer21-7b-q3_k_lgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0xdebe7c4 | +| 1 | output_norm.weight | 0xe469ca0 | 0x3800 | +| 2 | token_embd.weight | 0xe46d4a0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x1c32bc80 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x1c32c480 | 0x93000 | +| 5 | blk.0.attn_norm.weight | 0x1c3bf480 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x1c3c2c80 | 0x86c000 | +| 7 | blk.0.attn_q.bias | 0x1cc2ec80 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x1cc32480 | 0x405000 | +| 9 | blk.0.attn_v.bias | 0x1d037480 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x1d037c80 | 0xfc000 | +| 11 | blk.0.ffn_down.weight | 0x1d133c80 | 0x246c000 | +| 12 | blk.0.ffn_gate.weight | 0x1f59fc80 | 0x153f000 | +| 13 | blk.0.ffn_norm.weight | 0x20adec80 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x20ae2480 | 0x153f000 | +| 15 | blk.1.attn_k.bias | 0x22021480 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x22021c80 | 0x93000 | +| 17 | blk.1.attn_norm.weight | 0x220b4c80 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x220b8480 | 0x86c000 | +| 19 | blk.1.attn_q.bias | 0x22924480 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x22927c80 | 0x405000 | +| 21 | blk.1.attn_v.bias | 0x22d2cc80 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x22d2d480 | 0xfc000 | +| 23 | blk.1.ffn_down.weight | 0x22e29480 | 0x246c000 | +| 24 | blk.1.ffn_gate.weight | 0x25295480 | 0x1bd2800 | +| 25 | blk.1.ffn_norm.weight | 0x26e67c80 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x26e6b480 | 0x1bd2800 | +| 27 | blk.2.attn_k.bias | 0x28a3dc80 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x28a3e480 | 0x93000 | +| 29 | blk.2.attn_norm.weight | 0x28ad1480 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x28ad4c80 | 0x86c000 | +| 31 | blk.2.attn_q.bias | 0x29340c80 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x29344480 | 0x405000 | +| 33 | blk.2.attn_v.bias | 0x29749480 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x29749c80 | 0xfc000 | +| 35 | blk.2.ffn_down.weight | 0x29845c80 | 0x246c000 | +| 36 | blk.2.ffn_gate.weight | 0x2bcb1c80 | 0x1bd2800 | +| 37 | blk.2.ffn_norm.weight | 0x2d884480 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x2d887c80 | 0x1bd2800 | +| 39 | blk.3.attn_k.bias | 0x2f45a480 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x2f45ac80 | 0x93000 | +| 41 | blk.3.attn_norm.weight | 0x2f4edc80 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x2f4f1480 | 0x86c000 | +| 43 | blk.3.attn_q.bias | 0x2fd5d480 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x2fd60c80 | 0x405000 | +| 45 | blk.3.attn_v.bias | 0x30165c80 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x30166480 | 0xfc000 | +| 47 | blk.3.ffn_down.weight | 0x30262480 | 0x2c84000 | +| 48 | blk.3.ffn_gate.weight | 0x32ee6480 | 0x1bd2800 | +| 49 | blk.3.ffn_norm.weight | 0x34ab8c80 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x34abc480 | 0x1bd2800 | +| 51 | blk.4.attn_k.bias | 0x3668ec80 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x3668f480 | 0x93000 | +| 53 | blk.4.attn_norm.weight | 0x36722480 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x36725c80 | 0x86c000 | +| 55 | blk.4.attn_q.bias | 0x36f91c80 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x36f95480 | 0x405000 | +| 57 | blk.4.attn_v.bias | 0x3739a480 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x3739ac80 | 0xfc000 | +| 59 | blk.4.ffn_down.weight | 0x37496c80 | 0x246c000 | +| 60 | blk.4.ffn_gate.weight | 0x39902c80 | 0x1bd2800 | +| 61 | blk.4.ffn_norm.weight | 0x3b4d5480 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x3b4d8c80 | 0x1bd2800 | +| 63 | blk.5.attn_k.bias | 0x3d0ab480 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x3d0abc80 | 0x93000 | +| 65 | blk.5.attn_norm.weight | 0x3d13ec80 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x3d142480 | 0x86c000 | +| 67 | blk.5.attn_q.bias | 0x3d9ae480 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x3d9b1c80 | 0x405000 | +| 69 | blk.5.attn_v.bias | 0x3ddb6c80 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x3ddb7480 | 0xfc000 | +| 71 | blk.5.ffn_down.weight | 0x3deb3480 | 0x246c000 | +| 72 | blk.5.ffn_gate.weight | 0x4031f480 | 0x1bd2800 | +| 73 | blk.5.ffn_norm.weight | 0x41ef1c80 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x41ef5480 | 0x1bd2800 | +| 75 | blk.6.attn_k.bias | 0x43ac7c80 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x43ac8480 | 0x93000 | +| 77 | blk.6.attn_norm.weight | 0x43b5b480 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x43b5ec80 | 0x86c000 | +| 79 | blk.6.attn_q.bias | 0x443cac80 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x443ce480 | 0x405000 | +| 81 | blk.6.attn_v.bias | 0x447d3480 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x447d3c80 | 0xfc000 | +| 83 | blk.6.ffn_down.weight | 0x448cfc80 | 0x246c000 | +| 84 | blk.6.ffn_gate.weight | 0x46d3bc80 | 0x153f000 | +| 85 | blk.6.ffn_norm.weight | 0x4827ac80 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x4827e480 | 0x153f000 | +| 87 | blk.7.attn_k.bias | 0x497bd480 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x497bdc80 | 0x93000 | +| 89 | blk.7.attn_norm.weight | 0x49850c80 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x49854480 | 0x86c000 | +| 91 | blk.7.attn_q.bias | 0x4a0c0480 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x4a0c3c80 | 0x405000 | +| 93 | blk.7.attn_v.bias | 0x4a4c8c80 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x4a4c9480 | 0xfc000 | +| 95 | blk.7.ffn_down.weight | 0x4a5c5480 | 0x2c84000 | +| 96 | blk.7.ffn_gate.weight | 0x4d249480 | 0x153f000 | +| 97 | blk.7.ffn_norm.weight | 0x4e788480 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x4e78bc80 | 0x153f000 | +| 99 | blk.8.attn_k.bias | 0x4fccac80 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x4fccb480 | 0x93000 | +| 101 | blk.8.attn_norm.weight | 0x4fd5e480 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x4fd61c80 | 0x86c000 | +| 103 | blk.8.attn_q.bias | 0x505cdc80 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x505d1480 | 0x405000 | +| 105 | blk.8.attn_v.bias | 0x509d6480 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x509d6c80 | 0xfc000 | +| 107 | blk.8.ffn_down.weight | 0x50ad2c80 | 0x246c000 | +| 108 | blk.8.ffn_gate.weight | 0x52f3ec80 | 0x153f000 | +| 109 | blk.8.ffn_norm.weight | 0x5447dc80 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x54481480 | 0x153f000 | +| 111 | blk.9.attn_k.bias | 0x559c0480 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x559c0c80 | 0x93000 | +| 113 | blk.9.attn_norm.weight | 0x55a53c80 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x55a57480 | 0x86c000 | +| 115 | blk.9.attn_q.bias | 0x562c3480 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x562c6c80 | 0x405000 | +| 117 | blk.9.attn_v.bias | 0x566cbc80 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x566cc480 | 0xfc000 | +| 119 | blk.9.ffn_down.weight | 0x567c8480 | 0x2c84000 | +| 120 | blk.9.ffn_gate.weight | 0x5944c480 | 0x1bd2800 | +| 121 | blk.9.ffn_norm.weight | 0x5b01ec80 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x5b022480 | 0x1bd2800 | +| 123 | blk.10.attn_k.bias | 0x5cbf4c80 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x5cbf5480 | 0x93000 | +| 125 | blk.10.attn_norm.weight | 0x5cc88480 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x5cc8bc80 | 0x86c000 | +| 127 | blk.10.attn_q.bias | 0x5d4f7c80 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x5d4fb480 | 0x405000 | +| 129 | blk.10.attn_v.bias | 0x5d900480 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x5d900c80 | 0xfc000 | +| 131 | blk.10.ffn_down.weight | 0x5d9fcc80 | 0x2c84000 | +| 132 | blk.10.ffn_gate.weight | 0x60680c80 | 0x153f000 | +| 133 | blk.10.ffn_norm.weight | 0x61bbfc80 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x61bc3480 | 0x153f000 | +| 135 | blk.11.attn_k.bias | 0x63102480 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x63102c80 | 0x93000 | +| 137 | blk.11.attn_norm.weight | 0x63195c80 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x63199480 | 0x86c000 | +| 139 | blk.11.attn_q.bias | 0x63a05480 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x63a08c80 | 0x405000 | +| 141 | blk.11.attn_v.bias | 0x63e0dc80 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x63e0e480 | 0xfc000 | +| 143 | blk.11.ffn_down.weight | 0x63f0a480 | 0x246c000 | +| 144 | blk.11.ffn_gate.weight | 0x66376480 | 0x153f000 | +| 145 | blk.11.ffn_norm.weight | 0x678b5480 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x678b8c80 | 0x153f000 | +| 147 | blk.12.attn_k.bias | 0x68df7c80 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x68df8480 | 0x93000 | +| 149 | blk.12.attn_norm.weight | 0x68e8b480 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x68e8ec80 | 0x86c000 | +| 151 | blk.12.attn_q.bias | 0x696fac80 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x696fe480 | 0x405000 | +| 153 | blk.12.attn_v.bias | 0x69b03480 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x69b03c80 | 0xfc000 | +| 155 | blk.12.ffn_down.weight | 0x69bffc80 | 0x246c000 | +| 156 | blk.12.ffn_gate.weight | 0x6c06bc80 | 0x153f000 | +| 157 | blk.12.ffn_norm.weight | 0x6d5aac80 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x6d5ae480 | 0x153f000 | +| 159 | blk.13.attn_k.bias | 0x6eaed480 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x6eaedc80 | 0x93000 | +| 161 | blk.13.attn_norm.weight | 0x6eb80c80 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x6eb84480 | 0x86c000 | +| 163 | blk.13.attn_q.bias | 0x6f3f0480 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x6f3f3c80 | 0x405000 | +| 165 | blk.13.attn_v.bias | 0x6f7f8c80 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x6f7f9480 | 0xfc000 | +| 167 | blk.13.ffn_down.weight | 0x6f8f5480 | 0x246c000 | +| 168 | blk.13.ffn_gate.weight | 0x71d61480 | 0x153f000 | +| 169 | blk.13.ffn_norm.weight | 0x732a0480 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x732a3c80 | 0x153f000 | +| 171 | blk.14.attn_k.bias | 0x747e2c80 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x747e3480 | 0xc0800 | +| 173 | blk.14.attn_norm.weight | 0x748a3c80 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x748a7480 | 0x86c000 | +| 175 | blk.14.attn_q.bias | 0x75113480 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x75116c80 | 0x543800 | +| 177 | blk.14.attn_v.bias | 0x7565a480 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x7565ac80 | 0xfc000 | +| 179 | blk.14.ffn_down.weight | 0x75756c80 | 0x246c000 | +| 180 | blk.14.ffn_gate.weight | 0x77bc2c80 | 0x153f000 | +| 181 | blk.14.ffn_norm.weight | 0x79101c80 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x79105480 | 0x153f000 | +| 183 | blk.15.attn_k.bias | 0x7a644480 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x7a644c80 | 0xc0800 | +| 185 | blk.15.attn_norm.weight | 0x7a705480 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x7a708c80 | 0x86c000 | +| 187 | blk.15.attn_q.bias | 0x7af74c80 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x7af78480 | 0x543800 | +| 189 | blk.15.attn_v.bias | 0x7b4bbc80 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x7b4bc480 | 0xfc000 | +| 191 | blk.15.ffn_down.weight | 0x7b5b8480 | 0x246c000 | +| 192 | blk.15.ffn_gate.weight | 0x7da24480 | 0x153f000 | +| 193 | blk.15.ffn_norm.weight | 0x7ef63480 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x7ef66c80 | 0x153f000 | +| 195 | blk.16.attn_k.bias | 0x804a5c80 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x804a6480 | 0xc0800 | +| 197 | blk.16.attn_norm.weight | 0x80566c80 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x8056a480 | 0x86c000 | +| 199 | blk.16.attn_q.bias | 0x80dd6480 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x80dd9c80 | 0x543800 | +| 201 | blk.16.attn_v.bias | 0x8131d480 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x8131dc80 | 0xfc000 | +| 203 | blk.16.ffn_down.weight | 0x81419c80 | 0x246c000 | +| 204 | blk.16.ffn_gate.weight | 0x83885c80 | 0x153f000 | +| 205 | blk.16.ffn_norm.weight | 0x84dc4c80 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x84dc8480 | 0x153f000 | +| 207 | blk.17.attn_k.bias | 0x86307480 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x86307c80 | 0xc0800 | +| 209 | blk.17.attn_norm.weight | 0x863c8480 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x863cbc80 | 0x86c000 | +| 211 | blk.17.attn_q.bias | 0x86c37c80 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x86c3b480 | 0x543800 | +| 213 | blk.17.attn_v.bias | 0x8717ec80 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x8717f480 | 0xfc000 | +| 215 | blk.17.ffn_down.weight | 0x8727b480 | 0x246c000 | +| 216 | blk.17.ffn_gate.weight | 0x896e7480 | 0x153f000 | +| 217 | blk.17.ffn_norm.weight | 0x8ac26480 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x8ac29c80 | 0x153f000 | +| 219 | blk.18.attn_k.bias | 0x8c168c80 | 0x800 | +| 220 | blk.18.attn_k.weight | 0x8c169480 | 0xc0800 | +| 221 | blk.18.attn_norm.weight | 0x8c229c80 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0x8c22d480 | 0x86c000 | +| 223 | blk.18.attn_q.bias | 0x8ca99480 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0x8ca9cc80 | 0x543800 | +| 225 | blk.18.attn_v.bias | 0x8cfe0480 | 0x800 | +| 226 | blk.18.attn_v.weight | 0x8cfe0c80 | 0xfc000 | +| 227 | blk.18.ffn_down.weight | 0x8d0dcc80 | 0x2c84000 | +| 228 | blk.18.ffn_gate.weight | 0x8fd60c80 | 0x153f000 | +| 229 | blk.18.ffn_norm.weight | 0x9129fc80 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x912a3480 | 0x153f000 | +| 231 | blk.19.attn_k.bias | 0x927e2480 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x927e2c80 | 0xc0800 | +| 233 | blk.19.attn_norm.weight | 0x928a3480 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x928a6c80 | 0x86c000 | +| 235 | blk.19.attn_q.bias | 0x93112c80 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x93116480 | 0x543800 | +| 237 | blk.19.attn_v.bias | 0x93659c80 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x9365a480 | 0xfc000 | +| 239 | blk.19.ffn_down.weight | 0x93756480 | 0x2c84000 | +| 240 | blk.19.ffn_gate.weight | 0x963da480 | 0x153f000 | +| 241 | blk.19.ffn_norm.weight | 0x97919480 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x9791cc80 | 0x153f000 | +| 243 | blk.20.attn_k.bias | 0x98e5bc80 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x98e5c480 | 0xc0800 | +| 245 | blk.20.attn_norm.weight | 0x98f1cc80 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x98f20480 | 0x86c000 | +| 247 | blk.20.attn_q.bias | 0x9978c480 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x9978fc80 | 0x543800 | +| 249 | blk.20.attn_v.bias | 0x99cd3480 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x99cd3c80 | 0xfc000 | +| 251 | blk.20.ffn_down.weight | 0x99dcfc80 | 0x2c84000 | +| 252 | blk.20.ffn_gate.weight | 0x9ca53c80 | 0x1bd2800 | +| 253 | blk.20.ffn_norm.weight | 0x9e626480 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x9e629c80 | 0x1bd2800 | +| 255 | blk.21.attn_k.bias | 0xa01fc480 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xa01fcc80 | 0xc0800 | +| 257 | blk.21.attn_norm.weight | 0xa02bd480 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xa02c0c80 | 0x86c000 | +| 259 | blk.21.attn_q.bias | 0xa0b2cc80 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xa0b30480 | 0x543800 | +| 261 | blk.21.attn_v.bias | 0xa1073c80 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xa1074480 | 0xfc000 | +| 263 | blk.21.ffn_down.weight | 0xa1170480 | 0x2c84000 | +| 264 | blk.21.ffn_gate.weight | 0xa3df4480 | 0x1bd2800 | +| 265 | blk.21.ffn_norm.weight | 0xa59c6c80 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xa59ca480 | 0x1bd2800 | +| 267 | blk.22.attn_k.bias | 0xa759cc80 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xa759d480 | 0xc0800 | +| 269 | blk.22.attn_norm.weight | 0xa765dc80 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xa7661480 | 0x86c000 | +| 271 | blk.22.attn_q.bias | 0xa7ecd480 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xa7ed0c80 | 0x543800 | +| 273 | blk.22.attn_v.bias | 0xa8414480 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xa8414c80 | 0xfc000 | +| 275 | blk.22.ffn_down.weight | 0xa8510c80 | 0x2c84000 | +| 276 | blk.22.ffn_gate.weight | 0xab194c80 | 0x1bd2800 | +| 277 | blk.22.ffn_norm.weight | 0xacd67480 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xacd6ac80 | 0x1bd2800 | +| 279 | blk.23.attn_k.bias | 0xae93d480 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xae93dc80 | 0xc0800 | +| 281 | blk.23.attn_norm.weight | 0xae9fe480 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xaea01c80 | 0x86c000 | +| 283 | blk.23.attn_q.bias | 0xaf26dc80 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xaf271480 | 0x543800 | +| 285 | blk.23.attn_v.bias | 0xaf7b4c80 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xaf7b5480 | 0xfc000 | +| 287 | blk.23.ffn_down.weight | 0xaf8b1480 | 0x2c84000 | +| 288 | blk.23.ffn_gate.weight | 0xb2535480 | 0x1bd2800 | +| 289 | blk.23.ffn_norm.weight | 0xb4107c80 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xb410b480 | 0x1bd2800 | +| 291 | blk.24.attn_k.bias | 0xb5cddc80 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xb5cde480 | 0xc0800 | +| 293 | blk.24.attn_norm.weight | 0xb5d9ec80 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xb5da2480 | 0x86c000 | +| 295 | blk.24.attn_q.bias | 0xb660e480 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xb6611c80 | 0x543800 | +| 297 | blk.24.attn_v.bias | 0xb6b55480 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xb6b55c80 | 0xfc000 | +| 299 | blk.24.ffn_down.weight | 0xb6c51c80 | 0x2c84000 | +| 300 | blk.24.ffn_gate.weight | 0xb98d5c80 | 0x1bd2800 | +| 301 | blk.24.ffn_norm.weight | 0xbb4a8480 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xbb4abc80 | 0x1bd2800 | +| 303 | blk.25.attn_k.bias | 0xbd07e480 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xbd07ec80 | 0xc0800 | +| 305 | blk.25.attn_norm.weight | 0xbd13f480 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xbd142c80 | 0x86c000 | +| 307 | blk.25.attn_q.bias | 0xbd9aec80 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xbd9b2480 | 0x543800 | +| 309 | blk.25.attn_v.bias | 0xbdef5c80 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xbdef6480 | 0xfc000 | +| 311 | blk.25.ffn_down.weight | 0xbdff2480 | 0x2c84000 | +| 312 | blk.25.ffn_gate.weight | 0xc0c76480 | 0x1bd2800 | +| 313 | blk.25.ffn_norm.weight | 0xc2848c80 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xc284c480 | 0x1bd2800 | +| 315 | blk.26.attn_k.bias | 0xc441ec80 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xc441f480 | 0xc0800 | +| 317 | blk.26.attn_norm.weight | 0xc44dfc80 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xc44e3480 | 0x86c000 | +| 319 | blk.26.attn_q.bias | 0xc4d4f480 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xc4d52c80 | 0x543800 | +| 321 | blk.26.attn_v.bias | 0xc5296480 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xc5296c80 | 0xfc000 | +| 323 | blk.26.ffn_down.weight | 0xc5392c80 | 0x2c84000 | +| 324 | blk.26.ffn_gate.weight | 0xc8016c80 | 0x1bd2800 | +| 325 | blk.26.ffn_norm.weight | 0xc9be9480 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xc9becc80 | 0x1bd2800 | +| 327 | blk.27.attn_k.bias | 0xcb7bf480 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xcb7bfc80 | 0xc0800 | +| 329 | blk.27.attn_norm.weight | 0xcb880480 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xcb883c80 | 0x86c000 | +| 331 | blk.27.attn_q.bias | 0xcc0efc80 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xcc0f3480 | 0x543800 | +| 333 | blk.27.attn_v.bias | 0xcc636c80 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xcc637480 | 0xfc000 | +| 335 | blk.27.ffn_down.weight | 0xcc733480 | 0x2c84000 | +| 336 | blk.27.ffn_gate.weight | 0xcf3b7480 | 0x1bd2800 | +| 337 | blk.27.ffn_norm.weight | 0xd0f89c80 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xd0f8d480 | 0x1bd2800 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q3_K_M.md b/scores/Hammer2.1-7b-Q3_K_M.md new file mode 100644 index 0000000..60282d3 --- /dev/null +++ b/scores/Hammer2.1-7b-Q3_K_M.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q3_K_M.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 12 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q3\_K\_M.gguf - GGUF Internal File Dump](#hammer21-7b-q3_k_mgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0xdebe7c4 | +| 1 | output_norm.weight | 0xe469ca0 | 0x3800 | +| 2 | token_embd.weight | 0xe46d4a0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x1c32bc80 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x1c32c480 | 0x93000 | +| 5 | blk.0.attn_norm.weight | 0x1c3bf480 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x1c3c2c80 | 0x6e4000 | +| 7 | blk.0.attn_q.bias | 0x1caa6c80 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x1caaa480 | 0x405000 | +| 9 | blk.0.attn_v.bias | 0x1ceaf480 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x1ceafc80 | 0xc0800 | +| 11 | blk.0.ffn_down.weight | 0x1cf70480 | 0x246c000 | +| 12 | blk.0.ffn_gate.weight | 0x1f3dc480 | 0x153f000 | +| 13 | blk.0.ffn_norm.weight | 0x2091b480 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x2091ec80 | 0x153f000 | +| 15 | blk.1.attn_k.bias | 0x21e5dc80 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x21e5e480 | 0x93000 | +| 17 | blk.1.attn_norm.weight | 0x21ef1480 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x21ef4c80 | 0x6e4000 | +| 19 | blk.1.attn_q.bias | 0x225d8c80 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x225dc480 | 0x405000 | +| 21 | blk.1.attn_v.bias | 0x229e1480 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x229e1c80 | 0xc0800 | +| 23 | blk.1.ffn_down.weight | 0x22aa2480 | 0x246c000 | +| 24 | blk.1.ffn_gate.weight | 0x24f0e480 | 0x1bd2800 | +| 25 | blk.1.ffn_norm.weight | 0x26ae0c80 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x26ae4480 | 0x1bd2800 | +| 27 | blk.2.attn_k.bias | 0x286b6c80 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x286b7480 | 0x93000 | +| 29 | blk.2.attn_norm.weight | 0x2874a480 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x2874dc80 | 0x6e4000 | +| 31 | blk.2.attn_q.bias | 0x28e31c80 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x28e35480 | 0x405000 | +| 33 | blk.2.attn_v.bias | 0x2923a480 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x2923ac80 | 0xc0800 | +| 35 | blk.2.ffn_down.weight | 0x292fb480 | 0x246c000 | +| 36 | blk.2.ffn_gate.weight | 0x2b767480 | 0x1bd2800 | +| 37 | blk.2.ffn_norm.weight | 0x2d339c80 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x2d33d480 | 0x1bd2800 | +| 39 | blk.3.attn_k.bias | 0x2ef0fc80 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x2ef10480 | 0x93000 | +| 41 | blk.3.attn_norm.weight | 0x2efa3480 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x2efa6c80 | 0x6e4000 | +| 43 | blk.3.attn_q.bias | 0x2f68ac80 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x2f68e480 | 0x405000 | +| 45 | blk.3.attn_v.bias | 0x2fa93480 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x2fa93c80 | 0xc0800 | +| 47 | blk.3.ffn_down.weight | 0x2fb54480 | 0x246c000 | +| 48 | blk.3.ffn_gate.weight | 0x31fc0480 | 0x1bd2800 | +| 49 | blk.3.ffn_norm.weight | 0x33b92c80 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x33b96480 | 0x1bd2800 | +| 51 | blk.4.attn_k.bias | 0x35768c80 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x35769480 | 0x93000 | +| 53 | blk.4.attn_norm.weight | 0x357fc480 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x357ffc80 | 0x6e4000 | +| 55 | blk.4.attn_q.bias | 0x35ee3c80 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x35ee7480 | 0x405000 | +| 57 | blk.4.attn_v.bias | 0x362ec480 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x362ecc80 | 0xc0800 | +| 59 | blk.4.ffn_down.weight | 0x363ad480 | 0x246c000 | +| 60 | blk.4.ffn_gate.weight | 0x38819480 | 0x1bd2800 | +| 61 | blk.4.ffn_norm.weight | 0x3a3ebc80 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x3a3ef480 | 0x1bd2800 | +| 63 | blk.5.attn_k.bias | 0x3bfc1c80 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x3bfc2480 | 0x93000 | +| 65 | blk.5.attn_norm.weight | 0x3c055480 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x3c058c80 | 0x6e4000 | +| 67 | blk.5.attn_q.bias | 0x3c73cc80 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x3c740480 | 0x405000 | +| 69 | blk.5.attn_v.bias | 0x3cb45480 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x3cb45c80 | 0xc0800 | +| 71 | blk.5.ffn_down.weight | 0x3cc06480 | 0x246c000 | +| 72 | blk.5.ffn_gate.weight | 0x3f072480 | 0x1bd2800 | +| 73 | blk.5.ffn_norm.weight | 0x40c44c80 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x40c48480 | 0x1bd2800 | +| 75 | blk.6.attn_k.bias | 0x4281ac80 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x4281b480 | 0x93000 | +| 77 | blk.6.attn_norm.weight | 0x428ae480 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x428b1c80 | 0x6e4000 | +| 79 | blk.6.attn_q.bias | 0x42f95c80 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x42f99480 | 0x405000 | +| 81 | blk.6.attn_v.bias | 0x4339e480 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x4339ec80 | 0xc0800 | +| 83 | blk.6.ffn_down.weight | 0x4345f480 | 0x246c000 | +| 84 | blk.6.ffn_gate.weight | 0x458cb480 | 0x153f000 | +| 85 | blk.6.ffn_norm.weight | 0x46e0a480 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x46e0dc80 | 0x153f000 | +| 87 | blk.7.attn_k.bias | 0x4834cc80 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x4834d480 | 0x93000 | +| 89 | blk.7.attn_norm.weight | 0x483e0480 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x483e3c80 | 0x6e4000 | +| 91 | blk.7.attn_q.bias | 0x48ac7c80 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x48acb480 | 0x405000 | +| 93 | blk.7.attn_v.bias | 0x48ed0480 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x48ed0c80 | 0xc0800 | +| 95 | blk.7.ffn_down.weight | 0x48f91480 | 0x246c000 | +| 96 | blk.7.ffn_gate.weight | 0x4b3fd480 | 0x153f000 | +| 97 | blk.7.ffn_norm.weight | 0x4c93c480 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x4c93fc80 | 0x153f000 | +| 99 | blk.8.attn_k.bias | 0x4de7ec80 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x4de7f480 | 0x93000 | +| 101 | blk.8.attn_norm.weight | 0x4df12480 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x4df15c80 | 0x6e4000 | +| 103 | blk.8.attn_q.bias | 0x4e5f9c80 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x4e5fd480 | 0x405000 | +| 105 | blk.8.attn_v.bias | 0x4ea02480 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x4ea02c80 | 0xc0800 | +| 107 | blk.8.ffn_down.weight | 0x4eac3480 | 0x246c000 | +| 108 | blk.8.ffn_gate.weight | 0x50f2f480 | 0x153f000 | +| 109 | blk.8.ffn_norm.weight | 0x5246e480 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x52471c80 | 0x153f000 | +| 111 | blk.9.attn_k.bias | 0x539b0c80 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x539b1480 | 0x93000 | +| 113 | blk.9.attn_norm.weight | 0x53a44480 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x53a47c80 | 0x6e4000 | +| 115 | blk.9.attn_q.bias | 0x5412bc80 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x5412f480 | 0x405000 | +| 117 | blk.9.attn_v.bias | 0x54534480 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x54534c80 | 0xc0800 | +| 119 | blk.9.ffn_down.weight | 0x545f5480 | 0x246c000 | +| 120 | blk.9.ffn_gate.weight | 0x56a61480 | 0x1bd2800 | +| 121 | blk.9.ffn_norm.weight | 0x58633c80 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x58637480 | 0x1bd2800 | +| 123 | blk.10.attn_k.bias | 0x5a209c80 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x5a20a480 | 0x93000 | +| 125 | blk.10.attn_norm.weight | 0x5a29d480 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x5a2a0c80 | 0x6e4000 | +| 127 | blk.10.attn_q.bias | 0x5a984c80 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x5a988480 | 0x405000 | +| 129 | blk.10.attn_v.bias | 0x5ad8d480 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x5ad8dc80 | 0xc0800 | +| 131 | blk.10.ffn_down.weight | 0x5ae4e480 | 0x246c000 | +| 132 | blk.10.ffn_gate.weight | 0x5d2ba480 | 0x153f000 | +| 133 | blk.10.ffn_norm.weight | 0x5e7f9480 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x5e7fcc80 | 0x153f000 | +| 135 | blk.11.attn_k.bias | 0x5fd3bc80 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x5fd3c480 | 0x93000 | +| 137 | blk.11.attn_norm.weight | 0x5fdcf480 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x5fdd2c80 | 0x6e4000 | +| 139 | blk.11.attn_q.bias | 0x604b6c80 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x604ba480 | 0x405000 | +| 141 | blk.11.attn_v.bias | 0x608bf480 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x608bfc80 | 0xc0800 | +| 143 | blk.11.ffn_down.weight | 0x60980480 | 0x246c000 | +| 144 | blk.11.ffn_gate.weight | 0x62dec480 | 0x153f000 | +| 145 | blk.11.ffn_norm.weight | 0x6432b480 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x6432ec80 | 0x153f000 | +| 147 | blk.12.attn_k.bias | 0x6586dc80 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x6586e480 | 0x93000 | +| 149 | blk.12.attn_norm.weight | 0x65901480 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x65904c80 | 0x6e4000 | +| 151 | blk.12.attn_q.bias | 0x65fe8c80 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x65fec480 | 0x405000 | +| 153 | blk.12.attn_v.bias | 0x663f1480 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x663f1c80 | 0xc0800 | +| 155 | blk.12.ffn_down.weight | 0x664b2480 | 0x246c000 | +| 156 | blk.12.ffn_gate.weight | 0x6891e480 | 0x153f000 | +| 157 | blk.12.ffn_norm.weight | 0x69e5d480 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x69e60c80 | 0x153f000 | +| 159 | blk.13.attn_k.bias | 0x6b39fc80 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x6b3a0480 | 0x93000 | +| 161 | blk.13.attn_norm.weight | 0x6b433480 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x6b436c80 | 0x6e4000 | +| 163 | blk.13.attn_q.bias | 0x6bb1ac80 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x6bb1e480 | 0x405000 | +| 165 | blk.13.attn_v.bias | 0x6bf23480 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x6bf23c80 | 0xc0800 | +| 167 | blk.13.ffn_down.weight | 0x6bfe4480 | 0x246c000 | +| 168 | blk.13.ffn_gate.weight | 0x6e450480 | 0x153f000 | +| 169 | blk.13.ffn_norm.weight | 0x6f98f480 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x6f992c80 | 0x153f000 | +| 171 | blk.14.attn_k.bias | 0x70ed1c80 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x70ed2480 | 0xc0800 | +| 173 | blk.14.attn_norm.weight | 0x70f92c80 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x70f96480 | 0x6e4000 | +| 175 | blk.14.attn_q.bias | 0x7167a480 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x7167dc80 | 0x543800 | +| 177 | blk.14.attn_v.bias | 0x71bc1480 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x71bc1c80 | 0xfc000 | +| 179 | blk.14.ffn_down.weight | 0x71cbdc80 | 0x246c000 | +| 180 | blk.14.ffn_gate.weight | 0x74129c80 | 0x153f000 | +| 181 | blk.14.ffn_norm.weight | 0x75668c80 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x7566c480 | 0x153f000 | +| 183 | blk.15.attn_k.bias | 0x76bab480 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x76babc80 | 0xc0800 | +| 185 | blk.15.attn_norm.weight | 0x76c6c480 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x76c6fc80 | 0x6e4000 | +| 187 | blk.15.attn_q.bias | 0x77353c80 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x77357480 | 0x543800 | +| 189 | blk.15.attn_v.bias | 0x7789ac80 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x7789b480 | 0xfc000 | +| 191 | blk.15.ffn_down.weight | 0x77997480 | 0x246c000 | +| 192 | blk.15.ffn_gate.weight | 0x79e03480 | 0x153f000 | +| 193 | blk.15.ffn_norm.weight | 0x7b342480 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x7b345c80 | 0x153f000 | +| 195 | blk.16.attn_k.bias | 0x7c884c80 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x7c885480 | 0xc0800 | +| 197 | blk.16.attn_norm.weight | 0x7c945c80 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x7c949480 | 0x6e4000 | +| 199 | blk.16.attn_q.bias | 0x7d02d480 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x7d030c80 | 0x543800 | +| 201 | blk.16.attn_v.bias | 0x7d574480 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x7d574c80 | 0xfc000 | +| 203 | blk.16.ffn_down.weight | 0x7d670c80 | 0x246c000 | +| 204 | blk.16.ffn_gate.weight | 0x7fadcc80 | 0x153f000 | +| 205 | blk.16.ffn_norm.weight | 0x8101bc80 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x8101f480 | 0x153f000 | +| 207 | blk.17.attn_k.bias | 0x8255e480 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x8255ec80 | 0xc0800 | +| 209 | blk.17.attn_norm.weight | 0x8261f480 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x82622c80 | 0x6e4000 | +| 211 | blk.17.attn_q.bias | 0x82d06c80 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x82d0a480 | 0x543800 | +| 213 | blk.17.attn_v.bias | 0x8324dc80 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x8324e480 | 0xfc000 | +| 215 | blk.17.ffn_down.weight | 0x8334a480 | 0x246c000 | +| 216 | blk.17.ffn_gate.weight | 0x857b6480 | 0x153f000 | +| 217 | blk.17.ffn_norm.weight | 0x86cf5480 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x86cf8c80 | 0x153f000 | +| 219 | blk.18.attn_k.bias | 0x88237c80 | 0x800 | +| 220 | blk.18.attn_k.weight | 0x88238480 | 0xc0800 | +| 221 | blk.18.attn_norm.weight | 0x882f8c80 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0x882fc480 | 0x6e4000 | +| 223 | blk.18.attn_q.bias | 0x889e0480 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0x889e3c80 | 0x543800 | +| 225 | blk.18.attn_v.bias | 0x88f27480 | 0x800 | +| 226 | blk.18.attn_v.weight | 0x88f27c80 | 0xfc000 | +| 227 | blk.18.ffn_down.weight | 0x89023c80 | 0x246c000 | +| 228 | blk.18.ffn_gate.weight | 0x8b48fc80 | 0x153f000 | +| 229 | blk.18.ffn_norm.weight | 0x8c9cec80 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x8c9d2480 | 0x153f000 | +| 231 | blk.19.attn_k.bias | 0x8df11480 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x8df11c80 | 0xc0800 | +| 233 | blk.19.attn_norm.weight | 0x8dfd2480 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x8dfd5c80 | 0x6e4000 | +| 235 | blk.19.attn_q.bias | 0x8e6b9c80 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x8e6bd480 | 0x543800 | +| 237 | blk.19.attn_v.bias | 0x8ec00c80 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x8ec01480 | 0xfc000 | +| 239 | blk.19.ffn_down.weight | 0x8ecfd480 | 0x246c000 | +| 240 | blk.19.ffn_gate.weight | 0x91169480 | 0x153f000 | +| 241 | blk.19.ffn_norm.weight | 0x926a8480 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x926abc80 | 0x153f000 | +| 243 | blk.20.attn_k.bias | 0x93beac80 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x93beb480 | 0xc0800 | +| 245 | blk.20.attn_norm.weight | 0x93cabc80 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x93caf480 | 0x6e4000 | +| 247 | blk.20.attn_q.bias | 0x94393480 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x94396c80 | 0x543800 | +| 249 | blk.20.attn_v.bias | 0x948da480 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x948dac80 | 0xfc000 | +| 251 | blk.20.ffn_down.weight | 0x949d6c80 | 0x246c000 | +| 252 | blk.20.ffn_gate.weight | 0x96e42c80 | 0x1bd2800 | +| 253 | blk.20.ffn_norm.weight | 0x98a15480 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x98a18c80 | 0x1bd2800 | +| 255 | blk.21.attn_k.bias | 0x9a5eb480 | 0x800 | +| 256 | blk.21.attn_k.weight | 0x9a5ebc80 | 0xc0800 | +| 257 | blk.21.attn_norm.weight | 0x9a6ac480 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0x9a6afc80 | 0x6e4000 | +| 259 | blk.21.attn_q.bias | 0x9ad93c80 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0x9ad97480 | 0x543800 | +| 261 | blk.21.attn_v.bias | 0x9b2dac80 | 0x800 | +| 262 | blk.21.attn_v.weight | 0x9b2db480 | 0xfc000 | +| 263 | blk.21.ffn_down.weight | 0x9b3d7480 | 0x246c000 | +| 264 | blk.21.ffn_gate.weight | 0x9d843480 | 0x1bd2800 | +| 265 | blk.21.ffn_norm.weight | 0x9f415c80 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0x9f419480 | 0x1bd2800 | +| 267 | blk.22.attn_k.bias | 0xa0febc80 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xa0fec480 | 0xc0800 | +| 269 | blk.22.attn_norm.weight | 0xa10acc80 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xa10b0480 | 0x6e4000 | +| 271 | blk.22.attn_q.bias | 0xa1794480 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xa1797c80 | 0x543800 | +| 273 | blk.22.attn_v.bias | 0xa1cdb480 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xa1cdbc80 | 0xfc000 | +| 275 | blk.22.ffn_down.weight | 0xa1dd7c80 | 0x246c000 | +| 276 | blk.22.ffn_gate.weight | 0xa4243c80 | 0x1bd2800 | +| 277 | blk.22.ffn_norm.weight | 0xa5e16480 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xa5e19c80 | 0x1bd2800 | +| 279 | blk.23.attn_k.bias | 0xa79ec480 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xa79ecc80 | 0xc0800 | +| 281 | blk.23.attn_norm.weight | 0xa7aad480 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xa7ab0c80 | 0x6e4000 | +| 283 | blk.23.attn_q.bias | 0xa8194c80 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xa8198480 | 0x543800 | +| 285 | blk.23.attn_v.bias | 0xa86dbc80 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xa86dc480 | 0xfc000 | +| 287 | blk.23.ffn_down.weight | 0xa87d8480 | 0x246c000 | +| 288 | blk.23.ffn_gate.weight | 0xaac44480 | 0x1bd2800 | +| 289 | blk.23.ffn_norm.weight | 0xac816c80 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xac81a480 | 0x1bd2800 | +| 291 | blk.24.attn_k.bias | 0xae3ecc80 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xae3ed480 | 0xc0800 | +| 293 | blk.24.attn_norm.weight | 0xae4adc80 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xae4b1480 | 0x6e4000 | +| 295 | blk.24.attn_q.bias | 0xaeb95480 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xaeb98c80 | 0x543800 | +| 297 | blk.24.attn_v.bias | 0xaf0dc480 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xaf0dcc80 | 0xfc000 | +| 299 | blk.24.ffn_down.weight | 0xaf1d8c80 | 0x246c000 | +| 300 | blk.24.ffn_gate.weight | 0xb1644c80 | 0x1bd2800 | +| 301 | blk.24.ffn_norm.weight | 0xb3217480 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xb321ac80 | 0x1bd2800 | +| 303 | blk.25.attn_k.bias | 0xb4ded480 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xb4dedc80 | 0xc0800 | +| 305 | blk.25.attn_norm.weight | 0xb4eae480 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xb4eb1c80 | 0x6e4000 | +| 307 | blk.25.attn_q.bias | 0xb5595c80 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xb5599480 | 0x543800 | +| 309 | blk.25.attn_v.bias | 0xb5adcc80 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xb5add480 | 0xfc000 | +| 311 | blk.25.ffn_down.weight | 0xb5bd9480 | 0x246c000 | +| 312 | blk.25.ffn_gate.weight | 0xb8045480 | 0x1bd2800 | +| 313 | blk.25.ffn_norm.weight | 0xb9c17c80 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xb9c1b480 | 0x1bd2800 | +| 315 | blk.26.attn_k.bias | 0xbb7edc80 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xbb7ee480 | 0xc0800 | +| 317 | blk.26.attn_norm.weight | 0xbb8aec80 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xbb8b2480 | 0x6e4000 | +| 319 | blk.26.attn_q.bias | 0xbbf96480 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xbbf99c80 | 0x543800 | +| 321 | blk.26.attn_v.bias | 0xbc4dd480 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xbc4ddc80 | 0xfc000 | +| 323 | blk.26.ffn_down.weight | 0xbc5d9c80 | 0x246c000 | +| 324 | blk.26.ffn_gate.weight | 0xbea45c80 | 0x1bd2800 | +| 325 | blk.26.ffn_norm.weight | 0xc0618480 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xc061bc80 | 0x1bd2800 | +| 327 | blk.27.attn_k.bias | 0xc21ee480 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xc21eec80 | 0xc0800 | +| 329 | blk.27.attn_norm.weight | 0xc22af480 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xc22b2c80 | 0x6e4000 | +| 331 | blk.27.attn_q.bias | 0xc2996c80 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xc299a480 | 0x543800 | +| 333 | blk.27.attn_v.bias | 0xc2eddc80 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xc2ede480 | 0xfc000 | +| 335 | blk.27.ffn_down.weight | 0xc2fda480 | 0x246c000 | +| 336 | blk.27.ffn_gate.weight | 0xc5446480 | 0x1bd2800 | +| 337 | blk.27.ffn_norm.weight | 0xc7018c80 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xc701c480 | 0x1bd2800 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q3_K_S.md b/scores/Hammer2.1-7b-Q3_K_S.md new file mode 100644 index 0000000..714dafa --- /dev/null +++ b/scores/Hammer2.1-7b-Q3_K_S.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q3_K_S.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 11 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q3\_K\_S.gguf - GGUF Internal File Dump](#hammer21-7b-q3_k_sgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0xdebe7c4 | +| 1 | output_norm.weight | 0xe469ca0 | 0x3800 | +| 2 | token_embd.weight | 0xe46d4a0 | 0xaa18718 | +| 3 | blk.0.attn_k.bias | 0x18e85bc0 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x18e863c0 | 0x93000 | +| 5 | blk.0.attn_norm.weight | 0x18f193c0 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x18f1cbc0 | 0x543800 | +| 7 | blk.0.attn_q.bias | 0x194603c0 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x19463bc0 | 0x405000 | +| 9 | blk.0.attn_v.bias | 0x19868bc0 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x198693c0 | 0xc0800 | +| 11 | blk.0.ffn_down.weight | 0x19929bc0 | 0x1bd2800 | +| 12 | blk.0.ffn_gate.weight | 0x1b4fc3c0 | 0x153f000 | +| 13 | blk.0.ffn_norm.weight | 0x1ca3b3c0 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x1ca3ebc0 | 0x153f000 | +| 15 | blk.1.attn_k.bias | 0x1df7dbc0 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x1df7e3c0 | 0x93000 | +| 17 | blk.1.attn_norm.weight | 0x1e0113c0 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x1e014bc0 | 0x543800 | +| 19 | blk.1.attn_q.bias | 0x1e5583c0 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x1e55bbc0 | 0x405000 | +| 21 | blk.1.attn_v.bias | 0x1e960bc0 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x1e9613c0 | 0xc0800 | +| 23 | blk.1.ffn_down.weight | 0x1ea21bc0 | 0x1bd2800 | +| 24 | blk.1.ffn_gate.weight | 0x205f43c0 | 0x1bd2800 | +| 25 | blk.1.ffn_norm.weight | 0x221c6bc0 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x221ca3c0 | 0x1bd2800 | +| 27 | blk.2.attn_k.bias | 0x23d9cbc0 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x23d9d3c0 | 0x93000 | +| 29 | blk.2.attn_norm.weight | 0x23e303c0 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x23e33bc0 | 0x543800 | +| 31 | blk.2.attn_q.bias | 0x243773c0 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x2437abc0 | 0x405000 | +| 33 | blk.2.attn_v.bias | 0x2477fbc0 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x247803c0 | 0xc0800 | +| 35 | blk.2.ffn_down.weight | 0x24840bc0 | 0x1bd2800 | +| 36 | blk.2.ffn_gate.weight | 0x264133c0 | 0x1bd2800 | +| 37 | blk.2.ffn_norm.weight | 0x27fe5bc0 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x27fe93c0 | 0x1bd2800 | +| 39 | blk.3.attn_k.bias | 0x29bbbbc0 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x29bbc3c0 | 0x93000 | +| 41 | blk.3.attn_norm.weight | 0x29c4f3c0 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x29c52bc0 | 0x543800 | +| 43 | blk.3.attn_q.bias | 0x2a1963c0 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x2a199bc0 | 0x405000 | +| 45 | blk.3.attn_v.bias | 0x2a59ebc0 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x2a59f3c0 | 0xc0800 | +| 47 | blk.3.ffn_down.weight | 0x2a65fbc0 | 0x246c000 | +| 48 | blk.3.ffn_gate.weight | 0x2cacbbc0 | 0x1bd2800 | +| 49 | blk.3.ffn_norm.weight | 0x2e69e3c0 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x2e6a1bc0 | 0x1bd2800 | +| 51 | blk.4.attn_k.bias | 0x302743c0 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x30274bc0 | 0x93000 | +| 53 | blk.4.attn_norm.weight | 0x30307bc0 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x3030b3c0 | 0x543800 | +| 55 | blk.4.attn_q.bias | 0x3084ebc0 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x308523c0 | 0x405000 | +| 57 | blk.4.attn_v.bias | 0x30c573c0 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x30c57bc0 | 0xc0800 | +| 59 | blk.4.ffn_down.weight | 0x30d183c0 | 0x1bd2800 | +| 60 | blk.4.ffn_gate.weight | 0x328eabc0 | 0x1bd2800 | +| 61 | blk.4.ffn_norm.weight | 0x344bd3c0 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x344c0bc0 | 0x1bd2800 | +| 63 | blk.5.attn_k.bias | 0x360933c0 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x36093bc0 | 0x93000 | +| 65 | blk.5.attn_norm.weight | 0x36126bc0 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x3612a3c0 | 0x543800 | +| 67 | blk.5.attn_q.bias | 0x3666dbc0 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x366713c0 | 0x405000 | +| 69 | blk.5.attn_v.bias | 0x36a763c0 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x36a76bc0 | 0xc0800 | +| 71 | blk.5.ffn_down.weight | 0x36b373c0 | 0x1bd2800 | +| 72 | blk.5.ffn_gate.weight | 0x38709bc0 | 0x1bd2800 | +| 73 | blk.5.ffn_norm.weight | 0x3a2dc3c0 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x3a2dfbc0 | 0x1bd2800 | +| 75 | blk.6.attn_k.bias | 0x3beb23c0 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x3beb2bc0 | 0x93000 | +| 77 | blk.6.attn_norm.weight | 0x3bf45bc0 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x3bf493c0 | 0x543800 | +| 79 | blk.6.attn_q.bias | 0x3c48cbc0 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x3c4903c0 | 0x405000 | +| 81 | blk.6.attn_v.bias | 0x3c8953c0 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x3c895bc0 | 0xc0800 | +| 83 | blk.6.ffn_down.weight | 0x3c9563c0 | 0x1bd2800 | +| 84 | blk.6.ffn_gate.weight | 0x3e528bc0 | 0x153f000 | +| 85 | blk.6.ffn_norm.weight | 0x3fa67bc0 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x3fa6b3c0 | 0x153f000 | +| 87 | blk.7.attn_k.bias | 0x40faa3c0 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x40faabc0 | 0x93000 | +| 89 | blk.7.attn_norm.weight | 0x4103dbc0 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x410413c0 | 0x543800 | +| 91 | blk.7.attn_q.bias | 0x41584bc0 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x415883c0 | 0x405000 | +| 93 | blk.7.attn_v.bias | 0x4198d3c0 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x4198dbc0 | 0xc0800 | +| 95 | blk.7.ffn_down.weight | 0x41a4e3c0 | 0x246c000 | +| 96 | blk.7.ffn_gate.weight | 0x43eba3c0 | 0x153f000 | +| 97 | blk.7.ffn_norm.weight | 0x453f93c0 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x453fcbc0 | 0x153f000 | +| 99 | blk.8.attn_k.bias | 0x4693bbc0 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x4693c3c0 | 0x93000 | +| 101 | blk.8.attn_norm.weight | 0x469cf3c0 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x469d2bc0 | 0x543800 | +| 103 | blk.8.attn_q.bias | 0x46f163c0 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x46f19bc0 | 0x405000 | +| 105 | blk.8.attn_v.bias | 0x4731ebc0 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x4731f3c0 | 0xc0800 | +| 107 | blk.8.ffn_down.weight | 0x473dfbc0 | 0x1bd2800 | +| 108 | blk.8.ffn_gate.weight | 0x48fb23c0 | 0x153f000 | +| 109 | blk.8.ffn_norm.weight | 0x4a4f13c0 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x4a4f4bc0 | 0x153f000 | +| 111 | blk.9.attn_k.bias | 0x4ba33bc0 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x4ba343c0 | 0x93000 | +| 113 | blk.9.attn_norm.weight | 0x4bac73c0 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x4bacabc0 | 0x543800 | +| 115 | blk.9.attn_q.bias | 0x4c00e3c0 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x4c011bc0 | 0x405000 | +| 117 | blk.9.attn_v.bias | 0x4c416bc0 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x4c4173c0 | 0xc0800 | +| 119 | blk.9.ffn_down.weight | 0x4c4d7bc0 | 0x246c000 | +| 120 | blk.9.ffn_gate.weight | 0x4e943bc0 | 0x1bd2800 | +| 121 | blk.9.ffn_norm.weight | 0x505163c0 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x50519bc0 | 0x1bd2800 | +| 123 | blk.10.attn_k.bias | 0x520ec3c0 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x520ecbc0 | 0x93000 | +| 125 | blk.10.attn_norm.weight | 0x5217fbc0 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x521833c0 | 0x543800 | +| 127 | blk.10.attn_q.bias | 0x526c6bc0 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x526ca3c0 | 0x405000 | +| 129 | blk.10.attn_v.bias | 0x52acf3c0 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x52acfbc0 | 0xc0800 | +| 131 | blk.10.ffn_down.weight | 0x52b903c0 | 0x246c000 | +| 132 | blk.10.ffn_gate.weight | 0x54ffc3c0 | 0x153f000 | +| 133 | blk.10.ffn_norm.weight | 0x5653b3c0 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x5653ebc0 | 0x153f000 | +| 135 | blk.11.attn_k.bias | 0x57a7dbc0 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x57a7e3c0 | 0x93000 | +| 137 | blk.11.attn_norm.weight | 0x57b113c0 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x57b14bc0 | 0x543800 | +| 139 | blk.11.attn_q.bias | 0x580583c0 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x5805bbc0 | 0x405000 | +| 141 | blk.11.attn_v.bias | 0x58460bc0 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x584613c0 | 0xc0800 | +| 143 | blk.11.ffn_down.weight | 0x58521bc0 | 0x1bd2800 | +| 144 | blk.11.ffn_gate.weight | 0x5a0f43c0 | 0x153f000 | +| 145 | blk.11.ffn_norm.weight | 0x5b6333c0 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x5b636bc0 | 0x153f000 | +| 147 | blk.12.attn_k.bias | 0x5cb75bc0 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x5cb763c0 | 0x93000 | +| 149 | blk.12.attn_norm.weight | 0x5cc093c0 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x5cc0cbc0 | 0x543800 | +| 151 | blk.12.attn_q.bias | 0x5d1503c0 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x5d153bc0 | 0x405000 | +| 153 | blk.12.attn_v.bias | 0x5d558bc0 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x5d5593c0 | 0xc0800 | +| 155 | blk.12.ffn_down.weight | 0x5d619bc0 | 0x1bd2800 | +| 156 | blk.12.ffn_gate.weight | 0x5f1ec3c0 | 0x153f000 | +| 157 | blk.12.ffn_norm.weight | 0x6072b3c0 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x6072ebc0 | 0x153f000 | +| 159 | blk.13.attn_k.bias | 0x61c6dbc0 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x61c6e3c0 | 0x93000 | +| 161 | blk.13.attn_norm.weight | 0x61d013c0 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x61d04bc0 | 0x543800 | +| 163 | blk.13.attn_q.bias | 0x622483c0 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x6224bbc0 | 0x405000 | +| 165 | blk.13.attn_v.bias | 0x62650bc0 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x626513c0 | 0xc0800 | +| 167 | blk.13.ffn_down.weight | 0x62711bc0 | 0x1bd2800 | +| 168 | blk.13.ffn_gate.weight | 0x642e43c0 | 0x153f000 | +| 169 | blk.13.ffn_norm.weight | 0x658233c0 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x65826bc0 | 0x153f000 | +| 171 | blk.14.attn_k.bias | 0x66d65bc0 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x66d663c0 | 0xc0800 | +| 173 | blk.14.attn_norm.weight | 0x66e26bc0 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x66e2a3c0 | 0x543800 | +| 175 | blk.14.attn_q.bias | 0x6736dbc0 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x673713c0 | 0x543800 | +| 177 | blk.14.attn_v.bias | 0x678b4bc0 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x678b53c0 | 0xc0800 | +| 179 | blk.14.ffn_down.weight | 0x67975bc0 | 0x1bd2800 | +| 180 | blk.14.ffn_gate.weight | 0x695483c0 | 0x153f000 | +| 181 | blk.14.ffn_norm.weight | 0x6aa873c0 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x6aa8abc0 | 0x153f000 | +| 183 | blk.15.attn_k.bias | 0x6bfc9bc0 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x6bfca3c0 | 0xc0800 | +| 185 | blk.15.attn_norm.weight | 0x6c08abc0 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x6c08e3c0 | 0x543800 | +| 187 | blk.15.attn_q.bias | 0x6c5d1bc0 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x6c5d53c0 | 0x543800 | +| 189 | blk.15.attn_v.bias | 0x6cb18bc0 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x6cb193c0 | 0xc0800 | +| 191 | blk.15.ffn_down.weight | 0x6cbd9bc0 | 0x1bd2800 | +| 192 | blk.15.ffn_gate.weight | 0x6e7ac3c0 | 0x153f000 | +| 193 | blk.15.ffn_norm.weight | 0x6fceb3c0 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x6fceebc0 | 0x153f000 | +| 195 | blk.16.attn_k.bias | 0x7122dbc0 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x7122e3c0 | 0xc0800 | +| 197 | blk.16.attn_norm.weight | 0x712eebc0 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x712f23c0 | 0x543800 | +| 199 | blk.16.attn_q.bias | 0x71835bc0 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x718393c0 | 0x543800 | +| 201 | blk.16.attn_v.bias | 0x71d7cbc0 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x71d7d3c0 | 0xc0800 | +| 203 | blk.16.ffn_down.weight | 0x71e3dbc0 | 0x1bd2800 | +| 204 | blk.16.ffn_gate.weight | 0x73a103c0 | 0x153f000 | +| 205 | blk.16.ffn_norm.weight | 0x74f4f3c0 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x74f52bc0 | 0x153f000 | +| 207 | blk.17.attn_k.bias | 0x76491bc0 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x764923c0 | 0xc0800 | +| 209 | blk.17.attn_norm.weight | 0x76552bc0 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x765563c0 | 0x543800 | +| 211 | blk.17.attn_q.bias | 0x76a99bc0 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x76a9d3c0 | 0x543800 | +| 213 | blk.17.attn_v.bias | 0x76fe0bc0 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x76fe13c0 | 0xc0800 | +| 215 | blk.17.ffn_down.weight | 0x770a1bc0 | 0x1bd2800 | +| 216 | blk.17.ffn_gate.weight | 0x78c743c0 | 0x153f000 | +| 217 | blk.17.ffn_norm.weight | 0x7a1b33c0 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x7a1b6bc0 | 0x153f000 | +| 219 | blk.18.attn_k.bias | 0x7b6f5bc0 | 0x800 | +| 220 | blk.18.attn_k.weight | 0x7b6f63c0 | 0xc0800 | +| 221 | blk.18.attn_norm.weight | 0x7b7b6bc0 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0x7b7ba3c0 | 0x543800 | +| 223 | blk.18.attn_q.bias | 0x7bcfdbc0 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0x7bd013c0 | 0x543800 | +| 225 | blk.18.attn_v.bias | 0x7c244bc0 | 0x800 | +| 226 | blk.18.attn_v.weight | 0x7c2453c0 | 0xc0800 | +| 227 | blk.18.ffn_down.weight | 0x7c305bc0 | 0x246c000 | +| 228 | blk.18.ffn_gate.weight | 0x7e771bc0 | 0x153f000 | +| 229 | blk.18.ffn_norm.weight | 0x7fcb0bc0 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x7fcb43c0 | 0x153f000 | +| 231 | blk.19.attn_k.bias | 0x811f33c0 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x811f3bc0 | 0xc0800 | +| 233 | blk.19.attn_norm.weight | 0x812b43c0 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x812b7bc0 | 0x543800 | +| 235 | blk.19.attn_q.bias | 0x817fb3c0 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x817febc0 | 0x543800 | +| 237 | blk.19.attn_v.bias | 0x81d423c0 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x81d42bc0 | 0xc0800 | +| 239 | blk.19.ffn_down.weight | 0x81e033c0 | 0x246c000 | +| 240 | blk.19.ffn_gate.weight | 0x8426f3c0 | 0x153f000 | +| 241 | blk.19.ffn_norm.weight | 0x857ae3c0 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x857b1bc0 | 0x153f000 | +| 243 | blk.20.attn_k.bias | 0x86cf0bc0 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x86cf13c0 | 0xc0800 | +| 245 | blk.20.attn_norm.weight | 0x86db1bc0 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x86db53c0 | 0x543800 | +| 247 | blk.20.attn_q.bias | 0x872f8bc0 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x872fc3c0 | 0x543800 | +| 249 | blk.20.attn_v.bias | 0x8783fbc0 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x878403c0 | 0xc0800 | +| 251 | blk.20.ffn_down.weight | 0x87900bc0 | 0x246c000 | +| 252 | blk.20.ffn_gate.weight | 0x89d6cbc0 | 0x1bd2800 | +| 253 | blk.20.ffn_norm.weight | 0x8b93f3c0 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x8b942bc0 | 0x1bd2800 | +| 255 | blk.21.attn_k.bias | 0x8d5153c0 | 0x800 | +| 256 | blk.21.attn_k.weight | 0x8d515bc0 | 0xc0800 | +| 257 | blk.21.attn_norm.weight | 0x8d5d63c0 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0x8d5d9bc0 | 0x543800 | +| 259 | blk.21.attn_q.bias | 0x8db1d3c0 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0x8db20bc0 | 0x543800 | +| 261 | blk.21.attn_v.bias | 0x8e0643c0 | 0x800 | +| 262 | blk.21.attn_v.weight | 0x8e064bc0 | 0xc0800 | +| 263 | blk.21.ffn_down.weight | 0x8e1253c0 | 0x246c000 | +| 264 | blk.21.ffn_gate.weight | 0x905913c0 | 0x1bd2800 | +| 265 | blk.21.ffn_norm.weight | 0x92163bc0 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0x921673c0 | 0x1bd2800 | +| 267 | blk.22.attn_k.bias | 0x93d39bc0 | 0x800 | +| 268 | blk.22.attn_k.weight | 0x93d3a3c0 | 0xc0800 | +| 269 | blk.22.attn_norm.weight | 0x93dfabc0 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0x93dfe3c0 | 0x543800 | +| 271 | blk.22.attn_q.bias | 0x94341bc0 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0x943453c0 | 0x543800 | +| 273 | blk.22.attn_v.bias | 0x94888bc0 | 0x800 | +| 274 | blk.22.attn_v.weight | 0x948893c0 | 0xc0800 | +| 275 | blk.22.ffn_down.weight | 0x94949bc0 | 0x246c000 | +| 276 | blk.22.ffn_gate.weight | 0x96db5bc0 | 0x1bd2800 | +| 277 | blk.22.ffn_norm.weight | 0x989883c0 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0x9898bbc0 | 0x1bd2800 | +| 279 | blk.23.attn_k.bias | 0x9a55e3c0 | 0x800 | +| 280 | blk.23.attn_k.weight | 0x9a55ebc0 | 0xc0800 | +| 281 | blk.23.attn_norm.weight | 0x9a61f3c0 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0x9a622bc0 | 0x543800 | +| 283 | blk.23.attn_q.bias | 0x9ab663c0 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0x9ab69bc0 | 0x543800 | +| 285 | blk.23.attn_v.bias | 0x9b0ad3c0 | 0x800 | +| 286 | blk.23.attn_v.weight | 0x9b0adbc0 | 0xc0800 | +| 287 | blk.23.ffn_down.weight | 0x9b16e3c0 | 0x246c000 | +| 288 | blk.23.ffn_gate.weight | 0x9d5da3c0 | 0x1bd2800 | +| 289 | blk.23.ffn_norm.weight | 0x9f1acbc0 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0x9f1b03c0 | 0x1bd2800 | +| 291 | blk.24.attn_k.bias | 0xa0d82bc0 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xa0d833c0 | 0xc0800 | +| 293 | blk.24.attn_norm.weight | 0xa0e43bc0 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xa0e473c0 | 0x543800 | +| 295 | blk.24.attn_q.bias | 0xa138abc0 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xa138e3c0 | 0x543800 | +| 297 | blk.24.attn_v.bias | 0xa18d1bc0 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xa18d23c0 | 0xc0800 | +| 299 | blk.24.ffn_down.weight | 0xa1992bc0 | 0x246c000 | +| 300 | blk.24.ffn_gate.weight | 0xa3dfebc0 | 0x1bd2800 | +| 301 | blk.24.ffn_norm.weight | 0xa59d13c0 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xa59d4bc0 | 0x1bd2800 | +| 303 | blk.25.attn_k.bias | 0xa75a73c0 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xa75a7bc0 | 0xc0800 | +| 305 | blk.25.attn_norm.weight | 0xa76683c0 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xa766bbc0 | 0x543800 | +| 307 | blk.25.attn_q.bias | 0xa7baf3c0 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xa7bb2bc0 | 0x543800 | +| 309 | blk.25.attn_v.bias | 0xa80f63c0 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xa80f6bc0 | 0xc0800 | +| 311 | blk.25.ffn_down.weight | 0xa81b73c0 | 0x246c000 | +| 312 | blk.25.ffn_gate.weight | 0xaa6233c0 | 0x1bd2800 | +| 313 | blk.25.ffn_norm.weight | 0xac1f5bc0 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xac1f93c0 | 0x1bd2800 | +| 315 | blk.26.attn_k.bias | 0xaddcbbc0 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xaddcc3c0 | 0xc0800 | +| 317 | blk.26.attn_norm.weight | 0xade8cbc0 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xade903c0 | 0x543800 | +| 319 | blk.26.attn_q.bias | 0xae3d3bc0 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xae3d73c0 | 0x543800 | +| 321 | blk.26.attn_v.bias | 0xae91abc0 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xae91b3c0 | 0xc0800 | +| 323 | blk.26.ffn_down.weight | 0xae9dbbc0 | 0x246c000 | +| 324 | blk.26.ffn_gate.weight | 0xb0e47bc0 | 0x1bd2800 | +| 325 | blk.26.ffn_norm.weight | 0xb2a1a3c0 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xb2a1dbc0 | 0x1bd2800 | +| 327 | blk.27.attn_k.bias | 0xb45f03c0 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xb45f0bc0 | 0xc0800 | +| 329 | blk.27.attn_norm.weight | 0xb46b13c0 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xb46b4bc0 | 0x543800 | +| 331 | blk.27.attn_q.bias | 0xb4bf83c0 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xb4bfbbc0 | 0x543800 | +| 333 | blk.27.attn_v.bias | 0xb513f3c0 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xb513fbc0 | 0xc0800 | +| 335 | blk.27.ffn_down.weight | 0xb52003c0 | 0x246c000 | +| 336 | blk.27.ffn_gate.weight | 0xb766c3c0 | 0x1bd2800 | +| 337 | blk.27.ffn_norm.weight | 0xb923ebc0 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xb92423c0 | 0x1bd2800 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q2_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q2_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q2_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q3_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q2_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q4_K_M.md b/scores/Hammer2.1-7b-Q4_K_M.md new file mode 100644 index 0000000..65d1b47 --- /dev/null +++ b/scores/Hammer2.1-7b-Q4_K_M.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q4_K_M.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 15 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q4\_K\_M.gguf - GGUF Internal File Dump](#hammer21-7b-q4_k_mgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x123979e0 | +| 1 | output_norm.weight | 0x12942ea0 | 0x3800 | +| 2 | token_embd.weight | 0x129466a0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x20804e80 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x20805680 | 0xc0800 | +| 5 | blk.0.attn_norm.weight | 0x208c5e80 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x208c9680 | 0x6e4000 | +| 7 | blk.0.attn_q.bias | 0x20fad680 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x20fb0e80 | 0x543800 | +| 9 | blk.0.attn_v.bias | 0x214f4680 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x214f4e80 | 0xfc000 | +| 11 | blk.0.ffn_down.weight | 0x215f0e80 | 0x2c84000 | +| 12 | blk.0.ffn_gate.weight | 0x24274e80 | 0x1bd2800 | +| 13 | blk.0.ffn_norm.weight | 0x25e47680 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x25e4ae80 | 0x1bd2800 | +| 15 | blk.1.attn_k.bias | 0x27a1d680 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x27a1de80 | 0xc0800 | +| 17 | blk.1.attn_norm.weight | 0x27ade680 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x27ae1e80 | 0x6e4000 | +| 19 | blk.1.attn_q.bias | 0x281c5e80 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x281c9680 | 0x543800 | +| 21 | blk.1.attn_v.bias | 0x2870ce80 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x2870d680 | 0xfc000 | +| 23 | blk.1.ffn_down.weight | 0x28809680 | 0x2c84000 | +| 24 | blk.1.ffn_gate.weight | 0x2b48d680 | 0x246c000 | +| 25 | blk.1.ffn_norm.weight | 0x2d8f9680 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x2d8fce80 | 0x246c000 | +| 27 | blk.2.attn_k.bias | 0x2fd68e80 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x2fd69680 | 0xc0800 | +| 29 | blk.2.attn_norm.weight | 0x2fe29e80 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x2fe2d680 | 0x6e4000 | +| 31 | blk.2.attn_q.bias | 0x30511680 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x30514e80 | 0x543800 | +| 33 | blk.2.attn_v.bias | 0x30a58680 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x30a58e80 | 0xfc000 | +| 35 | blk.2.ffn_down.weight | 0x30b54e80 | 0x2c84000 | +| 36 | blk.2.ffn_gate.weight | 0x337d8e80 | 0x246c000 | +| 37 | blk.2.ffn_norm.weight | 0x35c44e80 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x35c48680 | 0x246c000 | +| 39 | blk.3.attn_k.bias | 0x380b4680 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x380b4e80 | 0xc0800 | +| 41 | blk.3.attn_norm.weight | 0x38175680 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x38178e80 | 0x6e4000 | +| 43 | blk.3.attn_q.bias | 0x3885ce80 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x38860680 | 0x543800 | +| 45 | blk.3.attn_v.bias | 0x38da3e80 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x38da4680 | 0xfc000 | +| 47 | blk.3.ffn_down.weight | 0x38ea0680 | 0x2c84000 | +| 48 | blk.3.ffn_gate.weight | 0x3bb24680 | 0x246c000 | +| 49 | blk.3.ffn_norm.weight | 0x3df90680 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x3df93e80 | 0x246c000 | +| 51 | blk.4.attn_k.bias | 0x403ffe80 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x40400680 | 0xc0800 | +| 53 | blk.4.attn_norm.weight | 0x404c0e80 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x404c4680 | 0x6e4000 | +| 55 | blk.4.attn_q.bias | 0x40ba8680 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x40babe80 | 0x543800 | +| 57 | blk.4.attn_v.bias | 0x410ef680 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x410efe80 | 0xfc000 | +| 59 | blk.4.ffn_down.weight | 0x411ebe80 | 0x2c84000 | +| 60 | blk.4.ffn_gate.weight | 0x43e6fe80 | 0x246c000 | +| 61 | blk.4.ffn_norm.weight | 0x462dbe80 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x462df680 | 0x246c000 | +| 63 | blk.5.attn_k.bias | 0x4874b680 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x4874be80 | 0xc0800 | +| 65 | blk.5.attn_norm.weight | 0x4880c680 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x4880fe80 | 0x6e4000 | +| 67 | blk.5.attn_q.bias | 0x48ef3e80 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x48ef7680 | 0x543800 | +| 69 | blk.5.attn_v.bias | 0x4943ae80 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x4943b680 | 0xfc000 | +| 71 | blk.5.ffn_down.weight | 0x49537680 | 0x2c84000 | +| 72 | blk.5.ffn_gate.weight | 0x4c1bb680 | 0x246c000 | +| 73 | blk.5.ffn_norm.weight | 0x4e627680 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x4e62ae80 | 0x246c000 | +| 75 | blk.6.attn_k.bias | 0x50a96e80 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x50a97680 | 0xc0800 | +| 77 | blk.6.attn_norm.weight | 0x50b57e80 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x50b5b680 | 0x6e4000 | +| 79 | blk.6.attn_q.bias | 0x5123f680 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x51242e80 | 0x543800 | +| 81 | blk.6.attn_v.bias | 0x51786680 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x51786e80 | 0xfc000 | +| 83 | blk.6.ffn_down.weight | 0x51882e80 | 0x2c84000 | +| 84 | blk.6.ffn_gate.weight | 0x54506e80 | 0x1bd2800 | +| 85 | blk.6.ffn_norm.weight | 0x560d9680 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x560dce80 | 0x1bd2800 | +| 87 | blk.7.attn_k.bias | 0x57caf680 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x57cafe80 | 0xc0800 | +| 89 | blk.7.attn_norm.weight | 0x57d70680 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x57d73e80 | 0x6e4000 | +| 91 | blk.7.attn_q.bias | 0x58457e80 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x5845b680 | 0x543800 | +| 93 | blk.7.attn_v.bias | 0x5899ee80 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x5899f680 | 0xfc000 | +| 95 | blk.7.ffn_down.weight | 0x58a9b680 | 0x2c84000 | +| 96 | blk.7.ffn_gate.weight | 0x5b71f680 | 0x1bd2800 | +| 97 | blk.7.ffn_norm.weight | 0x5d2f1e80 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x5d2f5680 | 0x1bd2800 | +| 99 | blk.8.attn_k.bias | 0x5eec7e80 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x5eec8680 | 0xc0800 | +| 101 | blk.8.attn_norm.weight | 0x5ef88e80 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x5ef8c680 | 0x6e4000 | +| 103 | blk.8.attn_q.bias | 0x5f670680 