metadata
library_name: transformers
license: llama3
base_model: meta-llama/Meta-Llama-3-8B-Instruct
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: IE_L3_1000steps_1e8rate_SFT
results: []
IE_L3_1000steps_1e8rate_SFT
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: 2.2129
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-08
- 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 |
|---|---|---|---|
| 2.213 | 0.4 | 50 | 2.2135 |
| 2.2719 | 0.8 | 100 | 2.2132 |
| 2.3303 | 1.2 | 150 | 2.2131 |
| 2.2516 | 1.6 | 200 | 2.2134 |
| 2.2888 | 2.0 | 250 | 2.2132 |
| 2.3253 | 2.4 | 300 | 2.2128 |
| 2.3392 | 2.8 | 350 | 2.2129 |
| 2.2267 | 3.2 | 400 | 2.2129 |
| 2.3535 | 3.6 | 450 | 2.2130 |
| 2.1823 | 4.0 | 500 | 2.2130 |
| 2.3604 | 4.4 | 550 | 2.2130 |
| 2.2783 | 4.8 | 600 | 2.2133 |
| 2.2905 | 5.2 | 650 | 2.2133 |
| 2.2602 | 5.6 | 700 | 2.2131 |
| 2.1896 | 6.0 | 750 | 2.2128 |
| 2.3313 | 6.4 | 800 | 2.2129 |
| 2.2222 | 6.8 | 850 | 2.2129 |
| 2.2568 | 7.2 | 900 | 2.2129 |
| 2.2812 | 7.6 | 950 | 2.2129 |
| 2.2885 | 8.0 | 1000 | 2.2129 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.0.0+cu117
- Datasets 3.0.0
- Tokenizers 0.19.1