K2Tilly/llama-finetune-qwen3-4b-MAP_math

<!-- 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. -->

train_run

This model is a fine-tuned version of Qwen/Qwen3-4B-Instruct-2507 on the qwen3mathmisconception_sharegpt dataset. It achieves the following results on the evaluation set:

• —Loss: 0.1027
• —Num Input Tokens Seen: 41149088

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-05
• —trainbatchsize: 4
• —evalbatchsize: 4
• —seed: 42
• —gradientaccumulationsteps: 8
• —totaltrainbatch_size: 32
• —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_steps: 200
• —num_epochs: 3.0

Training results

Framework versions

• —PEFT 0.17.1
• —Transformers 4.56.1
• —Pytorch 2.8.0+cu126
• —Datasets 4.0.0
• —Tokenizers 0.22.0