Downloads · 30 days
0
TinyBrowser-GNM
mariocde/TinyBrowser-GNM
TinyBrowser-GNM is a machine learning model from mariocde. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Lightweight CNN for on-device browser action prediction. Takes V3 semantic grid [1,32,64,8] + goal vector [1,64] → heatmap + 13 action classes.
Access
Public
Updated Mar 24, 2026
Repo size
17.6 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pth17.4 MB · 99%
At a glance
- Access
- Public
- Created
- Mar 24, 2026
- Updated
- Mar 24, 2026
- SHA
- 6fe002bc
Model card
From the Hugging Face model README
TinyBrowser GNM (Goal Navigation Model)
Lightweight CNN for on-device browser action prediction. Takes V3 semantic grid [1,32,64,8] + goal vector [1,64] → heatmap + 13 action classes.
Architecture
- Input: V3 Frame (uint8) + Goal embedding (float32)
- Output: Heatmap [1,1,32,64] (WHERE) + Action logits [1,13] (WHAT)
- Inference: <0.5ms on Hexagon HTP, <5ms on CPU
Files
gnm_best.pth— PyTorch checkpointgnm_fp32.onnx— ONNX for ONNX Runtime (QNN EP / NNAPI / CPU)
- Created
- Mar 24, 2026
- Updated
- Mar 24, 2026