This model is a fine-tuned version of Qwen/Qwen3-VL-8B-Instruct on the openai_sharegpt_reward dataset. It achieves the following results on the evaluation set:

Loss: 0.5400

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-05
train_batch_size: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 16
total_train_batch_size: 128
total_eval_batch_size: 8
optimizer: Use 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: 0.05
num_epochs: 8.0

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	4	0.7659
No log	2.0	8	0.6173
0.7412	3.0	12	0.5734
0.7412	4.0	16	0.5528
0.5315	5.0	20	0.5444
0.5315	6.0	24	0.5409
0.5315	7.0	28	0.5402
0.4632	8.0	32	0.5400

Framework versions

Transformers 4.57.1
Pytorch 2.4.0+cu124
Datasets 4.0.0
Tokenizers 0.22.1

ruitongl/qwen3-vl-8b-sftfull1

fullsft

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

ruitongl/
qwen3-vl-8b-sftfull1