Hugging Face's logo

mradermacher
/
Qwen-72B-i1-GGUF

Transformers
GGUF
Chinese
English
qwen
Inference Endpoints
imatrix
Model card Files Community
mradermacher commited on
Commit
de4629e
·
verified ·
1 Parent(s): e9eefd3

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -0
README.md CHANGED
@@ -34,17 +34,26 @@ more details, including on how to concatenate multi-part files.
34
 
35
  | Link | Type | Size/GB | Notes |
36
  |:-----|:-----|--------:|:------|
 
37
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ1_M.gguf) | i1-IQ1_M | 17.1 | mostly desperate |
38
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 19.5 | |
 
 
39
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ2_M.gguf) | i1-IQ2_M | 24.7 | |
40
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q2_K.gguf) | i1-Q2_K | 26.6 | IQ3_XXS probably better |
41
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 28.5 | lower quality |
 
 
42
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q3_K_S.gguf) | i1-Q3_K_S | 31.7 | IQ3_XS probably better |
43
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ3_M.gguf) | i1-IQ3_M | 34.8 | |
44
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q3_K_M.gguf) | i1-Q3_K_M | 36.8 | IQ3_S probably better |
45
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ4_XS.gguf) | i1-IQ4_XS | 38.9 | |
 
 
46
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q4_K_S.gguf) | i1-Q4_K_S | 41.4 | optimal size/speed/quality |
47
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q4_K_M.gguf) | i1-Q4_K_M | 45.2 | fast, recommended |
 
 
48
  | [PART 1](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 59.4 | practically like static Q6_K |
49
 
50
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
34
 
35
  | Link | Type | Size/GB | Notes |
36
  |:-----|:-----|--------:|:------|
37
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ1_S.gguf) | i1-IQ1_S | 15.6 | for the desperate |
38
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ1_M.gguf) | i1-IQ1_M | 17.1 | mostly desperate |
39
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 19.5 | |
40
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ2_XS.gguf) | i1-IQ2_XS | 21.6 | |
41
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ2_S.gguf) | i1-IQ2_S | 22.7 | |
42
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ2_M.gguf) | i1-IQ2_M | 24.7 | |
43
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q2_K.gguf) | i1-Q2_K | 26.6 | IQ3_XXS probably better |
44
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 28.5 | lower quality |
45
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ3_XS.gguf) | i1-IQ3_XS | 30.5 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ3_S.gguf) | i1-IQ3_S | 31.7 | beats Q3_K* |
47
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q3_K_S.gguf) | i1-Q3_K_S | 31.7 | IQ3_XS probably better |
48
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ3_M.gguf) | i1-IQ3_M | 34.8 | |
49
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q3_K_M.gguf) | i1-Q3_K_M | 36.8 | IQ3_S probably better |
50
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-IQ4_XS.gguf) | i1-IQ4_XS | 38.9 | |
51
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q3_K_L.gguf) | i1-Q3_K_L | 39.3 | IQ3_M probably better |
52
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q4_0.gguf) | i1-Q4_0 | 41.2 | fast, low quality |
53
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q4_K_S.gguf) | i1-Q4_K_S | 41.4 | optimal size/speed/quality |
54
  | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q4_K_M.gguf) | i1-Q4_K_M | 45.2 | fast, recommended |
55
+ | [GGUF](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q5_K_S.gguf) | i1-Q5_K_S | 50.0 | |
56
+ | [PART 1](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q5_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q5_K_M.gguf.part2of2) | i1-Q5_K_M | 53.2 | |
57
  | [PART 1](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Qwen-72B-i1-GGUF/resolve/main/Qwen-72B.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 59.4 | practically like static Q6_K |
58
 
59
  Here is a handy graph by ikawrakow comparing some lower-quality quant