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x5f673e80 | 0x543800 | +| 105 | blk.8.attn_v.bias | 0x5fbb7680 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x5fbb7e80 | 0xfc000 | +| 107 | blk.8.ffn_down.weight | 0x5fcb3e80 | 0x2c84000 | +| 108 | blk.8.ffn_gate.weight | 0x62937e80 | 0x1bd2800 | +| 109 | blk.8.ffn_norm.weight | 0x6450a680 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x6450de80 | 0x1bd2800 | +| 111 | blk.9.attn_k.bias | 0x660e0680 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x660e0e80 | 0xc0800 | +| 113 | blk.9.attn_norm.weight | 0x661a1680 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x661a4e80 | 0x6e4000 | +| 115 | blk.9.attn_q.bias | 0x66888e80 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x6688c680 | 0x543800 | +| 117 | blk.9.attn_v.bias | 0x66dcfe80 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x66dd0680 | 0xfc000 | +| 119 | blk.9.ffn_down.weight | 0x66ecc680 | 0x2c84000 | +| 120 | blk.9.ffn_gate.weight | 0x69b50680 | 0x246c000 | +| 121 | blk.9.ffn_norm.weight | 0x6bfbc680 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x6bfbfe80 | 0x246c000 | +| 123 | blk.10.attn_k.bias | 0x6e42be80 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x6e42c680 | 0xc0800 | +| 125 | blk.10.attn_norm.weight | 0x6e4ece80 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x6e4f0680 | 0x6e4000 | +| 127 | blk.10.attn_q.bias | 0x6ebd4680 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x6ebd7e80 | 0x543800 | +| 129 | blk.10.attn_v.bias | 0x6f11b680 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x6f11be80 | 0xfc000 | +| 131 | blk.10.ffn_down.weight | 0x6f217e80 | 0x2c84000 | +| 132 | blk.10.ffn_gate.weight | 0x71e9be80 | 0x1bd2800 | +| 133 | blk.10.ffn_norm.weight | 0x73a6e680 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x73a71e80 | 0x1bd2800 | +| 135 | blk.11.attn_k.bias | 0x75644680 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x75644e80 | 0xc0800 | +| 137 | blk.11.attn_norm.weight | 0x75705680 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x75708e80 | 0x6e4000 | +| 139 | blk.11.attn_q.bias | 0x75dece80 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x75df0680 | 0x543800 | +| 141 | blk.11.attn_v.bias | 0x76333e80 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x76334680 | 0xfc000 | +| 143 | blk.11.ffn_down.weight | 0x76430680 | 0x2c84000 | +| 144 | blk.11.ffn_gate.weight | 0x790b4680 | 0x1bd2800 | +| 145 | blk.11.ffn_norm.weight | 0x7ac86e80 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x7ac8a680 | 0x1bd2800 | +| 147 | blk.12.attn_k.bias | 0x7c85ce80 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x7c85d680 | 0xc0800 | +| 149 | blk.12.attn_norm.weight | 0x7c91de80 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x7c921680 | 0x6e4000 | +| 151 | blk.12.attn_q.bias | 0x7d005680 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x7d008e80 | 0x543800 | +| 153 | blk.12.attn_v.bias | 0x7d54c680 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x7d54ce80 | 0xfc000 | +| 155 | blk.12.ffn_down.weight | 0x7d648e80 | 0x2c84000 | +| 156 | blk.12.ffn_gate.weight | 0x802cce80 | 0x1bd2800 | +| 157 | blk.12.ffn_norm.weight | 0x81e9f680 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x81ea2e80 | 0x1bd2800 | +| 159 | blk.13.attn_k.bias | 0x83a75680 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x83a75e80 | 0xc0800 | +| 161 | blk.13.attn_norm.weight | 0x83b36680 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x83b39e80 | 0x6e4000 | +| 163 | blk.13.attn_q.bias | 0x8421de80 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x84221680 | 0x543800 | +| 165 | blk.13.attn_v.bias | 0x84764e80 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x84765680 | 0xfc000 | +| 167 | blk.13.ffn_down.weight | 0x84861680 | 0x2c84000 | +| 168 | blk.13.ffn_gate.weight | 0x874e5680 | 0x1bd2800 | +| 169 | blk.13.ffn_norm.weight | 0x890b7e80 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x890bb680 | 0x1bd2800 | +| 171 | blk.14.attn_k.bias | 0x8ac8de80 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x8ac8e680 | 0xfc000 | +| 173 | blk.14.attn_norm.weight | 0x8ad8a680 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x8ad8de80 | 0x6e4000 | +| 175 | blk.14.attn_q.bias | 0x8b471e80 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x8b475680 | 0x6e4000 | +| 177 | blk.14.attn_v.bias | 0x8bb59680 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x8bb59e80 | 0x134000 | +| 179 | blk.14.ffn_down.weight | 0x8bc8de80 | 0x2c84000 | +| 180 | blk.14.ffn_gate.weight | 0x8e911e80 | 0x1bd2800 | +| 181 | blk.14.ffn_norm.weight | 0x904e4680 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x904e7e80 | 0x1bd2800 | +| 183 | blk.15.attn_k.bias | 0x920ba680 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x920bae80 | 0xfc000 | +| 185 | blk.15.attn_norm.weight | 0x921b6e80 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x921ba680 | 0x6e4000 | +| 187 | blk.15.attn_q.bias | 0x9289e680 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x928a1e80 | 0x6e4000 | +| 189 | blk.15.attn_v.bias | 0x92f85e80 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x92f86680 | 0x134000 | +| 191 | blk.15.ffn_down.weight | 0x930ba680 | 0x2c84000 | +| 192 | blk.15.ffn_gate.weight | 0x95d3e680 | 0x1bd2800 | +| 193 | blk.15.ffn_norm.weight | 0x97910e80 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x97914680 | 0x1bd2800 | +| 195 | blk.16.attn_k.bias | 0x994e6e80 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x994e7680 | 0xfc000 | +| 197 | blk.16.attn_norm.weight | 0x995e3680 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x995e6e80 | 0x6e4000 | +| 199 | blk.16.attn_q.bias | 0x99ccae80 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x99cce680 | 0x6e4000 | +| 201 | blk.16.attn_v.bias | 0x9a3b2680 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x9a3b2e80 | 0x134000 | +| 203 | blk.16.ffn_down.weight | 0x9a4e6e80 | 0x2c84000 | +| 204 | blk.16.ffn_gate.weight | 0x9d16ae80 | 0x1bd2800 | +| 205 | blk.16.ffn_norm.weight | 0x9ed3d680 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x9ed40e80 | 0x1bd2800 | +| 207 | blk.17.attn_k.bias | 0xa0913680 | 0x800 | +| 208 | blk.17.attn_k.weight | 0xa0913e80 | 0xfc000 | +| 209 | blk.17.attn_norm.weight | 0xa0a0fe80 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0xa0a13680 | 0x6e4000 | +| 211 | blk.17.attn_q.bias | 0xa10f7680 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0xa10fae80 | 0x6e4000 | +| 213 | blk.17.attn_v.bias | 0xa17dee80 | 0x800 | +| 214 | blk.17.attn_v.weight | 0xa17df680 | 0x134000 | +| 215 | blk.17.ffn_down.weight | 0xa1913680 | 0x2c84000 | +| 216 | blk.17.ffn_gate.weight | 0xa4597680 | 0x1bd2800 | +| 217 | blk.17.ffn_norm.weight | 0xa6169e80 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0xa616d680 | 0x1bd2800 | +| 219 | blk.18.attn_k.bias | 0xa7d3fe80 | 0x800 | +| 220 | blk.18.attn_k.weight | 0xa7d40680 | 0xfc000 | +| 221 | blk.18.attn_norm.weight | 0xa7e3c680 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0xa7e3fe80 | 0x6e4000 | +| 223 | blk.18.attn_q.bias | 0xa8523e80 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0xa8527680 | 0x6e4000 | +| 225 | blk.18.attn_v.bias | 0xa8c0b680 | 0x800 | +| 226 | blk.18.attn_v.weight | 0xa8c0be80 | 0x134000 | +| 227 | blk.18.ffn_down.weight | 0xa8d3fe80 | 0x2c84000 | +| 228 | blk.18.ffn_gate.weight | 0xab9c3e80 | 0x1bd2800 | +| 229 | blk.18.ffn_norm.weight | 0xad596680 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0xad599e80 | 0x1bd2800 | +| 231 | blk.19.attn_k.bias | 0xaf16c680 | 0x800 | +| 232 | blk.19.attn_k.weight | 0xaf16ce80 | 0xfc000 | +| 233 | blk.19.attn_norm.weight | 0xaf268e80 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0xaf26c680 | 0x6e4000 | +| 235 | blk.19.attn_q.bias | 0xaf950680 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0xaf953e80 | 0x6e4000 | +| 237 | blk.19.attn_v.bias | 0xb0037e80 | 0x800 | +| 238 | blk.19.attn_v.weight | 0xb0038680 | 0x134000 | +| 239 | blk.19.ffn_down.weight | 0xb016c680 | 0x2c84000 | +| 240 | blk.19.ffn_gate.weight | 0xb2df0680 | 0x1bd2800 | +| 241 | blk.19.ffn_norm.weight | 0xb49c2e80 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0xb49c6680 | 0x1bd2800 | +| 243 | blk.20.attn_k.bias | 0xb6598e80 | 0x800 | +| 244 | blk.20.attn_k.weight | 0xb6599680 | 0xfc000 | +| 245 | blk.20.attn_norm.weight | 0xb6695680 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0xb6698e80 | 0x6e4000 | +| 247 | blk.20.attn_q.bias | 0xb6d7ce80 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0xb6d80680 | 0x6e4000 | +| 249 | blk.20.attn_v.bias | 0xb7464680 | 0x800 | +| 250 | blk.20.attn_v.weight | 0xb7464e80 | 0x134000 | +| 251 | blk.20.ffn_down.weight | 0xb7598e80 | 0x2c84000 | +| 252 | blk.20.ffn_gate.weight | 0xba21ce80 | 0x246c000 | +| 253 | blk.20.ffn_norm.weight | 0xbc688e80 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0xbc68c680 | 0x246c000 | +| 255 | blk.21.attn_k.bias | 0xbeaf8680 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xbeaf8e80 | 0xfc000 | +| 257 | blk.21.attn_norm.weight | 0xbebf4e80 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xbebf8680 | 0x6e4000 | +| 259 | blk.21.attn_q.bias | 0xbf2dc680 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xbf2dfe80 | 0x6e4000 | +| 261 | blk.21.attn_v.bias | 0xbf9c3e80 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xbf9c4680 | 0x134000 | +| 263 | blk.21.ffn_down.weight | 0xbfaf8680 | 0x2c84000 | +| 264 | blk.21.ffn_gate.weight | 0xc277c680 | 0x246c000 | +| 265 | blk.21.ffn_norm.weight | 0xc4be8680 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xc4bebe80 | 0x246c000 | +| 267 | blk.22.attn_k.bias | 0xc7057e80 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xc7058680 | 0xfc000 | +| 269 | blk.22.attn_norm.weight | 0xc7154680 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xc7157e80 | 0x6e4000 | +| 271 | blk.22.attn_q.bias | 0xc783be80 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xc783f680 | 0x6e4000 | +| 273 | blk.22.attn_v.bias | 0xc7f23680 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xc7f23e80 | 0x134000 | +| 275 | blk.22.ffn_down.weight | 0xc8057e80 | 0x2c84000 | +| 276 | blk.22.ffn_gate.weight | 0xcacdbe80 | 0x246c000 | +| 277 | blk.22.ffn_norm.weight | 0xcd147e80 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xcd14b680 | 0x246c000 | +| 279 | blk.23.attn_k.bias | 0xcf5b7680 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xcf5b7e80 | 0xfc000 | +| 281 | blk.23.attn_norm.weight | 0xcf6b3e80 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xcf6b7680 | 0x6e4000 | +| 283 | blk.23.attn_q.bias | 0xcfd9b680 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xcfd9ee80 | 0x6e4000 | +| 285 | blk.23.attn_v.bias | 0xd0482e80 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xd0483680 | 0x134000 | +| 287 | blk.23.ffn_down.weight | 0xd05b7680 | 0x2c84000 | +| 288 | blk.23.ffn_gate.weight | 0xd323b680 | 0x246c000 | +| 289 | blk.23.ffn_norm.weight | 0xd56a7680 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xd56aae80 | 0x246c000 | +| 291 | blk.24.attn_k.bias | 0xd7b16e80 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xd7b17680 | 0xfc000 | +| 293 | blk.24.attn_norm.weight | 0xd7c13680 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xd7c16e80 | 0x6e4000 | +| 295 | blk.24.attn_q.bias | 0xd82fae80 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xd82fe680 | 0x6e4000 | +| 297 | blk.24.attn_v.bias | 0xd89e2680 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xd89e2e80 | 0x134000 | +| 299 | blk.24.ffn_down.weight | 0xd8b16e80 | 0x2c84000 | +| 300 | blk.24.ffn_gate.weight | 0xdb79ae80 | 0x246c000 | +| 301 | blk.24.ffn_norm.weight | 0xddc06e80 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xddc0a680 | 0x246c000 | +| 303 | blk.25.attn_k.bias | 0xe0076680 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xe0076e80 | 0xfc000 | +| 305 | blk.25.attn_norm.weight | 0xe0172e80 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xe0176680 | 0x6e4000 | +| 307 | blk.25.attn_q.bias | 0xe085a680 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xe085de80 | 0x6e4000 | +| 309 | blk.25.attn_v.bias | 0xe0f41e80 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xe0f42680 | 0x134000 | +| 311 | blk.25.ffn_down.weight | 0xe1076680 | 0x2c84000 | +| 312 | blk.25.ffn_gate.weight | 0xe3cfa680 | 0x246c000 | +| 313 | blk.25.ffn_norm.weight | 0xe6166680 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xe6169e80 | 0x246c000 | +| 315 | blk.26.attn_k.bias | 0xe85d5e80 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xe85d6680 | 0xfc000 | +| 317 | blk.26.attn_norm.weight | 0xe86d2680 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xe86d5e80 | 0x6e4000 | +| 319 | blk.26.attn_q.bias | 0xe8db9e80 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xe8dbd680 | 0x6e4000 | +| 321 | blk.26.attn_v.bias | 0xe94a1680 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xe94a1e80 | 0x134000 | +| 323 | blk.26.ffn_down.weight | 0xe95d5e80 | 0x2c84000 | +| 324 | blk.26.ffn_gate.weight | 0xec259e80 | 0x246c000 | +| 325 | blk.26.ffn_norm.weight | 0xee6c5e80 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xee6c9680 | 0x246c000 | +| 327 | blk.27.attn_k.bias | 0xf0b35680 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xf0b35e80 | 0xfc000 | +| 329 | blk.27.attn_norm.weight | 0xf0c31e80 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xf0c35680 | 0x6e4000 | +| 331 | blk.27.attn_q.bias | 0xf1319680 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xf131ce80 | 0x6e4000 | +| 333 | blk.27.attn_v.bias | 0xf1a00e80 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xf1a01680 | 0x134000 | +| 335 | blk.27.ffn_down.weight | 0xf1b35680 | 0x2c84000 | +| 336 | blk.27.ffn_gate.weight | 0xf47b9680 | 0x246c000 | +| 337 | blk.27.ffn_norm.weight | 0xf6c25680 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xf6c28e80 | 0x246c000 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q4_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q4_K_S.md b/scores/Hammer2.1-7b-Q4_K_S.md new file mode 100644 index 0000000..fbb6dc9 --- /dev/null +++ b/scores/Hammer2.1-7b-Q4_K_S.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q4_K_S.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 14 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q4\_K\_S.gguf - GGUF Internal File Dump](#hammer21-7b-q4_k_sgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x123979e0 | +| 1 | output_norm.weight | 0x12942ea0 | 0x3800 | +| 2 | token_embd.weight | 0x129466a0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x20804e80 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x20805680 | 0xc0800 | +| 5 | blk.0.attn_norm.weight | 0x208c5e80 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x208c9680 | 0x6e4000 | +| 7 | blk.0.attn_q.bias | 0x20fad680 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x20fb0e80 | 0x543800 | +| 9 | blk.0.attn_v.bias | 0x214f4680 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x214f4e80 | 0xfc000 | +| 11 | blk.0.ffn_down.weight | 0x215f0e80 | 0x246c000 | +| 12 | blk.0.ffn_gate.weight | 0x23a5ce80 | 0x1bd2800 | +| 13 | blk.0.ffn_norm.weight | 0x2562f680 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x25632e80 | 0x1bd2800 | +| 15 | blk.1.attn_k.bias | 0x27205680 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x27205e80 | 0xc0800 | +| 17 | blk.1.attn_norm.weight | 0x272c6680 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x272c9e80 | 0x6e4000 | +| 19 | blk.1.attn_q.bias | 0x279ade80 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x279b1680 | 0x543800 | +| 21 | blk.1.attn_v.bias | 0x27ef4e80 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x27ef5680 | 0xfc000 | +| 23 | blk.1.ffn_down.weight | 0x27ff1680 | 0x246c000 | +| 24 | blk.1.ffn_gate.weight | 0x2a45d680 | 0x246c000 | +| 25 | blk.1.ffn_norm.weight | 0x2c8c9680 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x2c8cce80 | 0x246c000 | +| 27 | blk.2.attn_k.bias | 0x2ed38e80 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x2ed39680 | 0xc0800 | +| 29 | blk.2.attn_norm.weight | 0x2edf9e80 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x2edfd680 | 0x6e4000 | +| 31 | blk.2.attn_q.bias | 0x2f4e1680 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x2f4e4e80 | 0x543800 | +| 33 | blk.2.attn_v.bias | 0x2fa28680 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x2fa28e80 | 0xfc000 | +| 35 | blk.2.ffn_down.weight | 0x2fb24e80 | 0x246c000 | +| 36 | blk.2.ffn_gate.weight | 0x31f90e80 | 0x246c000 | +| 37 | blk.2.ffn_norm.weight | 0x343fce80 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x34400680 | 0x246c000 | +| 39 | blk.3.attn_k.bias | 0x3686c680 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x3686ce80 | 0xc0800 | +| 41 | blk.3.attn_norm.weight | 0x3692d680 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x36930e80 | 0x6e4000 | +| 43 | blk.3.attn_q.bias | 0x37014e80 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x37018680 | 0x543800 | +| 45 | blk.3.attn_v.bias | 0x3755be80 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x3755c680 | 0xfc000 | +| 47 | blk.3.ffn_down.weight | 0x37658680 | 0x2c84000 | +| 48 | blk.3.ffn_gate.weight | 0x3a2dc680 | 0x246c000 | +| 49 | blk.3.ffn_norm.weight | 0x3c748680 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x3c74be80 | 0x246c000 | +| 51 | blk.4.attn_k.bias | 0x3ebb7e80 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x3ebb8680 | 0xc0800 | +| 53 | blk.4.attn_norm.weight | 0x3ec78e80 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x3ec7c680 | 0x6e4000 | +| 55 | blk.4.attn_q.bias | 0x3f360680 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x3f363e80 | 0x543800 | +| 57 | blk.4.attn_v.bias | 0x3f8a7680 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x3f8a7e80 | 0xfc000 | +| 59 | blk.4.ffn_down.weight | 0x3f9a3e80 | 0x246c000 | +| 60 | blk.4.ffn_gate.weight | 0x41e0fe80 | 0x246c000 | +| 61 | blk.4.ffn_norm.weight | 0x4427be80 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x4427f680 | 0x246c000 | +| 63 | blk.5.attn_k.bias | 0x466eb680 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x466ebe80 | 0xc0800 | +| 65 | blk.5.attn_norm.weight | 0x467ac680 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x467afe80 | 0x6e4000 | +| 67 | blk.5.attn_q.bias | 0x46e93e80 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x46e97680 | 0x543800 | +| 69 | blk.5.attn_v.bias | 0x473dae80 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x473db680 | 0xfc000 | +| 71 | blk.5.ffn_down.weight | 0x474d7680 | 0x246c000 | +| 72 | blk.5.ffn_gate.weight | 0x49943680 | 0x246c000 | +| 73 | blk.5.ffn_norm.weight | 0x4bdaf680 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x4bdb2e80 | 0x246c000 | +| 75 | blk.6.attn_k.bias | 0x4e21ee80 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x4e21f680 | 0xc0800 | +| 77 | blk.6.attn_norm.weight | 0x4e2dfe80 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x4e2e3680 | 0x6e4000 | +| 79 | blk.6.attn_q.bias | 0x4e9c7680 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x4e9cae80 | 0x543800 | +| 81 | blk.6.attn_v.bias | 0x4ef0e680 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x4ef0ee80 | 0xfc000 | +| 83 | blk.6.ffn_down.weight | 0x4f00ae80 | 0x246c000 | +| 84 | blk.6.ffn_gate.weight | 0x51476e80 | 0x1bd2800 | +| 85 | blk.6.ffn_norm.weight | 0x53049680 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x5304ce80 | 0x1bd2800 | +| 87 | blk.7.attn_k.bias | 0x54c1f680 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x54c1fe80 | 0xc0800 | +| 89 | blk.7.attn_norm.weight | 0x54ce0680 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x54ce3e80 | 0x6e4000 | +| 91 | blk.7.attn_q.bias | 0x553c7e80 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x553cb680 | 0x543800 | +| 93 | blk.7.attn_v.bias | 0x5590ee80 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x5590f680 | 0xfc000 | +| 95 | blk.7.ffn_down.weight | 0x55a0b680 | 0x2c84000 | +| 96 | blk.7.ffn_gate.weight | 0x5868f680 | 0x1bd2800 | +| 97 | blk.7.ffn_norm.weight | 0x5a261e80 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x5a265680 | 0x1bd2800 | +| 99 | blk.8.attn_k.bias | 0x5be37e80 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x5be38680 | 0xc0800 | +| 101 | blk.8.attn_norm.weight | 0x5bef8e80 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x5befc680 | 0x6e4000 | +| 103 | blk.8.attn_q.bias | 0x5c5e0680 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x5c5e3e80 | 0x543800 | +| 105 | blk.8.attn_v.bias | 0x5cb27680 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x5cb27e80 | 0xfc000 | +| 107 | blk.8.ffn_down.weight | 0x5cc23e80 | 0x246c000 | +| 108 | blk.8.ffn_gate.weight | 0x5f08fe80 | 0x1bd2800 | +| 109 | blk.8.ffn_norm.weight | 0x60c62680 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x60c65e80 | 0x1bd2800 | +| 111 | blk.9.attn_k.bias | 0x62838680 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x62838e80 | 0xc0800 | +| 113 | blk.9.attn_norm.weight | 0x628f9680 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x628fce80 | 0x6e4000 | +| 115 | blk.9.attn_q.bias | 0x62fe0e80 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x62fe4680 | 0x543800 | +| 117 | blk.9.attn_v.bias | 0x63527e80 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x63528680 | 0xfc000 | +| 119 | blk.9.ffn_down.weight | 0x63624680 | 0x2c84000 | +| 120 | blk.9.ffn_gate.weight | 0x662a8680 | 0x246c000 | +| 121 | blk.9.ffn_norm.weight | 0x68714680 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x68717e80 | 0x246c000 | +| 123 | blk.10.attn_k.bias | 0x6ab83e80 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x6ab84680 | 0xc0800 | +| 125 | blk.10.attn_norm.weight | 0x6ac44e80 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x6ac48680 | 0x6e4000 | +| 127 | blk.10.attn_q.bias | 0x6b32c680 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x6b32fe80 | 0x543800 | +| 129 | blk.10.attn_v.bias | 0x6b873680 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x6b873e80 | 0xfc000 | +| 131 | blk.10.ffn_down.weight | 0x6b96fe80 | 0x2c84000 | +| 132 | blk.10.ffn_gate.weight | 0x6e5f3e80 | 0x1bd2800 | +| 133 | blk.10.ffn_norm.weight | 0x701c6680 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x701c9e80 | 0x1bd2800 | +| 135 | blk.11.attn_k.bias | 0x71d9c680 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x71d9ce80 | 0xc0800 | +| 137 | blk.11.attn_norm.weight | 0x71e5d680 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x71e60e80 | 0x6e4000 | +| 139 | blk.11.attn_q.bias | 0x72544e80 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x72548680 | 0x543800 | +| 141 | blk.11.attn_v.bias | 0x72a8be80 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x72a8c680 | 0xfc000 | +| 143 | blk.11.ffn_down.weight | 0x72b88680 | 0x246c000 | +| 144 | blk.11.ffn_gate.weight | 0x74ff4680 | 0x1bd2800 | +| 145 | blk.11.ffn_norm.weight | 0x76bc6e80 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x76bca680 | 0x1bd2800 | +| 147 | blk.12.attn_k.bias | 0x7879ce80 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x7879d680 | 0xc0800 | +| 149 | blk.12.attn_norm.weight | 0x7885de80 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x78861680 | 0x6e4000 | +| 151 | blk.12.attn_q.bias | 0x78f45680 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x78f48e80 | 0x543800 | +| 153 | blk.12.attn_v.bias | 0x7948c680 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x7948ce80 | 0xfc000 | +| 155 | blk.12.ffn_down.weight | 0x79588e80 | 0x246c000 | +| 156 | blk.12.ffn_gate.weight | 0x7b9f4e80 | 0x1bd2800 | +| 157 | blk.12.ffn_norm.weight | 0x7d5c7680 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x7d5cae80 | 0x1bd2800 | +| 159 | blk.13.attn_k.bias | 0x7f19d680 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x7f19de80 | 0xc0800 | +| 161 | blk.13.attn_norm.weight | 0x7f25e680 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x7f261e80 | 0x6e4000 | +| 163 | blk.13.attn_q.bias | 0x7f945e80 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x7f949680 | 0x543800 | +| 165 | blk.13.attn_v.bias | 0x7fe8ce80 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x7fe8d680 | 0xfc000 | +| 167 | blk.13.ffn_down.weight | 0x7ff89680 | 0x246c000 | +| 168 | blk.13.ffn_gate.weight | 0x823f5680 | 0x1bd2800 | +| 169 | blk.13.ffn_norm.weight | 0x83fc7e80 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0x83fcb680 | 0x1bd2800 | +| 171 | blk.14.attn_k.bias | 0x85b9de80 | 0x800 | +| 172 | blk.14.attn_k.weight | 0x85b9e680 | 0xfc000 | +| 173 | blk.14.attn_norm.weight | 0x85c9a680 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0x85c9de80 | 0x6e4000 | +| 175 | blk.14.attn_q.bias | 0x86381e80 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0x86385680 | 0x6e4000 | +| 177 | blk.14.attn_v.bias | 0x86a69680 | 0x800 | +| 178 | blk.14.attn_v.weight | 0x86a69e80 | 0xfc000 | +| 179 | blk.14.ffn_down.weight | 0x86b65e80 | 0x246c000 | +| 180 | blk.14.ffn_gate.weight | 0x88fd1e80 | 0x1bd2800 | +| 181 | blk.14.ffn_norm.weight | 0x8aba4680 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0x8aba7e80 | 0x1bd2800 | +| 183 | blk.15.attn_k.bias | 0x8c77a680 | 0x800 | +| 184 | blk.15.attn_k.weight | 0x8c77ae80 | 0xfc000 | +| 185 | blk.15.attn_norm.weight | 0x8c876e80 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0x8c87a680 | 0x6e4000 | +| 187 | blk.15.attn_q.bias | 0x8cf5e680 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0x8cf61e80 | 0x6e4000 | +| 189 | blk.15.attn_v.bias | 0x8d645e80 | 0x800 | +| 190 | blk.15.attn_v.weight | 0x8d646680 | 0xfc000 | +| 191 | blk.15.ffn_down.weight | 0x8d742680 | 0x246c000 | +| 192 | blk.15.ffn_gate.weight | 0x8fbae680 | 0x1bd2800 | +| 193 | blk.15.ffn_norm.weight | 0x91780e80 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x91784680 | 0x1bd2800 | +| 195 | blk.16.attn_k.bias | 0x93356e80 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x93357680 | 0xfc000 | +| 197 | blk.16.attn_norm.weight | 0x93453680 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x93456e80 | 0x6e4000 | +| 199 | blk.16.attn_q.bias | 0x93b3ae80 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x93b3e680 | 0x6e4000 | +| 201 | blk.16.attn_v.bias | 0x94222680 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x94222e80 | 0xfc000 | +| 203 | blk.16.ffn_down.weight | 0x9431ee80 | 0x246c000 | +| 204 | blk.16.ffn_gate.weight | 0x9678ae80 | 0x1bd2800 | +| 205 | blk.16.ffn_norm.weight | 0x9835d680 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x98360e80 | 0x1bd2800 | +| 207 | blk.17.attn_k.bias | 0x99f33680 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x99f33e80 | 0xfc000 | +| 209 | blk.17.attn_norm.weight | 0x9a02fe80 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x9a033680 | 0x6e4000 | +| 211 | blk.17.attn_q.bias | 0x9a717680 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x9a71ae80 | 0x6e4000 | +| 213 | blk.17.attn_v.bias | 0x9adfee80 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x9adff680 | 0xfc000 | +| 215 | blk.17.ffn_down.weight | 0x9aefb680 | 0x246c000 | +| 216 | blk.17.ffn_gate.weight | 0x9d367680 | 0x1bd2800 | +| 217 | blk.17.ffn_norm.weight | 0x9ef39e80 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x9ef3d680 | 0x1bd2800 | +| 219 | blk.18.attn_k.bias | 0xa0b0fe80 | 0x800 | +| 220 | blk.18.attn_k.weight | 0xa0b10680 | 0xfc000 | +| 221 | blk.18.attn_norm.weight | 0xa0c0c680 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0xa0c0fe80 | 0x6e4000 | +| 223 | blk.18.attn_q.bias | 0xa12f3e80 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0xa12f7680 | 0x6e4000 | +| 225 | blk.18.attn_v.bias | 0xa19db680 | 0x800 | +| 226 | blk.18.attn_v.weight | 0xa19dbe80 | 0xfc000 | +| 227 | blk.18.ffn_down.weight | 0xa1ad7e80 | 0x2c84000 | +| 228 | blk.18.ffn_gate.weight | 0xa475be80 | 0x1bd2800 | +| 229 | blk.18.ffn_norm.weight | 0xa632e680 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0xa6331e80 | 0x1bd2800 | +| 231 | blk.19.attn_k.bias | 0xa7f04680 | 0x800 | +| 232 | blk.19.attn_k.weight | 0xa7f04e80 | 0xfc000 | +| 233 | blk.19.attn_norm.weight | 0xa8000e80 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0xa8004680 | 0x6e4000 | +| 235 | blk.19.attn_q.bias | 0xa86e8680 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0xa86ebe80 | 0x6e4000 | +| 237 | blk.19.attn_v.bias | 0xa8dcfe80 | 0x800 | +| 238 | blk.19.attn_v.weight | 0xa8dd0680 | 0xfc000 | +| 239 | blk.19.ffn_down.weight | 0xa8ecc680 | 0x2c84000 | +| 240 | blk.19.ffn_gate.weight | 0xabb50680 | 0x1bd2800 | +| 241 | blk.19.ffn_norm.weight | 0xad722e80 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0xad726680 | 0x1bd2800 | +| 243 | blk.20.attn_k.bias | 0xaf2f8e80 | 0x800 | +| 244 | blk.20.attn_k.weight | 0xaf2f9680 | 0xfc000 | +| 245 | blk.20.attn_norm.weight | 0xaf3f5680 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0xaf3f8e80 | 0x6e4000 | +| 247 | blk.20.attn_q.bias | 0xafadce80 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0xafae0680 | 0x6e4000 | +| 249 | blk.20.attn_v.bias | 0xb01c4680 | 0x800 | +| 250 | blk.20.attn_v.weight | 0xb01c4e80 | 0xfc000 | +| 251 | blk.20.ffn_down.weight | 0xb02c0e80 | 0x2c84000 | +| 252 | blk.20.ffn_gate.weight | 0xb2f44e80 | 0x246c000 | +| 253 | blk.20.ffn_norm.weight | 0xb53b0e80 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0xb53b4680 | 0x246c000 | +| 255 | blk.21.attn_k.bias | 0xb7820680 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xb7820e80 | 0xfc000 | +| 257 | blk.21.attn_norm.weight | 0xb791ce80 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xb7920680 | 0x6e4000 | +| 259 | blk.21.attn_q.bias | 0xb8004680 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xb8007e80 | 0x6e4000 | +| 261 | blk.21.attn_v.bias | 0xb86ebe80 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xb86ec680 | 0xfc000 | +| 263 | blk.21.ffn_down.weight | 0xb87e8680 | 0x2c84000 | +| 264 | blk.21.ffn_gate.weight | 0xbb46c680 | 0x246c000 | +| 265 | blk.21.ffn_norm.weight | 0xbd8d8680 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xbd8dbe80 | 0x246c000 | +| 267 | blk.22.attn_k.bias | 0xbfd47e80 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xbfd48680 | 0xfc000 | +| 269 | blk.22.attn_norm.weight | 0xbfe44680 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xbfe47e80 | 0x6e4000 | +| 271 | blk.22.attn_q.bias | 0xc052be80 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xc052f680 | 0x6e4000 | +| 273 | blk.22.attn_v.bias | 0xc0c13680 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xc0c13e80 | 0xfc000 | +| 275 | blk.22.ffn_down.weight | 0xc0d0fe80 | 0x2c84000 | +| 276 | blk.22.ffn_gate.weight | 0xc3993e80 | 0x246c000 | +| 277 | blk.22.ffn_norm.weight | 0xc5dffe80 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xc5e03680 | 0x246c000 | +| 279 | blk.23.attn_k.bias | 0xc826f680 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xc826fe80 | 0xfc000 | +| 281 | blk.23.attn_norm.weight | 0xc836be80 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xc836f680 | 0x6e4000 | +| 283 | blk.23.attn_q.bias | 0xc8a53680 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xc8a56e80 | 0x6e4000 | +| 285 | blk.23.attn_v.bias | 0xc913ae80 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xc913b680 | 0xfc000 | +| 287 | blk.23.ffn_down.weight | 0xc9237680 | 0x2c84000 | +| 288 | blk.23.ffn_gate.weight | 0xcbebb680 | 0x246c000 | +| 289 | blk.23.ffn_norm.weight | 0xce327680 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xce32ae80 | 0x246c000 | +| 291 | blk.24.attn_k.bias | 0xd0796e80 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xd0797680 | 0xfc000 | +| 293 | blk.24.attn_norm.weight | 0xd0893680 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xd0896e80 | 0x6e4000 | +| 295 | blk.24.attn_q.bias | 0xd0f7ae80 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0xd0f7e680 | 0x6e4000 | +| 297 | blk.24.attn_v.bias | 0xd1662680 | 0x800 | +| 298 | blk.24.attn_v.weight | 0xd1662e80 | 0xfc000 | +| 299 | blk.24.ffn_down.weight | 0xd175ee80 | 0x2c84000 | +| 300 | blk.24.ffn_gate.weight | 0xd43e2e80 | 0x246c000 | +| 301 | blk.24.ffn_norm.weight | 0xd684ee80 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0xd6852680 | 0x246c000 | +| 303 | blk.25.attn_k.bias | 0xd8cbe680 | 0x800 | +| 304 | blk.25.attn_k.weight | 0xd8cbee80 | 0xfc000 | +| 305 | blk.25.attn_norm.weight | 0xd8dbae80 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0xd8dbe680 | 0x6e4000 | +| 307 | blk.25.attn_q.bias | 0xd94a2680 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0xd94a5e80 | 0x6e4000 | +| 309 | blk.25.attn_v.bias | 0xd9b89e80 | 0x800 | +| 310 | blk.25.attn_v.weight | 0xd9b8a680 | 0xfc000 | +| 311 | blk.25.ffn_down.weight | 0xd9c86680 | 0x2c84000 | +| 312 | blk.25.ffn_gate.weight | 0xdc90a680 | 0x246c000 | +| 313 | blk.25.ffn_norm.weight | 0xded76680 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0xded79e80 | 0x246c000 | +| 315 | blk.26.attn_k.bias | 0xe11e5e80 | 0x800 | +| 316 | blk.26.attn_k.weight | 0xe11e6680 | 0xfc000 | +| 317 | blk.26.attn_norm.weight | 0xe12e2680 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0xe12e5e80 | 0x6e4000 | +| 319 | blk.26.attn_q.bias | 0xe19c9e80 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0xe19cd680 | 0x6e4000 | +| 321 | blk.26.attn_v.bias | 0xe20b1680 | 0x800 | +| 322 | blk.26.attn_v.weight | 0xe20b1e80 | 0xfc000 | +| 323 | blk.26.ffn_down.weight | 0xe21ade80 | 0x2c84000 | +| 324 | blk.26.ffn_gate.weight | 0xe4e31e80 | 0x246c000 | +| 325 | blk.26.ffn_norm.weight | 0xe729de80 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0xe72a1680 | 0x246c000 | +| 327 | blk.27.attn_k.bias | 0xe970d680 | 0x800 | +| 328 | blk.27.attn_k.weight | 0xe970de80 | 0xfc000 | +| 329 | blk.27.attn_norm.weight | 0xe9809e80 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0xe980d680 | 0x6e4000 | +| 331 | blk.27.attn_q.bias | 0xe9ef1680 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0xe9ef4e80 | 0x6e4000 | +| 333 | blk.27.attn_v.bias | 0xea5d8e80 | 0x800 | +| 334 | blk.27.attn_v.weight | 0xea5d9680 | 0xfc000 | +| 335 | blk.27.ffn_down.weight | 0xea6d5680 | 0x2c84000 | +| 336 | blk.27.ffn_gate.weight | 0xed359680 | 0x246c000 | +| 337 | blk.27.ffn_norm.weight | 0xef7c5680 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0xef7c8e80 | 0x246c000 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q4_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q3_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q3_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q4_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q3_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q5_K_M.md b/scores/Hammer2.1-7b-Q5_K_M.md new file mode 100644 index 0000000..250a53a --- /dev/null +++ b/scores/Hammer2.1-7b-Q5_K_M.