commit 3f613f525a05648a07ed578f6be5e69c3a355214 Author: ModelHub XC Date: Tue Jun 16 18:58:16 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: mradermacher/granite-3.1-2b-Reasoning-i1-GGUF Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..24070d7 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,60 @@ +*.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 +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 +imatrix.dat filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ1_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ1_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ2_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ2_XS.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ2_XXS.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ3_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q2_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +granite-3.1-2b-Reasoning.i1-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..05ab656 --- /dev/null +++ b/README.md @@ -0,0 +1,83 @@ +--- +base_model: ruslanmv/granite-3.1-2b-Reasoning +language: +- en +library_name: transformers +license: apache-2.0 +quantized_by: mradermacher +tags: +- text-generation-inference +- transformers +- granite +- trl +- grpo +- ruslanmv +--- +## About + + + + + + +weighted/imatrix quants of https://huggingface.co/ruslanmv/granite-3.1-2b-Reasoning + + +static quants are available at https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-GGUF +## Usage + +If you are unsure how to use GGUF files, refer to one of [TheBloke's +READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for +more details, including on how to concatenate multi-part files. + +## Provided Quants + +(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) + +| Link | Type | Size/GB | Notes | +|:-----|:-----|--------:|:------| +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ1_S.gguf) | i1-IQ1_S | 0.7 | for the desperate | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ1_M.gguf) | i1-IQ1_M | 0.7 | mostly desperate | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 0.8 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ2_XS.gguf) | i1-IQ2_XS | 0.9 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ2_S.gguf) | i1-IQ2_S | 0.9 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ2_M.gguf) | i1-IQ2_M | 1.0 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q2_K_S.gguf) | i1-Q2_K_S | 1.0 | very low quality | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q2_K.gguf) | i1-Q2_K | 1.1 | IQ3_XXS probably better | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 1.1 | lower quality | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ3_XS.gguf) | i1-IQ3_XS | 1.2 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q3_K_S.gguf) | i1-Q3_K_S | 1.2 | IQ3_XS probably better | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ3_S.gguf) | i1-IQ3_S | 1.2 | beats Q3_K* | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ3_M.gguf) | i1-IQ3_M | 1.3 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q3_K_M.gguf) | i1-Q3_K_M | 1.4 | IQ3_S probably better | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q3_K_L.gguf) | i1-Q3_K_L | 1.5 | IQ3_M probably better | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ4_XS.gguf) | i1-IQ4_XS | 1.5 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-IQ4_NL.gguf) | i1-IQ4_NL | 1.6 | prefer IQ4_XS | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q4_0.gguf) | i1-Q4_0 | 1.6 | fast, low quality | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q4_K_S.gguf) | i1-Q4_K_S | 1.6 | optimal size/speed/quality | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q4_K_M.gguf) | i1-Q4_K_M | 1.6 | fast, recommended | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q4_1.gguf) | i1-Q4_1 | 1.7 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q5_K_S.gguf) | i1-Q5_K_S | 1.9 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q5_K_M.gguf) | i1-Q5_K_M | 1.9 | | +| [GGUF](https://huggingface.co/mradermacher/granite-3.1-2b-Reasoning-i1-GGUF/resolve/main/granite-3.1-2b-Reasoning.i1-Q6_K.gguf) | i1-Q6_K | 2.2 | practically like static Q6_K | + +Here is a handy graph by ikawrakow comparing some lower-quality quant +types (lower is better): + +![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png) + +And here are Artefact2's thoughts on the matter: +https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9 + +## FAQ / Model Request + +See https://huggingface.co/mradermacher/model_requests for some answers to +questions you might have and/or if you want some other model quantized. + +## Thanks + +I thank my company, [nethype GmbH](https://www.nethype.de/), for letting +me use its servers and providing upgrades to my workstation to enable +this work in my free time. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to. + + diff --git a/granite-3.1-2b-Reasoning.i1-IQ1_M.gguf b/granite-3.1-2b-Reasoning.i1-IQ1_M.gguf new file mode 100644 index 0000000..28f26ab --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ1_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7a2a6d14c342a1aca2c8b840f016776abf55cd59fe8317641b98188a94f5dcf +size 633919104 diff --git a/granite-3.1-2b-Reasoning.i1-IQ1_S.gguf b/granite-3.1-2b-Reasoning.i1-IQ1_S.gguf new file mode 100644 index 0000000..27ac48d --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ1_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b1f487e86414040a9f054d2701ab912a3a38d179a2a49d61c3e8a71d899e8cc +size 583784064 diff --git a/granite-3.1-2b-Reasoning.i1-IQ2_M.gguf b/granite-3.1-2b-Reasoning.i1-IQ2_M.gguf new file mode 100644 index 0000000..39845fe --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:04fe2178afa7bb75e66975f127f104c02c7ace2eba69541e3f8f53ff71f6e039 +size 888526464 diff --git a/granite-3.1-2b-Reasoning.i1-IQ2_S.gguf b/granite-3.1-2b-Reasoning.i1-IQ2_S.gguf new file mode 100644 index 0000000..736f922 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ2_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4bb52bebcdd835968352143dbf3d0c7895346c969e004890df8febd44496b371 +size 821679744 diff --git a/granite-3.1-2b-Reasoning.i1-IQ2_XS.gguf b/granite-3.1-2b-Reasoning.i1-IQ2_XS.gguf new file mode 100644 index 0000000..28b7f58 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ2_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:12d2683c3985571a19d6ed17ebd507a800bd3625b736f48fca86c8da057c3ec6 +size 789567104 diff --git a/granite-3.1-2b-Reasoning.i1-IQ2_XXS.gguf b/granite-3.1-2b-Reasoning.i1-IQ2_XXS.gguf new file mode 100644 index 0000000..292f96d --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ2_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:886565dc8b9cb49a3e61deb16711c0dfac520ed3d363b2d47bc622877643dd1c +size 717477504 diff --git a/granite-3.1-2b-Reasoning.i1-IQ3_M.gguf b/granite-3.1-2b-Reasoning.i1-IQ3_M.gguf new file mode 100644 index 0000000..1ad6441 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8fd16974d90be6170412721e94e82a3bb43302ed1ddbbcb12891d20627c6ac48 +size 1169283520 diff --git a/granite-3.1-2b-Reasoning.i1-IQ3_S.gguf b/granite-3.1-2b-Reasoning.i1-IQ3_S.gguf new file mode 100644 index 0000000..a4a2d23 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ3_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:738c1bdcd79ff7b1917150c5d396a4f294c091704373366283a4c1b11e6086f7 +size 1135860160 diff --git a/granite-3.1-2b-Reasoning.i1-IQ3_XS.gguf b/granite-3.1-2b-Reasoning.i1-IQ3_XS.gguf new file mode 100644 index 0000000..6211f4f --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f618f75e04210989d7d6dfa7169b0ebcf0268d52deb323e7732821865877effd +size 1078843840 diff --git a/granite-3.1-2b-Reasoning.i1-IQ3_XXS.gguf b/granite-3.1-2b-Reasoning.i1-IQ3_XXS.gguf new file mode 100644 index 0000000..9eb3a62 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d77f05ea196f859c821771dc6120bc2f043b4e95e7f46a1d644cac099f030625 +size 1005180544 diff --git a/granite-3.1-2b-Reasoning.i1-IQ4_NL.gguf b/granite-3.1-2b-Reasoning.i1-IQ4_NL.gguf new file mode 100644 index 0000000..17ac528 