update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-sa-4.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- te_dx_jp
|
| 7 |
+
model-index:
|
| 8 |
+
- name: t5-base-TEDxJP-2front-1body-2rear
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# t5-base-TEDxJP-2front-1body-2rear
|
| 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: 0.4507
|
| 20 |
+
- Wer: 0.1735
|
| 21 |
+
- Mer: 0.1673
|
| 22 |
+
- Wil: 0.2540
|
| 23 |
+
- Wip: 0.7460
|
| 24 |
+
- Hits: 55769
|
| 25 |
+
- Substitutions: 6394
|
| 26 |
+
- Deletions: 2424
|
| 27 |
+
- Insertions: 2390
|
| 28 |
+
- Cer: 0.1370
|
| 29 |
+
|
| 30 |
+
## Model description
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Intended uses & limitations
|
| 35 |
+
|
| 36 |
+
More information needed
|
| 37 |
+
|
| 38 |
+
## Training and evaluation data
|
| 39 |
+
|
| 40 |
+
More information needed
|
| 41 |
+
|
| 42 |
+
## Training procedure
|
| 43 |
+
|
| 44 |
+
### Training hyperparameters
|
| 45 |
+
|
| 46 |
+
The following hyperparameters were used during training:
|
| 47 |
+
- learning_rate: 0.0001
|
| 48 |
+
- train_batch_size: 32
|
| 49 |
+
- eval_batch_size: 32
|
| 50 |
+
- seed: 42
|
| 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
|
| 54 |
+
- num_epochs: 10
|
| 55 |
+
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Mer | Wil | Wip | Hits | Substitutions | Deletions | Insertions | Cer |
|
| 59 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:------:|:-----:|:-------------:|:---------:|:----------:|:------:|
|
| 60 |
+
| 0.6076 | 1.0 | 1457 | 0.4832 | 0.2119 | 0.1996 | 0.2902 | 0.7098 | 54866 | 6838 | 2883 | 3964 | 0.1884 |
|
| 61 |
+
| 0.4983 | 2.0 | 2914 | 0.4324 | 0.1806 | 0.1743 | 0.2618 | 0.7382 | 55275 | 6452 | 2860 | 2354 | 0.1477 |
|
| 62 |
+
| 0.4734 | 3.0 | 4371 | 0.4273 | 0.1781 | 0.1715 | 0.2592 | 0.7408 | 55577 | 6490 | 2520 | 2492 | 0.1396 |
|
| 63 |
+
| 0.4068 | 4.0 | 5828 | 0.4265 | 0.1756 | 0.1692 | 0.2568 | 0.7432 | 55696 | 6477 | 2414 | 2449 | 0.1380 |
|
| 64 |
+
| 0.378 | 5.0 | 7285 | 0.4281 | 0.1764 | 0.1697 | 0.2567 | 0.7433 | 55736 | 6436 | 2415 | 2539 | 0.1391 |
|
| 65 |
+
| 0.3395 | 6.0 | 8742 | 0.4337 | 0.1737 | 0.1675 | 0.2540 | 0.7460 | 55744 | 6376 | 2467 | 2374 | 0.1375 |
|
| 66 |
+
| 0.3282 | 7.0 | 10199 | 0.4388 | 0.1735 | 0.1675 | 0.2542 | 0.7458 | 55711 | 6396 | 2480 | 2331 | 0.1371 |
|
| 67 |
+
| 0.3371 | 8.0 | 11656 | 0.4443 | 0.1736 | 0.1674 | 0.2542 | 0.7458 | 55753 | 6403 | 2431 | 2377 | 0.1368 |
|
| 68 |
+
| 0.307 | 9.0 | 13113 | 0.4470 | 0.1725 | 0.1665 | 0.2528 | 0.7472 | 55770 | 6360 | 2457 | 2323 | 0.1361 |
|
| 69 |
+
| 0.2842 | 10.0 | 14570 | 0.4507 | 0.1735 | 0.1673 | 0.2540 | 0.7460 | 55769 | 6394 | 2424 | 2390 | 0.1370 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- Transformers 4.21.2
|
| 75 |
+
- Pytorch 1.12.1+cu116
|
| 76 |
+
- Datasets 2.4.0
|
| 77 |
+
- Tokenizers 0.12.1
|