COMMITS
April 2, 2026
P
🔒 pin tests.yml actions to commit SHAs (#721)
Pauline Bailly-Masson committed
July 17, 2025
Q
Update README.md
Quentin Gallouédec committed
July 7, 2025
Q
Update citation (#688)
Quentin Gallouédec committed
May 28, 2025
L
Bump vLLM and TRL (#665)
lewtun committed
L
Fix Weka refresh (#666)
lewtun committed
L
Set DP=2 for smol model evals (#664)
lewtun committed
May 27, 2025
L
Refresh Weka on Slurm (#662)
lewtun committed
L
Align EOS token ID between tokenizer and generation config (#663)
lewtun committed
L
Bump deps (#656)
lewtun committed
May 26, 2025
L
Add OpenR1-Distill recipe (#661)
lewtun committed
May 25, 2025
L
Add better logging defaults for GRPO (#657)
lewtun committed
G
GRPO with codeforces problems (#627)
Guilherme Penedo committed
May 22, 2025
L
Bump lower bound on liger-kernel (#654)
lewtun committed
May 21, 2025
L
Bump DeepSpeed to 0.16.8 to fix OOM on Qwen3 (#653)
lewtun committed
May 20, 2025
L
Add dataset mixer (#647)
lewtun committed
May 19, 2025
Q
use hf papers (#646)
Quentin Gallouédec committed
May 16, 2025
L
Bump LightEval (#643)
lewtun committed
E
Add dataset filtering script (#637)
Edward Beeching committed
May 9, 2025
L
Add time to Slurm (#639)
lewtun committed
L
Use pass@1 for all evals (#633)
lewtun committed
E
soft_overlong_punishment from DAPO paper (#638)
Edward Beeching committed
May 8, 2025
L
Fix style again :) (#636)
lewtun committed
A
Code Execution using Morph Cloud (#614)
Andrei committed
May 5, 2025
L
Fix style (#631)
lewtun committed
April 30, 2025
L
Update README.md
lewtun committed
B
🦜Enhance repetition penalty reward for language that cannot be split by whitespace (#516)
binary-husky committed
L
Bump LightEval to enable DP>1 (#629)
lewtun committed
April 26, 2025
L
Enable data and tensor parallelism for GRPO (#626)
lewtun committed
April 17, 2025
E
add back grad accumulations steps (#612)
Edward Beeching committed