Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
runanywhere
/
qwen3_vl_HNPU
like
1
Follow
RunAnywhere, Inc.
8
Image-Text-to-Text
hnpu
hexagon
npu
vlm
v75
v79
v81
License:
apache-2.0
Model card
Files
Files and versions
xet
Community
1
Copy to bucket
new
main
qwen3_vl_HNPU
10.2 GB
Ctrl+K
Ctrl+K
4 contributors
History:
38 commits
sanmonga22
v75 VLM: weight-share prefill+decode into llm_shared_512_w8.bin (2.57GB co-resident, fits app DSP ceiling); update manifest to 3 contexts
6658ada
verified
12 days ago
v75
v75 VLM: weight-share prefill+decode into llm_shared_512_w8.bin (2.57GB co-resident, fits app DSP ceiling); update manifest to 3 contexts
12 days ago
v79
Prune QHexRT HNPU bundle to runtime-minimum artifacts
13 days ago
v81
Prune QHexRT HNPU bundle to runtime-minimum artifacts
13 days ago
.gitattributes
Safe
1.57 kB
v75 (Galaxy S24 / SM8650, soc_model 57): Qwen3-VL-2B VLM (vision) bundle — fp16 vision tower + W8 shared decode (#1)
13 days ago
README.md
Safe
563 Bytes
v75: replace stale monolithic VLM bundle with device-validated decomposed VLM (vision+prefill+host weights+manifest)
12 days ago
