<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
distilbert_add_GLUE_Experiment_logit_kd_pretrain_stsb
This model is a fine-tuned version of gokuls/distilbert_add_pre-training-complete on the GLUE STSB dataset. It achieves the following results on the evaluation set:
- Loss: 1.2199
- Pearson: 0.0583
- Spearmanr: 0.0384
- Combined Score: 0.0484
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: 256
- eval_batch_size: 256
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Pearson | Spearmanr | Combined Score |
---|---|---|---|---|---|---|
1.8784 | 1.0 | 23 | 1.2688 | 0.0461 | 0.0259 | 0.0360 |
1.0833 | 2.0 | 46 | 1.3765 | 0.0190 | 0.0161 | 0.0176 |
1.0408 | 3.0 | 69 | 1.2379 | 0.0102 | 0.0246 | 0.0174 |
1.0289 | 4.0 | 92 | 1.2199 | 0.0583 | 0.0384 | 0.0484 |
1.0248 | 5.0 | 115 | 1.3059 | 0.0405 | 0.0470 | 0.0438 |
0.99 | 6.0 | 138 | 1.2294 | 0.0822 | 0.0782 | 0.0802 |
0.928 | 7.0 | 161 | 1.3883 | 0.1213 | 0.1310 | 0.1261 |
0.9014 | 8.0 | 184 | 1.4037 | 0.0892 | 0.1148 | 0.1020 |
0.8319 | 9.0 | 207 | 1.2495 | 0.1583 | 0.1653 | 0.1618 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2