Padomin commited on
Commit
24390d0
·
1 Parent(s): 6ec23e2

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +21 -21
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: 0.4377
20
- - Wer: 0.1695
21
- - Mer: 0.1637
22
- - Wil: 0.2493
23
- - Wip: 0.7507
24
- - Hits: 55897
25
- - Substitutions: 6293
26
- - Deletions: 2397
27
- - Insertions: 2255
28
- - Cer: 0.1336
29
 
30
  ## Model description
31
 
@@ -47,7 +47,7 @@ The following hyperparameters were used during training:
47
  - learning_rate: 0.0001
48
  - train_batch_size: 32
49
  - eval_batch_size: 32
50
- - seed: 20
51
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
  - lr_scheduler_type: linear
53
  - lr_scheduler_warmup_ratio: 0.1
@@ -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
- | 0.5953 | 1.0 | 1457 | 0.4689 | 0.2111 | 0.1986 | 0.2885 | 0.7115 | 55024 | 6790 | 2773 | 4070 | 0.1820 |
61
- | 0.5224 | 2.0 | 2914 | 0.4191 | 0.1744 | 0.1688 | 0.2541 | 0.7459 | 55445 | 6244 | 2898 | 2119 | 0.1369 |
62
- | 0.4837 | 3.0 | 4371 | 0.4121 | 0.1734 | 0.1675 | 0.2534 | 0.7466 | 55679 | 6322 | 2586 | 2292 | 0.1381 |
63
- | 0.4133 | 4.0 | 5828 | 0.4080 | 0.1696 | 0.1641 | 0.2495 | 0.7505 | 55810 | 6266 | 2511 | 2179 | 0.1325 |
64
- | 0.3846 | 5.0 | 7285 | 0.4125 | 0.1704 | 0.1646 | 0.2498 | 0.7502 | 55857 | 6257 | 2473 | 2277 | 0.1361 |
65
- | 0.3428 | 6.0 | 8742 | 0.4163 | 0.1693 | 0.1638 | 0.2493 | 0.7507 | 55829 | 6279 | 2479 | 2178 | 0.1329 |
66
- | 0.2926 | 7.0 | 10199 | 0.4240 | 0.1701 | 0.1642 | 0.2496 | 0.7504 | 55905 | 6277 | 2405 | 2302 | 0.1338 |
67
- | 0.2688 | 8.0 | 11656 | 0.4278 | 0.1698 | 0.1640 | 0.2496 | 0.7504 | 55888 | 6289 | 2410 | 2266 | 0.1343 |
68
- | 0.2678 | 9.0 | 13113 | 0.4335 | 0.1687 | 0.1632 | 0.2488 | 0.7512 | 55889 | 6290 | 2408 | 2199 | 0.1330 |
69
- | 0.2476 | 10.0 | 14570 | 0.4377 | 0.1695 | 0.1637 | 0.2493 | 0.7507 | 55897 | 6293 | 2397 | 2255 | 0.1336 |
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: 0.4384
20
+ - Wer: 0.1692
21
+ - Mer: 0.1635
22
+ - Wil: 0.2483
23
+ - Wip: 0.7517
24
+ - Hits: 55908
25
+ - Substitutions: 6222
26
+ - Deletions: 2457
27
+ - Insertions: 2249
28
+ - Cer: 0.1327
29
 
30
  ## Model description
31
 
 
47
  - learning_rate: 0.0001
48
  - train_batch_size: 32
49
  - eval_batch_size: 32
50
+ - seed: 30
51
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
  - lr_scheduler_type: linear
53
  - lr_scheduler_warmup_ratio: 0.1
 
57
 
58
  | Training Loss | Epoch | Step | Validation Loss | Wer | Mer | Wil | Wip | Hits | Substitutions | Deletions | Insertions | Cer |
59
  |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:------:|:-----:|:-------------:|:---------:|:----------:|:------:|
60
+ | 0.5881 | 1.0 | 1457 | 0.4572 | 0.2115 | 0.1995 | 0.2882 | 0.7118 | 54840 | 6661 | 3086 | 3916 | 0.1812 |
61
+ | 0.5018 | 2.0 | 2914 | 0.4167 | 0.1843 | 0.1765 | 0.2640 | 0.7360 | 55546 | 6494 | 2547 | 2863 | 0.1490 |
62
+ | 0.4633 | 3.0 | 4371 | 0.4110 | 0.1738 | 0.1679 | 0.2540 | 0.7460 | 55623 | 6327 | 2637 | 2260 | 0.1369 |
63
+ | 0.3971 | 4.0 | 5828 | 0.4068 | 0.1724 | 0.1666 | 0.2522 | 0.7478 | 55672 | 6278 | 2637 | 2218 | 0.1351 |
64
+ | 0.3907 | 5.0 | 7285 | 0.4131 | 0.1688 | 0.1635 | 0.2479 | 0.7521 | 55789 | 6180 | 2618 | 2106 | 0.1325 |
65
+ | 0.3305 | 6.0 | 8742 | 0.4147 | 0.1706 | 0.1649 | 0.2504 | 0.7496 | 55797 | 6281 | 2509 | 2227 | 0.1336 |
66
+ | 0.2937 | 7.0 | 10199 | 0.4236 | 0.1692 | 0.1636 | 0.2482 | 0.7518 | 55883 | 6207 | 2497 | 2223 | 0.1334 |
67
+ | 0.2649 | 8.0 | 11656 | 0.4307 | 0.1693 | 0.1638 | 0.2493 | 0.7507 | 55806 | 6272 | 2509 | 2154 | 0.1329 |
68
+ | 0.2914 | 9.0 | 13113 | 0.4319 | 0.1691 | 0.1634 | 0.2482 | 0.7518 | 55928 | 6230 | 2429 | 2262 | 0.1328 |
69
+ | 0.2598 | 10.0 | 14570 | 0.4384 | 0.1692 | 0.1635 | 0.2483 | 0.7517 | 55908 | 6222 | 2457 | 2249 | 0.1327 |
70
 
71
 
72
  ### Framework versions