From 859d9c70c98c0ddca2a7be272609db9f388aaef6 Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Sat, 5 Sep 2026 12:34:13 +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/janhq_Jan-code-4b-GGUF Source: Original Platform --- .gitattributes | 73 +++++++++++++ README.md | 182 +++++++++++++++++++++++++++++++++ configuration.json | 1 + janhq_Jan-code-4b-IQ2_M.gguf | 3 + janhq_Jan-code-4b-IQ3_M.gguf | 3 + janhq_Jan-code-4b-IQ3_XS.gguf | 3 + janhq_Jan-code-4b-IQ3_XXS.gguf | 3 + janhq_Jan-code-4b-IQ4_NL.gguf | 3 + janhq_Jan-code-4b-IQ4_XS.gguf | 3 + janhq_Jan-code-4b-Q2_K.gguf | 3 + janhq_Jan-code-4b-Q2_K_L.gguf | 3 + janhq_Jan-code-4b-Q3_K_L.gguf | 3 + janhq_Jan-code-4b-Q3_K_M.gguf | 3 + janhq_Jan-code-4b-Q3_K_S.gguf | 3 + janhq_Jan-code-4b-Q3_K_XL.gguf | 3 + janhq_Jan-code-4b-Q4_0.gguf | 3 + janhq_Jan-code-4b-Q4_1.gguf | 3 + janhq_Jan-code-4b-Q4_K_L.gguf | 3 + janhq_Jan-code-4b-Q4_K_M.gguf | 3 + janhq_Jan-code-4b-Q4_K_S.gguf | 3 + janhq_Jan-code-4b-Q5_K_L.gguf | 3 + janhq_Jan-code-4b-Q5_K_M.gguf | 3 + janhq_Jan-code-4b-Q5_K_S.gguf | 3 + janhq_Jan-code-4b-Q6_K.gguf | 3 + janhq_Jan-code-4b-Q6_K_L.gguf | 3 + janhq_Jan-code-4b-Q8_0.gguf | 3 + janhq_Jan-code-4b-bf16.gguf | 3 + janhq_Jan-code-4b-imatrix.gguf | 3 + 28 files changed, 331 insertions(+) create mode 100644 .gitattributes create mode 100644 README.md create mode 100644 configuration.json create mode 100644 janhq_Jan-code-4b-IQ2_M.gguf create mode 100644 janhq_Jan-code-4b-IQ3_M.gguf create mode 100644 janhq_Jan-code-4b-IQ3_XS.gguf create mode 100644 janhq_Jan-code-4b-IQ3_XXS.gguf create mode 100644 janhq_Jan-code-4b-IQ4_NL.gguf create mode 100644 janhq_Jan-code-4b-IQ4_XS.gguf create mode 100644 janhq_Jan-code-4b-Q2_K.gguf create mode 100644 janhq_Jan-code-4b-Q2_K_L.gguf create mode 100644 janhq_Jan-code-4b-Q3_K_L.gguf create mode 100644 janhq_Jan-code-4b-Q3_K_M.gguf create mode 100644 janhq_Jan-code-4b-Q3_K_S.gguf create mode 100644 janhq_Jan-code-4b-Q3_K_XL.gguf create mode 100644 janhq_Jan-code-4b-Q4_0.gguf create mode 100644 janhq_Jan-code-4b-Q4_1.gguf create mode 100644 janhq_Jan-code-4b-Q4_K_L.gguf create mode 100644 janhq_Jan-code-4b-Q4_K_M.gguf create mode 100644 janhq_Jan-code-4b-Q4_K_S.gguf create mode 100644 janhq_Jan-code-4b-Q5_K_L.gguf create mode 100644 janhq_Jan-code-4b-Q5_K_M.gguf create mode 100644 janhq_Jan-code-4b-Q5_K_S.gguf create mode 100644 janhq_Jan-code-4b-Q6_K.gguf create mode 100644 janhq_Jan-code-4b-Q6_K_L.gguf create mode 100644 janhq_Jan-code-4b-Q8_0.gguf create mode 100644 janhq_Jan-code-4b-bf16.gguf create mode 100644 janhq_Jan-code-4b-imatrix.gguf diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..be784de --- /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 + +janhq_Jan-code-4b-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-bf16.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q5_K_L.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q6_K_L.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q4_K_L.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +janhq_Jan-code-4b-imatrix.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..3893b78 --- /dev/null +++ b/README.md @@ -0,0 +1,182 @@ +--- +quantized_by: bartowski +pipeline_tag: text-generation +base_model_relation: quantized +tags: +- agent +license: apache-2.0 +base_model: janhq/Jan-code-4b +language: +- en +--- + +## Llamacpp imatrix Quantizations of Jan-code-4b by janhq + +Using llama.cpp release b8184 for quantization. + +Original model: https://huggingface.co/janhq/Jan-code-4b + +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) +- [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 | +| -------- | ---------- | --------- | ----- | ----------- | +| [Jan-code-4b-bf16.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-bf16.gguf) | bf16 | 8.83GB | false | Full BF16 weights. | +| [Jan-code-4b-Q8_0.