Search results

Is GRPOTrainer compatible?

Status: Open.

#25 In anhvth/opensloth;

· Jovanqing opened on Nov 13, 2025

[Feature] Support Python 3.10 and 3.11

Status: Open.

#24 In anhvth/opensloth;

· steveepreston opened on Oct 13, 2025

[Feature] TPU Support

Status: Open.

#23 In anhvth/opensloth;

· steveepreston opened on Oct 11, 2025

GRPO Patch

Status: Open.

#22 In anhvth/opensloth;

· tsunghan-wu opened on Oct 3, 2025

Distributed Training Crash: torchrun DDP SIGSEGV with Multi-GPU and Unsloth Patch

Status: Open.

#21 In anhvth/opensloth;

· wangxiaodong1021 opened on Oct 1, 2025

How can I turn off logging?

Status: Open.

#20 In anhvth/opensloth;

· kaen2891 opened on Sep 22, 2025

How to train VLM model like Qwen2.5 VL?

Status: Open.

#19 In anhvth/opensloth;

· songtang209 opened on Aug 31, 2025

Fine Tune Qwen3 235B

Status: Open.

#18 In anhvth/opensloth;

· etemiz opened on Aug 18, 2025

destroy_process_group() was not called before program exit, which can leak resources

Status: Open.

#17 In anhvth/opensloth;

· ai-bond opened on Jul 30, 2025

TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType _inner_training_loop and os.path.exists(self.args.resume_from_checkpoint)

Status: Open.

#15 In anhvth/opensloth;

· ai-bond opened on Jul 29, 2025

Unable run train if in system was disabled IPV6 thru kernel params.

Status: Open.

#14 In anhvth/opensloth;

· ai-bond opened on Jul 29, 2025

How To Display Eval/Loss Diagram in WanDB?

Status: Open.

#13 In anhvth/opensloth;

· devianl2 opened on Jul 18, 2025