Search results

GRPO with reward model. CUDA out of memory. How to fix? Thank you very much.

🏋 GRPORelated to GRPORelated to GRPO

🏋 RewardRelated to Reward modellingRelated to Reward modelling

Status: Open.

#4419 In huggingface/trl;

· guotong1988 opened on Nov 1, 2025

SFTTrainer creates empty adapter.safetensors file while saving when training with LoRA and Deepspeed Zero3

⚡ PEFTRelated to PEFTRelated to PEFT

🏋 SFTRelated to SFTRelated to SFT

🚀 deepspeedRelated to deepspeedRelated to deepspeed

Status: Open.

#4416 In huggingface/trl;

· maximus-21 opened on Oct 31, 2025

Complete paper index

📚 documentationImprovements or additions to documentationImprovements or additions to documentation

Status: Open.

#4407 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Remove setting chat template in most trainers

🏋 SFTRelated to SFTRelated to SFT

Status: Open.

#4404 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Move <code>masked_mean</code>, <code>masked_var</code> and <code>masked_whiten</code> to <code>ppo.py</code>

🏋 PPORelated to PPORelated to PPO

Status: Open.

#4403 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Make vLLM server OpenAI-compatible

✨ enhancementNew feature or requestNew feature or request

Status: Open.

#4402 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Move everything related to Judges to <code>trl.experimental</code>

Status: Open.

#4400 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Update or remove some of the notebooks

📚 documentationImprovements or additions to documentationImprovements or additions to documentation

Status: Open.

#4399 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Add kernels to Docker images

✨ enhancementNew feature or requestNew feature or request

Status: Open.

#4398 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Remove or move Multi Adapter RL

⚡ PEFTRelated to PEFTRelated to PEFT

📚 documentationImprovements or additions to documentationImprovements or additions to documentation

Status: Open.

#4397 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Remove Sentiment Tuning Examples

Status: Open.

#4396 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025

Move everything related to Mergekit to the experimental submodule

Status: Open.

#4395 In huggingface/trl;

· qgallouedec opened on Oct 30, 2025