unslothai/unsloth Issues | Good First Issue

Python (64,271 stars) (5,658 forks) (51 indexed issues) (PR metrics pending)Last commit May 15, 2026

Issues

51 open indexed issues

RSS Atom

Open

Docs: current Codex local llama.cpp guide does not work as written with Responses API

good first issue

unslothai/unsloth #5,141 opened last month · Python · 64,271 stars

4 comments1 reaction0 assignees

Open

[Feature] Request permission before downloading model on first run

feature requesthelp wanted

unslothai/unsloth #5,125 opened last month · Python · 64,271 stars

4 comments0 reactions0 assignees

Open

[Feature] Model Management in Unsloth Studio - Delete & Update Models

feature requestgood first issuehelp wanted

unslothai/unsloth #5,098 opened last month · Python · 64,271 stars

1 comment2 reactions0 assignees

Open

[Feature] Running Usloth Studion in kaggle

feature requesthelp wanted

unslothai/unsloth #4,944 opened 2 months ago · Python · 64,271 stars

0 comments0 reactions0 assignees

Open

[Feature] Unsloth studio: Set dataset streaming mode

feature requestgood first issuehelp wanted

unslothai/unsloth #4,903 opened 2 months ago · Python · 64,271 stars

8 comments0 reactions1 assignee

Open

[Feature] Draft model / speculative decoding

feature requestgood first issuehelp wanted

unslothai/unsloth #4,753 opened 2 months ago · Python · 64,271 stars

4 comments2 reactions0 assignees

Open

[Feature] Unsloth/ Whisper/Large-v3 - S3 Bucket connection

feature requestgood first issuehelp wanted

unslothai/unsloth #4,539 opened 2 months ago · Python · 64,271 stars

3 comments0 reactions0 assignees

Open

[Bug] Qwen3.5 vision training , error "only for image model"

help wanted

unslothai/unsloth #4,202 opened 3 months ago · Python · 64,271 stars

4 comments0 reactions0 assignees

Open

[Feature] add aarch64 container images

feature requesthelp wanted

unslothai/unsloth #4,198 opened 3 months ago · Python · 64,271 stars

0 comments0 reactions0 assignees

Open

[Docs] Update DPO example to use DPOConfig instead of TrainingArguments

good first issuehelp wanted

unslothai/unsloth #4,155 opened 3 months ago · Python · 64,271 stars

4 comments0 reactions0 assignees

Open

[Feature Request] Add Idefics3 architecture support (Granite Docling VLM)

help wanted

unslothai/unsloth #4,079 opened 3 months ago · Python · 64,271 stars

3 comments0 reactions0 assignees

Open

[Feature Request] fast inference for LFM (and Mamba models)

good first issuehelp wanted

unslothai/unsloth #4,073 opened 3 months ago · Python · 64,271 stars

5 comments0 reactions0 assignees

Open

Request for Notebook to Fine-Tune Qwen TTS (or Alternatives Using Existing Notebooks)

feature requestgood first issuehelp wanted

unslothai/unsloth #3,961 opened 4 months ago · Python · 64,271 stars

0 comments0 reactions0 assignees

Open

How can I increase the GPU utilization while training / finetune an LLM with unsloth

help wanted

unslothai/unsloth #3,957 opened 4 months ago · Python · 64,271 stars

1 comment0 reactions0 assignees

Open

[Feature] Add FT support for the Qwen3-TTS model.

feature requestgood first issuehelp wanted

unslothai/unsloth #3,951 opened 4 months ago · Python · 64,271 stars

6 comments1 reaction0 assignees

Open

How to set "reasoning_effort" of GPT-OSS during GRPO rollouts?

help wanted

unslothai/unsloth #3,949 opened 4 months ago · Python · 64,271 stars

3 comments0 reactions0 assignees

Open

[Bug] After adding import unsloth to the first line of the script, the GRPOTrainer fails to run properly; however, it works normally again once this import is removed. The Sophia optimizer interface being used was generated by an AI.

help wanted

unslothai/unsloth #3,591 opened 7 months ago · Python · 64,271 stars

11 comments0 reactions0 assignees

Open

[Bug] Forced coupling between num_generations and per_device_train_batch_size in GRPOTrainer resulting in OOM

help wanted

unslothai/unsloth #3,572 opened 7 months ago · Python · 64,271 stars

3 comments1 reaction0 assignees

Open

[Bug] Cannot load qwen3-vl series with lora adapter on vllm.

good first issue

unslothai/unsloth #3,560 opened 7 months ago · Python · 64,271 stars

5 comments0 reactions0 assignees

Open

reinforce(gspo) training didn't yield any improments

help wantedunsure bug?

unslothai/unsloth #3,485 opened 7 months ago · Python · 64,271 stars

1 comment0 reactions0 assignees