未验证 提交 8af0d85e 编写于 作者: A Aurelius84 提交者: GitHub

fix unittest failed on windows GPU (#29072)

上级 dfaf6b5e
......@@ -410,7 +410,7 @@ test_parallel_executor_transformer^|test_parallel_executor_transformer_auto_grow
test_fuse_bn_add_act_pass^|test_activation_mkldnn_op^|test_tsm^|test_gru_rnn_op^|test_rnn_op^|test_simple_rnn_op^|test_pass_builder^|test_lstm_cudnn_op^|test_inplace_addto_strategy^|^
test_ir_inplace_pass^|test_ir_memory_optimize_pass^|test_memory_reuse_exclude_feed_var^|test_mix_precision_all_reduce_fuse^|test_parallel_executor_pg^|test_print_op^|test_py_func_op^|^
test_weight_decay^|test_mobile_net^|test_graph^|test_imperative_out_scale^|test_imperative_qat^|test_imperative_qat_channelwise^|test_moving_average_abs_max_scale_op^|^
test_quantization_pass^|test_quantization_scale_pass^|test_user_defined_quantization^|test_matmul_v2_op^|test_sentiment^|test_conv2d_int8_mkldnn_op^|^
test_quantization_pass^|test_quantization_scale_pass^|test_user_defined_quantization^|test_matmul_v2_op^|test_conv2d_int8_mkldnn_op^|^
test_crypto^|test_callbacks^|test_program_prune_backward^|test_train_recognize_digits^|test_imperative_ocr_attention_model
rem /*===============================================================*/
......
......@@ -275,7 +275,7 @@ class Args(object):
lr = 0.01
vocab_size = 1000
padding_size = 50
log_step = 2
log_step = 5
train_step = 10
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册