|
--- |
|
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_shuffleTrue_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_shuffleTrue_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: 0.5035 |
|
- Accuracy: 0.7281 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| 0.6954 | 0.04 | 250 | 0.6616 | 0.5961 | |
|
| 0.6049 | 0.08 | 500 | 0.5946 | 0.6491 | |
|
| 0.559 | 0.13 | 750 | 0.5750 | 0.6657 | |
|
| 0.5508 | 0.17 | 1000 | 0.5561 | 0.6800 | |
|
| 0.5562 | 0.21 | 1250 | 0.5426 | 0.6886 | |
|
| 0.5167 | 0.25 | 1500 | 0.5325 | 0.6916 | |
|
| 0.5202 | 0.29 | 1750 | 0.5269 | 0.7029 | |
|
| 0.4852 | 0.33 | 2000 | 0.5255 | 0.7070 | |
|
| 0.4796 | 0.38 | 2250 | 0.5265 | 0.7093 | |
|
| 0.5037 | 0.42 | 2500 | 0.5196 | 0.7119 | |
|
| 0.4928 | 0.46 | 2750 | 0.5221 | 0.7146 | |
|
| 0.5138 | 0.5 | 3000 | 0.5257 | 0.7206 | |
|
| 0.4909 | 0.54 | 3250 | 0.5216 | 0.7202 | |
|
| 0.5197 | 0.59 | 3500 | 0.5130 | 0.7273 | |
|
| 0.5086 | 0.63 | 3750 | 0.5107 | 0.7277 | |
|
| 0.5366 | 0.67 | 4000 | 0.5075 | 0.7281 | |
|
| 0.5121 | 0.71 | 4250 | 0.5043 | 0.7292 | |
|
| 0.4945 | 0.75 | 4500 | 0.5048 | 0.7285 | |
|
| 0.5112 | 0.79 | 4750 | 0.5039 | 0.7303 | |
|
| 0.4971 | 0.84 | 5000 | 0.5022 | 0.7300 | |
|
| 0.5118 | 0.88 | 5250 | 0.5031 | 0.7285 | |
|
| 0.4899 | 0.92 | 5500 | 0.5029 | 0.7288 | |
|
| 0.4671 | 0.96 | 5750 | 0.5035 | 0.7281 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.10.0 |
|
- Transformers 4.38.2 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
