Downloads · 30 days
0
qwen2vl-mermaid-lora
passawee-kmitl/qwen2vl-mermaid-lora
qwen2vl-mermaid-lora is a machine learning model from passawee-kmitl. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft. The card lists the license as apache-2.0.
LoRA adapters that convert flowchart images into Mermaid flowchart TD code.
Access
Public
Updated Sep 14, 2026
Repo size
252 MB
Likes
2
Public
Repo makeup
Click a slice to open those files.
.safetensors222 MB · 84%
At a glance
- Library
- peft
- License
- apache-2.0
- Access
- Public
- Created
- Sep 12, 2026
- Updated
- Sep 14, 2026
- SHA
- 25876028
Base models
Model card
From the Hugging Face model README
Image-to-Mermaid QLoRA adapters (Qwen/Qwen2-VL-2B-Instruct)
LoRA adapters that convert flowchart images into Mermaid flowchart TD code.
adapter_final/-- v1, block-format targets, pad-only loss maskadapter_final_v2/-- v2, block-format targets, assistant-only loss maskadapter_final_v3/-- v3, inline-format targets, assistant-only loss maskdataset.zip-- 270 synthetic samples across simple/complex/messy tiersresults_*.json-- raw predictions for every arm, so metrics can be recomputed on CPU
Trained on a single Colab T4: 4-bit NF4 quantisation, fp16, LoRA r=16, 3 epochs.
- Library
- peft
- License
- apache-2.0
- Created
- Sep 12, 2026
- Updated
- Sep 14, 2026