Commits

Branch selector

tool-use-fixes

User selector

All users

All time

Commit history

Commits on Dec 24, 2025

• .

  joschu

  committed
  72b82bdView commit details

  Copy full SHA for 72b82bd

  Browse repository at this point

• .

  joschu

  committed
  d68ae5cView commit details

  Copy full SHA for d68ae5c

  Browse repository at this point

• Add HuggingFace reference link to Qwen3Renderer docstring

  Show description for 1fd12b5
  

  joschu

  and

  claude

  committed
  1fd12b5View commit details

  Copy full SHA for 1fd12b5

  Browse repository at this point

• Make test docstrings self-contained and timeless

  Show description for fcc8240
  

  joschu

  and

  claude

  committed
  fcc8240View commit details

  Copy full SHA for fcc8240

  Browse repository at this point

• Fix type check errors in test_tool_calling.py

  Show description for aebf840
  

  joschu

  and

  claude

  committed
  aebf840View commit details

  Copy full SHA for aebf840

  Browse repository at this point

• Add unit tests for tool calling bug fixes

  Show description for 264f314
  

  joschu

  and

  claude

  committed
  264f314View commit details

  Copy full SHA for 264f314

  Browse repository at this point

• Fix tool calling bugs in renderers

  Show description for 88e5ad5
  

  joschu

  and

  claude

  committed
  88e5ad5View commit details

  Copy full SHA for 88e5ad5

  Browse repository at this point

• Improve phrasing in tool_use/search README (#213)

  Show description for e3e3746
  

  claudeaceae

  and

  claude

  authored
  e3e3746View commit details

  Copy full SHA for e3e3746

  Browse repository at this point

• Remove duplicate word in llms-full.txt (#215)

  Show description for c6aaada
  

  claudeaceae

  and

  claude

  authored
  c6aaadaView commit details

  Copy full SHA for c6aaada

  Browse repository at this point

Commits on Dec 20, 2025

• Cookbook Recipe of Using Rubric (#166)

  Show description for be77c0c
  

  ruiqi-zhong

  and

  joschu

  authored
  be77c0cView commit details

  Copy full SHA for be77c0c

  Browse repository at this point

• Fix numerical instability in DPO loss calculation (#188)

  atgctg

  authored
  6fdb5f8View commit details

  Copy full SHA for 6fdb5f8

  Browse repository at this point

• fix(wandb): recursively serialize nested objects in dump_config (#196)

  Show description for c3bf74a
  

  bledden

  and

  claude

  authored
  c3bf74aView commit details

  Copy full SHA for c3bf74a

  Browse repository at this point

Commits on Dec 19, 2025

• Add open source checkpoints for on-policy distillation (#191)

  kzl-tml

  authored
  8f32fc8View commit details

  Copy full SHA for 8f32fc8

  Browse repository at this point

• Add Qwen3VLInstructRenderer to correctly omit <think> tags for VL Instruct models (#189)

  tyler-griggs

  authored
  92d3693View commit details

  Copy full SHA for 92d3693

  Browse repository at this point

• fix: correct pig latin training examples for proper consonant cluster… (#169)

  nathon-lee

  authored
  62ba5c3View commit details

  Copy full SHA for 62ba5c3

  Browse repository at this point

• fix: correct margin calculation in DPO training (#172)

  nathon-lee

  authored
  bc22ba7View commit details

  Copy full SHA for bc22ba7

  Browse repository at this point

• Set min transformers version to 4.57.1 (#185)

  opherlieber

  authored
  5f70d5dView commit details

  Copy full SHA for 5f70d5d

  Browse repository at this point

Commits on Dec 18, 2025

• Revise visualization example in training-sampling.mdx (#182)

  Show description for a0675ee
  

  nintang

  and

  joschu

  authored
  a0675eeView commit details

  Copy full SHA for a0675ee

  Browse repository at this point

• Sync typo fixes from tinker-docs (#187)

  Show description for 41875b6
  

  joschu

  and

  claude

  authored
  41875b6View commit details

  Copy full SHA for 41875b6

  Browse repository at this point

Commits on Dec 17, 2025

• WIP: Improve renderer test coverage and document HF compatibility (#186)

  Show description for 60be6cf
  

  3 peopleauthored
  60be6cfView commit details

  Copy full SHA for 60be6cf

  Browse repository at this point

• Fix Qwen3Renderer stripping out last <think>, and Qwen3DisableThinkingRenderer not adding <think>\n\n</think> (#178)

  thejaminator

  authored
  fa6fc32View commit details

  Copy full SHA for fa6fc32

  Browse repository at this point

Commits on Dec 16, 2025

• Support 'fast' ImageProcessor (#184)

  opherlieber

  authored
  eb7382fView commit details

  Copy full SHA for eb7382f

  Browse repository at this point

• update sign-up link (#183)

  jshanson7

  authored
  e418cb6View commit details

  Copy full SHA for e418cb6

  Browse repository at this point

• Add pillow to pyproject.toml dependencies (#164)

  Show description for 95771a6
  

  Copilot

  and

  rowanz

  authored
  95771a6View commit details

  Copy full SHA for 95771a6

  Browse repository at this point

• fix: fix grammar in Search-R1 README (#171)

  nathon-lee

  authored
  037afebView commit details

  Copy full SHA for 037afeb

  Browse repository at this point

Commits on Dec 14, 2025

• Rewrite AGENTS.md to reference the documentation that we copied in (#175)

  joschu

  authored
  1d30325View commit details

  Copy full SHA for 1d30325

  Browse repository at this point

• Add API reference documentation (#174)

  joschu

  authored
  65d0111View commit details

  Copy full SHA for 65d0111

  Browse repository at this point

• Add docs from (separate) documentation repo. Update llms.txt (#173)

  joschu

  authored
  28ca48fView commit details

  Copy full SHA for 28ca48f

  Browse repository at this point

Commits on Dec 13, 2025

• Simplify clock cycle management code for better readability (#167)

  Show description for 281c82f
  

  joschu

  and

  claude

  authored
  281c82fView commit details

  Copy full SHA for 281c82f

  Browse repository at this point

Commits on Dec 12, 2025

• Improve clock cycle management for RL (#162)

  erikwijmans

  authored
  3b4d2f4View commit details

  Copy full SHA for 3b4d2f4

  Browse repository at this point

• Fix constant group removal (#163)

  erikwijmans

  authored
  900bc33View commit details

  Copy full SHA for 900bc33

  Browse repository at this point

• Add vision input recipe (#161)

  rowanz

  authored
  944af9aView commit details

  Copy full SHA for 944af9a

  Browse repository at this point

Commits on Dec 6, 2025

• create a changelog (#152)

  joschu

  authored
  989f849View commit details

  Copy full SHA for 989f849

  Browse repository at this point

• feat(qwen3): Add strip_thinking_from_history option to Qwen3Renderer (#142)

  Show description for 0924ba8
  

  joschu

  and

  claude

  authored
  0924ba8View commit details

  Copy full SHA for 0924ba8

  Browse repository at this point

• Allow training runs with saving disabled via `save_every=0` (#149)

  Show description for 365634d
  

  4 peopleauthored
  365634dView commit details

  Copy full SHA for 365634d

  Browse repository at this point