Commit ce915c5
authored
Update Paddle submodule to latest develop (PaddlePaddle#2273)
Co-authored-by: tianshuo78520a <[email protected]>1 parent 6730b0e commit ce915c5
1 file changed
+1
-1
lines changed- ci/rules/no-int32-type-dims.yml+4-4
- cmake/external/magma.cmake+24-10
- cmake/external/warprnnt.cmake+2-3
- cmake/external/xpu.cmake+1-1
- paddle/cinn/backends/compiler.cc+2-4
- paddle/cinn/backends/compiler.h+5-6
- paddle/cinn/backends/llvm/execution_engine.cc+5-4
- paddle/cinn/backends/llvm/execution_engine.h+2-2
- paddle/cinn/hlir/framework/pir/compilation_cache.h+5-5
- paddle/cinn/hlir/framework/pir/compilation_task.cc+4-4
- paddle/cinn/hlir/framework/pir/compilation_task.h+3
- paddle/cinn/hlir/framework/pir_compiler.cc+12-11
- paddle/cinn/runtime/flags.cc+1-1
- paddle/fluid/framework/fleet/fleet_wrapper.cc+3-3
- paddle/fluid/framework/fleet/nccl_wrapper.cc+2-1
- paddle/fluid/framework/ir/cutlass_teller.h+39-24
- paddle/fluid/framework/ir/onednn/compute_propagate_scales_onednn_pass.cc+4-1
- paddle/fluid/framework/ir/xpu/quant_utils.cc+3
- paddle/fluid/inference/tensorrt/convert/conv3d_op.cc+20-5
- paddle/fluid/inference/tensorrt/convert/deformable_conv_op.cc+20-5
- paddle/fluid/inference/tensorrt/convert/dequantize_linear_op.cc+5-3
- paddle/fluid/inference/tensorrt/convert/quantize_linear_op.cc+5-3
- paddle/fluid/inference/tensorrt/convert/sparse_fc_op.cc+10-9
- paddle/fluid/operators/collective/send_v2_op.cu.cc+1-1
- paddle/fluid/prim/api/composite_backward/composite_backward_api.h+15-4
- paddle/fluid/primitive/decomp_rule/decomp_rule/composite.h+2-2
- paddle/fluid/primitive/decomp_rule/decomp_vjp/details.h+8-2
- paddle/fluid/pybind/cuda_streams_py.cc+27
- paddle/phi/backends/onednn/onednn_reuse.h+4-1
- paddle/phi/kernels/CMakeLists.txt+5-1
- paddle/phi/kernels/cpu/grid_sample_utils.h+148-37
- paddle/phi/kernels/cpu/rnn_functor.h+10-9
- paddle/phi/kernels/cpu/rnn_kernel.cc+2-1
- paddle/phi/kernels/cpu/send_u_recv_grad_kernel.cc+4-1
- paddle/phi/kernels/cpu/send_u_recv_kernel.cc+4-1
- paddle/phi/kernels/cpu/send_ue_recv_grad_kernel.cc+4-1
- paddle/phi/kernels/cpu/send_ue_recv_kernel.cc+4-1
- paddle/phi/kernels/cpu/send_uv_kernel.cc+4-1
- paddle/phi/kernels/cpu/tdm_child_kernel.cc+5-2
- paddle/phi/kernels/cpu/tdm_sampler_kernel.cc+5-3
- paddle/phi/kernels/funcs/cross_entropy.cc+7-2
- paddle/phi/kernels/funcs/cross_entropy.cu+11-8
- paddle/phi/kernels/funcs/diag_functor.h+2-3
- paddle/phi/kernels/funcs/math/unpooling.cu+112-28
- paddle/phi/kernels/fusion/gpu/block_attn.h+22-6
- paddle/phi/kernels/fusion/gpu/block_multi_head_attention_kernel.cu+8-2
