Home
last modified time | relevance | path

Searched refs:batch_count (Results 1 – 24 of 24) sorted by relevance

/external/tensorflow/tensorflow/stream_executor/cuda/
Dcuda_fft.cc81 fft::Type type, int batch_count, ScratchAllocator *scratch_allocator) { in Initialize() argument
100 if (batch_count == 1 && input_embed == nullptr && output_embed == nullptr) { in Initialize()
198 output_stride, output_distance, CUDAFftType(type), batch_count); in Initialize()
222 output_stride, output_distance, CUDAFftType(type), batch_count, in Initialize()
396 bool in_place_fft, int batch_count) { in CreateBatchedPlan() argument
401 batch_count, /*scratch_allocator=*/nullptr); in CreateBatchedPlan()
411 << " batch_count: " << batch_count; in CreateBatchedPlan()
423 bool in_place_fft, int batch_count, ScratchAllocator *scratch_allocator) { in CreateBatchedPlanWithScratchAllocator() argument
428 batch_count, scratch_allocator); in CreateBatchedPlanWithScratchAllocator()
438 << " batch_count: " << batch_count; in CreateBatchedPlanWithScratchAllocator()
Dcuda_blas.cc2370 int ldc, int batch_count, ScratchAllocator *scratch_allocator) { in DoBlasInternalImpl() argument
2372 for (int i = 0; i < batch_count; ++i) { in DoBlasInternalImpl()
2380 const size_t size = batch_count * sizeof(CUDA_T *); in DoBlasInternalImpl()
2412 stream->AllocateTemporaryArray<CUDA_T *>(batch_count)); in DoBlasInternalImpl()
2414 stream->AllocateTemporaryArray<CUDA_T *>(batch_count)); in DoBlasInternalImpl()
2416 stream->AllocateTemporaryArray<CUDA_T *>(batch_count)); in DoBlasInternalImpl()
2473 batch_count, compute_type, algo); in DoBlasInternalImpl()
2490 const_cast<CUDA_T **>(GpuMemory(c)), ldc, batch_count); in DoBlasInternalImpl()
2498 for (int b = 0; b < batch_count; ++b) { in DoBlasInternalImpl()
2519 int ldc, int batch_count, ScratchAllocator *scratch_allocator) { in DoBlasInternalImpl() argument
[all …]
Dcuda_fft.h73 int batch_count, ScratchAllocator* scratch_allocator);
Dcuda_blas.h113 int batch_count, ScratchAllocator *scratch_allocator);
Dcuda_dnn.cc4153 const int batch_count = output_dimensions.NodesPerFeatureMap(); in DoMatMul() local
4155 std::vector<DeviceMemory<float>> a(batch_count); in DoMatMul()
4156 std::vector<DeviceMemory<float>> b(batch_count); in DoMatMul()
4157 std::vector<DeviceMemory<float>> c(batch_count); in DoMatMul()
4158 for (int i = 0; i < batch_count; ++i) { in DoMatMul()
4187 ldc, batch_count); in DoMatMul()
/external/tensorflow/tensorflow/stream_executor/
Dfft.h143 bool in_place_fft, int batch_count) = 0;
166 bool in_place_fft, int batch_count,
237 bool in_place_fft, int batch_count) override; \
242 bool in_place_fft, int batch_count, ScratchAllocator *scratch_allocator) \
Dblas.h228 int batch_count = 1; member
1162 int ldc, int batch_count, ScratchAllocator *scratch_allocator) = 0;
1169 int batch_count, ScratchAllocator *scratch_allocator) = 0;
1176 int batch_count, ScratchAllocator *scratch_allocator) = 0;
1184 int batch_count, ScratchAllocator *scratch_allocator) = 0;
1192 int batch_count, ScratchAllocator *scratch_allocator) = 0;
1200 int64 stride_c, int batch_count) = 0;
1206 int batch_count) = 0;
1212 int batch_count) = 0;
1219 int64 stride_c, int batch_count) = 0;
[all …]
Dstream.cc4046 int batch_count) { in ThenBlasGemmBatched() argument
4048 b, ldb, beta, c, ldc, batch_count, in ThenBlasGemmBatched()
4058 int batch_count, ScratchAllocator *scratch_allocator) { in ThenBlasGemmBatchedWithScratch() argument
4061 PARAM(beta), PARAM(c), PARAM(ldc), PARAM(batch_count)); in ThenBlasGemmBatchedWithScratch()
4070 k, alpha, a, lda, b, ldb, beta, c, ldc, batch_count, in ThenBlasGemmBatchedWithScratch()
4079 int batch_count) { in ThenBlasGemmBatched() argument
4081 b, ldb, beta, c, ldc, batch_count, in ThenBlasGemmBatched()
4090 int batch_count, ScratchAllocator *scratch_allocator) { in ThenBlasGemmBatchedWithScratch() argument
4093 PARAM(beta), PARAM(c), PARAM(ldc), PARAM(batch_count)); in ThenBlasGemmBatchedWithScratch()
4102 k, alpha, a, lda, b, ldb, beta, c, ldc, batch_count, in ThenBlasGemmBatchedWithScratch()
[all …]
Dstream.h1405 int ldc, int batch_count);
1413 int ldc, int batch_count);
1421 int ldc, int batch_count);
1429 int batch_count);
1437 int batch_count);
1444 int ldc, int batch_count, ScratchAllocator *scratch_allocator);
1450 int batch_count, ScratchAllocator *scratch_allocator);
1456 int batch_count, ScratchAllocator *scratch_allocator);
1464 int batch_count, ScratchAllocator *scratch_allocator);
1472 int batch_count, ScratchAllocator *scratch_allocator);
[all …]
/external/tensorflow/tensorflow/core/kernels/
Ddepthwise_conv_ops_test.cc59 const int batch_count = 1; in Run() local
