Transformers
GGUF
English
256k context
Qwen3
Mixture of Experts
MOE
MOE Dense
2 experts
4Bx12
All use cases
bfloat16
Merge
thinking
reasoning
GPT-5.1-High-Reasoning-Distill
Gemini-3-Pro-Preview-High-Reasoning-Distill
Claude-4.5-Opus-High-Reasoning-Distill
Claude-Sonnet-4-Reasoning-Distill
Kimi-K2-Thinking-Distill
Gemini-2.5-Flash-Distill
Gemini-2.5-Flash-Lite-Preview-Distill
gpt-oss-120b-Distill
GLM-Flash-4.6-Distill
Open-R1-Distill
Command-A-Reasoning-Distill
imatrix
conversational
auto-patch README.md
Browse files
README.md
CHANGED
|
@@ -63,10 +63,12 @@ more details, including on how to concatenate multi-part files.
|
|
| 63 |
|:-----|:-----|--------:|:------|
|
| 64 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.imatrix.gguf) | imatrix | 0.1 | imatrix file (for creating your own quants) |
|
| 65 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ1_S.gguf) | i1-IQ1_S | 7.1 | for the desperate |
|
|
|
|
| 66 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 9.1 | |
|
| 67 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_XS.gguf) | i1-IQ2_XS | 10.1 | |
|
| 68 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_S.gguf) | i1-IQ2_S | 10.2 | |
|
| 69 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_M.gguf) | i1-IQ2_M | 11.2 | |
|
|
|
|
| 70 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q3_K_S.gguf) | i1-Q3_K_S | 14.7 | IQ3_XS probably better |
|
| 71 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ3_S.gguf) | i1-IQ3_S | 14.7 | beats Q3_K* |
|
| 72 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q3_K_L.gguf) | i1-Q3_K_L | 17.6 | IQ3_M probably better |
|
|
@@ -74,6 +76,7 @@ more details, including on how to concatenate multi-part files.
|
|
| 74 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q4_K_M.gguf) | i1-Q4_K_M | 20.5 | fast, recommended |
|
| 75 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q5_K_S.gguf) | i1-Q5_K_S | 23.3 | |
|
| 76 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q5_K_M.gguf) | i1-Q5_K_M | 24.0 | |
|
|
|
|
| 77 |
|
| 78 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
| 79 |
types (lower is better):
|
|
|
|
| 63 |
|:-----|:-----|--------:|:------|
|
| 64 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.imatrix.gguf) | imatrix | 0.1 | imatrix file (for creating your own quants) |
|
| 65 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ1_S.gguf) | i1-IQ1_S | 7.1 | for the desperate |
|
| 66 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ1_M.gguf) | i1-IQ1_M | 7.8 | mostly desperate |
|
| 67 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 9.1 | |
|
| 68 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_XS.gguf) | i1-IQ2_XS | 10.1 | |
|
| 69 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_S.gguf) | i1-IQ2_S | 10.2 | |
|
| 70 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ2_M.gguf) | i1-IQ2_M | 11.2 | |
|
| 71 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 13.1 | lower quality |
|
| 72 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q3_K_S.gguf) | i1-Q3_K_S | 14.7 | IQ3_XS probably better |
|
| 73 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-IQ3_S.gguf) | i1-IQ3_S | 14.7 | beats Q3_K* |
|
| 74 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q3_K_L.gguf) | i1-Q3_K_L | 17.6 | IQ3_M probably better |
|
|
|
|
| 76 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q4_K_M.gguf) | i1-Q4_K_M | 20.5 | fast, recommended |
|
| 77 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q5_K_S.gguf) | i1-Q5_K_S | 23.3 | |
|
| 78 |
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q5_K_M.gguf) | i1-Q5_K_M | 24.0 | |
|
| 79 |
+
| [GGUF](https://huggingface.co/mradermacher/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill-i1-GGUF/resolve/main/Qwen3-48B-A4B-Savant-Commander-GATED-12x-Closed-Open-Source-Distill.i1-Q6_K.gguf) | i1-Q6_K | 27.7 | practically like static Q6_K |
|
| 80 |
|
| 81 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
| 82 |
types (lower is better):
|