mradermacher commited on
Commit
57e6ce4
·
verified ·
1 Parent(s): 64b188d

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -0
README.md CHANGED
@@ -56,14 +56,19 @@ more details, including on how to concatenate multi-part files.
56
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q2_K_S.gguf) | i1-Q2_K_S | 12.0 | very low quality |
57
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q2_K.gguf) | i1-Q2_K | 12.9 | IQ3_XXS probably better |
58
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 13.4 | lower quality |
 
59
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q3_K_S.gguf) | i1-Q3_K_S | 15.1 | IQ3_XS probably better |
 
60
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ3_M.gguf) | i1-IQ3_M | 15.7 | |
61
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q3_K_M.gguf) | i1-Q3_K_M | 16.8 | IQ3_S probably better |
62
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q3_K_L.gguf) | i1-Q3_K_L | 18.2 | IQ3_M probably better |
63
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ4_XS.gguf) | i1-IQ4_XS | 18.6 | |
 
64
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q4_K_S.gguf) | i1-Q4_K_S | 19.7 | optimal size/speed/quality |
65
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q4_K_M.gguf) | i1-Q4_K_M | 20.8 | fast, recommended |
 
66
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q5_K_S.gguf) | i1-Q5_K_S | 23.8 | |
 
67
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q6_K.gguf) | i1-Q6_K | 28.3 | practically like static Q6_K |
68
 
69
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
56
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q2_K_S.gguf) | i1-Q2_K_S | 12.0 | very low quality |
57
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q2_K.gguf) | i1-Q2_K | 12.9 | IQ3_XXS probably better |
58
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 13.4 | lower quality |
59
+ | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ3_XS.gguf) | i1-IQ3_XS | 14.3 | |
60
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q3_K_S.gguf) | i1-Q3_K_S | 15.1 | IQ3_XS probably better |
61
+ | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ3_S.gguf) | i1-IQ3_S | 15.1 | beats Q3_K* |
62
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ3_M.gguf) | i1-IQ3_M | 15.7 | |
63
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q3_K_M.gguf) | i1-Q3_K_M | 16.8 | IQ3_S probably better |
64
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q3_K_L.gguf) | i1-Q3_K_L | 18.2 | IQ3_M probably better |
65
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-IQ4_XS.gguf) | i1-IQ4_XS | 18.6 | |
66
+ | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q4_0.gguf) | i1-Q4_0 | 19.6 | fast, low quality |
67
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q4_K_S.gguf) | i1-Q4_K_S | 19.7 | optimal size/speed/quality |
68
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q4_K_M.gguf) | i1-Q4_K_M | 20.8 | fast, recommended |
69
+ | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q4_1.gguf) | i1-Q4_1 | 21.7 | |
70
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q5_K_S.gguf) | i1-Q5_K_S | 23.8 | |
71
+ | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q5_K_M.gguf) | i1-Q5_K_M | 24.4 | |
72
  | [GGUF](https://huggingface.co/mradermacher/CapyberaHermesYi-34B-ChatML-200K-i1-GGUF/resolve/main/CapyberaHermesYi-34B-ChatML-200K.i1-Q6_K.gguf) | i1-Q6_K | 28.3 | practically like static Q6_K |
73
 
74
  Here is a handy graph by ikawrakow comparing some lower-quality quant