forum
Stuck at download [unsloth studio], Running on colab with T4 15gib vram
Open
Fine-tuned model never predicts EOT token
Open
Qwen3.5 + fast_inference=True (vLLM) — supported yet?
Open
[Windows 11] [Unsloth Studio] keeps trying to download an already downloaded model
Open
Small typing mistakes
Open
MiniMax models can't do websearch in Unsloth Studio
Open
Stuck on downloading model
Open
Subject: Crash Report: Training process exited unexpectedly with Gemma-4 E4B on free Colab (Unsloth
Dependencies not there for merge export
Unsloth studio doesnt retain my recipes after shutting down
Agent style support in unsloth studio
Too Many Requests for url
Installed gemma-4-26B-A4B-it-GGUF in hyperlink wont work
Failing to merge weights into 16-bit
Open
Add AdamW Beta1 + Beta2 to Studio
Gemma4 Log: `<|tool_response>' was not control-type; this is probably a bug in the model`
unsloth on akashml (gpu renting)
expected scalar type Float but found BFloat16
Open
Studio not running on export or Load model in chat
MLX Support
Qwen3-Coder-Next-GGUF:UD-Q3_K_M does not support tools
Open
Load model with lower context length in Unsloth Studio
Gemma4 training
Open
Gemma4 works on only one GPU
GPT-OSS-20b with Flex Attention
Unsloth Studio not working for Qwen3.5?
Open
Llama.cpp server error after update: "Failed to load model" & HTTP Error 422 during install
Ridiculous levels of loss (Gemma 4, colab training)
Solved
How can I disable thinking via curl when send request to unsloth/Qwen3.5-27B-GGUF?
Nemotron 3 nano 30b a3b bf16 support for fast inference using vllm and weight sharing for doing GRPO
Open
How to connect unsloth studio recipies with openweb UI
Unsloth + Docker + Windows
Open
Exporting Nemotron-3-nano-4B LoRa as GGUF failure
http://localhost:8888/ shows nothing
Can't find LM Studio downloaded models?
Unsloth Studio detects my GPU (24GB Vram), but only uses CPU/RAM?
Qwen3-30B-A3B: inconsistent VRAM usage and very different generation latency across two environments
Open
how to add unlsoth sutdio exprted models to ollama
Any ideas why I can't make a gguf file for 3.5-35B
Error loading checkpoint: '-'
unsloth studio failed to start
Solved
Found a ui glitch in unsloth studio:
Installing unsloth studio on linux keeps crashing
when training: Failed to import ML libraries: No module named 'unsloth_studio.losses'
Open
cant upload own dataset file?
Solved
Trying to train and unsloth studio :One of the subprocesses has abruptly died during map operation.
~8 tok/s on Pixel 8 following exact Colab notebook — expected ~40 tok/s
Got this error installing in windows
Long wait for simple Hi?
Open
Issue loading CohereLabs/tiny-aya-global
Open