hbertv2-Massive-intent_48_emb_compress
This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new_emb_compress_48 on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.9335
- Accuracy: 0.8490
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 |
---|---|---|---|---|
2.2651 | 1.0 | 180 | 1.2606 | 0.6606 |
1.0649 | 2.0 | 360 | 1.0114 | 0.7383 |
0.7609 | 3.0 | 540 | 0.8142 | 0.7959 |
0.5705 | 4.0 | 720 | 0.8234 | 0.7919 |
0.4182 | 5.0 | 900 | 0.7970 | 0.8013 |
0.3164 | 6.0 | 1080 | 0.7823 | 0.8141 |
0.233 | 7.0 | 1260 | 0.8068 | 0.8151 |
0.1623 | 8.0 | 1440 | 0.8618 | 0.8234 |
0.1208 | 9.0 | 1620 | 0.8545 | 0.8239 |
0.0823 | 10.0 | 1800 | 0.9072 | 0.8288 |
0.0508 | 11.0 | 1980 | 0.8755 | 0.8431 |
0.0329 | 12.0 | 2160 | 0.9474 | 0.8318 |
0.0181 | 13.0 | 2340 | 0.9236 | 0.8436 |
0.0084 | 14.0 | 2520 | 0.9424 | 0.8470 |
0.0045 | 15.0 | 2700 | 0.9335 | 0.8490 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3
- Downloads last month
- 9