|
--- |
|
license: llama3 |
|
base_model: meta-llama/Meta-Llama-3-8B-Instruct |
|
tags: |
|
- trl |
|
- sft |
|
- generated_from_trainer |
|
model-index: |
|
- name: UTI_L3_1000steps_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. --> |
|
|
|
# UTI_L3_1000steps_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: 1.6055 |
|
|
|
## 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 | |
|
|:-------------:|:-------:|:----:|:---------------:| |
|
| 2.4485 | 0.3333 | 25 | 2.4666 | |
|
| 2.4645 | 0.6667 | 50 | 2.4522 | |
|
| 2.452 | 1.0 | 75 | 2.4164 | |
|
| 2.391 | 1.3333 | 100 | 2.3529 | |
|
| 2.2816 | 1.6667 | 125 | 2.2866 | |
|
| 2.175 | 2.0 | 150 | 2.2255 | |
|
| 2.2168 | 2.3333 | 175 | 2.1683 | |
|
| 2.1574 | 2.6667 | 200 | 2.1166 | |
|
| 2.1107 | 3.0 | 225 | 2.0679 | |
|
| 2.0126 | 3.3333 | 250 | 2.0229 | |
|
| 1.9353 | 3.6667 | 275 | 1.9810 | |
|
| 1.9552 | 4.0 | 300 | 1.9445 | |
|
| 1.9759 | 4.3333 | 325 | 1.9100 | |
|
| 1.8721 | 4.6667 | 350 | 1.8773 | |
|
| 1.8928 | 5.0 | 375 | 1.8491 | |
|
| 1.8331 | 5.3333 | 400 | 1.8236 | |
|
| 1.8221 | 5.6667 | 425 | 1.7980 | |
|
| 1.7615 | 6.0 | 450 | 1.7762 | |
|
| 1.7701 | 6.3333 | 475 | 1.7562 | |
|
| 1.7034 | 6.6667 | 500 | 1.7327 | |
|
| 1.7471 | 7.0 | 525 | 1.7064 | |
|
| 1.7317 | 7.3333 | 550 | 1.6831 | |
|
| 1.6897 | 7.6667 | 575 | 1.6645 | |
|
| 1.6452 | 8.0 | 600 | 1.6476 | |
|
| 1.6675 | 8.3333 | 625 | 1.6327 | |
|
| 1.569 | 8.6667 | 650 | 1.6238 | |
|
| 1.705 | 9.0 | 675 | 1.6163 | |
|
| 1.6025 | 9.3333 | 700 | 1.6121 | |
|
| 1.6224 | 9.6667 | 725 | 1.6083 | |
|
| 1.6976 | 10.0 | 750 | 1.6074 | |
|
| 1.6031 | 10.3333 | 775 | 1.6059 | |
|
| 1.5703 | 10.6667 | 800 | 1.6046 | |
|
| 1.6563 | 11.0 | 825 | 1.6055 | |
|
| 1.6464 | 11.3333 | 850 | 1.6059 | |
|
| 1.6075 | 11.6667 | 875 | 1.6055 | |
|
| 1.6453 | 12.0 | 900 | 1.6057 | |
|
| 1.5754 | 12.3333 | 925 | 1.6054 | |
|
| 1.5962 | 12.6667 | 950 | 1.6055 | |
|
| 1.6333 | 13.0 | 975 | 1.6055 | |
|
| 1.6086 | 13.3333 | 1000 | 1.6055 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.41.2 |
|
- Pytorch 2.0.0+cu117 |
|
- Datasets 2.19.2 |
|
- Tokenizers 0.19.1 |
|
|
