auto-patch README.md
Browse files
README.md
CHANGED
@@ -58,18 +58,28 @@ more details, including on how to concatenate multi-part files.
|
|
58 |
|
59 |
| Link | Type | Size/GB | Notes |
|
60 |
|:-----|:-----|--------:|:------|
|
|
|
61 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ1_M.gguf) | i1-IQ1_M | 27.2 | mostly desperate |
|
62 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 29.2 | |
|
|
|
|
|
63 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ2_M.gguf) | i1-IQ2_M | 33.6 | |
|
64 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q2_K_S.gguf) | i1-Q2_K_S | 33.8 | very low quality |
|
65 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q2_K.gguf) | i1-Q2_K | 34.1 | IQ3_XXS probably better |
|
66 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 36.4 | lower quality |
|
|
|
67 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q3_K_S.gguf) | i1-Q3_K_S | 39.3 | IQ3_XS probably better |
|
|
|
68 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ3_M.gguf) | i1-IQ3_M | 40.6 | |
|
69 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q3_K_M.gguf) | i1-Q3_K_M | 43.1 | IQ3_S probably better |
|
|
|
70 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ4_XS.gguf) | i1-IQ4_XS | 45.5 | |
|
|
|
71 |
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_S.gguf.part2of2) | i1-Q4_K_S | 50.2 | optimal size/speed/quality |
|
|
|
72 |
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_M.gguf.part2of2) | i1-Q4_K_M | 54.3 | fast, recommended |
|
|
|
|
|
73 |
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 73.8 | practically like static Q6_K |
|
74 |
|
75 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|
|
|
58 |
|
59 |
| Link | Type | Size/GB | Notes |
|
60 |
|:-----|:-----|--------:|:------|
|
61 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ1_S.gguf) | i1-IQ1_S | 26.0 | for the desperate |
|
62 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ1_M.gguf) | i1-IQ1_M | 27.2 | mostly desperate |
|
63 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 29.2 | |
|
64 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ2_XS.gguf) | i1-IQ2_XS | 31.0 | |
|
65 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ2_S.gguf) | i1-IQ2_S | 32.0 | |
|
66 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ2_M.gguf) | i1-IQ2_M | 33.6 | |
|
67 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q2_K_S.gguf) | i1-Q2_K_S | 33.8 | very low quality |
|
68 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q2_K.gguf) | i1-Q2_K | 34.1 | IQ3_XXS probably better |
|
69 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 36.4 | lower quality |
|
70 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ3_XS.gguf) | i1-IQ3_XS | 37.5 | |
|
71 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q3_K_S.gguf) | i1-Q3_K_S | 39.3 | IQ3_XS probably better |
|
72 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ3_S.gguf) | i1-IQ3_S | 39.4 | beats Q3_K* |
|
73 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ3_M.gguf) | i1-IQ3_M | 40.6 | |
|
74 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q3_K_M.gguf) | i1-Q3_K_M | 43.1 | IQ3_S probably better |
|
75 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q3_K_L.gguf) | i1-Q3_K_L | 45.3 | IQ3_M probably better |
|
76 |
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-IQ4_XS.gguf) | i1-IQ4_XS | 45.5 | |
|
77 |
+
| [GGUF](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_0.gguf) | i1-Q4_0 | 47.4 | fast, low quality |
|
78 |
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_S.gguf.part2of2) | i1-Q4_K_S | 50.2 | optimal size/speed/quality |
|
79 |
+
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_1.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_1.gguf.part2of2) | i1-Q4_1 | 52.4 | |
|
80 |
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q4_K_M.gguf.part2of2) | i1-Q4_K_M | 54.3 | fast, recommended |
|
81 |
+
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q5_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q5_K_S.gguf.part2of2) | i1-Q5_K_S | 58.9 | |
|
82 |
+
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q5_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q5_K_M.gguf.part2of2) | i1-Q5_K_M | 62.4 | |
|
83 |
| [PART 1](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Babel-83B-Chat-i1-GGUF/resolve/main/Babel-83B-Chat.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 73.8 | practically like static Q6_K |
|
84 |
|
85 |
Here is a handy graph by ikawrakow comparing some lower-quality quant
|