From 6ade4ed31ce14d52d78ec356fc55ef410939fe6f Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Sun, 6 Sep 2026 07:54:12 +0800 Subject: [PATCH] =?UTF-8?q?=E5=88=9D=E5=A7=8B=E5=8C=96=E9=A1=B9=E7=9B=AE?= =?UTF-8?q?=EF=BC=8C=E7=94=B1ModelHub=20XC=E7=A4=BE=E5=8C=BA=E6=8F=90?= =?UTF-8?q?=E4=BE=9B=E6=A8=A1=E5=9E=8B?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Model: bartowski/prism-ml_Bonsai-8B-unpacked-GGUF Source: Original Platform --- .gitattributes | 73 +++++++++ README.md | 185 +++++++++++++++++++++++ configuration.json | 1 + prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf | 3 + prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf | 3 + prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf | 3 + prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf | 3 + prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf | 3 + prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q2_K.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q4_0.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q4_1.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q6_K.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf | 3 + prism-ml_Bonsai-8B-unpacked-Q8_0.gguf | 3 + prism-ml_Bonsai-8B-unpacked-bf16.gguf | 3 + prism-ml_Bonsai-8B-unpacked-imatrix.gguf | 3 + 28 files changed, 334 insertions(+) create mode 100644 .gitattributes create mode 100644 README.md create mode 100644 configuration.json create mode 100644 prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q2_K.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q4_0.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q4_1.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q6_K.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-Q8_0.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-bf16.gguf create mode 100644 prism-ml_Bonsai-8B-unpacked-imatrix.gguf diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..8a9a5e8 --- /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 + +prism-ml_Bonsai-8B-unpacked-bf16.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-imatrix.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +prism-ml_Bonsai-8B-unpacked-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text \ No newline at end of file diff --git a/README.md b/README.md new file mode 100644 index 0000000..6f206d7 --- /dev/null +++ b/README.md @@ -0,0 +1,185 @@ +--- +quantized_by: bartowski +pipeline_tag: text-generation +base_model_relation: quantized +tags: +- prismml +- bonsai +license: apache-2.0 +base_model: prism-ml/Bonsai-8B-unpacked +--- + +## Llamacpp imatrix Quantizations of Bonsai-8B-unpacked by prism-ml + +Using llama.cpp release b8586 for quantization. + +Original model: https://huggingface.co/prism-ml/Bonsai-8B-unpacked + +All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/82ae9b520227f57d79ba04add13d0d0d) + +Run them in your choice of tools: + +- [llama.cpp](https://github.com/ggml-org/llama.cpp) +- [ramalama](https://github.com/containers/ramalama) +- [LM Studio](https://lmstudio.ai/) +- [koboldcpp](https://github.com/LostRuins/koboldcpp) +- [Jan AI](https://www.jan.ai/) +- [Text Generation Web UI](https://github.com/oobabooga/text-generation-webui) +- [LoLLMs](https://github.com/ParisNeo/lollms) + +Note: if it's a newly supported model, you may need to wait for an update from the developers. + +## Prompt format + +``` +<|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 | +| -------- | ---------- | --------- | ----- | ----------- | +| [Bonsai-8B-unpacked-bf16.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-bf16.gguf) | bf16 | 16.38GB | false | Full BF16 weights. | +| [Bonsai-8B-unpacked-Q8_0.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q8_0.gguf) | Q8_0 | 8.71GB | false | Extremely high quality, generally unneeded but max available quant. | +| [Bonsai-8B-unpacked-Q6_K_L.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf) | Q6_K_L | 7.30GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. | +| [Bonsai-8B-unpacked-Q6_K.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q6_K.gguf) | Q6_K | 7.00GB | false | Very high quality, near perfect, *recommended*. | +| [Bonsai-8B-unpacked-Q5_K_L.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf) | Q5_K_L | 6.37GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. | +| [Bonsai-8B-unpacked-Q5_K_M.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf) | Q5_K_M | 5.99GB | false | High quality, *recommended*. | +| [Bonsai-8B-unpacked-Q5_K_S.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf) | Q5_K_S | 5.76GB | false | High quality, *recommended*. | +| [Bonsai-8B-unpacked-Q4_K_L.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf) | Q4_K_L | 5.66GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. | +| [Bonsai-8B-unpacked-Q4_1.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q4_1.gguf) | Q4_1 | 5.31GB | false | Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon. | +| [Bonsai-8B-unpacked-Q4_K_M.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf) | Q4_K_M | 5.20GB | false | Good quality, default size for most use cases, *recommended*. | +| [Bonsai-8B-unpacked-Q3_K_XL.