- paddle/phi/kernels/fusion/gpu/fused_multi_transformer_op.cu.h+15-4
- paddle/phi/kernels/fusion/xpu/blha_get_max_len_kernel.cc+4-1
- paddle/phi/kernels/fusion/xpu/block_multi_head_attention_kernel.cc+16-4
- paddle/phi/kernels/gpu/batch_norm_kernel.cu+4-1
- paddle/phi/kernels/gpu/binomial_kernel.cu+4-1
- paddle/phi/kernels/gpu/cholesky_kernel.cu+13-7
- paddle/phi/kernels/gpu/cross_entropy_bwd_w_downcast.cu-1
- paddle/phi/kernels/gpu/cross_entropy_grad_kernel.cu-1
- paddle/phi/kernels/gpu/cudnn_lstm_grad_kernel.cu+7-5
- paddle/phi/kernels/gpu/cudnn_lstm_kernel.cu+10-8
- paddle/phi/kernels/gpu/depthwise_conv.h+47-13
- paddle/phi/kernels/gpu/eig_grad_kernel.cu+523
- paddle/phi/kernels/gpu/eig_kernel.cu+132
- paddle/phi/kernels/gpu/flash_attn_v3_kernel.cu+8-2
- paddle/phi/kernels/gpu/graph_reindex_kernel.cu+6-2
- paddle/phi/kernels/gpu/graph_sample_neighbors_kernel.cu+2-3
- paddle/phi/kernels/gpu/instance_norm_kernel.cu+4-1
- paddle/phi/kernels/gpu/lookup_table_grad_kernel.cu-3
- paddle/phi/kernels/gpu/moe_permute_kernel.cu+4-1
- paddle/phi/kernels/gpu/psroi_pool_grad_kernel.cu-4
- paddle/phi/kernels/gpu/rnn_kernel.cu.cc+13-11
- paddle/phi/kernels/gpu/row_conv_grad_kernel.cu+30-20
- paddle/phi/kernels/gpu/row_conv_kernel.cu+16-15
- paddle/phi/kernels/gpu/send_u_recv_grad_kernel.cu+3-1
- paddle/phi/kernels/gpu/send_u_recv_kernel.cu+4-1
- paddle/phi/kernels/gpu/send_ue_recv_grad_kernel.cu+3-1
- paddle/phi/kernels/gpu/send_ue_recv_kernel.cu+4-1
- paddle/phi/kernels/gpu/svd_kernel.cu+14-8
- paddle/phi/kernels/gpu/unpool_grad_kernel.cu+56-14
- paddle/phi/kernels/gpu/unpool_kernel.cu+56-14
- paddle/phi/kernels/gpu/weight_dequantize_kernel.cu+5-3
- paddle/phi/kernels/gpu/weight_only_linear_grad_kernel.cu-2
- paddle/phi/kernels/gpu/yolo_box_post_kernel.cu+1-2
- paddle/phi/kernels/impl/kldiv_loss_grad_kernel_impl.h+4-1
- paddle/phi/kernels/sparse/cpu/conv.h+11-11
- paddle/phi/kernels/sparse/gpu/conv_kernel_impl.cuh+3-3
- paddle/phi/kernels/sparse/gpu/mv_grad_kernel.cu+2-2
- paddle/phi/kernels/sparse/gpu/pool_grad_kernel.cu+5-2
- paddle/phi/kernels/sparse/gpu/sparse_attention_kernel.cu+3-3
- paddle/phi/kernels/stride/elementwise_stride_base.cu.h+12-3
- paddle/phi/kernels/xpu/index_elementwise_get_kernel.cc+1
- paddle/phi/kernels/xpu/index_elementwise_put_kernel.cc+3-1
- python/paddle/base/core.py+1
- python/paddle/base/dygraph/tensor_patch_methods.py+57-12
- python/paddle/nn/functional/flash_attention.py-6
- python/paddle/nn/functional/sdpa.py+9-8
- python/setup.py.in+2-2
- setup.py+9
- test/legacy_test/test_dlpack_basic.py+24-3
- test/legacy_test/test_eig_op.py+129-67
- test/legacy_test/test_flash_attention.py+25
0 commit comments