commit d3163e29fc9a5de0122b1e13c9252a384b32f9b5 Author: ModelHub XC Date: Sun Jun 14 13:58:17 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..e9a0681 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,60 @@ +*.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 +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K_L.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_L.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_L.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-bf16.gguf filter=lfs diff=lfs merge=lfs -text +Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-imatrix.gguf filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..972e5ac --- /dev/null +++ b/README.md @@ -0,0 +1,194 @@ +--- +quantized_by: bartowski +pipeline_tag: text-generation +base_model_relation: quantized +language: +- en +tags: +- qwen3 +- qwen +- qwen3-4b +- merged +- lora +- fp16 +- no-4bit +- transformers +- text-generation +- instruction-tuned +license: apache-2.0 +datasets: +- TeichAI/claude-haiku-4.5-1700x +base_model: Tralalabs/Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-FP16 +--- + +## Llamacpp imatrix Quantizations of Qwen3-2507-4B-Instruct-Haiku-4.5-Merged by Tralalabs + +Using llama.cpp release b9061 for quantization. + +Original model: https://huggingface.co/Tralalabs/Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-FP16 + +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 | +| -------- | ---------- | --------- | ----- | ----------- | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-bf16.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-bf16.gguf) | bf16 | 8.05GB | false | Full BF16 weights. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0.gguf) | Q8_0 | 4.28GB | false | Extremely high quality, generally unneeded but max available quant. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K_L.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K_L.gguf) | Q6_K_L | 3.57GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K.gguf) | Q6_K | 3.48GB | false | Very high quality, near perfect, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_L.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_L.gguf) | Q5_K_L | 3.07GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_M.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_M.gguf) | Q5_K_M | 2.98GB | false | High quality, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_S.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_S.gguf) | Q5_K_S | 2.85GB | false | High quality, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_L.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_L.gguf) | Q4_K_L | 2.70GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_1.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_1.gguf) | Q4_1 | 2.63GB | false | Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_M.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_M.gguf) | Q4_K_M | 2.60GB | false | Good quality, default size for most use cases, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_S.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_S.gguf) | Q4_K_S | 2.44GB | false | Slightly lower quality with more space savings, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_NL.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_NL.gguf) | IQ4_NL | 2.44GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_0.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_0.gguf) | Q4_0 | 2.42GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_XL.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_XL.gguf) | Q3_K_XL | 2.36GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_XS.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_XS.gguf) | IQ4_XS | 2.34GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_L.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_L.gguf) | Q3_K_L | 2.27GB | false | Lower quality but usable, good for low RAM availability. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_M.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_M.gguf) | Q3_K_M | 2.15GB | false | Low quality. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_M.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_M.gguf) | IQ3_M | 2.05GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_S.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_S.gguf) | Q3_K_S | 1.96GB | false | Low quality, not recommended. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XS.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XS.gguf) | IQ3_XS | 1.91GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K_L.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K_L.gguf) | Q2_K_L | 1.86GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XXS.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XXS.gguf) | IQ3_XXS | 1.78GB | false | Lower quality, new method with decent performance, comparable to Q3 quants. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K.gguf) | Q2_K | 1.76GB | false | Very low quality but surprisingly usable. | +| [Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ2_M.gguf](https://huggingface.co/bartowski/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF/blob/main/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ2_M.gguf) | IQ2_M | 1.69GB | 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/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF --include "Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-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/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-GGUF --include "Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0/*" --local-dir ./ +``` + +You can either specify a new local-dir (Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-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/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ2_M.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ2_M.gguf new file mode 100644 index 0000000..69f5780 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ2_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9784f5a11dca342f556c14cab32854c0f9ceab0eb556b39557f798560f0dda97 +size 1693904032 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_M.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_M.gguf new file mode 100644 index 0000000..69a16c1 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7005d57609d5260733f3fc8245d4128e4b11895a5cb05808ac869439479b24bf +size 2048420512 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XS.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XS.gguf new file mode 100644 index 0000000..087986b --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5843a3ddf468b0a866a8a8c1569bcaa4256b1337d5989b6b3859412755aa0fad +size 1910221472 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XXS.