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q5_K_M.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 17 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q5\_K\_M.gguf - GGUF Internal File Dump](#hammer21-7b-q5_k_mgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x16463fa0 | +| 1 | output_norm.weight | 0x16a0f460 | 0x3800 | +| 2 | token_embd.weight | 0x16a12c60 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x248d1440 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x248d1c40 | 0xfc000 | +| 5 | blk.0.attn_norm.weight | 0x249cdc40 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x249d1440 | 0x86c000 | +| 7 | blk.0.attn_q.bias | 0x2523d440 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x25240c40 | 0x6e4000 | +| 9 | blk.0.attn_v.bias | 0x25924c40 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x25925440 | 0x134000 | +| 11 | blk.0.ffn_down.weight | 0x25a59440 | 0x351d800 | +| 12 | blk.0.ffn_gate.weight | 0x28f76c40 | 0x246c000 | +| 13 | blk.0.ffn_norm.weight | 0x2b3e2c40 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x2b3e6440 | 0x246c000 | +| 15 | blk.1.attn_k.bias | 0x2d852440 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x2d852c40 | 0xfc000 | +| 17 | blk.1.attn_norm.weight | 0x2d94ec40 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x2d952440 | 0x86c000 | +| 19 | blk.1.attn_q.bias | 0x2e1be440 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x2e1c1c40 | 0x6e4000 | +| 21 | blk.1.attn_v.bias | 0x2e8a5c40 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x2e8a6440 | 0x134000 | +| 23 | blk.1.ffn_down.weight | 0x2e9da440 | 0x351d800 | +| 24 | blk.1.ffn_gate.weight | 0x31ef7c40 | 0x2c84000 | +| 25 | blk.1.ffn_norm.weight | 0x34b7bc40 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x34b7f440 | 0x2c84000 | +| 27 | blk.2.attn_k.bias | 0x37803440 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x37803c40 | 0xfc000 | +| 29 | blk.2.attn_norm.weight | 0x378ffc40 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x37903440 | 0x86c000 | +| 31 | blk.2.attn_q.bias | 0x3816f440 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x38172c40 | 0x6e4000 | +| 33 | blk.2.attn_v.bias | 0x38856c40 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x38857440 | 0x134000 | +| 35 | blk.2.ffn_down.weight | 0x3898b440 | 0x351d800 | +| 36 | blk.2.ffn_gate.weight | 0x3bea8c40 | 0x2c84000 | +| 37 | blk.2.ffn_norm.weight | 0x3eb2cc40 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x3eb30440 | 0x2c84000 | +| 39 | blk.3.attn_k.bias | 0x417b4440 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x417b4c40 | 0xfc000 | +| 41 | blk.3.attn_norm.weight | 0x418b0c40 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x418b4440 | 0x86c000 | +| 43 | blk.3.attn_q.bias | 0x42120440 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x42123c40 | 0x6e4000 | +| 45 | blk.3.attn_v.bias | 0x42807c40 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x42808440 | 0x134000 | +| 47 | blk.3.ffn_down.weight | 0x4293c440 | 0x351d800 | +| 48 | blk.3.ffn_gate.weight | 0x45e59c40 | 0x2c84000 | +| 49 | blk.3.ffn_norm.weight | 0x48addc40 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x48ae1440 | 0x2c84000 | +| 51 | blk.4.attn_k.bias | 0x4b765440 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x4b765c40 | 0xfc000 | +| 53 | blk.4.attn_norm.weight | 0x4b861c40 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x4b865440 | 0x86c000 | +| 55 | blk.4.attn_q.bias | 0x4c0d1440 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x4c0d4c40 | 0x6e4000 | +| 57 | blk.4.attn_v.bias | 0x4c7b8c40 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x4c7b9440 | 0x134000 | +| 59 | blk.4.ffn_down.weight | 0x4c8ed440 | 0x351d800 | +| 60 | blk.4.ffn_gate.weight | 0x4fe0ac40 | 0x2c84000 | +| 61 | blk.4.ffn_norm.weight | 0x52a8ec40 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x52a92440 | 0x2c84000 | +| 63 | blk.5.attn_k.bias | 0x55716440 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x55716c40 | 0xfc000 | +| 65 | blk.5.attn_norm.weight | 0x55812c40 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x55816440 | 0x86c000 | +| 67 | blk.5.attn_q.bias | 0x56082440 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x56085c40 | 0x6e4000 | +| 69 | blk.5.attn_v.bias | 0x56769c40 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x5676a440 | 0x134000 | +| 71 | blk.5.ffn_down.weight | 0x5689e440 | 0x351d800 | +| 72 | blk.5.ffn_gate.weight | 0x59dbbc40 | 0x2c84000 | +| 73 | blk.5.ffn_norm.weight | 0x5ca3fc40 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x5ca43440 | 0x2c84000 | +| 75 | blk.6.attn_k.bias | 0x5f6c7440 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x5f6c7c40 | 0xfc000 | +| 77 | blk.6.attn_norm.weight | 0x5f7c3c40 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x5f7c7440 | 0x86c000 | +| 79 | blk.6.attn_q.bias | 0x60033440 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x60036c40 | 0x6e4000 | +| 81 | blk.6.attn_v.bias | 0x6071ac40 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x6071b440 | 0x134000 | +| 83 | blk.6.ffn_down.weight | 0x6084f440 | 0x351d800 | +| 84 | blk.6.ffn_gate.weight | 0x63d6cc40 | 0x246c000 | +| 85 | blk.6.ffn_norm.weight | 0x661d8c40 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x661dc440 | 0x246c000 | +| 87 | blk.7.attn_k.bias | 0x68648440 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x68648c40 | 0xfc000 | +| 89 | blk.7.attn_norm.weight | 0x68744c40 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x68748440 | 0x86c000 | +| 91 | blk.7.attn_q.bias | 0x68fb4440 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x68fb7c40 | 0x6e4000 | +| 93 | blk.7.attn_v.bias | 0x6969bc40 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x6969c440 | 0x134000 | +| 95 | blk.7.ffn_down.weight | 0x697d0440 | 0x351d800 | +| 96 | blk.7.ffn_gate.weight | 0x6ccedc40 | 0x246c000 | +| 97 | blk.7.ffn_norm.weight | 0x6f159c40 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x6f15d440 | 0x246c000 | +| 99 | blk.8.attn_k.bias | 0x715c9440 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x715c9c40 | 0xfc000 | +| 101 | blk.8.attn_norm.weight | 0x716c5c40 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x716c9440 | 0x86c000 | +| 103 | blk.8.attn_q.bias | 0x71f35440 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x71f38c40 | 0x6e4000 | +| 105 | blk.8.attn_v.bias | 0x7261cc40 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x7261d440 | 0x134000 | +| 107 | blk.8.ffn_down.weight | 0x72751440 | 0x351d800 | +| 108 | blk.8.ffn_gate.weight | 0x75c6ec40 | 0x246c000 | +| 109 | blk.8.ffn_norm.weight | 0x780dac40 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x780de440 | 0x246c000 | +| 111 | blk.9.attn_k.bias | 0x7a54a440 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x7a54ac40 | 0xfc000 | +| 113 | blk.9.attn_norm.weight | 0x7a646c40 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x7a64a440 | 0x86c000 | +| 115 | blk.9.attn_q.bias | 0x7aeb6440 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x7aeb9c40 | 0x6e4000 | +| 117 | blk.9.attn_v.bias | 0x7b59dc40 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x7b59e440 | 0x134000 | +| 119 | blk.9.ffn_down.weight | 0x7b6d2440 | 0x351d800 | +| 120 | blk.9.ffn_gate.weight | 0x7ebefc40 | 0x2c84000 | +| 121 | blk.9.ffn_norm.weight | 0x81873c40 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x81877440 | 0x2c84000 | +| 123 | blk.10.attn_k.bias | 0x844fb440 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x844fbc40 | 0xfc000 | +| 125 | blk.10.attn_norm.weight | 0x845f7c40 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x845fb440 | 0x86c000 | +| 127 | blk.10.attn_q.bias | 0x84e67440 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x84e6ac40 | 0x6e4000 | +| 129 | blk.10.attn_v.bias | 0x8554ec40 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x8554f440 | 0x134000 | +| 131 | blk.10.ffn_down.weight | 0x85683440 | 0x351d800 | +| 132 | blk.10.ffn_gate.weight | 0x88ba0c40 | 0x246c000 | +| 133 | blk.10.ffn_norm.weight | 0x8b00cc40 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x8b010440 | 0x246c000 | +| 135 | blk.11.attn_k.bias | 0x8d47c440 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x8d47cc40 | 0xfc000 | +| 137 | blk.11.attn_norm.weight | 0x8d578c40 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x8d57c440 | 0x86c000 | +| 139 | blk.11.attn_q.bias | 0x8dde8440 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x8ddebc40 | 0x6e4000 | +| 141 | blk.11.attn_v.bias | 0x8e4cfc40 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x8e4d0440 | 0x134000 | +| 143 | blk.11.ffn_down.weight | 0x8e604440 | 0x351d800 | +| 144 | blk.11.ffn_gate.weight | 0x91b21c40 | 0x246c000 | +| 145 | blk.11.ffn_norm.weight | 0x93f8dc40 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x93f91440 | 0x246c000 | +| 147 | blk.12.attn_k.bias | 0x963fd440 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x963fdc40 | 0xfc000 | +| 149 | blk.12.attn_norm.weight | 0x964f9c40 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x964fd440 | 0x86c000 | +| 151 | blk.12.attn_q.bias | 0x96d69440 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x96d6cc40 | 0x6e4000 | +| 153 | blk.12.attn_v.bias | 0x97450c40 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x97451440 | 0x134000 | +| 155 | blk.12.ffn_down.weight | 0x97585440 | 0x351d800 | +| 156 | blk.12.ffn_gate.weight | 0x9aaa2c40 | 0x246c000 | +| 157 | blk.12.ffn_norm.weight | 0x9cf0ec40 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x9cf12440 | 0x246c000 | +| 159 | blk.13.attn_k.bias | 0x9f37e440 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x9f37ec40 | 0xfc000 | +| 161 | blk.13.attn_norm.weight | 0x9f47ac40 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x9f47e440 | 0x86c000 | +| 163 | blk.13.attn_q.bias | 0x9fcea440 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x9fcedc40 | 0x6e4000 | +| 165 | blk.13.attn_v.bias | 0xa03d1c40 | 0x800 | +| 166 | blk.13.attn_v.weight | 0xa03d2440 | 0x134000 | +| 167 | blk.13.ffn_down.weight | 0xa0506440 | 0x351d800 | +| 168 | blk.13.ffn_gate.weight | 0xa3a23c40 | 0x246c000 | +| 169 | blk.13.ffn_norm.weight | 0xa5e8fc40 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0xa5e93440 | 0x246c000 | +| 171 | blk.14.attn_k.bias | 0xa82ff440 | 0x800 | +| 172 | blk.14.attn_k.weight | 0xa82ffc40 | 0x134000 | +| 173 | blk.14.attn_norm.weight | 0xa8433c40 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0xa8437440 | 0x86c000 | +| 175 | blk.14.attn_q.bias | 0xa8ca3440 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0xa8ca6c40 | 0x86c000 | +| 177 | blk.14.attn_v.bias | 0xa9512c40 | 0x800 | +| 178 | blk.14.attn_v.weight | 0xa9513440 | 0x16f800 | +| 179 | blk.14.ffn_down.weight | 0xa9682c40 | 0x351d800 | +| 180 | blk.14.ffn_gate.weight | 0xacba0440 | 0x246c000 | +| 181 | blk.14.ffn_norm.weight | 0xaf00c440 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0xaf00fc40 | 0x246c000 | +| 183 | blk.15.attn_k.bias | 0xb147bc40 | 0x800 | +| 184 | blk.15.attn_k.weight | 0xb147c440 | 0x134000 | +| 185 | blk.15.attn_norm.weight | 0xb15b0440 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0xb15b3c40 | 0x86c000 | +| 187 | blk.15.attn_q.bias | 0xb1e1fc40 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0xb1e23440 | 0x86c000 | +| 189 | blk.15.attn_v.bias | 0xb268f440 | 0x800 | +| 190 | blk.15.attn_v.weight | 0xb268fc40 | 0x16f800 | +| 191 | blk.15.ffn_down.weight | 0xb27ff440 | 0x351d800 | +| 192 | blk.15.ffn_gate.weight | 0xb5d1cc40 | 0x246c000 | +| 193 | blk.15.ffn_norm.weight | 0xb8188c40 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0xb818c440 | 0x246c000 | +| 195 | blk.16.attn_k.bias | 0xba5f8440 | 0x800 | +| 196 | blk.16.attn_k.weight | 0xba5f8c40 | 0x134000 | +| 197 | blk.16.attn_norm.weight | 0xba72cc40 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0xba730440 | 0x86c000 | +| 199 | blk.16.attn_q.bias | 0xbaf9c440 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0xbaf9fc40 | 0x86c000 | +| 201 | blk.16.attn_v.bias | 0xbb80bc40 | 0x800 | +| 202 | blk.16.attn_v.weight | 0xbb80c440 | 0x16f800 | +| 203 | blk.16.ffn_down.weight | 0xbb97bc40 | 0x351d800 | +| 204 | blk.16.ffn_gate.weight | 0xbee99440 | 0x246c000 | +| 205 | blk.16.ffn_norm.weight | 0xc1305440 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0xc1308c40 | 0x246c000 | +| 207 | blk.17.attn_k.bias | 0xc3774c40 | 0x800 | +| 208 | blk.17.attn_k.weight | 0xc3775440 | 0x134000 | +| 209 | blk.17.attn_norm.weight | 0xc38a9440 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0xc38acc40 | 0x86c000 | +| 211 | blk.17.attn_q.bias | 0xc4118c40 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0xc411c440 | 0x86c000 | +| 213 | blk.17.attn_v.bias | 0xc4988440 | 0x800 | +| 214 | blk.17.attn_v.weight | 0xc4988c40 | 0x16f800 | +| 215 | blk.17.ffn_down.weight | 0xc4af8440 | 0x351d800 | +| 216 | blk.17.ffn_gate.weight | 0xc8015c40 | 0x246c000 | +| 217 | blk.17.ffn_norm.weight | 0xca481c40 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0xca485440 | 0x246c000 | +| 219 | blk.18.attn_k.bias | 0xcc8f1440 | 0x800 | +| 220 | blk.18.attn_k.weight | 0xcc8f1c40 | 0x134000 | +| 221 | blk.18.attn_norm.weight | 0xcca25c40 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0xcca29440 | 0x86c000 | +| 223 | blk.18.attn_q.bias | 0xcd295440 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0xcd298c40 | 0x86c000 | +| 225 | blk.18.attn_v.bias | 0xcdb04c40 | 0x800 | +| 226 | blk.18.attn_v.weight | 0xcdb05440 | 0x16f800 | +| 227 | blk.18.ffn_down.weight | 0xcdc74c40 | 0x351d800 | +| 228 | blk.18.ffn_gate.weight | 0xd1192440 | 0x246c000 | +| 229 | blk.18.ffn_norm.weight | 0xd35fe440 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0xd3601c40 | 0x246c000 | +| 231 | blk.19.attn_k.bias | 0xd5a6dc40 | 0x800 | +| 232 | blk.19.attn_k.weight | 0xd5a6e440 | 0x134000 | +| 233 | blk.19.attn_norm.weight | 0xd5ba2440 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0xd5ba5c40 | 0x86c000 | +| 235 | blk.19.attn_q.bias | 0xd6411c40 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0xd6415440 | 0x86c000 | +| 237 | blk.19.attn_v.bias | 0xd6c81440 | 0x800 | +| 238 | blk.19.attn_v.weight | 0xd6c81c40 | 0x16f800 | +| 239 | blk.19.ffn_down.weight | 0xd6df1440 | 0x351d800 | +| 240 | blk.19.ffn_gate.weight | 0xda30ec40 | 0x246c000 | +| 241 | blk.19.ffn_norm.weight | 0xdc77ac40 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0xdc77e440 | 0x246c000 | +| 243 | blk.20.attn_k.bias | 0xdebea440 | 0x800 | +| 244 | blk.20.attn_k.weight | 0xdebeac40 | 0x134000 | +| 245 | blk.20.attn_norm.weight | 0xded1ec40 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0xded22440 | 0x86c000 | +| 247 | blk.20.attn_q.bias | 0xdf58e440 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0xdf591c40 | 0x86c000 | +| 249 | blk.20.attn_v.bias | 0xdfdfdc40 | 0x800 | +| 250 | blk.20.attn_v.weight | 0xdfdfe440 | 0x16f800 | +| 251 | blk.20.ffn_down.weight | 0xdff6dc40 | 0x351d800 | +| 252 | blk.20.ffn_gate.weight | 0xe348b440 | 0x2c84000 | +| 253 | blk.20.ffn_norm.weight | 0xe610f440 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0xe6112c40 | 0x2c84000 | +| 255 | blk.21.attn_k.bias | 0xe8d96c40 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xe8d97440 | 0x134000 | +| 257 | blk.21.attn_norm.weight | 0xe8ecb440 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xe8ecec40 | 0x86c000 | +| 259 | blk.21.attn_q.bias | 0xe973ac40 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xe973e440 | 0x86c000 | +| 261 | blk.21.attn_v.bias | 0xe9faa440 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xe9faac40 | 0x16f800 | +| 263 | blk.21.ffn_down.weight | 0xea11a440 | 0x351d800 | +| 264 | blk.21.ffn_gate.weight | 0xed637c40 | 0x2c84000 | +| 265 | blk.21.ffn_norm.weight | 0xf02bbc40 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xf02bf440 | 0x2c84000 | +| 267 | blk.22.attn_k.bias | 0xf2f43440 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xf2f43c40 | 0x134000 | +| 269 | blk.22.attn_norm.weight | 0xf3077c40 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xf307b440 | 0x86c000 | +| 271 | blk.22.attn_q.bias | 0xf38e7440 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xf38eac40 | 0x86c000 | +| 273 | blk.22.attn_v.bias | 0xf4156c40 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xf4157440 | 0x16f800 | +| 275 | blk.22.ffn_down.weight | 0xf42c6c40 | 0x351d800 | +| 276 | blk.22.ffn_gate.weight | 0xf77e4440 | 0x2c84000 | +| 277 | blk.22.ffn_norm.weight | 0xfa468440 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xfa46bc40 | 0x2c84000 | +| 279 | blk.23.attn_k.bias | 0xfd0efc40 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xfd0f0440 | 0x134000 | +| 281 | blk.23.attn_norm.weight | 0xfd224440 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xfd227c40 | 0x86c000 | +| 283 | blk.23.attn_q.bias | 0xfda93c40 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xfda97440 | 0x86c000 | +| 285 | blk.23.attn_v.bias | 0xfe303440 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xfe303c40 | 0x16f800 | +| 287 | blk.23.ffn_down.weight | 0xfe473440 | 0x351d800 | +| 288 | blk.23.ffn_gate.weight | 0x101990c40 | 0x2c84000 | +| 289 | blk.23.ffn_norm.weight | 0x104614c40 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0x104618440 | 0x2c84000 | +| 291 | blk.24.attn_k.bias | 0x10729c440 | 0x800 | +| 292 | blk.24.attn_k.weight | 0x10729cc40 | 0x134000 | +| 293 | blk.24.attn_norm.weight | 0x1073d0c40 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0x1073d4440 | 0x86c000 | +| 295 | blk.24.attn_q.bias | 0x107c40440 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0x107c43c40 | 0x86c000 | +| 297 | blk.24.attn_v.bias | 0x1084afc40 | 0x800 | +| 298 | blk.24.attn_v.weight | 0x1084b0440 | 0x16f800 | +| 299 | blk.24.ffn_down.weight | 0x10861fc40 | 0x351d800 | +| 300 | blk.24.ffn_gate.weight | 0x10bb3d440 | 0x2c84000 | +| 301 | blk.24.ffn_norm.weight | 0x10e7c1440 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0x10e7c4c40 | 0x2c84000 | +| 303 | blk.25.attn_k.bias | 0x111448c40 | 0x800 | +| 304 | blk.25.attn_k.weight | 0x111449440 | 0x134000 | +| 305 | blk.25.attn_norm.weight | 0x11157d440 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0x111580c40 | 0x86c000 | +| 307 | blk.25.attn_q.bias | 0x111decc40 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0x111df0440 | 0x86c000 | +| 309 | blk.25.attn_v.bias | 0x11265c440 | 0x800 | +| 310 | blk.25.attn_v.weight | 0x11265cc40 | 0x16f800 | +| 311 | blk.25.ffn_down.weight | 0x1127cc440 | 0x351d800 | +| 312 | blk.25.ffn_gate.weight | 0x115ce9c40 | 0x2c84000 | +| 313 | blk.25.ffn_norm.weight | 0x11896dc40 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0x118971440 | 0x2c84000 | +| 315 | blk.26.attn_k.bias | 0x11b5f5440 | 0x800 | +| 316 | blk.26.attn_k.weight | 0x11b5f5c40 | 0x134000 | +| 317 | blk.26.attn_norm.weight | 0x11b729c40 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0x11b72d440 | 0x86c000 | +| 319 | blk.26.attn_q.bias | 0x11bf99440 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0x11bf9cc40 | 0x86c000 | +| 321 | blk.26.attn_v.bias | 0x11c808c40 | 0x800 | +| 322 | blk.26.attn_v.weight | 0x11c809440 | 0x16f800 | +| 323 | blk.26.ffn_down.weight | 0x11c978c40 | 0x351d800 | +| 324 | blk.26.ffn_gate.weight | 0x11fe96440 | 0x2c84000 | +| 325 | blk.26.ffn_norm.weight | 0x122b1a440 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0x122b1dc40 | 0x2c84000 | +| 327 | blk.27.attn_k.bias | 0x1257a1c40 | 0x800 | +| 328 | blk.27.attn_k.weight | 0x1257a2440 | 0x134000 | +| 329 | blk.27.attn_norm.weight | 0x1258d6440 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0x1258d9c40 | 0x86c000 | +| 331 | blk.27.attn_q.bias | 0x126145c40 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0x126149440 | 0x86c000 | +| 333 | blk.27.attn_v.bias | 0x1269b5440 | 0x800 | +| 334 | blk.27.attn_v.weight | 0x1269b5c40 | 0x16f800 | +| 335 | blk.27.ffn_down.weight | 0x126b25440 | 0x351d800 | +| 336 | blk.27.ffn_gate.weight | 0x12a042c40 | 0x2c84000 | +| 337 | blk.27.ffn_norm.weight | 0x12ccc6c40 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0x12ccca440 | 0x2c84000 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q5_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q5_K_S.md b/scores/Hammer2.1-7b-Q5_K_S.md new file mode 100644 index 0000000..81e6f7d --- /dev/null +++ b/scores/Hammer2.1-7b-Q5_K_S.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q5_K_S.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 16 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q5\_K\_S.gguf - GGUF Internal File Dump](#hammer21-7b-q5_k_sgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x16463fa0 | +| 1 | output_norm.weight | 0x16a0f460 | 0x3800 | +| 2 | token_embd.weight | 0x16a12c60 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x248d1440 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x248d1c40 | 0xfc000 | +| 5 | blk.0.attn_norm.weight | 0x249cdc40 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x249d1440 | 0x86c000 | +| 7 | blk.0.attn_q.bias | 0x2523d440 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x25240c40 | 0x6e4000 | +| 9 | blk.0.attn_v.bias | 0x25924c40 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x25925440 | 0x134000 | +| 11 | blk.0.ffn_down.weight | 0x25a59440 | 0x2c84000 | +| 12 | blk.0.ffn_gate.weight | 0x286dd440 | 0x246c000 | +| 13 | blk.0.ffn_norm.weight | 0x2ab49440 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x2ab4cc40 | 0x246c000 | +| 15 | blk.1.attn_k.bias | 0x2cfb8c40 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x2cfb9440 | 0xfc000 | +| 17 | blk.1.attn_norm.weight | 0x2d0b5440 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x2d0b8c40 | 0x86c000 | +| 19 | blk.1.attn_q.bias | 0x2d924c40 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x2d928440 | 0x6e4000 | +| 21 | blk.1.attn_v.bias | 0x2e00c440 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x2e00cc40 | 0x134000 | +| 23 | blk.1.ffn_down.weight | 0x2e140c40 | 0x2c84000 | +| 24 | blk.1.ffn_gate.weight | 0x30dc4c40 | 0x2c84000 | +| 25 | blk.1.ffn_norm.weight | 0x33a48c40 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x33a4c440 | 0x2c84000 | +| 27 | blk.2.attn_k.bias | 0x366d0440 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x366d0c40 | 0xfc000 | +| 29 | blk.2.attn_norm.weight | 0x367ccc40 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x367d0440 | 0x86c000 | +| 31 | blk.2.attn_q.bias | 0x3703c440 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x3703fc40 | 0x6e4000 | +| 33 | blk.2.attn_v.bias | 0x37723c40 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x37724440 | 0x134000 | +| 35 | blk.2.ffn_down.weight | 0x37858440 | 0x2c84000 | +| 36 | blk.2.ffn_gate.weight | 0x3a4dc440 | 0x2c84000 | +| 37 | blk.2.ffn_norm.weight | 0x3d160440 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x3d163c40 | 0x2c84000 | +| 39 | blk.3.attn_k.bias | 0x3fde7c40 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x3fde8440 | 0xfc000 | +| 41 | blk.3.attn_norm.weight | 0x3fee4440 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x3fee7c40 | 0x86c000 | +| 43 | blk.3.attn_q.bias | 0x40753c40 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x40757440 | 0x6e4000 | +| 45 | blk.3.attn_v.bias | 0x40e3b440 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x40e3bc40 | 0x134000 | +| 47 | blk.3.ffn_down.weight | 0x40f6fc40 | 0x351d800 | +| 48 | blk.3.ffn_gate.weight | 0x4448d440 | 0x2c84000 | +| 49 | blk.3.ffn_norm.weight | 0x47111440 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x47114c40 | 0x2c84000 | +| 51 | blk.4.attn_k.bias | 0x49d98c40 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x49d99440 | 0xfc000 | +| 53 | blk.4.attn_norm.weight | 0x49e95440 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x49e98c40 | 0x86c000 | +| 55 | blk.4.attn_q.bias | 0x4a704c40 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x4a708440 | 0x6e4000 | +| 57 | blk.4.attn_v.bias | 0x4adec440 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x4adecc40 | 0x134000 | +| 59 | blk.4.ffn_down.weight | 0x4af20c40 | 0x2c84000 | +| 60 | blk.4.ffn_gate.weight | 0x4dba4c40 | 0x2c84000 | +| 61 | blk.4.ffn_norm.weight | 0x50828c40 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x5082c440 | 0x2c84000 | +| 63 | blk.5.attn_k.bias | 0x534b0440 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x534b0c40 | 0xfc000 | +| 65 | blk.5.attn_norm.weight | 0x535acc40 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x535b0440 | 0x86c000 | +| 67 | blk.5.attn_q.bias | 0x53e1c440 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x53e1fc40 | 0x6e4000 | +| 69 | blk.5.attn_v.bias | 0x54503c40 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x54504440 | 0x134000 | +| 71 | blk.5.ffn_down.weight | 0x54638440 | 0x2c84000 | +| 72 | blk.5.ffn_gate.weight | 0x572bc440 | 0x2c84000 | +| 73 | blk.5.ffn_norm.weight | 0x59f40440 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x59f43c40 | 0x2c84000 | +| 75 | blk.6.attn_k.bias | 0x5cbc7c40 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x5cbc8440 | 0xfc000 | +| 77 | blk.6.attn_norm.weight | 0x5ccc4440 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x5ccc7c40 | 0x86c000 | +| 79 | blk.6.attn_q.bias | 0x5d533c40 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x5d537440 | 0x6e4000 | +| 81 | blk.6.attn_v.bias | 0x5dc1b440 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x5dc1bc40 | 0x134000 | +| 83 | blk.6.ffn_down.weight | 0x5dd4fc40 | 0x2c84000 | +| 84 | blk.6.ffn_gate.weight | 0x609d3c40 | 0x246c000 | +| 85 | blk.6.ffn_norm.weight | 0x62e3fc40 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x62e43440 | 0x246c000 | +| 87 | blk.7.attn_k.bias | 0x652af440 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x652afc40 | 0xfc000 | +| 89 | blk.7.attn_norm.weight | 0x653abc40 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x653af440 | 0x86c000 | +| 91 | blk.7.attn_q.bias | 0x65c1b440 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x65c1ec40 | 0x6e4000 | +| 93 | blk.7.attn_v.bias | 0x66302c40 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x66303440 | 0x134000 | +| 95 | blk.7.ffn_down.weight | 0x66437440 | 0x351d800 | +| 96 | blk.7.ffn_gate.weight | 0x69954c40 | 0x246c000 | +| 97 | blk.7.ffn_norm.weight | 0x6bdc0c40 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x6bdc4440 | 0x246c000 | +| 99 | blk.8.attn_k.bias | 0x6e230440 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x6e230c40 | 0xfc000 | +| 101 | blk.8.attn_norm.weight | 0x6e32cc40 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x6e330440 | 0x86c000 | +| 103 | blk.8.attn_q.bias | 0x6eb9c440 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x6eb9fc40 | 0x6e4000 | +| 105 | blk.8.attn_v.bias | 0x6f283c40 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x6f284440 | 0x134000 | +| 107 | blk.8.ffn_down.weight | 0x6f3b8440 | 0x2c84000 | +| 108 | blk.8.ffn_gate.weight | 0x7203c440 | 0x246c000 | +| 109 | blk.8.ffn_norm.weight | 0x744a8440 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x744abc40 | 0x246c000 | +| 111 | blk.9.attn_k.bias | 0x76917c40 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x76918440 | 0xfc000 | +| 113 | blk.9.attn_norm.weight | 0x76a14440 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x76a17c40 | 0x86c000 | +| 115 | blk.9.attn_q.bias | 0x77283c40 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x77287440 | 0x6e4000 | +| 117 | blk.9.attn_v.bias | 0x7796b440 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x7796bc40 | 0x134000 | +| 119 | blk.9.ffn_down.weight | 0x77a9fc40 | 0x351d800 | +| 120 | blk.9.ffn_gate.weight | 0x7afbd440 | 0x2c84000 | +| 121 | blk.9.ffn_norm.weight | 0x7dc41440 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x7dc44c40 | 0x2c84000 | +| 123 | blk.10.attn_k.bias | 0x808c8c40 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x808c9440 | 0xfc000 | +| 125 | blk.10.attn_norm.weight | 0x809c5440 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x809c8c40 | 0x86c000 | +| 127 | blk.10.attn_q.bias | 0x81234c40 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x81238440 | 0x6e4000 | +| 129 | blk.10.attn_v.bias | 0x8191c440 | 0x800 | +| 130 | blk.10.attn_v.weight | 0x8191cc40 | 0x134000 | +| 131 | blk.10.ffn_down.weight | 0x81a50c40 | 0x351d800 | +| 132 | blk.10.ffn_gate.weight | 0x84f6e440 | 0x246c000 | +| 133 | blk.10.ffn_norm.weight | 0x873da440 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0x873ddc40 | 0x246c000 | +| 135 | blk.11.attn_k.bias | 0x89849c40 | 0x800 | +| 136 | blk.11.attn_k.weight | 0x8984a440 | 0xfc000 | +| 137 | blk.11.attn_norm.weight | 0x89946440 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0x89949c40 | 0x86c000 | +| 139 | blk.11.attn_q.bias | 0x8a1b5c40 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0x8a1b9440 | 0x6e4000 | +| 141 | blk.11.attn_v.bias | 0x8a89d440 | 0x800 | +| 142 | blk.11.attn_v.weight | 0x8a89dc40 | 0x134000 | +| 143 | blk.11.ffn_down.weight | 0x8a9d1c40 | 0x2c84000 | +| 144 | blk.11.ffn_gate.weight | 0x8d655c40 | 0x246c000 | +| 145 | blk.11.ffn_norm.weight | 0x8fac1c40 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0x8fac5440 | 0x246c000 | +| 147 | blk.12.attn_k.bias | 0x91f31440 | 0x800 | +| 148 | blk.12.attn_k.weight | 0x91f31c40 | 0xfc000 | +| 149 | blk.12.attn_norm.weight | 0x9202dc40 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0x92031440 | 0x86c000 | +| 151 | blk.12.attn_q.bias | 0x9289d440 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0x928a0c40 | 0x6e4000 | +| 153 | blk.12.attn_v.bias | 0x92f84c40 | 0x800 | +| 154 | blk.12.attn_v.weight | 0x92f85440 | 0x134000 | +| 155 | blk.12.ffn_down.weight | 0x930b9440 | 0x2c84000 | +| 156 | blk.12.ffn_gate.weight | 0x95d3d440 | 0x246c000 | +| 157 | blk.12.ffn_norm.weight | 0x981a9440 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0x981acc40 | 0x246c000 | +| 159 | blk.13.attn_k.bias | 0x9a618c40 | 0x800 | +| 160 | blk.13.attn_k.weight | 0x9a619440 | 0xfc000 | +| 161 | blk.13.attn_norm.weight | 0x9a715440 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0x9a718c40 | 0x86c000 | +| 163 | blk.13.attn_q.bias | 0x9af84c40 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0x9af88440 | 0x6e4000 | +| 165 | blk.13.attn_v.bias | 0x9b66c440 | 0x800 | +| 166 | blk.13.attn_v.weight | 0x9b66cc40 | 0x134000 | +| 167 | blk.13.ffn_down.weight | 0x9b7a0c40 | 0x2c84000 | +| 168 | blk.13.ffn_gate.weight | 0x9e424c40 | 0x246c000 | +| 169 | blk.13.ffn_norm.weight | 0xa0890c40 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0xa0894440 | 0x246c000 | +| 171 | blk.14.attn_k.bias | 0xa2d00440 | 0x800 | +| 172 | blk.14.attn_k.weight | 0xa2d00c40 | 0x134000 | +| 173 | blk.14.attn_norm.weight | 0xa2e34c40 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0xa2e38440 | 0x86c000 | +| 175 | blk.14.attn_q.bias | 0xa36a4440 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0xa36a7c40 | 0x86c000 | +| 177 | blk.14.attn_v.bias | 0xa3f13c40 | 0x800 | +| 178 | blk.14.attn_v.weight | 0xa3f14440 | 0x134000 | +| 179 | blk.14.ffn_down.weight | 0xa4048440 | 0x2c84000 | +| 180 | blk.14.ffn_gate.weight | 0xa6ccc440 | 0x246c000 | +| 181 | blk.14.ffn_norm.weight | 0xa9138440 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0xa913bc40 | 0x246c000 | +| 183 | blk.15.attn_k.bias | 0xab5a7c40 | 0x800 | +| 184 | blk.15.attn_k.weight | 0xab5a8440 | 0x134000 | +| 185 | blk.15.attn_norm.weight | 0xab6dc440 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0xab6dfc40 | 0x86c000 | +| 187 | blk.15.attn_q.bias | 0xabf4bc40 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0xabf4f440 | 0x86c000 | +| 189 | blk.15.attn_v.bias | 0xac7bb440 | 0x800 | +| 190 | blk.15.attn_v.weight | 0xac7bbc40 | 0x134000 | +| 191 | blk.15.ffn_down.weight | 0xac8efc40 | 0x2c84000 | +| 192 | blk.15.ffn_gate.weight | 0xaf573c40 | 0x246c000 | +| 193 | blk.15.ffn_norm.weight | 0xb19dfc40 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0xb19e3440 | 0x246c000 | +| 195 | blk.16.attn_k.bias | 0xb3e4f440 | 0x800 | +| 196 | blk.16.attn_k.weight | 0xb3e4fc40 | 0x134000 | +| 197 | blk.16.attn_norm.weight | 0xb3f83c40 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0xb3f87440 | 0x86c000 | +| 199 | blk.16.attn_q.bias | 0xb47f3440 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0xb47f6c40 | 0x86c000 | +| 201 | blk.16.attn_v.bias | 0xb5062c40 | 0x800 | +| 202 | blk.16.attn_v.weight | 0xb5063440 | 0x134000 | +| 203 | blk.16.ffn_down.weight | 0xb5197440 | 0x2c84000 | +| 204 | blk.16.ffn_gate.weight | 0xb7e1b440 | 0x246c000 | +| 205 | blk.16.ffn_norm.weight | 0xba287440 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0xba28ac40 | 0x246c000 | +| 207 | blk.17.attn_k.bias | 0xbc6f6c40 | 0x800 | +| 208 | blk.17.attn_k.weight | 0xbc6f7440 | 0x134000 | +| 209 | blk.17.attn_norm.weight | 0xbc82b440 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0xbc82ec40 | 0x86c000 | +| 211 | blk.17.attn_q.bias | 0xbd09ac40 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0xbd09e440 | 0x86c000 | +| 213 | blk.17.attn_v.bias | 0xbd90a440 | 0x800 | +| 214 | blk.17.attn_v.weight | 0xbd90ac40 | 0x134000 | +| 215 | blk.17.ffn_down.weight | 0xbda3ec40 | 0x2c84000 | +| 216 | blk.17.ffn_gate.weight | 0xc06c2c40 | 0x246c000 | +| 217 | blk.17.ffn_norm.weight | 0xc2b2ec40 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0xc2b32440 | 0x246c000 | +| 219 | blk.18.attn_k.bias | 0xc4f9e440 | 0x800 | +| 220 | blk.18.attn_k.weight | 0xc4f9ec40 | 0x134000 | +| 221 | blk.18.attn_norm.weight | 0xc50d2c40 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0xc50d6440 | 0x86c000 | +| 223 | blk.18.attn_q.bias | 0xc5942440 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0xc5945c40 | 0x86c000 | +| 225 | blk.18.attn_v.bias | 0xc61b1c40 | 0x800 | +| 226 | blk.18.attn_v.weight | 0xc61b2440 | 0x134000 | +| 227 | blk.18.ffn_down.weight | 0xc62e6440 | 0x351d800 | +| 228 | blk.18.ffn_gate.weight | 0xc9803c40 | 0x246c000 | +| 229 | blk.18.ffn_norm.weight | 0xcbc6fc40 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0xcbc73440 | 0x246c000 | +| 231 | blk.19.attn_k.bias | 0xce0df440 | 0x800 | +| 232 | blk.19.attn_k.weight | 0xce0dfc40 | 0x134000 | +| 233 | blk.19.attn_norm.weight | 0xce213c40 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0xce217440 | 0x86c000 | +| 235 | blk.19.attn_q.bias | 0xcea83440 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0xcea86c40 | 0x86c000 | +| 237 | blk.19.attn_v.bias | 0xcf2f2c40 | 0x800 | +| 238 | blk.19.attn_v.weight | 0xcf2f3440 | 0x134000 | +| 239 | blk.19.ffn_down.weight | 0xcf427440 | 0x351d800 | +| 240 | blk.19.ffn_gate.weight | 0xd2944c40 | 0x246c000 | +| 241 | blk.19.ffn_norm.weight | 0xd4db0c40 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0xd4db4440 | 0x246c000 | +| 243 | blk.20.attn_k.bias | 0xd7220440 | 0x800 | +| 244 | blk.20.attn_k.weight | 0xd7220c40 | 0x134000 | +| 245 | blk.20.attn_norm.weight | 0xd7354c40 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0xd7358440 | 0x86c000 | +| 247 | blk.20.attn_q.bias | 0xd7bc4440 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0xd7bc7c40 | 0x86c000 | +| 249 | blk.20.attn_v.bias | 0xd8433c40 | 0x800 | +| 250 | blk.20.attn_v.weight | 0xd8434440 | 0x134000 | +| 251 | blk.20.ffn_down.weight | 0xd8568440 | 0x351d800 | +| 252 | blk.20.ffn_gate.weight | 0xdba85c40 | 0x2c84000 | +| 253 | blk.20.ffn_norm.weight | 0xde709c40 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0xde70d440 | 0x2c84000 | +| 255 | blk.21.attn_k.bias | 0xe1391440 | 0x800 | +| 256 | blk.21.attn_k.weight | 0xe1391c40 | 0x134000 | +| 257 | blk.21.attn_norm.weight | 0xe14c5c40 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0xe14c9440 | 0x86c000 | +| 259 | blk.21.attn_q.bias | 0xe1d35440 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0xe1d38c40 | 0x86c000 | +| 261 | blk.21.attn_v.bias | 0xe25a4c40 | 0x800 | +| 262 | blk.21.attn_v.weight | 0xe25a5440 | 0x134000 | +| 263 | blk.21.ffn_down.weight | 0xe26d9440 | 0x351d800 | +| 264 | blk.21.ffn_gate.weight | 0xe5bf6c40 | 0x2c84000 | +| 265 | blk.21.ffn_norm.weight | 0xe887ac40 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0xe887e440 | 0x2c84000 | +| 267 | blk.22.attn_k.bias | 0xeb502440 | 0x800 | +| 268 | blk.22.attn_k.weight | 0xeb502c40 | 0x134000 | +| 269 | blk.22.attn_norm.weight | 0xeb636c40 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0xeb63a440 | 0x86c000 | +| 271 | blk.22.attn_q.bias | 0xebea6440 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0xebea9c40 | 0x86c000 | +| 273 | blk.22.attn_v.bias | 0xec715c40 | 0x800 | +| 274 | blk.22.attn_v.weight | 0xec716440 | 0x134000 | +| 275 | blk.22.ffn_down.weight | 0xec84a440 | 0x351d800 | +| 276 | blk.22.ffn_gate.weight | 0xefd67c40 | 0x2c84000 | +| 277 | blk.22.ffn_norm.weight | 0xf29ebc40 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0xf29ef440 | 0x2c84000 | +| 279 | blk.23.attn_k.bias | 0xf5673440 | 0x800 | +| 280 | blk.23.attn_k.weight | 0xf5673c40 | 0x134000 | +| 281 | blk.23.attn_norm.weight | 0xf57a7c40 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0xf57ab440 | 0x86c000 | +| 283 | blk.23.attn_q.bias | 0xf6017440 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0xf601ac40 | 0x86c000 | +| 285 | blk.23.attn_v.bias | 0xf6886c40 | 0x800 | +| 286 | blk.23.attn_v.weight | 0xf6887440 | 0x134000 | +| 287 | blk.23.ffn_down.weight | 0xf69bb440 | 0x351d800 | +| 288 | blk.23.ffn_gate.weight | 0xf9ed8c40 | 0x2c84000 | +| 289 | blk.23.ffn_norm.weight | 0xfcb5cc40 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0xfcb60440 | 0x2c84000 | +| 291 | blk.24.attn_k.bias | 0xff7e4440 | 0x800 | +| 292 | blk.24.attn_k.weight | 0xff7e4c40 | 0x134000 | +| 293 | blk.24.attn_norm.weight | 0xff918c40 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0xff91c440 | 0x86c000 | +| 295 | blk.24.attn_q.bias | 0x100188440 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0x10018bc40 | 0x86c000 | +| 297 | blk.24.attn_v.bias | 0x1009f7c40 | 0x800 | +| 298 | blk.24.attn_v.weight | 0x1009f8440 | 0x134000 | +| 299 | blk.24.ffn_down.weight | 0x100b2c440 | 0x351d800 | +| 300 | blk.24.ffn_gate.weight | 0x104049c40 | 0x2c84000 | +| 301 | blk.24.ffn_norm.weight | 0x106ccdc40 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0x106cd1440 | 0x2c84000 | +| 303 | blk.25.attn_k.bias | 0x109955440 | 0x800 | +| 304 | blk.25.attn_k.weight | 0x109955c40 | 0x134000 | +| 305 | blk.25.attn_norm.weight | 0x109a89c40 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0x109a8d440 | 0x86c000 | +| 307 | blk.25.attn_q.bias | 0x10a2f9440 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0x10a2fcc40 | 0x86c000 | +| 309 | blk.25.attn_v.bias | 0x10ab68c40 | 0x800 | +| 310 | blk.25.attn_v.weight | 0x10ab69440 | 0x134000 | +| 311 | blk.25.ffn_down.weight | 0x10ac9d440 | 0x351d800 | +| 312 | blk.25.ffn_gate.weight | 0x10e1bac40 | 0x2c84000 | +| 313 | blk.25.ffn_norm.weight | 0x110e3ec40 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0x110e42440 | 0x2c84000 | +| 315 | blk.26.attn_k.bias | 0x113ac6440 | 0x800 | +| 316 | blk.26.attn_k.weight | 0x113ac6c40 | 0x134000 | +| 317 | blk.26.attn_norm.weight | 0x113bfac40 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0x113bfe440 | 0x86c000 | +| 319 | blk.26.attn_q.bias | 0x11446a440 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0x11446dc40 | 0x86c000 | +| 321 | blk.26.attn_v.bias | 0x114cd9c40 | 0x800 | +| 322 | blk.26.attn_v.weight | 0x114cda440 | 0x134000 | +| 323 | blk.26.ffn_down.weight | 0x114e0e440 | 0x351d800 | +| 324 | blk.26.ffn_gate.weight | 0x11832bc40 | 0x2c84000 | +| 325 | blk.26.ffn_norm.weight | 0x11afafc40 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0x11afb3440 | 0x2c84000 | +| 327 | blk.27.attn_k.bias | 0x11dc37440 | 0x800 | +| 328 | blk.27.attn_k.weight | 0x11dc37c40 | 0x134000 | +| 329 | blk.27.attn_norm.weight | 0x11dd6bc40 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0x11dd6f440 | 0x86c000 | +| 331 | blk.27.attn_q.bias | 0x11e5db440 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0x11e5dec40 | 0x86c000 | +| 333 | blk.27.attn_v.bias | 0x11ee4ac40 | 0x800 | +| 334 | blk.27.attn_v.weight | 0x11ee4b440 | 0x134000 | +| 335 | blk.27.ffn_down.weight | 0x11ef7f440 | 0x351d800 | +| 336 | blk.27.ffn_gate.weight | 0x12249cc40 | 0x2c84000 | +| 337 | blk.27.ffn_norm.weight | 0x125120c40 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0x125124440 | 0x2c84000 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q5_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q4_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q4_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q5_K | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q4_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q6_K | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q6_K.md b/scores/Hammer2.1-7b-Q6_K.md new file mode 100644 index 0000000..ba16839 --- /dev/null +++ b/scores/Hammer2.1-7b-Q6_K.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q6_K.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 18 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q6\_K.gguf - GGUF Internal File Dump](#hammer21-7b-q6_kgguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x1a93d1bc | +| 1 | output_norm.weight | 0x1aee8680 | 0x3800 | +| 2 | token_embd.weight | 0x1aeebe80 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x28daa660 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x28daae60 | 0x134000 | +| 5 | blk.0.attn_norm.weight | 0x28edee60 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x28ee2660 | 0xa0c800 | +| 7 | blk.0.attn_q.bias | 0x298eee60 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x298f2660 | 0x86c000 | +| 9 | blk.0.attn_v.bias | 0x2a15e660 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x2a15ee60 | 0x16f800 | +| 11 | blk.0.ffn_down.weight | 0x2a2ce660 | 0x44cc000 | +| 12 | blk.0.ffn_gate.weight | 0x2e79a660 | 0x2c84000 | +| 13 | blk.0.ffn_norm.weight | 0x3141e660 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x31421e60 | 0x2c84000 | +| 15 | blk.1.attn_k.bias | 0x340a5e60 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x340a6660 | 0x134000 | +| 17 | blk.1.attn_norm.weight | 0x341da660 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x341dde60 | 0xa0c800 | +| 19 | blk.1.attn_q.bias | 0x34bea660 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x34bede60 | 0x86c000 | +| 21 | blk.1.attn_v.bias | 0x35459e60 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x3545a660 | 0x16f800 | +| 23 | blk.1.ffn_down.weight | 0x355c9e60 | 0x44cc000 | +| 24 | blk.1.ffn_gate.weight | 0x39a95e60 | 0x351d800 | +| 25 | blk.1.ffn_norm.weight | 0x3cfb3660 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x3cfb6e60 | 0x351d800 | +| 27 | blk.2.attn_k.bias | 0x404d4660 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x404d4e60 | 0x134000 | +| 29 | blk.2.attn_norm.weight | 0x40608e60 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x4060c660 | 0xa0c800 | +| 31 | blk.2.attn_q.bias | 0x41018e60 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x4101c660 | 0x86c000 | +| 33 | blk.2.attn_v.bias | 0x41888660 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x41888e60 | 0x16f800 | +| 35 | blk.2.ffn_down.weight | 0x419f8660 | 0x44cc000 | +| 36 | blk.2.ffn_gate.weight | 0x45ec4660 | 0x351d800 | +| 37 | blk.2.ffn_norm.weight | 0x493e1e60 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x493e5660 | 0x351d800 | +| 39 | blk.3.attn_k.bias | 0x4c902e60 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x4c903660 | 0x134000 | +| 41 | blk.3.attn_norm.weight | 0x4ca37660 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x4ca3ae60 | 0xa0c800 | +| 43 | blk.3.attn_q.bias | 0x4d447660 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x4d44ae60 | 0x86c000 | +| 45 | blk.3.attn_v.bias | 0x4dcb6e60 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x4dcb7660 | 0x16f800 | +| 47 | blk.3.ffn_down.weight | 0x4de26e60 | 0x44cc000 | +| 48 | blk.3.ffn_gate.weight | 0x522f2e60 | 0x351d800 | +| 49 | blk.3.ffn_norm.weight | 0x55810660 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x55813e60 | 0x351d800 | +| 51 | blk.4.attn_k.bias | 0x58d31660 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x58d31e60 | 0x134000 | +| 53 | blk.4.attn_norm.weight | 0x58e65e60 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x58e69660 | 0xa0c800 | +| 55 | blk.4.attn_q.bias | 0x59875e60 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x59879660 | 0x86c000 | +| 57 | blk.4.attn_v.bias | 0x5a0e5660 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x5a0e5e60 | 0x16f800 | +| 59 | blk.4.ffn_down.weight | 0x5a255660 | 0x44cc000 | +| 60 | blk.4.ffn_gate.weight | 0x5e721660 | 0x351d800 | +| 61 | blk.4.ffn_norm.weight | 0x61c3ee60 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x61c42660 | 0x351d800 | +| 63 | blk.5.attn_k.bias | 0x6515fe60 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x65160660 | 0x134000 | +| 65 | blk.5.attn_norm.weight | 0x65294660 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x65297e60 | 0xa0c800 | +| 67 | blk.5.attn_q.bias | 0x65ca4660 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x65ca7e60 | 0x86c000 | +| 69 | blk.5.attn_v.bias | 0x66513e60 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x66514660 | 0x16f800 | +| 71 | blk.5.ffn_down.weight | 0x66683e60 | 0x44cc000 | +| 72 | blk.5.ffn_gate.weight | 0x6ab4fe60 | 0x351d800 | +| 73 | blk.5.ffn_norm.weight | 0x6e06d660 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x6e070e60 | 0x351d800 | +| 75 | blk.6.attn_k.bias | 0x7158e660 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x7158ee60 | 0x134000 | +| 77 | blk.6.attn_norm.weight | 0x716c2e60 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x716c6660 | 0xa0c800 | +| 79 | blk.6.attn_q.bias | 0x720d2e60 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x720d6660 | 0x86c000 | +| 81 | blk.6.attn_v.bias | 0x72942660 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x72942e60 | 0x16f800 | +| 83 | blk.6.ffn_down.weight | 0x72ab2660 | 0x44cc000 | +| 84 | blk.6.ffn_gate.weight | 0x76f7e660 | 0x2c84000 | +| 85 | blk.6.ffn_norm.weight | 0x79c02660 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x79c05e60 | 0x2c84000 | +| 87 | blk.7.attn_k.bias | 0x7c889e60 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x7c88a660 | 0x134000 | +| 89 | blk.7.attn_norm.weight | 0x7c9be660 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x7c9c1e60 | 0xa0c800 | +| 91 | blk.7.attn_q.bias | 0x7d3ce660 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x7d3d1e60 | 0x86c000 | +| 93 | blk.7.attn_v.bias | 0x7dc3de60 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x7dc3e660 | 0x16f800 | +| 95 | blk.7.ffn_down.weight | 0x7ddade60 | 0x44cc000 | +| 96 | blk.7.ffn_gate.weight | 0x82279e60 | 0x2c84000 | +| 97 | blk.7.ffn_norm.weight | 0x84efde60 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x84f01660 | 0x2c84000 | +| 99 | blk.8.attn_k.bias | 0x87b85660 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x87b85e60 | 0x134000 | +| 101 | blk.8.attn_norm.weight | 0x87cb9e60 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x87cbd660 | 0xa0c800 | +| 103 | blk.8.attn_q.bias | 0x886c9e60 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0x886cd660 | 0x86c000 | +| 105 | blk.8.attn_v.bias | 0x88f39660 | 0x800 | +| 106 | blk.8.attn_v.weight | 0x88f39e60 | 0x16f800 | +| 107 | blk.8.ffn_down.weight | 0x890a9660 | 0x44cc000 | +| 108 | blk.8.ffn_gate.weight | 0x8d575660 | 0x2c84000 | +| 109 | blk.8.ffn_norm.weight | 0x901f9660 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0x901fce60 | 0x2c84000 | +| 111 | blk.9.attn_k.bias | 0x92e80e60 | 0x800 | +| 112 | blk.9.attn_k.weight | 0x92e81660 | 0x134000 | +| 113 | blk.9.attn_norm.weight | 0x92fb5660 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0x92fb8e60 | 0xa0c800 | +| 115 | blk.9.attn_q.bias | 0x939c5660 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0x939c8e60 | 0x86c000 | +| 117 | blk.9.attn_v.bias | 0x94234e60 | 0x800 | +| 118 | blk.9.attn_v.weight | 0x94235660 | 0x16f800 | +| 119 | blk.9.ffn_down.weight | 0x943a4e60 | 0x44cc000 | +| 120 | blk.9.ffn_gate.weight | 0x98870e60 | 0x351d800 | +| 121 | blk.9.ffn_norm.weight | 0x9bd8e660 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0x9bd91e60 | 0x351d800 | +| 123 | blk.10.attn_k.bias | 0x9f2af660 | 0x800 | +| 124 | blk.10.attn_k.weight | 0x9f2afe60 | 0x134000 | +| 125 | blk.10.attn_norm.weight | 0x9f3e3e60 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0x9f3e7660 | 0xa0c800 | +| 127 | blk.10.attn_q.bias | 0x9fdf3e60 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0x9fdf7660 | 0x86c000 | +| 129 | blk.10.attn_v.bias | 0xa0663660 | 0x800 | +| 130 | blk.10.attn_v.weight | 0xa0663e60 | 0x16f800 | +| 131 | blk.10.ffn_down.weight | 0xa07d3660 | 0x44cc000 | +| 132 | blk.10.ffn_gate.weight | 0xa4c9f660 | 0x2c84000 | +| 133 | blk.10.ffn_norm.weight | 0xa7923660 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0xa7926e60 | 0x2c84000 | +| 135 | blk.11.attn_k.bias | 0xaa5aae60 | 0x800 | +| 136 | blk.11.attn_k.weight | 0xaa5ab660 | 0x134000 | +| 137 | blk.11.attn_norm.weight | 0xaa6df660 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0xaa6e2e60 | 0xa0c800 | +| 139 | blk.11.attn_q.bias | 0xab0ef660 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0xab0f2e60 | 0x86c000 | +| 141 | blk.11.attn_v.bias | 0xab95ee60 | 0x800 | +| 142 | blk.11.attn_v.weight | 0xab95f660 | 0x16f800 | +| 143 | blk.11.ffn_down.weight | 0xabacee60 | 0x44cc000 | +| 144 | blk.11.ffn_gate.weight | 0xaff9ae60 | 0x2c84000 | +| 145 | blk.11.ffn_norm.weight | 0xb2c1ee60 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0xb2c22660 | 0x2c84000 | +| 147 | blk.12.attn_k.bias | 0xb58a6660 | 0x800 | +| 148 | blk.12.attn_k.weight | 0xb58a6e60 | 0x134000 | +| 149 | blk.12.attn_norm.weight | 0xb59dae60 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0xb59de660 | 0xa0c800 | +| 151 | blk.12.attn_q.bias | 0xb63eae60 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0xb63ee660 | 0x86c000 | +| 153 | blk.12.attn_v.bias | 0xb6c5a660 | 0x800 | +| 154 | blk.12.attn_v.weight | 0xb6c5ae60 | 0x16f800 | +| 155 | blk.12.ffn_down.weight | 0xb6dca660 | 0x44cc000 | +| 156 | blk.12.ffn_gate.weight | 0xbb296660 | 0x2c84000 | +| 157 | blk.12.ffn_norm.weight | 0xbdf1a660 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0xbdf1de60 | 0x2c84000 | +| 159 | blk.13.attn_k.bias | 0xc0ba1e60 | 0x800 | +| 160 | blk.13.attn_k.weight | 0xc0ba2660 | 0x134000 | +| 161 | blk.13.attn_norm.weight | 0xc0cd6660 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0xc0cd9e60 | 0xa0c800 | +| 163 | blk.13.attn_q.bias | 0xc16e6660 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0xc16e9e60 | 0x86c000 | +| 165 | blk.13.attn_v.bias | 0xc1f55e60 | 0x800 | +| 166 | blk.13.attn_v.weight | 0xc1f56660 | 0x16f800 | +| 167 | blk.13.ffn_down.weight | 0xc20c5e60 | 0x44cc000 | +| 168 | blk.13.ffn_gate.weight | 0xc6591e60 | 0x2c84000 | +| 169 | blk.13.ffn_norm.weight | 0xc9215e60 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0xc9219660 | 0x2c84000 | +| 171 | blk.14.attn_k.bias | 0xcbe9d660 | 0x800 | +| 172 | blk.14.attn_k.weight | 0xcbe9de60 | 0x16f800 | +| 173 | blk.14.attn_norm.weight | 0xcc00d660 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0xcc010e60 | 0xa0c800 | +| 175 | blk.14.attn_q.bias | 0xcca1d660 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0xcca20e60 | 0xa0c800 | +| 177 | blk.14.attn_v.bias | 0xcd42d660 | 0x800 | +| 178 | blk.14.attn_v.weight | 0xcd42de60 | 0x1dc000 | +| 179 | blk.14.ffn_down.weight | 0xcd609e60 | 0x44cc000 | +| 180 | blk.14.ffn_gate.weight | 0xd1ad5e60 | 0x2c84000 | +| 181 | blk.14.ffn_norm.weight | 0xd4759e60 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0xd475d660 | 0x2c84000 | +| 183 | blk.15.attn_k.bias | 0xd73e1660 | 0x800 | +| 184 | blk.15.attn_k.weight | 0xd73e1e60 | 0x16f800 | +| 185 | blk.15.attn_norm.weight | 0xd7551660 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0xd7554e60 | 0xa0c800 | +| 187 | blk.15.attn_q.bias | 0xd7f61660 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0xd7f64e60 | 0xa0c800 | +| 189 | blk.15.attn_v.bias | 0xd8971660 | 0x800 | +| 190 | blk.15.attn_v.weight | 0xd8971e60 | 0x1dc000 | +| 191 | blk.15.ffn_down.weight | 0xd8b4de60 | 0x44cc000 | +| 192 | blk.15.ffn_gate.weight | 0xdd019e60 | 0x2c84000 | +| 193 | blk.15.ffn_norm.weight | 0xdfc9de60 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0xdfca1660 | 0x2c84000 | +| 195 | blk.16.attn_k.bias | 0xe2925660 | 0x800 | +| 196 | blk.16.attn_k.weight | 0xe2925e60 | 0x16f800 | +| 197 | blk.16.attn_norm.weight | 0xe2a95660 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0xe2a98e60 | 0xa0c800 | +| 199 | blk.16.attn_q.bias | 0xe34a5660 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0xe34a8e60 | 0xa0c800 | +| 201 | blk.16.attn_v.bias | 0xe3eb5660 | 0x800 | +| 202 | blk.16.attn_v.weight | 0xe3eb5e60 | 0x1dc000 | +| 203 | blk.16.ffn_down.weight | 0xe4091e60 | 0x44cc000 | +| 204 | blk.16.ffn_gate.weight | 0xe855de60 | 0x2c84000 | +| 205 | blk.16.ffn_norm.weight | 0xeb1e1e60 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0xeb1e5660 | 0x2c84000 | +| 207 | blk.17.attn_k.bias | 0xede69660 | 0x800 | +| 208 | blk.17.attn_k.weight | 0xede69e60 | 0x16f800 | +| 209 | blk.17.attn_norm.weight | 0xedfd9660 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0xedfdce60 | 0xa0c800 | +| 211 | blk.17.attn_q.bias | 0xee9e9660 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0xee9ece60 | 0xa0c800 | +| 213 | blk.17.attn_v.bias | 0xef3f9660 | 0x800 | +| 214 | blk.17.attn_v.weight | 0xef3f9e60 | 0x1dc000 | +| 215 | blk.17.ffn_down.weight | 0xef5d5e60 | 0x44cc000 | +| 216 | blk.17.ffn_gate.weight | 0xf3aa1e60 | 0x2c84000 | +| 217 | blk.17.ffn_norm.weight | 0xf6725e60 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0xf6729660 | 0x2c84000 | +| 219 | blk.18.attn_k.bias | 0xf93ad660 | 0x800 | +| 220 | blk.18.attn_k.weight | 0xf93ade60 | 0x16f800 | +| 221 | blk.18.attn_norm.weight | 0xf951d660 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0xf9520e60 | 0xa0c800 | +| 223 | blk.18.attn_q.bias | 0xf9f2d660 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0xf9f30e60 | 0xa0c800 | +| 225 | blk.18.attn_v.bias | 0xfa93d660 | 0x800 | +| 226 | blk.18.attn_v.weight | 0xfa93de60 | 0x1dc000 | +| 227 | blk.18.ffn_down.weight | 0xfab19e60 | 0x44cc000 | +| 228 | blk.18.ffn_gate.weight | 0xfefe5e60 | 0x2c84000 | +| 229 | blk.18.ffn_norm.weight | 0x101c69e60 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x101c6d660 | 0x2c84000 | +| 231 | blk.19.attn_k.bias | 0x1048f1660 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x1048f1e60 | 0x16f800 | +| 233 | blk.19.attn_norm.weight | 0x104a61660 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x104a64e60 | 0xa0c800 | +| 235 | blk.19.attn_q.bias | 0x105471660 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x105474e60 | 0xa0c800 | +| 237 | blk.19.attn_v.bias | 0x105e81660 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x105e81e60 | 0x1dc000 | +| 239 | blk.19.ffn_down.weight | 0x10605de60 | 0x44cc000 | +| 240 | blk.19.ffn_gate.weight | 0x10a529e60 | 0x2c84000 | +| 241 | blk.19.ffn_norm.weight | 0x10d1ade60 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x10d1b1660 | 0x2c84000 | +| 243 | blk.20.attn_k.bias | 0x10fe35660 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x10fe35e60 | 0x16f800 | +| 245 | blk.20.attn_norm.weight | 0x10ffa5660 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x10ffa8e60 | 0xa0c800 | +| 247 | blk.20.attn_q.bias | 0x1109b5660 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x1109b8e60 | 0xa0c800 | +| 249 | blk.20.attn_v.bias | 0x1113c5660 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x1113c5e60 | 0x1dc000 | +| 251 | blk.20.ffn_down.weight | 0x1115a1e60 | 0x44cc000 | +| 252 | blk.20.ffn_gate.weight | 0x115a6de60 | 0x351d800 | +| 253 | blk.20.ffn_norm.weight | 0x118f8b660 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x118f8ee60 | 0x351d800 | +| 255 | blk.21.attn_k.bias | 0x11c4ac660 | 0x800 | +| 256 | blk.21.attn_k.weight | 0x11c4ace60 | 0x16f800 | +| 257 | blk.21.attn_norm.weight | 0x11c61c660 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0x11c61fe60 | 0xa0c800 | +| 259 | blk.21.attn_q.bias | 0x11d02c660 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0x11d02fe60 | 0xa0c800 | +| 261 | blk.21.attn_v.bias | 0x11da3c660 | 0x800 | +| 262 | blk.21.attn_v.weight | 0x11da3ce60 | 0x1dc000 | +| 263 | blk.21.ffn_down.weight | 0x11dc18e60 | 0x44cc000 | +| 264 | blk.21.ffn_gate.weight | 0x1220e4e60 | 0x351d800 | +| 265 | blk.21.ffn_norm.weight | 0x125602660 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0x125605e60 | 0x351d800 | +| 267 | blk.22.attn_k.bias | 0x128b23660 | 0x800 | +| 268 | blk.22.attn_k.weight | 0x128b23e60 | 0x16f800 | +| 269 | blk.22.attn_norm.weight | 0x128c93660 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0x128c96e60 | 0xa0c800 | +| 271 | blk.22.attn_q.bias | 0x1296a3660 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0x1296a6e60 | 0xa0c800 | +| 273 | blk.22.attn_v.bias | 0x12a0b3660 | 0x800 | +| 274 | blk.22.attn_v.weight | 0x12a0b3e60 | 0x1dc000 | +| 275 | blk.22.ffn_down.weight | 0x12a28fe60 | 0x44cc000 | +| 276 | blk.22.ffn_gate.weight | 0x12e75be60 | 0x351d800 | +| 277 | blk.22.ffn_norm.weight | 0x131c79660 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0x131c7ce60 | 0x351d800 | +| 279 | blk.23.attn_k.bias | 0x13519a660 | 0x800 | +| 280 | blk.23.attn_k.weight | 0x13519ae60 | 0x16f800 | +| 281 | blk.23.attn_norm.weight | 0x13530a660 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0x13530de60 | 0xa0c800 | +| 283 | blk.23.attn_q.bias | 0x135d1a660 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0x135d1de60 | 0xa0c800 | +| 285 | blk.23.attn_v.bias | 0x13672a660 | 0x800 | +| 286 | blk.23.attn_v.weight | 0x13672ae60 | 0x1dc000 | +| 287 | blk.23.ffn_down.weight | 0x136906e60 | 0x44cc000 | +| 288 | blk.23.ffn_gate.weight | 0x13add2e60 | 0x351d800 | +| 289 | blk.23.ffn_norm.weight | 0x13e2f0660 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0x13e2f3e60 | 0x351d800 | +| 291 | blk.24.attn_k.bias | 0x141811660 | 0x800 | +| 292 | blk.24.attn_k.weight | 0x141811e60 | 0x16f800 | +| 293 | blk.24.attn_norm.weight | 0x141981660 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0x141984e60 | 0xa0c800 | +| 295 | blk.24.attn_q.bias | 0x142391660 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0x142394e60 | 0xa0c800 | +| 297 | blk.24.attn_v.bias | 0x142da1660 | 0x800 | +| 298 | blk.24.attn_v.weight | 0x142da1e60 | 0x1dc000 | +| 299 | blk.24.ffn_down.weight | 0x142f7de60 | 0x44cc000 | +| 300 | blk.24.ffn_gate.weight | 0x147449e60 | 0x351d800 | +| 301 | blk.24.ffn_norm.weight | 0x14a967660 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0x14a96ae60 | 0x351d800 | +| 303 | blk.25.attn_k.bias | 0x14de88660 | 0x800 | +| 304 | blk.25.attn_k.weight | 0x14de88e60 | 0x16f800 | +| 305 | blk.25.attn_norm.weight | 0x14dff8660 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0x14dffbe60 | 0xa0c800 | +| 307 | blk.25.attn_q.bias | 0x14ea08660 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0x14ea0be60 | 0xa0c800 | +| 309 | blk.25.attn_v.bias | 0x14f418660 | 0x800 | +| 310 | blk.25.attn_v.weight | 0x14f418e60 | 0x1dc000 | +| 311 | blk.25.ffn_down.weight | 0x14f5f4e60 | 0x44cc000 | +| 312 | blk.25.ffn_gate.weight | 0x153ac0e60 | 0x351d800 | +| 313 | blk.25.ffn_norm.weight | 0x156fde660 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0x156fe1e60 | 0x351d800 | +| 315 | blk.26.attn_k.bias | 0x15a4ff660 | 0x800 | +| 316 | blk.26.attn_k.weight | 0x15a4ffe60 | 0x16f800 | +| 317 | blk.26.attn_norm.weight | 0x15a66f660 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0x15a672e60 | 0xa0c800 | +| 319 | blk.26.attn_q.bias | 0x15b07f660 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0x15b082e60 | 0xa0c800 | +| 321 | blk.26.attn_v.bias | 0x15ba8f660 | 0x800 | +| 322 | blk.26.attn_v.weight | 0x15ba8fe60 | 0x1dc000 | +| 323 | blk.26.ffn_down.weight | 0x15bc6be60 | 0x44cc000 | +| 324 | blk.26.ffn_gate.weight | 0x160137e60 | 0x351d800 | +| 325 | blk.26.ffn_norm.weight | 0x163655660 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0x163658e60 | 0x351d800 | +| 327 | blk.27.attn_k.bias | 0x166b76660 | 0x800 | +| 328 | blk.27.attn_k.weight | 0x166b76e60 | 0x16f800 | +| 329 | blk.27.attn_norm.weight | 0x166ce6660 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0x166ce9e60 | 0xa0c800 | +| 331 | blk.27.attn_q.bias | 0x1676f6660 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0x1676f9e60 | 0xa0c800 | +| 333 | blk.27.attn_v.bias | 0x168106660 | 0x800 | +| 334 | blk.27.attn_v.weight | 0x168106e60 | 0x1dc000 | +| 335 | blk.27.ffn_down.weight | 0x1682e2e60 | 0x44cc000 | +| 336 | blk.27.ffn_gate.weight | 0x16c7aee60 | 0x351d800 | +| 337 | blk.27.ffn_norm.weight | 0x16fccc660 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0x16fccfe60 | 0x351d800 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q6_K | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q5_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q5_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q5_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-Q8_0.md b/scores/Hammer2.1-7b-Q8_0.md new file mode 100644 index 0000000..807f0cd --- /dev/null +++ b/scores/Hammer2.1-7b-Q8_0.md @@ -0,0 +1,1029 @@ +# Hammer2.1-7b-Q8_0.gguf - GGUF Internal File Dump + +- Endian: LITTLE endian + +## Key Value Metadata Store + +There are 36 key-value pairs in this file + +| POS | TYPE | Count | Key | Value | +|----:|:---------|-------:|:-------------------------------------------|:--------------------------------------------------------------------| +| 1 | UINT32 | 1 | GGUF.version | 3 | +| 2 | UINT64 | 1 | GGUF.tensor_count | 339 | +| 3 | UINT64 | 1 | GGUF.kv_count | 33 | +| 4 | STRING | 1 | general.architecture | `qwen2` | +| 5 | STRING | 1 | general.type | `model` | +| 6 | STRING | 1 | general.name | `Hammer2.1 7b GGUF` | +| 7 | STRING | 1 | general.finetune | `GGUF` | +| 8 | STRING | 1 | general.basename | `Hammer2.1` | +| 9 | STRING | 1 | general.size_label | `7B` | +| 10 | UINT32 | 1 | qwen2.block_count | 28 | +| 11 | UINT32 | 1 | qwen2.context_length | 32768 | +| 12 | UINT32 | 1 | qwen2.embedding_length | 3584 | +| 13 | UINT32 | 1 | qwen2.feed_forward_length | 18944 | +| 14 | UINT32 | 1 | qwen2.attention.head_count | 28 | +| 15 | UINT32 | 1 | qwen2.attention.head_count_kv | 4 | +| 16 | FLOAT32 | 1 | qwen2.rope.freq_base | 1000000.0 | +| 17 | FLOAT32 | 1 | qwen2.attention.layer_norm_rms_epsilon | 1e-06 | +| 18 | STRING | 1 | qwen2.rope.scaling.type | `yarn` | +| 19 | FLOAT32 | 1 | qwen2.rope.scaling.factor | 4.0 | +| 20 | UINT32 | 1 | qwen2.rope.scaling.original_context_length | 32768 | +| 21 | STRING | 1 | tokenizer.ggml.model | `gpt2` | +| 22 | STRING | 1 | tokenizer.ggml.pre | `qwen2` | +| 23 | [STRING] | 151665 | tokenizer.ggml.tokens | [ `!`, `"`, `#`, `$`, `%`, ... ] | +| 24 | [INT32] | 151665 | tokenizer.ggml.token_type | [ 1, 1, 1, 1, 1, 1, 1, ... ] | +| 25 | [STRING] | 151387 | tokenizer.ggml.merges | [ `Ġ Ġ`, `ĠĠ ĠĠ`, `i n`, `Ġ t`, `ĠĠĠĠ ĠĠĠĠ`, ... ] | +| 26 | UINT32 | 1 | tokenizer.ggml.eos_token_id | 151645 | +| 27 | UINT32 | 1 | tokenizer.ggml.padding_token_id | 151643 | +| 28 | UINT32 | 1 | tokenizer.ggml.bos_token_id | 151643 | +| 29 | BOOL | 1 | tokenizer.ggml.add_bos_token | False | +| 30 | STRING | 1 | tokenizer.chat_template | `{%- set system_message = 'You `...