Lance Norskog
Lance Norskog
Lucene is another database you might be interested in. It is the major open source text search engine, and has a modular "codec" plugin design for the actual key-value storage...
Is the Dockerization project finished, or in a functioning state?
@vsoch @robinsloan As Theano is an end-of-life technology base, I found a VAE written in Keras and ported it to Google Colab. This one simplifies the problem by limiting it...
Used this code block at the beginning: !pip uninstall -y tensorflow keras !pip install tensorflow==2.9.0 !pip install keras==2.9.0 Cheers!
If this is a dead project, could you mark it as such instead of wasting people's time?
@lzfelix To do top-k, would this need a beam search implementation? Where the beam search takes transitional probabilities into account? I've seen beam search implementations with bigram models that do...
@tapos12 : argsort gives indices of a sorted array >>> import numpy as np >>> a = [-1, -2, 0, 4, 2] >>> print(np.argsort(a)) [1 0 2 4 3]
This is probably good enough.
@tagomaris- another possibility: the 'forest' plugin could implement one disk-based buffer for all events. When any "child" blocks, forest blocks all events to all children.
You can convert between channels-first and channels-last with an n-dimensional matrix transpose. The Keras 'Permute' layer does exactly this: https://www.tensorflow.org/api_docs/python/tf/keras/layers/Permute