zephyr-7b-dpo-qlora
This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-qlora on the ilharco/altruism_1k dataset. It achieves the following results on the evaluation set:
- Loss: 0.0420
- Rewards/chosen: -5.5143
- Rewards/rejected: -13.0608
- Rewards/accuracies: 0.9949
- Rewards/margins: 7.5465
- Logps/rejected: -1865.7168
- Logps/chosen: -1348.6318
- Logits/rejected: -2.6274
- Logits/chosen: -2.6531
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
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.1963 | 0.51 | 100 | 0.1328 | -1.0972 | -3.5537 | 1.0 | 2.4565 | -915.0070 | -906.9143 | -2.8398 | -2.8256 |
Framework versions
- PEFT 0.7.1
- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.14.6
- Tokenizers 0.15.2
- Downloads last month
- 1
Unable to determine this model’s pipeline type. Check the
docs
.