Results are not same

[varsha-devi]

Hello, I successfully loaded the pre-trained model given for mrw dataset using the following command in eval.py script

"checkpoint = torch.load(args.ckpt) model.load_state_dict(checkpoint, strict=False)"

but the results are not same as given in the article or in github repo.

The results generated by evaluating pre-trained model "mrw_pvse.pth"

Loading dataset Computing results... (eval_on_gpu=True) Images: 5000, Sentences: 5000 rsum: 0.48 Average i2t Recall: 0.07 Image to text: 0.02 0.06 0.14 2509.00 (0.50) 2509.74 (0.50) Average t2i Recall: 0.09 Text to image: 0.02 0.06 0.18 2500.00 (0.50) 2500.98 (0.50)

Am I doing any mistake