From 2579c5f5ca9a4c8893c6b065a0cb1aa31da9ce13 Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Fri, 10 Apr 2026 12:51:02 +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/mlabonne_Qwen3-14B-abliterated-GGUF Source: Original Platform --- .gitattributes | 61 +++++++ README.md | 180 ++++++++++++++++++++ mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf | 3 + mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf | 3 + mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf | 3 + mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf | 3 + mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf | 3 + mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf | 3 + mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q2_K.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q4_0.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q4_1.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q6_K.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf | 3 + mlabonne_Qwen3-14B-abliterated-Q8_0.gguf | 3 + mlabonne_Qwen3-14B-abliterated-bf16.gguf | 3 + mlabonne_Qwen3-14B-abliterated.imatrix | 3 + 28 files changed, 319 insertions(+) create mode 100644 .gitattributes create mode 100644 README.md create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q2_K.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q4_0.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q4_1.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q6_K.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-Q8_0.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated-bf16.gguf create mode 100644 mlabonne_Qwen3-14B-abliterated.imatrix diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..05db70d --- /dev/null +++ b/.gitattributes @@ -0,0 +1,61 @@ +*.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 +mlabonne_Qwen3-14B-abliterated-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-bf16.gguf filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated.imatrix filter=lfs diff=lfs merge=lfs -text +mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..90f9206 --- /dev/null +++ b/README.md @@ -0,0 +1,180 @@ +--- +quantized_by: bartowski +pipeline_tag: text-generation +base_model: mlabonne/Qwen3-14B-abliterated +tags: +- abliteration +- abliterated +license: apache-2.0 +license_link: https://huggingface.co/Qwen/Qwen3-14B/blob/main/LICENSE +base_model_relation: quantized +--- + +## Llamacpp imatrix Quantizations of Qwen3-14B-abliterated by mlabonne + +Using llama.cpp release b5415 for quantization. + +Original model: https://huggingface.co/mlabonne/Qwen3-14B-abliterated + +All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/eb213dccb3571f863da82e99418f81e8) + +Run them in [LM Studio](https://lmstudio.ai/) + +Run them directly with [llama.cpp](https://github.com/ggerganov/llama.cpp), or any other llama.cpp based project + +## Prompt format + +``` +<|im_start|>system +{system_prompt}<|im_end|> +<|im_start|>user +{prompt}<|im_end|> +<|im_start|>assistant +``` + +## What's new: + +Original model updated + +## Download a file (not the whole branch) from below: + +| Filename | Quant type | File Size | Split | Description | +| -------- | ---------- | --------- | ----- | ----------- | +| [Qwen3-14B-abliterated-bf16.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-bf16.gguf) | bf16 | 29.54GB | false | Full BF16 weights. | +| [Qwen3-14B-abliterated-Q8_0.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q8_0.gguf) | Q8_0 | 15.70GB | false | Extremely high quality, generally unneeded but max available quant. | +| [Qwen3-14B-abliterated-Q6_K_L.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf) | Q6_K_L | 12.50GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. | +| [Qwen3-14B-abliterated-Q6_K.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q6_K.gguf) | Q6_K | 12.12GB | false | Very high quality, near perfect, *recommended*. | +| [Qwen3-14B-abliterated-Q5_K_L.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf) | Q5_K_L | 10.99GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. | +| [Qwen3-14B-abliterated-Q5_K_M.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf) | Q5_K_M | 10.51GB | false | High quality, *recommended*. | +| [Qwen3-14B-abliterated-Q5_K_S.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf) | Q5_K_S | 10.26GB | false | High quality, *recommended*. | +| [Qwen3-14B-abliterated-Q4_K_L.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf) | Q4_K_L | 9.58GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. | +| [Qwen3-14B-abliterated-Q4_1.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q4_1.gguf) | Q4_1 | 9.39GB | false | Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon. | +| [Qwen3-14B-abliterated-Q4_K_M.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf) | Q4_K_M | 9.00GB | false | Good quality, default size for most use cases, *recommended*. | +| [Qwen3-14B-abliterated-Q3_K_XL.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf) | Q3_K_XL | 8.58GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. | +| [Qwen3-14B-abliterated-Q4_K_S.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf) | Q4_K_S | 8.57GB | false | Slightly lower quality with more space savings, *recommended*. | +| [Qwen3-14B-abliterated-Q4_0.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q4_0.gguf) | Q4_0 | 8.54GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. | +| [Qwen3-14B-abliterated-IQ4_NL.