metadata
library_name: peft
tags:
- trl
- dpo
- DPO
- WeniGPT
- generated_from_trainer
base_model: Weni/WeniGPT-Agents-Mistral-1.0.0-SFT-merged
model-index:
- name: WeniGPT-Agents-Mistral-1.0.0-SFT-1.0.25-DPO
results: []
WeniGPT-Agents-Mistral-1.0.0-SFT-1.0.25-DPO
This model is a fine-tuned version of Weni/WeniGPT-Agents-Mistral-1.0.0-SFT-merged on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0048
- Rewards/chosen: 3.9169
- Rewards/rejected: -6.2067
- Rewards/accuracies: 1.0
- Rewards/margins: 10.1235
- Logps/rejected: -240.9936
- Logps/chosen: -117.4673
- Logits/rejected: -1.9027
- Logits/chosen: -1.8786
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: 5e-06
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.03
- training_steps: 180
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.225 | 0.9677 | 30 | 0.2227 | 2.6681 | -1.0085 | 1.0 | 3.6765 | -215.0026 | -123.7113 | -1.9566 | -1.9326 |
| 0.0853 | 1.9355 | 60 | 0.1095 | 3.7286 | -2.2208 | 1.0 | 5.9494 | -221.0642 | -118.4086 | -1.9400 | -1.9170 |
| 0.0285 | 2.9032 | 90 | 0.0545 | 4.1460 | -3.9811 | 1.0 | 8.1270 | -229.8655 | -116.3218 | -1.9301 | -1.9063 |
| 0.001 | 3.8710 | 120 | 0.0468 | 4.1806 | -5.0175 | 1.0 | 9.1980 | -235.0477 | -116.1489 | -1.9141 | -1.8902 |
| 0.0021 | 4.8387 | 150 | 0.0087 | 3.9958 | -5.9294 | 1.0 | 9.9252 | -239.6072 | -117.0728 | -1.9056 | -1.8815 |
| 0.0014 | 5.8065 | 180 | 0.0048 | 3.9169 | -6.2067 | 1.0 | 10.1235 | -240.9936 | -117.4673 | -1.9027 | -1.8786 |
Framework versions
- PEFT 0.10.0
- Transformers 4.40.0
- Pytorch 2.1.0+cu118
- Datasets 2.18.0
- Tokenizers 0.19.1