Hugging Face's logo 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

litert-community
/
TIPSv2-B14-DPT-LiteRT

Depth Estimation
LiteRT
LiteRT
android
on-device
gpu
tipsv2
vit
dpt
surface-normals
semantic-segmentation
ade20k
Model card Files Files and versions
xet
Community

Instructions to use litert-community/TIPSv2-B14-DPT-LiteRT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

  • Libraries
  • LiteRT

    How to use litert-community/TIPSv2-B14-DPT-LiteRT with LiteRT:

    # No code snippets available yet for this library.
    
    # To use this model, check the repository files and the library's documentation.
    
    # Want to help? PRs adding snippets are welcome at:
    # https://github.com/huggingface/huggingface.js
  • Notebooks
  • Google Colab
  • Kaggle
TIPSv2-B14-DPT-LiteRT
637 MB
Ctrl+K
Ctrl+K
  • 1 contributor
History: 13 commits
mlboydaisuke's picture
mlboydaisuke
NPU section: cause of the NPU loss found (tanh-GELU decomposition); erf-GELU variant flips it
3554546 verified 1 day ago
  • .gitattributes
    1.56 kB
    hero 7 days ago
  • README.md
    8.04 kB
    NPU section: cause of the NPU loss found (tanh-GELU decomposition); erf-GELU variant flips it 1 day ago
  • build_tipsv2.py
    20.2 kB
    range fold: bias carries absolute lambda (bit-exact fp32) 7 days ago
  • hero.png
    540 kB
    xet
    hero: indoor scene (two chairs) — the NYU/ADE20K heads' home domain 7 days ago
  • tipsv2_b14_dpt_erf_fp16.tflite
    318 MB
    xet
    Add erf-GELU variant: 2.30x faster on Hexagon NPU (142.0 ms vs 326.9 ms), outputs corr 0.99998+ 1 day ago
  • tipsv2_b14_dpt_fp16.tflite
    318 MB
    xet
    fp16 model: corrected depth-head bias fold 7 days ago