Skip to main content
Ctrl+K
Trinity-RFT 0.4.0 文档 - Home
  • v0.3.1

    • latest
    • v0.4.0
    • v0.3.3
    • v0.3.2
    • v0.3.1

教程

  • 安装指南
  • 开发者指南
  • Workflow 开发指南
  • Algorithm 开发指南
  • Algorithm 进阶开发
  • Operator 开发指南
  • 参数配置指南
  • Synchronizer 介绍

样例

  • 快速上手
  • Off-Policy RFT
  • 异步 RFT
  • 多轮对话拼接任务
  • 通用多轮 RFT
  • ReAct Agent 训练
  • 邮件搜索例子
  • 离线 DPO 和 SFT
  • Megatron-LM 支持
  • 数据处理

常见问题

  • 常见问题

API Reference

  • API Reference
    • trinity.buffer package
      • trinity.buffer.operators package
        • trinity.buffer.operators.filters package
        • trinity.buffer.operators.mappers package
        • trinity.buffer.operators.data_juicer_operator module
        • trinity.buffer.operators.experience_operator module
      • trinity.buffer.pipelines package
        • trinity.buffer.pipelines.experience_pipeline module
        • trinity.buffer.pipelines.task_pipeline module
      • trinity.buffer.reader package
        • trinity.buffer.reader.file_reader module
        • trinity.buffer.reader.queue_reader module
        • trinity.buffer.reader.sql_reader module
      • trinity.buffer.schema package
        • trinity.buffer.schema.formatter module
        • trinity.buffer.schema.sql_schema module
      • trinity.buffer.storage package
        • trinity.buffer.storage.file module
        • trinity.buffer.storage.queue module
        • trinity.buffer.storage.sql module
      • trinity.buffer.writer package
        • trinity.buffer.writer.file_writer module
        • trinity.buffer.writer.queue_writer module
        • trinity.buffer.writer.sql_writer module
      • trinity.buffer.buffer module
      • trinity.buffer.buffer_reader module
      • trinity.buffer.buffer_writer module
      • trinity.buffer.utils module
    • trinity.explorer package
      • trinity.explorer.api package
        • trinity.explorer.api.api module
        • trinity.explorer.api.service module
      • trinity.explorer.explorer module
      • trinity.explorer.explorer_client module
      • trinity.explorer.scheduler module
      • trinity.explorer.workflow_runner module
    • trinity.trainer package
      • trinity.trainer.verl package
        • trinity.trainer.verl.dp_actor module
        • trinity.trainer.verl.fsdp_checkpoint_manager module
        • trinity.trainer.verl.fsdp_workers module
        • trinity.trainer.verl.megatron_actor module
        • trinity.trainer.verl.megatron_checkpoint_manager module
        • trinity.trainer.verl.megatron_workers module
        • trinity.trainer.verl.utils module
      • trinity.trainer.trainer module
      • trinity.trainer.verl_trainer module
    • trinity.algorithm package
      • trinity.algorithm.advantage_fn package
        • trinity.algorithm.advantage_fn.advantage_fn module
        • trinity.algorithm.advantage_fn.asymre_advantage module
        • trinity.algorithm.advantage_fn.grpo_advantage module
        • trinity.algorithm.advantage_fn.multi_step_grpo_advantage module
        • trinity.algorithm.advantage_fn.opmd_advantage module
        • trinity.algorithm.advantage_fn.ppo_advantage module
        • trinity.algorithm.advantage_fn.rec_advantage module
        • trinity.algorithm.advantage_fn.reinforce_advantage module
        • trinity.algorithm.advantage_fn.reinforce_plus_plus_advantage module
        • trinity.algorithm.advantage_fn.remax_advantage module
        • trinity.algorithm.advantage_fn.rloo_advantage module
      • trinity.algorithm.entropy_loss_fn package
        • trinity.algorithm.entropy_loss_fn.entropy_loss_fn module
      • trinity.algorithm.kl_fn package
        • trinity.algorithm.kl_fn.kl_fn module
      • trinity.algorithm.policy_loss_fn package
        • trinity.algorithm.policy_loss_fn.chord_policy_loss module
        • trinity.algorithm.policy_loss_fn.cispo_policy_loss module
        • trinity.algorithm.policy_loss_fn.dpo_loss module
        • trinity.algorithm.policy_loss_fn.gspo_policy_loss module
        • trinity.algorithm.policy_loss_fn.mix_policy_loss module
        • trinity.algorithm.policy_loss_fn.opmd_policy_loss module
        • trinity.algorithm.policy_loss_fn.policy_loss_fn module
        • trinity.algorithm.policy_loss_fn.ppo_policy_loss module
        • trinity.algorithm.policy_loss_fn.rec_policy_loss module
        • trinity.algorithm.policy_loss_fn.sft_loss module
        • trinity.algorithm.policy_loss_fn.sppo_loss_fn module
        • trinity.algorithm.policy_loss_fn.topr_policy_loss module
      • trinity.algorithm.sample_strategy package
        • trinity.algorithm.sample_strategy.mix_sample_strategy module
        • trinity.algorithm.sample_strategy.sample_strategy module
        • trinity.algorithm.sample_strategy.utils module
      • trinity.algorithm.algorithm module
      • trinity.algorithm.key_mapper module
      • trinity.algorithm.utils module
    • trinity.manager package
      • trinity.manager.config_registry package
        • trinity.manager.config_registry.algorithm_config_manager module
        • trinity.manager.config_registry.buffer_config_manager module
        • trinity.manager.config_registry.config_registry module
        • trinity.manager.config_registry.explorer_config_manager module
        • trinity.manager.config_registry.model_config_manager module
        • trinity.manager.config_registry.trainer_config_manager module
      • trinity.manager.config_manager module
      • trinity.manager.state_manager module
      • trinity.manager.synchronizer module
    • trinity.common package
      • trinity.common.models package
        • trinity.common.models.mm_utils module
        • trinity.common.models.model module
        • trinity.common.models.utils module
        • trinity.common.models.vllm_model module
        • trinity.common.models.vllm_worker module
      • trinity.common.rewards package
        • trinity.common.rewards.accuracy_reward module
        • trinity.common.rewards.agents_reward module
        • trinity.common.rewards.countdown_reward module
        • trinity.common.rewards.dapo_reward module
        • trinity.common.rewards.format_reward module
        • trinity.common.rewards.human_reward module
        • trinity.common.rewards.math_reward module
        • trinity.common.rewards.reward_fn module
        • trinity.common.rewards.tool_reward module
        • trinity.common.rewards.utils module
      • trinity.common.workflows package
        • trinity.common.workflows.agentscope package
        • trinity.common.workflows.agentscope_workflow module
        • trinity.common.workflows.customized_math_workflows module
        • trinity.common.workflows.customized_toolcall_workflows module
        • trinity.common.workflows.eval_workflow module
        • trinity.common.workflows.math_rm_workflow module
        • trinity.common.workflows.math_ruler_workflow module
        • trinity.common.workflows.math_trainable_ruler_workflow module
        • trinity.common.workflows.rubric_judge_workflow module
        • trinity.common.workflows.simple_mm_workflow module
        • trinity.common.workflows.step_wise_workflow module
        • trinity.common.workflows.workflow module
      • trinity.common.config module
      • trinity.common.constants module
      • trinity.common.experience module
      • trinity.common.verl_config module
    • trinity.utils package
      • trinity.utils.annotations module
      • trinity.utils.distributed module
      • trinity.utils.dlc_utils module
      • trinity.utils.eval_utils module
      • trinity.utils.log module
      • trinity.utils.lora_utils module
      • trinity.utils.math_eval_utils module
      • trinity.utils.monitor module
      • trinity.utils.plugin_loader module
      • trinity.utils.registry module
      • trinity.utils.timer module
  • .rst

