metadata
library_name: transformers
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: ssc-hch-mms-model-mix-adapt-max3-devtrain
results: []
ssc-hch-mms-model-mix-adapt-max3-devtrain
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2614
- Cer: 0.1088
- Wer: 0.6374
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: 0.0005
- train_batch_size: 1
- eval_batch_size: 6
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 2
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer | Wer |
|---|---|---|---|---|---|
| 0.3141 | 0.4739 | 200 | 0.2800 | 0.1156 | 0.6619 |
| 0.3367 | 0.9479 | 400 | 0.2720 | 0.1125 | 0.6523 |
| 0.2839 | 1.4218 | 600 | 0.2836 | 0.1166 | 0.6643 |
| 0.3091 | 1.8957 | 800 | 0.2665 | 0.1116 | 0.6487 |
| 0.2442 | 2.3697 | 1000 | 0.2818 | 0.1131 | 0.6497 |
| 0.249 | 2.8436 | 1200 | 0.2785 | 0.1124 | 0.6550 |
| 0.2169 | 3.3175 | 1400 | 0.2674 | 0.1117 | 0.6481 |
| 0.2255 | 3.7915 | 1600 | 0.2593 | 0.1092 | 0.6400 |
| 0.2075 | 4.2654 | 1800 | 0.2625 | 0.1087 | 0.6307 |
| 0.2047 | 4.7393 | 2000 | 0.2614 | 0.1088 | 0.6374 |
Framework versions
- Transformers 4.52.1
- Pytorch 2.9.1+cu128
- Datasets 3.6.0
- Tokenizers 0.21.4