Jianyu Huang 5c67cc7a9e [caffe2] Enable fp16 for SparseNormalize op (#45551) ... Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/45551 The FP16 version of SparseNormalize op in Caffe2 is missing. This Diff adds FP16 support to unblock MC process of adding FP16 to Dper3. Check https://fb.quip.com/L0T2AXGwUY3n#EReACAeifk3 . One question is whether the pure FP16 Sparse Normalized op will affect the accuracy? Maybe we should do it in FP32 domain. ghstack-source-id: 114184398 Test Plan: ``` buck run mode/opt //caffe2/caffe2/python/operator_test:sparse_normalize_test ``` ``` buck run mode/opt -c python.package_style=inplace mode/no-gpu //caffe2/caffe2/python/benchmarks:sparse_normalize_benchmark -- --fp16 ``` Reviewed By: jspark1105 Differential Revision: D24005618 fbshipit-source-id: 8b918ec4063fdaafa444779b95206ba2b7b38537		2020-10-13 15:35:22 -07:00
..
fused_rowwise_nbit_conversion_bench.py	Remove __future__ imports for legacy Python2 supports (#45033)	2020-09-23 17:57:02 -07:00
sparse_lengths_sum_nbit_benchmark.py	Remove __future__ imports for legacy Python2 supports (#45033)	2020-09-23 17:57:02 -07:00
sparse_normalize_benchmark.py	[caffe2] Enable fp16 for SparseNormalize op (#45551)	2020-10-13 15:35:22 -07:00