zephyr-7b-dpo-full-ultrabin-high-margin
This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-full on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:
- Loss: 0.5523
- Rewards/chosen: -0.7174
- Rewards/rejected: -1.6454
- Rewards/accuracies: 0.7461
- Rewards/margins: 0.9280
- Logps/rejected: -427.1986
- Logps/chosen: -334.3585
- Logits/rejected: 0.5161
- Logits/chosen: 0.0387
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: 55
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 64
- 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 |
---|---|---|---|---|---|---|---|---|---|---|---|
0.4719 | 0.3484 | 50 | 0.5899 | -0.3895 | -0.8981 | 0.7070 | 0.5086 | -352.4708 | -301.5678 | -1.9397 | -1.9963 |
0.3224 | 0.6969 | 100 | 0.5598 | -0.6746 | -1.5654 | 0.75 | 0.8907 | -419.1961 | -330.0835 | 0.2134 | -0.2417 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.1.2
- Datasets 2.20.0
- Tokenizers 0.19.1
- Downloads last month
- 0
Model tree for sfulay/zephyr-7b-dpo-full-ultrabin-high-margin
Base model
mistralai/Mistral-7B-v0.1
Finetuned
alignment-handbook/zephyr-7b-sft-full