PerceptiLabs
@perceptilabs
A developer tool for applied #deeplearning π Build, train, analyze and deploy models faster and easier than ever. π
ID: 927902541622497280
http://www.perceptilabs.com 07-11-2017 14:16:20
660 Tweet
759 Followers
314 Following
.PerceptiLabs is #MLOps made easy π #Lowcode is the future, check it out: β Versioning π Load and pre-process data β‘ Modeling π Train & interpret π Evaluate & benchmark π Deploy
You asked how to deploy, use, & test a model that has been exported by PerceptiLabs. So, we added a Gradio deployment target. This blog discusses the rationale behind it & how you can use it to test your model. #DeepLearning #MachineLearning #MLOps blog.perceptilabs.com/integrations-gβ¦
Together with our community, we are gathering public datasets and State-of-the-Art models for everyone to use & easily apply in PerceptiLabs π We've integrated Hugging Face, Kaggle, TensorFlow, Papers with Code since they are all amazing sources π
.PerceptiLabs now supports different data storage structures. #ImageClassification supports subdirectories of images, #SegmentImages supports subdirectories for images & masks, & #MultiModel supports images in relative paths defined in your CSV file.π«
Join us for our next #LiveCoding event TOMORROW at 9AM PST! In this live coding event our CTO + Co-Founder, Robert Lundberg , will build a Multi-Modal Regression model that can predict the price of houses using both meta data & images. Join here: youtu.be/Qy9qxAMFnzc
Export your model as a pure TensorFlow model or export our different bundles including scripts to easily deploy on your custom infrastructure π More options will be added to the public version soon π #ML #MachineLearning #TensorFlow
PerceptiLabs calculates Precision (aka positive predictive value), Recall (aka sensitivity), and F1 Score for you (aka F-score or F-measure). You can see all three updated in real-time during training, & the first two in our Evaluate View.ππΌ Read more: docs.perceptilabs.com/perceptilabs/aβ¦
Photos go in, reality comes outβ¦ and fast! π See you how Two Minute Papers turns a few photos into video quickly & with no AI. youtube.com/watch?v=yptwRRβ¦
Need to quickly benchmark your model? Use PerceptiLabs' Overview screen to rapidly create multiple models for comparison from the same dataset. Then modify each model & use the Training & Evaluate views to make comparisons. Learn more: docs.perceptilabs.com/perceptilabs
Together with our community, we are gathering public datasets and State-of-the-Art models for everyone to use & easily apply in PerceptiLabs π We've integrated Hugging Face, Kaggle, TensorFlow, Papers with Code since they are all amazing sources π
PerceptiLabs The fastest way to build, debug, and interpret neural networks. perceptilabs.com