64 Tensor image(dtype, {batch_count, image_height, image_width, depth}); in Run()
/external/llvm-project/compiler-rt/lib/sanitizer_common/
Dsanitizer_quarantine.h277 uptr batch_count = 0; in PrintStats() local
282 batch_count++; in PrintStats()
287 uptr quarantine_chunks_capacity = batch_count * QuarantineBatch::kSize; in PrintStats()
296 batch_count, total_bytes, total_quarantined_bytes, in PrintStats()
/external/tensorflow/tensorflow/stream_executor/rocm/
Drocm_fft.cc161 fft::Type type, int batch_count, ScratchAllocator *scratch_allocator) { in Initialize() argument
178 if (batch_count == 1 && input_embed == nullptr && output_embed == nullptr) { in Initialize()
280 output_distance, ROCMFftType(type), batch_count); in Initialize()
305 output_distance, ROCMFftType(type), batch_count, in Initialize()
465 bool in_place_fft, int batch_count) { in CreateBatchedPlan() argument
470 batch_count, /*scratch_allocator=*/nullptr); in CreateBatchedPlan()
483 bool in_place_fft, int batch_count, ScratchAllocator *scratch_allocator) { in CreateBatchedPlanWithScratchAllocator() argument
488 batch_count, scratch_allocator); in CreateBatchedPlanWithScratchAllocator()
Drocm_blas.cc1940 int batch_count, uint64_t batch_stride, in ReorganizeMemory() argument
1942 assert(batch_count > 0); in ReorganizeMemory()
1949 for (int i = 1; i < batch_count; ++i) { in ReorganizeMemory()
1986 int batch_count, uint64_t batch_stride, ScratchAllocator *scratch_allocator, in AllocateStridedBuffer() argument
1998 for (int i = 1; i < batch_count; ++i) { in AllocateStridedBuffer()
2007 size_t matrix_batch_byte_size = matrix_byte_size * batch_count; in AllocateStridedBuffer()
2033 return ReorganizeMemory(stream, device_memory, raw_ptrs, batch_count, in AllocateStridedBuffer()
2045 int ldc, int batch_count, ScratchAllocator *scratch_allocator) { in DoBlasGemmBatchedInternal() argument
2074 for (int i = 0; i < batch_count; ++i) { in DoBlasGemmBatchedInternal()
2090 a_raw_ptrs, batch_count, batch_stride_a, scratch_allocator, stream, in DoBlasGemmBatchedInternal()
[all …]
Drocm_blas.h129 int batch_count, uint64_t batch_stride,
160 int ldc, int batch_count, ScratchAllocator *scratch_allocator);
Drocm_fft.h72 int batch_count, ScratchAllocator *scratch_allocator);
Drocm_dnn.cc3906 const int batch_count = output_dimensions.NodesPerFeatureMap(); in DoMatMul() local
3908 std::vector<DeviceMemory<float>> a(batch_count); in DoMatMul()
3909 std::vector<DeviceMemory<float>> b(batch_count); in DoMatMul()
3910 std::vector<DeviceMemory<float>> c(batch_count); in DoMatMul()
3911 for (int i = 0; i < batch_count; ++i) { in DoMatMul()
3940 ldc, batch_count); in DoMatMul()
/external/igt-gpu-tools/scripts/
Dtrace.pl750 my (%running, %runnable, %queued, %batch_avg, %batch_total_avg, %batch_count);
791 $batch_count{$ring}++;
802 $batch_avg{$ring} /= $batch_count{$ring};
803 $batch_total_avg{$ring} /= $batch_count{$ring};
804 $submit_avg{$ring} /= $batch_count{$ring};
805 $execute_avg{$ring} /= $batch_count{$ring};
806 $ctxsave_avg{$ring} /= $batch_count{$ring};
1112 $stats{'count'} = $batch_count{$ring};
/external/tensorflow/tensorflow/python/keras/layers/preprocessing/
Dnormalization.py188 batch_count = math_ops.reduce_prod(batch_reduce_shape)
190 total_count = batch_count + self.count
192 math_ops.cast(batch_count, dtype=self.dtype) /
/external/tensorflow/tensorflow/python/keras/preprocessing/
Dtext_dataset_test.py87 batch_count = 0
90 batch_count += 1
92 self.assertEqual(batch_count, 2)
Dimage_dataset_test.py113 batch_count = 0
116 batch_count += 1
118 self.assertEqual(batch_count, 2)
/external/blktrace/iowatcher/
Dmain.c1104 int batch_count; in plot_io_movie() local
1115 batch_count = cols / total_frames; in plot_io_movie()
1117 if (batch_count == 0) in plot_io_movie()
1118 batch_count = 1; in plot_io_movie()
1168 while (i < cols && batch_i < batch_count) { in plot_io_movie()
/external/tensorflow/tensorflow/python/keras/engine/
Dtraining_utils_v1.py937 batch_count = int(len(index_array) / batch_size)
940 last_batch = index_array[batch_count * batch_size:]
941 index_array = index_array[:batch_count * batch_size]
942 index_array = index_array.reshape((batch_count, batch_size))
/external/angle/src/tests/
Dcapture_replay_tests.py658 assert len(self.tests) <= self.args.batch_count
857 test_batch_num = int(math.ceil(len(test_names) / float(args.batch_count)))
/external/tensorflow/tensorflow/compiler/xla/service/cpu/
Ddot_op_emitter.cc1149 int64 batch_count = lhs_array_reshaped.GetShape().dimensions(0); in EmitBatchDotOperation() local
1154 llvm_ir::IrName(&dot, "bdot"), /*start=*/0, /*end=*/batch_count, in EmitBatchDotOperation()