<!-- 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. -->
comp_4070
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 4.0419
- Cer: 0.9957
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: 3e-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 981
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 46.3138 | 0.31 | 100 | 52.4882 | 1.0565 |
| 46.8456 | 0.61 | 200 | 52.3881 | 1.0195 |
| 45.1111 | 0.92 | 300 | 52.2131 | 1.0098 |
| 42.1887 | 1.22 | 400 | 51.9607 | 1.0074 |
| 39.8973 | 1.53 | 500 | 51.6479 | 1.0044 |
| 40.3098 | 1.83 | 600 | 51.2534 | 0.9954 |
| 43.0428 | 2.14 | 700 | 50.7402 | 1.0144 |
| 41.0924 | 2.45 | 800 | 49.0088 | 0.9957 |
| 23.9534 | 2.75 | 900 | 22.1701 | 0.9957 |
| 13.6958 | 3.06 | 1000 | 13.5678 | 0.9957 |
| 10.6346 | 3.36 | 1100 | 10.6896 | 0.9957 |
| 9.2081 | 3.67 | 1200 | 9.2513 | 0.9957 |
| 8.1931 | 3.98 | 1300 | 8.3136 | 0.9957 |
| 7.411 | 4.28 | 1400 | 7.6162 | 0.9957 |
| 7.0001 | 4.59 | 1500 | 7.0569 | 0.9957 |
| 6.4564 | 4.89 | 1600 | 6.6027 | 0.9957 |
| 5.9763 | 5.2 | 1700 | 6.2126 | 0.9957 |
| 5.7513 | 5.5 | 1800 | 5.8822 | 0.9957 |
| 5.3524 | 5.81 | 1900 | 5.5997 | 0.9957 |
| 5.1859 | 6.12 | 2000 | 5.3638 | 0.9957 |
| 5.0409 | 6.42 | 2100 | 5.1582 | 0.9957 |
| 4.8325 | 6.73 | 2200 | 4.9859 | 0.9957 |
| 4.7159 | 7.03 | 2300 | 4.8386 | 0.9957 |
| 4.5141 | 7.34 | 2400 | 4.7151 | 0.9957 |
| 4.5168 | 7.65 | 2500 | 4.6124 | 0.9957 |
| 4.391 | 7.95 | 2600 | 4.5273 | 0.9957 |
| 4.3214 | 8.26 | 2700 | 4.4586 | 0.9957 |
| 4.2589 | 8.56 | 2800 | 4.4045 | 0.9957 |
| 4.2121 | 8.87 | 2900 | 4.3628 | 0.9957 |
| 4.1945 | 9.17 | 3000 | 4.3291 | 0.9957 |
| 4.1998 | 9.48 | 3100 | 4.3051 | 0.9957 |
| 4.1552 | 9.79 | 3200 | 4.2893 | 0.9957 |
| 4.1489 | 10.09 | 3300 | 4.2744 | 0.9957 |
| 4.1448 | 10.4 | 3400 | 4.2620 | 0.9957 |
| 4.1337 | 10.7 | 3500 | 4.2535 | 0.9957 |
| 4.0886 | 11.01 | 3600 | 4.2446 | 0.9957 |
| 4.1142 | 11.31 | 3700 | 4.2376 | 0.9957 |
| 4.0867 | 11.62 | 3800 | 4.2323 | 0.9957 |
| 4.0829 | 11.93 | 3900 | 4.2275 | 0.9957 |
| 4.1394 | 12.23 | 4000 | 4.2231 | 0.9957 |
| 4.1104 | 12.54 | 4100 | 4.2181 | 0.9957 |
| 4.1217 | 12.84 | 4200 | 4.2162 | 0.9957 |
| 4.0601 | 13.15 | 4300 | 4.2118 | 0.9957 |
| 4.0797 | 13.46 | 4400 | 4.2086 | 0.9957 |
| 4.0236 | 13.76 | 4500 | 4.2027 | 0.9957 |
| 4.1162 | 14.07 | 4600 | 4.1970 | 0.9957 |
| 4.1238 | 14.37 | 4700 | 4.1884 | 0.9957 |