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q8_0.gguf) | Q8_0 | 4.69GB | false | Extremely high quality, generally unneeded but max available quant. | +| [Jan-code-4b-Q6_K_L.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q6_K_L.gguf) | Q6_K_L | 3.92GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. | +| [Jan-code-4b-Q6_K.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q6_K.gguf) | Q6_K | 3.73GB | false | Very high quality, near perfect, *recommended*. | +| [Jan-code-4b-Q5_K_L.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q5_K_L.gguf) | Q5_K_L | 3.54GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. | +| [Jan-code-4b-Q5_K_M.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q5_K_M.gguf) | Q5_K_M | 3.30GB | false | High quality, *recommended*. | +| [Jan-code-4b-Q5_K_S.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q5_K_S.gguf) | Q5_K_S | 3.14GB | false | High quality, *recommended*. | +| [Jan-code-4b-Q4_K_L.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q4_K_L.gguf) | Q4_K_L | 3.10GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. | +| [Jan-code-4b-Q4_1.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q4_1.gguf) | Q4_1 | 2.84GB | false | Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon. | +| [Jan-code-4b-Q4_K_M.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q4_K_M.gguf) | Q4_K_M | 2.81GB | false | Good quality, default size for most use cases, *recommended*. | +| [Jan-code-4b-Q3_K_XL.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q3_K_XL.gguf) | Q3_K_XL | 2.75GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. | +| [Jan-code-4b-Q4_K_S.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q4_K_S.gguf) | Q4_K_S | 2.65GB | false | Slightly lower quality with more space savings, *recommended*. | +| [Jan-code-4b-IQ4_NL.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-IQ4_NL.gguf) | IQ4_NL | 2.60GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. | +| [Jan-code-4b-Q4_0.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q4_0.gguf) | Q4_0 | 2.59GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. | +| [Jan-code-4b-IQ4_XS.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-IQ4_XS.gguf) | IQ4_XS | 2.48GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. | +| [Jan-code-4b-Q3_K_L.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q3_K_L.gguf) | Q3_K_L | 2.41GB | false | Lower quality but usable, good for low RAM availability. | +| [Jan-code-4b-Q3_K_M.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q3_K_M.gguf) | Q3_K_M | 2.24GB | false | Low quality. | +| [Jan-code-4b-Q2_K_L.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q2_K_L.gguf) | Q2_K_L | 2.18GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. | +| [Jan-code-4b-IQ3_M.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-IQ3_M.gguf) | IQ3_M | 2.13GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. | +| [Jan-code-4b-Q3_K_S.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q3_K_S.gguf) | Q3_K_S | 2.05GB | false | Low quality, not recommended. | +| [Jan-code-4b-IQ3_XS.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-IQ3_XS.gguf) | IQ3_XS | 1.98GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. | +| [Jan-code-4b-IQ3_XXS.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-IQ3_XXS.gguf) | IQ3_XXS | 1.82GB | false | Lower quality, new method with decent performance, comparable to Q3 quants. | +| [Jan-code-4b-Q2_K.