Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on Jul 10, 2026

• feat(online_dpo): support AGENT_RUNNER selection and mixed datasets

  Show description for 7c7ade3
  

  lvhaoyu

  and

  claude

  committed
  7c7ade3View commit details

  Copy full SHA for 7c7ade3

  Browse repository at this point

• refactor(online_dpo): remove reward/ directory, use uni_agent.reward.llm_judge directly

  Show description for f401b5f
  

  lvhaoyu

  and

  claude

  committed
  f401b5fView commit details

  Copy full SHA for f401b5f

  Browse repository at this point

• refactor(online_dpo): simplify to sandbox runners + pluggable scoring

  Show description for 6ee5953
  

  lvhaoyu

  and

  claude

  committed
  6ee5953View commit details

  Copy full SHA for 6ee5953

  Browse repository at this point

• fix: use urlparse for repo URL validation to prevent SSRF bypass

  Show description for a7f4f07
  

  lvhaoyu

  and

  claude

  committed
  a7f4f07View commit details

  Copy full SHA for a7f4f07

  Browse repository at this point

• feat: add Claude-Code-style agent runner for online DPO

  Show description for 4b02e13
  

  lvhaoyu

  and

  claude

  committed
  4b02e13View commit details

  Copy full SHA for 4b02e13

  Browse repository at this point

• feat: add dataset-specific batch LLM judge for online DPO

  Show description for 1e012cc
  

  lvhaoyu

  and

  claude

  committed
  1e012ccView commit details

  Copy full SHA for 1e012cc

  Browse repository at this point

• fix: repair online DPO + uni-agent gateway integration and remove redundant code

  Show description for a632bed
  

  lvhaoyu

  and

  claude

  committed
  a632bedView commit details

  Copy full SHA for a632bed

  Browse repository at this point

• update doc for online dpo

  

  lvhaoyu

  committed
  a2341fcView commit details

  Copy full SHA for a2341fc

  Browse repository at this point

• support uniagent for dpo

  

  lvhaoyu

  committed
  f7cdc12View commit details

  Copy full SHA for f7cdc12

  Browse repository at this point

Commits on Jul 9, 2026

• fix: sanitize workspace paths, avoid shell injection in scripts

  Show description for 86842fb
  

  lvhaoyu

  and

  claude

  committed
  86842fbView commit details

  Copy full SHA for 86842fb

  Browse repository at this point

• feat: add sandbox tools for online DPO with multi-backend support

  Show description for 1dc98e5
  

  lvhaoyu

  and

  claude

  committed
  1dc98e5View commit details

  Copy full SHA for 1dc98e5

  Browse repository at this point

Commits on Jul 8, 2026

• update batch dpo with model judge

  

  lvhaoyu

  committed
  d12964dView commit details

  Copy full SHA for d12964d

  Browse repository at this point

• update

  

  lvhaoyu

  committed
  e36e338View commit details

  Copy full SHA for e36e338

  Browse repository at this point

• update

  

  lvhaoyu

  committed
  384daf9View commit details

  Copy full SHA for 384daf9

  Browse repository at this point

• add agent online dpo

  

  lvhaoyu

  committed
  f47a653View commit details

  Copy full SHA for f47a653

  Browse repository at this point

• new

  

  lvhaoyu

  committed
  c5ad556View commit details

  Copy full SHA for c5ad556

  Browse repository at this point

• Clarify DPO reward calculation requirements

  

  frelam

  authored and

  lvhaoyu

  committed
  03b2ff2View commit details

  Copy full SHA for 03b2ff2

  Browse repository at this point

• Expand agent abilities with new insights on DPO and SFT

  Show description for b5c7e6e
  

  frelam

  authored and

  lvhaoyu

  committed
  b5c7e6eView commit details

  Copy full SHA for b5c7e6e

  Browse repository at this point

• Add honesty and humility as an ability

  

  frelam

  authored and

  lvhaoyu

  committed
  29bc16cView commit details

  Copy full SHA for 29bc16c

  Browse repository at this point

• add agent reject sampling

  

  lvhaoyu

  committed
  ad4c644View commit details

  Copy full SHA for ad4c644

  Browse repository at this point

• add agent sft

  

  lvhaoyu

  committed
  a8cda8dView commit details

  Copy full SHA for a8cda8d

  Browse repository at this point

• update

  

  lvhaoyu

  committed
  b1f5cc9View commit details

  Copy full SHA for b1f5cc9

  Browse repository at this point

• add optimizer reset per iter feature

  

  lvhaoyu

  committed
  d57fceaView commit details

  Copy full SHA for d57fcea

  Browse repository at this point

• Update agent abilities with clearer descriptions

  Show description for 3f7812c
  

  frelam

  authored and

  lvhaoyu

  committed
  3f7812cView commit details

  Copy full SHA for 3f7812c

  Browse repository at this point

• Add new abilities for agent in agent_ability.md

  

  frelam

  authored and

  lvhaoyu

  committed
  870aaadView commit details

  Copy full SHA for 870aaad

  Browse repository at this point

• Create agent_ability.md

  

  frelam

  authored and

  lvhaoyu

  committed
  feca7edView commit details

  Copy full SHA for feca7ed

  Browse repository at this point

• Refactor math solution extraction to use new method

  

  frelam

  authored and

  lvhaoyu

  committed
  405d8f7View commit details

  Copy full SHA for 405d8f7

  Browse repository at this point

• Refactor math_reward.py for clarity and efficiency

  Show description for edbd3d5
  

  frelam

  authored and

  lvhaoyu

  committed
  edbd3d5View commit details

  Copy full SHA for edbd3d5

  Browse repository at this point

• Add reward calculation functions in reward.py

  

  frelam

  authored and

  lvhaoyu

  committed
  e30fcdfView commit details

  Copy full SHA for e30fcdf

  Browse repository at this point

• Add max effort handling in mixed reasoning dataset

  

  frelam

  authored and

  lvhaoyu

  committed
  5c310f4View commit details

  Copy full SHA for 5c310f4

  Browse repository at this point

• Update max_occurrences and add data_source parameter

  

  frelam

  authored and

  lvhaoyu

  committed
  f5850a4View commit details

  Copy full SHA for f5850a4

  Browse repository at this point

• Update thinking format pattern in check_thinking_format

  

  frelam

  authored and

  lvhaoyu

  committed
  37033fdView commit details

  Copy full SHA for 37033fd

  Browse repository at this point

• Add mixed reasoning dataset generation script

  

  frelam

  authored and

  lvhaoyu

  committed
  736b934View commit details

  Copy full SHA for 736b934

  Browse repository at this point

• Update length bonus logic for max-effort data source

  Show description for 1d188d0
  

  frelam

  authored and

  lvhaoyu

  committed
  1d188d0View commit details

  Copy full SHA for 1d188d0

  Browse repository at this point

• Add mix_sharegpt_agentinstruct_sft.py for dataset mixing

  Show description for 0d35c3e
  

  frelam

  authored and

  lvhaoyu

  committed
  0d35c3eView commit details

  Copy full SHA for 0d35c3e

  Browse repository at this point