mradermacher commited on
Commit
80302a9
·
verified ·
1 Parent(s): 1c7d6a3

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +17 -0
README.md CHANGED
@@ -37,11 +37,28 @@ more details, including on how to concatenate multi-part files.
37
 
38
  | Link | Type | Size/GB | Notes |
39
  |:-----|:-----|--------:|:------|
 
 
 
 
 
 
 
40
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q2_K.gguf) | i1-Q2_K | 25.7 | IQ3_XXS probably better |
41
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 26.8 | lower quality |
 
 
 
42
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ3_M.gguf) | i1-IQ3_M | 31.2 | |
43
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q3_K_M.gguf) | i1-Q3_K_M | 33.5 | IQ3_S probably better |
 
 
 
44
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q4_K_S.gguf) | i1-Q4_K_S | 39.5 | optimal size/speed/quality |
 
 
 
 
45
 
46
  Here is a handy graph by ikawrakow comparing some lower-quality quant
47
  types (lower is better):
 
37
 
38
  | Link | Type | Size/GB | Notes |
39
  |:-----|:-----|--------:|:------|
40
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ1_S.gguf) | i1-IQ1_S | 14.7 | for the desperate |
41
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ1_M.gguf) | i1-IQ1_M | 16.2 | mostly desperate |
42
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 18.5 | |
43
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ2_XS.gguf) | i1-IQ2_XS | 20.5 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ2_S.gguf) | i1-IQ2_S | 21.6 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ2_M.gguf) | i1-IQ2_M | 23.5 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q2_K_S.gguf) | i1-Q2_K_S | 23.8 | very low quality |
47
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q2_K.gguf) | i1-Q2_K | 25.7 | IQ3_XXS probably better |
48
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 26.8 | lower quality |
49
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ3_XS.gguf) | i1-IQ3_XS | 28.6 | |
50
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ3_S.gguf) | i1-IQ3_S | 30.2 | beats Q3_K* |
51
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q3_K_S.gguf) | i1-Q3_K_S | 30.2 | IQ3_XS probably better |
52
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ3_M.gguf) | i1-IQ3_M | 31.2 | |
53
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q3_K_M.gguf) | i1-Q3_K_M | 33.5 | IQ3_S probably better |
54
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q3_K_L.gguf) | i1-Q3_K_L | 36.4 | IQ3_M probably better |
55
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-IQ4_XS.gguf) | i1-IQ4_XS | 37.1 | |
56
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q4_0.gguf) | i1-Q4_0 | 39.3 | fast, low quality |
57
  | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q4_K_S.gguf) | i1-Q4_K_S | 39.5 | optimal size/speed/quality |
58
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q4_K_M.gguf) | i1-Q4_K_M | 41.7 | fast, recommended |
59
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q5_K_S.gguf) | i1-Q5_K_S | 47.7 | |
60
+ | [GGUF](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q5_K_M.gguf) | i1-Q5_K_M | 49.0 | |
61
+ | [PART 1](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/karakuri-lm-70b-chat-v0.1-i1-GGUF/resolve/main/karakuri-lm-70b-chat-v0.1.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 56.9 | practically like static Q6_K |
62
 
63
  Here is a handy graph by ikawrakow comparing some lower-quality quant
64
  types (lower is better):