hBERTv1_new_pretrain_w_init_48_ver2_mnli
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48 on the GLUE MNLI dataset. It achieves the following results on the evaluation set:
- Loss: 1.0270
- Accuracy: 0.4595
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: 4e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 10
- 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 |
|---|---|---|---|---|
| 1.0624 | 1.0 | 6136 | 1.0715 | 0.3840 |
| 1.0497 | 2.0 | 12272 | 1.0548 | 0.4072 |
| 1.0421 | 3.0 | 18408 | 1.0476 | 0.4432 |
| 1.0485 | 4.0 | 24544 | 1.0547 | 0.4414 |
| 1.0473 | 5.0 | 30680 | 1.0516 | 0.4553 |
| 1.0498 | 6.0 | 36816 | 1.0556 | 0.4427 |
| 1.0531 | 7.0 | 42952 | 1.0556 | 0.4381 |
| 1.0609 | 8.0 | 49088 | 1.0687 | 0.4028 |
Framework versions
- Transformers 4.34.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.14.1
- Downloads last month
- 11
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.