|
--- |
|
license: gemma |
|
library_name: peft |
|
tags: |
|
- trl |
|
- reward-trainer |
|
- generated_from_trainer |
|
base_model: google/gemma-2b |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: RM-HH-AllMix_harmless_gpt3_20000_gemma2b_shuffleFalse_extractchosenFalse |
|
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. --> |
|
|
|
# RM-HH-AllMix_harmless_gpt3_20000_gemma2b_shuffleFalse_extractchosenFalse |
|
|
|
This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0554 |
|
- Accuracy: 0.9722 |
|
|
|
## 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: 1.41e-05 |
|
- train_batch_size: 1 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 4 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 1.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| 0.6806 | 0.04 | 250 | 0.4207 | 0.8447 | |
|
| 0.6465 | 0.08 | 500 | 0.2075 | 0.9485 | |
|
| 0.5999 | 0.13 | 750 | 0.1133 | 0.9725 | |
|
| 0.5725 | 0.17 | 1000 | 0.0804 | 0.9786 | |
|
| 0.6223 | 0.21 | 1250 | 0.0783 | 0.9778 | |
|
| 0.5595 | 0.25 | 1500 | 0.0632 | 0.9789 | |
|
| 0.5956 | 0.29 | 1750 | 0.0589 | 0.9778 | |
|
| 0.557 | 0.33 | 2000 | 0.0599 | 0.9756 | |
|
| 0.5641 | 0.38 | 2250 | 0.0558 | 0.9767 | |
|
| 0.5871 | 0.42 | 2500 | 0.0589 | 0.9744 | |
|
| 0.5512 | 0.46 | 2750 | 0.0568 | 0.9741 | |
|
| 0.5775 | 0.5 | 3000 | 0.0529 | 0.9756 | |
|
| 0.5923 | 0.54 | 3250 | 0.0555 | 0.9748 | |
|
| 0.548 | 0.59 | 3500 | 0.0577 | 0.9722 | |
|
| 0.564 | 0.63 | 3750 | 0.0579 | 0.9722 | |
|
| 0.563 | 0.67 | 4000 | 0.0599 | 0.9718 | |
|
| 0.5932 | 0.71 | 4250 | 0.0561 | 0.9729 | |
|
| 0.5247 | 0.75 | 4500 | 0.0569 | 0.9725 | |
|
| 0.5472 | 0.79 | 4750 | 0.0579 | 0.9718 | |
|
| 0.5704 | 0.84 | 5000 | 0.0556 | 0.9729 | |
|
| 0.5456 | 0.88 | 5250 | 0.0550 | 0.9725 | |
|
| 0.5563 | 0.92 | 5500 | 0.0545 | 0.9729 | |
|
| 0.552 | 0.96 | 5750 | 0.0554 | 0.9722 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.10.0 |
|
- Transformers 4.38.2 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
