BladeDISC
BladeDISC copied to clipboard
BladeDISC is an end-to-end DynamIc Shape Compiler project for machine learning workloads.
We're now planing to start support quantization in BladeDISC. The basic idea is: - introduce a fake_quant op. The op will be used to pass quantization related info (e.g. scale/zero_point)...
wind up stuff
Also apply linter to the codes
The overall process can be divided into the following steps: - [x] make each subgraph executable, this allows us to collect the inputs of each subgraph when inference with the...
# Why we want to support `tf.feature_column`s We have found plenty of models with `tf.feature_column` ops in many industrial models, such as in CTR models. Normally the `tf.feature_column` ops are...
### Why TorchBench? 1. A benchmark maintained by the PyTorch community officially 2. `torchbenchmark/models` contains copies of popular or exemplary workloads 3. `torch.utils.benchmark.Timer` has priorities than `timeit.Timer` in PyTorch 4....