Search results

CLI help fails: TypeError: not enough arguments for format string

Status: Open.

Bug

#5099 In huggingface/trl;

· albertvillanova opened on Feb 13, 2026

Agent Skills for TRL

Status: Open.

#5095 In huggingface/trl;

· albertvillanova opened on Feb 13, 2026

Training crash with "uncorrectable ECC error" and NCCL timeout?

Status: Open.

#5094 In huggingface/trl;

· 2402227288 opened on Feb 13, 2026

RFC: Defining the TRL Public API (towards v1.0)

Status: Open.

#5090 In huggingface/trl;

· qgallouedec opened on Feb 12, 2026

GRPOTrainer still has unresolved bugs in VLM / VL setting – any plan for fixing?

Status: Open.

#5041 In huggingface/trl;

· wswaq opened on Feb 10, 2026

Source to GRPO online example

Status: Open.

#5033 In huggingface/trl;

· ttrung149 opened on Feb 9, 2026

Fine-Tuning gemma-3-4b-pt raises "ValueError: <code>token_type_ids</code> is required as a model input when training"

🏋 SFTRelated to SFTRelated to SFT

🐛 bugSomething isn't workingSomething isn't working

Status: Open.

#5032 In huggingface/trl;

· jesperschlegel opened on Feb 9, 2026

Maximum Likelihood Reinforcement Learning

Status: Open.

#5025 In huggingface/trl;

· catherinelee274 opened on Feb 9, 2026

DPPO - Divergence Proximal Policy Optimization

Status: Open.

#4998 In huggingface/trl;

· ErykCh opened on Feb 6, 2026

Support multi-turn training in GRPOTrainer

Status: Open.

Feature

#4974 In huggingface/trl;

· albertvillanova opened on Feb 5, 2026

[GRPO] AssertionError in vLLM when using QLoRA (4-bit) due to weight shape mismatch during merge_adapter

Status: Open.

#4973 In huggingface/trl;

· ChengHu98 opened on Feb 5, 2026

Simple Quick Start examples are great – but on what hardware? A note in the doc would be lovely.

Status: Open.

#4968 In huggingface/trl;

· jchwenger opened on Feb 4, 2026