cgus commited on
Commit
93b386b
·
1 Parent(s): 72f3170

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -10
README.md CHANGED
@@ -13,18 +13,16 @@ Original model: [MiniChat-3B](https://huggingface.co/GeneZC/MiniChat-3B)
13
  Model creator: [GeneZC](https://huggingface.co/GeneZC)
14
 
15
  [4bpw h6 (main)](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/main)
16
- 4.65bpw h6
17
- 5bpw h6
18
- 5.5bpw h6
19
- 6bpw h6
20
- 4bpw h8
21
- 4.65 h8
22
- 5bpw h8
23
  5.5bpw h8
24
- 6bpw h8
25
- 8bpw h8
26
-
27
 
 
 
 
28
  # Original model card:
29
 
30
  ## MiniChat-3B
 
13
  Model creator: [GeneZC](https://huggingface.co/GeneZC)
14
 
15
  [4bpw h6 (main)](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/main)
16
+ [4bpw h8](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/4bpw-h8)
17
+ [4.65 h8](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/4.65bpw-h8)
18
+ [5bpw h8](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/5bpw-h8)
 
 
 
 
19
  5.5bpw h8
20
+ [6bpw h8](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/6bpw-h8)
21
+ [8bpw h8](https://huggingface.co/cgus/MiniChat-3B-exl2/tree/8bpw-h8)
 
22
 
23
+ I originally planned to make both h6 and h8 versions for each quant but there was consistent 30MB difference between h6 and h8.
24
+ So I don't see much of a reason to upload the rest of h6.
25
+
26
  # Original model card:
27
 
28
  ## MiniChat-3B