mediocreatbestMEnglish · 2 years ago[StackOverflow] If a PCI device is completely non-responsive, it's possible to completely remove the device and then re-scan it, hopefully re-initializing the device so it works again.plus-squareunix.stackexchange.comexternal-linkmessage-square0fedilinkarrow-up13arrow-down10
arrow-up13arrow-down1external-link[StackOverflow] If a PCI device is completely non-responsive, it's possible to completely remove the device and then re-scan it, hopefully re-initializing the device so it works again.plus-squareunix.stackexchange.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years ago[GitHub] bduggan/raku-jupyter-kernel allows you to run Raku (né Perl 6) within a Jupyter Notebook environment. In terms of onboarding, this seems to be one of the easiest ways to start using Raku.plus-squaregithub.comexternal-linkmessage-square0fedilinkarrow-up11arrow-down10
arrow-up11arrow-down1external-link[GitHub] bduggan/raku-jupyter-kernel allows you to run Raku (né Perl 6) within a Jupyter Notebook environment. In terms of onboarding, this seems to be one of the easiest ways to start using Raku.plus-squaregithub.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years ago[Paper] Optimizing Deep Learning Models For Raspberry Pi. Custom CNN (on MNIST data) performance from 114ms to 3.75ms. ResNet50 (on "flowers" data): from 1.1s to 1.0s (lowest) or 1.6s (highest).plus-squarearxiv.orgexternal-linkmessage-square0fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-link[Paper] Optimizing Deep Learning Models For Raspberry Pi. Custom CNN (on MNIST data) performance from 114ms to 3.75ms. ResNet50 (on "flowers" data): from 1.1s to 1.0s (lowest) or 1.6s (highest).plus-squarearxiv.orgmediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoTinyNeuralNetwork is a library to compress machine learning models through pruning, quantization, and more. Can also convert PyTorch models to TF Lite models.plus-squaregithub.comexternal-linkmessage-square0fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-linkTinyNeuralNetwork is a library to compress machine learning models through pruning, quantization, and more. Can also convert PyTorch models to TF Lite models.plus-squaregithub.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoOverview of machine learning frameworks that are supported on Raspberry Pi: OpenCV, TF Lite, Tencent ncnn, Tencent TNN, Alibaba MNN, Paddle Lite, ARMnn, MXNet + Gluon, PyTorch, and Caffe.plus-squareqengineering.euexternal-linkmessage-square0fedilinkarrow-up14arrow-down10
arrow-up14arrow-down1external-linkOverview of machine learning frameworks that are supported on Raspberry Pi: OpenCV, TF Lite, Tencent ncnn, Tencent TNN, Alibaba MNN, Paddle Lite, ARMnn, MXNet + Gluon, PyTorch, and Caffe.plus-squareqengineering.eumediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoArm NN is an optimized library of tensor operators for machine learning models to use. Support for TF Lite / ONNX models and Raspberry Pi 4 / armv7.plus-squaregithub.comexternal-linkmessage-square0fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-linkArm NN is an optimized library of tensor operators for machine learning models to use. Support for TF Lite / ONNX models and Raspberry Pi 4 / armv7.plus-squaregithub.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoTextSynth is a hosted service for generating text completions using language models. Free and paid tiers. Could be useful to play with LLMs without a strong computer (Pricing discussion in body text).plus-squaretextsynth.comexternal-linkmessage-square0fedilinkarrow-up11arrow-down11
arrow-up10arrow-down1external-linkTextSynth is a hosted service for generating text completions using language models. Free and paid tiers. Could be useful to play with LLMs without a strong computer (Pricing discussion in body text).plus-squaretextsynth.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoLaMini-LM is a collection of small language models that are accessible to run on local hardware without lots of resources. Models range from 250MB to 6.3GB.plus-squaregithub.comexternal-linkmessage-square0fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-linkLaMini-LM is a collection of small language models that are accessible to run on local hardware without lots of resources. Models range from 250MB to 6.3GB.plus-squaregithub.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agojncraton/languagemodels is a simple Python library for running LLMs locally. Supports instruction and embedding use cases. Chooses models according to available RAM.plus-squaregithub.comexternal-linkmessage-square0fedilinkarrow-up13arrow-down10
arrow-up13arrow-down1external-linkjncraton/languagemodels is a simple Python library for running LLMs locally. Supports instruction and embedding use cases. Chooses models according to available RAM.plus-squaregithub.commediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoAltoids tin for watercolor using sculpey modeling clay to create a custom tray for the paintsplus-squarewww.instructables.comexternal-linkmessage-square1fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-linkAltoids tin for watercolor using sculpey modeling clay to create a custom tray for the paintsplus-squarewww.instructables.commediocreatbestMEnglish · 2 years agomessage-square1fedilink
mediocreatbestMEnglish · 2 years agoTaming AI Bots: Prevent LLMs from entering "bad" states using continuous guidance from the LLM ("is this good? bad?") to avoid bad states.arxiv.orgexternal-linkmessage-square3fedilinkarrow-up15arrow-down10
arrow-up15arrow-down1external-linkTaming AI Bots: Prevent LLMs from entering "bad" states using continuous guidance from the LLM ("is this good? bad?") to avoid bad states.arxiv.orgmediocreatbestMEnglish · 2 years agomessage-square3fedilink
mediocreatbestMEnglish · 2 years ago"Prompt Gisting:" Train two models such that given inputs "Translate French<G1><G2>" and "<G1>G2>The cat," then G1 and G2 represent the entire instruction.plus-squarearxiv.orgexternal-linkmessage-square0fedilinkarrow-up12arrow-down10
arrow-up12arrow-down1external-link"Prompt Gisting:" Train two models such that given inputs "Translate French<G1><G2>" and "<G1>G2>The cat," then G1 and G2 represent the entire instruction.plus-squarearxiv.orgmediocreatbestMEnglish · 2 years agomessage-square0fedilink
mediocreatbestMEnglish · 2 years agoAn LLM prompt that is a special kind of summarizer for compressing an idea into as short a text ("tweet") as possible. Includes decompressor.plus-squareredd.itexternal-linkmessage-square0fedilinkarrow-up11arrow-down10
arrow-up11arrow-down1external-linkAn LLM prompt that is a special kind of summarizer for compressing an idea into as short a text ("tweet") as possible. Includes decompressor.plus-squareredd.itmediocreatbestMEnglish · 2 years agomessage-square0fedilink
