YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

OmniASR Fine-tuned Model

Training Summary

Best checkpoint: step_5000

Best validation CTC Loss: 131.9508

Best validation WER: 37.78

Best validation UER: 12.11

Validation Metrics

Step CTC Loss WER UER
5000 131.9508 37.78 12.11
7500 135.5030 38.83 13.06
10000 139.1468 39.06 13.09

Training Metrics

Step CTC Loss UER WER Learning Rate
500 1449.5287 0.00000504
1000 234.3588 0.00000999
1500 95.6657 0.00001000
2000 77.7399 0.00001000
2500 70.4134 0.00001000
3000 64.9089 0.00001000
3500 60.0571 0.00001000
4000 56.9815 0.00001000
4500 54.4134 0.00001000
5000 50.6973 0.00001000
5500 48.7766 0.00000742
6000 44.3841 0.00000550
6500 43.2830 0.00000407
7000 41.3124 0.00000302
7500 38.5571 0.00000224
8000 38.7908 0.00000166
8500 38.4408 0.00000123
9000 38.1090 0.00000091
9500 37.1977 0.00000068
10000 37.4088 0.00000050
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support