From 8ea3af373e9a61bea9f78d2b8bbfdc3e6dd43682 Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Thu, 16 Jul 2026 16:35:06 +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/soob3123_amoral-gemma3-12B-v2-GGUF Source: Original Platform --- .gitattributes | 49 ++++++ README.md | 183 +++++++++++++++++++++ configuration.json | 1 + soob3123_amoral-gemma3-12B-v2-IQ2_M.gguf | 3 + soob3123_amoral-gemma3-12B-v2-IQ2_S.gguf | 3 + soob3123_amoral-gemma3-12B-v2-IQ3_M.gguf | 3 + soob3123_amoral-gemma3-12B-v2-IQ3_XS.gguf | 3 + soob3123_amoral-gemma3-12B-v2-IQ3_XXS.gguf | 3 + soob3123_amoral-gemma3-12B-v2-IQ4_NL.gguf | 3 + soob3123_amoral-gemma3-12B-v2-IQ4_XS.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q2_K.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q2_K_L.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q3_K_L.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q3_K_M.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q3_K_S.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q3_K_XL.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q4_0.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q4_1.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q4_K_L.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q4_K_M.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q4_K_S.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q5_K_L.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q5_K_M.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q5_K_S.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q6_K.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q6_K_L.gguf | 3 + soob3123_amoral-gemma3-12B-v2-Q8_0.gguf | 3 + soob3123_amoral-gemma3-12B-v2-bf16.gguf | 3 + soob3123_amoral-gemma3-12B-v2.imatrix | 3 + 29 files changed, 311 insertions(+) create mode 100644 .gitattributes create mode 100644 README.md create mode 100644 configuration.json create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ2_M.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ2_S.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ3_M.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ3_XS.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ3_XXS.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ4_NL.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-IQ4_XS.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q2_K.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q2_K_L.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q3_K_L.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q3_K_M.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q3_K_S.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q3_K_XL.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q4_0.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q4_1.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q4_K_L.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q4_K_M.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q4_K_S.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q5_K_L.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q5_K_M.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q5_K_S.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q6_K.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q6_K_L.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-Q8_0.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2-bf16.gguf create mode 100644 soob3123_amoral-gemma3-12B-v2.imatrix diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..6d5fac8 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,49 @@ +*.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 +*.gguf* 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 + +soob3123_amoral-gemma3-12B-v2.imatrix 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..a2d89d2 --- /dev/null +++ b/README.md @@ -0,0 +1,183 @@ +--- +quantized_by: bartowski +pipeline_tag: text-generation +tags: +- text-generation-inference +- transformers +- gemma3 +- analytical-tasks +- bias-neutralization +- uncensored +base_model: soob3123/amoral-gemma3-12B-v2 +language: +- en +license: apache-2.0 +base_model_relation: quantized +--- + +## Llamacpp imatrix Quantizations of amoral-gemma3-12B-v2 by soob3123 + +Using llama.cpp release b4978 for quantization. + +Original model: https://huggingface.co/soob3123/amoral-gemma3-12B-v2 + +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 + +``` +user +{system_prompt} + +{prompt} +model + +model +``` + +## Download a file (not the whole branch) from below: + +| Filename | Quant type | File Size | Split | Description | +| -------- | ---------- | --------- | ----- | ----------- | +| [amoral-gemma3-12B-v2-bf16.