filmov
tv
OpendTect Machine Learning Developers Q&A Webinar: how can I add my own trained model and other Q+A
Показать описание
0:00 Introduction
3:17 How can I add my own trained model to the OpendTect database?
9:49 Upcoming Machine Learning Plugin improvements
15:59 Will this new pre-trained models facility be available in the next OpendTect public release?
17:22 How can I read a training example file in OpendTect?
28:26 What are the developments around SynthRock linked with the Machine Learning plugin?
32:10 Can pseudo wells be exported as .las files?
34:56 Can one also import a trained model with mlio?
42:02 Can I use data augmentation as quickly in Keras TensorFlow?
48:09 How can I write my own new Tensorflow - Keras model architecture code inside OpendTect environment and make it available in the Machine Learning plugin UI?
55:47 The Python Model scripts filename need to start with mlmodel_
57:30 What is the status of pytorch development?
58:28 What are the resources one can refer to when having questions about working with the Machine Learning Plugin and/or have suggestions for improvements?
59:28 - What is the easiest way to become part of the OpendTect Machine Learning Community on Discord?
1:00:21 If someone has an idea for a webinar that shows how their company is using the Machine Learning development environment as a success story, or has a model that they want to make publicly available, how can they make that happen?
1:04:31 Discord OpendTect Machine Learning Developers Community
1:05:47 Outro
Keywords: OpendTect, Machine Learning, Python, Spyder, Webinar, Development, Tensorflow, Keras, Scikit-learn, Trained models, Pre-trained models, Discord, Seismic, F3 Demo, Seismic Classification, Salt, Limestone, Cloud, On-premise, UNet, odpy, dgbpy, mlio, mlapply, datasets, input shape, output shape, training data, seismic image transformation, workflow, SynthRock, hdf5, Rock Property Prediction, Well data, Well logs, Pseudo wells, .las files, Nvidia, CUDA, GPU, Data augmentation, 2D images, 3D images, Segmentation, Pytorch, Distribute models