metadata
tags:
- trl
- dpo
- generated_from_trainer
model-index:
- name: zephyr-7b-dpo-full
results: []
zephyr-7b-dpo-full
This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0449
- Rewards/chosen: -0.6529
- Rewards/rejected: -1.2907
- Rewards/accuracies: 0.7383
- Rewards/margins: 0.6378
- Logps/rejected: -518.3444
- Logps/chosen: -448.4197
- Logits/rejected: 1.6154
- Logits/chosen: 1.4915
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-07
- train_batch_size: 4
- eval_batch_size: 8
- seed: 2
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.09 | 0.21 | 100 | 0.0844 | -0.0945 | -0.2957 | 0.6992 | 0.2012 | -418.8424 | -392.5748 | 0.5303 | 0.4926 |
| 0.0405 | 0.42 | 200 | 0.0452 | -0.6625 | -1.1657 | 0.7539 | 0.5032 | -505.8466 | -449.3789 | 1.2734 | 1.1859 |
| 0.0479 | 0.63 | 300 | 0.0477 | -0.5208 | -1.1204 | 0.7383 | 0.5995 | -501.3084 | -435.2133 | 1.4093 | 1.2982 |
| 0.0457 | 0.84 | 400 | 0.0449 | -0.6529 | -1.2907 | 0.7383 | 0.6378 | -518.3444 | -448.4197 | 1.6154 | 1.4915 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1