r"""Add the user input to chatbot. Inputs: infer.chatbot, infer.messages, infer.role, infer.query, infer.escape_html Output: infer.ch

Returns a new ToolResult with the given fields replaced.

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

Pop a subset of the DataProto via `batch_keys` and `meta_info_keys` Args: batch_keys (list, optional): a list of strings indicati

r"""Register a chat template. To add the following chat template: ``` <s><user>user prompt here <model>model response here</s> <u

统一搜索接口 Args: keywords: 搜索关键词 max_results: 最大结果数 Returns: 搜索结果列表，每个结果包含 title, url, desc

Factory method to create and properly initialize an MCPAgent instance.

r"""Forms a list of slots according to the inputs to encode.

Converts a sequence of ids in a string, using the tokenizer and vocabulary with options to remove special tokens and clean up tokeniz

If distributed is initialized, print only on rank 0.

生成回答。 参数: prompt: 包含system和user字段的字典 max_new_tokens: 最大生成token数 temperature: 温度参数

r"""Get plugin for multimodal inputs.

Run the MCP server.

r"""Optionally check the package version.

r"""Get element by id. Example: top.lang, train.dataset

Convert message to dictionary format

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

r"""Validate if this model accepts the input modalities.

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.

Converts a string to a sequence of ids (integer), using the tokenizer and vocabulary. Args: text (`str`, `List[str]` or

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

r"""Get chat template and fixes the tokenizer.

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

r"""Load pretrained tokenizer and optionally loads processor. Note: including inplace operation of model_args.

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

Load a dataset from a JSON file. Loads and parses a JSON dataset file based on the provided dataset enum. The file path is constructed based

Create a huggingface pretrained tokenizer which correctness handles eos and pad tokens. Args: name (str): The name of the tokenizer.

Repeat the batch data a specified number of times. Args: repeat_times (int): Number of times to repeat the data.

r"""Check if the environment variable is enabled.

Union with another DataProto. Union batch and meta_info separately. Throw an error if - there are conflict keys in batch and they are

r"""Register a new composite model. Args: model_type: model type projector_key: multi_modal_projector vision_model_keys:

A func to sync embedding input with alltoall in sequence parallel gather sequence dimension and scatter head dim: e.g. seq_dim: 1, head_d

r"""Validate if the number of images, videos and audios match the number of placeholders in messages.

Add a message to memory

r"""Get the path to saved model checkpoints.

r"""Load pretrained model.

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

Save data to a JSONL (JSON Lines) file. Args: data: List of objects to save as JSON lines file_path: Target file path

r"""Pre-process input messages before tokenization for VLMs.

Count the number of tokens in a text using GPT-4 tokenizer. Args: text: Text to measure Returns: int: Numbe

r"""Return a single pair of token ids representing prompt and response respectively.

Evaluate a single sample against its ground truth. Args: sample: Dictionary containing 'ground truth' and 'p

Get ulysses sequence parallel world size.

r"""Get the train dataset and optionally gets the evaluation dataset.

Compute mean of tensor with a masked values.

This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, num_key_value_heads, seqlen, he

Create a system message

关闭客户端

r"""Works like shutil.copy() for file, and shutil.copytree for dir, and supports hdfs. Copy data and mode bits ("cp src dst"). Return the file's