biomistral-7b-iter-sft-dpo-step1-wo-kqa_golden-iter-dpo-step1
This model is a fine-tuned version of Minbyul/biomistral-7b-wo-kqa_golden-iter-sft-step1 on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:
- Loss: 0.5984
- Rewards/chosen: 0.0255
- Rewards/rejected: -0.2139
- Rewards/accuracies: 0.8438
- Rewards/margins: 0.2394
- Logps/rejected: -352.9643
- Logps/chosen: -123.9539
- Logits/rejected: -3.0295
- Logits/chosen: -3.1903
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: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- 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: 1
Training results
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.1.2
- Datasets 2.14.6
- Tokenizers 0.15.2