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-Q2_K.gguf) | Q2_K | 1.80GB | false | Very low quality but surprisingly usable. | +| [Jan-code-4b-IQ2_M.gguf](https://huggingface.co/bartowski/janhq_Jan-code-4b-GGUF/blob/main/janhq_Jan-code-4b-IQ2_M.gguf) | IQ2_M | 1.68GB | 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/janhq_Jan-code-4b-GGUF --include "janhq_Jan-code-4b-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/janhq_Jan-code-4b-GGUF --include "janhq_Jan-code-4b-Q8_0/*" --local-dir ./ +``` + +You can either specify a new local-dir (janhq_Jan-code-4b-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/janhq_Jan-code-4b-IQ2_M.gguf b/janhq_Jan-code-4b-IQ2_M.gguf new file mode 100644 index 0000000..65e8b6f --- /dev/null +++ b/janhq_Jan-code-4b-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3d2de75469807b8d3e4dfd10b0d4764cf4a2cbfb0cac157de0cfeff8a8309e09 +size 1680112192 diff --git a/janhq_Jan-code-4b-IQ3_M.gguf b/janhq_Jan-code-4b-IQ3_M.gguf new file mode 100644 index 0000000..92bac55 --- /dev/null +++ b/janhq_Jan-code-4b-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:354e98263ee23ba4e984534055d054311e1838635ba3b3914f1a39672da0cdef +size 2130024512 diff --git a/janhq_Jan-code-4b-IQ3_XS.gguf b/janhq_Jan-code-4b-IQ3_XS.gguf new file mode 100644 index 0000000..725da92 --- /dev/null +++ b/janhq_Jan-code-4b-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3831759489521b5626f5b2a73b7b3869a311bbeedf8339460c4256754d317943 +size 1981503552 diff --git a/janhq_Jan-code-4b-IQ3_XXS.gguf b/janhq_Jan-code-4b-IQ3_XXS.gguf new file mode 100644 index 0000000..ccdb89b --- /dev/null +++ b/janhq_Jan-code-4b-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:810a711ed2c7b93f94e343fd3d6bf1442503647c075e18bbccd5b61815e7b430 +size 1824782912 diff --git a/janhq_Jan-code-4b-IQ4_NL.gguf b/janhq_Jan-code-4b-IQ4_NL.gguf new file mode 100644 index 0000000..91f0d3c --- /dev/null +++ b/janhq_Jan-code-4b-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4526d9062351e09d9447d2be3ca0881aaa3be54c963a132c86001ff9334158f +size 2600130112 diff --git a/janhq_Jan-code-4b-IQ4_XS.gguf b/janhq_Jan-code-4b-IQ4_XS.gguf new file mode 100644 index 0000000..905cfdc --- /dev/null +++ b/janhq_Jan-code-4b-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ec432948835587bf5a0adf3d5ed231fe53948bda22fee83d5ee52b6f2764db4 +size 2477383232 diff --git a/janhq_Jan-code-4b-Q2_K.gguf b/janhq_Jan-code-4b-Q2_K.gguf new file mode 100644 index 0000000..31fe855 --- /dev/null +++ b/janhq_Jan-code-4b-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0900776478d879dd6b5dc80571151f0ad6c5c1b31fcf261d2db3e7db068bc0d2 +size 1797124672 diff --git a/janhq_Jan-code-4b-Q2_K_L.gguf b/janhq_Jan-code-4b-Q2_K_L.gguf new file mode 100644 index 0000000..e0fb10a --- /dev/null +++ b/janhq_Jan-code-4b-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d04bb3aebb9b04a264d79b8b429d5c37c4ca3cb8528a4b177b635a689f07c89 +size 2176964672 diff --git a/janhq_Jan-code-4b-Q3_K_L.gguf b/janhq_Jan-code-4b-Q3_K_L.gguf new file mode 100644 index 0000000..3f0297c --- /dev/null +++ b/janhq_Jan-code-4b-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:28f64eec6da44991260f432d2f26cdd57881f850be7afecb7363d3e62a2e3074 +size 2406914112 diff --git a/janhq_Jan-code-4b-Q3_K_M.gguf b/janhq_Jan-code-4b-Q3_K_M.gguf new file mode 100644 index 0000000..e5a8de9 --- /dev/null +++ b/janhq_Jan-code-4b-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:600e14194e7f104ff815333bb758dc0f9f8635dcaa44ab60148da595f28e24a0 +size 2242746432 diff --git a/janhq_Jan-code-4b-Q3_K_S.gguf b/janhq_Jan-code-4b-Q3_K_S.gguf new file mode 100644 index 0000000..7c77638 --- /dev/null +++ b/janhq_Jan-code-4b-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51c560882a18d37cb4610b3d1f09b030330c7dba36ff411c1ee2ce9dd387fb32 +size 2054125632 diff --git a/janhq_Jan-code-4b-Q3_K_XL.gguf b/janhq_Jan-code-4b-Q3_K_XL.gguf new file mode 100644 index 0000000..2d33223 --- /dev/null +++ b/janhq_Jan-code-4b-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c93f7b4cbf2731c77e599e194f08723ea4707ca8fb96eb96d537ed2c6f6b9e5f +size 