Release 1.3.10
Compatible with Keras !
- NEWS
- TensorLayer is now compatible with Keras, see tutorial_keras.py
tl.layers.get_layers_with_name
: get a list of tensor outputs from a network
- Update
tl.cost.sigmoid_cross_entropy
for TF0.11tl.file.load_and_assign_npz
- log of
tl.file.save_npz
EmbeddingAttentionSeq2seqWrapper
for TF1.0 by @pnpnpntl.nlpVocabulary
withpad_id
- Deprecate
tl.prepro.distorted_images
andtl.prepro.crop_central_whiten_images
, please usetf.map_fn
ortl.prepro.threading_data
instead
- Documentation
This version works well under TF 0.10, 0.11, 0.12 and 1.0 backend, feel free to report if you found a bug.
Many thanks!