commit c627a18de3960fb8b18df5e32c340ac9f83c2762 Author: ModelHub XC Date: Thu Jul 23 09:32:06 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: bartowski/LiquidAI_LFM2-8B-A1B-GGUF Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..02ff3af --- /dev/null +++ b/.gitattributes @@ -0,0 +1,73 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bin.* filter=lfs diff=lfs merge=lfs -text +*.bz2 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 +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack 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 +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +saved_model/**/* 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 +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zstandard filter=lfs diff=lfs merge=lfs -text +*.tfevents* filter=lfs diff=lfs merge=lfs -text +*.db* filter=lfs diff=lfs merge=lfs -text +*.ark* filter=lfs diff=lfs merge=lfs -text +**/*ckpt*data* filter=lfs diff=lfs merge=lfs -text +**/*ckpt*.meta filter=lfs diff=lfs merge=lfs -text +**/*ckpt*.index filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text + +*.ggml filter=lfs diff=lfs merge=lfs -text +*.llamafile* filter=lfs diff=lfs merge=lfs -text +*.pt2 filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text + +LiquidAI_LFM2-8B-A1B-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-imatrix.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q4_K_L.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-bf16.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q6_K_L.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q5_K_L.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text +LiquidAI_LFM2-8B-A1B-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text \ No newline at end of file diff --git a/LiquidAI_LFM2-8B-A1B-IQ2_M.gguf b/LiquidAI_LFM2-8B-A1B-IQ2_M.gguf new file mode 100644 index 0000000..263fd17 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b95448d7f31a12736fbf91b081f839ff4074d8bd21bbdfc494b54bd64841dede +size 2650602432 diff --git a/LiquidAI_LFM2-8B-A1B-IQ3_M.gguf b/LiquidAI_LFM2-8B-A1B-IQ3_M.gguf new file mode 100644 index 0000000..10b1212 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7fb4705e813c7e0b6a3814c94f8bd13c30f751ba9d806fd368022168270884e2 +size 3815472064 diff --git a/LiquidAI_LFM2-8B-A1B-IQ3_XS.gguf b/LiquidAI_LFM2-8B-A1B-IQ3_XS.gguf new file mode 100644 index 0000000..b120c30 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c20f3f96e01d40547bb8e81b73596aafa770b2a1d654fdd2e0ba675c8e3f08cf +size 3455400896 diff --git a/LiquidAI_LFM2-8B-A1B-IQ3_XXS.gguf b/LiquidAI_LFM2-8B-A1B-IQ3_XXS.gguf new file mode 100644 index 0000000..24645e4 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9916185bffb1d4b7d109e7c4cb65cf3de312b8d2b0d5ab49929e4c98de47cfad +size 3307322304 diff --git a/LiquidAI_LFM2-8B-A1B-IQ4_NL.gguf b/LiquidAI_LFM2-8B-A1B-IQ4_NL.gguf new file mode 100644 index 0000000..325e5a5 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:454bdeaa30f1520e00b6113d85a3051950df5d3aa5eb7ace4c3b09dc0089eaba +size 4740185024 diff --git a/LiquidAI_LFM2-8B-A1B-IQ4_XS.gguf b/LiquidAI_LFM2-8B-A1B-IQ4_XS.gguf new file mode 100644 index 0000000..1316b84 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea3b93fb618d0def7bb702c100c05924ab5b7cd631c3cfdccefd5dfa2c73ba06 +size 4484791232 diff --git a/LiquidAI_LFM2-8B-A1B-Q2_K.gguf b/LiquidAI_LFM2-8B-A1B-Q2_K.gguf new file mode 100644 index 0000000..b69d88b --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c5ca2740a85b8e4e052d416818550204ea90e5b07c358a189994a64347e549c +size 2947627968 diff --git a/LiquidAI_LFM2-8B-A1B-Q2_K_L.gguf b/LiquidAI_LFM2-8B-A1B-Q2_K_L.gguf new file mode 100644 index 0000000..5b887c0 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35b58aa042b2d3f512de73a56de101065c2d161be0b29cd63d17cf1a0e785e43 +size 2980133824 diff --git