Hugging Face's logo

mradermacher
/
Mixtral_34Bx2_MoE_60B-GGUF

Transformers
GGUF
English
yi
Mixture of Experts
Inference Endpoints
Model card Files Community
mradermacher commited on
Commit
1963939
•
1 Parent(s): bdfd847

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -2
README.md CHANGED
@@ -39,14 +39,13 @@ more details, including on how to concatenate multi-part files.
39
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.IQ4_XS.gguf) | IQ4_XS | 33.5 | |
40
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q4_0.gguf) | Q4_0 | 35.0 | fast, low quality |
41
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q4_K_S.gguf) | Q4_K_S | 35.2 | fast, recommended |
42
- | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.IQ4_NL.gguf) | IQ4_NL | 35.3 | slightly worse than Q4_K_S |
43
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q4_K_M.gguf) | Q4_K_M | 37.3 | fast, recommended |
44
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q5_K_S.gguf) | Q5_K_S | 42.5 | |
45
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q5_K_M.gguf) | Q5_K_M | 43.7 | |
46
  | [PART 1](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q6_K.gguf.part2of2) | Q6_K | 50.5 | very good quality |
47
  | [PART 1](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q8_0.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q8_0.gguf.part2of2) | Q8_0 | 65.1 | fast, best quality |
48
 
49
-
50
  Here is a handy graph by ikawrakow comparing some lower-quality quant
51
  types (lower is better):
52
 
 
39
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.IQ4_XS.gguf) | IQ4_XS | 33.5 | |
40
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q4_0.gguf) | Q4_0 | 35.0 | fast, low quality |
41
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q4_K_S.gguf) | Q4_K_S | 35.2 | fast, recommended |
42
+ | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.IQ4_NL.gguf) | IQ4_NL | 35.3 | prefer IQ4_XS |
43
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q4_K_M.gguf) | Q4_K_M | 37.3 | fast, recommended |
44
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q5_K_S.gguf) | Q5_K_S | 42.5 | |
45
  | [GGUF](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q5_K_M.gguf) | Q5_K_M | 43.7 | |
46
  | [PART 1](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q6_K.gguf.part2of2) | Q6_K | 50.5 | very good quality |
47
  | [PART 1](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q8_0.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Mixtral_34Bx2_MoE_60B-GGUF/resolve/main/Mixtral_34Bx2_MoE_60B.Q8_0.gguf.part2of2) | Q8_0 | 65.1 | fast, best quality |
48
 
 
49
  Here is a handy graph by ikawrakow comparing some lower-quality quant
50
  types (lower is better):
51