andyc03/Qwen3.5-9B-attack-v2.1

<!-- 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. -->

output_saber

This model is a fine-tuned version of /scratch/aszalay1/nanxi/Models/Qwen3.5-9B on the saberattacktrain dataset. It achieves the following results on the evaluation set:

• —Loss: 0.9440

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: 2e-06
• —trainbatchsize: 1
• —evalbatchsize: 1
• —seed: 42
• —distributed_type: multi-GPU
• —num_devices: 4
• —gradientaccumulationsteps: 16
• —totaltrainbatch_size: 64
• —totalevalbatch_size: 4
• —optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_steps: 0.03
• —num_epochs: 1.0

Training results

Framework versions

• —Transformers 5.5.3
• —Pytorch 2.11.0+cu129
• —Datasets 3.6.0
• —Tokenizers 0.22.2