Hugging Face's logo

mradermacher
/
WinterGoliath-123b-i1-GGUF

Transformers
GGUF
English
Merge
Inference Endpoints
Model card Files Community
mradermacher commited on
Commit
2723684
•
1 Parent(s): ec7e149

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -0
README.md CHANGED
@@ -27,15 +27,20 @@ more details, including on how to concatenate multi-part files.
27
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ1_S.gguf) | i1-IQ1_S | 26.4 | for the desperate |
28
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 33.2 | |
29
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_XS.gguf) | i1-IQ2_XS | 36.8 | |
 
30
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_M.gguf) | i1-IQ2_M | 42.0 | |
31
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q2_K.gguf) | i1-Q2_K | 45.8 | IQ3_XXS probably better |
32
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 47.9 | fast, lower quality |
33
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_XS.gguf.part2of2) | i1-IQ3_XS | 50.8 | |
 
34
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_S.gguf.part2of2) | i1-IQ3_S | 53.9 | fast, beats Q3_K* |
35
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_M.gguf.part2of2) | i1-IQ3_M | 55.7 | |
36
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_M.gguf.part2of2) | i1-Q3_K_M | 59.9 | IQ3_S probably better |
 
37
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_S.gguf.part2of2) | i1-Q4_K_S | 70.6 | almost as good as Q4_K_M |
38
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_M.gguf.part2of2) | i1-Q4_K_M | 74.6 | fast, medium quality |
 
 
39
 
40
 
41
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
27
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ1_S.gguf) | i1-IQ1_S | 26.4 | for the desperate |
28
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 33.2 | |
29
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_XS.gguf) | i1-IQ2_XS | 36.8 | |
30
+ | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_S.gguf) | i1-IQ2_S | 38.6 | |
31
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ2_M.gguf) | i1-IQ2_M | 42.0 | |
32
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q2_K.gguf) | i1-Q2_K | 45.8 | IQ3_XXS probably better |
33
  | [GGUF](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 47.9 | fast, lower quality |
34
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_XS.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_XS.gguf.part2of2) | i1-IQ3_XS | 50.8 | |
35
+ | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_S.gguf.part2of2) | i1-Q3_K_S | 53.7 | IQ3_XS probably better |
36
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_S.gguf.part2of2) | i1-IQ3_S | 53.9 | fast, beats Q3_K* |
37
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-IQ3_M.gguf.part2of2) | i1-IQ3_M | 55.7 | |
38
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_M.gguf.part2of2) | i1-Q3_K_M | 59.9 | IQ3_S probably better |
39
+ | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_L.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q3_K_L.gguf.part2of2) | i1-Q3_K_L | 65.2 | IQ3_M probably better |
40
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_S.gguf.part2of2) | i1-Q4_K_S | 70.6 | almost as good as Q4_K_M |
41
  | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q4_K_M.gguf.part2of2) | i1-Q4_K_M | 74.6 | fast, medium quality |
42
+ | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q5_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q5_K_S.gguf.part2of2) | i1-Q5_K_S | 85.5 | |
43
+ | [PART 1](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q5_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/WinterGoliath-123b-i1-GGUF/resolve/main/WinterGoliath-123b.i1-Q5_K_M.gguf.part2of2) | i1-Q5_K_M | 87.8 | best weighted quant |
44
 
45
 
46
  Here is a handy graph by ikawrakow comparing some lower-quality quant