scenario-TCR-XLMV_data-AmazonScience_massive_all_1_1_beta2
This model is a fine-tuned version of facebook/xlm-v-base on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.8678
- Accuracy: 0.8495
- F1: 0.8258
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: 32
- eval_batch_size: 32
- seed: 67
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 500
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
---|---|---|---|---|---|
0.6252 | 0.27 | 5000 | 0.7387 | 0.8183 | 0.7743 |
0.4497 | 0.53 | 10000 | 0.6721 | 0.8363 | 0.7908 |
0.3806 | 0.8 | 15000 | 0.6702 | 0.8451 | 0.8090 |
0.303 | 1.07 | 20000 | 0.7162 | 0.8457 | 0.8130 |
0.2732 | 1.34 | 25000 | 0.7250 | 0.8475 | 0.8178 |
0.2574 | 1.6 | 30000 | 0.7626 | 0.8449 | 0.8188 |
0.2565 | 1.87 | 35000 | 0.7255 | 0.8506 | 0.8251 |
0.2074 | 2.14 | 40000 | 0.7439 | 0.8524 | 0.8268 |
0.2139 | 2.41 | 45000 | 0.8088 | 0.8478 | 0.8233 |
0.2007 | 2.67 | 50000 | 0.7556 | 0.8476 | 0.8223 |
0.2012 | 2.94 | 55000 | 0.7599 | 0.8505 | 0.8250 |
0.1698 | 3.21 | 60000 | 0.8283 | 0.8481 | 0.8255 |
0.1728 | 3.47 | 65000 | 0.7996 | 0.8521 | 0.8320 |
0.1711 | 3.74 | 70000 | 0.7974 | 0.8520 | 0.8292 |
0.1623 | 4.01 | 75000 | 0.8819 | 0.8485 | 0.8223 |
0.1502 | 4.28 | 80000 | 0.8330 | 0.8534 | 0.8320 |
0.1605 | 4.54 | 85000 | 0.8250 | 0.8499 | 0.8264 |
0.1659 | 4.81 | 90000 | 0.8318 | 0.8493 | 0.8237 |
0.1241 | 5.08 | 95000 | 0.9368 | 0.8518 | 0.8191 |
0.1361 | 5.34 | 100000 | 0.9396 | 0.8510 | 0.8237 |
0.1481 | 5.61 | 105000 | 0.8678 | 0.8495 | 0.8258 |
Framework versions
- Transformers 4.33.3
- Pytorch 2.1.1+cu121
- Datasets 2.14.5
- Tokenizers 0.13.3
- Downloads last month
- 7
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.
Model tree for haryoaw/scenario-TCR-XLMV_data-AmazonScience_massive_all_1_1_beta2
Base model
facebook/xlm-v-base
Finetuned
this model
Evaluation results
- Accuracy on massivevalidation set self-reported0.850
- F1 on massivevalidation set self-reported0.826