From fa5f556f88e0bd0f41778e599ffc083aa77feb45 Mon Sep 17 00:00:00 2001 From: LifengWang Date: Fri, 21 Mar 2025 16:46:00 +0000 Subject: [PATCH] [CI] enable operator benchmark on CPU (#143733) This is to enable operator benchmark for CPU to track op level performance. This PR is motivated by PR: https://github.com/pytorch/pytorch/issues/120982 and investigate feasibility in https://github.com/pytorch/pytorch/pull/127216 Pull Request resolved: https://github.com/pytorch/pytorch/pull/143733 Approved by: https://github.com/leslie-fang-intel, https://github.com/atalman, https://github.com/huydhn, https://github.com/malfet Co-authored-by: diwei sun Co-authored-by: chuanqiw --- .ci/pytorch/test.sh | 34 + .github/pytorch-probot.yml | 1 + .github/workflows/operator_benchmark.yml | 56 + .../operator_benchmark/benchmark_core.py | 57 +- .../operator_benchmark/benchmark_runner.py | 11 + .../operator_benchmark/check_perf_csv.py | 116 ++ ...i_operator_benchmark_eager_float32_cpu.csv | 1319 +++++++++++++++++ .../operator_benchmark/pt/nan_to_num_test.py | 4 + .../pt/quantization_test.py | 8 +- 9 files changed, 1601 insertions(+), 5 deletions(-) create mode 100644 .github/workflows/operator_benchmark.yml create mode 100644 benchmarks/operator_benchmark/check_perf_csv.py create mode 100644 benchmarks/operator_benchmark/expected_ci_operator_benchmark_eager_float32_cpu.csv diff --git a/.ci/pytorch/test.sh b/.ci/pytorch/test.sh index ad32638dd246..2607346c1ba6 100755 --- a/.ci/pytorch/test.sh +++ b/.ci/pytorch/test.sh @@ -1527,6 +1527,27 @@ test_linux_aarch64() { --shard "$SHARD_NUMBER" "$NUM_TEST_SHARDS" --verbose } +test_operator_benchmark() { + TEST_REPORTS_DIR=$(pwd)/test/test-reports + mkdir -p "$TEST_REPORTS_DIR" + TEST_DIR=$(pwd) + + test_inductor_set_cpu_affinity + + cd benchmarks/operator_benchmark/pt_extension + python setup.py install + + cd "${TEST_DIR}"/benchmarks/operator_benchmark + $TASKSET python -m benchmark_all_test --device "$1" --tag-filter "$2" \ + --output-dir "${TEST_REPORTS_DIR}/operator_benchmark_eager_float32_cpu.csv" + + pip_install pandas + python check_perf_csv.py \ + --actual "${TEST_REPORTS_DIR}/operator_benchmark_eager_float32_cpu.csv" \ + --expected "expected_ci_operator_benchmark_eager_float32_cpu.csv" +} + + if ! [[ "${BUILD_ENVIRONMENT}" == *libtorch* || "${BUILD_ENVIRONMENT}" == *-bazel-* ]]; then (cd test && python -c "import torch; print(torch.__config__.show())") (cd test && python -c "import torch; print(torch.__config__.parallel_info())") @@ -1557,6 +1578,19 @@ elif [[ "$TEST_CONFIG" == distributed ]]; then if [[ "${SHARD_NUMBER}" == 1 ]]; then test_rpc fi +elif [[ "${TEST_CONFIG}" == *operator_benchmark* ]]; then + TEST_MODE="short" + + if [[ "${TEST_CONFIG}" == *cpu* ]]; then + if [[ "${TEST_CONFIG}" == *long* ]]; then + TEST_MODE="long" + elif [[ "${TEST_CONFIG}" == *all* ]]; then + TEST_MODE="all" + fi + + test_operator_benchmark cpu ${TEST_MODE} + + fi elif [[ "${TEST_CONFIG}" == *inductor_distributed* ]]; then test_inductor_distributed elif [[ "${TEST_CONFIG}" == *inductor-halide* ]]; then diff --git a/.github/pytorch-probot.yml b/.github/pytorch-probot.yml index ccb71e6a9bf0..6a8ae50b68f4 100644 --- a/.github/pytorch-probot.yml +++ b/.github/pytorch-probot.yml @@ -25,6 +25,7 @@ ciflow_push_tags: - ciflow/xpu - ciflow/torchbench - ciflow/autoformat +- ciflow/op-benchmark retryable_workflows: - pull - trunk diff --git a/.github/workflows/operator_benchmark.yml b/.github/workflows/operator_benchmark.yml new file mode 100644 index 000000000000..7da1b438c7e9 --- /dev/null +++ b/.github/workflows/operator_benchmark.yml @@ -0,0 +1,56 @@ +name: operator_benchmark + +on: + push: + tags: + - ciflow/op-benchmark/* + workflow_dispatch: + inputs: + test_mode: + required: false + type: string + default: 'short' + description: tag filter for operator benchmarks, options from long, short, all + +concurrency: + group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.sha }}-${{ github.event_name == 'workflow_dispatch' }} + cancel-in-progress: true + +permissions: read-all + +jobs: + linux-jammy-cpu-py3_9-gcc11-opbenchmark-build: + if: github.repository_owner == 'pytorch' + name: linux-jammy-cpu-py3.9-gcc11-opbenchmark + uses: ./.github/workflows/_linux-build.yml + with: + build-environment: linux-jammy-py3.9-gcc11-build + docker-image-name: pytorch-linux-jammy-py3.9-gcc11-inductor-benchmarks + test-matrix: | + { include: [ + { config: "cpu_operator_benchmark_short", shard: 1, num_shards: 1, runner: "linux.12xlarge" }, + ]} + secrets: inherit + + linux-jammy-cpu-py3_9-gcc11-opbenchmark-on-demand-build: + if: ${{ github.event_name == 'workflow_dispatch' && github.repository_owner == 'pytorch' }} + name: linux-jammy-cpu-py3.9-gcc11-opbenchmark + uses: ./.github/workflows/_linux-build.yml + with: + build-environment: linux-jammy-py3.9-gcc11-build + docker-image-name: pytorch-linux-jammy-py3.9-gcc11-inductor-benchmarks + test-matrix: | + { include: [ + { config: "cpu_operator_benchmark_${{ inputs.test_mode }}", shard: 1, num_shards: 1, runner: "linux.12xlarge" }, + ]} + secrets: inherit + + linux-jammy-cpu-py3_9-gcc11-opbenchmark-test: + name: linux-jammy-cpu-py3.9-gcc11-opbenchmark + uses: ./.github/workflows/_linux-test.yml + needs: linux-jammy-cpu-py3_9-gcc11-opbenchmark-build + with: + build-environment: linux-jammy-py3.9-gcc11-build + docker-image: ${{ needs.linux-jammy-cpu-py3_9-gcc11-opbenchmark-build.outputs.docker-image }} + test-matrix: ${{ needs.linux-jammy-cpu-py3_9-gcc11-opbenchmark-build.outputs.test-matrix }} + secrets: inherit diff --git a/benchmarks/operator_benchmark/benchmark_core.py b/benchmarks/operator_benchmark/benchmark_core.py index 045a52a5076b..47f346f2933c 100644 --- a/benchmarks/operator_benchmark/benchmark_core.py +++ b/benchmarks/operator_benchmark/benchmark_core.py @@ -1,7 +1,9 @@ import ast import copy +import csv import functools import json +import os import timeit from collections import namedtuple @@ -31,6 +33,8 @@ TestConfig = namedtuple("TestConfig", "test_name input_config tag run_backward") BENCHMARK_TESTER = [] +SKIP_OP_LISTS = ["weight_norm_sparsifier_step"] + def _register_test(*test_metainfo): """save the metainfo needed to create a test. Currently test_metainfo @@ -187,7 +191,9 @@ class BenchmarkRunner: self.use_jit = args.use_jit self.num_runs = args.num_runs self.print_per_iter = False + self.output_dir = args.output_dir self.operator_range = benchmark_utils.get_operator_range(args.operator_range) + self.disable_output = args.disable_output # 100 is the default warmup iterations if self.args.warmup_iterations == -1: self.args.warmup_iterations = 100 @@ -397,6 +403,9 @@ class BenchmarkRunner: test_flag == cmd_flag for cmd_flag in cmd_flag_list ) + def _check_skip(self, test_module, cmd_flag): + return cmd_flag is None or (test_module not in cmd_flag) + def _keep_test(self, test_case): # TODO: consider regex matching for test filtering. # Currently, this is a sub-string matching. @@ -412,6 +421,7 @@ class BenchmarkRunner: return ( self._check_keep(op_test_config.test_name, self.args.test_name) and self._check_keep_list(test_case.op_bench.module_name(), operators) + and self._check_skip(test_case.op_bench.module_name(), SKIP_OP_LISTS) and self._check_operator_first_char( test_case.op_bench.module_name(), self.operator_range ) @@ -446,8 +456,36 @@ class BenchmarkRunner: return False + def _output_csv(self, filename, headers, row): + if self.args.disable_output is True: + return + if os.path.exists(filename): + with open(filename) as fd: + lines = list(csv.reader(fd)) or [[]] + if headers and len(headers) > len(lines[0]): + # if prior results failed the header might not be filled in yet + lines[0] = headers + else: + headers = lines[0] + else: + lines = [headers] + lines.append([(f"{x:.6f}" if isinstance(x, float) else x) for x in row]) + with open(filename, "w") as fd: + writer = csv.writer(fd, lineterminator="\n") + for line in lines: + writer.writerow(list(line) + ["0"] * (len(headers) - len(line))) + def run(self): self._print_header() + output_filename = self.args.output_dir + headers = [ + "Benchmarking Framework", + "Benchamrking Module Name", + "Case Name", + "tag", + "run_backward", + "Execution Time", + ] if self.args.output_json: perf_list = [] @@ -490,8 +528,25 @@ class BenchmarkRunner: ) for _ in range(self.num_runs) ] - self._print_perf_result(reported_time, test_case) + + # output results to csv + self._output_csv( + output_filename, + headers, + [ + test_case.framework, + test_case.op_bench.module_name(), + ( + test_case.test_config.test_name + "_BACKWARD" + if test_case.test_config.run_backward is True + else test_case.test_config.test_name + ), + test_case.test_config.tag, + test_case.test_config.run_backward, + reported_time[0], + ], + ) if self.args.output_json: perf_list.append( self._perf_result_to_dict(reported_time, test_case) diff --git a/benchmarks/operator_benchmark/benchmark_runner.py b/benchmarks/operator_benchmark/benchmark_runner.py index 1004bd4d575c..21d8cfd94bb6 100644 --- a/benchmarks/operator_benchmark/benchmark_runner.py +++ b/benchmarks/operator_benchmark/benchmark_runner.py @@ -150,6 +150,17 @@ def parse_args(): default="None", ) + parser.add_argument( + "--output-dir", + help="Choose the output directory to save the logs", + default="benchmark_logs", + ) + parser.add_argument( + "--disable-output", + help="Disable log output to csv file", + default="False", + ) + args, _ = parser.parse_known_args() if args.omp_num_threads: diff --git a/benchmarks/operator_benchmark/check_perf_csv.py b/benchmarks/operator_benchmark/check_perf_csv.py new file mode 100644 index 000000000000..585a41e3beaf --- /dev/null +++ b/benchmarks/operator_benchmark/check_perf_csv.py @@ -0,0 +1,116 @@ +import argparse +import sys +import textwrap + +import pandas as pd + + +SKIP_TEST_LISTS = [ + # https://github.com/pytorch/pytorch/issues/143852 + "channel_shuffle_batch_size4_channels_per_group64_height64_width64_groups4_channel_lastTrue", + "batchnorm_N3136_C256_cpu_trainingTrue_cudnnFalse", + "index_add__M256_N512_K1_dim1_cpu_dtypetorch.float32", + "interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastTrue_modelinear", + "original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu", + "original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu", +] + + +def get_field(csv, case: str, field: str): + try: + return csv.loc[csv["Case Name"] == case][field].item() + except Exception: + return None + + +def check_perf(actual_csv, expected_csv, expected_filename, threshold): + failed = [] + improved = [] + baseline_not_found = [] + + actual_csv = actual_csv[~actual_csv["Case Name"].isin(set(SKIP_TEST_LISTS))] + + for case in actual_csv["Case Name"]: + perf = get_field(actual_csv, case, "Execution Time") + expected_perf = get_field(expected_csv, case, "Execution Time") + + if expected_perf is None: + status = "Baseline Not Found" + print(f"{case:34} {status}") + baseline_not_found.append(case) + continue + + speed_up = expected_perf / perf + + if (1 - threshold) <= speed_up < (1 + threshold): + status = "PASS" + print(f"{case:34} {status}") + continue + elif speed_up >= 1 + threshold: + status = "IMPROVED:" + improved.append(case) + else: + status = "FAILED:" + failed.append(case) + print(f"{case:34} {status:9} perf={perf}, expected={expected_perf}") + + msg = "" + if failed or improved or baseline_not_found: + if failed: + msg += textwrap.dedent( + f""" + Error: {len(failed)} models have performance status regressed: + {" ".join(failed)} + + """ + ) + if improved: + msg += textwrap.dedent( + f""" + Improvement: {len(improved)} models have performance status improved: + {" ".join(improved)} + + """ + ) + + if baseline_not_found: + msg += textwrap.dedent( + f""" + Baseline Not Found: {len(baseline_not_found)} models don't have the baseline data: + {" ".join(baseline_not_found)} + + """ + ) + + msg += textwrap.dedent( + f""" + If this change is expected, you can update `{expected_filename}` to reflect the new baseline. + """ + ) + return failed or improved or baseline_not_found, msg + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument("--actual", type=str, required=True) + parser.add_argument("--expected", type=str, required=True) + parser.add_argument( + "--threshold", + type=float, + default=0.5, + help="threshold to define regression/improvement", + ) + args = parser.parse_args() + + actual = pd.read_csv(args.actual) + actual.drop_duplicates(subset=["Case Name"], keep="first", inplace=True) + expected = pd.read_csv(args.expected) + + failed, msg = check_perf(actual, expected, args.expected, args.threshold) + if failed: + print(msg) + sys.exit(1) + + +if __name__ == "__main__": + main() diff --git a/benchmarks/operator_benchmark/expected_ci_operator_benchmark_eager_float32_cpu.csv b/benchmarks/operator_benchmark/expected_ci_operator_benchmark_eager_float32_cpu.csv new file mode 100644 index 000000000000..6289dbda597e --- /dev/null +++ b/benchmarks/operator_benchmark/expected_ci_operator_benchmark_eager_float32_cpu.csv @@ -0,0 +1,1319 @@ +Benchmarking Framework,Benchamrking Module Name,Case Name,tag,run_backward,Execution Time +PyTorch,add,add_M1_N1_K1_cpu,short,FALSE,3.9497 +PyTorch,add,add_M64_N64_K64_cpu,short,FALSE,14.3181 +PyTorch,add,add_M64_N64_K128_cpu,short,FALSE,14.6826 +PyTorch,add,add_M1_N1_K1_cpu_bwdall_BACKWARD,short,TRUE,58.1449 +PyTorch,add,add_M1_N1_K1_cpu_bwd1_BACKWARD,short,TRUE,57.765 +PyTorch,add,add_M1_N1_K1_cpu_bwd2_BACKWARD,short,TRUE,57.8035 +PyTorch,add,add_M64_N64_K64_cpu_bwdall_BACKWARD,short,TRUE,135.2775 +PyTorch,add,add_M64_N64_K64_cpu_bwd1_BACKWARD,short,TRUE,135.1988 +PyTorch,add,add_M64_N64_K64_cpu_bwd2_BACKWARD,short,TRUE,135.1905 +PyTorch,add,add_M64_N64_K128_cpu_bwdall_BACKWARD,short,TRUE,135.9341 +PyTorch,add,add_M64_N64_K128_cpu_bwd1_BACKWARD,short,TRUE,136.0071 +PyTorch,add,add_M64_N64_K128_cpu_bwd2_BACKWARD,short,TRUE,135.7898 +PyTorch,addmm,addmm_M1_N1_K1_cpu,short,FALSE,6.5832 +PyTorch,addmm,addmm_M64_N64_K64_cpu,short,FALSE,18.623 +PyTorch,addmm,addmm_M64_N64_K128_cpu,short,FALSE,19.0005 +PyTorch,addmm,addmm_M1_N1_K1_cpu_bwdall_BACKWARD,short,TRUE,86.4294 +PyTorch,addmm,addmm_M1_N1_K1_cpu_bwd1_BACKWARD,short,TRUE,86.5513 +PyTorch,addmm,addmm_M1_N1_K1_cpu_bwd2_BACKWARD,short,TRUE,86.5072 +PyTorch,addmm,addmm_M1_N1_K1_cpu_bwd3_BACKWARD,short,TRUE,86.4965 +PyTorch,addmm,addmm_M64_N64_K64_cpu_bwdall_BACKWARD,short,TRUE,145.9072 +PyTorch,addmm,addmm_M64_N64_K64_cpu_bwd1_BACKWARD,short,TRUE,145.2227 +PyTorch,addmm,addmm_M64_N64_K64_cpu_bwd2_BACKWARD,short,TRUE,145.3786 +PyTorch,addmm,addmm_M64_N64_K64_cpu_bwd3_BACKWARD,short,TRUE,145.2559 +PyTorch,addmm,addmm_M64_N64_K128_cpu_bwdall_BACKWARD,short,TRUE,155.407 +PyTorch,addmm,addmm_M64_N64_K128_cpu_bwd1_BACKWARD,short,TRUE,155.4162 +PyTorch,addmm,addmm_M64_N64_K128_cpu_bwd2_BACKWARD,short,TRUE,155.485 +PyTorch,addmm,addmm_M64_N64_K128_cpu_bwd3_BACKWARD,short,TRUE,155.5694 +PyTorch,as_strided,"as_strided_M8_N8_size(2,2)_stride(1,1)_storage_offset0_cpu",short,FALSE,4.4039 +PyTorch,as_strided,"as_strided_M256_N256_size(32,32)_stride(1,1)_storage_offset0_cpu",short,FALSE,4.4316 +PyTorch,as_strided,"as_strided_M512_N512_size(64,64)_stride(2,2)_storage_offset1_cpu",short,FALSE,4.3663 +PyTorch,batchnorm,batchnorm_M1_N256_K3136_cpu_trainingTrue_cudnnFalse,short,FALSE,153.1791 +PyTorch,batchnorm,batchnorm_M1_N256_K3136_cpu_trainingFalse_cudnnFalse,short,FALSE,36.8686 +PyTorch,batchnorm,batchnorm_M1_N256_K3136_cpu_trainingTrue_cudnnFalse_bwdall_BACKWARD,short,TRUE,171.3087 +PyTorch,batchnorm,batchnorm_M1_N256_K3136_cpu_trainingTrue_cudnnFalse_bwd1_BACKWARD,short,TRUE,171.5833 +PyTorch,batchnorm,batchnorm_M1_N256_K3136_cpu_trainingFalse_cudnnFalse_bwdall_BACKWARD,short,TRUE,169.3315 +PyTorch,batchnorm,batchnorm_M1_N256_K3136_cpu_trainingFalse_cudnnFalse_bwd1_BACKWARD,short,TRUE,169.9856 +PyTorch,batchnorm,batchnorm_N3136_C256_cpu_trainingFalse_cudnnFalse,short,FALSE,37.001 +PyTorch,batchnorm,batchnorm_N3136_C256_cpu_trainingTrue_cudnnFalse_bwdall_BACKWARD,short,TRUE,707.938 +PyTorch,batchnorm,batchnorm_N3136_C256_cpu_trainingTrue_cudnnFalse_bwd1_BACKWARD,short,TRUE,705.6394 +PyTorch,batchnorm,batchnorm_N3136_C256_cpu_trainingFalse_cudnnFalse_bwdall_BACKWARD,short,TRUE,228.8024 +PyTorch,batchnorm,batchnorm_N3136_C256_cpu_trainingFalse_cudnnFalse_bwd1_BACKWARD,short,TRUE,229.531 +PyTorch,add,"add_in_one[64,1,64]_in_two[1,64,1]_cpu_dtypetorch.float32",short,FALSE,14.6918 +PyTorch,add,add_M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,3.8735 +PyTorch,add,add_M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,14.1837 +PyTorch,add,add_M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,14.262 +PyTorch,copy_,copy__M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,2.3689 +PyTorch,copy_,copy__M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,8.467 +PyTorch,copy_,copy__M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,8.3493 +PyTorch,cat,"cat_sizes(1,1,1)_N2_dim0_cpu",short,FALSE,4.084 +PyTorch,cat,"cat_sizes(512,512,2)_N2_dim1_cpu",short,FALSE,19.7723 +PyTorch,cat,"cat_sizes(128,1024,2)_N2_dim1_cpu",short,FALSE,19.0845 +PyTorch,channel_shuffle,channel_shuffle_batch_size2_channels_per_group16_height16_width16_groups2_channel_lastTrue,short,FALSE,11.9975 +PyTorch,channel_shuffle,channel_shuffle_batch_size2_channels_per_group16_height16_width16_groups2_channel_lastFalse,short,FALSE,10.2126 +PyTorch,channel_shuffle,channel_shuffle_batch_size2_channels_per_group32_height32_width32_groups2_channel_lastTrue,short,FALSE,13.4302 +PyTorch,channel_shuffle,channel_shuffle_batch_size2_channels_per_group32_height32_width32_groups2_channel_lastFalse,short,FALSE,10.7077 +PyTorch,channel_shuffle,channel_shuffle_batch_size4_channels_per_group32_height32_width32_groups4_channel_lastTrue,short,FALSE,16.3329 +PyTorch,channel_shuffle,channel_shuffle_batch_size4_channels_per_group32_height32_width32_groups4_channel_lastFalse,short,FALSE,12.5681 +PyTorch,channel_shuffle,channel_shuffle_batch_size4_channels_per_group64_height64_width64_groups4_channel_lastTrue,short,FALSE,125.6838 +PyTorch,channel_shuffle,channel_shuffle_batch_size4_channels_per_group64_height64_width64_groups4_channel_lastFalse,short,FALSE,120.3468 +PyTorch,channel_shuffle,channel_shuffle_batch_size8_channels_per_group64_height64_width64_groups8_channel_lastTrue,short,FALSE,1682.1261 +PyTorch,channel_shuffle,channel_shuffle_batch_size8_channels_per_group64_height64_width64_groups8_channel_lastFalse,short,FALSE,1669.6469 +PyTorch,channel_shuffle,channel_shuffle_batch_size16_channels_per_group64_height64_width64_groups16_channel_lastTrue,short,FALSE,7362.4572 +PyTorch,channel_shuffle,channel_shuffle_batch_size16_channels_per_group64_height64_width64_groups16_channel_lastFalse,short,FALSE,7460.3745 +PyTorch,chunk,chunk_M8_N8_chunks2_cpu,short,FALSE,6.7387 +PyTorch,chunk,chunk_M256_N512_chunks2_cpu,short,FALSE,6.7331 +PyTorch,chunk,chunk_M512_N512_chunks2_cpu,short,FALSE,6.7612 +PyTorch,Conv1d,Conv1d_IC128_OC256_kernel3_stride1_N1_L64_cpu,short,FALSE,84.5187 +PyTorch,Conv1d,Conv1d_IC256_OC256_kernel3_stride2_N4_L64_cpu,short,FALSE,190.8436 +PyTorch,ConvTranspose1d,ConvTranspose1d_IC2016_OC1026_kernel1024_stride256_N1_L224_cpu,short,FALSE,2746443.218 +PyTorch,ConvTranspose1d,ConvTranspose1d_IC128_OC256_kernel3_stride1_N1_L64_cpu,short,FALSE,211.5399 +PyTorch,ConvTranspose1d,ConvTranspose1d_IC256_OC256_kernel3_stride2_N4_L64_cpu,short,FALSE,337.3341 +PyTorch,Conv2d,Conv2d_IC256_OC256_kernel3_stride1_N1_H16_W16_G1_pad0_cpu,short,FALSE,255.67 +PyTorch,ConvTranspose2d,ConvTranspose2d_IC256_OC256_kernel3_stride1_N1_H16_W16_G1_pad0_cpu,short,FALSE,335.2168 +PyTorch,Conv2dPointwise,Conv2dPointwise_IC256_OC256_stride1_N1_H16_W16_G1_pad0_cpu,short,FALSE,154.9221 +PyTorch,Conv3d,Conv3d_IC64_OC64_kernel3_stride1_N8_D4_H16_W16_cpu,short,FALSE,546.3879 +PyTorch,ConvTranspose3d,ConvTranspose3d_IC64_OC64_kernel3_stride1_N8_D4_H16_W16_cpu,short,FALSE,1085.1947 +PyTorch,diag,diag_dim1_M64_N64_diagonal0_outTrue_cpu,short,FALSE,10.5764 +PyTorch,diag,diag_dim2_M128_N128_diagonal-10_outFalse_cpu,short,FALSE,7.56 +PyTorch,diag,diag_dim1_M256_N256_diagonal20_outTrue_cpu,short,FALSE,23.1775 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,29.0592 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,43.3839 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.0397 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.3495 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,29.689 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,43.9881 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.5983 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.8823 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,30.6063 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,47.841 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,30.6428 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,47.8691 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,28.9983 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,43.4274 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.1056 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.4201 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,29.3727 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,45.1213 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.4613 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,45.4844 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,30.9368 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,48.3018 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,30.9671 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,48.6072 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,29.0466 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,43.3651 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.07 