Highlights
Pinned Loading
-
ExperimentalMachines/openweights
ExperimentalMachines/openweights PublicRun open-weight models from Hugging Face on your Android phone. No account, no cloud, no telemetry. Native Kotlin and Compose, inference by llama.cpp & ExecuTorch
-
ExperimentalMachines/execupack
ExperimentalMachines/execupack PublicExports small dense open-weight LLMs (≤4B) from Hugging Face to ExecuTorch .pte files (XNNPACK, Vulkan, QNN, MediaTek) on CI, for the openweights Android app
Python
-
ExperimentalMachines/execubench
ExperimentalMachines/execubench PublicBenchmarks for compiled ExecuTorch models on real phones in AWS Device Farm: speed, memory, heat and answer quality, single and multi turn, every number traceable to a raw device log
Python
-
ExperimentalMachines/execuserve
ExperimentalMachines/execuserve PublicServe ExecuTorch .pte models from an Android phone: an OpenAI- and Anthropic-compatible HTTP server that runs in the background, with multi-model hosting, KV-cache reuse and a built-in browser chat…
Kotlin
-
penge-passport-ph
penge-passport-ph PublicRead-only, rate-limited checker for DFA Philippine passport appointment availability, for Node.js and Python, and a free alerts website.
TypeScript 3
If the problem persists, check the GitHub status page or contact support.




