zephyr-7b-dpo-full-magpi-low-bleu-3-epochs

This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-full on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0004
• Rewards/chosen: -1.8844
• Rewards/rejected: -46.8077
• Rewards/accuracies: 1.0
• Rewards/margins: 44.9232
• Logps/rejected: -5321.5576
• Logps/chosen: -555.4259
• Logits/rejected: 2.7529
• Logits/chosen: -1.2323

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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.0066	0.4739	50	0.0028	-1.0908	-33.9616	0.9980	32.8709	-4036.9529	-476.0595	-1.2144	-1.9103
0.0177	0.9479	100	0.0006	-1.6117	-43.9541	1.0	42.3424	-5036.1978	-528.1522	1.4562	-2.1299
0.0006	1.4218	150	0.0004	-1.7244	-46.1666	1.0	44.4422	-5257.4517	-539.4232	1.6969	-1.9837
0.0002	1.8957	200	0.0005	-1.7575	-44.7450	1.0	42.9875	-5115.2886	-542.7341	2.1634	-2.0033
0.0001	2.3697	250	0.0004	-1.8985	-46.5225	1.0	44.6240	-5293.0405	-556.8339	2.7114	-1.2429
0.0001	2.8436	300	0.0004	-1.8844	-46.8077	1.0	44.9232	-5321.5576	-555.4259	2.7529	-1.2323

Framework versions

• Transformers 4.44.0.dev0
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1