2747250752 diff --git a/janhq_Jan-code-4b-Q4_0.gguf b/janhq_Jan-code-4b-Q4_0.gguf new file mode 100644 index 0000000..91b690b --- /dev/null +++ b/janhq_Jan-code-4b-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:413320e5f18ec7cfc83ff602be36c75e3aaf52fab7198dda0f39770e0e41e138 +size 2594559552 diff --git a/janhq_Jan-code-4b-Q4_1.gguf b/janhq_Jan-code-4b-Q4_1.gguf new file mode 100644 index 0000000..b8edf67 --- /dev/null +++ b/janhq_Jan-code-4b-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3712556917fc0ca31aa99e0244ef45aab99a1ef20f4a7e12cc5c92c08e546d92 +size 2839725632 diff --git a/janhq_Jan-code-4b-Q4_K_L.gguf b/janhq_Jan-code-4b-Q4_K_L.gguf new file mode 100644 index 0000000..2df2006 --- /dev/null +++ b/janhq_Jan-code-4b-Q4_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0025fac9b7cc82f3a4c2211d60ebbe2fd34ef796028bbda99c16679be627b02e +size 3102066752 diff --git a/janhq_Jan-code-4b-Q4_K_M.gguf b/janhq_Jan-code-4b-Q4_K_M.gguf new file mode 100644 index 0000000..618aeb8 --- /dev/null +++ b/janhq_Jan-code-4b-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9938eb10674535ee43207c8ea8cb52b0605d9401dfb56387b8c1020b6ae2e56e +size 2813388352 diff --git a/janhq_Jan-code-4b-Q4_K_S.gguf b/janhq_Jan-code-4b-Q4_K_S.gguf new file mode 100644 index 0000000..cfd8a1d --- /dev/null +++ b/janhq_Jan-code-4b-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:82fdb40c529b8ab1432cdc5e3ee6ee6ba0bea34fc8fc4783a5b5d5bb21dfc8e9 +size 2649282112 diff --git a/janhq_Jan-code-4b-Q5_K_L.gguf b/janhq_Jan-code-4b-Q5_K_L.gguf new file mode 100644 index 0000000..123e4f9 --- /dev/null +++ b/janhq_Jan-code-4b-Q5_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be6465e397692c6d963d60bbda7ae1bc838272a3a79d8a8e8edfc394b391dffb +size 3538536512 diff --git a/janhq_Jan-code-4b-Q5_K_M.gguf b/janhq_Jan-code-4b-Q5_K_M.gguf new file mode 100644 index 0000000..b0cea92 --- /dev/null +++ b/janhq_Jan-code-4b-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5537c58d9e0a4e5b62a9bfc2c0ceb1fb046b3406633ac8b24c4dc5a455d64e57 +size 3298477632 diff --git a/janhq_Jan-code-4b-Q5_K_S.gguf b/janhq_Jan-code-4b-Q5_K_S.gguf new file mode 100644 index 0000000..8bbec5f --- /dev/null +++ b/janhq_Jan-code-4b-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdd9fe743ee40e741230352605506977cfc9fab9703f90f7d9c7723250f8f332 +size 3141252672 diff --git a/janhq_Jan-code-4b-Q6_K.gguf b/janhq_Jan-code-4b-Q6_K.gguf new file mode 100644 index 0000000..5cea891 --- /dev/null +++ b/janhq_Jan-code-4b-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0416327ef3d88582d12d5433ead674d49d5bbce5e0205f9eb6621e3e297abff2 +size 3728176192 diff --git a/janhq_Jan-code-4b-Q6_K_L.gguf b/janhq_Jan-code-4b-Q6_K_L.gguf new file mode 100644 index 0000000..b2518b2 --- /dev/null +++ b/janhq_Jan-code-4b-Q6_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f74341897225fa42f64a541e023fc8c5bee324ebbfa5b8e51745be5c436abc5 +size 3916576832 diff --git a/janhq_Jan-code-4b-Q8_0.gguf b/janhq_Jan-code-4b-Q8_0.gguf new file mode 100644 index 0000000..ac65788 --- /dev/null +++ b/janhq_Jan-code-4b-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bda699510efdaaa9ffbd984cc5dd03c4fe42ca882b2420471f589609163a07ad +size 4693669952 diff --git a/janhq_Jan-code-4b-bf16.gguf b/janhq_Jan-code-4b-bf16.gguf new file mode 100644 index 0000000..d1d90dc --- /dev/null +++ b/janhq_Jan-code-4b-bf16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a2e024f1dc4956c89258d12cb1e3ea5ecf3897756e64b9541d94e8cf8c9e019b +size 8829196096 diff --git a/janhq_Jan-code-4b-imatrix.gguf b/janhq_Jan-code-4b-imatrix.gguf new file mode 100644 index 0000000..c806aaf --- /dev/null +++ b/janhq_Jan-code-4b-imatrix.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98fa0d48f5c6337031cb1ff076294c55990772e9feff2e6fd4ff1e540f7c8f95 +size 3872640