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XXS.gguf new file mode 100644 index 0000000..c480ca3 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ3_XXS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:642035690797d5647a203666f25578464d00cfe440dec36d624739d597b0607d +size 1782254752 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_NL.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_NL.gguf new file mode 100644 index 0000000..a177136 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_NL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ba9499bb027d1c9f8f74f014eb36ed98e7db37f32ec3334a06dea2e8287883e +size 2442537632 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_XS.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_XS.gguf new file mode 100644 index 0000000..7e33ebe --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-IQ4_XS.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:53344df249c8ae428c9b4116e87a29680e6712fcfe067d67ce449eb632ced1a8 +size 2337843872 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K.gguf new file mode 100644 index 0000000..686c943 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b47dcc05fc4470b789a674045207623e44809d1120b90df50792497f0f4d90a7 +size 1762396832 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K_L.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K_L.gguf new file mode 100644 index 0000000..5692fb7 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q2_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea92dcab89f91c333ab775c40087559f02f45e9053225e5674beb598d042fdd8 +size 1856597152 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_L.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_L.gguf new file mode 100644 index 0000000..d161e17 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6927ee28d6e75cc101e74d95bff520316890d55fb6b4d1d76e72750837fae3d +size 2266327712 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_M.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_M.gguf new file mode 100644 index 0000000..a4739da --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05e7a50f64413f8bb358c5d3ac59f98a8110cfd9be01fc741995f9556918dd65 +size 2154588832 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_S.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_S.gguf new file mode 100644 index 0000000..f00e81c --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9259680fe30d9366edc04919a78f016adeb38d36de1f5e8560aa3d9d8c10aaf6 +size 1960725152 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_XL.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_XL.gguf new file mode 100644 index 0000000..f44467a --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q3_K_XL.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eebc52f1dd771bc13d43b511df92fc9768ca4004b7055ae8585a0c25a68674c3 +size 2360528032 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_0.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_0.gguf new file mode 100644 index 0000000..91e0ba6 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aa68f0229d84882a6ff9dcd6a81b0d860d4abc6972d49c3d327dcd2c8732467b +size 2424433312 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_1.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_1.gguf new file mode 100644 index 0000000..f8e00ed --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1ae48ee0f8899b0bcb7a2afa33bb7df64e15cb213023821e6c8aeee2d43c0067 +size 2633493152 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_L.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_L.gguf new file mode 100644 index 0000000..2b541e8 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ee00b081a7d42a1f0283029bdf87f6869de0b2317f0d79296f4b447a6c00e52 +size 2699123872 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_M.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_M.gguf new file mode 100644 index 0000000..a74b737 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:25291c657bee019dccea3f947611443077d7e2943251a26b7d7b1bc40304891e +size 2604923552 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_S.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_S.gguf new file mode 100644 index 0000000..d18bf47 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9206a84a18698b113fe5577cdfc2d497903c575fca6979bcbf109127a19f850e +size 2436557472 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_L.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_L.gguf new file mode 100644 index 0000000..665d939 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4192e01437f5e64fe7b775cfe8192593dde43b9f9e7026ef4f91625f698c0320 +size 3071450272 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_M.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_M.gguf new file mode 100644 index 0000000..b854a45 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06442828e314d568c65cf79cf086df305760d3d7256ebb703c0eeae865da9844 +size 2977249952 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_S.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_S.gguf new file mode 100644 index 0000000..8cf2cff --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae03a9cb732159a67b754156f3b1ef24ed394f2ad53794ca90911c28cc9d3931 +size 2848778912 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K.gguf new file mode 100644 index 0000000..7547f4f --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:433560f934eb88409128d0d4b2eb3b96229abf2b7075aff2487ad8924e665b1b +size 3477678752 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K_L.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K_L.gguf new file mode 100644 index 0000000..e2693f9 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q6_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd66d1da9726090035bd75a18eb27b48663117f3bbb7ef1a534b4f564c16318f +size 3571879072 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0.gguf new file mode 100644 index 0000000..8105e44 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b95e44c7190692bc98c6d4b8bbac730cb20847e681718648bd2a913630c1c905 +size 4280405152 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-bf16.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-bf16.gguf new file mode 100644 index 0000000..19c13d2 --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-bf16.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a1ea1a4c7085f84804e24270b39fcce33a94301bad3b3c790c00a83cbf48e7f +size 8051284800 diff --git a/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-imatrix.gguf b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-imatrix.gguf new file mode 100644 index 0000000..3cf9f0f --- /dev/null +++ b/Tralalabs_Qwen3-2507-4B-Instruct-Haiku-4.5-Merged-imatrix.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fe323566f1949f4a04699ac7a3b611789f40626270e181e0fffae5f0025aa07 +size 3872640