mradermacher commited on
Commit
3c25cc3
·
verified ·
1 Parent(s): f71fa12

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -0
README.md CHANGED
@@ -41,20 +41,28 @@ more details, including on how to concatenate multi-part files.
41
  | Link | Type | Size/GB | Notes |
42
  |:-----|:-----|--------:|:------|
43
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.imatrix.gguf) | imatrix | 0.1 | imatrix file (for creating your own qwuants) |
 
44
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ1_M.gguf) | i1-IQ1_M | 5.1 | mostly desperate |
45
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 5.9 | |
46
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_XS.gguf) | i1-IQ2_XS | 6.6 | |
 
47
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_M.gguf) | i1-IQ2_M | 7.3 | |
48
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q2_K_S.gguf) | i1-Q2_K_S | 7.5 | very low quality |
49
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q2_K.gguf) | i1-Q2_K | 8.2 | IQ3_XXS probably better |
50
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 8.6 | lower quality |
 
51
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q3_K_S.gguf) | i1-Q3_K_S | 9.6 | IQ3_XS probably better |
 
52
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ3_M.gguf) | i1-IQ3_M | 9.7 | |
53
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q3_K_M.gguf) | i1-Q3_K_M | 10.6 | IQ3_S probably better |
54
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q3_K_L.gguf) | i1-Q3_K_L | 11.5 | IQ3_M probably better |
55
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ4_XS.gguf) | i1-IQ4_XS | 11.8 | |
 
56
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q4_K_S.gguf) | i1-Q4_K_S | 12.5 | optimal size/speed/quality |
57
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q4_K_M.gguf) | i1-Q4_K_M | 13.3 | fast, recommended |
 
 
 
58
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q6_K.gguf) | i1-Q6_K | 18.0 | practically like static Q6_K |
59
 
60
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
41
  | Link | Type | Size/GB | Notes |
42
  |:-----|:-----|--------:|:------|
43
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.imatrix.gguf) | imatrix | 0.1 | imatrix file (for creating your own qwuants) |
44
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ1_S.gguf) | i1-IQ1_S | 4.6 | for the desperate |
45
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ1_M.gguf) | i1-IQ1_M | 5.1 | mostly desperate |
46
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 5.9 | |
47
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_XS.gguf) | i1-IQ2_XS | 6.6 | |
48
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_S.gguf) | i1-IQ2_S | 6.6 | |
49
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ2_M.gguf) | i1-IQ2_M | 7.3 | |
50
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q2_K_S.gguf) | i1-Q2_K_S | 7.5 | very low quality |
51
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q2_K.gguf) | i1-Q2_K | 8.2 | IQ3_XXS probably better |
52
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 8.6 | lower quality |
53
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ3_XS.gguf) | i1-IQ3_XS | 9.1 | |
54
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q3_K_S.gguf) | i1-Q3_K_S | 9.6 | IQ3_XS probably better |
55
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ3_S.gguf) | i1-IQ3_S | 9.6 | beats Q3_K* |
56
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ3_M.gguf) | i1-IQ3_M | 9.7 | |
57
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q3_K_M.gguf) | i1-Q3_K_M | 10.6 | IQ3_S probably better |
58
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q3_K_L.gguf) | i1-Q3_K_L | 11.5 | IQ3_M probably better |
59
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-IQ4_XS.gguf) | i1-IQ4_XS | 11.8 | |
60
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q4_0.gguf) | i1-Q4_0 | 12.5 | fast, low quality |
61
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q4_K_S.gguf) | i1-Q4_K_S | 12.5 | optimal size/speed/quality |
62
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q4_K_M.gguf) | i1-Q4_K_M | 13.3 | fast, recommended |
63
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q4_1.gguf) | i1-Q4_1 | 13.8 | |
64
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q5_K_S.gguf) | i1-Q5_K_S | 15.2 | |
65
+ | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q5_K_M.gguf) | i1-Q5_K_M | 15.6 | |
66
  | [GGUF](https://huggingface.co/mradermacher/ERNIE-4.5-21B-A3B-Thinking-i1-GGUF/resolve/main/ERNIE-4.5-21B-A3B-Thinking.i1-Q6_K.gguf) | i1-Q6_K | 18.0 | practically like static Q6_K |
67
 
68
  Here is a handy graph by ikawrakow comparing some lower-quality quant