LinjieMu/Qwen2.5-VL-7B-StageI

TEXT_SFT_r1

This model is a fine-tuned version of Qwen/Qwen2.5-VL-7B-Instruct on the TEXT_r1 dataset. It achieves the following results on the evaluation set:

• Loss: 0.7760

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
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• total_eval_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.9311	0.3484	25	0.9495
0.8454	0.6969	50	0.8349
0.7698	1.0557	75	0.8031
0.767	1.4042	100	0.7898
0.7408	1.7526	125	0.7821
0.7278	2.1115	150	0.7778
0.7398	2.4599	175	0.7762
0.729	2.8084	200	0.7760

Framework versions

• Transformers 4.51.2
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.21.1