`{- '<|im_start|>assistant' }}` | +| 31 | UINT32 | 1 | general.quantization_version | 2 | +| 32 | UINT32 | 1 | general.file_type | 7 | +| 33 | STRING | 1 | quantize.imatrix.file | `./imatrix/imatrix-Hammer2.1-7b-small.dat` | +| 34 | STRING | 1 | quantize.imatrix.dataset | `../../datasets/imatrix/calibration_eur_small.txt` | +| 35 | INT32 | 1 | quantize.imatrix.entries_count | 197 | +| 36 | INT32 | 1 | quantize.imatrix.chunks_count | 968 | + +## Tensors Overview ~8B Elements + +Total number of elements in all tensors: 7612756480 Elements + +- [Hammer2.1-7b-Q8\_0.gguf - GGUF Internal File Dump](#hammer21-7b-q8_0gguf---gguf-internal-file-dump) + - [Key Value Metadata Store](#key-value-metadata-store) + - [Tensors Overview ~8B Elements](#tensors-overview-8b-elements) + - [Tensor Data Offset](#tensor-data-offset) + - [Base Tensor Group : ~1B Elements](#base-tensor-group--1b-elements) + - [Block 0 Tensor Group : ~233M Elements](#block-0-tensor-group--233m-elements) + - [Block 1 Tensor Group : ~233M Elements](#block-1-tensor-group--233m-elements) + - [Block 2 Tensor Group : ~233M Elements](#block-2-tensor-group--233m-elements) + - [Block 3 Tensor Group : ~233M Elements](#block-3-tensor-group--233m-elements) + - [Block 4 Tensor Group : ~233M Elements](#block-4-tensor-group--233m-elements) + - [Block 5 Tensor Group : ~233M Elements](#block-5-tensor-group--233m-elements) + - [Block 6 Tensor Group : ~233M Elements](#block-6-tensor-group--233m-elements) + - [Block 7 Tensor Group : ~233M Elements](#block-7-tensor-group--233m-elements) + - [Block 8 Tensor Group : ~233M Elements](#block-8-tensor-group--233m-elements) + - [Block 9 Tensor Group : ~233M Elements](#block-9-tensor-group--233m-elements) + - [Block 10 Tensor Group : ~233M Elements](#block-10-tensor-group--233m-elements) + - [Block 11 Tensor Group : ~233M Elements](#block-11-tensor-group--233m-elements) + - [Block 12 Tensor Group : ~233M Elements](#block-12-tensor-group--233m-elements) + - [Block 13 Tensor Group : ~233M Elements](#block-13-tensor-group--233m-elements) + - [Block 14 Tensor Group : ~233M Elements](#block-14-tensor-group--233m-elements) + - [Block 15 Tensor Group : ~233M Elements](#block-15-tensor-group--233m-elements) + - [Block 16 Tensor Group : ~233M Elements](#block-16-tensor-group--233m-elements) + - [Block 17 Tensor Group : ~233M Elements](#block-17-tensor-group--233m-elements) + - [Block 18 Tensor Group : ~233M Elements](#block-18-tensor-group--233m-elements) + - [Block 19 Tensor Group : ~233M Elements](#block-19-tensor-group--233m-elements) + - [Block 20 Tensor Group : ~233M Elements](#block-20-tensor-group--233m-elements) + - [Block 21 Tensor Group : ~233M Elements](#block-21-tensor-group--233m-elements) + - [Block 22 Tensor Group : ~233M Elements](#block-22-tensor-group--233m-elements) + - [Block 23 Tensor Group : ~233M Elements](#block-23-tensor-group--233m-elements) + - [Block 24 Tensor Group : ~233M Elements](#block-24-tensor-group--233m-elements) + - [Block 25 Tensor Group : ~233M Elements](#block-25-tensor-group--233m-elements) + - [Block 26 Tensor Group : ~233M Elements](#block-26-tensor-group--233m-elements) + - [Block 27 Tensor Group : ~233M Elements](#block-27-tensor-group--233m-elements) + +### Tensor Data Offset + +This table contains the offset and data segment relative to start of file + +| T_ID | Tensor Layer Name | Data Offset (B) | Data Size (B) | +|-----:|:--------------------------|-----------------:|-----------------:| +| 0 | output.weight | 0x5ab4c0 | 0x226c90e0 | +| 1 | output_norm.weight | 0x22c745a0 | 0x3800 | +| 2 | token_embd.weight | 0x22c77da0 | 0xdebe7c4 | +| 3 | blk.0.attn_k.bias | 0x30b36580 | 0x800 | +| 4 | blk.0.attn_k.weight | 0x30b36d80 | 0x16f800 | +| 5 | blk.0.attn_norm.weight | 0x30ca6580 | 0x3800 | +| 6 | blk.0.attn_output.weight | 0x30ca9d80 | 0xd04000 | +| 7 | blk.0.attn_q.bias | 0x319add80 | 0x3800 | +| 8 | blk.0.attn_q.weight | 0x319b1580 | 0xa0c800 | +| 9 | blk.0.attn_v.bias | 0x323bdd80 | 0x800 | +| 10 | blk.0.attn_v.weight | 0x323be580 | 0x1dc000 | +| 11 | blk.0.ffn_down.weight | 0x3259a580 | 0x44cc000 | +| 12 | blk.0.ffn_gate.weight | 0x36a66580 | 0x351d800 | +| 13 | blk.0.ffn_norm.weight | 0x39f83d80 | 0x3800 | +| 14 | blk.0.ffn_up.weight | 0x39f87580 | 0x351d800 | +| 15 | blk.1.attn_k.bias | 0x3d4a4d80 | 0x800 | +| 16 | blk.1.attn_k.weight | 0x3d4a5580 | 0x16f800 | +| 17 | blk.1.attn_norm.weight | 0x3d614d80 | 0x3800 | +| 18 | blk.1.attn_output.weight | 0x3d618580 | 0xd04000 | +| 19 | blk.1.attn_q.bias | 0x3e31c580 | 0x3800 | +| 20 | blk.1.attn_q.weight | 0x3e31fd80 | 0xa0c800 | +| 21 | blk.1.attn_v.bias | 0x3ed2c580 | 0x800 | +| 22 | blk.1.attn_v.weight | 0x3ed2cd80 | 0x1dc000 | +| 23 | blk.1.ffn_down.weight | 0x3ef08d80 | 0x44cc000 | +| 24 | blk.1.ffn_gate.weight | 0x433d4d80 | 0x44cc000 | +| 25 | blk.1.ffn_norm.weight | 0x478a0d80 | 0x3800 | +| 26 | blk.1.ffn_up.weight | 0x478a4580 | 0x44cc000 | +| 27 | blk.2.attn_k.bias | 0x4bd70580 | 0x800 | +| 28 | blk.2.attn_k.weight | 0x4bd70d80 | 0x16f800 | +| 29 | blk.2.attn_norm.weight | 0x4bee0580 | 0x3800 | +| 30 | blk.2.attn_output.weight | 0x4bee3d80 | 0xd04000 | +| 31 | blk.2.attn_q.bias | 0x4cbe7d80 | 0x3800 | +| 32 | blk.2.attn_q.weight | 0x4cbeb580 | 0xa0c800 | +| 33 | blk.2.attn_v.bias | 0x4d5f7d80 | 0x800 | +| 34 | blk.2.attn_v.weight | 0x4d5f8580 | 0x1dc000 | +| 35 | blk.2.ffn_down.weight | 0x4d7d4580 | 0x44cc000 | +| 36 | blk.2.ffn_gate.weight | 0x51ca0580 | 0x44cc000 | +| 37 | blk.2.ffn_norm.weight | 0x5616c580 | 0x3800 | +| 38 | blk.2.ffn_up.weight | 0x5616fd80 | 0x44cc000 | +| 39 | blk.3.attn_k.bias | 0x5a63bd80 | 0x800 | +| 40 | blk.3.attn_k.weight | 0x5a63c580 | 0x16f800 | +| 41 | blk.3.attn_norm.weight | 0x5a7abd80 | 0x3800 | +| 42 | blk.3.attn_output.weight | 0x5a7af580 | 0xd04000 | +| 43 | blk.3.attn_q.bias | 0x5b4b3580 | 0x3800 | +| 44 | blk.3.attn_q.weight | 0x5b4b6d80 | 0xa0c800 | +| 45 | blk.3.attn_v.bias | 0x5bec3580 | 0x800 | +| 46 | blk.3.attn_v.weight | 0x5bec3d80 | 0x1dc000 | +| 47 | blk.3.ffn_down.weight | 0x5c09fd80 | 0x44cc000 | +| 48 | blk.3.ffn_gate.weight | 0x6056bd80 | 0x44cc000 | +| 49 | blk.3.ffn_norm.weight | 0x64a37d80 | 0x3800 | +| 50 | blk.3.ffn_up.weight | 0x64a3b580 | 0x44cc000 | +| 51 | blk.4.attn_k.bias | 0x68f07580 | 0x800 | +| 52 | blk.4.attn_k.weight | 0x68f07d80 | 0x16f800 | +| 53 | blk.4.attn_norm.weight | 0x69077580 | 0x3800 | +| 54 | blk.4.attn_output.weight | 0x6907ad80 | 0xd04000 | +| 55 | blk.4.attn_q.bias | 0x69d7ed80 | 0x3800 | +| 56 | blk.4.attn_q.weight | 0x69d82580 | 0xa0c800 | +| 57 | blk.4.attn_v.bias | 0x6a78ed80 | 0x800 | +| 58 | blk.4.attn_v.weight | 0x6a78f580 | 0x1dc000 | +| 59 | blk.4.ffn_down.weight | 0x6a96b580 | 0x44cc000 | +| 60 | blk.4.ffn_gate.weight | 0x6ee37580 | 0x44cc000 | +| 61 | blk.4.ffn_norm.weight | 0x73303580 | 0x3800 | +| 62 | blk.4.ffn_up.weight | 0x73306d80 | 0x44cc000 | +| 63 | blk.5.attn_k.bias | 0x777d2d80 | 0x800 | +| 64 | blk.5.attn_k.weight | 0x777d3580 | 0x16f800 | +| 65 | blk.5.attn_norm.weight | 0x77942d80 | 0x3800 | +| 66 | blk.5.attn_output.weight | 0x77946580 | 0xd04000 | +| 67 | blk.5.attn_q.bias | 0x7864a580 | 0x3800 | +| 68 | blk.5.attn_q.weight | 0x7864dd80 | 0xa0c800 | +| 69 | blk.5.attn_v.bias | 0x7905a580 | 0x800 | +| 70 | blk.5.attn_v.weight | 0x7905ad80 | 0x1dc000 | +| 71 | blk.5.ffn_down.weight | 0x79236d80 | 0x44cc000 | +| 72 | blk.5.ffn_gate.weight | 0x7d702d80 | 0x44cc000 | +| 73 | blk.5.ffn_norm.weight | 0x81bced80 | 0x3800 | +| 74 | blk.5.ffn_up.weight | 0x81bd2580 | 0x44cc000 | +| 75 | blk.6.attn_k.bias | 0x8609e580 | 0x800 | +| 76 | blk.6.attn_k.weight | 0x8609ed80 | 0x16f800 | +| 77 | blk.6.attn_norm.weight | 0x8620e580 | 0x3800 | +| 78 | blk.6.attn_output.weight | 0x86211d80 | 0xd04000 | +| 79 | blk.6.attn_q.bias | 0x86f15d80 | 0x3800 | +| 80 | blk.6.attn_q.weight | 0x86f19580 | 0xa0c800 | +| 81 | blk.6.attn_v.bias | 0x87925d80 | 0x800 | +| 82 | blk.6.attn_v.weight | 0x87926580 | 0x1dc000 | +| 83 | blk.6.ffn_down.weight | 0x87b02580 | 0x44cc000 | +| 84 | blk.6.ffn_gate.weight | 0x8bfce580 | 0x351d800 | +| 85 | blk.6.ffn_norm.weight | 0x8f4ebd80 | 0x3800 | +| 86 | blk.6.ffn_up.weight | 0x8f4ef580 | 0x351d800 | +| 87 | blk.7.attn_k.bias | 0x92a0cd80 | 0x800 | +| 88 | blk.7.attn_k.weight | 0x92a0d580 | 0x16f800 | +| 89 | blk.7.attn_norm.weight | 0x92b7cd80 | 0x3800 | +| 90 | blk.7.attn_output.weight | 0x92b80580 | 0xd04000 | +| 91 | blk.7.attn_q.bias | 0x93884580 | 0x3800 | +| 92 | blk.7.attn_q.weight | 0x93887d80 | 0xa0c800 | +| 93 | blk.7.attn_v.bias | 0x94294580 | 0x800 | +| 94 | blk.7.attn_v.weight | 0x94294d80 | 0x1dc000 | +| 95 | blk.7.ffn_down.weight | 0x94470d80 | 0x44cc000 | +| 96 | blk.7.ffn_gate.weight | 0x9893cd80 | 0x351d800 | +| 97 | blk.7.ffn_norm.weight | 0x9be5a580 | 0x3800 | +| 98 | blk.7.ffn_up.weight | 0x9be5dd80 | 0x351d800 | +| 99 | blk.8.attn_k.bias | 0x9f37b580 | 0x800 | +| 100 | blk.8.attn_k.weight | 0x9f37bd80 | 0x16f800 | +| 101 | blk.8.attn_norm.weight | 0x9f4eb580 | 0x3800 | +| 102 | blk.8.attn_output.weight | 0x9f4eed80 | 0xd04000 | +| 103 | blk.8.attn_q.bias | 0xa01f2d80 | 0x3800 | +| 104 | blk.8.attn_q.weight | 0xa01f6580 | 0xa0c800 | +| 105 | blk.8.attn_v.bias | 0xa0c02d80 | 0x800 | +| 106 | blk.8.attn_v.weight | 0xa0c03580 | 0x1dc000 | +| 107 | blk.8.ffn_down.weight | 0xa0ddf580 | 0x44cc000 | +| 108 | blk.8.ffn_gate.weight | 0xa52ab580 | 0x351d800 | +| 109 | blk.8.ffn_norm.weight | 0xa87c8d80 | 0x3800 | +| 110 | blk.8.ffn_up.weight | 0xa87cc580 | 0x351d800 | +| 111 | blk.9.attn_k.bias | 0xabce9d80 | 0x800 | +| 112 | blk.9.attn_k.weight | 0xabcea580 | 0x16f800 | +| 113 | blk.9.attn_norm.weight | 0xabe59d80 | 0x3800 | +| 114 | blk.9.attn_output.weight | 0xabe5d580 | 0xd04000 | +| 115 | blk.9.attn_q.bias | 0xacb61580 | 0x3800 | +| 116 | blk.9.attn_q.weight | 0xacb64d80 | 0xa0c800 | +| 117 | blk.9.attn_v.bias | 0xad571580 | 0x800 | +| 118 | blk.9.attn_v.weight | 0xad571d80 | 0x1dc000 | +| 119 | blk.9.ffn_down.weight | 0xad74dd80 | 0x44cc000 | +| 120 | blk.9.ffn_gate.weight | 0xb1c19d80 | 0x44cc000 | +| 121 | blk.9.ffn_norm.weight | 0xb60e5d80 | 0x3800 | +| 122 | blk.9.ffn_up.weight | 0xb60e9580 | 0x44cc000 | +| 123 | blk.10.attn_k.bias | 0xba5b5580 | 0x800 | +| 124 | blk.10.attn_k.weight | 0xba5b5d80 | 0x16f800 | +| 125 | blk.10.attn_norm.weight | 0xba725580 | 0x3800 | +| 126 | blk.10.attn_output.weight | 0xba728d80 | 0xd04000 | +| 127 | blk.10.attn_q.bias | 0xbb42cd80 | 0x3800 | +| 128 | blk.10.attn_q.weight | 0xbb430580 | 0xa0c800 | +| 129 | blk.10.attn_v.bias | 0xbbe3cd80 | 0x800 | +| 130 | blk.10.attn_v.weight | 0xbbe3d580 | 0x1dc000 | +| 131 | blk.10.ffn_down.weight | 0xbc019580 | 0x44cc000 | +| 132 | blk.10.ffn_gate.weight | 0xc04e5580 | 0x351d800 | +| 133 | blk.10.ffn_norm.weight | 0xc3a02d80 | 0x3800 | +| 134 | blk.10.ffn_up.weight | 0xc3a06580 | 0x351d800 | +| 135 | blk.11.attn_k.bias | 0xc6f23d80 | 0x800 | +| 136 | blk.11.attn_k.weight | 0xc6f24580 | 0x16f800 | +| 137 | blk.11.attn_norm.weight | 0xc7093d80 | 0x3800 | +| 138 | blk.11.attn_output.weight | 0xc7097580 | 0xd04000 | +| 139 | blk.11.attn_q.bias | 0xc7d9b580 | 0x3800 | +| 140 | blk.11.attn_q.weight | 0xc7d9ed80 | 0xa0c800 | +| 141 | blk.11.attn_v.bias | 0xc87ab580 | 0x800 | +| 142 | blk.11.attn_v.weight | 0xc87abd80 | 0x1dc000 | +| 143 | blk.11.ffn_down.weight | 0xc8987d80 | 0x44cc000 | +| 144 | blk.11.ffn_gate.weight | 0xcce53d80 | 0x351d800 | +| 145 | blk.11.ffn_norm.weight | 0xd0371580 | 0x3800 | +| 146 | blk.11.ffn_up.weight | 0xd0374d80 | 0x351d800 | +| 147 | blk.12.attn_k.bias | 0xd3892580 | 0x800 | +| 148 | blk.12.attn_k.weight | 0xd3892d80 | 0x16f800 | +| 149 | blk.12.attn_norm.weight | 0xd3a02580 | 0x3800 | +| 150 | blk.12.attn_output.weight | 0xd3a05d80 | 0xd04000 | +| 151 | blk.12.attn_q.bias | 0xd4709d80 | 0x3800 | +| 152 | blk.12.attn_q.weight | 0xd470d580 | 0xa0c800 | +| 153 | blk.12.attn_v.bias | 0xd5119d80 | 0x800 | +| 154 | blk.12.attn_v.weight | 0xd511a580 | 0x1dc000 | +| 155 | blk.12.ffn_down.weight | 0xd52f6580 | 0x44cc000 | +| 156 | blk.12.ffn_gate.weight | 0xd97c2580 | 0x351d800 | +| 157 | blk.12.ffn_norm.weight | 0xdccdfd80 | 0x3800 | +| 158 | blk.12.ffn_up.weight | 0xdcce3580 | 0x351d800 | +| 159 | blk.13.attn_k.bias | 0xe0200d80 | 0x800 | +| 160 | blk.13.attn_k.weight | 0xe0201580 | 0x16f800 | +| 161 | blk.13.attn_norm.weight | 0xe0370d80 | 0x3800 | +| 162 | blk.13.attn_output.weight | 0xe0374580 | 0xd04000 | +| 163 | blk.13.attn_q.bias | 0xe1078580 | 0x3800 | +| 164 | blk.13.attn_q.weight | 0xe107bd80 | 0xa0c800 | +| 165 | blk.13.attn_v.bias | 0xe1a88580 | 0x800 | +| 166 | blk.13.attn_v.weight | 0xe1a88d80 | 0x1dc000 | +| 167 | blk.13.ffn_down.weight | 0xe1c64d80 | 0x44cc000 | +| 168 | blk.13.ffn_gate.weight | 0xe6130d80 | 0x351d800 | +| 169 | blk.13.ffn_norm.weight | 0xe964e580 | 0x3800 | +| 170 | blk.13.ffn_up.weight | 0xe9651d80 | 0x351d800 | +| 171 | blk.14.attn_k.bias | 0xecb6f580 | 0x800 | +| 172 | blk.14.attn_k.weight | 0xecb6fd80 | 0x1dc000 | +| 173 | blk.14.attn_norm.weight | 0xecd4bd80 | 0x3800 | +| 174 | blk.14.attn_output.weight | 0xecd4f580 | 0xd04000 | +| 175 | blk.14.attn_q.bias | 0xeda53580 | 0x3800 | +| 176 | blk.14.attn_q.weight | 0xeda56d80 | 0xd04000 | +| 177 | blk.14.attn_v.bias | 0xee75ad80 | 0x800 | +| 178 | blk.14.attn_v.weight | 0xee75b580 | 0x380000 | +| 179 | blk.14.ffn_down.weight | 0xeeadb580 | 0x44cc000 | +| 180 | blk.14.ffn_gate.weight | 0xf2fa7580 | 0x351d800 | +| 181 | blk.14.ffn_norm.weight | 0xf64c4d80 | 0x3800 | +| 182 | blk.14.ffn_up.weight | 0xf64c8580 | 0x351d800 | +| 183 | blk.15.attn_k.bias | 0xf99e5d80 | 0x800 | +| 184 | blk.15.attn_k.weight | 0xf99e6580 | 0x1dc000 | +| 185 | blk.15.attn_norm.weight | 0xf9bc2580 | 0x3800 | +| 186 | blk.15.attn_output.weight | 0xf9bc5d80 | 0xd04000 | +| 187 | blk.15.attn_q.bias | 0xfa8c9d80 | 0x3800 | +| 188 | blk.15.attn_q.weight | 0xfa8cd580 | 0xd04000 | +| 189 | blk.15.attn_v.bias | 0xfb5d1580 | 0x800 | +| 190 | blk.15.attn_v.weight | 0xfb5d1d80 | 0x380000 | +| 191 | blk.15.ffn_down.weight | 0xfb951d80 | 0x44cc000 | +| 192 | blk.15.ffn_gate.weight | 0xffe1dd80 | 0x351d800 | +| 193 | blk.15.ffn_norm.weight | 0x10333b580 | 0x3800 | +| 194 | blk.15.ffn_up.weight | 0x10333ed80 | 0x351d800 | +| 195 | blk.16.attn_k.bias | 0x10685c580 | 0x800 | +| 196 | blk.16.attn_k.weight | 0x10685cd80 | 0x1dc000 | +| 197 | blk.16.attn_norm.weight | 0x106a38d80 | 0x3800 | +| 198 | blk.16.attn_output.weight | 0x106a3c580 | 0xd04000 | +| 199 | blk.16.attn_q.bias | 0x107740580 | 0x3800 | +| 200 | blk.16.attn_q.weight | 0x107743d80 | 0xd04000 | +| 201 | blk.16.attn_v.bias | 0x108447d80 | 0x800 | +| 202 | blk.16.attn_v.weight | 0x108448580 | 0x380000 | +| 203 | blk.16.ffn_down.weight | 0x1087c8580 | 0x44cc000 | +| 204 | blk.16.ffn_gate.weight | 0x10cc94580 | 0x351d800 | +| 205 | blk.16.ffn_norm.weight | 0x1101b1d80 | 0x3800 | +| 206 | blk.16.ffn_up.weight | 0x1101b5580 | 0x351d800 | +| 207 | blk.17.attn_k.bias | 0x1136d2d80 | 0x800 | +| 208 | blk.17.attn_k.weight | 0x1136d3580 | 0x1dc000 | +| 209 | blk.17.attn_norm.weight | 0x1138af580 | 0x3800 | +| 210 | blk.17.attn_output.weight | 0x1138b2d80 | 0xd04000 | +| 211 | blk.17.attn_q.bias | 0x1145b6d80 | 0x3800 | +| 212 | blk.17.attn_q.weight | 0x1145ba580 | 0xd04000 | +| 213 | blk.17.attn_v.bias | 0x1152be580 | 0x800 | +| 214 | blk.17.attn_v.weight | 0x1152bed80 | 0x380000 | +| 215 | blk.17.ffn_down.weight | 0x11563ed80 | 0x44cc000 | +| 216 | blk.17.ffn_gate.weight | 0x119b0ad80 | 0x351d800 | +| 217 | blk.17.ffn_norm.weight | 0x11d028580 | 0x3800 | +| 218 | blk.17.ffn_up.weight | 0x11d02bd80 | 0x351d800 | +| 219 | blk.18.attn_k.bias | 0x120549580 | 0x800 | +| 220 | blk.18.attn_k.weight | 0x120549d80 | 0x1dc000 | +| 221 | blk.18.attn_norm.weight | 0x120725d80 | 0x3800 | +| 222 | blk.18.attn_output.weight | 0x120729580 | 0xd04000 | +| 223 | blk.18.attn_q.bias | 0x12142d580 | 0x3800 | +| 224 | blk.18.attn_q.weight | 0x121430d80 | 0xd04000 | +| 225 | blk.18.attn_v.bias | 0x122134d80 | 0x800 | +| 226 | blk.18.attn_v.weight | 0x122135580 | 0x380000 | +| 227 | blk.18.ffn_down.weight | 0x1224b5580 | 0x44cc000 | +| 228 | blk.18.ffn_gate.weight | 0x126981580 | 0x351d800 | +| 229 | blk.18.ffn_norm.weight | 0x129e9ed80 | 0x3800 | +| 230 | blk.18.ffn_up.weight | 0x129ea2580 | 0x351d800 | +| 231 | blk.19.attn_k.bias | 0x12d3bfd80 | 0x800 | +| 232 | blk.19.attn_k.weight | 0x12d3c0580 | 0x1dc000 | +| 233 | blk.19.attn_norm.weight | 0x12d59c580 | 0x3800 | +| 234 | blk.19.attn_output.weight | 0x12d59fd80 | 0xd04000 | +| 235 | blk.19.attn_q.bias | 0x12e2a3d80 | 0x3800 | +| 236 | blk.19.attn_q.weight | 0x12e2a7580 | 0xd04000 | +| 237 | blk.19.attn_v.bias | 0x12efab580 | 0x800 | +| 238 | blk.19.attn_v.weight | 0x12efabd80 | 0x380000 | +| 239 | blk.19.ffn_down.weight | 0x12f32bd80 | 0x44cc000 | +| 240 | blk.19.ffn_gate.weight | 0x1337f7d80 | 0x351d800 | +| 241 | blk.19.ffn_norm.weight | 0x136d15580 | 0x3800 | +| 242 | blk.19.ffn_up.weight | 0x136d18d80 | 0x351d800 | +| 243 | blk.20.attn_k.bias | 0x13a236580 | 0x800 | +| 244 | blk.20.attn_k.weight | 0x13a236d80 | 0x1dc000 | +| 245 | blk.20.attn_norm.weight | 0x13a412d80 | 0x3800 | +| 246 | blk.20.attn_output.weight | 0x13a416580 | 0xd04000 | +| 247 | blk.20.attn_q.bias | 0x13b11a580 | 0x3800 | +| 248 | blk.20.attn_q.weight | 0x13b11dd80 | 0xd04000 | +| 249 | blk.20.attn_v.bias | 0x13be21d80 | 0x800 | +| 250 | blk.20.attn_v.weight | 0x13be22580 | 0x380000 | +| 251 | blk.20.ffn_down.weight | 0x13c1a2580 | 0x44cc000 | +| 252 | blk.20.ffn_gate.weight | 0x14066e580 | 0x44cc000 | +| 253 | blk.20.ffn_norm.weight | 0x144b3a580 | 0x3800 | +| 254 | blk.20.ffn_up.weight | 0x144b3dd80 | 0x44cc000 | +| 255 | blk.21.attn_k.bias | 0x149009d80 | 0x800 | +| 256 | blk.21.attn_k.weight | 0x14900a580 | 0x1dc000 | +| 257 | blk.21.attn_norm.weight | 0x1491e6580 | 0x3800 | +| 258 | blk.21.attn_output.weight | 0x1491e9d80 | 0xd04000 | +| 259 | blk.21.attn_q.bias | 0x149eedd80 | 0x3800 | +| 260 | blk.21.attn_q.weight | 0x149ef1580 | 0xd04000 | +| 261 | blk.21.attn_v.bias | 0x14abf5580 | 0x800 | +| 262 | blk.21.attn_v.weight | 0x14abf5d80 | 0x380000 | +| 263 | blk.21.ffn_down.weight | 0x14af75d80 | 0x44cc000 | +| 264 | blk.21.ffn_gate.weight | 0x14f441d80 | 0x44cc000 | +| 265 | blk.21.ffn_norm.weight | 0x15390dd80 | 0x3800 | +| 266 | blk.21.ffn_up.weight | 0x153911580 | 0x44cc000 | +| 267 | blk.22.attn_k.bias | 0x157ddd580 | 0x800 | +| 268 | blk.22.attn_k.weight | 0x157dddd80 | 0x1dc000 | +| 269 | blk.22.attn_norm.weight | 0x157fb9d80 | 0x3800 | +| 270 | blk.22.attn_output.weight | 0x157fbd580 | 0xd04000 | +| 271 | blk.22.attn_q.bias | 0x158cc1580 | 0x3800 | +| 272 | blk.22.attn_q.weight | 0x158cc4d80 | 0xd04000 | +| 273 | blk.22.attn_v.bias | 0x1599c8d80 | 0x800 | +| 274 | blk.22.attn_v.weight | 0x1599c9580 | 0x380000 | +| 275 | blk.22.ffn_down.weight | 0x159d49580 | 0x44cc000 | +| 276 | blk.22.ffn_gate.weight | 0x15e215580 | 0x44cc000 | +| 277 | blk.22.ffn_norm.weight | 0x1626e1580 | 0x3800 | +| 278 | blk.22.ffn_up.weight | 0x1626e4d80 | 0x44cc000 | +| 279 | blk.23.attn_k.bias | 0x166bb0d80 | 0x800 | +| 280 | blk.23.attn_k.weight | 0x166bb1580 | 0x1dc000 | +| 281 | blk.23.attn_norm.weight | 0x166d8d580 | 0x3800 | +| 282 | blk.23.attn_output.weight | 0x166d90d80 | 0xd04000 | +| 283 | blk.23.attn_q.bias | 0x167a94d80 | 0x3800 | +| 284 | blk.23.attn_q.weight | 0x167a98580 | 0xd04000 | +| 285 | blk.23.attn_v.bias | 0x16879c580 | 0x800 | +| 286 | blk.23.attn_v.weight | 0x16879cd80 | 0x380000 | +| 287 | blk.23.ffn_down.weight | 0x168b1cd80 | 0x44cc000 | +| 288 | blk.23.ffn_gate.weight | 0x16cfe8d80 | 0x44cc000 | +| 289 | blk.23.ffn_norm.weight | 0x1714b4d80 | 0x3800 | +| 290 | blk.23.ffn_up.weight | 0x1714b8580 | 0x44cc000 | +| 291 | blk.24.attn_k.bias | 0x175984580 | 0x800 | +| 292 | blk.24.attn_k.weight | 0x175984d80 | 0x1dc000 | +| 293 | blk.24.attn_norm.weight | 0x175b60d80 | 0x3800 | +| 294 | blk.24.attn_output.weight | 0x175b64580 | 0xd04000 | +| 295 | blk.24.attn_q.bias | 0x176868580 | 0x3800 | +| 296 | blk.24.attn_q.weight | 0x17686bd80 | 0xd04000 | +| 297 | blk.24.attn_v.bias | 0x17756fd80 | 0x800 | +| 298 | blk.24.attn_v.weight | 0x177570580 | 0x380000 | +| 299 | blk.24.ffn_down.weight | 0x1778f0580 | 0x44cc000 | +| 300 | blk.24.ffn_gate.weight | 0x17bdbc580 | 0x44cc000 | +| 301 | blk.24.ffn_norm.weight | 0x180288580 | 0x3800 | +| 302 | blk.24.ffn_up.weight | 0x18028bd80 | 0x44cc000 | +| 303 | blk.25.attn_k.bias | 0x184757d80 | 0x800 | +| 304 | blk.25.attn_k.weight | 0x184758580 | 0x1dc000 | +| 305 | blk.25.attn_norm.weight | 0x184934580 | 0x3800 | +| 306 | blk.25.attn_output.weight | 0x184937d80 | 0xd04000 | +| 307 | blk.25.attn_q.bias | 0x18563bd80 | 0x3800 | +| 308 | blk.25.attn_q.weight | 0x18563f580 | 0xd04000 | +| 309 | blk.25.attn_v.bias | 0x186343580 | 0x800 | +| 310 | blk.25.attn_v.weight | 0x186343d80 | 0x380000 | +| 311 | blk.25.ffn_down.weight | 0x1866c3d80 | 0x44cc000 | +| 312 | blk.25.ffn_gate.weight | 0x18ab8fd80 | 0x44cc000 | +| 313 | blk.25.ffn_norm.weight | 0x18f05bd80 | 0x3800 | +| 314 | blk.25.ffn_up.weight | 0x18f05f580 | 0x44cc000 | +| 315 | blk.26.attn_k.bias | 0x19352b580 | 0x800 | +| 316 | blk.26.attn_k.weight | 0x19352bd80 | 0x1dc000 | +| 317 | blk.26.attn_norm.weight | 0x193707d80 | 0x3800 | +| 318 | blk.26.attn_output.weight | 0x19370b580 | 0xd04000 | +| 319 | blk.26.attn_q.bias | 0x19440f580 | 0x3800 | +| 320 | blk.26.attn_q.weight | 0x194412d80 | 0xd04000 | +| 321 | blk.26.attn_v.bias | 0x195116d80 | 0x800 | +| 322 | blk.26.attn_v.weight | 0x195117580 | 0x380000 | +| 323 | blk.26.ffn_down.weight | 0x195497580 | 0x44cc000 | +| 324 | blk.26.ffn_gate.weight | 0x199963580 | 0x44cc000 | +| 325 | blk.26.ffn_norm.weight | 0x19de2f580 | 0x3800 | +| 326 | blk.26.ffn_up.weight | 0x19de32d80 | 0x44cc000 | +| 327 | blk.27.attn_k.bias | 0x1a22fed80 | 0x800 | +| 328 | blk.27.attn_k.weight | 0x1a22ff580 | 0x1dc000 | +| 329 | blk.27.attn_norm.weight | 0x1a24db580 | 0x3800 | +| 330 | blk.27.attn_output.weight | 0x1a24ded80 | 0xd04000 | +| 331 | blk.27.attn_q.bias | 0x1a31e2d80 | 0x3800 | +| 332 | blk.27.attn_q.weight | 0x1a31e6580 | 0xd04000 | +| 333 | blk.27.attn_v.bias | 0x1a3eea580 | 0x800 | +| 334 | blk.27.attn_v.weight | 0x1a3eead80 | 0x380000 | +| 335 | blk.27.ffn_down.weight | 0x1a426ad80 | 0x44cc000 | +| 336 | blk.27.ffn_gate.weight | 0x1a8736d80 | 0x44cc000 | +| 337 | blk.27.ffn_norm.weight | 0x1acc02d80 | 0x3800 | +| 338 | blk.27.ffn_up.weight | 0x1acc06580 | 0x44cc000 | + +### Base Tensor Group : ~1B Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------|:---------------------------------|:------------------|:----------------------|:-----| +| 0 | output.weight | Output (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q8_0 | +| 1 | output_norm.weight | Output Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 2 | token_embd.weight | Token Embedding (W) | (~544M) 543567360 | 3584 x 151665 x 1 x 1 | Q3_K | + +- Total elements in base: ( ~1B) 1087138304 +- Percentage of total elements: 14.28% + + +### Block 0 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 3 | blk.0.attn_k.bias | Block 0 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 4 | blk.0.attn_k.weight | Block 0 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 5 | blk.0.attn_norm.weight | Block 0 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 6 | blk.0.attn_output.weight | Block 0 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 7 | blk.0.attn_q.bias | Block 0 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 8 | blk.0.attn_q.weight | Block 0 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 9 | blk.0.attn_v.bias | Block 0 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 10 | blk.0.attn_v.weight | Block 0 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 11 | blk.0.ffn_down.weight | Block 0 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 12 | blk.0.ffn_gate.weight | Block 0 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 13 | blk.0.ffn_norm.weight | Block 0 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 14 | blk.0.ffn_up.weight | Block 0 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.0: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 1 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 15 | blk.1.attn_k.bias | Block 1 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 16 | blk.1.attn_k.weight | Block 1 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 17 | blk.1.attn_norm.weight | Block 1 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 18 | blk.1.attn_output.weight | Block 1 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 19 | blk.1.attn_q.bias | Block 1 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 20 | blk.1.attn_q.weight | Block 1 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 21 | blk.1.attn_v.bias | Block 1 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 22 | blk.1.attn_v.weight | Block 1 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 23 | blk.1.ffn_down.weight | Block 1 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 24 | blk.1.ffn_gate.weight | Block 1 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 25 | blk.1.ffn_norm.weight | Block 1 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 26 | blk.1.ffn_up.weight | Block 1 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.1: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 2 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 27 | blk.2.attn_k.bias | Block 2 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 28 | blk.2.attn_k.weight | Block 2 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 29 | blk.2.attn_norm.weight | Block 2 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 30 | blk.2.attn_output.weight | Block 2 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 31 | blk.2.attn_q.bias | Block 2 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 32 | blk.2.attn_q.weight | Block 2 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 33 | blk.2.attn_v.bias | Block 2 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 34 | blk.2.attn_v.weight | Block 2 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 35 | blk.2.ffn_down.weight | Block 2 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 36 | blk.2.ffn_gate.weight | Block 2 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 37 | blk.2.ffn_norm.weight | Block 2 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 38 | blk.2.ffn_up.weight | Block 2 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.2: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 3 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 39 | blk.3.attn_k.bias | Block 3 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 40 | blk.3.attn_k.weight | Block 3 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 41 | blk.3.attn_norm.weight | Block 3 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 42 | blk.3.attn_output.weight | Block 3 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 43 | blk.3.attn_q.bias | Block 3 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 44 | blk.3.attn_q.weight | Block 3 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 45 | blk.3.attn_v.bias | Block 3 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 46 | blk.3.attn_v.weight | Block 3 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 47 | blk.3.ffn_down.weight | Block 3 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 48 | blk.3.ffn_gate.weight | Block 3 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 49 | blk.3.ffn_norm.weight | Block 3 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 50 | blk.3.ffn_up.weight | Block 3 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.3: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 4 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 51 | blk.4.attn_k.bias | Block 4 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 52 | blk.4.attn_k.weight | Block 4 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 53 | blk.4.attn_norm.weight | Block 4 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 54 | blk.4.attn_output.weight | Block 4 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 55 | blk.4.attn_q.bias | Block 4 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 56 | blk.4.attn_q.weight | Block 4 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 57 | blk.4.attn_v.bias | Block 4 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 58 | blk.4.attn_v.weight | Block 4 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 59 | blk.4.ffn_down.weight | Block 4 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 60 | blk.4.ffn_gate.weight | Block 4 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 61 | blk.4.ffn_norm.weight | Block 4 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 62 | blk.4.ffn_up.weight | Block 4 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.4: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 5 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 63 | blk.5.attn_k.bias | Block 5 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 64 | blk.5.attn_k.weight | Block 5 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 65 | blk.5.attn_norm.weight | Block 5 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 66 | blk.5.attn_output.weight | Block 5 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 67 | blk.5.attn_q.bias | Block 5 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 68 | blk.5.attn_q.weight | Block 5 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 69 | blk.5.attn_v.bias | Block 5 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 70 | blk.5.attn_v.weight | Block 5 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 71 | blk.5.ffn_down.weight | Block 5 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 72 | blk.5.ffn_gate.weight | Block 5 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 73 | blk.5.ffn_norm.weight | Block 5 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 74 | blk.5.ffn_up.weight | Block 5 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.5: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 6 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 75 | blk.6.attn_k.bias | Block 6 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 76 | blk.6.attn_k.weight | Block 6 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 77 | blk.6.attn_norm.weight | Block 6 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 78 | blk.6.attn_output.weight | Block 6 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 79 | blk.6.attn_q.bias | Block 6 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 80 | blk.6.attn_q.weight | Block 6 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 81 | blk.6.attn_v.bias | Block 6 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 82 | blk.6.attn_v.weight | Block 6 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 83 | blk.6.ffn_down.weight | Block 6 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 84 | blk.6.ffn_gate.weight | Block 6 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 85 | blk.6.ffn_norm.weight | Block 6 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 86 | blk.6.ffn_up.weight | Block 6 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.6: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 7 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 87 | blk.7.attn_k.bias | Block 7 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 88 | blk.7.attn_k.weight | Block 7 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 89 | blk.7.attn_norm.weight | Block 7 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 90 | blk.7.attn_output.weight | Block 7 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 91 | blk.7.attn_q.bias | Block 7 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 92 | blk.7.attn_q.weight | Block 7 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 93 | blk.7.attn_v.bias | Block 7 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 94 | blk.7.attn_v.weight | Block 7 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 95 | blk.7.ffn_down.weight | Block 7 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 96 | blk.7.ffn_gate.weight | Block 7 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 97 | blk.7.ffn_norm.weight | Block 7 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 98 | blk.7.ffn_up.weight | Block 7 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.7: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 8 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 99 | blk.8.attn_k.bias | Block 8 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 100 | blk.8.attn_k.weight | Block 8 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 101 | blk.8.attn_norm.weight | Block 8 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 102 | blk.8.attn_output.weight | Block 8 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 103 | blk.8.attn_q.bias | Block 8 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 104 | blk.8.attn_q.weight | Block 8 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 105 | blk.8.attn_v.bias | Block 8 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 106 | blk.8.attn_v.weight | Block 8 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 107 | blk.8.ffn_down.weight | Block 8 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 108 | blk.8.ffn_gate.weight | Block 8 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 109 | blk.8.ffn_norm.weight | Block 8 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 110 | blk.8.ffn_up.weight | Block 8 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.8: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 9 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:-------------------------|:-----------------------------------------------|:----------------|:----------------------|:-----| +| 111 | blk.9.attn_k.bias | Block 9 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 112 | blk.9.attn_k.weight | Block 9 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 113 | blk.9.attn_norm.weight | Block 9 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 114 | blk.9.attn_output.weight | Block 9 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 115 | blk.9.attn_q.bias | Block 9 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 116 | blk.9.attn_q.weight | Block 9 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 117 | blk.9.attn_v.bias | Block 9 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 118 | blk.9.attn_v.weight | Block 9 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 119 | blk.9.ffn_down.weight | Block 9 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 120 | blk.9.ffn_gate.weight | Block 9 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 121 | blk.9.ffn_norm.weight | Block 9 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 122 | blk.9.ffn_up.weight | Block 9 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.9: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 10 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 123 | blk.10.attn_k.bias | Block 10 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 124 | blk.10.attn_k.weight | Block 10 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 125 | blk.10.attn_norm.weight | Block 10 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 126 | blk.10.attn_output.weight | Block 10 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 127 | blk.10.attn_q.bias | Block 10 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 128 | blk.10.attn_q.weight | Block 10 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 129 | blk.10.attn_v.bias | Block 10 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 130 | blk.10.attn_v.weight | Block 10 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 131 | blk.10.ffn_down.weight | Block 10 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 132 | blk.10.ffn_gate.weight | Block 10 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 133 | blk.10.ffn_norm.weight | Block 10 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 134 | blk.10.ffn_up.weight | Block 10 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.10: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 11 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 135 | blk.11.attn_k.bias | Block 11 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 136 | blk.11.attn_k.weight | Block 11 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 137 | blk.11.attn_norm.weight | Block 11 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 138 | blk.11.attn_output.weight | Block 11 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 