File size: 2,118 Bytes
659176d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 |
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: xls-r-1b-bemba-15hrs
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. -->
# xls-r-1b-bemba-15hrs
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2134
- Wer: 0.3485
## 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-05
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 400
- num_epochs: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 2.3016 | 0.36 | 400 | 0.6032 | 0.9932 |
| 0.5196 | 0.71 | 800 | 0.3089 | 0.5020 |
| 0.4397 | 1.07 | 1200 | 0.2562 | 0.4223 |
| 0.3617 | 1.43 | 1600 | 0.2269 | 0.4009 |
| 0.36 | 1.79 | 2000 | 0.2106 | 0.3896 |
| 0.3404 | 2.14 | 2400 | 0.2079 | 0.3681 |
| 0.2915 | 2.5 | 2800 | 0.2024 | 0.3488 |
| 0.2869 | 2.86 | 3200 | 0.2068 | 0.3550 |
| 0.2492 | 3.22 | 3600 | 0.1925 | 0.3273 |
| 0.2542 | 3.57 | 4000 | 0.2041 | 0.3446 |
| 0.2333 | 3.93 | 4400 | 0.1985 | 0.3386 |
| 0.2023 | 4.29 | 4800 | 0.2134 | 0.3485 |
### Framework versions
- Transformers 4.19.0.dev0
- Pytorch 1.10.0+cu111
- Datasets 2.1.0
- Tokenizers 0.12.1
|