Hugging Face's logo

hfl
/
chinese-alpaca-2-1.3b-rlhf-gguf

GGUF
Chinese
English
Inference Endpoints
Model card Files Community
hfl-rc commited on
Commit
23bddf0
·
verified ·
1 Parent(s): 59957e7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -0
README.md CHANGED
@@ -20,9 +20,12 @@ Metric: PPL, lower is better
20
  |-----|------|------|
21
  | Q2_K | 20.1066 +/- 0.29236 | 18.8209 +/- 0.27561 |
22
  | Q3_K | 16.9214 +/- 0.26133 | 16.5729 +/- 0.25706 |
 
23
  | Q4_K | 16.1579 +/- 0.25064 | 15.7746 +/- 0.24476 |
 
24
  | Q5_K | 15.3198 +/- 0.23627 | 15.4791 +/- 0.23959 |
25
  | Q6_K | 15.3718 +/- 0.23764 | 15.2572 +/- 0.23549 |
 
26
 
27
  *The model with `-im` suffix is generated with important matrix, which has generally better performance (not always though).*
28
 
 
20
  |-----|------|------|
21
  | Q2_K | 20.1066 +/- 0.29236 | 18.8209 +/- 0.27561 |
22
  | Q3_K | 16.9214 +/- 0.26133 | 16.5729 +/- 0.25706 |
23
+ | Q4_0 | 15.8056 +/- 0.23749 | - |
24
  | Q4_K | 16.1579 +/- 0.25064 | 15.7746 +/- 0.24476 |
25
+ | Q5_0 | 15.4528 +/- 0.23911 | - |
26
  | Q5_K | 15.3198 +/- 0.23627 | 15.4791 +/- 0.23959 |
27
  | Q6_K | 15.3718 +/- 0.23764 | 15.2572 +/- 0.23549 |
28
+ | Q8_0 | 15.3302 +/- 0.23727 | - |
29
 
30
  *The model with `-im` suffix is generated with important matrix, which has generally better performance (not always though).*
31