a/LiquidAI_LFM2-8B-A1B-Q3_K_L.gguf b/LiquidAI_LFM2-8B-A1B-Q3_K_L.gguf new file mode 100644 index 0000000..ea71b59 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20646b5c085b4722135566633f941484a3e6535ca5344a656020cd6dddde65f6 +size 3960241088 diff --git a/LiquidAI_LFM2-8B-A1B-Q3_K_M.gguf b/LiquidAI_LFM2-8B-A1B-Q3_K_M.gguf new file mode 100644 index 0000000..7d55633 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4069ab6aa215d15ff151bc02471b95664575967383d0d097ac2e34cb1685144 +size 3817569216 diff --git a/LiquidAI_LFM2-8B-A1B-Q3_K_S.gguf b/LiquidAI_LFM2-8B-A1B-Q3_K_S.gguf new file mode 100644 index 0000000..6e2e9d8 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84dd851ceef226f6d523416c1634e64d22ccdf4f4f52d40e1b0f7b4f072c83aa +size 3654761408 diff --git a/LiquidAI_LFM2-8B-A1B-Q3_K_XL.gguf b/LiquidAI_LFM2-8B-A1B-Q3_K_XL.gguf new file mode 100644 index 0000000..9d14396 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9af4f52a7a29d0cb9a9fae3dc86e76e85938a8ec9bd1c19d7a1ffabf9c9d1301 +size 3992746944 diff --git a/LiquidAI_LFM2-8B-A1B-Q4_0.gguf b/LiquidAI_LFM2-8B-A1B-Q4_0.gguf new file mode 100644 index 0000000..a5b4f93 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:382d8b20813803f82209810cbdf562f4391bf70e3c0324d86b9088f4954352cb +size 4812274624 diff --git a/LiquidAI_LFM2-8B-A1B-Q4_1.gguf b/LiquidAI_LFM2-8B-A1B-Q4_1.gguf new file mode 100644 index 0000000..3270ae2 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca0000439d19ca188a34fda75b91bf087c0d90a18db5a79eeb43283f816e8d81 +size 5249399744 diff --git a/LiquidAI_LFM2-8B-A1B-Q4_K_L.gguf b/LiquidAI_LFM2-8B-A1B-Q4_K_L.gguf new file mode 100644 index 0000000..e48da25 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q4_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be475ac67025430fff3ba5d46a123f8c47c1b20628924fcc95c6460444dbb8dd +size 5083577280 diff --git a/LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf b/LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf new file mode 100644 index 0000000..5bb0038 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efb59182eca2424126e9f8bde8513a1736e92d3b9a3187a2afc67968bd44512a +size 5051071424 diff --git a/LiquidAI_LFM2-8B-A1B-Q4_K_S.gguf b/LiquidAI_LFM2-8B-A1B-Q4_K_S.gguf new file mode 100644 index 0000000..7b76522 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9205a193e65c8f5f8f18ba1f4206adec5e98d097768f24230b2719a2d24ea0e +size 4891179968 diff --git a/LiquidAI_LFM2-8B-A1B-Q5_K_L.gguf b/LiquidAI_LFM2-8B-A1B-Q5_K_L.gguf new file mode 100644 index 0000000..cb31d89 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q5_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be7507dcb97630c964ec1cf92330c7245b5cd80bc669069352e1362a7d3a34c7 +size 5954419648 diff --git a/LiquidAI_LFM2-8B-A1B-Q5_K_M.gguf b/LiquidAI_LFM2-8B-A1B-Q5_K_M.gguf new file mode 100644 index 0000000..35d54a8 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:44b1ef132884fc890d21881d292e0d95c0f9d7920c391c57a2e292e1537819a2 +size 5921913792 diff --git a/LiquidAI_LFM2-8B-A1B-Q5_K_S.gguf b/LiquidAI_LFM2-8B-A1B-Q5_K_S.gguf new file mode 100644 index 0000000..ac6244d --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca87c8106ab9dca8c9e6af0d82a414f5f8aee1da4eb429402ad7270511b92914 +size 5761760192 diff --git a/LiquidAI_LFM2-8B-A1B-Q6_K.gguf b/LiquidAI_LFM2-8B-A1B-Q6_K.gguf new file mode 100644 index 0000000..3064032 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:169ea61e8bfcdb3539e5a946ca3787e0d94d26f76d39a370b41ad4d0c87a73ee +size 6850526144 diff --git a/LiquidAI_LFM2-8B-A1B-Q6_K_L.gguf b/LiquidAI_LFM2-8B-A1B-Q6_K_L.gguf new file mode 100644 index 0000000..b7513af --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q6_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4dc498edf024f3866fca51236fef4371a18fa1d54383d6e29f51891523e0d55 +size 6883032000 diff --git