hbertv1-Massive-intent_w_in
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_wt_init on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.7790
- Accuracy: 0.8746
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 33
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
2.2877 | 1.0 | 180 | 0.9877 | 0.7329 |
0.8514 | 2.0 | 360 | 0.7403 | 0.7993 |
0.5896 | 3.0 | 540 | 0.6955 | 0.8239 |
0.4058 | 4.0 | 720 | 0.6778 | 0.8313 |
0.3003 | 5.0 | 900 | 0.6345 | 0.8505 |
0.2236 | 6.0 | 1080 | 0.6567 | 0.8583 |
0.1615 | 7.0 | 1260 | 0.7163 | 0.8460 |
0.1159 | 8.0 | 1440 | 0.7450 | 0.8519 |
0.0976 | 9.0 | 1620 | 0.7533 | 0.8490 |
0.061 | 10.0 | 1800 | 0.7502 | 0.8642 |
0.0438 | 11.0 | 1980 | 0.7729 | 0.8618 |
0.0309 | 12.0 | 2160 | 0.7790 | 0.8746 |
0.0191 | 13.0 | 2340 | 0.8302 | 0.8682 |
0.0101 | 14.0 | 2520 | 0.8224 | 0.8721 |
0.0057 | 15.0 | 2700 | 0.8229 | 0.8716 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3
- Downloads last month
- 9