HybridBackend
HybridBackend copied to clipboard
DeepRec-AI
A high-performance framework for training wide-and-deep recommender systems on heterogeneous cluster
# Current behavior when hb read some nested lists with ragged_rank > 1，the read Value cannot be transformed to SparseTensor by function hb.data.to_sparse. For example: dense_feature is one of the...
# System information - OS Platform: Ubuntu 18.04.5 LTS - Docker version: 18.09.5 - GCC version: 7.5.0 - Python version: 3.6.9 - TensorFlow/PyTorch version: tf1.15.5 # Willing to contribute Yes
# Current behavior In distributed mode, deeprec works fine when training on one hour of data, but hangs when training on one day or more. Log：  Nvidia-smi:  cpu:...
feature_column bucket_size is 6, use 8 gpus, then worker-5 and worker-6 'save/RestoreV2' failed; backtrace: Traceback (most recent call last): File "neg_feedback_multi.py", line 1252, in tf.app.run() File "/home/pai/lib/python3.6/site-packages/tensorflow_core/python/platform/app.py", line 40, in...
# Current behavior If there is only one worker ,training with EarlyStopping callback is ok. When multi workers with EarlyStopping callback doing distribute training, all workers will be hanging and...
# Current behavior I am using hybridBackend to do data parallelism, I create a dataset and make it an iterator, when I use hybridBackend scope to wrap the whole pipeline,...
# User Story The fixed-length features in TFRecord support configuration with default values（https://www.tensorflow.org/api_docs/python/tf/io/FixedLenFeature）, but currently, Parquet does not support this feature. If encountering a non-existent feature, an error will be...
Metadata
Owner
Metadata
A high-performance framework for training wide-and-deep recommender systems on heterogeneous cluster