TensorBoard is the best visualization for training model. It can working offline. Most clusters aren't have online access.
TensorBoard is the best visualization for training model. It can working offline. Most clusters aren't have online access.
Конвейер машинного обучения для классификации рукописных цифр MNIST с использованием TensorFlow Extended (TFX)
В рамках данной статьи будет рассмотрено построение конвейера машинного обучения для классификации рукописных цифр из базы данных MNIST с использованием фреймворка TensorFlow и TFX, а также мониторинг процесса обучения с помощью TensorBoard и выпуск модели с автоматической генерацией к ней API с помощью TensorFlow Model Server. Основная цель данной статьи заключается в восполнении информационного пробела по теме использования TFX в контексте решения задачи классификации рукописных цифр MNIST, которая уже была представлена TensorFlow в виде одноименного репозитория tfx/tfx/examples/mnist .
https://habr.com/ru/articles/898140/
#python #tensorflow #machine_learning #конвейеры #tfx #tensorflowmodelserver #tensorboard
Once you dig deep enough, there's something really addictive/exciting about seeing the loss curve dipping lower than the last time.
Give me your tasty predictions, machine!
#ml #tensorboard
On my blog: Training Models locally via Containers
#Tensorboard #FineTuning #LLM #AI
https://heidloff.net/article/training-models-locally-containers-tensorboard/
I was surprised to learn that you cannot do ROC curves on #tensorboard directly... but after banging my head against the keyboard, I've found that you can send #matplotlib grahps to it.
(And also torchmetrics gives you nice matplotlib graphs for your metrics.)
RT @stackdpodcast: @dhinojosa and @kito99 @frankgreco and @zsevarac. Visual Recognition ML API for #Java, #DeepNets, #LLMs, Stack #OverflowAI, #StableDiffusion, #ChatGPT, #OpenJDK Panama, #Llma 2, #PromptEngineering, #ApacheZeppelin, #Tensorboard, #H2O.ai, #Spark, and more! https://www.pubhouse.net/2023/10/stackd-67-ai-nullpointers.html
RT @stackdpodcast: #67: @dhinojosa and @kito99 dive into #ai with fellow @Java_Champions @frankgreco and @zsevarac: Visual Recognition ML API for Java, #DeepNets, #StableDiffusion, #ChatGPT, #OpenJDK Panama, #Llma 2, #PromptEngineering, #ApacheZeppelin, #Tensorboard, #H2O.ai, #Spark, and more! https://www.pubhouse.net/2023/10/stackd-67-ai-nullpointers.html
New optimization run in progress!
Will the new run, with more hidden neurons, more stupid bugs, be able to over-take the old one?
Find out tomorrow, I guess. Or wait, let's just check once more... #tensorboard addiction.
Hello, world!
I'm currently working on #tools for #ML engineers at #Meta [#mlops, #pytorch, #tensorboard, #debugging], tend to write for developers at explog.in [#devtools, #python] and spend way too much time configuring #emacs. Past lives have involved #android, #maps and #reliability.
[wow, edits; so many hashtags] #introduction
tbparse: Load tensorboard event logs as pandas DataFrames for scientific plotting
- Supports parsing tensorboard event scalars, tensors, histograms, images, audio, hparams, and text.
- Supports event generated by PyTorch, Tensorboard/Keras, and TensorboardX.
- Allows parsing multiple tensorboard event files in a hierarchical directory structure.
https://github.com/j3soon/tbparse
#Python #TensorBoard #Pandas #PyTorch #TensorFlow #MachineLearning #DeepLearning #DataScience