mradermacher commited on
Commit
c627f61
·
verified ·
1 Parent(s): 0ece267

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -0
README.md CHANGED
@@ -44,6 +44,18 @@ more details, including on how to concatenate multi-part files.
44
  |:-----|:-----|--------:|:------|
45
  | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.mmproj-Q8_0.gguf) | mmproj-Q8_0 | 0.9 | multi-modal supplement |
46
  | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.mmproj-f16.gguf) | mmproj-f16 | 1.4 | multi-modal supplement |
 
 
 
 
 
 
 
 
 
 
 
 
47
 
48
  Here is a handy graph by ikawrakow comparing some lower-quality quant
49
  types (lower is better):
 
44
  |:-----|:-----|--------:|:------|
45
  | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.mmproj-Q8_0.gguf) | mmproj-Q8_0 | 0.9 | multi-modal supplement |
46
  | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.mmproj-f16.gguf) | mmproj-f16 | 1.4 | multi-modal supplement |
47
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q2_K.gguf) | Q2_K | 1.5 | |
48
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q3_K_S.gguf) | Q3_K_S | 1.7 | |
49
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q3_K_M.gguf) | Q3_K_M | 1.8 | lower quality |
50
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q3_K_L.gguf) | Q3_K_L | 1.9 | |
51
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.IQ4_XS.gguf) | IQ4_XS | 2.0 | |
52
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q4_K_S.gguf) | Q4_K_S | 2.1 | fast, recommended |
53
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q4_K_M.gguf) | Q4_K_M | 2.2 | fast, recommended |
54
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q5_K_S.gguf) | Q5_K_S | 2.5 | |
55
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q5_K_M.gguf) | Q5_K_M | 2.5 | |
56
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q6_K.gguf) | Q6_K | 2.9 | very good quality |
57
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.Q8_0.gguf) | Q8_0 | 3.7 | fast, best quality |
58
+ | [GGUF](https://huggingface.co/mradermacher/InfiGUI-G1-3B-GGUF/resolve/main/InfiGUI-G1-3B.f16.gguf) | f16 | 6.9 | 16 bpw, overkill |
59
 
60
  Here is a handy graph by ikawrakow comparing some lower-quality quant
61
  types (lower is better):