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-bf16.gguf) | bf16 | 23.54GB | false | Full BF16 weights. | +| [amoral-gemma3-12B-v2-Q8_0.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q8_0.gguf) | Q8_0 | 12.51GB | false | Extremely high quality, generally unneeded but max available quant. | +| [amoral-gemma3-12B-v2-Q6_K_L.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q6_K_L.gguf) | Q6_K_L | 9.90GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. | +| [amoral-gemma3-12B-v2-Q6_K.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q6_K.gguf) | Q6_K | 9.66GB | false | Very high quality, near perfect, *recommended*. | +| [amoral-gemma3-12B-v2-Q5_K_L.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q5_K_L.gguf) | Q5_K_L | 8.69GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. | +| [amoral-gemma3-12B-v2-Q5_K_M.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q5_K_M.gguf) | Q5_K_M | 8.45GB | false | High quality, *recommended*. | +| [amoral-gemma3-12B-v2-Q5_K_S.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q5_K_S.gguf) | Q5_K_S | 8.23GB | false | High quality, *recommended*. | +| [amoral-gemma3-12B-v2-Q4_1.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q4_1.gguf) | Q4_1 | 7.56GB | false | Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon. | +| [amoral-gemma3-12B-v2-Q4_K_L.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q4_K_L.gguf) | Q4_K_L | 7.54GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. | +| [amoral-gemma3-12B-v2-Q4_K_M.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q4_K_M.gguf) | Q4_K_M | 7.30GB | false | Good quality, default size for most use cases, *recommended*. | +| [amoral-gemma3-12B-v2-Q4_K_S.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q4_K_S.gguf) | Q4_K_S | 6.94GB | false | Slightly lower quality with more space savings, *recommended*. | +| [amoral-gemma3-12B-v2-Q4_0.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q4_0.gguf) | Q4_0 | 6.91GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. | +| [amoral-gemma3-12B-v2-IQ4_NL.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ4_NL.gguf) | IQ4_NL | 6.89GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. | +| [amoral-gemma3-12B-v2-Q3_K_XL.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q3_K_XL.gguf) | Q3_K_XL | 6.72GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. | +| [amoral-gemma3-12B-v2-IQ4_XS.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ4_XS.gguf) | IQ4_XS | 6.55GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. | +| [amoral-gemma3-12B-v2-Q3_K_L.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q3_K_L.gguf) | Q3_K_L | 6.48GB | false | Lower quality but usable, good for low RAM availability. | +| [amoral-gemma3-12B-v2-Q3_K_M.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q3_K_M.gguf) | Q3_K_M | 6.01GB | false | Low quality. | +| [amoral-gemma3-12B-v2-IQ3_M.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ3_M.gguf) | IQ3_M | 5.66GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. | +| [amoral-gemma3-12B-v2-Q3_K_S.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q3_K_S.gguf) | Q3_K_S | 5.46GB | false | Low quality, not recommended. | +| [amoral-gemma3-12B-v2-IQ3_XS.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ3_XS.gguf) | IQ3_XS | 5.21GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. | +| [amoral-gemma3-12B-v2-Q2_K_L.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q2_K_L.gguf) | Q2_K_L | 5.01GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. | +| [amoral-gemma3-12B-v2-IQ3_XXS.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ3_XXS.gguf) | IQ3_XXS | 4.78GB | false | Lower quality, new method with decent performance, comparable to Q3 quants. | +| [amoral-gemma3-12B-v2-Q2_K.