Converts an id to a token, special tokens included

Converts a token to an id using the vocab, special tokens included

Returns a sampler for evaluation.

load the models and its relative processor, tokenizer. Args: model_name (str): the model name or model path. Returns: N

Initialize the weights.

Master loop to dispatch tasks to workers

监控线程，定期报告状态

节点队列同步线程

Converts a string in a sequence of tokens (string), using the tokenizer. Split in words for word-based vocabulary or sub-words for

Convert AITZ action to Qwen2.5 action format Args: aitz_action (dict): AITZ format action, contains ACTION and ARGS resi

Determines if two actions are considered to be the same. Two actions being "the same" is defined here as two actions that would result in a s

Converts a sequence of tokens in a single string.

Modified from the origin code. Do the inference based on the params. Args: image (Image.Image) instruction (str) pre

Generate batch from context tokens.

Get response from Qwen model Args: user_query: User query text screenshot: Screenshot path history_actions: Hist

Returns the training [`~torch.utils.data.DataLoader`]. Will use no sampler if `train_dataset` does not implement `__len__`, a random

Manage Global Task Synchronization Signal And Assign Advantages.

Args: predict_bbox: Tensor of shape [bs, 2], format [x, y] (coordinates of predicted points). target_bbox: Tensor of shape [bs, 4

Args: predict_bbox: Tensor of shape [bs, 2], format [x, y] (coordinates of predicted points). target_bbox: Tensor of shape [bs, 4