Hi! First of all, thank you for such a complete open release — scripts, data, and checkpoints all working out of the box is rare, and it's been genuinely helpful.
One small request: would it be possible to also share the intermediate reasoning-SFT checkpoints (before the GRPO stage) for Games / Office / Industrial? As far as I can tell, only the final post-GRPO models are currently up.
Thanks again for the great work!
Hi! First of all, thank you for such a complete open release — scripts, data, and checkpoints all working out of the box is rare, and it's been genuinely helpful.
One small request: would it be possible to also share the intermediate reasoning-SFT checkpoints (before the GRPO stage) for Games / Office / Industrial? As far as I can tell, only the final post-GRPO models are currently up.
Thanks again for the great work!