|
--- |
|
base_model: meta-llama/Meta-Llama-3-8B |
|
datasets: |
|
- generator |
|
library_name: peft |
|
license: llama3 |
|
tags: |
|
- trl |
|
- sft |
|
- generated_from_trainer |
|
model-index: |
|
- name: Meta-Llama-3-8B_AviationQA-cosine |
|
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. --> |
|
|
|
# Meta-Llama-3-8B_AviationQA-cosine |
|
|
|
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the generator dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6061 |
|
|
|
## 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: 0.0001 |
|
- train_batch_size: 3 |
|
- eval_batch_size: 6 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 6 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_ratio: 0.03 |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 0.7872 | 0.0590 | 50 | 0.7652 | |
|
| 0.7373 | 0.1181 | 100 | 0.7328 | |
|
| 0.7242 | 0.1771 | 150 | 0.7182 | |
|
| 0.7143 | 0.2361 | 200 | 0.7107 | |
|
| 0.73 | 0.2952 | 250 | 0.7046 | |
|
| 0.7159 | 0.3542 | 300 | 0.6973 | |
|
| 0.7211 | 0.4132 | 350 | 0.6921 | |
|
| 0.7096 | 0.4723 | 400 | 0.6873 | |
|
| 0.6845 | 0.5313 | 450 | 0.6824 | |
|
| 0.7251 | 0.5903 | 500 | 0.6783 | |
|
| 0.6685 | 0.6494 | 550 | 0.6720 | |
|
| 0.697 | 0.7084 | 600 | 0.6667 | |
|
| 0.7006 | 0.7674 | 650 | 0.6639 | |
|
| 0.6952 | 0.8264 | 700 | 0.6618 | |
|
| 0.6649 | 0.8855 | 750 | 0.6596 | |
|
| 0.6877 | 0.9445 | 800 | 0.6553 | |
|
| 0.6673 | 1.0035 | 850 | 0.6531 | |
|
| 0.6611 | 1.0626 | 900 | 0.6487 | |
|
| 0.6971 | 1.1216 | 950 | 0.6452 | |
|
| 0.6652 | 1.1806 | 1000 | 0.6423 | |
|
| 0.645 | 1.2397 | 1050 | 0.6397 | |
|
| 0.6494 | 1.2987 | 1100 | 0.6388 | |
|
| 0.6623 | 1.3577 | 1150 | 0.6359 | |
|
| 0.6552 | 1.4168 | 1200 | 0.6334 | |
|
| 0.6465 | 1.4758 | 1250 | 0.6297 | |
|
| 0.6495 | 1.5348 | 1300 | 0.6285 | |
|
| 0.6521 | 1.5939 | 1350 | 0.6272 | |
|
| 0.6505 | 1.6529 | 1400 | 0.6261 | |
|
| 0.6773 | 1.7119 | 1450 | 0.6238 | |
|
| 0.6487 | 1.7710 | 1500 | 0.6225 | |
|
| 0.639 | 1.8300 | 1550 | 0.6208 | |
|
| 0.6465 | 1.8890 | 1600 | 0.6194 | |
|
| 0.6528 | 1.9481 | 1650 | 0.6182 | |
|
| 0.6265 | 2.0071 | 1700 | 0.6164 | |
|
| 0.6161 | 2.0661 | 1750 | 0.6137 | |
|
| 0.6236 | 2.1251 | 1800 | 0.6118 | |
|
| 0.6371 | 2.1842 | 1850 | 0.6111 | |
|
| 0.6294 | 2.2432 | 1900 | 0.6093 | |
|
| 0.6257 | 2.3022 | 1950 | 0.6087 | |
|
| 0.6204 | 2.3613 | 2000 | 0.6081 | |
|
| 0.6133 | 2.4203 | 2050 | 0.6073 | |
|
| 0.6108 | 2.4793 | 2100 | 0.6068 | |
|
| 0.622 | 2.5384 | 2150 | 0.6066 | |
|
| 0.6233 | 2.5974 | 2200 | 0.6064 | |
|
| 0.6183 | 2.6564 | 2250 | 0.6063 | |
|
| 0.6237 | 2.7155 | 2300 | 0.6062 | |
|
| 0.6388 | 2.7745 | 2350 | 0.6062 | |
|
| 0.6236 | 2.8335 | 2400 | 0.6062 | |
|
| 0.6236 | 2.8926 | 2450 | 0.6062 | |
|
| 0.6205 | 2.9516 | 2500 | 0.6061 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.11.1 |
|
- Transformers 4.41.2 |
|
- Pytorch 2.3.0+cu121 |
|
- Datasets 2.20.0 |
|
- Tokenizers 0.19.1 |
