|
--- |
|
license: llama3 |
|
base_model: meta-llama/Meta-Llama-3-8B-Instruct |
|
tags: |
|
- trl |
|
- sft |
|
- generated_from_trainer |
|
model-index: |
|
- name: Transaminitis_L3_1000rate_1e8_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. --> |
|
|
|
# Transaminitis_L3_1000rate_1e8_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: 2.6870 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 2.684 | 0.2 | 25 | 2.6901 | |
|
| 2.6773 | 0.4 | 50 | 2.6883 | |
|
| 2.6627 | 0.6 | 75 | 2.6887 | |
|
| 2.6575 | 0.8 | 100 | 2.6912 | |
|
| 2.6624 | 1.0 | 125 | 2.6897 | |
|
| 2.6725 | 1.2 | 150 | 2.6884 | |
|
| 2.6661 | 1.4 | 175 | 2.6891 | |
|
| 2.692 | 1.6 | 200 | 2.6879 | |
|
| 2.6801 | 1.8 | 225 | 2.6855 | |
|
| 2.6683 | 2.0 | 250 | 2.6867 | |
|
| 2.6812 | 2.2 | 275 | 2.6857 | |
|
| 2.6786 | 2.4 | 300 | 2.6862 | |
|
| 2.6726 | 2.6 | 325 | 2.6863 | |
|
| 2.6733 | 2.8 | 350 | 2.6870 | |
|
| 2.664 | 3.0 | 375 | 2.6880 | |
|
| 2.665 | 3.2 | 400 | 2.6871 | |
|
| 2.671 | 3.4 | 425 | 2.6854 | |
|
| 2.6788 | 3.6 | 450 | 2.6870 | |
|
| 2.673 | 3.8 | 475 | 2.6880 | |
|
| 2.648 | 4.0 | 500 | 2.6863 | |
|
| 2.6661 | 4.2 | 525 | 2.6866 | |
|
| 2.6707 | 4.4 | 550 | 2.6856 | |
|
| 2.6799 | 4.6 | 575 | 2.6870 | |
|
| 2.673 | 4.8 | 600 | 2.6874 | |
|
| 2.6757 | 5.0 | 625 | 2.6856 | |
|
| 2.6658 | 5.2 | 650 | 2.6874 | |
|
| 2.6712 | 5.4 | 675 | 2.6869 | |
|
| 2.674 | 5.6 | 700 | 2.6866 | |
|
| 2.6804 | 5.8 | 725 | 2.6866 | |
|
| 2.6755 | 6.0 | 750 | 2.6872 | |
|
| 2.685 | 6.2 | 775 | 2.6870 | |
|
| 2.6701 | 6.4 | 800 | 2.6870 | |
|
| 2.6893 | 6.6 | 825 | 2.6870 | |
|
| 2.6722 | 6.8 | 850 | 2.6870 | |
|
| 2.6783 | 7.0 | 875 | 2.6870 | |
|
| 2.6671 | 7.2 | 900 | 2.6870 | |
|
| 2.6691 | 7.4 | 925 | 2.6870 | |
|
| 2.6947 | 7.6 | 950 | 2.6870 | |
|
| 2.6773 | 7.8 | 975 | 2.6870 | |
|
| 2.6737 | 8.0 | 1000 | 2.6870 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.40.2 |
|
- Pytorch 2.0.0+cu117 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|