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c9767d61817514040653315d3b5a47c99241c14a1b179301c3d85231e54e50a9 +size 1458624960 diff --git a/granite-3.1-2b-Reasoning.i1-IQ4_XS.gguf b/granite-3.1-2b-Reasoning.i1-IQ4_XS.gguf new file mode 100644 index 0000000..baf5211 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:748ecfd3cbd010366706ecf4f17061d3def64bf08f56c9afe25707b5c60feaaf +size 1383913920 diff --git a/granite-3.1-2b-Reasoning.i1-Q2_K.gguf b/granite-3.1-2b-Reasoning.i1-Q2_K.gguf new file mode 100644 index 0000000..b304c9c --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5da04857dbf3a157a3785bcbf44f55a9ebef98a0e176ba30fa7b191eb44eeb86 +size 978246080 diff --git a/granite-3.1-2b-Reasoning.i1-Q2_K_S.gguf b/granite-3.1-2b-Reasoning.i1-Q2_K_S.gguf new file mode 100644 index 0000000..ab7b5a5 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q2_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fc13c7b1bdbb8209a22b3fcd1202a2e0b33797da1b79ede86268c8700559b00b +size 912710080 diff --git a/granite-3.1-2b-Reasoning.i1-Q3_K_L.gguf b/granite-3.1-2b-Reasoning.i1-Q3_K_L.gguf new file mode 100644 index 0000000..41871ca --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2cd5378c6a2402b09e74bfd725df869b273189741e16b04abe60e16cd45b815 +size 1357371840 diff --git a/granite-3.1-2b-Reasoning.i1-Q3_K_M.gguf b/granite-3.1-2b-Reasoning.i1-Q3_K_M.gguf new file mode 100644 index 0000000..a4ac0ab --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5559f7c1c3d3eba529b49fbc054dedb9b5f7aeab118d22137bb2b1d6566e3fa5 +size 1251727808 diff --git a/granite-3.1-2b-Reasoning.i1-Q3_K_S.gguf b/granite-3.1-2b-Reasoning.i1-Q3_K_S.gguf new file mode 100644 index 0000000..312a70e --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df21c333b4bb5fd2a549803b69039b15fd188154941682219c48ca0c03194afa +size 1130289600 diff --git a/granite-3.1-2b-Reasoning.i1-Q4_0.gguf b/granite-3.1-2b-Reasoning.i1-Q4_0.gguf new file mode 100644 index 0000000..62067a1 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e6f22b2c81d295f3d2431afb5e8f34252441814cd521dd29318d9f0de8ed7f73 +size 1458624960 diff --git a/granite-3.1-2b-Reasoning.i1-Q4_1.gguf b/granite-3.1-2b-Reasoning.i1-Q4_1.gguf new file mode 100644 index 0000000..15e8742 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fb6e4eea2f095378bc213b762707e428f24d1e44aaa279c8d92cf02230d08c3 +size 1605425600 diff --git a/granite-3.1-2b-Reasoning.i1-Q4_K_M.gguf b/granite-3.1-2b-Reasoning.i1-Q4_K_M.gguf new file mode 100644 index 0000000..acd7dbf --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f2cb0165f91895d9ba14d22b923d5d7d312144253e3202d7c5ac8e5c9678814 +size 1545296320 diff --git a/granite-3.1-2b-Reasoning.i1-Q4_K_S.gguf b/granite-3.1-2b-Reasoning.i1-Q4_K_S.gguf new file mode 100644 index 0000000..0f60fbb --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e86c36f7815fd1378e396ddf2dc195d62703ebf982faee071bd38621ee970d1c +size 1464392128 diff --git a/granite-3.1-2b-Reasoning.i1-Q5_K_M.gguf b/granite-3.1-2b-Reasoning.i1-Q5_K_M.gguf new file mode 100644 index 0000000..8f1abfd --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a7264923eade9c3f5e3f5d47c2f21d856c3c93754e9c1d4d18b69a6cf08c0a8 +size 1804818880 diff --git a/granite-3.1-2b-Reasoning.i1-Q5_K_S.gguf b/granite-3.1-2b-Reasoning.i1-Q5_K_S.gguf new file mode 100644 index 0000000..1951b10 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2617063e4f6ed26d277f4c3d93e15bb707ebb81dcb92251cb5c45914902ef62 +size 1757469120 diff --git a/granite-3.1-2b-Reasoning.i1-Q6_K.gguf b/granite-3.1-2b-Reasoning.i1-Q6_K.gguf new file mode 100644 index 0000000..6e957e7 --- /dev/null +++ b/granite-3.1-2b-Reasoning.i1-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90ed457c235ee0dc5fa9937c3bd907957294d9a1b2179ee3df2116a3c44d211f +size 2080561600 diff --git a/imatrix.dat b/imatrix.dat new file mode 100644 index 0000000..4de5aea --- /dev/null +++ b/imatrix.dat @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01e5b8ee1b8e0c888a4e996f7db1c900e2ae5ba5ec58ef310218b050bfdc2561 +size 3286085