mradermacher commited on
Commit
c140491
1 Parent(s): 986331e

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +16 -1
README.md CHANGED
@@ -19,7 +19,6 @@ tags:
19
  <!-- ### vocab_type: -->
20
  weighted/imatrix quants of https://huggingface.co/ChuckMcSneed/Gembo-v1-70b
21
 
22
-
23
  <!-- provided-files -->
24
  static quants are available at https://huggingface.co/mradermacher/Gembo-v1-70b-GGUF
25
  ## Usage
@@ -34,8 +33,24 @@ more details, including on how to concatenate multi-part files.
34
 
35
  | Link | Type | Size/GB | Notes |
36
  |:-----|:-----|--------:|:------|
 
 
 
37
  | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q2_K.gguf) | i1-Q2_K | 25.6 | IQ3_XXS probably better |
 
 
 
 
 
 
 
 
 
38
  | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q4_K_S.gguf) | i1-Q4_K_S | 39.3 | optimal size/speed/quality |
 
 
 
 
39
 
40
 
41
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
19
  <!-- ### vocab_type: -->
20
  weighted/imatrix quants of https://huggingface.co/ChuckMcSneed/Gembo-v1-70b
21
 
 
22
  <!-- provided-files -->
23
  static quants are available at https://huggingface.co/mradermacher/Gembo-v1-70b-GGUF
24
  ## Usage
 
33
 
34
  | Link | Type | Size/GB | Notes |
35
  |:-----|:-----|--------:|:------|
36
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 18.4 | |
37
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ2_XS.gguf) | i1-IQ2_XS | 20.4 | |
38
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ2_M.gguf) | i1-IQ2_M | 23.3 | |
39
  | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q2_K.gguf) | i1-Q2_K | 25.6 | IQ3_XXS probably better |
40
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 26.7 | lower quality |
41
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ3_XS.gguf) | i1-IQ3_XS | 28.4 | |
42
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ3_S.gguf) | i1-IQ3_S | 30.0 | beats Q3_K* |
43
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q3_K_S.gguf) | i1-Q3_K_S | 30.0 | IQ3_XS probably better |
44
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ3_M.gguf) | i1-IQ3_M | 31.0 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q3_K_M.gguf) | i1-Q3_K_M | 33.4 | IQ3_S probably better |
46
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q3_K_L.gguf) | i1-Q3_K_L | 36.2 | IQ3_M probably better |
47
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-IQ4_XS.gguf) | i1-IQ4_XS | 36.9 | |
48
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q4_0.gguf) | i1-Q4_0 | 39.1 | fast, low quality |
49
  | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q4_K_S.gguf) | i1-Q4_K_S | 39.3 | optimal size/speed/quality |
50
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q4_K_M.gguf) | i1-Q4_K_M | 41.5 | fast, recommended |
51
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q5_K_S.gguf) | i1-Q5_K_S | 47.6 | |
52
+ | [GGUF](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q5_K_M.gguf) | i1-Q5_K_M | 48.9 | |
53
+ | [PART 1](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Gembo-v1-70b-i1-GGUF/resolve/main/Gembo-v1-70b.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 56.7 | practically like static Q6_K |
54
 
55
 
56
  Here is a handy graph by ikawrakow comparing some lower-quality quant