leobianco/npov_RM_organic_Mistral-7B-Instruct-v0_3_S130104_epo15_lr2_7e-04_r32_2609211259

<!-- 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. -->

npovRMorganicMistral-7B-Instruct-v03S130104epo15lr27e-04r322609211259

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on an unknown dataset. It achieves the following results on the evaluation set:

• —Loss: 1.7617
• —Roc Auc: 0.9269
• —Best Threshold: 0.9892
• —Tpr At Best Threshold: 0.8974
• —Fpr At Best Threshold: 0.1
• —Accuracy At Best Threshold: 0.8980
• —Avg Score True Positives: 0.9260
• —Avg Score True Negatives: 0.2980

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: 0.000267175562302475
• —trainbatchsize: 16
• —evalbatchsize: 32
• —seed: 130104
• —distributed_type: multi-GPU
• —num_devices: 2
• —totaltrainbatch_size: 32
• —totalevalbatch_size: 64
• —optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_steps: 0.1
• —num_epochs: 15.0

Training results

Framework versions

• —PEFT 0.20.0
• —Transformers 5.14.1
• —Pytorch 2.11.0+cu130
• —Datasets 5.0.1
• —Tokenizers 0.22.2