r"""Initialize the adapters. Support full-parameter, freeze and LoRA training. Note that the trainable parameters must be cast to float32.

Init resource pool and worker group

Initialize a single MCP connection.

列出报告

List all available tools.

加载输出和参考数据。 返回: 输出DataFrame和参考DataFrame的元组

从配置文件加载服务器配置

r"""Make an iterator from the DataProto. This is built upon that TensorDict can be used as a normal Pytorch dataset. See https://pytorch.org/t

Pad a DataProto to size divisible by size_divisor Args: size_divisor (int): size divisor Returns: data: (DataProto): the pad

r"""Extract a chat template from the tokenizer.

Note that this function only rename the key in the batch

Note that this operation is in-place

Rotates half the hidden dims of the input.

Rotates half the hidden dims of the input.

保存从MCP获取的数据为CSV文件 参数: text (str): JSON格式的原始数据 file_name (str): 保存的文件名（不含扩展名） path_name (str, opti

Set ulysses sequence parallel process group.

Execute a single step in the agent's workflow. Must be implemented by subclasses to define specific behavior.

尝试调用指定的 OpenAI API 客户端，并进行重试。

broadcast tensor in tp shards across mp_group

broadcast tensor in tp shards across mp_group

r"""Configure root logger using a stdout stream handler with an explicit format.

r"""Encode formatted inputs to pairs of token ids. Turn 0: prefix + system + query resp Turn t: query

评估控盘程度

评估抛售压力

评估套牢深度

Fix potential encoding issues in HTML content

生成主力成本分析文本

生成套牢分析文本

r"""Get arguments to load config/tokenizer/model. Note: including inplace operation of model_args.

r"""Preprocesses the dataset, including format checking and tokenization.

Get type information for a single type.

处理代码生成和执行动作

r"""Validate the configuration.

Refresh the list of available tools from the MCP server.

通用的报告保存方法

r"""Update gradio components according to the (elem_id, properties) mapping.

Alias for register_agent

Build the memory buffer given weight_buffer_meta Args: weight_buffer_meta: contains mapping from name to a dictionary containing shape an

for cuda memory alignment, make sure alignment by 128-bits

计算每个问题的最短COT长度。 返回: 问题ID到最短COT长度的字典

同步调用LLM

基于数据目录的key与handler配置,获取明细数据

与AI模型进行对话，支持工具调用 参数: messages (List[Dict]): 对话消息列表 返回: tuple: (响应消息, 消息类型) - 消息类型为"message"或"tool_cal

Check if token limits are exceeded

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

Converts a single index or a sequence of indices in a token or a sequence of tokens, using the vocabulary and added tokens.

r"""Create reference model for PPO/DPO training. Evaluation mode is not supported. The valuehead parameter is randomly initialized since it is us

Disconnect from a specific MCP server or all servers if no server_id provided.

r"""Return multiple pairs of token ids representing prompts and responses respectively.

Execute a Web search and return detailed search results. Args: query: The search query to submit to the search engine

Execute the tool with given parameters.

从文本中提取\boxed{}中的答案

r"""Find all available modules to apply LoRA, GaLore or APOLLO.

r"""Fix proxy settings for gradio ui.