forked from jiuyuan/InfiniTensor
Compare commits
merge into: p32581460:master
p32581460:master
p32581460:cuda-attention
p32581460:ascend
p32581460:dev-leakyrelu
p32581460:instance_norm
p32581460:cuda-transpose
p32581460:kvcache_backup
p32581460:kvcache_attention_fp16
p32581460:kunlun_temp
p32581460:dist/graph
p32581460:dist_bench
p32581460:bang-softmax
p32581460:dropout
p32581460:update_pybind11
p32581460:support_fp16
p32581460:add_paddle_model
p32581460:point2point
p32581460:NNET_231111_from_master
p32581460:NNET_231111
p32581460:allocator_memPool
p32581460:test_codegen
p32581460:change_path
p32581460:xpu_allreduce
p32581460:dev-dynamic-graph
p32581460:dev-dynamic-graph-allocator
p32581460:dump/init
p32581460:gpt
p32581460:nnet_e2e_for_merge
p32581460:gencode
p32581460:optimization-pass
p32581460:dev-memory
p32581460:conv_half
p32581460:benchmark_conv
p32581460:benchmark_softmax
p32581460:benchmark
p32581460:dcj/for_multiple_datatype
p32581460:Conv_NHWC
p32581460:NNET_bias
p32581460:NNET_bias_0630
p32581460:constroy/doc_on_ares
p32581460:pure_engine
p32581460:v0630
p32581460:NNET_e2e
p32581460:update_doc
p32581460:model_test
p32581460:TC_revision
p32581460:NNET_gcn
p32581460:NNET_op_test
p32581460:NNET_OpSearch
p32581460:NNET_gcn_fuse
p32581460:fsrcnn-conv-bias-act-fuse
p32581460:NNET_e2e_for_merge
p32581460:NNET_eliminateOP
p32581460:NNET_anyOp
p32581460:NNET_transpose
p32581460:cpu_backend2
p32581460:NNET_e2e_fix
p32581460:NNET_GAN
p32581460:test_onnx
p32581460:activation
p32581460:ddp
p32581460:search_engine
p32581460:power-fusion
p32581460:op_timer
p32581460:graph-onnx
p32581460:graphFactory
p32581460:case-fsrcnn
p32581460:testAccuracy
p32581460:train_wanghailu_1010
p32581460:broadcast_wanghailu_0916
jiuyuan:test-models
pull from: p32581460:master
p32581460:cuda-attention
p32581460:ascend
p32581460:dev-leakyrelu
p32581460:instance_norm
p32581460:cuda-transpose
p32581460:master
p32581460:kvcache_backup
p32581460:kvcache_attention_fp16
p32581460:kunlun_temp
p32581460:dist/graph
p32581460:dist_bench
p32581460:bang-softmax
p32581460:dropout
p32581460:update_pybind11
p32581460:support_fp16
p32581460:add_paddle_model
p32581460:point2point
p32581460:NNET_231111_from_master
p32581460:NNET_231111
p32581460:allocator_memPool
p32581460:test_codegen
p32581460:change_path
p32581460:xpu_allreduce
p32581460:dev-dynamic-graph
p32581460:dev-dynamic-graph-allocator
p32581460:dump/init
p32581460:gpt
p32581460:nnet_e2e_for_merge
p32581460:gencode
p32581460:optimization-pass
p32581460:dev-memory
p32581460:conv_half
p32581460:benchmark_conv
p32581460:benchmark_softmax
p32581460:benchmark
p32581460:dcj/for_multiple_datatype
p32581460:Conv_NHWC
p32581460:NNET_bias
p32581460:NNET_bias_0630
p32581460:constroy/doc_on_ares
p32581460:pure_engine
p32581460:v0630
p32581460:NNET_e2e
p32581460:update_doc
p32581460:model_test
p32581460:TC_revision
p32581460:NNET_gcn
p32581460:NNET_op_test
p32581460:NNET_OpSearch
p32581460:NNET_gcn_fuse
p32581460:fsrcnn-conv-bias-act-fuse
p32581460:NNET_e2e_for_merge
p32581460:NNET_eliminateOP
p32581460:NNET_anyOp
p32581460:NNET_transpose
p32581460:cpu_backend2
p32581460:NNET_e2e_fix
p32581460:NNET_GAN
p32581460:test_onnx
p32581460:activation
p32581460:ddp
p32581460:search_engine
p32581460:power-fusion
p32581460:op_timer
p32581460:graph-onnx
p32581460:graphFactory
p32581460:case-fsrcnn
p32581460:testAccuracy
p32581460:train_wanghailu_1010
p32581460:broadcast_wanghailu_0916
jiuyuan:code_generate
jiuyuan:support_ascend_fp16_zyz
jiuyuan:aug-cuda-op-need
jiuyuan:master
jiuyuan:fix_runtime
jiuyuan:support_ascend_fp16
jiuyuan:cxjj
jiuyuan:fix_fp16_matmul
jiuyuan:bang-rmsnorm
jiuyuan:cuda-attention
jiuyuan:operator-test
jiuyuan:bang-rms-soft
jiuyuan:cuda-transpose
jiuyuan:add_leaky_relu
jiuyuan:instance_norm
jiuyuan:kvcache_backup
jiuyuan:kvcache_attention_fp16
jiuyuan:kunlun_temp
jiuyuan:dist/graph
jiuyuan:dist_bench
jiuyuan:dropout
jiuyuan:update_pybind11
jiuyuan:support_fp16
jiuyuan:add_paddle_model
jiuyuan:point2point
jiuyuan:NNET_231111_from_master
jiuyuan:NNET_231111
jiuyuan:allocator_memPool
jiuyuan:test_codegen
jiuyuan:change_path
jiuyuan:xpu_allreduce
jiuyuan:dev-dynamic-graph
jiuyuan:dev-dynamic-graph-allocator
jiuyuan:dump/init
jiuyuan:gpt
jiuyuan:nnet_e2e_for_merge
jiuyuan:gencode
jiuyuan:optimization-pass
jiuyuan:dev-memory
jiuyuan:conv_half
jiuyuan:benchmark_conv
jiuyuan:benchmark_softmax
jiuyuan:benchmark
jiuyuan:dcj/for_multiple_datatype
jiuyuan:Conv_NHWC
jiuyuan:NNET_bias
jiuyuan:NNET_bias_0630
jiuyuan:constroy/doc_on_ares
jiuyuan:pure_engine
jiuyuan:v0630
jiuyuan:NNET_e2e
jiuyuan:update_doc
jiuyuan:model_test
jiuyuan:TC_revision
jiuyuan:NNET_gcn
jiuyuan:NNET_op_test
jiuyuan:NNET_OpSearch
jiuyuan:NNET_gcn_fuse
jiuyuan:fsrcnn-conv-bias-act-fuse
jiuyuan:NNET_e2e_for_merge
jiuyuan:NNET_eliminateOP
jiuyuan:NNET_anyOp
jiuyuan:NNET_transpose
jiuyuan:cpu_backend2
jiuyuan:NNET_e2e_fix
jiuyuan:NNET_GAN
jiuyuan:test_onnx
jiuyuan:activation
jiuyuan:ddp
jiuyuan:search_engine
jiuyuan:power-fusion
jiuyuan:op_timer
jiuyuan:graph-onnx
jiuyuan:graphFactory
jiuyuan:case-fsrcnn
jiuyuan:testAccuracy
jiuyuan:train_wanghailu_1010
jiuyuan:broadcast_wanghailu_0916
jiuyuan:test-models
These branches are equal. There is no need to create a pull request.