trinity.manager.config_manager module

目录

  • ConfigManager
    • ConfigManager.__init__()
    • ConfigManager.reset_session_state()
    • ConfigManager.maintain_session_state()
    • ConfigManager.get_configs()
    • ConfigManager.beginner_mode()
    • ConfigManager.expert_mode()
    • ConfigManager.generate_config()
    • ConfigManager.run_config()

trinity.manager.config_manager module#

class trinity.manager.config_manager.ConfigManager[源代码]#

基类:object

__init__()[源代码]#
reset_session_state()[源代码]#
maintain_session_state()[源代码]#
get_configs(*config_names: str, columns_spec: List[int] = None)[源代码]#
beginner_mode()[源代码]#
expert_mode()[源代码]#
generate_config()[源代码]#
run_config(parent, yaml_config: str) → None[源代码]#

上一页

trinity.manager.config_registry.trainer_config_manager module

下一页

trinity.manager.state_manager module

目录
  • ConfigManager
    • ConfigManager.__init__()
    • ConfigManager.reset_session_state()
    • ConfigManager.maintain_session_state()
    • ConfigManager.get_configs()
    • ConfigManager.beginner_mode()
    • ConfigManager.expert_mode()
    • ConfigManager.generate_config()
    • ConfigManager.run_config()

作者: Trinity-RFT Team

© Copyright 2025, Trinity-RFT Team.