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: 0.4621
- Rewards/chosen: -2.9012
- Rewards/rejected: -4.5213
- Rewards/accuracies: 0.7292
- Rewards/margins: 1.6201
- Logps/rejected: -450.6970
- Logps/chosen: -420.5302
- Logits/rejected: 89.5207
- Logits/chosen: 95.5635
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.1919 | 1.9 | 100 | 0.4767 | -2.8814 | -4.4710 | 0.7188 | 1.5897 | -449.6920 | -420.1344 | 89.6189 | 95.6517 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.2
- Datasets 2.14.6
- Tokenizers 0.15.2
- Downloads last month
- 1