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-Q2_K.gguf) | Q2_K | 4.77GB | false | Very low quality but surprisingly usable. | +| [amoral-gemma3-12B-v2-IQ2_M.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ2_M.gguf) | IQ2_M | 4.31GB | false | Relatively low quality, uses SOTA techniques to be surprisingly usable. | +| [amoral-gemma3-12B-v2-IQ2_S.gguf](https://huggingface.co/bartowski/soob3123_amoral-gemma3-12B-v2-GGUF/blob/main/soob3123_amoral-gemma3-12B-v2-IQ2_S.gguf) | IQ2_S | 4.02GB | 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/soob3123_amoral-gemma3-12B-v2-GGUF --include "soob3123_amoral-gemma3-12B-v2-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/soob3123_amoral-gemma3-12B-v2-GGUF --include "soob3123_amoral-gemma3-12B-v2-Q8_0/*" --local-dir ./ +``` + +You can either specify a new local-dir (soob3123_amoral-gemma3-12B-v2-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/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/soob3123_amoral-gemma3-12B-v2-IQ2_M.gguf b/soob3123_amoral-gemma3-12B-v2-IQ2_M.gguf new file mode 100644 index 0000000..c87c5cf --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5ff2becae2e5f869a55e88d852b2aba82b8de7e447b114a9947953ad275da522 +size 4310461376 diff --git a/soob3123_amoral-gemma3-12B-v2-IQ2_S.gguf b/soob3123_amoral-gemma3-12B-v2-IQ2_S.gguf new file mode 100644 index 0000000..a675b4f --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ2_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c969713c17b6bd94aed7cb7a6004d69cc4f163f4482372750eefe7611859cb04 +size 4020710336 diff --git a/soob3123_amoral-gemma3-12B-v2-IQ3_M.gguf b/soob3123_amoral-gemma3-12B-v2-IQ3_M.gguf new file mode 100644 index 0000000..69ce369 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ddf20ff224ac508d01b73aba485d4e9952aaa02de6baa5cfa8657db5ce2c1a9 +size 5655722816 diff --git a/soob3123_amoral-gemma3-12B-v2-IQ3_XS.gguf b/soob3123_amoral-gemma3-12B-v2-IQ3_XS.gguf new file mode 100644 index 0000000..d28afdc --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90878f89af8065afad790e91980388beb66b40df431e2f436f2176fcc7feb2d4 +size 5206166336 diff --git a/soob3123_amoral-gemma3-12B-v2-IQ3_XXS.gguf b/soob3123_amoral-gemma3-12B-v2-IQ3_XXS.gguf new file mode 100644 index 0000000..0da54b6 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:423c38861cfdd51eca3e5b0c23f1941f9281a88d503c7e03817f5fc061f6b553 +size 4784901056 diff --git a/soob3123_amoral-gemma3-12B-v2-IQ4_NL.gguf b/soob3123_amoral-gemma3-12B-v2-IQ4_NL.gguf new file mode 100644 index 0000000..ddbc650 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27904d0ed08c846fc0073b427835a9e66ac78ce07785dff36d68e75a2a8aa523 +size 6887164736 diff --git a/soob3123_amoral-gemma3-12B-v2-IQ4_XS.gguf b/soob3123_amoral-gemma3-12B-v2-IQ4_XS.gguf new file mode 100644 index 0000000..4bd9fa2 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:342226b1e896e4705e4a70af23636cf972c0f20f8b9a15dc2aa1ace7e6b64560 +size 6550965056 diff --git a/soob3123_amoral-gemma3-12B-v2-Q2_K.gguf b/soob3123_amoral-gemma3-12B-v2-Q2_K.gguf new file mode 100644 index 0000000..5e32f0a --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06420e06a2c0691be5017eb6e429a48f541e4ab0a0c71619b64cb16543312a61 +size 4768222016 diff --git a/soob3123_amoral-gemma3-12B-v2-Q2_K_L.gguf b/soob3123_amoral-gemma3-12B-v2-Q2_K_L.gguf new file mode 100644 index 0000000..ca480b9 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ad581fa861778be0c50cdf82ca940ad20fa9379a8dd8c5572f2f935c9cbe84a9 +size 5012075456 diff --git a/soob3123_amoral-gemma3-12B-v2-Q3_K_L.gguf b/soob3123_amoral-gemma3-12B-v2-Q3_K_L.gguf new file mode 100644 index 0000000..04180dc --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbb5c8a675de153663ac1c68d35203ee335e0ac65aefc0013e6b2f2ae2bc4e81 +size 6480186176 diff --git a/soob3123_amoral-gemma3-12B-v2-Q3_K_M.gguf b/soob3123_amoral-gemma3-12B-v2-Q3_K_M.gguf new file mode 100644 index 0000000..11b4a4f --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:518004705d082a2cb3c67322d650278e53a5ec95d7287e33088ca90c94bfcbaf +size 6008818496 diff --git a/soob3123_amoral-gemma3-12B-v2-Q3_K_S.gguf b/soob3123_amoral-gemma3-12B-v2-Q3_K_S.gguf new file mode 100644 index 0000000..d13c261 