---
library_name: transformers
license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: mms-1b-bemgen-combined-model
  results: []
---

<!-- 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. -->

# mms-1b-bemgen-combined-model

This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2478
- Wer: 0.3899

## 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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Use 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: 30.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 6.8762        | 0.0516 | 100  | 0.9801          | 0.9386 |
| 0.5788        | 0.1031 | 200  | 0.3466          | 0.5014 |
| 0.4891        | 0.1547 | 300  | 0.3220          | 0.4820 |
| 0.4386        | 0.2063 | 400  | 0.3071          | 0.4802 |
| 0.4272        | 0.2579 | 500  | 0.3056          | 0.4988 |
| 0.3982        | 0.3094 | 600  | 0.2981          | 0.4626 |
| 0.425         | 0.3610 | 700  | 0.2977          | 0.4631 |
| 0.4036        | 0.4126 | 800  | 0.2897          | 0.4438 |
| 0.3903        | 0.4642 | 900  | 0.2878          | 0.4627 |
| 0.3758        | 0.5157 | 1000 | 0.2926          | 0.4523 |
| 0.3861        | 0.5673 | 1100 | 0.2807          | 0.4410 |
| 0.3763        | 0.6189 | 1200 | 0.2790          | 0.4331 |
| 0.3984        | 0.6704 | 1300 | 0.2803          | 0.4312 |
| 0.373         | 0.7220 | 1400 | 0.2802          | 0.4246 |
| 0.3848        | 0.7736 | 1500 | 0.2759          | 0.4752 |
| 0.4235        | 0.8252 | 1600 | 0.2738          | 0.4268 |
| 0.3704        | 0.8767 | 1700 | 0.2688          | 0.4219 |
| 0.3911        | 0.9283 | 1800 | 0.2653          | 0.4201 |
| 0.3954        | 0.9799 | 1900 | 0.2697          | 0.4482 |
| 0.352         | 1.0315 | 2000 | 0.2654          | 0.4154 |
| 0.3808        | 1.0830 | 2100 | 0.2631          | 0.4051 |
| 0.3681        | 1.1346 | 2200 | 0.2610          | 0.4219 |
| 0.3355        | 1.1862 | 2300 | 0.2608          | 0.4098 |
| 0.342         | 1.2378 | 2400 | 0.2602          | 0.4082 |
| 0.347         | 1.2893 | 2500 | 0.2628          | 0.4055 |
| 0.3409        | 1.3409 | 2600 | 0.2588          | 0.4129 |
| 0.3423        | 1.3925 | 2700 | 0.2617          | 0.4192 |
| 0.3341        | 1.4440 | 2800 | 0.2578          | 0.4055 |
| 0.3425        | 1.4956 | 2900 | 0.2580          | 0.3988 |
| 0.337         | 1.5472 | 3000 | 0.2568          | 0.4071 |
| 0.3412        | 1.5988 | 3100 | 0.2552          | 0.3993 |
| 0.3837        | 1.6503 | 3200 | 0.2622          | 0.4084 |
| 0.3372        | 1.7019 | 3300 | 0.2548          | 0.3991 |
| 0.3394        | 1.7535 | 3400 | 0.2535          | 0.4061 |
| 0.3542        | 1.8051 | 3500 | 0.2512          | 0.3927 |
| 0.3368        | 1.8566 | 3600 | 0.2580          | 0.4004 |
| 0.3807        | 1.9082 | 3700 | 0.2490          | 0.3975 |
| 0.3454        | 1.9598 | 3800 | 0.2514          | 0.4002 |
| 0.3456        | 2.0113 | 3900 | 0.2457          | 0.3931 |
| 0.3202        | 2.0629 | 4000 | 0.2466          | 0.3916 |
| 0.3233        | 2.1145 | 4100 | 0.2495          | 0.3975 |
| 0.3052        | 2.1661 | 4200 | 0.2478          | 0.3899 |


### Framework versions

- Transformers 4.47.1
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0