How to use from
Unsloth StudioInstall Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for MakiAi/qwen35-4b-codex-mobile-colab-t4-lora to start chattingUsing HuggingFace Spaces for Unsloth
# No setup required# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for MakiAi/qwen35-4b-codex-mobile-colab-t4-lora to start chattingLoad model with FastModel
pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
model_name="MakiAi/qwen35-4b-codex-mobile-colab-t4-lora",
max_seq_length=2048,
)Quick Links
Qwen3.5-4B Codex Mobile Colab T4 LoRA
Small experimental LoRA adapter trained with Unsloth on Google Colab T4 via Google Colab CLI.
Base model
unsloth/Qwen3.5-4B
Experiment
- Session:
unsloth-qwen35-4b-t4 - GPU observed:
Tesla T4, 15360 MiB, 14910 MiB - Training examples:
20 - Max sequence length:
1024 - Max steps:
20 - LoRA rank:
16 - Seed:
3407 - Output folder in Colab Drive:
/content/drive/MyDrive/colab-cli-unsloth-qwen35-4b
Intended behavior
Tiny Japanese experiment-report style adapter. It nudges answers toward concise conclusions, observations, next actions, Drive-mounted saves, evidence, and reproducibility.
See comparison.md and comparison.json for before/after response checks.
- Downloads last month
- 28
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Install Unsloth Studio (macOS, Linux, WSL)
# Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for MakiAi/qwen35-4b-codex-mobile-colab-t4-lora to start chatting