metadata
license: llama3
base_model: meta-llama/Meta-Llama-3-8B
tags:
- trl
- reward-trainer
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: rm_llama3_8B_helpsteer2
results: []
rm_llama3_8B_helpsteer2
This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.1203
- Accuracy: 0.6339
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.8471 | 0.1572 | 50 | 0.7326 | 0.5819 |
| 0.7455 | 0.3145 | 100 | 0.6821 | 0.5549 |
| 0.7059 | 0.4717 | 150 | 0.6642 | 0.6050 |
| 0.6926 | 0.6289 | 200 | 0.6707 | 0.5915 |
| 0.6683 | 0.7862 | 250 | 0.6506 | 0.6320 |
| 0.6727 | 0.9434 | 300 | 0.6456 | 0.6224 |
| 0.629 | 1.1006 | 350 | 0.6218 | 0.6551 |
| 0.5446 | 1.2579 | 400 | 0.6604 | 0.6281 |
| 0.5377 | 1.4151 | 450 | 0.6345 | 0.6455 |
| 0.5555 | 1.5723 | 500 | 0.6145 | 0.6320 |
| 0.5645 | 1.7296 | 550 | 0.6178 | 0.6474 |
| 0.5392 | 1.8868 | 600 | 0.6323 | 0.6532 |
| 0.4505 | 2.0440 | 650 | 0.7539 | 0.6455 |
| 0.1406 | 2.2013 | 700 | 1.0884 | 0.6339 |
| 0.1487 | 2.3585 | 750 | 1.1136 | 0.6339 |
| 0.1493 | 2.5157 | 800 | 1.1202 | 0.6358 |
| 0.1259 | 2.6730 | 850 | 1.1253 | 0.6320 |
| 0.1382 | 2.8302 | 900 | 1.1189 | 0.6320 |
| 0.1448 | 2.9874 | 950 | 1.1203 | 0.6339 |
Framework versions
- Transformers 4.43.4
- Pytorch 2.2.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1