mradermacher
commited on
auto-patch README.md
Browse files
README.md
CHANGED
@@ -46,7 +46,17 @@ more details, including on how to concatenate multi-part files.
|
|
46 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_XS.gguf.part2of2) | i1-IQ2_XS | 64.1 | |
|
47 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_S.gguf.part2of2) | i1-IQ2_S | 68.2 | |
|
48 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_M.gguf.part2of2) | i1-IQ2_M | 74.0 | |
|
|
|
49 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_XXS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_XXS.gguf.part2of2) | i1-IQ3_XXS | 83.6 | lower quality |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
|
51 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
52 |
types (lower is better):
|
|
|
46 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_XS.gguf.part2of2) | i1-IQ2_XS | 64.1 | |
|
47 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_S.gguf.part2of2) | i1-IQ2_S | 68.2 | |
|
48 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ2_M.gguf.part2of2) | i1-IQ2_M | 74.0 | |
|
49 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q2_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q2_K.gguf.part2of2) | i1-Q2_K | 80.4 | IQ3_XXS probably better |
|
50 |
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_XXS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_XXS.gguf.part2of2) | i1-IQ3_XXS | 83.6 | lower quality |
|
51 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_XS.gguf.part2of2) | i1-IQ3_XS | 89.2 | |
|
52 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_S.gguf.part2of2) | i1-Q3_K_S | 94.0 | IQ3_XS probably better |
|
53 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_S.gguf.part2of2) | i1-IQ3_S | 94.3 | beats Q3_K* |
|
54 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-IQ3_M.gguf.part2of2) | i1-IQ3_M | 98.3 | |
|
55 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_M.gguf.part1of3) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_M.gguf.part2of3) [PART 3](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_M.gguf.part3of3) | i1-Q3_K_M | 105.2 | IQ3_S probably better |
|
56 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_L.gguf.part1of3) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_L.gguf.part2of3) [PART 3](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q3_K_L.gguf.part3of3) | i1-Q3_K_L | 114.9 | IQ3_M probably better |
|
57 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q4_K_S.gguf.part1of3) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q4_K_S.gguf.part2of3) [PART 3](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q4_K_S.gguf.part3of3) | i1-Q4_K_S | 123.8 | optimal size/speed/quality |
|
58 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q4_K_M.gguf.part1of3) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q4_K_M.gguf.part2of3) [PART 3](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q4_K_M.gguf.part3of3) | i1-Q4_K_M | 130.2 | fast, recommended |
|
59 |
+
| [PART 1](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q6_K.gguf.part1of4) [PART 2](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q6_K.gguf.part2of4) [PART 3](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q6_K.gguf.part3of4) [PART 4](https://huggingface.co/mradermacher/Mistral-Large-218B-Instruct-i1-GGUF/resolve/main/Mistral-Large-218B-Instruct.i1-Q6_K.gguf.part4of4) | i1-Q6_K | 179.0 | practically like static Q6_K |
|
60 |
|
61 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
62 |
types (lower is better):
|