a/LiquidAI_LFM2-8B-A1B-Q8_0.gguf b/LiquidAI_LFM2-8B-A1B-Q8_0.gguf new file mode 100644 index 0000000..c890c4b --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac7cf13235b2f2b00387cc4e91d1085137fe23b5552991f352fbfefac6c6ddf5 +size 8868428736 diff --git a/LiquidAI_LFM2-8B-A1B-bf16.gguf b/LiquidAI_LFM2-8B-A1B-bf16.gguf new file mode 100644 index 0000000..cb30804 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-bf16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b71d63488b8aceac56c91bf641ab48160459c576ec506a472289eba66ad8074c +size 16685562560 diff --git a/LiquidAI_LFM2-8B-A1B-imatrix.gguf b/LiquidAI_LFM2-8B-A1B-imatrix.gguf new file mode 100644 index 0000000..663a881 --- /dev/null +++ b/LiquidAI_LFM2-8B-A1B-imatrix.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb7ccf3855d1cbcbda25ded6228465ab79689804dfb67f0f5431acffb3d105c7 +size 17375232 diff --git a/README.md b/README.md new file mode 100644 index 0000000..9718154 --- /dev/null +++ b/README.md @@ -0,0 +1,172 @@ +--- +quantized_by: bartowski +pipeline_tag: text-generation +base_model: LiquidAI/LFM2-8B-A1B +base_model_relation: quantized +--- + +## Llamacpp imatrix Quantizations of LFM2-8B-A1B by LiquidAI + +Using llama.cpp release b6714 for quantization. + +Original model: https://huggingface.co/LiquidAI/LFM2-8B-A1B + +All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/eb213dccb3571f863da82e99418f81e8) combined with a subset of combined_all_small.parquet from Ed Addario [here](https://huggingface.co/datasets/eaddario/imatrix-calibration/blob/main/combined_all_small.parquet) + +Run them in [LM Studio](https://lmstudio.ai/) + +Run them directly with [llama.cpp](https://github.com/ggml-org/llama.cpp), or any other llama.cpp based project + +## Prompt format + +No chat template specified so default is used. This may be incorrect, check original model card for details. + +``` +<|startoftext|><|im_start|>system +{system_prompt}<|im_end|> +<|im_start|>user +{prompt}<|im_end|> +<|im_start|>assistant +``` + +## Download a file (not the whole branch) from below: + +| Filename | Quant type | File Size | Split | Description | +| -------- | ---------- | --------- | ----- | ----------- | +| [LFM2-8B-A1B-bf16.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-bf16.gguf) | bf16 | 16.69GB | false | Full BF16 weights. | +| [LFM2-8B-A1B-Q8_0.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q8_0.gguf) | Q8_0 | 8.87GB | false | Extremely high quality, generally unneeded but max available quant. | +| [LFM2-8B-A1B-Q6_K_L.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q6_K_L.gguf) | Q6_K_L | 6.88GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. | +| [LFM2-8B-A1B-Q6_K.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q6_K.gguf) | Q6_K | 6.85GB | false | Very high quality, near perfect, *recommended*. | +| [LFM2-8B-A1B-Q5_K_L.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q5_K_L.gguf) | Q5_K_L | 5.95GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. | +| [LFM2-8B-A1B-Q5_K_M.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q5_K_M.gguf) | Q5_K_M | 5.92GB | false | High quality, *recommended*. | +| [LFM2-8B-A1B-Q5_K_S.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q5_K_S.gguf) | Q5_K_S | 5.76GB | false | High quality, *recommended*. | +| [LFM2-8B-A1B-Q4_1.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q4_1.gguf) | Q4_1 | 5.25GB | false | Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon. | +| [LFM2-8B-A1B-Q4_K_L.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q4_K_L.gguf) | Q4_K_L | 5.08GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. | +| [LFM2-8B-A1B-Q4_K_M.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf) | Q4_K_M | 5.05GB | false | Good quality, default size for most use cases, *recommended*. | +| [LFM2-8B-A1B-Q4_K_S.