fblgit commited on
Commit
7bab67e
1 Parent(s): c25c343

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -25
README.md CHANGED
@@ -11,37 +11,15 @@ model-index:
11
  results: []
12
  ---
13
 
14
- OMA, OneManArmy presents, `una-neural-chat-v3-3` **PHASE 2**. Powered by UNA (Uniform Neural Alignment), using zephyr trainer, allenai/ultrafeedback cleaned.. and JUST THAT.
 
 
15
  Outperforming its base model, not adding any data.. just UNA Algorythm on Transformers Lib.
16
  UNA Settings:
17
  * MLP : 0.05
18
  * ATT : 0.03
19
  * LNOR : 0.02
20
 
21
- # una-neural-chat-v3-3-phase2
22
-
23
- This model is a fine-tuned version of [Intel/neural-chat-7b-v3-3](https://huggingface.co/Intel/neural-chat-7b-v3-3) on the allenai/ultrafeedback_binarized_cleaned dataset.
24
- It achieves the following results on the evaluation set:
25
- - Loss: 0.4524
26
- - Rewards/chosen: -0.7101
27
- - Rewards/rejected: -2.0953
28
- - Rewards/accuracies: 0.7831
29
- - Rewards/margins: 1.3852
30
- - Logps/rejected: -321.5471
31
- - Logps/chosen: -327.5048
32
- - Logits/rejected: -2.6445
33
- - Logits/chosen: -2.6674
34
-
35
- ### Training results
36
-
37
- | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
38
- |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
39
- | 0.5431 | 0.2 | 380 | 0.4900 | -0.6823 | -1.6613 | 0.7607 | 0.9790 | -317.2069 | -327.2263 | -2.6478 | -2.6651 |
40
- | 0.4369 | 0.4 | 760 | 0.4783 | -0.7562 | -2.1298 | 0.7719 | 1.3737 | -321.8924 | -327.9652 | -2.7370 | -2.7562 |
41
- | 0.4005 | 0.6 | 1140 | 0.4697 | -0.6913 | -2.0134 | 0.7770 | 1.3221 | -320.7278 | -327.3167 | -2.7067 | -2.7224 |
42
- | 0.3759 | 0.8 | 1520 | 0.4568 | -0.7387 | -2.0643 | 0.7882 | 1.3256 | -321.2370 | -327.7909 | -2.6626 | -2.6829 |
43
- | 0.5213 | 1.0 | 1900 | 0.4524 | -0.7101 | -2.0953 | 0.7831 | 1.3852 | -321.5471 | -327.5048 | -2.6445 | -2.6674 |
44
-
45
 
46
  ### Framework versions
47
 
 
11
  results: []
12
  ---
13
 
14
+ # una-neural-chat-v3-3-phase2
15
+
16
+ OMA, OneManArmy proudly presents, `una-neural-chat-v3-3` **PHASE 2**. Powered by UNA (Uniform Neural Alignment), using zephyr trainer, allenai/ultrafeedback cleaned.. and JUST THAT.
17
  Outperforming its base model, not adding any data.. just UNA Algorythm on Transformers Lib.
18
  UNA Settings:
19
  * MLP : 0.05
20
  * ATT : 0.03
21
  * LNOR : 0.02
22
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
23
 
24
  ### Framework versions
25