139 | blk.11.attn_q.bias | Block 11 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 140 | blk.11.attn_q.weight | Block 11 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 141 | blk.11.attn_v.bias | Block 11 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 142 | blk.11.attn_v.weight | Block 11 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 143 | blk.11.ffn_down.weight | Block 11 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 144 | blk.11.ffn_gate.weight | Block 11 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 145 | blk.11.ffn_norm.weight | Block 11 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 146 | blk.11.ffn_up.weight | Block 11 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.11: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 12 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 147 | blk.12.attn_k.bias | Block 12 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 148 | blk.12.attn_k.weight | Block 12 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 149 | blk.12.attn_norm.weight | Block 12 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 150 | blk.12.attn_output.weight | Block 12 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 151 | blk.12.attn_q.bias | Block 12 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 152 | blk.12.attn_q.weight | Block 12 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 153 | blk.12.attn_v.bias | Block 12 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 154 | blk.12.attn_v.weight | Block 12 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 155 | blk.12.ffn_down.weight | Block 12 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 156 | blk.12.ffn_gate.weight | Block 12 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 157 | blk.12.ffn_norm.weight | Block 12 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 158 | blk.12.ffn_up.weight | Block 12 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.12: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 13 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 159 | blk.13.attn_k.bias | Block 13 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 160 | blk.13.attn_k.weight | Block 13 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q6_K | +| 161 | blk.13.attn_norm.weight | Block 13 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 162 | blk.13.attn_output.weight | Block 13 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 163 | blk.13.attn_q.bias | Block 13 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 164 | blk.13.attn_q.weight | Block 13 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q6_K | +| 165 | blk.13.attn_v.bias | Block 13 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 166 | blk.13.attn_v.weight | Block 13 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 167 | blk.13.ffn_down.weight | Block 13 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 168 | blk.13.ffn_gate.weight | Block 13 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 169 | blk.13.ffn_norm.weight | Block 13 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 170 | blk.13.ffn_up.weight | Block 13 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.13: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 14 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 171 | blk.14.attn_k.bias | Block 14 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 172 | blk.14.attn_k.weight | Block 14 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 173 | blk.14.attn_norm.weight | Block 14 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 174 | blk.14.attn_output.weight | Block 14 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 175 | blk.14.attn_q.bias | Block 14 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 176 | blk.14.attn_q.weight | Block 14 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 177 | blk.14.attn_v.bias | Block 14 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 178 | blk.14.attn_v.weight | Block 14 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 179 | blk.14.ffn_down.weight | Block 14 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 180 | blk.14.ffn_gate.weight | Block 14 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 181 | blk.14.ffn_norm.weight | Block 14 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 182 | blk.14.ffn_up.weight | Block 14 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.14: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 15 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 183 | blk.15.attn_k.bias | Block 15 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 184 | blk.15.attn_k.weight | Block 15 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 185 | blk.15.attn_norm.weight | Block 15 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 186 | blk.15.attn_output.weight | Block 15 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 187 | blk.15.attn_q.bias | Block 15 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 188 | blk.15.attn_q.weight | Block 15 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 189 | blk.15.attn_v.bias | Block 15 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 190 | blk.15.attn_v.weight | Block 15 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 191 | blk.15.ffn_down.weight | Block 15 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 192 | blk.15.ffn_gate.weight | Block 15 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 193 | blk.15.ffn_norm.weight | Block 15 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 194 | blk.15.ffn_up.weight | Block 15 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.15: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 16 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 195 | blk.16.attn_k.bias | Block 16 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 196 | blk.16.attn_k.weight | Block 16 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 197 | blk.16.attn_norm.weight | Block 16 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 198 | blk.16.attn_output.weight | Block 16 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 199 | blk.16.attn_q.bias | Block 16 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 200 | blk.16.attn_q.weight | Block 16 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 201 | blk.16.attn_v.bias | Block 16 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 202 | blk.16.attn_v.weight | Block 16 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 203 | blk.16.ffn_down.weight | Block 16 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 204 | blk.16.ffn_gate.weight | Block 16 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 205 | blk.16.ffn_norm.weight | Block 16 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 206 | blk.16.ffn_up.weight | Block 16 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.16: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 17 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 207 | blk.17.attn_k.bias | Block 17 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 208 | blk.17.attn_k.weight | Block 17 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 209 | blk.17.attn_norm.weight | Block 17 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 210 | blk.17.attn_output.weight | Block 17 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 211 | blk.17.attn_q.bias | Block 17 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 212 | blk.17.attn_q.weight | Block 17 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 213 | blk.17.attn_v.bias | Block 17 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 214 | blk.17.attn_v.weight | Block 17 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 215 | blk.17.ffn_down.weight | Block 17 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 216 | blk.17.ffn_gate.weight | Block 17 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 217 | blk.17.ffn_norm.weight | Block 17 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 218 | blk.17.ffn_up.weight | Block 17 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.17: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 18 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 219 | blk.18.attn_k.bias | Block 18 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 220 | blk.18.attn_k.weight | Block 18 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 221 | blk.18.attn_norm.weight | Block 18 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 222 | blk.18.attn_output.weight | Block 18 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 223 | blk.18.attn_q.bias | Block 18 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 224 | blk.18.attn_q.weight | Block 18 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 225 | blk.18.attn_v.bias | Block 18 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 226 | blk.18.attn_v.weight | Block 18 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 227 | blk.18.ffn_down.weight | Block 18 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 228 | blk.18.ffn_gate.weight | Block 18 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 229 | blk.18.ffn_norm.weight | Block 18 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 230 | blk.18.ffn_up.weight | Block 18 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.18: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 19 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 231 | blk.19.attn_k.bias | Block 19 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 232 | blk.19.attn_k.weight | Block 19 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 233 | blk.19.attn_norm.weight | Block 19 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 234 | blk.19.attn_output.weight | Block 19 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 235 | blk.19.attn_q.bias | Block 19 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 236 | blk.19.attn_q.weight | Block 19 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 237 | blk.19.attn_v.bias | Block 19 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 238 | blk.19.attn_v.weight | Block 19 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 239 | blk.19.ffn_down.weight | Block 19 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 240 | blk.19.ffn_gate.weight | Block 19 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | +| 241 | blk.19.ffn_norm.weight | Block 19 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 242 | blk.19.ffn_up.weight | Block 19 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q6_K | + +- Total elements in blk.19: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 20 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 243 | blk.20.attn_k.bias | Block 20 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 244 | blk.20.attn_k.weight | Block 20 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 245 | blk.20.attn_norm.weight | Block 20 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 246 | blk.20.attn_output.weight | Block 20 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 247 | blk.20.attn_q.bias | Block 20 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 248 | blk.20.attn_q.weight | Block 20 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 249 | blk.20.attn_v.bias | Block 20 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 250 | blk.20.attn_v.weight | Block 20 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 251 | blk.20.ffn_down.weight | Block 20 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 252 | blk.20.ffn_gate.weight | Block 20 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 253 | blk.20.ffn_norm.weight | Block 20 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 254 | blk.20.ffn_up.weight | Block 20 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.20: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 21 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 255 | blk.21.attn_k.bias | Block 21 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 256 | blk.21.attn_k.weight | Block 21 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 257 | blk.21.attn_norm.weight | Block 21 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 258 | blk.21.attn_output.weight | Block 21 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 259 | blk.21.attn_q.bias | Block 21 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 260 | blk.21.attn_q.weight | Block 21 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 261 | blk.21.attn_v.bias | Block 21 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 262 | blk.21.attn_v.weight | Block 21 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 263 | blk.21.ffn_down.weight | Block 21 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 264 | blk.21.ffn_gate.weight | Block 21 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 265 | blk.21.ffn_norm.weight | Block 21 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 266 | blk.21.ffn_up.weight | Block 21 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.21: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 22 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 267 | blk.22.attn_k.bias | Block 22 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 268 | blk.22.attn_k.weight | Block 22 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 269 | blk.22.attn_norm.weight | Block 22 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 270 | blk.22.attn_output.weight | Block 22 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 271 | blk.22.attn_q.bias | Block 22 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 272 | blk.22.attn_q.weight | Block 22 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 273 | blk.22.attn_v.bias | Block 22 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 274 | blk.22.attn_v.weight | Block 22 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 275 | blk.22.ffn_down.weight | Block 22 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 276 | blk.22.ffn_gate.weight | Block 22 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 277 | blk.22.ffn_norm.weight | Block 22 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 278 | blk.22.ffn_up.weight | Block 22 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.22: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 23 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 279 | blk.23.attn_k.bias | Block 23 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 280 | blk.23.attn_k.weight | Block 23 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 281 | blk.23.attn_norm.weight | Block 23 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 282 | blk.23.attn_output.weight | Block 23 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 283 | blk.23.attn_q.bias | Block 23 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 284 | blk.23.attn_q.weight | Block 23 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 285 | blk.23.attn_v.bias | Block 23 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 286 | blk.23.attn_v.weight | Block 23 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 287 | blk.23.ffn_down.weight | Block 23 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 288 | blk.23.ffn_gate.weight | Block 23 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 289 | blk.23.ffn_norm.weight | Block 23 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 290 | blk.23.ffn_up.weight | Block 23 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.23: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 24 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 291 | blk.24.attn_k.bias | Block 24 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 292 | blk.24.attn_k.weight | Block 24 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 293 | blk.24.attn_norm.weight | Block 24 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 294 | blk.24.attn_output.weight | Block 24 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 295 | blk.24.attn_q.bias | Block 24 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 296 | blk.24.attn_q.weight | Block 24 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 297 | blk.24.attn_v.bias | Block 24 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 298 | blk.24.attn_v.weight | Block 24 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 299 | blk.24.ffn_down.weight | Block 24 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 300 | blk.24.ffn_gate.weight | Block 24 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 301 | blk.24.ffn_norm.weight | Block 24 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 302 | blk.24.ffn_up.weight | Block 24 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.24: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 25 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 303 | blk.25.attn_k.bias | Block 25 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 304 | blk.25.attn_k.weight | Block 25 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 305 | blk.25.attn_norm.weight | Block 25 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 306 | blk.25.attn_output.weight | Block 25 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 307 | blk.25.attn_q.bias | Block 25 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 308 | blk.25.attn_q.weight | Block 25 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 309 | blk.25.attn_v.bias | Block 25 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 310 | blk.25.attn_v.weight | Block 25 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 311 | blk.25.ffn_down.weight | Block 25 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 312 | blk.25.ffn_gate.weight | Block 25 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 313 | blk.25.ffn_norm.weight | Block 25 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 314 | blk.25.ffn_up.weight | Block 25 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.25: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 26 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 315 | blk.26.attn_k.bias | Block 26 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 316 | blk.26.attn_k.weight | Block 26 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 317 | blk.26.attn_norm.weight | Block 26 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 318 | blk.26.attn_output.weight | Block 26 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 319 | blk.26.attn_q.bias | Block 26 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 320 | blk.26.attn_q.weight | Block 26 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 321 | blk.26.attn_v.bias | Block 26 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 322 | blk.26.attn_v.weight | Block 26 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 323 | blk.26.ffn_down.weight | Block 26 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 324 | blk.26.ffn_gate.weight | Block 26 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 325 | blk.26.ffn_norm.weight | Block 26 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 326 | blk.26.ffn_up.weight | Block 26 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.26: (~233M) 233057792 +- Percentage of total elements: 3.06% + + +### Block 27 Tensor Group : ~233M Elements + +| T_ID | Tensor Layer Name | Human Friendly Tensor Layer Name | Elements | Shape | Type | +|-----:|:--------------------------|:------------------------------------------------|:----------------|:----------------------|:-----| +| 327 | blk.27.attn_k.bias | Block 27 Attention Key (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 328 | blk.27.attn_k.weight | Block 27 Attention Key (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | Q8_0 | +| 329 | blk.27.attn_norm.weight | Block 27 Attention Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 330 | blk.27.attn_output.weight | Block 27 Attention Output (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 331 | blk.27.attn_q.bias | Block 27 Attention Query (B) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 332 | blk.27.attn_q.weight | Block 27 Attention Query (W) | (~13M) 12845056 | 3584 x 3584 x 1 x 1 | Q8_0 | +| 333 | blk.27.attn_v.bias | Block 27 Attention Value (B) | ( 512) 512 | 512 x 1 x 1 x 1 | F32 | +| 334 | blk.27.attn_v.weight | Block 27 Attention Value (W) | ( ~2M) 1835008 | 3584 x 512 x 1 x 1 | F16 | +| 335 | blk.27.ffn_down.weight | Block 27 Feed-Forward Network "Down" (W) | (~68M) 67895296 | 18944 x 3584 x 1 x 1 | Q8_0 | +| 336 | blk.27.ffn_gate.weight | Block 27 Feed-Forward Network "Gate" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | +| 337 | blk.27.ffn_norm.weight | Block 27 Feed-Forward Network Normalization (W) | ( ~4K) 3584 | 3584 x 1 x 1 x 1 | F32 | +| 338 | blk.27.ffn_up.weight | Block 27 Feed-Forward Network "Up" (W) | (~68M) 67895296 | 3584 x 18944 x 1 x 1 | Q8_0 | + +- Total elements in blk.27: (~233M) 233057792 +- Percentage of total elements: 3.06% diff --git a/scores/Hammer2.1-7b-iq3_m.arc b/scores/Hammer2.1-7b-iq3_m.arc new file mode 100644 index 0000000..b85da29 --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_m.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_M.gguf (version GGUF V3 (latest)) + +Final result: 56.5333 +/- 1.8113 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1682.53 ms +llama_perf_context_print: prompt eval time = 106353.34 ms / 35972 tokens ( 2.96 ms per token, 338.23 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 107373.11 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_m.hsw b/scores/Hammer2.1-7b-iq3_m.hsw new file mode 100644 index 0000000..efc4a98 --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_m.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_M.gguf (version GGUF V3 (latest)) + +750 71.46666667% [68.1319%, 74.5827%] + + +llama_perf_context_print: load time = 253.78 ms +llama_perf_context_print: prompt eval time = 375081.86 ms / 126038 tokens ( 2.98 ms per token, 336.03 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 380601.65 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_m.mmlu b/scores/Hammer2.1-7b-iq3_m.mmlu new file mode 100644 index 0000000..1150f9d --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_m.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_M.gguf (version GGUF V3 (latest)) + +Final result: 38.0000 +/- 1.7736 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 252.71 ms +llama_perf_context_print: prompt eval time = 195840.26 ms / 67719 tokens ( 2.89 ms per token, 345.79 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 197498.82 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_m.ppx b/scores/Hammer2.1-7b-iq3_m.ppx new file mode 100644 index 0000000..59e4a0f --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_m.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.827809 ± 0.071236 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 98.97% +Mean ln(PPL(Q)/PPL(base)) : 0.048068 ± 0.001042 +Mean PPL(Q)/PPL(base) : 1.049242 ± 0.001093 +Mean PPL(Q)-PPL(base) : 0.461232 ± 0.011000 + +====== KL divergence statistics ====== +Mean KLD: 0.059970 ± 0.000226 +Maximum KLD: 5.830572 +99.9% KLD: 0.899464 +99.0% KLD: 0.388867 +99.0% KLD: 0.388867 +Median KLD: 0.039933 +10.0% KLD: 0.000901 + 5.0% KLD: 0.000182 + 1.0% KLD: 0.000013 +Minimum KLD: -0.000098 + +====== Token probability statistics ====== +Mean Δp: -0.217 ± 0.017 % +Maximum Δp: 90.326% +99.9% Δp: 34.151% +99.0% Δp: 18.912% +95.0% Δp: 9.734% +90.0% Δp: 5.703% +75.0% Δp: 1.068% +Median Δp: -0.002% +25.0% Δp: -1.292% +10.0% Δp: -6.110% + 5.0% Δp: -10.513% + 1.0% Δp: -22.188% + 0.1% Δp: -43.583% +Minimum Δp: -81.978% +RMS Δp : 6.574 ± 0.032 % +Same top p: 87.097 ± 0.087 % diff --git a/scores/Hammer2.1-7b-iq3_m.tqa b/scores/Hammer2.1-7b-iq3_m.tqa new file mode 100644 index 0000000..7ab1b63 --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_m.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_M.gguf (version GGUF V3 (latest)) + +Final result: 29.8667 +/- 1.6723 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 252.53 ms +llama_perf_context_print: prompt eval time = 151286.74 ms / 49696 tokens ( 3.04 ms per token, 328.49 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 153386.94 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_m.wng b/scores/Hammer2.1-7b-iq3_m.wng new file mode 100644 index 0000000..6d1878f --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_m.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_M.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.6667 +/- 1.6636 + +llama_perf_context_print: load time = 254.13 ms +llama_perf_context_print: prompt eval time = 63679.99 ms / 21448 tokens ( 2.97 ms per token, 336.81 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 64320.92 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_s.arc b/scores/Hammer2.1-7b-iq3_s.arc new file mode 100644 index 0000000..38cb110 --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_s.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_S.gguf (version GGUF V3 (latest)) + +Final result: 55.8667 +/- 1.8143 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1546.80 ms +llama_perf_context_print: prompt eval time = 106501.37 ms / 35972 tokens ( 2.96 ms per token, 337.76 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 107504.12 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_s.hsw b/scores/Hammer2.1-7b-iq3_s.hsw new file mode 100644 index 0000000..c7dee38 --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_s.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_S.gguf (version GGUF V3 (latest)) + +750 70.26666667% [66.8989%, 73.4279%] + + +llama_perf_context_print: load time = 251.76 ms +llama_perf_context_print: prompt eval time = 375564.87 ms / 126038 tokens ( 2.98 ms per token, 335.60 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 381034.82 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_s.mmlu b/scores/Hammer2.1-7b-iq3_s.mmlu new file mode 100644 index 0000000..2b5829d --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_s.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_S.gguf (version GGUF V3 (latest)) + +Final result: 38.6667 +/- 1.7794 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 252.58 ms +llama_perf_context_print: prompt eval time = 196145.96 ms / 67719 tokens ( 2.90 ms per token, 345.25 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 197800.20 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_s.ppx b/scores/Hammer2.1-7b-iq3_s.ppx new file mode 100644 index 0000000..49cc202 --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_s.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.942560 ± 0.071243 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 98.64% +Mean ln(PPL(Q)/PPL(base)) : 0.059677 ± 0.001178 +Mean PPL(Q)/PPL(base) : 1.061493 ± 0.001251 +Mean PPL(Q)-PPL(base) : 0.575982 ± 0.012339 + +====== KL divergence statistics ====== +Mean KLD: 0.079935 ± 0.000289 +Maximum KLD: 5.765717 +99.9% KLD: 1.150883 +99.0% KLD: 0.493128 +99.0% KLD: 0.493128 +Median KLD: 0.053589 +10.0% KLD: 0.001480 + 5.0% KLD: 0.000323 + 1.0% KLD: 0.000027 +Minimum KLD: -0.000038 + +====== Token probability statistics ====== +Mean Δp: -0.885 ± 0.019 % +Maximum Δp: 91.523% +99.9% Δp: 34.667% +99.0% Δp: 19.093% +95.0% Δp: 9.310% +90.0% Δp: 5.250% +75.0% Δp: 0.738% +Median Δp: -0.033% +25.0% Δp: -2.046% +10.0% Δp: -7.921% + 5.0% Δp: -13.116% + 1.0% Δp: -26.866% + 0.1% Δp: -52.295% +Minimum Δp: -90.266% +RMS Δp : 7.445 ± 0.038 % +Same top p: 85.475 ± 0.091 % diff --git a/scores/Hammer2.1-7b-iq3_s.tqa b/scores/Hammer2.1-7b-iq3_s.tqa new file mode 100644 index 0000000..407463a --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_s.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_S.gguf (version GGUF V3 (latest)) + +Final result: 30.9333 +/- 1.6889 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 248.10 ms +llama_perf_context_print: prompt eval time = 151449.03 ms / 49696 tokens ( 3.05 ms per token, 328.14 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 153545.12 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq3_s.wng b/scores/Hammer2.1-7b-iq3_s.wng new file mode 100644 index 0000000..870bd1a --- /dev/null +++ b/scores/Hammer2.1-7b-iq3_s.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ3_S.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.4000 +/- 1.6680 + +llama_perf_context_print: load time = 250.05 ms +llama_perf_context_print: prompt eval time = 63750.53 ms / 21448 tokens ( 2.97 ms per token, 336.44 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 64386.61 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq4_nl.arc b/scores/Hammer2.1-7b-iq4_nl.arc new file mode 100644 index 0000000..b64b04e --- /dev/null +++ b/scores/Hammer2.1-7b-iq4_nl.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ4_NL.gguf (version GGUF V3 (latest)) + +Final result: 57.3333 +/- 1.8072 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1954.91 ms +llama_perf_context_print: prompt eval time = 110236.28 ms / 35972 tokens ( 3.06 ms per token, 326.32 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 111260.63 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq4_nl.hsw b/scores/Hammer2.1-7b-iq4_nl.hsw new file mode 100644 index 0000000..dd1d987 --- /dev/null +++ b/scores/Hammer2.1-7b-iq4_nl.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ4_NL.gguf (version GGUF V3 (latest)) + +750 71.20000000% [67.8576%, 74.3263%] + + +llama_perf_context_print: load time = 270.76 ms +llama_perf_context_print: prompt eval time = 388008.19 ms / 126038 tokens ( 3.08 ms per token, 324.83 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 393490.42 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq4_nl.mmlu b/scores/Hammer2.1-7b-iq4_nl.mmlu new file mode 100644 index 0000000..f979dd7 --- /dev/null +++ b/scores/Hammer2.1-7b-iq4_nl.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ4_NL.gguf (version GGUF V3 (latest)) + +Final result: 38.5333 +/- 1.7783 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 262.83 ms +llama_perf_context_print: prompt eval time = 202888.48 ms / 67719 tokens ( 3.00 ms per token, 333.77 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 204561.85 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq4_nl.ppx b/scores/Hammer2.1-7b-iq4_nl.ppx new file mode 100644 index 0000000..955450d --- /dev/null +++ b/scores/Hammer2.1-7b-iq4_nl.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.585928 ± 0.068060 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.51% +Mean ln(PPL(Q)/PPL(base)) : 0.023148 ± 0.000703 +Mean PPL(Q)/PPL(base) : 1.023418 ± 0.000719 +Mean PPL(Q)-PPL(base) : 0.219350 ± 0.006862 + +====== KL divergence statistics ====== +Mean KLD: 0.026627 ± 0.000108 +Maximum KLD: 2.930657 +99.9% KLD: 0.451824 +99.0% KLD: 0.171826 +99.0% KLD: 0.171826 +Median KLD: 0.017210 +10.0% KLD: 0.000451 + 5.0% KLD: 0.000095 + 1.0% KLD: 0.000007 +Minimum KLD: -0.000060 + +====== Token probability statistics ====== +Mean Δp: -0.523 ± 0.011 % +Maximum Δp: 54.223% +99.9% Δp: 22.680% +99.0% Δp: 11.163% +95.0% Δp: 5.267% +90.0% Δp: 2.899% +75.0% Δp: 0.419% +Median Δp: -0.014% +25.0% Δp: -1.188% +10.0% Δp: -4.623% + 5.0% Δp: -7.618% + 1.0% Δp: -15.106% + 0.1% Δp: -30.059% +Minimum Δp: -89.467% +RMS Δp : 4.346 ± 0.026 % +Same top p: 91.424 ± 0.073 % diff --git a/scores/Hammer2.1-7b-iq4_nl.tqa b/scores/Hammer2.1-7b-iq4_nl.tqa new file mode 100644 index 0000000..2567163 --- /dev/null +++ b/scores/Hammer2.1-7b-iq4_nl.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ4_NL.gguf (version GGUF V3 (latest)) + +Final result: 30.1333 +/- 1.6766 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 260.43 ms +llama_perf_context_print: prompt eval time = 156558.33 ms / 49696 tokens ( 3.15 ms per token, 317.43 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 158668.65 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-iq4_nl.wng b/scores/Hammer2.1-7b-iq4_nl.wng new file mode 100644 index 0000000..adc4eb5 --- /dev/null +++ b/scores/Hammer2.1-7b-iq4_nl.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-IQ4_NL.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 69.8667 +/- 1.6766 + +llama_perf_context_print: load time = 266.77 ms +llama_perf_context_print: prompt eval time = 65947.52 ms / 21448 tokens ( 3.07 ms per token, 325.23 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 66587.55 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_l.arc b/scores/Hammer2.1-7b-q3_k_l.arc new file mode 100644 index 0000000..52784ec --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_l.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_L.gguf (version GGUF V3 (latest)) + +Final result: 57.2000 +/- 1.8079 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1728.15 ms +llama_perf_context_print: prompt eval time = 114247.06 ms / 35972 tokens ( 3.18 ms per token, 314.86 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 115262.68 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_l.hsw b/scores/Hammer2.1-7b-q3_k_l.hsw new file mode 100644 index 0000000..f7ba24b --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_l.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_L.gguf (version GGUF V3 (latest)) + +750 69.86666667% [66.4884%, 73.0424%] + + +llama_perf_context_print: load time = 256.26 ms +llama_perf_context_print: prompt eval time = 403213.28 ms / 126038 tokens ( 3.20 ms per token, 312.58 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 408748.80 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_l.mmlu b/scores/Hammer2.1-7b-q3_k_l.mmlu new file mode 100644 index 0000000..0d8a271 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_l.