Return a tensor with the input `shape` as a view into the 1-D data starting at `start_index`.

move the batch to device Args: device (torch.device, str): torch device Returns: DataProto: the current Data

Take action for the robot.

Copy src from hdfs to local if src is on hdfs or directly return src. If cache_dir is None, we will use the default cache dir of the system. Note

Initialization TODO - `self.FRAME_IDX`: The index of the file saved for the current scene. - `self.fcitx5-configtool`: Left g

Execute action in environment

Concat a list of DataProto. The batch is concatenated among dim=0. The meta_info is assumed to be identical and will use the first one.

Create a huggingface pretrained tokenizer. Args: name (str): The name of the tokenizer. correct_pad_token (bool): Whether to corr

更新点的位置

Select a subset of the DataProto via batch_keys and meta_info_keys Args: batch_keys (list, optional): a list of strings indicatin

Close the environment

Split the batch among dim=0 into chunks. The meta_info is passed to each DataProto after split. Args: chunks (int): the number of

Create a DataProto from a dict of tensors. This assumes that 1. All the tensor in tensors have the same dim0 2. Only dim0 is the batch

See: https://github.com/pytorch/pytorch/issues/563#issuecomment-330103591

Add entity to scene. Add bias to z axis if scene is not sapien.Scene.

Return a tensor with the input `shape` as a view into the 1-D data starting at `start_index`.

Update rendering to refresh the camera's RGBD information (rendering must be updated even when disabled, otherwise data cannot be col

Get ulysses sequence parallel process group.

Compute mean of tensor with a masked values.

pad a 2D tensors (e.g. responses, logprobs) in the last dim to max_seq_length. input shape: [bs, seq_length] output shape: [bs, max_seq_l

Set gripper posture - `left_pos`: Left gripper pose - `right_pos`: Right gripper pose - `set_tag`: "left" to set the

given TP,DP,PP rank to get the global rank.

Split the batch into a list of micro_batches, where the max_token_len is smaller than max_token_len and the number of valid tokens in each micro b

broadcast tensor from rank0 across mp_group

Compute logits given a batch of data. Args: data (DataProto): a batch of data represented by DataProto. It must contain key ```in

Plan multiple paths - matching CuroboPlanner's interface

Interpolative planning with screw motion. Will not avoid collision and will fail if the path contains collision.

Update point cloud for collision checking

TODO: optimize this. Technically, we only need one broadcast

Build the model.

Performs one decoding iteration and returns newly generated results. This function performs one decoding iteration of the engine. It first

broadcast tensor across mp_group