fmehvash commited on
Commit
00a0b49
·
verified ·
1 Parent(s): 6d0382e

End of training

Browse files
Files changed (1) hide show
  1. README.md +15 -15
README.md CHANGED
@@ -18,11 +18,11 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 1.1755
22
- - Rouge1: 35.8333
23
- - Rouge2: 21.1765
24
- - Rougel: 36.2544
25
- - Rougelsum: 36.2544
26
 
27
  ## Model description
28
 
@@ -53,16 +53,16 @@ The following hyperparameters were used during training:
53
 
54
  | Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
55
  |:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|
56
- | 5.8034 | 1.0 | 46 | 2.2364 | 22.9783 | 15.7143 | 23.7391 | 23.7391 |
57
- | 5.3384 | 2.0 | 92 | 1.9892 | 29.4009 | 19.5677 | 29.4009 | 29.4009 |
58
- | 4.345 | 3.0 | 138 | 1.7570 | 39.2143 | 25.7143 | 39.2143 | 39.3421 |
59
- | 3.726 | 4.0 | 184 | 1.5852 | 37.683 | 21.1765 | 37.683 | 37.683 |
60
- | 3.4372 | 5.0 | 230 | 1.4693 | 38.5621 | 21.1765 | 38.5621 | 38.5621 |
61
- | 3.2692 | 6.0 | 276 | 1.3816 | 42.1754 | 28.231 | 42.4082 | 42.4082 |
62
- | 2.9275 | 7.0 | 322 | 1.2673 | 35.8333 | 21.1765 | 36.2544 | 36.2544 |
63
- | 2.8377 | 8.0 | 368 | 1.2202 | 38.5621 | 21.1765 | 38.5621 | 38.5621 |
64
- | 2.9022 | 9.0 | 414 | 1.1861 | 38.5621 | 21.1765 | 38.5621 | 38.5621 |
65
- | 2.8109 | 10.0 | 460 | 1.1755 | 35.8333 | 21.1765 | 36.2544 | 36.2544 |
66
 
67
 
68
  ### Framework versions
 
18
 
19
  This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.0014
22
+ - Rouge1: 99.9541
23
+ - Rouge2: 87.8299
24
+ - Rougel: 99.9541
25
+ - Rougelsum: 99.9541
26
 
27
  ## Model description
28
 
 
53
 
54
  | Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
55
  |:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|
56
+ | 4.0259 | 1.0 | 376 | 2.0056 | 39.5517 | 23.9488 | 39.1195 | 39.0817 |
57
+ | 1.4477 | 2.0 | 752 | 0.6119 | 77.1176 | 63.7739 | 77.6351 | 77.12 |
58
+ | 0.7437 | 3.0 | 1128 | 0.0754 | 95.5267 | 84.31 | 95.5087 | 95.4086 |
59
+ | 0.3729 | 4.0 | 1504 | 0.0207 | 99.2424 | 86.3148 | 99.1965 | 99.1965 |
60
+ | 0.3327 | 5.0 | 1880 | 0.0119 | 99.2424 | 86.3148 | 99.1965 | 99.1965 |
61
+ | 0.2315 | 6.0 | 2256 | 0.0060 | 99.2424 | 86.3148 | 99.1965 | 99.1965 |
62
+ | 0.2843 | 7.0 | 2632 | 0.0029 | 99.9541 | 87.8299 | 99.9541 | 99.9541 |
63
+ | 0.1362 | 8.0 | 3008 | 0.0024 | 99.9541 | 87.8299 | 99.9541 | 99.9541 |
64
+ | 0.1723 | 9.0 | 3384 | 0.0016 | 99.9541 | 87.8299 | 99.9541 | 99.9541 |
65
+ | 0.1149 | 10.0 | 3760 | 0.0014 | 99.9541 | 87.8299 | 99.9541 | 99.9541 |
66
 
67
 
68
  ### Framework versions