metadata
library_name: transformers
license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
- generated_from_trainer
model-index:
- name: mms-1b-all-gui
results: []
mms-1b-all-gui
This model is a fine-tuned version of facebook/mms-1b-all on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.8980
- Cer: 0.2397
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 7.8584 | 0.4329 | 100 | 1.5099 | 0.3558 |
| 2.5647 | 0.8658 | 200 | 1.1562 | 0.3514 |
| 2.3995 | 1.2987 | 300 | 0.8952 | 0.2601 |
| 2.0466 | 1.7316 | 400 | 0.8955 | 0.2396 |
| 1.8489 | 2.1645 | 500 | 0.8439 | 0.2397 |
| 1.6392 | 2.5974 | 600 | 0.8097 | 0.2385 |
| 1.7354 | 3.0303 | 700 | 0.7468 | 0.2136 |
| 1.3546 | 3.4632 | 800 | 0.7213 | 0.2236 |
| 1.3797 | 3.8961 | 900 | 0.7014 | 0.2238 |
| 1.2238 | 4.3290 | 1000 | 0.6774 | 0.1970 |
| 1.2138 | 4.7619 | 1100 | 0.7162 | 0.1940 |
| 1.2993 | 5.1948 | 1200 | 0.8006 | 0.2100 |
| 1.7973 | 5.6277 | 1300 | 1.0200 | 0.2625 |
| 2.4040 | 6.0606 | 1400 | 1.2120 | 0.2859 |
| 2.1873 | 6.4935 | 1500 | 1.0882 | 0.3209 |
| 1.9755 | 6.9264 | 1600 | 1.1028 | 0.4591 |
| 1.8324 | 7.3593 | 1700 | 0.9170 | 0.3022 |
| 1.6503 | 7.7922 | 1800 | 0.8828 | 0.2572 |
| 1.6647 | 8.2251 | 1900 | 0.8677 | 0.2544 |
| 1.6586 | 8.6580 | 2000 | 0.9189 | 0.2337 |
| 1.6733 | 9.0909 | 2100 | 0.8938 | 0.2419 |
| 1.6615 | 9.5238 | 2200 | 0.8966 | 0.2384 |
| 1.6580 | 9.9567 | 2300 | 0.8980 | 0.2397 |
Framework versions
- Transformers 5.1.0
- Pytorch 2.9.1+cu128
- Datasets 3.6.0
- Tokenizers 0.22.2