You can deploy this model and start earning money today!

namhokaist/
appgen-qwen3-vl-8b-sft-ngc-only-M-clean-lr2p5e7-gbs32-1ep

Name: appgen-qwen3-vl-8b-sft-ngc-only-M-clean-lr2p5e7-gbs32-1ep
Availability: InStock

Qwen/Qwen3-VL-8B-Instruct

finetune

apache-2.0

endpoints_compatible

Views

No views yet

AppGen Qwen3-VL NGC-only arm M

Frozen-vision supervised fine-tuning on only luca0621/appgen-sft-ngc-v1 (ngc-f-qwen3-normalized, pinned revision 769ea99dbc4ff190048ae0db37eb6310dba595e0).

Final checkpoint: 112
Dataset variant: clean
Dataset SHA-256: af678b5cb244b6ca21a232180f201e188afbd5369eda1570b4a06cfc769cedee
Learning rate: 2.5e-07
Global batch size: 32
Vision encoder and aligner frozen: True
System prompt SHA-256: 67ff8adb0e78a617f3d0edcf196d4e4cc3239967a8c30619fc5afc14484ee8c0

Exact training, data-composition, overlap, and numerical-verification receipts are included.

namhokaist/appgen-qwen3-vl-8b-sft-ngc-only-M-clean-lr2p5e7-gbs32-1ep

AppGen Qwen3-VL NGC-only arm M

namhokaist/
appgen-qwen3-vl-8b-sft-ngc-only-M-clean-lr2p5e7-gbs32-1ep