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf) | Q3_K_XL | 5.02GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. | +| [Bonsai-8B-unpacked-Q4_K_S.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf) | Q4_K_S | 4.89GB | false | Slightly lower quality with more space savings, *recommended*. | +| [Bonsai-8B-unpacked-IQ4_NL.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf) | IQ4_NL | 4.89GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. | +| [Bonsai-8B-unpacked-Q4_0.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q4_0.gguf) | Q4_0 | 4.86GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. | +| [Bonsai-8B-unpacked-IQ4_XS.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf) | IQ4_XS | 4.67GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. | +| [Bonsai-8B-unpacked-Q3_K_L.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf) | Q3_K_L | 4.47GB | false | Lower quality but usable, good for low RAM availability. | +| [Bonsai-8B-unpacked-Q3_K_M.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf) | Q3_K_M | 4.25GB | false | Low quality. | +| [Bonsai-8B-unpacked-Q2_K_L.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf) | Q2_K_L | 4.04GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. | +| [Bonsai-8B-unpacked-IQ3_M.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf) | IQ3_M | 4.03GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. | +| [Bonsai-8B-unpacked-Q3_K_S.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf) | Q3_K_S | 3.89GB | false | Low quality, not recommended. | +| [Bonsai-8B-unpacked-IQ3_XS.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf) | IQ3_XS | 3.78GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. | +| [Bonsai-8B-unpacked-IQ3_XXS.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf) | IQ3_XXS | 3.55GB | false | Lower quality, new method with decent performance, comparable to Q3 quants. | +| [Bonsai-8B-unpacked-Q2_K.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-Q2_K.gguf) | Q2_K | 3.43GB | false | Very low quality but surprisingly usable. | +| [Bonsai-8B-unpacked-IQ2_M.gguf](https://huggingface.co/bartowski/prism-ml_Bonsai-8B-unpacked-GGUF/blob/main/prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf) | IQ2_M | 3.36GB | 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/prism-ml_Bonsai-8B-unpacked-GGUF --include "prism-ml_Bonsai-8B-unpacked-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/prism-ml_Bonsai-8B-unpacked-GGUF --include "prism-ml_Bonsai-8B-unpacked-Q8_0/*" --local-dir ./ +``` + +You can either specify a new local-dir (prism-ml_Bonsai-8B-unpacked-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 diff --git a/prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf b/prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf new file mode 100644 index 0000000..1652b44 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df56ab714f789be8bd1b7b19814777d7febe0b77886e95c0bed7552f808df072 +size 3363096768 diff --git a/prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf b/prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf new file mode 100644 index 0000000..6cc9b09 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c795143e99e7f4a46d2448593aefc581077ab10abac2cf032379a6f11a18f311 +size 4032086368 diff --git a/prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf b/prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf new file mode 100644 index 0000000..8bbda37 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98684e9926360a18927b4ccddef96fae894e05b3f6b47067a53c0485105cf141 +size 3778855264 diff --git a/prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf b/prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf new file mode 100644 index 0000000..126726a --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0317ac115d7149080629a78c60078c9c0e8b81a07d890ba56483568d6959a8d0 +size 3547711680 diff --git a/prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf b/prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf new file mode 100644 index 0000000..3b261f4 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90c8cb2b4425dbb77c44ff6de5a1d23b57413a38ec518e52d1c6fb6c360e9267 +size 4890016256 diff --git a/prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf b/prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf new file mode 100644 index 0000000..f5ce9d8 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e500d1b4a6cf72d9bcd95b36a23b77b612f30dfc9e88d93aac2fc9bd5c07b7c0 +size 4667703168 diff --git a/prism-ml_Bonsai-8B-unpacked-Q2_K.gguf b/prism-ml_Bonsai-8B-unpacked-Q2_K.gguf new file mode 100644 index 0000000..02237d3 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:848ce0768d59945dc26c04a854d12f4ae6a72ddb843b483311f7da9664364b0b +size 3429106496 diff --git a/prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf b/prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf new file mode 100644 index 0000000..5852550 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b0aee01efd1a9a5afdf72b7b333613a0c8bf121d8f36647b74dcfc1298345b7c +size 4035782496 diff --git a/prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf b/prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf new file mode 100644 index 0000000..245f8c8 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3cd86dabec2e08f9fa6cff80c1b99c0614c8d54f7f6cb0cb4d213f41dd791611 +size 4472488288 diff --git a/prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf b/prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf new file mode 100644 index 0000000..4015bb7 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:202d7b382f034bba9a7e4c2359c6b832c9d768daff3b80d0ef524f63ca799407 +size 4249141600 diff --git a/prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf b/prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf new file mode 100644 index 0000000..278159e --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ead2787358dccea3d01511c4e0f22ca5f766057c676bd49e148907d31a7276f6 +size 3886203232 diff --git a/prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf b/prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf new file mode 100644 index 0000000..eff1e25 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8ae2402d2ccba3954560e79a3eb673ca03cac2605048f381981ad857cc67fdb0 +size 5016069984 diff --git a/prism-ml_Bonsai-8B-unpacked-Q4_0.gguf b/prism-ml_Bonsai-8B-unpacked-Q4_0.gguf new file mode 100644 index 0000000..a0713f4 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73431a69ce38f28e016546a7e7fc48614479df399099ea849d7d475ef8a0e99e +size 4863670784 diff --git a/prism-ml_Bonsai-8B-unpacked-Q4_1.gguf b/prism-ml_Bonsai-8B-unpacked-Q4_1.gguf new file mode 100644 index 0000000..0a40d04 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0194b0210076a2a644c9871fb5c523efa7dae791f9ed247059ae666d1cf5c2af +size 5305151232 diff --git a/prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf b/prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf new file mode 100644 index 0000000..0d9c0b6 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q4_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf94979b90a4591fece6dfc6f0393136ff7ce93819fc9503b49a425bc7be5455 +size 5659567968 diff --git a/prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf b/prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf new file mode 100644 index 0000000..7d9f56d --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70e3ea7628d1b0efee84365ad9fc5df0c9d183fd556b3291ceaee52d5c88ac35 +size 5198494208 diff --git a/prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf b/prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf new file mode 100644 index 0000000..7d510f0 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:192ac5429d06bd5de9b9c3290b0af2ceaec30e45fc46111208fae1ef3f6ee888 +size 4885690880 diff --git a/prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf b/prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf new file mode 100644 index 0000000..7823249 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q5_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47e0cc8ed153bcc9c0269f6121917601767b4987ed3f9313110dd30e9ddefff9 +size 6373255008 diff --git a/prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf b/prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf new file mode 100644 index 0000000..db72bab --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9bf5585b9a36c451c380804e219a18ceff7a946c16155051d694733197dafbc5 +size 5989835776 diff --git a/prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf b/prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf new file mode 100644 index 0000000..e892d8f --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a423ca907fd8e1ba017353bbe05ab6ef22d7c542acf253fe0701b04e2b79022 +size 5759214592 diff --git a/prism-ml_Bonsai-8B-unpacked-Q6_K.gguf b/prism-ml_Bonsai-8B-unpacked-Q6_K.gguf new file mode 100644 index 0000000..8c44129 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cca0aabab239069e11da3e9e438a6f0e95972d6065fd6465bdca2613987324bd +size 6998367392 diff --git a/prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf b/prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf new file mode 100644 index 0000000..4402eb4 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q6_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2eebf86fc963d666a649e3c469de9596fdb731d5794958be2dc591177cbc7694 +size 7299278688 diff --git a/prism-ml_Bonsai-8B-unpacked-Q8_0.gguf b/prism-ml_Bonsai-8B-unpacked-Q8_0.gguf new file mode 100644 index 0000000..6f12030 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b6f9f91fbf1e59288648214eb9bef3a13030caa5dbed41968064fb769ad424e +size 8707188576 diff --git a/prism-ml_Bonsai-8B-unpacked-bf16.gguf b/prism-ml_Bonsai-8B-unpacked-bf16.gguf new file mode 100644 index 0000000..a5a4993 --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-bf16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f326c84402c15f41c75c297787832f62864652d0fb5cd10aae82cd17799b3d3e +size 16383663168 diff --git a/prism-ml_Bonsai-8B-unpacked-imatrix.gguf b/prism-ml_Bonsai-8B-unpacked-imatrix.gguf new file mode 100644 index 0000000..20f8a3c --- /dev/null +++ b/prism-ml_Bonsai-8B-unpacked-imatrix.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be3d83b42304f9532713ed452a7825be0239de7de25a2ae237f9b2cc2e7b585d +size 5347200