metadata
license: other
base_model: HuggingFaceH4/zephyr-7b-gemma-sft-v0.1
tags:
- alignment-handbook
- trl
- dpo
- generated_from_trainer
- trl
- dpo
- generated_from_trainer
datasets:
- argilla/dpo-mix-7k
model-index:
- name: zephyr-7b-gemma-dpo
results: []
zephyr-7b-gemma-dpo
This model is a fine-tuned version of HuggingFaceH4/zephyr-7b-gemma-sft-v0.1 on the argilla/dpo-mix-7k dataset. It achieves the following results on the evaluation set:
- Loss: 0.4694
- Rewards/chosen: -3.9430
- Rewards/rejected: -5.5568
- Rewards/accuracies: 0.7292
- Rewards/margins: 1.6137
- Logps/rejected: -473.0876
- Logps/chosen: -441.5945
- Logits/rejected: 88.3937
- Logits/chosen: 94.2780
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: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2
Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.1534 | 1.8957 | 100 | 0.4631 | -3.9815 | -5.5353 | 0.7083 | 1.5538 | -472.6590 | -442.3637 | 88.4677 | 94.3385 |
Framework versions
- Transformers 4.41.0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.19.1