zhijiang 4dc4470cc7 Fix fusion for two LayerNorm sharing same input but with different weights (#15919) ... in gpt_j_residual(https://arxiv.org/pdf/2204.06745.pdf), there are 2 LN nodes will share one same input, and ORT does CSE graph optimization before LN fusion, which will modify the LN graph pattern and thus make LN fusion failure. 		2023-05-22 08:26:36 +08:00
..
approximation
computation_reduction
cse
fusion	Fix fusion for two LayerNorm sharing same input but with different weights (#15919)	2023-05-22 08:26:36 +08:00
gemm_activation_fusion
matmul_add_fusion
model_parallel
propagate_cast
recompute
runtime_optimization
abs-2id-max.onnx
abs-id-max.onnx
abs-id.onnx
cast_elimination.onnx
cast_elimination.py
computation_reduction.py
computation_reduction_transformer.onnx
concat_graph_gen.py
concat_slice_basic_test.onnx
concat_slice_elimination.py
concat_trainable.onnx
constant-subgraph.onnx
dropout.onnx
dropout_ratio.onnx
dropout_zeroratio_elimination.py
expand_elimination.onnx
expand_elimination.py
fp16model_loop.onnx
id-elim.onnx
id-elim.py
id-scan9_sum.py
insert_cast_twice.onnx
model_creation_for_testing.ipynb
multinomial_float16.onnx
noop-add.onnx
noop-add.py
qdq_conv.onnx
qdq_conv_gen.py
qdq_with_multi_consumer_dq_nodes.fixed.onnx
qdq_with_multi_consumer_dq_nodes.fixed.txt
random_normal_like_float16.onnx
scalar_const_not_share.onnx
scalar_const_not_share.py
scan9_sum.onnx
shape-add.onnx
slice-v1-elim.onnx
slice-v11-elim.onnx
triple-cast.onnx