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_L.gguf (version GGUF V3 (latest)) + +Final result: 36.6667 +/- 1.7608 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 252.71 ms +llama_perf_context_print: prompt eval time = 210303.01 ms / 67719 tokens ( 3.11 ms per token, 322.01 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 211970.11 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_l.ppx b/scores/Hammer2.1-7b-q3_k_l.ppx new file mode 100644 index 0000000..502aed0 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_l.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 10.010842 ± 0.071669 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 98.55% +Mean ln(PPL(Q)/PPL(base)) : 0.066521 ± 0.001216 +Mean PPL(Q)/PPL(base) : 1.068783 ± 0.001299 +Mean PPL(Q)-PPL(base) : 0.644264 ± 0.012880 + +====== KL divergence statistics ====== +Mean KLD: 0.080230 ± 0.000317 +Maximum KLD: 5.784035 +99.9% KLD: 1.294543 +99.0% KLD: 0.528458 +99.0% KLD: 0.528458 +Median KLD: 0.051107 +10.0% KLD: 0.001336 + 5.0% KLD: 0.000286 + 1.0% KLD: 0.000024 +Minimum KLD: -0.000067 + +====== Token probability statistics ====== +Mean Δp: -1.245 ± 0.019 % +Maximum Δp: 81.317% +99.9% Δp: 34.733% +99.0% Δp: 17.968% +95.0% Δp: 8.306% +90.0% Δp: 4.460% +75.0% Δp: 0.555% +Median Δp: -0.050% +25.0% Δp: -2.346% +10.0% Δp: -8.488% + 5.0% Δp: -13.739% + 1.0% Δp: -28.248% + 0.1% Δp: -54.584% +Minimum Δp: -91.716% +RMS Δp : 7.556 ± 0.040 % +Same top p: 85.515 ± 0.091 % diff --git a/scores/Hammer2.1-7b-q3_k_l.tqa b/scores/Hammer2.1-7b-q3_k_l.tqa new file mode 100644 index 0000000..5c49314 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_l.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_L.gguf (version GGUF V3 (latest)) + +Final result: 29.6000 +/- 1.6680 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 253.28 ms +llama_perf_context_print: prompt eval time = 162637.78 ms / 49696 tokens ( 3.27 ms per token, 305.56 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 164733.51 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_l.wng b/scores/Hammer2.1-7b-q3_k_l.wng new file mode 100644 index 0000000..464ed90 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_l.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_L.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.9333 +/- 1.6591 + +llama_perf_context_print: load time = 254.30 ms +llama_perf_context_print: prompt eval time = 68423.01 ms / 21448 tokens ( 3.19 ms per token, 313.46 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 69062.22 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_m.arc b/scores/Hammer2.1-7b-q3_k_m.arc new file mode 100644 index 0000000..ff2815f --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_m.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_M.gguf (version GGUF V3 (latest)) + +Final result: 55.7333 +/- 1.8149 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1623.86 ms +llama_perf_context_print: prompt eval time = 111857.08 ms / 35972 tokens ( 3.11 ms per token, 321.59 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 112882.88 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_m.hsw b/scores/Hammer2.1-7b-q3_k_m.hsw new file mode 100644 index 0000000..ed9d15d --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_m.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_M.gguf (version GGUF V3 (latest)) + +750 69.86666667% [66.4884%, 73.0424%] + + +llama_perf_context_print: load time = 263.27 ms +llama_perf_context_print: prompt eval time = 395258.72 ms / 126038 tokens ( 3.14 ms per token, 318.87 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 400785.88 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_m.mmlu b/scores/Hammer2.1-7b-q3_k_m.mmlu new file mode 100644 index 0000000..c0c197a --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_m.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_M.gguf (version GGUF V3 (latest)) + +Final result: 37.2000 +/- 1.7661 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 248.74 ms +llama_perf_context_print: prompt eval time = 206021.07 ms / 67719 tokens ( 3.04 ms per token, 328.70 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 207679.96 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_m.ppx b/scores/Hammer2.1-7b-q3_k_m.ppx new file mode 100644 index 0000000..88b164f --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_m.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 10.109660 ± 0.072549 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 98.44% +Mean ln(PPL(Q)/PPL(base)) : 0.076344 ± 0.001263 +Mean PPL(Q)/PPL(base) : 1.079333 ± 0.001363 +Mean PPL(Q)-PPL(base) : 0.743083 ± 0.013719 + +====== KL divergence statistics ====== +Mean KLD: 0.086444 ± 0.000342 +Maximum KLD: 6.291467 +99.9% KLD: 1.404206 +99.0% KLD: 0.571756 +99.0% KLD: 0.571756 +Median KLD: 0.054763 +10.0% KLD: 0.001465 + 5.0% KLD: 0.000315 + 1.0% KLD: 0.000028 +Minimum KLD: -0.000235 + +====== Token probability statistics ====== +Mean Δp: -1.387 ± 0.020 % +Maximum Δp: 82.526% +99.9% Δp: 35.574% +99.0% Δp: 18.394% +95.0% Δp: 8.338% +90.0% Δp: 4.407% +75.0% Δp: 0.508% +Median Δp: -0.066% +25.0% Δp: -2.517% +10.0% Δp: -8.904% + 5.0% Δp: -14.322% + 1.0% Δp: -29.340% + 0.1% Δp: -57.123% +Minimum Δp: -93.887% +RMS Δp : 7.845 ± 0.041 % +Same top p: 85.114 ± 0.092 % diff --git a/scores/Hammer2.1-7b-q3_k_m.tqa b/scores/Hammer2.1-7b-q3_k_m.tqa new file mode 100644 index 0000000..e912db0 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_m.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_M.gguf (version GGUF V3 (latest)) + +Final result: 29.8667 +/- 1.6723 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 254.73 ms +llama_perf_context_print: prompt eval time = 159391.21 ms / 49696 tokens ( 3.21 ms per token, 311.79 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 161494.93 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_m.wng b/scores/Hammer2.1-7b-q3_k_m.wng new file mode 100644 index 0000000..1f37620 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_m.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_M.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.4000 +/- 1.6680 + +llama_perf_context_print: load time = 248.84 ms +llama_perf_context_print: prompt eval time = 67078.03 ms / 21448 tokens ( 3.13 ms per token, 319.75 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 67715.29 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_s.arc b/scores/Hammer2.1-7b-q3_k_s.arc new file mode 100644 index 0000000..f6a8ce5 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_s.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_S.gguf (version GGUF V3 (latest)) + +Final result: 56.5333 +/- 1.8113 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1475.41 ms +llama_perf_context_print: prompt eval time = 113557.50 ms / 35972 tokens ( 3.16 ms per token, 316.77 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 114575.72 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_s.hsw b/scores/Hammer2.1-7b-q3_k_s.hsw new file mode 100644 index 0000000..df91c70 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_s.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_S.gguf (version GGUF V3 (latest)) + +750 68.80000000% [65.3955%, 72.0129%] + + +llama_perf_context_print: load time = 247.18 ms +llama_perf_context_print: prompt eval time = 400796.81 ms / 126038 tokens ( 3.18 ms per token, 314.47 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 406267.35 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_s.mmlu b/scores/Hammer2.1-7b-q3_k_s.mmlu new file mode 100644 index 0000000..4e2654f --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_s.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_S.gguf (version GGUF V3 (latest)) + +Final result: 37.3333 +/- 1.7674 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 252.83 ms +llama_perf_context_print: prompt eval time = 209065.30 ms / 67719 tokens ( 3.09 ms per token, 323.91 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 210718.75 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_s.ppx b/scores/Hammer2.1-7b-q3_k_s.ppx new file mode 100644 index 0000000..14b55e6 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_s.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 10.224307 ± 0.073520 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 98.16% +Mean ln(PPL(Q)/PPL(base)) : 0.087620 ± 0.001374 +Mean PPL(Q)/PPL(base) : 1.091573 ± 0.001499 +Mean PPL(Q)-PPL(base) : 0.857730 ± 0.015181 + +====== KL divergence statistics ====== +Mean KLD: 0.101951 ± 0.000397 +Maximum KLD: 6.073646 +99.9% KLD: 1.662331 +99.0% KLD: 0.673299 +99.0% KLD: 0.673299 +Median KLD: 0.064775 +10.0% KLD: 0.001756 + 5.0% KLD: 0.000385 + 1.0% KLD: 0.000033 +Minimum KLD: -0.000036 + +====== Token probability statistics ====== +Mean Δp: -1.602 ± 0.022 % +Maximum Δp: 88.808% +99.9% Δp: 37.347% +99.0% Δp: 19.623% +95.0% Δp: 8.796% +90.0% Δp: 4.589% +75.0% Δp: 0.491% +Median Δp: -0.086% +25.0% Δp: -2.849% +10.0% Δp: -9.793% + 5.0% Δp: -15.740% + 1.0% Δp: -31.943% + 0.1% Δp: -61.801% +Minimum Δp: -98.023% +RMS Δp : 8.487 ± 0.044 % +Same top p: 84.095 ± 0.095 % diff --git a/scores/Hammer2.1-7b-q3_k_s.tqa b/scores/Hammer2.1-7b-q3_k_s.tqa new file mode 100644 index 0000000..8dc7c85 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_s.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_S.gguf (version GGUF V3 (latest)) + +Final result: 29.0667 +/- 1.6591 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 251.28 ms +llama_perf_context_print: prompt eval time = 161651.93 ms / 49696 tokens ( 3.25 ms per token, 307.43 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 163734.79 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q3_k_s.wng b/scores/Hammer2.1-7b-q3_k_s.wng new file mode 100644 index 0000000..11ed992 --- /dev/null +++ b/scores/Hammer2.1-7b-q3_k_s.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q3_K_S.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 69.3333 +/- 1.6849 + +llama_perf_context_print: load time = 249.55 ms +llama_perf_context_print: prompt eval time = 68038.26 ms / 21448 tokens ( 3.17 ms per token, 315.23 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 68675.52 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_m.arc b/scores/Hammer2.1-7b-q4_k_m.arc new file mode 100644 index 0000000..b468a26 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_m.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_M.gguf (version GGUF V3 (latest)) + +Final result: 57.8667 +/- 1.8042 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1953.65 ms +llama_perf_context_print: prompt eval time = 115394.35 ms / 35972 tokens ( 3.21 ms per token, 311.73 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 116424.56 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_m.hsw b/scores/Hammer2.1-7b-q4_k_m.hsw new file mode 100644 index 0000000..72d1841 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_m.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_M.gguf (version GGUF V3 (latest)) + +750 71.20000000% [67.8576%, 74.3263%] + + +llama_perf_context_print: load time = 260.90 ms +llama_perf_context_print: prompt eval time = 405989.14 ms / 126038 tokens ( 3.22 ms per token, 310.45 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 411436.74 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_m.mmlu b/scores/Hammer2.1-7b-q4_k_m.mmlu new file mode 100644 index 0000000..5b70615 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_m.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_M.gguf (version GGUF V3 (latest)) + +Final result: 38.0000 +/- 1.7736 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 260.22 ms +llama_perf_context_print: prompt eval time = 212336.08 ms / 67719 tokens ( 3.14 ms per token, 318.92 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 214000.87 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_m.ppx b/scores/Hammer2.1-7b-q4_k_m.ppx new file mode 100644 index 0000000..5702523 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_m.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.513981 ± 0.067589 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.56% +Mean ln(PPL(Q)/PPL(base)) : 0.015615 ± 0.000668 +Mean PPL(Q)/PPL(base) : 1.015737 ± 0.000679 +Mean PPL(Q)-PPL(base) : 0.147404 ± 0.006420 + +====== KL divergence statistics ====== +Mean KLD: 0.023792 ± 0.000107 +Maximum KLD: 5.229656 +99.9% KLD: 0.422463 +99.0% KLD: 0.159748 +99.0% KLD: 0.159748 +Median KLD: 0.014752 +10.0% KLD: 0.000388 + 5.0% KLD: 0.000078 + 1.0% KLD: 0.000005 +Minimum KLD: -0.000044 + +====== Token probability statistics ====== +Mean Δp: -0.337 ± 0.010 % +Maximum Δp: 67.097% +99.9% Δp: 22.363% +99.0% Δp: 11.173% +95.0% Δp: 5.170% +90.0% Δp: 2.864% +75.0% Δp: 0.466% +Median Δp: -0.008% +25.0% Δp: -0.987% +10.0% Δp: -3.948% + 5.0% Δp: -6.542% + 1.0% Δp: -13.717% + 0.1% Δp: -28.959% +Minimum Δp: -91.120% +RMS Δp : 4.057 ± 0.026 % +Same top p: 92.039 ± 0.070 % diff --git a/scores/Hammer2.1-7b-q4_k_m.tqa b/scores/Hammer2.1-7b-q4_k_m.tqa new file mode 100644 index 0000000..96cfa94 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_m.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_M.gguf (version GGUF V3 (latest)) + +Final result: 29.8667 +/- 1.6723 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 273.17 ms +llama_perf_context_print: prompt eval time = 163807.79 ms / 49696 tokens ( 3.30 ms per token, 303.38 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 165910.76 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_m.wng b/scores/Hammer2.1-7b-q4_k_m.wng new file mode 100644 index 0000000..49af288 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_m.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_M.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 69.8667 +/- 1.6766 + +llama_perf_context_print: load time = 263.91 ms +llama_perf_context_print: prompt eval time = 69022.90 ms / 21448 tokens ( 3.22 ms per token, 310.74 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 69661.76 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_s.arc b/scores/Hammer2.1-7b-q4_k_s.arc new file mode 100644 index 0000000..6320cdf --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_s.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_S.gguf (version GGUF V3 (latest)) + +Final result: 57.8667 +/- 1.8042 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 1866.16 ms +llama_perf_context_print: prompt eval time = 113639.43 ms / 35972 tokens ( 3.16 ms per token, 316.55 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 114663.24 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_s.hsw b/scores/Hammer2.1-7b-q4_k_s.hsw new file mode 100644 index 0000000..985c8dd --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_s.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_S.gguf (version GGUF V3 (latest)) + +750 71.06666667% [67.7206%, 74.1981%] + + +llama_perf_context_print: load time = 262.17 ms +llama_perf_context_print: prompt eval time = 400033.89 ms / 126038 tokens ( 3.17 ms per token, 315.07 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 405454.55 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_s.mmlu b/scores/Hammer2.1-7b-q4_k_s.mmlu new file mode 100644 index 0000000..dc2bb47 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_s.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_S.gguf (version GGUF V3 (latest)) + +Final result: 37.8667 +/- 1.7724 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 258.75 ms +llama_perf_context_print: prompt eval time = 209132.23 ms / 67719 tokens ( 3.09 ms per token, 323.81 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 210787.92 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_s.ppx b/scores/Hammer2.1-7b-q4_k_s.ppx new file mode 100644 index 0000000..ca1ee41 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_s.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.529467 ± 0.067693 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.53% +Mean ln(PPL(Q)/PPL(base)) : 0.017241 ± 0.000692 +Mean PPL(Q)/PPL(base) : 1.017391 ± 0.000704 +Mean PPL(Q)-PPL(base) : 0.162890 ± 0.006661 + +====== KL divergence statistics ====== +Mean KLD: 0.025429 ± 0.000116 +Maximum KLD: 6.250518 +99.9% KLD: 0.457138 +99.0% KLD: 0.169697 +99.0% KLD: 0.169697 +Median KLD: 0.015727 +10.0% KLD: 0.000429 + 5.0% KLD: 0.000085 + 1.0% KLD: 0.000006 +Minimum KLD: -0.000044 + +====== Token probability statistics ====== +Mean Δp: -0.378 ± 0.011 % +Maximum Δp: 64.248% +99.9% Δp: 23.260% +99.0% Δp: 11.368% +95.0% Δp: 5.279% +90.0% Δp: 2.904% +75.0% Δp: 0.455% +Median Δp: -0.010% +25.0% Δp: -1.037% +10.0% Δp: -4.112% + 5.0% Δp: -6.788% + 1.0% Δp: -14.402% + 0.1% Δp: -30.405% +Minimum Δp: -91.281% +RMS Δp : 4.191 ± 0.026 % +Same top p: 91.792 ± 0.071 % diff --git a/scores/Hammer2.1-7b-q4_k_s.tqa b/scores/Hammer2.1-7b-q4_k_s.tqa new file mode 100644 index 0000000..59dfc9e --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_s.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_S.gguf (version GGUF V3 (latest)) + +Final result: 29.6000 +/- 1.6680 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 264.10 ms +llama_perf_context_print: prompt eval time = 161357.40 ms / 49696 tokens ( 3.25 ms per token, 307.99 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 163438.71 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q4_k_s.wng b/scores/Hammer2.1-7b-q4_k_s.wng new file mode 100644 index 0000000..d026664 --- /dev/null +++ b/scores/Hammer2.1-7b-q4_k_s.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q4_K_S.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 69.6000 +/- 1.6807 + +llama_perf_context_print: load time = 262.82 ms +llama_perf_context_print: prompt eval time = 67973.27 ms / 21448 tokens ( 3.17 ms per token, 315.54 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 68612.34 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_m.arc b/scores/Hammer2.1-7b-q5_k_m.arc new file mode 100644 index 0000000..3cd9380 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_m.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_M.gguf (version GGUF V3 (latest)) + +Final result: 58.0000 +/- 1.8034 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 2350.20 ms +llama_perf_context_print: prompt eval time = 114769.46 ms / 35972 tokens ( 3.19 ms per token, 313.43 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 115764.61 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_m.hsw b/scores/Hammer2.1-7b-q5_k_m.hsw new file mode 100644 index 0000000..e405f8c --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_m.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_M.gguf (version GGUF V3 (latest)) + +750 71.33333333% [67.9947%, 74.4545%] + + +llama_perf_context_print: load time = 287.60 ms +llama_perf_context_print: prompt eval time = 405854.89 ms / 126038 tokens ( 3.22 ms per token, 310.55 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 411351.99 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_m.mmlu b/scores/Hammer2.1-7b-q5_k_m.mmlu new file mode 100644 index 0000000..b6a3bf3 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_m.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_M.gguf (version GGUF V3 (latest)) + +Final result: 38.6667 +/- 1.7794 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 281.78 ms +llama_perf_context_print: prompt eval time = 211602.07 ms / 67719 tokens ( 3.12 ms per token, 320.03 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 213278.72 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_m.ppx b/scores/Hammer2.1-7b-q5_k_m.ppx new file mode 100644 index 0000000..6827688 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_m.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.411472 ± 0.066924 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.88% +Mean ln(PPL(Q)/PPL(base)) : 0.004782 ± 0.000343 +Mean PPL(Q)/PPL(base) : 1.004793 ± 0.000345 +Mean PPL(Q)-PPL(base) : 0.044894 ± 0.003261 + +====== KL divergence statistics ====== +Mean KLD: 0.005881 ± 0.000035 +Maximum KLD: 2.356096 +99.9% KLD: 0.110341 +99.0% KLD: 0.039253 +99.0% KLD: 0.039253 +Median KLD: 0.003644 +10.0% KLD: 0.000086 + 5.0% KLD: 0.000015 + 1.0% KLD: -0.000000 +Minimum KLD: -0.000153 + +====== Token probability statistics ====== +Mean Δp: -0.012 ± 0.005 % +Maximum Δp: 40.226% +99.9% Δp: 11.947% +99.0% Δp: 6.009% +95.0% Δp: 2.947% +90.0% Δp: 1.726% +75.0% Δp: 0.356% +Median Δp: -0.000% +25.0% Δp: -0.356% +10.0% Δp: -1.732% + 5.0% Δp: -2.967% + 1.0% Δp: -6.302% + 0.1% Δp: -13.317% +Minimum Δp: -62.412% +RMS Δp : 2.028 ± 0.015 % +Same top p: 95.904 ± 0.051 % diff --git a/scores/Hammer2.1-7b-q5_k_m.tqa b/scores/Hammer2.1-7b-q5_k_m.tqa new file mode 100644 index 0000000..47cc0a1 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_m.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_M.gguf (version GGUF V3 (latest)) + +Final result: 30.2667 +/- 1.6787 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 279.86 ms +llama_perf_context_print: prompt eval time = 163647.89 ms / 49696 tokens ( 3.29 ms per token, 303.68 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 165759.65 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_m.wng b/scores/Hammer2.1-7b-q5_k_m.wng new file mode 100644 index 0000000..769edc5 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_m.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_M.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.6667 +/- 1.6636 + +llama_perf_context_print: load time = 283.78 ms +llama_perf_context_print: prompt eval time = 68831.62 ms / 21448 tokens ( 3.21 ms per token, 311.60 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 69470.43 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_s.arc b/scores/Hammer2.1-7b-q5_k_s.arc new file mode 100644 index 0000000..289e90a --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_s.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_S.gguf (version GGUF V3 (latest)) + +Final result: 58.4000 +/- 1.8010 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 2246.68 ms +llama_perf_context_print: prompt eval time = 116329.15 ms / 35972 tokens ( 3.23 ms per token, 309.23 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 117348.24 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_s.hsw b/scores/Hammer2.1-7b-q5_k_s.hsw new file mode 100644 index 0000000..f4ecd2f --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_s.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_S.gguf (version GGUF V3 (latest)) + +750 71.46666667% [68.1319%, 74.5827%] + + +llama_perf_context_print: load time = 275.63 ms +llama_perf_context_print: prompt eval time = 410573.07 ms / 126038 tokens ( 3.26 ms per token, 306.98 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 416121.78 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_s.mmlu b/scores/Hammer2.1-7b-q5_k_s.mmlu new file mode 100644 index 0000000..c52041b --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_s.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_S.gguf (version GGUF V3 (latest)) + +Final result: 38.8000 +/- 1.7805 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 277.93 ms +llama_perf_context_print: prompt eval time = 214179.80 ms / 67719 tokens ( 3.16 ms per token, 316.18 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 215843.77 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_s.ppx b/scores/Hammer2.1-7b-q5_k_s.ppx new file mode 100644 index 0000000..b700502 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_s.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.410802 ± 0.066883 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.88% +Mean ln(PPL(Q)/PPL(base)) : 0.004710 ± 0.000355 +Mean PPL(Q)/PPL(base) : 1.004722 ± 0.000357 +Mean PPL(Q)-PPL(base) : 0.044224 ± 0.003363 + +====== KL divergence statistics ====== +Mean KLD: 0.006252 ± 0.000036 +Maximum KLD: 2.202329 +99.9% KLD: 0.113542 +99.0% KLD: 0.041912 +99.0% KLD: 0.041912 +Median KLD: 0.003874 +10.0% KLD: 0.000095 + 5.0% KLD: 0.000017 + 1.0% KLD: 0.000000 +Minimum KLD: -0.000135 + +====== Token probability statistics ====== +Mean Δp: -0.049 ± 0.005 % +Maximum Δp: 48.849% +99.9% Δp: 12.232% +99.0% Δp: 6.155% +95.0% Δp: 2.940% +90.0% Δp: 1.698% +75.0% Δp: 0.326% +Median Δp: -0.000% +25.0% Δp: -0.399% +10.0% Δp: -1.846% + 5.0% Δp: -3.125% + 1.0% Δp: -6.593% + 0.1% Δp: -13.502% +Minimum Δp: -61.344% +RMS Δp : 2.085 ± 0.015 % +Same top p: 95.765 ± 0.052 % diff --git a/scores/Hammer2.1-7b-q5_k_s.tqa b/scores/Hammer2.1-7b-q5_k_s.tqa new file mode 100644 index 0000000..49f7418 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_s.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_S.gguf (version GGUF V3 (latest)) + +Final result: 30.5333 +/- 1.6828 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 285.20 ms +llama_perf_context_print: prompt eval time = 165611.98 ms / 49696 tokens ( 3.33 ms per token, 300.07 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 167719.47 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q5_k_s.wng b/scores/Hammer2.1-7b-q5_k_s.wng new file mode 100644 index 0000000..81f8216 --- /dev/null +++ b/scores/Hammer2.1-7b-q5_k_s.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q5_K_S.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.6667 +/- 1.6636 + +llama_perf_context_print: load time = 282.74 ms +llama_perf_context_print: prompt eval time = 69663.19 ms / 21448 tokens ( 3.25 ms per token, 307.88 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 70297.96 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q6_k.arc b/scores/Hammer2.1-7b-q6_k.arc new file mode 100644 index 0000000..5da0741 --- /dev/null +++ b/scores/Hammer2.1-7b-q6_k.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q6_K.gguf (version GGUF V3 (latest)) + +Final result: 58.8000 +/- 1.7984 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 2767.04 ms +llama_perf_context_print: prompt eval time = 112965.93 ms / 35972 tokens ( 3.14 ms per token, 318.43 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 113983.11 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q6_k.hsw b/scores/Hammer2.1-7b-q6_k.hsw new file mode 100644 index 0000000..b29d5f7 --- /dev/null +++ b/scores/Hammer2.1-7b-q6_k.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q6_K.gguf (version GGUF V3 (latest)) + +750 72.00000000% [68.6807%, 75.0950%] + + +llama_perf_context_print: load time = 303.64 ms +llama_perf_context_print: prompt eval time = 399056.52 ms / 126038 tokens ( 3.17 ms per token, 315.84 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 404573.51 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q6_k.mmlu b/scores/Hammer2.1-7b-q6_k.mmlu new file mode 100644 index 0000000..b124d24 --- /dev/null +++ b/scores/Hammer2.1-7b-q6_k.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q6_K.gguf (version GGUF V3 (latest)) + +Final result: 38.9333 +/- 1.7816 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 309.26 ms +llama_perf_context_print: prompt eval time = 208947.61 ms / 67719 tokens ( 3.09 ms per token, 324.10 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 210629.00 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q6_k.ppx b/scores/Hammer2.1-7b-q6_k.ppx new file mode 100644 index 0000000..c4834eb --- /dev/null +++ b/scores/Hammer2.1-7b-q6_k.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.384561 ± 0.066615 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.96% +Mean ln(PPL(Q)/PPL(base)) : 0.001918 ± 0.000206 +Mean PPL(Q)/PPL(base) : 1.001920 ± 0.000207 +Mean PPL(Q)-PPL(base) : 0.017984 ± 0.001945 + +====== KL divergence statistics ====== +Mean KLD: 0.001779 ± 0.000018 +Maximum KLD: 1.977490 +99.9% KLD: 0.040032 +99.0% KLD: 0.012342 +99.0% KLD: 0.012342 +Median KLD: 0.001053 +10.0% KLD: 0.000028 + 5.0% KLD: 0.000005 + 1.0% KLD: -0.000001 +Minimum KLD: -0.000152 + +====== Token probability statistics ====== +Mean Δp: -0.027 ± 0.003 % +Maximum Δp: 49.334% +99.9% Δp: 7.267% +99.0% Δp: 3.236% +95.0% Δp: 1.482% +90.0% Δp: 0.845% +75.0% Δp: 0.164% +Median Δp: -0.000% +25.0% Δp: -0.202% +10.0% Δp: -0.938% + 5.0% Δp: -1.620% + 1.0% Δp: -3.443% + 0.1% Δp: -7.356% +Minimum Δp: -46.022% +RMS Δp : 1.101 ± 0.012 % +Same top p: 97.794 ± 0.038 % diff --git a/scores/Hammer2.1-7b-q6_k.tqa b/scores/Hammer2.1-7b-q6_k.tqa new file mode 100644 index 0000000..0e98d80 --- /dev/null +++ b/scores/Hammer2.1-7b-q6_k.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q6_K.gguf (version GGUF V3 (latest)) + +Final result: 29.7333 +/- 1.6702 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 318.58 ms +llama_perf_context_print: prompt eval time = 161421.27 ms / 49696 tokens ( 3.25 ms per token, 307.87 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 163513.36 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q6_k.wng b/scores/Hammer2.1-7b-q6_k.wng new file mode 100644 index 0000000..f20eda4 --- /dev/null +++ b/scores/Hammer2.1-7b-q6_k.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q6_K.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.4000 +/- 1.6680 + +llama_perf_context_print: load time = 313.52 ms +llama_perf_context_print: prompt eval time = 67409.09 ms / 21448 tokens ( 3.14 ms per token, 318.18 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 68029.15 ms / 21449 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q8_0.arc b/scores/Hammer2.1-7b-q8_0.arc new file mode 100644 index 0000000..b057cbd --- /dev/null +++ b/scores/Hammer2.1-7b-q8_0.arc @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q8_0.gguf (version GGUF V3 (latest)) + +Final result: 58.5333 +/- 1.8002 +Random chance: 25.0083 +/- 1.5824 + + +llama_perf_context_print: load time = 3247.05 ms +llama_perf_context_print: prompt eval time = 106606.41 ms / 35972 tokens ( 2.96 ms per token, 337.43 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 107624.02 ms / 35973 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q8_0.hsw b/scores/Hammer2.1-7b-q8_0.hsw new file mode 100644 index 0000000..b6cf1fe --- /dev/null +++ b/scores/Hammer2.1-7b-q8_0.hsw @@ -0,0 +1,12 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q8_0.gguf (version GGUF V3 (latest)) + +750 72.00000000% [68.6807%, 75.0950%] + + +llama_perf_context_print: load time = 318.54 ms +llama_perf_context_print: prompt eval time = 375734.27 ms / 126038 tokens ( 2.98 ms per token, 335.44 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 381209.78 ms / 126039 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q8_0.mmlu b/scores/Hammer2.1-7b-q8_0.mmlu new file mode 100644 index 0000000..e55cdca --- /dev/null +++ b/scores/Hammer2.1-7b-q8_0.mmlu @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q8_0.gguf (version GGUF V3 (latest)) + +Final result: 38.9333 +/- 1.7816 +Random chance: 25.0000 +/- 1.5822 + + +llama_perf_context_print: load time = 316.60 ms +llama_perf_context_print: prompt eval time = 197007.43 ms / 67719 tokens ( 2.91 ms per token, 343.74 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 198698.88 ms / 67720 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q8_0.ppx b/scores/Hammer2.1-7b-q8_0.ppx new file mode 100644 index 0000000..65ceb10 --- /dev/null +++ b/scores/Hammer2.1-7b-q8_0.ppx @@ -0,0 +1,37 @@ +====== Perplexity statistics ====== +Mean PPL(Q) : 9.379380 ± 0.066571 +Mean PPL(base) : 9.366577 ± 0.066397 +Cor(ln(PPL(Q)), ln(PPL(base))): 99.98% +Mean ln(PPL(Q)/PPL(base)) : 0.001366 ± 0.000154 +Mean PPL(Q)/PPL(base) : 1.001367 ± 0.000154 +Mean PPL(Q)-PPL(base) : 0.012802 ± 0.001454 + +====== KL divergence statistics ====== +Mean KLD: 0.000692 ± 0.000012 +Maximum KLD: 1.498512 +99.9% KLD: 0.024913 +99.0% KLD: 0.005779 +99.0% KLD: 0.005779 +Median KLD: 0.000331 +10.0% KLD: 0.000011 + 5.0% KLD: 0.000002 + 1.0% KLD: -0.000002 +Minimum KLD: -0.000134 + +====== Token probability statistics ====== +Mean Δp: -0.028 ± 0.002 % +Maximum Δp: 21.822% +99.9% Δp: 4.935% +99.0% Δp: 1.949% +95.0% Δp: 0.826% +90.0% Δp: 0.458% +75.0% Δp: 0.085% +Median Δp: -0.000% +25.0% Δp: -0.129% +10.0% Δp: -0.558% + 5.0% Δp: -0.952% + 1.0% Δp: -2.125% + 0.1% Δp: -5.152% +Minimum Δp: -39.341% +RMS Δp : 0.706 ± 0.010 % +Same top p: 98.711 ± 0.029 % diff --git a/scores/Hammer2.1-7b-q8_0.tqa b/scores/Hammer2.1-7b-q8_0.tqa new file mode 100644 index 0000000..4b62319 --- /dev/null +++ b/scores/Hammer2.1-7b-q8_0.tqa @@ -0,0 +1,13 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q8_0.gguf (version GGUF V3 (latest)) + +Final result: 30.1333 +/- 1.6766 +Random chance: 19.8992 +/- 1.4588 + + +llama_perf_context_print: load time = 324.70 ms +llama_perf_context_print: prompt eval time = 151947.90 ms / 49696 tokens ( 3.06 ms per token, 327.06 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 154045.90 ms / 49697 tokens +ggml_metal_free: deallocating diff --git a/scores/Hammer2.1-7b-q8_0.wng b/scores/Hammer2.1-7b-q8_0.wng new file mode 100644 index 0000000..d138ac4 --- /dev/null +++ b/scores/Hammer2.1-7b-q8_0.wng @@ -0,0 +1,11 @@ +build: 5170 (658987cf) with Apple clang version 17.0.0 (clang-1700.0.13.3) for arm64-apple-darwin24.4.0 +llama_model_load_from_file_impl: using device Metal (Apple M3 Pro) - 27647 MiB free +llama_model_loader: loaded meta data with 33 key-value pairs and 339 tensors from ./Hammer2.1-7b-Q8_0.gguf (version GGUF V3 (latest)) + +Final Winogrande score(750 tasks): 70.2667 +/- 1.6702 + +llama_perf_context_print: load time = 321.25 ms +llama_perf_context_print: prompt eval time = 63938.77 ms / 21448 tokens ( 2.98 ms per token, 335.45 tokens per second) +llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) +llama_perf_context_print: total time = 64599.27 ms / 21449 tokens +ggml_metal_free: deallocating