`torch_dtype` is deprecated! Use `dtype` instead! You are attempting to use Flash Attention 2 without specifying a torch dtype. This might lead to unexpected behaviour /root/eval/work/Isaac-GR00T/gr00t/model/modules/dit.py:255: FutureWarning: Accessing config attribute `compute_dtype` directly via 'AlternateVLDiT' object attribute is deprecated. Please access 'compute_dtype' over 'AlternateVLDiT's config object instead, e.g. 'unet.config.compute_dtype'. embedding_dim=self.inner_dim, compute_dtype=self.compute_dtype /root/eval/work/Isaac-GR00T/gr00t/model/modules/dit.py:286: FutureWarning: Accessing config attribute `output_dim` directly via 'AlternateVLDiT' object attribute is deprecated. Please access 'output_dim' over 'AlternateVLDiT's config object instead, e.g. 'unet.config.output_dim'. self.proj_out_2 = nn.Linear(self.inner_dim, self.output_dim) Total number of DiT parameters: 1091722240 Total number of SelfAttentionTransformer parameters: 201433088 Loading checkpoint shards: 0%| | 0/3 [00:00