metadata
license: apache-2.0
base_model: distilbert/distilbert-base-uncased
tags:
- trl
- reward-trainer
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: distilbert_social_reasoning_reward_model
results: []
distilbert_social_reasoning_reward_model
This model is a fine-tuned version of distilbert/distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6145
- Accuracy: 0.6871
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: 0.0005
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.6733 | 0.24 | 10 | 0.6564 | 0.6725 |
| 0.6475 | 0.48 | 20 | 0.6191 | 0.6708 |
| 0.655 | 0.72 | 30 | 0.6216 | 0.6708 |
| 0.6489 | 0.96 | 40 | 0.6311 | 0.6708 |
| 0.6204 | 1.2 | 50 | 0.6837 | 0.6147 |
| 0.5924 | 1.44 | 60 | 0.6329 | 0.6988 |
| 0.6124 | 1.68 | 70 | 0.6220 | 0.6620 |
| 0.6123 | 1.92 | 80 | 0.6366 | 0.6515 |
| 0.562 | 2.16 | 90 | 0.6584 | 0.6532 |
| 0.5169 | 2.4 | 100 | 0.6956 | 0.6410 |
| 0.5045 | 2.63 | 110 | 0.6823 | 0.6392 |
| 0.4712 | 2.87 | 120 | 0.6927 | 0.6375 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2