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 an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.7740
- Rewards/chosen: -1.9500
- Rewards/rejected: -2.1880
- Rewards/accuracies: 0.5664
- Rewards/margins: 0.2381
- Logps/rejected: -737.1887
- Logps/chosen: -585.4705
- Logits/rejected: -5.6032
- Logits/chosen: -5.3452
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: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- 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 |
---|---|---|---|---|---|---|---|---|---|---|---|
0.4622 | 0.2558 | 100 | 0.7195 | -0.4883 | -0.5174 | 0.4883 | 0.0292 | -570.1285 | -439.2997 | -5.0069 | -4.7883 |
0.3618 | 0.5115 | 200 | 0.7717 | -1.6081 | -1.7766 | 0.5703 | 0.1685 | -696.0472 | -551.2876 | -5.4044 | -5.1462 |
0.3249 | 0.7673 | 300 | 0.7740 | -1.9500 | -2.1880 | 0.5664 | 0.2381 | -737.1887 | -585.4705 | -5.6032 | -5.3452 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.1.2+cu118
- Datasets 2.19.1
- Tokenizers 0.19.1