Downloads · 30 days
211
100% of all-time downloads
qwen3.5-4b-posture
4fhct4sd/qwen3.5-4b-posture
qwen3.5-4b-posture is a text generation model from 4fhct4sd. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.
Final phase 2 LoRA adapter merged into Qwen/Qwen3.5-4B, exported as 16-bit safetensors. The adapter continues phase 1 (scatter) with phase 2 (crystal) training.
All-time downloads
211
Public
Parameters
4.7B
9.3 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors9.3 GB · 100%
Parameter types
How the weights are stored.
BF164.7B · 100%
Try a prompt
Base models
Model card
From the Hugging Face model README
Qwen3.5-4B Posture — merged model
Final phase 2 LoRA adapter merged into Qwen/Qwen3.5-4B, exported as 16-bit safetensors. The adapter continues phase 1 (scatter) with phase 2 (crystal) training.
Adapters and resumable checkpoints: https://huggingface.co/4fhct4sd/qwen3.5-4b-posture-checkpoints
Use Transformers 5.5.0 for the tested inference runtime. The local Studio 5.3.0 sidecar produced invalid-probability CUDA assertions during multi-turn chat; 5.5.0 passed a five-turn test. This is limited validation, not a general quality or stability guarantee. The model may invent conversational turns; generated claims about itself are not verified facts.
No evaluation benchmark has been run. Training data is not included in this repository.
- Task
- Text Generation
- Library
- transformers
- Type
- qwen3_5
- License
- apache-2.0
- Created
- Sep 24, 2026
- Updated
- Sep 24, 2026