| 4.1031 | 14.68 | 4800 | 4.1753 | 0.9957 |
| 4.1089 | 14.98 | 4900 | 4.1638 | 0.9957 |
| 4.02 | 15.29 | 5000 | 4.1517 | 0.9957 |
| 3.9624 | 15.6 | 5100 | 4.1411 | 0.9957 |
| 3.9731 | 15.9 | 5200 | 4.1312 | 0.9957 |
| 4.0816 | 16.21 | 5300 | 4.1222 | 0.9957 |
| 4.0718 | 16.51 | 5400 | 4.1139 | 0.9957 |
| 4.0561 | 16.82 | 5500 | 4.1086 | 0.9957 |
| 3.9548 | 17.13 | 5600 | 4.1050 | 0.9957 |
| 3.972 | 17.43 | 5700 | 4.0987 | 0.9957 |
| 3.9433 | 17.74 | 5800 | 4.0931 | 0.9957 |
| 4.0043 | 18.04 | 5900 | 4.0871 | 0.9957 |
| 4.0505 | 18.35 | 6000 | 4.0863 | 0.9957 |
| 4.0064 | 18.65 | 6100 | 4.0859 | 0.9957 |
| 3.9995 | 18.96 | 6200 | 4.0803 | 0.9957 |
| 3.9386 | 19.27 | 6300 | 4.0712 | 0.9957 |
| 3.9644 | 19.57 | 6400 | 4.0683 | 0.9957 |
| 3.9286 | 19.88 | 6500 | 4.0667 | 0.9957 |
| 3.9873 | 20.18 | 6600 | 4.0676 | 0.9957 |
| 3.9687 | 20.49 | 6700 | 4.0658 | 0.9957 |
| 3.9949 | 20.8 | 6800 | 4.0623 | 0.9957 |
| 3.9831 | 21.1 | 6900 | 4.0570 | 0.9957 |
| 3.9596 | 21.41 | 7000 | 4.0561 | 0.9957 |
| 3.9585 | 21.71 | 7100 | 4.0533 | 0.9957 |
| 3.97 | 22.02 | 7200 | 4.0537 | 0.9957 |
| 3.9665 | 22.32 | 7300 | 4.0550 | 0.9957 |
| 3.9263 | 22.63 | 7400 | 4.0547 | 0.9957 |
| 3.9257 | 22.94 | 7500 | 4.0538 | 0.9957 |
| 3.9945 | 23.24 | 7600 | 4.0493 | 0.9957 |
| 3.9555 | 23.55 | 7700 | 4.0479 | 0.9957 |
| 3.9997 | 23.85 | 7800 | 4.0472 | 0.9957 |
| 3.8985 | 24.16 | 7900 | 4.0477 | 0.9957 |
| 3.9051 | 24.46 | 8000 | 4.0449 | 0.9957 |
| 3.8753 | 24.77 | 8100 | 4.0449 | 0.9957 |
| 4.0476 | 25.08 | 8200 | 4.0437 | 0.9957 |
| 4.0331 | 25.38 | 8300 | 4.0447 | 0.9957 |
| 4.0025 | 25.69 | 8400 | 4.0442 | 0.9957 |
| 3.9211 | 25.99 | 8500 | 4.0440 | 0.9957 |
| 3.8675 | 26.3 | 8600 | 4.0429 | 0.9957 |
| 3.9078 | 26.61 | 8700 | 4.0447 | 0.9957 |
| 3.8914 | 26.91 | 8800 | 4.0428 | 0.9957 |
| 4.0395 | 27.22 | 8900 | 4.0421 | 0.9957 |
| 4.006 | 27.52 | 9000 | 4.0425 | 0.9957 |
| 4.0034 | 27.83 | 9100 | 4.0417 | 0.9957 |
| 3.89 | 28.13 | 9200 | 4.0419 | 0.9957 |
| 3.8887 | 28.44 | 9300 | 4.0423 | 0.9957 |
| 3.8958 | 28.75 | 9400 | 4.0422 | 0.9957 |
| 3.9894 | 29.05 | 9500 | 4.0419 | 0.9957 |
| 4.0323 | 29.36 | 9600 | 4.0419 | 0.9957 |
| 3.9905 | 29.66 | 9700 | 4.0419 | 0.9957 |
| 3.9555 | 29.97 | 9800 | 4.0419 | 0.9957 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1
- Datasets 2.12.0
- Tokenizers 0.13.2