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q4_K_S.gguf) | Q4_K_S | 4.89GB | false | Slightly lower quality with more space savings, *recommended*. | +| [LFM2-8B-A1B-Q4_0.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q4_0.gguf) | Q4_0 | 4.81GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. | +| [LFM2-8B-A1B-IQ4_NL.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-IQ4_NL.gguf) | IQ4_NL | 4.74GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. | +| [LFM2-8B-A1B-IQ4_XS.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-IQ4_XS.gguf) | IQ4_XS | 4.48GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. | +| [LFM2-8B-A1B-Q3_K_XL.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q3_K_XL.gguf) | Q3_K_XL | 3.99GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. | +| [LFM2-8B-A1B-Q3_K_L.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q3_K_L.gguf) | Q3_K_L | 3.96GB | false | Lower quality but usable, good for low RAM availability. | +| [LFM2-8B-A1B-Q3_K_M.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q3_K_M.gguf) | Q3_K_M | 3.82GB | false | Low quality. | +| [LFM2-8B-A1B-IQ3_M.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-IQ3_M.gguf) | IQ3_M | 3.82GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. | +| [LFM2-8B-A1B-Q3_K_S.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q3_K_S.gguf) | Q3_K_S | 3.65GB | false | Low quality, not recommended. | +| [LFM2-8B-A1B-IQ3_XS.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-IQ3_XS.gguf) | IQ3_XS | 3.46GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. | +| [LFM2-8B-A1B-IQ3_XXS.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-IQ3_XXS.gguf) | IQ3_XXS | 3.31GB | false | Lower quality, new method with decent performance, comparable to Q3 quants. | +| [LFM2-8B-A1B-Q2_K_L.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q2_K_L.gguf) | Q2_K_L | 2.98GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. | +| [LFM2-8B-A1B-Q2_K.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-Q2_K.gguf) | Q2_K | 2.95GB | false | Very low quality but surprisingly usable. | +| [LFM2-8B-A1B-IQ2_M.gguf](https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF/blob/main/LiquidAI_LFM2-8B-A1B-IQ2_M.gguf) | IQ2_M | 2.65GB | false | Relatively low quality, uses SOTA techniques to be surprisingly usable. | + +## Embed/output weights + +Some of these quants (Q3_K_XL, Q4_K_L etc) are the standard quantization method with the embeddings and output weights quantized to Q8_0 instead of what they would normally default to. + +## Downloading using huggingface-cli + +
+ Click to view download instructions + +First, make sure you have hugginface-cli installed: + +``` +pip install -U "huggingface_hub[cli]" +``` + +Then, you can target the specific file you want: + +``` +huggingface-cli download bartowski/LiquidAI_LFM2-8B-A1B-GGUF --include "LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf" --local-dir ./ +``` + +If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run: + +``` +huggingface-cli download bartowski/LiquidAI_LFM2-8B-A1B-GGUF --include "LiquidAI_LFM2-8B-A1B-Q8_0/*" --local-dir ./ +``` + +You can either specify a new local-dir (LiquidAI_LFM2-8B-A1B-Q8_0) or download them all in place (./) + +
+ +## ARM/AVX information + +Previously, you would download Q4_0_4_4/4_8/8_8, and these would have their weights interleaved in memory in order to improve performance on ARM and AVX machines by loading up more data in one pass. + +Now, however, there is something called "online repacking" for weights. details in [this PR](https://github.com/ggml-org/llama.cpp/pull/9921). If you use Q4_0 and your hardware would benefit from repacking weights, it will do it automatically on the fly. + +As of llama.cpp build [b4282](https://github.com/ggml-org/llama.cpp/releases/tag/b4282) you will not be able to run the Q4_0_X_X files and will instead need to use Q4_0. + +Additionally, if you want to get slightly better quality for , you can use IQ4_NL thanks to [this PR](https://github.com/ggml-org/llama.cpp/pull/10541) which will also repack the weights for ARM, though only the 4_4 for now. The loading time may be slower but it will result in an overall speed incrase. + +
