Jiahang Xu

PNNX无法转换torchvision.vit model

1

## error log | 日志或报错信息 | ログ 我使用的python代码如下： ``` python import torch import torchvision.models as models model = models.vit_l_16(pretrained=True) model = model.eval() model_input = torch.rand(1, 3, 224, 224) mod =...

Hi Xiusu, I hope this message finds you well. I'm wondering if there is any information of your searched ViTAS_Twin subnet, including T, S, B, L. Actually, I found the...

Hi, I'm Jiahang from Microsoft LongRoPE team. Thank you for your interest in our work and for creating a comprehensive implementation of it on GitHub. We truly appreciate your effort...

Add a `trace_agg_mode` option to control whether the trace aggregation is performed per transition or per trajectory (by adding response mask for multi-turn). Default remains 'transition' for backward compatibility.