Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 26, 2025

• Small fix on contributing docs (#4753)

  murilo-cunha

  authored
  2337cc9

  Copy full SHA for 2337cc9

Commits on Dec 24, 2025

• docs: fix "Good Second Issue" redirection link (#4749)

  casinca

  authored
  c04fdc0

  Copy full SHA for c04fdc0

• Change tiny model dtype from float16 to bfloat16 to fix CUDA error (#4745)

  albertvillanova

  authored
  51113d1

  Copy full SHA for 51113d1

Commits on Dec 23, 2025

• Avoid docstyle formatting for `TestParseResponse` (#4736)

  qgallouedec

  authored
  19021f8

  Copy full SHA for 19021f8

• Hotfix for browsergym openenv notebook (#4740)

  sergiopaniego

  authored
  5e6ceed

  Copy full SHA for 5e6ceed

• RLOO supports async rewards. (#4718)

  Show description for 044058e
  

  pramodith

  and

  qgallouedec

  authored
  044058e

  Copy full SHA for 044058e

• Add GRPO QLoRA free notebook (#4660)

  sergiopaniego

  authored
  c9a9e25

  Copy full SHA for c9a9e25

• Add uv/hf jobs support to OpenEnv scripts (#4720)

  sergiopaniego

  authored
  11c9867

  Copy full SHA for 11c9867

• Include data type for tiny models and update tests (#4728)

  qgallouedec

  authored
  32bf84d

  Copy full SHA for 32bf84d

Commits on Dec 22, 2025

• Fix: handle multiple tool calls in `qwen3_schema` (#4709)

  mattbui

  authored
  22b386a

  Copy full SHA for 22b386a

Commits on Dec 21, 2025

• Upgrade GitHub Actions for Node 24 compatibility (#4733)

  salmanmkc

  authored
  490f4c7

  Copy full SHA for 490f4c7

• Upgrade GitHub Actions to latest versions (#4734)

  Show description for f0bce66
  salmanmkc

  authored
  f0bce66

  Copy full SHA for f0bce66

Commits on Dec 20, 2025

• Remove experimental imports from testing_utils (#4727)

  albertvillanova

  authored
  38e1a35

  Copy full SHA for 38e1a35

Commits on Dec 19, 2025

• Apply docstyle

  qgallouedec

  committed
  a515085

  Copy full SHA for a515085

• Refactor vLLM generation [3/N]: Decouple profiling from trainer (#4717)

  Show description for 29a39ab
  

  albertvillanova

  and

  qgallouedec

  authored
  29a39ab

  Copy full SHA for 29a39ab

• Fix deprecation version for RLOO max_prompt_length (#4726)

  albertvillanova

  authored
  8611eba

  Copy full SHA for 8611eba

• Disallow `PeftModel` + `peft_config` in trainers (#4713)

  qgallouedec

  authored
  bb5cd8a

  Copy full SHA for bb5cd8a

• Update agents notebook dependencies (#4724)

  sergiopaniego

  authored
  20691d0

  Copy full SHA for 20691d0

Commits on Dec 18, 2025

• Upload FunctionGemma notebook (#4721)

  sergiopaniego

  authored
  1dc8bbc

  Copy full SHA for 1dc8bbc

• Overwrite model default generation config used by model.generate (#4647)

  albertvillanova

  authored
  8918c98

  Copy full SHA for 8918c98

• Support async reward functions and parallelize call to reward functions. (#4567)

  Show description for 50bc248
  

  pramodith

  and

  qgallouedec

  authored
  50bc248

  Copy full SHA for 50bc248

• Add inference example to GRPO agent training notebook (#4710)

  sergiopaniego

  authored
  13ad537

  Copy full SHA for 13ad537

• Fix test assertion for `top_k` parameter in `OnlineDPOTrainer` (#4714)

  qgallouedec

  authored
  0ead75b

  Copy full SHA for 0ead75b

• [docs] Fix RapidFire AI position in documentation (#4715)

  qgallouedec

  authored
  23a941a

  Copy full SHA for 23a941a

• Fix KeyError with transformers 5.0.0+ where push_to_hub_token is removed (#4691)

  Show description for 976a2c0
  

  Manodeepray

  and

  qgallouedec

  authored
  976a2c0

  Copy full SHA for 976a2c0

• docs: Add RapidFire AI cross-references to DPO and GRPO trainer docs (#4705)

  Show description for 30e9894
  

  3 peopleauthored
  30e9894

  Copy full SHA for 30e9894

Commits on Dec 17, 2025

• Align RLOO with GRPO (#4706)

  Show description for 157cd63
  

  qgallouedec

  and

  albertvillanova

  authored
  157cd63

  Copy full SHA for 157cd63

• BrowserGym example for LLMs (no vision) (#4696)

  Show description for 06a897d
  

  sergiopaniego

  and

  qgallouedec

  authored
  06a897d

  Copy full SHA for 06a897d

• Deprecate max_prompt_length in RLOOTrainer (#4703)

  albertvillanova

  authored
  ca54e24

  Copy full SHA for ca54e24

Commits on Dec 16, 2025

• Include `generation_config` for tiny model uploads (#4643)

  qgallouedec

  authored
  997368b

  Copy full SHA for 997368b

• Preserve truncated tokens in BFD packing (#4632)

  Show description for ec70ef2
  

  qgallouedec

  and

  albertvillanova

  authored
  ec70ef2

  Copy full SHA for ec70ef2

• Move `get_reward` function to `experimental.utils` (#4683)

  qgallouedec

  authored
  a12faa5

  Copy full SHA for a12faa5

• Align use of vllm_max_model_length in RLOOTrainer (#4702)

  albertvillanova

  authored
  61c9921

  Copy full SHA for 61c9921

• Align GRPO and RLOO initialization (#4685)

  qgallouedec

  authored
  00da046

  Copy full SHA for 00da046

• Move `prepare_model_for_kbit_training`, `enable_gradient_checkpointing`, `prepare_peft_model` to `experimental.utils` (#4704)

  qgallouedec

  authored
  20cc2e1

  Copy full SHA for 20cc2e1

Pagination

Previous

Next