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9aa8e1f032ce7d3446f32813737cc8b5e2c9cf4d3f2fd6f177ccad6341ffe44 +size 5458316096 diff --git a/soob3123_amoral-gemma3-12B-v2-Q3_K_XL.gguf b/soob3123_amoral-gemma3-12B-v2-Q3_K_XL.gguf new file mode 100644 index 0000000..92aff45 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eab6b0a9985f048953b034e04d08e1886c375dbb9b4b0446acd7de63ac6e3d16 +size 6724039616 diff --git a/soob3123_amoral-gemma3-12B-v2-Q4_0.gguf b/soob3123_amoral-gemma3-12B-v2-Q4_0.gguf new file mode 100644 index 0000000..74b7986 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:983b19e049b1a464fc99003d23cb5b52d03eb17162cd98ed0130cc8d0263608f +size 6909283136 diff --git a/soob3123_amoral-gemma3-12B-v2-Q4_1.gguf b/soob3123_amoral-gemma3-12B-v2-Q4_1.gguf new file mode 100644 index 0000000..85cd226 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a30a023c5d26aaedad7e363a01cafea6fde7c6c8e6644b5b785ec820f142ac74 +size 7559564096 diff --git a/soob3123_amoral-gemma3-12B-v2-Q4_K_L.gguf b/soob3123_amoral-gemma3-12B-v2-Q4_K_L.gguf new file mode 100644 index 0000000..782d5f2 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q4_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:55025bc42d68ab78bf899aec176b7a7670812d06058dc4d45af998718e9b5fad +size 7544632256 diff --git a/soob3123_amoral-gemma3-12B-v2-Q4_K_M.gguf b/soob3123_amoral-gemma3-12B-v2-Q4_K_M.gguf new file mode 100644 index 0000000..c6ef565 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb5792cf73bac3dbaa39e3a79ec01a056affff4607b96f96c9b911c877d5a50a +size 7300778816 diff --git a/soob3123_amoral-gemma3-12B-v2-Q4_K_S.gguf b/soob3123_amoral-gemma3-12B-v2-Q4_K_S.gguf new file mode 100644 index 0000000..ccdf4ca --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dec90529c7babe2df0b566bcaf0aedd72a82027d000497bb8609f6d0cf0360e1 +size 6935333696 diff --git a/soob3123_amoral-gemma3-12B-v2-Q5_K_L.gguf b/soob3123_amoral-gemma3-12B-v2-Q5_K_L.gguf new file mode 100644 index 0000000..efb0e2e --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q5_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66ffc3eabb84d4a664fcd38113e2776060123fb79bf0075ba9f3fa0b8bb124c9 +size 8688890816 diff --git a/soob3123_amoral-gemma3-12B-v2-Q5_K_M.gguf b/soob3123_amoral-gemma3-12B-v2-Q5_K_M.gguf new file mode 100644 index 0000000..2982f67 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a91e3d1226a1871d607e81d726e979f803e15ee90d2e16efb01a834a71c86cbb +size 8445037376 diff --git a/soob3123_amoral-gemma3-12B-v2-Q5_K_S.gguf b/soob3123_amoral-gemma3-12B-v2-Q5_K_S.gguf new file mode 100644 index 0000000..815e17f --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8cdc99b96f96c9b3838304e99708167b6463753b9acf1f173207b050d3a35ce +size 8231963456 diff --git a/soob3123_amoral-gemma3-12B-v2-Q6_K.gguf b/soob3123_amoral-gemma3-12B-v2-Q6_K.gguf new file mode 100644 index 0000000..c9eb035 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca211c797032077bad38bba0750699d042dfe5e1cff9de0edd4cb4ff0c1abc7c +size 9660812096 diff --git a/soob3123_amoral-gemma3-12B-v2-Q6_K_L.gguf b/soob3123_amoral-gemma3-12B-v2-Q6_K_L.gguf new file mode 100644 index 0000000..e54b565 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q6_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b48b676e42ad7700e0dfd31451e426ef1bf9abc64d6bd47b87d3ded13127f22 +size 9904665536 diff --git a/soob3123_amoral-gemma3-12B-v2-Q8_0.gguf b/soob3123_amoral-gemma3-12B-v2-Q8_0.gguf new file mode 100644 index 0000000..0e0238c --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0cac1ed3e20f9f2fd5bb97557c1479103b5cc7a37b50264b77cfc5c649ce4e32 +size 12510213056 diff --git a/soob3123_amoral-gemma3-12B-v2-bf16.gguf b/soob3123_amoral-gemma3-12B-v2-bf16.gguf new file mode 100644 index 0000000..ed62564 --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2-bf16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bec1c03196cd27ff40e68c1800e39dc6b10d8e8dfe84dea0666ac5fdd8d730f6 +size 23540151968 diff --git a/soob3123_amoral-gemma3-12B-v2.imatrix b/soob3123_amoral-gemma3-12B-v2.imatrix new file mode 100644 index 0000000..6c7b09f --- /dev/null +++ b/soob3123_amoral-gemma3-12B-v2.imatrix @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f985799b855f2ade34f47e039810b2a8ad493b2eb0714ffbdb0372ef4c3cc7a1 +size 7433114