metadata
license: apache-2.0
library_name: peft
tags:
- trl
- dpo
- generated_from_trainer
base_model: norallm/normistral-7b-warm
model-index:
- name: ap-normistral-7b-align-scan
results: []
ap-normistral-7b-align-scan
This model is a fine-tuned version of norallm/normistral-7b-warm on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4970
- Rewards/chosen: 0.0579
- Rewards/rejected: 0.0502
- Rewards/accuracies: 0.5565
- Rewards/margins: 0.0077
- Logps/rejected: -35.8660
- Logps/chosen: -32.3273
- Logits/rejected: 98.3156
- Logits/chosen: 98.3222
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: 4
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.4665 | 0.26 | 100 | 0.4971 | 0.0771 | 0.0651 | 0.5187 | 0.0119 | -35.8363 | -32.2890 | 98.7124 | 98.7206 |
| 0.3678 | 0.52 | 200 | 0.4983 | -0.0042 | -0.0160 | 0.5104 | 0.0118 | -35.9986 | -32.4516 | 98.3577 | 98.3655 |
| 0.3546 | 0.78 | 300 | 0.4970 | 0.0579 | 0.0502 | 0.5565 | 0.0077 | -35.8660 | -32.3273 | 98.3156 | 98.3222 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1