+ Click to view Q4_0_X_X information (deprecated + +I'm keeping this section to show the potential theoretical uplift in performance from using the Q4_0 with online repacking. + +
+ Click to view benchmarks on an AVX2 system (EPYC7702) + +| model | size | params | backend | threads | test | t/s | % (vs Q4_0) | +| ------------------------------ | ---------: | ---------: | ---------- | ------: | ------------: | -------------------: |-------------: | +| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | pp512 | 204.03 ± 1.03 | 100% | +| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | pp1024 | 282.92 ± 0.19 | 100% | +| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | pp2048 | 259.49 ± 0.44 | 100% | +| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | tg128 | 39.12 ± 0.27 | 100% | +| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | tg256 | 39.31 ± 0.69 | 100% | +| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | tg512 | 40.52 ± 0.03 | 100% | +| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | pp512 | 301.02 ± 1.74 | 147% | +| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | pp1024 | 287.23 ± 0.20 | 101% | +| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | pp2048 | 262.77 ± 1.81 | 101% | +| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | tg128 | 18.80 ± 0.99 | 48% | +| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | tg256 | 24.46 ± 3.04 | 83% | +| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | tg512 | 36.32 ± 3.59 | 90% | +| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | pp512 | 271.71 ± 3.53 | 133% | +| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | pp1024 | 279.86 ± 45.63 | 100% | +| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | pp2048 | 320.77 ± 5.00 | 124% | +| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | tg128 | 43.51 ± 0.05 | 111% | +| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | tg256 | 43.35 ± 0.09 | 110% | +| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | tg512 | 42.60 ± 0.31 | 105% | + +Q4_0_8_8 offers a nice bump to prompt processing and a small bump to text generation + +
+ +
+ +## Which file should I choose? + +
+ Click here for details + +A great write up with charts showing various performances is provided by Artefact2 [here](https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9) + +The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have. + +If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM. + +If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total. + +Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'. + +If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M. + +If you want to get more into the weeds, you can check out this extremely useful feature chart: + +[llama.cpp feature matrix](https://github.com/ggml-org/llama.cpp/wiki/Feature-matrix) + +But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size. + +These I-quants can also be used on CPU, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide. + +
+ +## Credits + +Thank you kalomaze and Dampf for assistance in creating the imatrix calibration dataset. + +Thank you ZeroWw for the inspiration to experiment with embed/output. + +Thank you to LM Studio for sponsoring my work. + +Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski diff --git a/configuration.json b/configuration.json new file mode 100644 index 0000000..bbeeda1 --- /dev/null +++ b/configuration.json @@ -0,0 +1 @@ +{"framework": "pytorch", "task": "text-generation", "allow_remote": true} \ No newline at end of file