metadata
license: llama3
base_model: meta-llama/Meta-Llama-3-8B-Instruct
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: Summary_L3_1000steps_1e7rate_SFT2
results: []
Summary_L3_1000steps_1e7rate_SFT2
This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.5908
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-07
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.1137 | 0.2 | 50 | 2.1001 |
| 2.0888 | 0.4 | 100 | 2.0502 |
| 1.9941 | 0.6 | 150 | 1.9720 |
| 1.9206 | 0.8 | 200 | 1.9029 |
| 1.8477 | 1.0 | 250 | 1.8416 |
| 1.7846 | 1.2 | 300 | 1.7881 |
| 1.7997 | 1.4 | 350 | 1.7414 |
| 1.6961 | 1.6 | 400 | 1.7028 |
| 1.6667 | 1.8 | 450 | 1.6706 |
| 1.6768 | 2.0 | 500 | 1.6449 |
| 1.6485 | 2.2 | 550 | 1.6250 |
| 1.6208 | 2.4 | 600 | 1.6107 |
| 1.6199 | 2.6 | 650 | 1.6006 |
| 1.6081 | 2.8 | 700 | 1.5947 |
| 1.5993 | 3.0 | 750 | 1.5916 |
| 1.5986 | 3.2 | 800 | 1.5910 |
| 1.5963 | 3.4 | 850 | 1.5907 |
| 1.6348 | 3.6 | 900 | 1.5907 |
| 1.6064 | 3.8 | 950 | 1.5908 |
| 1.5811 | 4.0 | 1000 | 1.5908 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.0.0+cu117
- Datasets 2.19.2
- Tokenizers 0.19.1