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf) | IQ4_NL | 8.54GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. | +| [Qwen3-14B-abliterated-IQ4_XS.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf) | IQ4_XS | 8.11GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. | +| [Qwen3-14B-abliterated-Q3_K_L.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf) | Q3_K_L | 7.90GB | false | Lower quality but usable, good for low RAM availability. | +| [Qwen3-14B-abliterated-Q3_K_M.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf) | Q3_K_M | 7.32GB | false | Low quality. | +| [Qwen3-14B-abliterated-IQ3_M.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf) | IQ3_M | 6.88GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. | +| [Qwen3-14B-abliterated-Q3_K_S.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf) | Q3_K_S | 6.66GB | false | Low quality, not recommended. | +| [Qwen3-14B-abliterated-Q2_K_L.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf) | Q2_K_L | 6.51GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. | +| [Qwen3-14B-abliterated-IQ3_XS.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf) | IQ3_XS | 6.38GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. | +| [Qwen3-14B-abliterated-IQ3_XXS.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf) | IQ3_XXS | 5.94GB | false | Lower quality, new method with decent performance, comparable to Q3 quants. | +| [Qwen3-14B-abliterated-Q2_K.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-Q2_K.gguf) | Q2_K | 5.75GB | false | Very low quality but surprisingly usable. | +| [Qwen3-14B-abliterated-IQ2_M.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf) | IQ2_M | 5.32GB | false | Relatively low quality, uses SOTA techniques to be surprisingly usable. | +| [Qwen3-14B-abliterated-IQ2_S.gguf](https://huggingface.co/bartowski/mlabonne_Qwen3-14B-abliterated-GGUF/blob/main/mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf) | IQ2_S | 4.96GB | false | Low quality, uses SOTA techniques to be 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/mlabonne_Qwen3-14B-abliterated-GGUF --include "mlabonne_Qwen3-14B-abliterated-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/mlabonne_Qwen3-14B-abliterated-GGUF --include "mlabonne_Qwen3-14B-abliterated-Q8_0/*" --local-dir ./ +``` + +You can either specify a new local-dir (mlabonne_Qwen3-14B-abliterated-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/ggerganov/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/ggerganov/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/ggerganov/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/ggerganov/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/mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf b/mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf new file mode 100644 index 0000000..1f5f000 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c05b774d4af43f748862331af0e5cf6b7c4f4b0ab995491fe9f954a54b80cb4 +size 5322941536 diff --git a/mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf b/mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf new file mode 100644 index 0000000..da70317 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ2_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:672e0b086d8f56537a60bd108f28f819ac7cbe7f67f34fc259f09dbef7ad4ae7 +size 4963312736 diff --git a/mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf b/mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf new file mode 100644 index 0000000..1ca1876 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5cee38b6201e572084d13f65fe648a298b2328154e369f271640b601b4a7f176 +size 6883410016 diff --git a/mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf b/mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf new file mode 100644 index 0000000..3cd078e --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eea8f9e4433a903abb1e75eb0452089e425fc6dddc6d11eb21071e5fe8246eee +size 6375301216 diff --git a/mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf b/mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf new file mode 100644 index 0000000..db19d00 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9df30e4013cbf732b55d6437cfce3765d7e8f9567001aea80b3ccceb53fd698f +size 5942666336 diff --git a/mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf b/mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf new file mode 100644 index 0000000..b0288cf --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1702dab3f9957cf71f601ea4a1beb48da06b6e4bdcc9f52cc3b28db4e4b9d041 +size 8541363296 diff --git a/mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf b/mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf new file mode 100644 index 0000000..3f0db1c --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a84ad678a549b38ae0bcd285d99b62711cf808d3e1dd5d6b0b2203c8f5e7f69 +size 8110730336 diff --git a/mlabonne_Qwen3-14B-abliterated-Q2_K.gguf b/mlabonne_Qwen3-14B-abliterated-Q2_K.gguf new file mode 100644 index 0000000..cce931d --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c87c0c198e0b3f7861f868176935916ef7f487965b1a781cf76f2b04f483b15 +size 5753984096 diff --git a/mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf b/mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf new file mode 100644 index 0000000..90a897b --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fee726573fb376291bbf08acdcca961bdbd4cb8b85891b3b979f4feef5c73092 +size 6513664096 diff --git a/mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf b/mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf new file mode 100644 index 0000000..38dc2fd --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a64bfd94ddd40fc7a9764bf1fdef7f9dd1a9d9cbf555f52c7458792fecbcc5eb +size 7900651616 diff --git a/mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf b/mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf new file mode 100644 index 0000000..5afe894 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c449ac4b31e325be266d652ea516650c6a05e68b09d374b9e51ef652a91a376d +size 7321313376 diff --git a/mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf b/mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf new file mode 100644 index 0000000..9c11665 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:43f45742bb89b6aab5d053b8df5604c06ff93b1c41f41de2338e86a276049824 +size 6657106016 diff --git a/mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf b/mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf new file mode 100644 index 0000000..bd22b20 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:074deb9caccf667e3ddaa89de100b763d5cb4f95f605a2884b8991b06449f581 +size 8581324896 diff --git a/mlabonne_Qwen3-14B-abliterated-Q4_0.gguf b/mlabonne_Qwen3-14B-abliterated-Q4_0.gguf new file mode 100644 index 0000000..37ca28a --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe68566263d6414494ffa827d1a6116b26f499b25efe67782cb102d2c800c46c +size 8543001696 diff --git a/mlabonne_Qwen3-14B-abliterated-Q4_1.gguf b/mlabonne_Qwen3-14B-abliterated-Q4_1.gguf new file mode 100644 index 0000000..a156d82 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:39c1739893f283c21707fd29e52ca82346d6d4fe58cebc1c48222291275cc57d +size 9389522016 diff --git a/mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf b/mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf new file mode 100644 index 0000000..1fc4ed0 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q4_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8eaa2980edbabdd16de66da92e00e87d9bf93103b71df23b4c220e0471d8e7b8 +size 9579110496 diff --git a/mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf b/mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf new file mode 100644 index 0000000..05e47e3 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fe972a7c6e847ec791453b89a7333d369fbde329cbd4cc9a4f0598854db5d54 +size 9001753696 diff --git a/mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf b/mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf new file mode 100644 index 0000000..2207dfb --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b87c4038cebcc654413f63fed97c75a5b9dc9a1bb018af809bb3e3a3b5db496 +size 8573475936 diff --git a/mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf b/mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf new file mode 100644 index 0000000..b979ca0 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q5_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8e5182452022f17376865add82def239e6424ded3dfc2122ce756d5c4c687cf +size 10994688096 diff --git a/mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf b/mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf new file mode 100644 index 0000000..84c4f1e --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ba161c229e6e6442864bea87b311ce36b62c0c664f0c8eac75cc43ee0fb619b +size 10514570336 diff --git a/mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf b/mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf new file mode 100644 index 0000000..36c177b --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1baf807508e5f3e2e9bc7e3b7587623ad35141bbd6d1a09bb62032f9084a89de +size 10263895136 diff --git a/mlabonne_Qwen3-14B-abliterated-Q6_K.gguf b/mlabonne_Qwen3-14B-abliterated-Q6_K.gguf new file mode 100644 index 0000000..b239988 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:db037b17caecc0f5be69fb8e0fd759810acce911fe64d61a0116d46029509a16 +size 12121938016 diff --git a/mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf b/mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf new file mode 100644 index 0000000..ee1c6c1 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q6_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1fa4ba6f3af071be5d70b6654a18702e5086b0783652268bed0ebc5904db707 +size 12498739296 diff --git a/mlabonne_Qwen3-14B-abliterated-Q8_0.gguf b/mlabonne_Qwen3-14B-abliterated-Q8_0.gguf new file mode 100644 index 0000000..4368763 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98244eaddf92baf49c9c820d0b1108d8c7906e5abd1575c7f8942873c578f9d8 +size 15698534496 diff --git a/mlabonne_Qwen3-14B-abliterated-bf16.gguf b/mlabonne_Qwen3-14B-abliterated-bf16.gguf new file mode 100644 index 0000000..0cc9f4c --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated-bf16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84a64243f08fbaf3b279e237659b61797e49623a092d21f131aab990ebdbc7d3 +size 29543423808 diff --git a/mlabonne_Qwen3-14B-abliterated.imatrix b/mlabonne_Qwen3-14B-abliterated.imatrix new file mode 100644 index 0000000..f89c5b4 --- /dev/null +++ b/mlabonne_Qwen3-14B-abliterated.imatrix @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b7caf7292d6e4309bceb90b21a43321be48ad56b05c8ac665e31b920dcd0105 +size 7709778