Hugging Face's logo

Padomin
/
t5-base-TEDxJP-9front-1body-0rear

Text2Text Generation
Transformers
PyTorch
TensorBoard
t5
Generated from Trainer
text-generation-inference
Model card Files Metrics Community
Padomin commited on
Commit
74b0bcd
·
1 Parent(s): 452a530

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +19 -19
README.md CHANGED
@@ -16,16 +16,16 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model is a fine-tuned version of [sonoisa/t5-base-japanese](https://huggingface.co/sonoisa/t5-base-japanese) on the te_dx_jp dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 0.4586
20
  - Wer: 0.1728
21
- - Mer: 0.1668
22
- - Wil: 0.2532
23
- - Wip: 0.7468
24
- - Hits: 55746
25
- - Substitutions: 6364
26
- - Deletions: 2477
27
- - Insertions: 2321
28
- - Cer: 0.1365
29
 
30
  ## Model description
31
 
@@ -57,16 +57,16 @@ The following hyperparameters were used during training:
57
 
58
  | Training Loss | Epoch | Step | Validation Loss | Wer | Mer | Wil | Wip | Hits | Substitutions | Deletions | Insertions | Cer |
59
  |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:------:|:-----:|:-------------:|:---------:|:----------:|:------:|
60
- | 0.6036 | 1.0 | 1457 | 0.4842 | 0.2211 | 0.2070 | 0.2960 | 0.7040 | 54723 | 6721 | 3143 | 4418 | 0.1900 |
61
- | 0.5567 | 2.0 | 2914 | 0.4414 | 0.1873 | 0.1792 | 0.2663 | 0.7337 | 55441 | 6469 | 2677 | 2954 | 0.1532 |
62
- | 0.4666 | 3.0 | 4371 | 0.4339 | 0.1793 | 0.1724 | 0.2600 | 0.7400 | 55595 | 6486 | 2506 | 2587 | 0.1432 |
63
- | 0.3673 | 4.0 | 5828 | 0.4316 | 0.1735 | 0.1678 | 0.2540 | 0.7460 | 55576 | 6330 | 2681 | 2196 | 0.1376 |
64
- | 0.3758 | 5.0 | 7285 | 0.4393 | 0.1719 | 0.1663 | 0.2531 | 0.7469 | 55662 | 6383 | 2542 | 2178 | 0.1341 |
65
- | 0.3668 | 6.0 | 8742 | 0.4357 | 0.1708 | 0.1653 | 0.2514 | 0.7486 | 55701 | 6329 | 2557 | 2144 | 0.1346 |
66
- | 0.3512 | 7.0 | 10199 | 0.4430 | 0.1723 | 0.1664 | 0.2527 | 0.7473 | 55732 | 6349 | 2506 | 2272 | 0.1360 |
67
- | 0.3173 | 8.0 | 11656 | 0.4505 | 0.1719 | 0.1660 | 0.2520 | 0.7480 | 55767 | 6322 | 2498 | 2282 | 0.1367 |
68
- | 0.2982 | 9.0 | 13113 | 0.4548 | 0.1723 | 0.1663 | 0.2524 | 0.7476 | 55778 | 6338 | 2471 | 2317 | 0.1361 |
69
- | 0.3362 | 10.0 | 14570 | 0.4586 | 0.1728 | 0.1668 | 0.2532 | 0.7468 | 55746 | 6364 | 2477 | 2321 | 0.1365 |
70
 
71
 
72
  ### Framework versions
 
16
 
17
  This model is a fine-tuned version of [sonoisa/t5-base-japanese](https://huggingface.co/sonoisa/t5-base-japanese) on the te_dx_jp dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.4576
20
  - Wer: 0.1728
21
+ - Mer: 0.1669
22
+ - Wil: 0.2543
23
+ - Wip: 0.7457
24
+ - Hits: 55705
25
+ - Substitutions: 6444
26
+ - Deletions: 2438
27
+ - Insertions: 2281
28
+ - Cer: 0.1351
29
 
30
  ## Model description
31
 
 
57
 
58
  | Training Loss | Epoch | Step | Validation Loss | Wer | Mer | Wil | Wip | Hits | Substitutions | Deletions | Insertions | Cer |
59
  |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:------:|:-----:|:-------------:|:---------:|:----------:|:------:|
60
+ | 0.649 | 1.0 | 1457 | 0.4844 | 0.2290 | 0.2126 | 0.3015 | 0.6985 | 54758 | 6748 | 3081 | 4959 | 0.2080 |
61
+ | 0.5319 | 2.0 | 2914 | 0.4385 | 0.1804 | 0.1741 | 0.2614 | 0.7386 | 55298 | 6437 | 2852 | 2364 | 0.1465 |
62
+ | 0.4819 | 3.0 | 4371 | 0.4338 | 0.1760 | 0.1698 | 0.2569 | 0.7431 | 55558 | 6419 | 2610 | 2336 | 0.1389 |
63
+ | 0.4307 | 4.0 | 5828 | 0.4328 | 0.1759 | 0.1696 | 0.2569 | 0.7431 | 55649 | 6454 | 2484 | 2424 | 0.1390 |
64
+ | 0.3735 | 5.0 | 7285 | 0.4331 | 0.1740 | 0.1680 | 0.2549 | 0.7451 | 55652 | 6398 | 2537 | 2306 | 0.1367 |
65
+ | 0.3495 | 6.0 | 8742 | 0.4380 | 0.1740 | 0.1681 | 0.2552 | 0.7448 | 55619 | 6420 | 2548 | 2267 | 0.1356 |
66
+ | 0.3679 | 7.0 | 10199 | 0.4437 | 0.1741 | 0.1682 | 0.2556 | 0.7444 | 55621 | 6441 | 2525 | 2281 | 0.1354 |
67
+ | 0.3035 | 8.0 | 11656 | 0.4494 | 0.1727 | 0.1669 | 0.2542 | 0.7458 | 55672 | 6433 | 2482 | 2237 | 0.1350 |
68
+ | 0.3041 | 9.0 | 13113 | 0.4541 | 0.1736 | 0.1677 | 0.2550 | 0.7450 | 55674 | 6441 | 2472 | 2302 | 0.1383 |
69
+ | 0.2948 | 10.0 | 14570 | 0.4576 | 0.1728 | 0.1669 | 0.2543 | 0.7457 | 55705 | 6444 | 2438 | 2281 | 0.1351 |
70
 
71
 
72
  ### Framework versions