Skip to content

P
pytensor
提交 1fe82a1f authored 作者: Frédéric Bastien's avatar Frédéric Bastien
浏览文件
操作

Merge pull request #4533 from nouiz/fix_both_gpu_backend

Fix gh-4530, we can't have 2 opt with the same tag
上级 0e312ca3 86bb3e87
隐藏空白字符变更
内嵌 并排
正在显示 包含 1 行增加 和 1 行删除
theano/gpuarray/opt.py
浏览文件 @ 1fe82a1f
...@@ -975,7 +975,7 @@ def local_assert_no_cpu_op(node): ...@@ -975,7 +975,7 @@ def local_assert_no_cpu_op(node):
assert_no_cpu_op = theano.tensor.opt.in2out(local_assert_no_cpu_op, assert_no_cpu_op = theano.tensor.opt.in2out(local_assert_no_cpu_op,
name='assert_no_cpu_op') name='assert_no_cpu_op')
# 49.2 is after device specialization & fusion optimizations for last transfers # 49.2 is after device specialization & fusion optimizations for last transfers
optdb.register('assert_no_cpu_op', assert_no_cpu_op, 49.2) optdb.register('gpua_assert_no_cpu_op', assert_no_cpu_op, 49.2)
def tensor_to_gpu(x, context_name): def tensor_to_gpu(x, context_name):
......
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论