Skip to content

Actions: huggingface/trl

Actions

Hugging Face Issue Labeler

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
issue_auto_labeller.yml
725 workflow runs
725 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

PPO grad_norm is 0.0 Hugging Face Issue Labeler #651: Issue #3961 opened by faker52
27s
27s
sft_video_llm example fail Hugging Face Issue Labeler #650: Issue #3958 opened by yao-matrix
22s
22s
sft_gemma3 example doesn't work Hugging Face Issue Labeler #649: Issue #3957 opened by yao-matrix
27s
27s
Why not use AutoModel for ref_model in grpo trainer? Hugging Face Issue Labeler #648: Issue #3948 opened by csshihao
30s
30s
LoRA-GA: Low-Rank Adaptation with Gradient Approximation Hugging Face Issue Labeler #647: Issue #3945 opened by electroglyph
28s
28s
Can we avoid saving the optimization stage? Hugging Face Issue Labeler #646: Issue #3944 opened by HelloWorldLTY
22s
22s
Should the reference model of GRPO be an actual model? Hugging Face Issue Labeler #645: Issue #3939 opened by shaojun0
26s
26s
Bugs when sft with mixing pure text and multimodal data Hugging Face Issue Labeler #644: Issue #3934 opened by shiym2000
31s
31s
GRPOTrainer with top_entropy_quntile < 1 causes hang with multi gpu training Hugging Face Issue Labeler #643: Issue #3933 opened by avishaiElmakies
27s
27s
Add support for RLPR Hugging Face Issue Labeler #642: Issue #3928 opened by mitchelldehaven
23s
23s
Using assistant_only_loss=True with sequence length > max_length fails silently Hugging Face Issue Labeler #641: Issue #3927 opened by jonnyli1125
31s
31s
SFTTrainer freezes LoRA adapter when using PEFT model as argument Hugging Face Issue Labeler #640: Issue #3926 opened by dvgodoy
32s
32s
When I want to use client to request trl vllm-serve, it always timeout when init_communicator() Hugging Face Issue Labeler #639: Issue #3925 opened by JackeyZhang1001
28s
28s
Feature request for GRPO trainer: vLLM guided decoding with xgrammar (JSON) Hugging Face Issue Labeler #638: Issue #3924 opened by sanghak123
27s
27s
Does the trl library support the Lite PPO algorithm? Hugging Face Issue Labeler #637: Issue #3920 opened by ArcherShirou
29s
29s
How to use trl-SFTTrainer to train Qwen-30B-A3B? Hugging Face Issue Labeler #636: Issue #3918 opened by JeffWb
23s
23s
Inconsistent apply_chat_template behaviour for multimodal dataset between SFT and GRPO Hugging Face Issue Labeler #635: Issue #3915 opened by ishaan-rawal-ai
36s
36s
max_length causing training loss issues not found in v0.19.1 Hugging Face Issue Labeler #634: Issue #3910 opened by AmazingGabriel16
28s
28s
[GRPO Trainer] Accuracy reward stays 0 Hugging Face Issue Labeler #633: Issue #3903 opened by Revist
25s
25s
sft_gemma3 example fail Hugging Face Issue Labeler #632: Issue #3901 opened by yao-matrix
30s
30s
How to gather completions before computing rewards in GRPOTrainer Hugging Face Issue Labeler #631: Issue #3896 opened by rubickkcibur
28s
28s
What are the important metrics that help gauge accuracy and validity of training? Hugging Face Issue Labeler #630: Issue #3893 opened by RageItalic
28s
28s
Is there a standard GRPO GSM8k process for TRL? Hugging Face Issue Labeler #629: Issue #3892 opened by mitchelldehaven
25s
25s
Bug in BFD packing Hugging Face Issue Labeler #628: Issue #3887 opened by RicardoDominguez
29s
29s
GRPO fails with the "server" vllm_mode when num_processes > 1 Hugging Face Issue Labeler #627: Issue #3885 opened by Kirill-Kravtsov
22s
22s