Search results

[Docs] Ministral 3 Spelling & Qwen3-2507 & Qwen3.5 Issues

Status: Open.

#4176 In unslothai/unsloth;

· MyzTyn opened on Mar 7, 2026

[Bug] Official 500K gpt-oss-20b notebook now fails with CUDA OOM

bug

Status: Open.

#4175 In unslothai/unsloth;

· Minkyuman opened on Mar 7, 2026

Is Vision DPO Training supported?

Status: Open.

#4174 In unslothai/unsloth;

· HamzaSulaiman1996 opened on Mar 6, 2026

[Bug] Unable to finetune older models like <code>SmolLM2</code> and <code>Qwen2</code> with unsloth

bug

Status: Open.

#4172 In unslothai/unsloth;

· ParagEkbote opened on Mar 6, 2026

[Feature] Phi-4-multimodal

feature requestFeature request pending on roadmapFeature request pending on roadmap

Status: Open.

#4169 In unslothai/unsloth;

· Halfbreed86 opened on Mar 6, 2026

[Bug] AttributeError: 'Embedding' object has no attribute 'in_features'

bug

Status: Open.

#4166 In unslothai/unsloth;

· ferreroal opened on Mar 6, 2026

model.generate very slow

Status: Open.

#4165 In unslothai/unsloth;

· FilippoBoni1921 opened on Mar 6, 2026

[Bug] Qwen 3.5 - 27B (others?) - missing tensors on final build.

bug

Status: Open.

#4164 In unslothai/unsloth;

· David-AU-github opened on Mar 6, 2026

[Bug] TypeError: ModernBertForSequenceClassification.__init__() got an unexpected keyword argument 'num_labels'

bug

Status: Open.

#4163 In unslothai/unsloth;

· suntaiwu opened on Mar 6, 2026

[Bug] version 2026.3.1 has multiple GPU bug when doing CPT with Qwen3.5

bug

Status: Open.

#4162 In unslothai/unsloth;

· WY19940327 opened on Mar 6, 2026

[Bug] Qwen3.5 Packing leads to unstable grads

bug

Status: Open.

#4160 In unslothai/unsloth;

· kirawi opened on Mar 5, 2026

[Feature] Support newer torch 2.9.0 or above

feature requestFeature request pending on roadmapFeature request pending on roadmap

Status: Open.

#4159 In unslothai/unsloth;

· alexdong opened on Mar 4, 2026