htlou/backup_0202_llamafactory_RLAIF-V-Cosi-q0_25_preference-llava-mistral

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

RLAIF-V-Cosi-q025preference

This model is a fine-tuned version of llava-hf/llava-v1.6-mistral-7b-hf on the RLAIF-V-Cosi-q025preference dataset. It achieves the following results on the evaluation set:

• —Loss: 0.5909
• —Rewards/chosen: -2.7988
• —Rewards/rejected: -4.2628
• —Rewards/accuracies: 0.7266
• —Rewards/margins: 1.4639
• —Logps/rejected: -211.2034
• —Logps/chosen: -194.0678
• —Logits/rejected: -2.6060
• —Logits/chosen: -2.6047

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: 1e-06
• —trainbatchsize: 8
• —evalbatchsize: 8
• —seed: 42
• —distributed_type: multi-GPU
• —num_devices: 8
• —gradientaccumulationsteps: 4
• —totaltrainbatch_size: 256
• —totalevalbatch_size: 64
• —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• —lrschedulertype: cosine
• —lrschedulerwarmup_steps: 10
• —num_epochs: 3.0

Training results

Framework versions

• —Transformers 4.45.2
• —Pytorch 2.4.0+cu121
• —Datasets 2.21.0
• —Tokenizers 0.20.3