|
--- |
|
license: llama3 |
|
base_model: meta-llama/Meta-Llama-3-8B-Instruct |
|
tags: |
|
- trl |
|
- sft |
|
- generated_from_trainer |
|
model-index: |
|
- name: MedQA_L3_500steps_1e7rate_SFT |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# MedQA_L3_500steps_1e7rate_SFT |
|
|
|
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.3157 |
|
|
|
## 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: 500 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 1.774 | 0.0489 | 50 | 1.7867 | |
|
| 1.7099 | 0.0977 | 100 | 1.6989 | |
|
| 1.5892 | 0.1466 | 150 | 1.5687 | |
|
| 1.4868 | 0.1954 | 200 | 1.4685 | |
|
| 1.4001 | 0.2443 | 250 | 1.3929 | |
|
| 1.3564 | 0.2931 | 300 | 1.3457 | |
|
| 1.3261 | 0.3420 | 350 | 1.3226 | |
|
| 1.3101 | 0.3908 | 400 | 1.3163 | |
|
| 1.3032 | 0.4397 | 450 | 1.3159 | |
|
| 1.3189 | 0.4885 | 500 | 1.3157 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.41.0 |
|
- Pytorch 2.0.0+cu117 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|