xls-r-1b-bigcgen-combined-20hrs-model
This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the BIGCGEN - NA dataset. It achieves the following results on the evaluation set:
- Loss: 0.6477
- Wer: 0.7021
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: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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: 500
- num_epochs: 30.0
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
No log | 0.1054 | 100 | 3.7150 | 1.0 |
No log | 0.2107 | 200 | 2.7910 | 1.0 |
No log | 0.3161 | 300 | 1.6095 | 1.0 |
No log | 0.4215 | 400 | 0.9536 | 0.9673 |
5.9793 | 0.5269 | 500 | 0.8478 | 0.9195 |
5.9793 | 0.6322 | 600 | 0.7154 | 0.7658 |
5.9793 | 0.7376 | 700 | 0.7038 | 0.7147 |
5.9793 | 0.8430 | 800 | 0.6687 | 0.6510 |
5.9793 | 0.9484 | 900 | 0.6481 | 0.6681 |
1.3585 | 1.0537 | 1000 | 0.7987 | 0.8549 |
1.3585 | 1.1591 | 1100 | 0.6608 | 0.6476 |
1.3585 | 1.2645 | 1200 | 0.6284 | 0.6383 |
1.3585 | 1.3699 | 1300 | 0.6126 | 0.6017 |
1.3585 | 1.4752 | 1400 | 0.6486 | 0.6449 |
1.1993 | 1.5806 | 1500 | 0.6060 | 0.6444 |
1.1993 | 1.6860 | 1600 | 0.5986 | 0.5791 |
1.1993 | 1.7914 | 1700 | 0.6037 | 0.6004 |
1.1993 | 1.8967 | 1800 | 0.7841 | 0.8145 |
1.1993 | 2.0021 | 1900 | 0.5611 | 0.5845 |
1.074 | 2.1075 | 2000 | 0.6489 | 0.7152 |
1.074 | 2.2129 | 2100 | 0.5867 | 0.6060 |
1.074 | 2.3182 | 2200 | 0.6211 | 0.6363 |
1.074 | 2.4236 | 2300 | 0.6476 | 0.7017 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0
- Downloads last month
- 10
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Model tree for csikasote/xls-r-1b-bigcgen-combined-20hrs-model
Base model
facebook/wav2vec2-xls-r-1b