<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
nmt-ted-id-en-lr_1e-3-ep_30-seq_128-bs_64
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.6505
- Bleu: 18.2883
- Gen Len: 16.2262
- Meteor: 0.3649
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len | Meteor |
|---|---|---|---|---|---|---|
| 2.4121 | 1.0 | 625 | 1.7491 | 12.8148 | 16.6354 | 0.293 |
| 1.9046 | 2.0 | 1250 | 1.5805 | 15.2033 | 16.2966 | 0.3249 |
| 1.7145 | 3.0 | 1875 | 1.4995 | 16.1248 | 16.363 | 0.3378 |
| 1.5134 | 4.0 | 2500 | 1.4677 | 16.5428 | 16.3165 | 0.3442 |
| 1.4192 | 5.0 | 3125 | 1.4495 | 17.0497 | 16.2708 | 0.3508 |
| 1.3584 | 6.0 | 3750 | 1.4328 | 17.4935 | 16.3636 | 0.3515 |
| 1.3161 | 7.0 | 4375 | 1.4375 | 17.6288 | 16.2925 | 0.3553 |
| 1.2355 | 8.0 | 5000 | 1.4354 | 17.8579 | 16.3658 | 0.3581 |
| 1.1749 | 9.0 | 5625 | 1.4395 | 18.063 | 16.3039 | 0.3595 |
| 1.1464 | 10.0 | 6250 | 1.4489 | 18.2212 | 16.2542 | 0.3614 |
| 1.1117 | 11.0 | 6875 | 1.4535 | 18.0255 | 16.2545 | 0.361 |
| 1.0608 | 12.0 | 7500 | 1.4656 | 18.3451 | 16.2774 | 0.3624 |
| 1.0179 | 13.0 | 8125 | 1.4685 | 18.3879 | 16.2881 | 0.3647 |
| 0.9986 | 14.0 | 8750 | 1.4824 | 18.3278 | 16.2391 | 0.3643 |
| 0.9716 | 15.0 | 9375 | 1.4983 | 18.4478 | 16.2554 | 0.3649 |
| 0.9355 | 16.0 | 10000 | 1.5088 | 18.5089 | 16.2369 | 0.3655 |
| 0.9027 | 17.0 | 10625 | 1.5139 | 18.2406 | 16.2369 | 0.3644 |
| 0.8845 | 18.0 | 11250 | 1.5294 | 18.2969 | 16.2133 | 0.3634 |
| 0.8659 | 19.0 | 11875 | 1.5457 | 18.4358 | 16.2142 | 0.365 |
| 0.8378 | 20.0 | 12500 | 1.5733 | 18.3323 | 16.2447 | 0.3645 |
| 0.8138 | 21.0 | 13125 | 1.5718 | 18.1424 | 16.2048 | 0.3634 |
| 0.798 | 22.0 | 13750 | 1.5802 | 18.3342 | 16.2394 | 0.3633 |
| 0.7878 | 23.0 | 14375 | 1.5854 | 18.3571 | 16.2472 | 0.3651 |
| 0.7656 | 24.0 | 15000 | 1.6036 | 18.1945 | 16.2403 | 0.3644 |
| 0.7445 | 25.0 | 15625 | 1.6124 | 18.2537 | 16.2328 | 0.3645 |
| 0.7406 | 26.0 | 16250 | 1.6291 | 18.3761 | 16.1935 | 0.3658 |
| 0.7284 | 27.0 | 16875 | 1.6328 | 18.2859 | 16.2504 | 0.3644 |
| 0.7149 | 28.0 | 17500 | 1.6389 | 18.2636 | 16.2413 | 0.3645 |
| 0.7059 | 29.0 | 18125 | 1.6463 | 18.304 | 16.2271 | 0.3652 |
| 0.7038 | 30.0 | 18750 | 1.6505 | 18.2883 | 16.2262 | 0.3649 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.2