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.4114 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,30.312 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,43.7561 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.5243 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.8939 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,30.9467 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,48.6129 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,30.9435 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,48.225 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,29.0514 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,43.3893 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,28.9891 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.4042 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,29.4403 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,44.0072 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,29.5882 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,43.8711 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,31.001 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,49.1249 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,31.0196 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,49.3764 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,99.4515 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,99.5995 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,94.7753 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,94.8932 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,100.8096 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,100.7441 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,95.4789 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,95.8774 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,115.1715 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,115.2046 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,98.8603 +PyTorch,embeddingbag,embeddingbag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,99.0584 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,99.526 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,99.4308 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,103.1781 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,103.3991 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,100.831 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,100.7591 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,103.9218 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,104.1436 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,115.0634 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,115.2862 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,108.1062 +PyTorch,embeddingbag,embeddingbag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,108.1646 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,99.5187 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,99.5428 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,143.7975 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,142.3146 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,100.8197 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,100.7614 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,145.1106 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,143.4348 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,115.1787 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,115.1834 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,153.0874 +PyTorch,embeddingbag,embeddingbag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,152.9047 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,99.4746 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,99.4203 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,142.9623 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,141.5405 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,100.7833 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,100.7818 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,144.5829 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,142.8518 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,115.2002 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,115.3467 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,154.0366 +PyTorch,embeddingbag,embeddingbag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,153.7648 +PyTorch,embedding,embedding_num_embeddings10_embedding_dim64_input_size8_cpu,short,FALSE,15.2528 +PyTorch,embedding,embedding_num_embeddings10_embedding_dim64_input_size16_cpu,short,FALSE,15.324 +PyTorch,embedding,embedding_num_embeddings10_embedding_dim64_input_size64_cpu,short,FALSE,16.6834 +PyTorch,embedding,embedding_num_embeddings120_embedding_dim64_input_size8_cpu,short,FALSE,15.2863 +PyTorch,embedding,embedding_num_embeddings120_embedding_dim64_input_size16_cpu,short,FALSE,15.3453 +PyTorch,embedding,embedding_num_embeddings120_embedding_dim64_input_size64_cpu,short,FALSE,16.7499 +PyTorch,embedding,embedding_num_embeddings1000_embedding_dim64_input_size8_cpu,short,FALSE,15.285 +PyTorch,embedding,embedding_num_embeddings1000_embedding_dim64_input_size16_cpu,short,FALSE,15.3442 +PyTorch,embedding,embedding_num_embeddings1000_embedding_dim64_input_size64_cpu,short,FALSE,16.7735 +PyTorch,embedding,embedding_num_embeddings2300_embedding_dim64_input_size8_cpu,short,FALSE,15.2839 +PyTorch,embedding,embedding_num_embeddings2300_embedding_dim64_input_size16_cpu,short,FALSE,15.34 +PyTorch,embedding,embedding_num_embeddings2300_embedding_dim64_input_size64_cpu,short,FALSE,16.7823 +PyTorch,embedding,embedding_num_embeddings10_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,64.2787 +PyTorch,embedding,embedding_num_embeddings10_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,65.3662 +PyTorch,embedding,embedding_num_embeddings10_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,73.7048 +PyTorch,embedding,embedding_num_embeddings120_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,70.0031 +PyTorch,embedding,embedding_num_embeddings120_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,71.1541 +PyTorch,embedding,embedding_num_embeddings120_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,79.6225 +PyTorch,embedding,embedding_num_embeddings1000_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,99.6097 +PyTorch,embedding,embedding_num_embeddings1000_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,100.9448 +PyTorch,embedding,embedding_num_embeddings1000_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,116.043 +PyTorch,embedding,embedding_num_embeddings2300_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,100.1301 +PyTorch,embedding,embedding_num_embeddings2300_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,101.7065 +PyTorch,embedding,embedding_num_embeddings2300_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,113.1131 +PyTorch,fill_,fill__N1_cpu_dtypetorch.int32,short,FALSE,1.5661 +PyTorch,fill_,fill__N1024_cpu_dtypetorch.int32,short,FALSE,2.491 +PyTorch,fill_,fill__N2048_cpu_dtypetorch.int32,short,FALSE,2.6027 +PyTorch,gather,gather_M256_N512_dim0_cpu,short,FALSE,95.7869 +PyTorch,gather,gather_M512_N512_dim1_cpu,short,FALSE,56.6895 +PyTorch,GroupNormBenchmark,"GroupNormBenchmark_dims(32,8,16)_num_groups2",short,FALSE,16.3328 +PyTorch,GroupNormBenchmark,"GroupNormBenchmark_dims(32,8,16)_num_groups4",short,FALSE,16.7871 +PyTorch,GroupNormBenchmark,"GroupNormBenchmark_dims(32,8,56,56)_num_groups2",short,FALSE,33.1706 +PyTorch,GroupNormBenchmark,"GroupNormBenchmark_dims(32,8,56,56)_num_groups4",short,FALSE,33.2813 +PyTorch,Hardsigmoid,Hardsigmoid_N1_C3_H256_W256_cpu,short,FALSE,22.0699 +PyTorch,Hardsigmoid,Hardsigmoid_N4_C3_H256_W256_cpu,short,FALSE,22.4028 +PyTorch,Hardswish,Hardswish_N1_C3_H256_W256_cpu,short,FALSE,21.769 +PyTorch,Hardswish,Hardswish_N4_C3_H256_W256_cpu,short,FALSE,22.496 +PyTorch,InstanceNormBenchmark,"InstanceNormBenchmark_dims(32,8,16)",short,FALSE,43.7271 +PyTorch,InstanceNormBenchmark,"InstanceNormBenchmark_dims(32,8,56,56)",short,FALSE,155.211 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastTrue_modenearest",short,FALSE,14.6238 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastTrue_modelinear",short,FALSE,17.9502 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastTrue_modebicubic",short,FALSE,62.1607 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastFalse_modenearest",short,FALSE,29.3439 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastFalse_modelinear",short,FALSE,32.6772 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastFalse_modebicubic",short,FALSE,49.7318 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastTrue_modenearest",short,FALSE,124.5945 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastTrue_modelinear",short,FALSE,224.6027 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastTrue_modebicubic",short,FALSE,677.6898 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastFalse_modenearest",short,FALSE,47.635 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastFalse_modelinear",short,FALSE,84.5371 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastFalse_modebicubic",short,FALSE,453.9032 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastTrue_modenearest",short,FALSE,137.7183 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastTrue_modelinear",short,FALSE,250.3212 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastTrue_modebicubic",short,FALSE,763.5104 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastFalse_modenearest",short,FALSE,51.1275 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastFalse_modelinear",short,FALSE,90.131 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastFalse_modebicubic",short,FALSE,503.6091 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastTrue_modenearest",short,FALSE,13.9714 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastTrue_modelinear",short,FALSE,17.504 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastTrue_modebicubic",short,FALSE,34.1424 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastFalse_modenearest",short,FALSE,13.986 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastFalse_modelinear",short,FALSE,17.5393 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastFalse_modebicubic",short,FALSE,34.149 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastTrue_modenearest",short,FALSE,47.9375 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastTrue_modelinear",short,FALSE,84.0489 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastTrue_modebicubic",short,FALSE,453.023 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastFalse_modenearest",short,FALSE,47.88 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastFalse_modelinear",short,FALSE,83.8355 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastFalse_modebicubic",short,FALSE,453.0895 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastTrue_modenearest",short,FALSE,51.024 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastTrue_modelinear",short,FALSE,89.7526 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastTrue_modebicubic",short,FALSE,502.907 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastFalse_modenearest",short,FALSE,51.233 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastFalse_modelinear",short,FALSE,89.4653 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastFalse_modebicubic",short,FALSE,503.0013 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastTrue_modenearest_dtypetorch.uint8",short,FALSE,14.5385 +PyTorch,interpolate,"interpolate_input_size(1,3,60,40)_output_size(24,24)_channels_lastFalse_modenearest_dtypetorch.uint8",short,FALSE,33.8506 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastTrue_modenearest_dtypetorch.uint8",short,FALSE,127.4482 +PyTorch,interpolate,"interpolate_input_size(1,3,600,400)_output_size(240,240)_channels_lastFalse_modenearest_dtypetorch.uint8",short,FALSE,64.694 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastTrue_modenearest_dtypetorch.uint8",short,FALSE,141.7011 +PyTorch,interpolate,"interpolate_input_size(1,3,320,320)_output_size(256,256)_channels_lastFalse_modenearest_dtypetorch.uint8",short,FALSE,70.1855 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastTrue_modenearest_dtypetorch.uint8",short,FALSE,13.7804 +PyTorch,interpolate,"interpolate_input_size(1,1,60,40)_output_size(24,24)_channels_lastFalse_modenearest_dtypetorch.uint8",short,FALSE,13.7883 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastTrue_modenearest_dtypetorch.uint8",short,FALSE,64.5975 +PyTorch,interpolate,"interpolate_input_size(1,1,600,400)_output_size(240,240)_channels_lastFalse_modenearest_dtypetorch.uint8",short,FALSE,64.778 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastTrue_modenearest_dtypetorch.uint8",short,FALSE,70.2955 +PyTorch,interpolate,"interpolate_input_size(1,1,320,320)_output_size(256,256)_channels_lastFalse_modenearest_dtypetorch.uint8",short,FALSE,70.5035 +PyTorch,LayerNormBenchmark,"LayerNormBenchmark_dims(1,8,16)",short,FALSE,10.2883 +PyTorch,LayerNormBenchmark,"LayerNormBenchmark_dims(8,8,16)",short,FALSE,14.6892 +PyTorch,LayerNormBenchmark,"LayerNormBenchmark_dims(32,8,16)",short,FALSE,16.7842 +PyTorch,LayerNormBenchmark,"LayerNormBenchmark_dims(64,128,56,56)",short,FALSE,3613.5986 +PyTorch,linear,linear_N1_IN1_OUT1_cpu,short,FALSE,17.7464 +PyTorch,linear,linear_N4_IN256_OUT128_cpu,short,FALSE,25.6582 +PyTorch,linear,linear_N16_IN512_OUT256_cpu,short,FALSE,40.8298 +PyTorch,matmul,matmul_M1_N1_K1_trans_aTrue_trans_bFalse_cpu,short,FALSE,5.6513 +PyTorch,matmul,matmul_M128_N128_K128_trans_aTrue_trans_bFalse_cpu,short,FALSE,15.9769 +PyTorch,matmul,matmul_M256_N256_K256_trans_aFalse_trans_bTrue_cpu,short,FALSE,36.2135 +PyTorch,nan_to_num,nan_to_num_M16_N64_dtypetorch.float32_replace_infTrue,short,FALSE,6.1434 +PyTorch,nan_to_num,nan_to_num_M16_N64_dtypetorch.float32_replace_infFalse,short,FALSE,6.6946 +PyTorch,nan_to_num,nan_to_num_M16_N64_dtypetorch.float64_replace_infTrue,short,FALSE,6.4431 +PyTorch,nan_to_num,nan_to_num_M16_N64_dtypetorch.float64_replace_infFalse,short,FALSE,6.9691 +PyTorch,nan_to_num,nan_to_num_M64_N64_dtypetorch.float32_replace_infTrue,short,FALSE,7.7852 +PyTorch,nan_to_num,nan_to_num_M64_N64_dtypetorch.float32_replace_infFalse,short,FALSE,8.2928 +PyTorch,nan_to_num,nan_to_num_M64_N64_dtypetorch.float64_replace_infTrue,short,FALSE,8.6692 +PyTorch,nan_to_num,nan_to_num_M64_N64_dtypetorch.float64_replace_infFalse,short,FALSE,9.1626 +PyTorch,nan_to_num_,nan_to_num__M16_N64_dtypetorch.float32_replace_infTrue,short,FALSE,2.98 +PyTorch,nan_to_num_,nan_to_num__M16_N64_dtypetorch.float32_replace_infFalse,short,FALSE,3.4837 +PyTorch,nan_to_num_,nan_to_num__M16_N64_dtypetorch.float64_replace_infTrue,short,FALSE,3.2369 +PyTorch,nan_to_num_,nan_to_num__M16_N64_dtypetorch.float64_replace_infFalse,short,FALSE,3.7523 +PyTorch,nan_to_num_,nan_to_num__M64_N64_dtypetorch.float32_replace_infTrue,short,FALSE,3.6271 +PyTorch,nan_to_num_,nan_to_num__M64_N64_dtypetorch.float32_replace_infFalse,short,FALSE,4.1246 +PyTorch,nan_to_num_,nan_to_num__M64_N64_dtypetorch.float64_replace_infTrue,short,FALSE,4.4276 +PyTorch,nan_to_num_,nan_to_num__M64_N64_dtypetorch.float64_replace_infFalse,short,FALSE,4.9627 +PyTorch,MaxPool1d,MaxPool1d_kernel3_stride1_N8_C256_L256_cpu,short,FALSE,29.9109 +PyTorch,AvgPool1d,AvgPool1d_kernel3_stride1_N8_C256_L256_cpu,short,FALSE,222.0801 +PyTorch,MaxPool2d,"MaxPool2d_kernel[3,1]_stride[2,1]_N1_C16_H32_W32_cpu",short,FALSE,22.5565 +PyTorch,AvgPool2d,"AvgPool2d_kernel[3,1]_stride[2,1]_N1_C16_H32_W32_cpu",short,FALSE,16.4938 +PyTorch,AdaptiveMaxPool2d,"AdaptiveMaxPool2d_kernel[3,1]_stride[2,1]_N1_C16_H32_W32_cpu",short,FALSE,16.1588 +PyTorch,FractionalMaxPool2d,"FractionalMaxPool2d_kernel[3,1]_stride[2,1]_N1_C16_H32_W32_cpu",short,FALSE,22.3268 +PyTorch,MaxPool3d,"MaxPool3d_kernel[3,1,3]_stride[2,1,2]_N1_C16_D16_H32_W32_cpu",short,FALSE,158.2535 +PyTorch,AvgPool3d,"AvgPool3d_kernel[3,1,3]_stride[2,1,2]_N1_C16_D16_H32_W32_cpu",short,FALSE,46.2965 +PyTorch,AdaptiveMaxPool3d,"AdaptiveMaxPool3d_kernel[3,1,3]_stride[2,1,2]_N1_C16_D16_H32_W32_cpu",short,FALSE,32.4115 +PyTorch,FractionalMaxPool3d,"FractionalMaxPool3d_kernel[3,1,3]_stride[2,1,2]_N1_C16_D16_H32_W32_cpu",short,FALSE,22.1037 +PyTorch,fmod,fmod_M1_N1_K1_cpu_dtypetorch.int32,short,FALSE,3.8882 +PyTorch,fmod,fmod_M1_N1_K1_cpu_dtypetorch.float32,short,FALSE,3.9922 +PyTorch,fmod,fmod_M1_N1_K1_cpu_dtypetorch.float64,short,FALSE,3.9689 +PyTorch,fmod,fmod_M64_N64_K64_cpu_dtypetorch.int32,short,FALSE,81.9179 +PyTorch,fmod,fmod_M64_N64_K64_cpu_dtypetorch.float32,short,FALSE,101.8454 +PyTorch,fmod,fmod_M64_N64_K64_cpu_dtypetorch.float64,short,FALSE,192.0626 +PyTorch,fmod,fmod_M64_N64_K128_cpu_dtypetorch.int32,short,FALSE,83.1678 +PyTorch,fmod,fmod_M64_N64_K128_cpu_dtypetorch.float32,short,FALSE,104.5534 +PyTorch,fmod,fmod_M64_N64_K128_cpu_dtypetorch.float64,short,FALSE,195.5447 +PyTorch,remainder,remainder_M1_N1_K1_cpu_dtypetorch.int32,short,FALSE,3.8216 +PyTorch,remainder,remainder_M1_N1_K1_cpu_dtypetorch.float32,short,FALSE,3.9375 +PyTorch,remainder,remainder_M1_N1_K1_cpu_dtypetorch.float64,short,FALSE,3.9714 +PyTorch,remainder,remainder_M64_N64_K64_cpu_dtypetorch.int32,short,FALSE,160.44 +PyTorch,remainder,remainder_M64_N64_K64_cpu_dtypetorch.float32,short,FALSE,112.3482 +PyTorch,remainder,remainder_M64_N64_K64_cpu_dtypetorch.float64,short,FALSE,212.7387 +PyTorch,remainder,remainder_M64_N64_K128_cpu_dtypetorch.int32,short,FALSE,151.3565 +PyTorch,remainder,remainder_M64_N64_K128_cpu_dtypetorch.float32,short,FALSE,116.2934 +PyTorch,remainder,remainder_M64_N64_K128_cpu_dtypetorch.float64,short,FALSE,219.1238 +PyTorch,Softmax,Softmax_N1_C3_H256_W256_cpu,short,FALSE,34.4325 +PyTorch,Softmax,Softmax_N4_C3_H256_W256_cpu,short,FALSE,76.2321 +PyTorch,Softmax2d,Softmax2d_N1_C3_H256_W256_cpu,short,FALSE,32.0697 +PyTorch,Softmax2d,Softmax2d_N4_C3_H256_W256_cpu,short,FALSE,73.5653 +PyTorch,LogSoftmax,LogSoftmax_N1_C3_H256_W256_cpu,short,FALSE,60.9824 +PyTorch,LogSoftmax,LogSoftmax_N4_C3_H256_W256_cpu,short,FALSE,106.1641 +PyTorch,split,split_M8_N8_parts2_cpu,short,FALSE,7.7648 +PyTorch,split,split_M256_N512_parts2_cpu,short,FALSE,7.7645 +PyTorch,split,split_M512_N512_parts2_cpu,short,FALSE,7.7457 +PyTorch,sum,sum_R64_V32_dim0_contiguousTrue_cpu,short,FALSE,7.4003 +PyTorch,sum,sum_R64_V32_dim0_contiguousFalse_cpu,short,FALSE,7.7515 +PyTorch,sum,sum_R64_V32_dim1_contiguousTrue_cpu,short,FALSE,7.6422 +PyTorch,sum,sum_R64_V32_dim1_contiguousFalse_cpu,short,FALSE,8.0323 +PyTorch,sum,sum_R64_V512_dim0_contiguousTrue_cpu,short,FALSE,12.4306 +PyTorch,sum,sum_R64_V512_dim0_contiguousFalse_cpu,short,FALSE,13.821 +PyTorch,sum,sum_R64_V512_dim1_contiguousTrue_cpu,short,FALSE,15.0937 +PyTorch,sum,sum_R64_V512_dim1_contiguousFalse_cpu,short,FALSE,14.2064 +PyTorch,sum,sum_R256_V32_dim0_contiguousTrue_cpu,short,FALSE,7.76 +PyTorch,sum,sum_R256_V32_dim0_contiguousFalse_cpu,short,FALSE,9.7236 +PyTorch,sum,sum_R256_V32_dim1_contiguousTrue_cpu,short,FALSE,7.9835 +PyTorch,sum,sum_R256_V32_dim1_contiguousFalse_cpu,short,FALSE,9.6207 +PyTorch,sum,sum_R256_V512_dim0_contiguousTrue_cpu,short,FALSE,12.5841 +PyTorch,sum,sum_R256_V512_dim0_contiguousFalse_cpu,short,FALSE,20.8765 +PyTorch,sum,sum_R256_V512_dim1_contiguousTrue_cpu,short,FALSE,15.4414 +PyTorch,sum,sum_R256_V512_dim1_contiguousFalse_cpu,short,FALSE,15.3287 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M8_N16_cpu,short,FALSE,5.0499 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M8_N64_cpu,short,FALSE,5.3229 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M8_N128_cpu,short,FALSE,5.4418 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M16_N16_cpu,short,FALSE,5.0868 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M16_N64_cpu,short,FALSE,5.4495 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M16_N128_cpu,short,FALSE,5.5578 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M32_N16_cpu,short,FALSE,5.2631 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M32_N64_cpu,short,FALSE,5.5646 +PyTorch,FloatToHalfTensorConversionBenchmark,FloatToHalfTensorConversionBenchmark_M32_N128_cpu,short,FALSE,5.7898 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M8_N16_cpu,short,FALSE,5.0228 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M8_N64_cpu,short,FALSE,5.3692 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M8_N128_cpu,short,FALSE,5.4006 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M16_N16_cpu,short,FALSE,5.1107 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M16_N64_cpu,short,FALSE,5.4119 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M16_N128_cpu,short,FALSE,5.5583 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M32_N16_cpu,short,FALSE,5.3818 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M32_N64_cpu,short,FALSE,5.5742 +PyTorch,HalfToFloatTensorConversionBenchmark,HalfToFloatTensorConversionBenchmark_M32_N128_cpu,short,FALSE,6.8414 +PyTorch,relu,"relu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,9.4657 +PyTorch,relu,"relu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,9.4625 +PyTorch,relu,"relu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,9.4165 +PyTorch,relu,"relu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,10.0753 +PyTorch,relu,"relu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,10.0801 +PyTorch,relu,"relu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,9.9056 +PyTorch,relu,"relu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,37.4143 +PyTorch,relu,"relu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,37.4995 +PyTorch,relu,"relu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,47.061 +PyTorch,relu,"relu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,38.4561 +PyTorch,relu,"relu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,38.6113 +PyTorch,relu,"relu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,60.9784 +PyTorch,relu6,"relu6_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,9.0443 +PyTorch,relu6,"relu6_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,8.9833 +PyTorch,relu6,"relu6_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,8.9762 +PyTorch,relu6,"relu6_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,9.6588 +PyTorch,relu6,"relu6_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,9.5969 +PyTorch,relu6,"relu6_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,9.547 +PyTorch,relu6,"relu6_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,68.739 +PyTorch,relu6,"relu6_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,45.14133333 +PyTorch,relu6,"relu6_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,52.6664 +PyTorch,relu6,"relu6_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,69.1875 +PyTorch,relu6,"relu6_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,48.3458 +PyTorch,relu6,"relu6_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,62.0719 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,7.5728 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,7.5451 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,7.4914 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,8.1647 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,8.1768 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,8.0619 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,67.118 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,43.702 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,50.3613 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,67.436 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,46.9813 +PyTorch,functional.hardtanh,"functional.hardtanh_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,59.2295 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,6.5189 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,6.5208 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,6.5417 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,7.514 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,7.4671 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,7.5016 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,423.648 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,432.648 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,477.7001 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,428.6677 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,438.0222 +PyTorch,functional.hardsigmoid,"functional.hardsigmoid_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,492.3953 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,10.6166 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,10.6037 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,10.5716 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,11.9313 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,11.9191 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,11.4355 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,57.1153 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,57.19 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,226.2822 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,57.4159 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,57.3784 +PyTorch,functional.leaky_relu,"functional.leaky_relu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,238.6827 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,6.2392 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,6.2414 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,6.2808 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,7.6169 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,7.552 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,7.7053 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,733.8272 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,881.4968 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,649.5353 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,736.2685 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,889.0958 +PyTorch,functional.sigmoid,"functional.sigmoid_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,672.2981 +PyTorch,functional.tanh,"functional.tanh_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,6.2062 +PyTorch,functional.tanh,"functional.tanh_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,6.355 +PyTorch,functional.tanh,"functional.tanh_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,6.2835 +PyTorch,functional.tanh,"functional.tanh_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,7.6287 +PyTorch,functional.tanh,"functional.tanh_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,7.7579 +PyTorch,functional.tanh,"functional.tanh_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,7.6012 +PyTorch,functional.tanh,"functional.tanh_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,584.3268 +PyTorch,functional.tanh,"functional.tanh_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,680.9102 +PyTorch,functional.tanh,"functional.tanh_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,701.6249 +PyTorch,functional.tanh,"functional.tanh_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,591.4621 +PyTorch,functional.tanh,"functional.tanh_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,687.5734 +PyTorch,functional.tanh,"functional.tanh_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,710.5012 +PyTorch,functional.hardswish,"functional.hardswish_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,10.6018 +PyTorch,functional.hardswish,"functional.hardswish_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,10.5699 +PyTorch,functional.hardswish,"functional.hardswish_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,10.5695 +PyTorch,functional.hardswish,"functional.hardswish_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,11.5372 +PyTorch,functional.hardswish,"functional.hardswish_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,11.5601 +PyTorch,functional.hardswish,"functional.hardswish_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,11.6734 +PyTorch,functional.hardswish,"functional.hardswish_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,248.9016 +PyTorch,functional.hardswish,"functional.hardswish_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,251.228 +PyTorch,functional.hardswish,"functional.hardswish_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,311.6496 +PyTorch,functional.hardswish,"functional.hardswish_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,254.0779 +PyTorch,functional.hardswish,"functional.hardswish_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,256.7338 +PyTorch,functional.hardswish,"functional.hardswish_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,311.0197 +PyTorch,functional.elu,"functional.elu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,11.1139 +PyTorch,functional.elu,"functional.elu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,11.3503 +PyTorch,functional.elu,"functional.elu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,11.3583 +PyTorch,functional.elu,"functional.elu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,11.9967 +PyTorch,functional.elu,"functional.elu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,12.4721 +PyTorch,functional.elu,"functional.elu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,12.5357 +PyTorch,functional.elu,"functional.elu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,221.9019 +PyTorch,functional.elu,"functional.elu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,562.2536 +PyTorch,functional.elu,"functional.elu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,631.6971 +PyTorch,functional.elu,"functional.elu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,224.0514 +PyTorch,functional.elu,"functional.elu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,575.5199 +PyTorch,functional.elu,"functional.elu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,644.9067 +PyTorch,functional.celu,"functional.celu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,10.8539 +PyTorch,functional.celu,"functional.celu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,11.0591 +PyTorch,functional.celu,"functional.celu_dims(3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,11.098 +PyTorch,functional.celu,"functional.celu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,11.738 +PyTorch,functional.celu,"functional.celu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,12.2373 +PyTorch,functional.celu,"functional.celu_dims(2,3,4,5)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,12.2706 +PyTorch,functional.celu,"functional.celu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,221.5425 +PyTorch,functional.celu,"functional.celu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,562.3881 +PyTorch,functional.celu,"functional.celu_dims(512,512)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,627.6411 +PyTorch,functional.celu,"functional.celu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.quint8",short,FALSE,222.5929 +PyTorch,functional.celu,"functional.celu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint8",short,FALSE,575.7397 +PyTorch,functional.celu,"functional.celu_dims(256,1024)_contigFalse_inplaceFalse_dtypetorch.qint32",short,FALSE,645.4033 +PyTorch,add,add_N2_dtypetorch.quint8_contigFalse,short,FALSE,10.5944 +PyTorch,add,add_N2_dtypetorch.quint8_contigTrue,short,FALSE,10.3022 +PyTorch,add,add_N2_dtypetorch.qint8_contigFalse,short,FALSE,10.5853 +PyTorch,add,add_N2_dtypetorch.qint8_contigTrue,short,FALSE,10.381 +PyTorch,add,add_N2_dtypetorch.qint32_contigFalse,short,FALSE,10.6032 +PyTorch,add,add_N2_dtypetorch.qint32_contigTrue,short,FALSE,10.2916 +PyTorch,add,add_N8_dtypetorch.quint8_contigFalse,short,FALSE,11.1193 +PyTorch,add,add_N8_dtypetorch.quint8_contigTrue,short,FALSE,10.8562 +PyTorch,add,add_N8_dtypetorch.qint8_contigFalse,short,FALSE,11.1441 +PyTorch,add,add_N8_dtypetorch.qint8_contigTrue,short,FALSE,10.8855 +PyTorch,add,add_N8_dtypetorch.qint32_contigFalse,short,FALSE,11.1265 +PyTorch,add,add_N8_dtypetorch.qint32_contigTrue,short,FALSE,10.6666 +PyTorch,add,add_N64_dtypetorch.quint8_contigFalse,short,FALSE,42.827 +PyTorch,add,add_N64_dtypetorch.quint8_contigTrue,short,FALSE,12.1777 +PyTorch,add,add_N64_dtypetorch.qint8_contigFalse,short,FALSE,71.8321 +PyTorch,add,add_N64_dtypetorch.qint8_contigTrue,short,FALSE,12.2144 +PyTorch,add,add_N64_dtypetorch.qint32_contigFalse,short,FALSE,45.3253 +PyTorch,add,add_N64_dtypetorch.qint32_contigTrue,short,FALSE,31.7538 +PyTorch,add,add_N512_dtypetorch.quint8_contigFalse,short,FALSE,282.9102 +PyTorch,add,add_N512_dtypetorch.quint8_contigTrue,short,FALSE,29.0446 +PyTorch,add,add_N512_dtypetorch.qint8_contigFalse,short,FALSE,557.6633 +PyTorch,add,add_N512_dtypetorch.qint8_contigTrue,short,FALSE,28.9897 +PyTorch,add,add_N512_dtypetorch.qint32_contigFalse,short,FALSE,332.7038 +PyTorch,add,add_N512_dtypetorch.qint32_contigTrue,short,FALSE,186.5795 +PyTorch,add_relu,add_relu_N2_dtypetorch.quint8_contigFalse,short,FALSE,10.5985 +PyTorch,add_relu,add_relu_N2_dtypetorch.quint8_contigTrue,short,FALSE,10.2837 +PyTorch,add_relu,add_relu_N2_dtypetorch.qint8_contigFalse,short,FALSE,10.6095 +PyTorch,add_relu,add_relu_N2_dtypetorch.qint8_contigTrue,short,FALSE,10.2838 +PyTorch,add_relu,add_relu_N2_dtypetorch.qint32_contigFalse,short,FALSE,10.6035 +PyTorch,add_relu,add_relu_N2_dtypetorch.qint32_contigTrue,short,FALSE,10.2648 +PyTorch,add_relu,add_relu_N8_dtypetorch.quint8_contigFalse,short,FALSE,11.1752 +PyTorch,add_relu,add_relu_N8_dtypetorch.quint8_contigTrue,short,FALSE,10.8657 +PyTorch,add_relu,add_relu_N8_dtypetorch.qint8_contigFalse,short,FALSE,11.2 +PyTorch,add_relu,add_relu_N8_dtypetorch.qint8_contigTrue,short,FALSE,10.8263 +PyTorch,add_relu,add_relu_N8_dtypetorch.qint32_contigFalse,short,FALSE,11.1316 +PyTorch,add_relu,add_relu_N8_dtypetorch.qint32_contigTrue,short,FALSE,10.6437 +PyTorch,add_relu,add_relu_N64_dtypetorch.quint8_contigFalse,short,FALSE,44.7881 +PyTorch,add_relu,add_relu_N64_dtypetorch.quint8_contigTrue,short,FALSE,12.43 +PyTorch,add_relu,add_relu_N64_dtypetorch.qint8_contigFalse,short,FALSE,57.4703 +PyTorch,add_relu,add_relu_N64_dtypetorch.qint8_contigTrue,short,FALSE,12.4346 +PyTorch,add_relu,add_relu_N64_dtypetorch.qint32_contigFalse,short,FALSE,45.4349 +PyTorch,add_relu,add_relu_N64_dtypetorch.qint32_contigTrue,short,FALSE,31.8962 +PyTorch,add_relu,add_relu_N512_dtypetorch.quint8_contigFalse,short,FALSE,300.9877 +PyTorch,add_relu,add_relu_N512_dtypetorch.quint8_contigTrue,short,FALSE,31.4974 +PyTorch,add_relu,add_relu_N512_dtypetorch.qint8_contigFalse,short,FALSE,410.9462 +PyTorch,add_relu,add_relu_N512_dtypetorch.qint8_contigTrue,short,FALSE,31.4363 +PyTorch,add_relu,add_relu_N512_dtypetorch.qint32_contigFalse,short,FALSE,344.9523 +PyTorch,add_relu,add_relu_N512_dtypetorch.qint32_contigTrue,short,FALSE,187.2273 +PyTorch,mul,mul_N2_dtypetorch.quint8_contigFalse,short,FALSE,10.6211 +PyTorch,mul,mul_N2_dtypetorch.quint8_contigTrue,short,FALSE,10.2957 +PyTorch,mul,mul_N2_dtypetorch.qint8_contigFalse,short,FALSE,26.7876 +PyTorch,mul,mul_N2_dtypetorch.qint8_contigTrue,short,FALSE,11.8583 +PyTorch,mul,mul_N2_dtypetorch.qint32_contigFalse,short,FALSE,10.6105 +PyTorch,mul,mul_N2_dtypetorch.qint32_contigTrue,short,FALSE,10.253 +PyTorch,mul,mul_N8_dtypetorch.quint8_contigFalse,short,FALSE,10.8428 +PyTorch,mul,mul_N8_dtypetorch.quint8_contigTrue,short,FALSE,10.2394 +PyTorch,mul,mul_N8_dtypetorch.qint8_contigFalse,short,FALSE,482666.2954 +PyTorch,mul,mul_N8_dtypetorch.qint8_contigTrue,short,FALSE,482636.4966 +PyTorch,mul,mul_N8_dtypetorch.qint32_contigFalse,short,FALSE,10.8113 +PyTorch,mul,mul_N8_dtypetorch.qint32_contigTrue,short,FALSE,10.5465 +PyTorch,mul,mul_N64_dtypetorch.quint8_contigFalse,short,FALSE,24.7507 +PyTorch,mul,mul_N64_dtypetorch.quint8_contigTrue,short,FALSE,11.7326 +PyTorch,mul,mul_N64_dtypetorch.qint8_contigFalse,short,FALSE,482828.4171 +PyTorch,mul,mul_N64_dtypetorch.qint8_contigTrue,short,FALSE,481824.9386 +PyTorch,mul,mul_N64_dtypetorch.qint32_contigFalse,short,FALSE,26.0202 +PyTorch,mul,mul_N64_dtypetorch.qint32_contigTrue,short,FALSE,12.8726 +PyTorch,mul,mul_N512_dtypetorch.quint8_contigFalse,short,FALSE,145.5759 +PyTorch,mul,mul_N512_dtypetorch.quint8_contigTrue,short,FALSE,32.8526 +PyTorch,mul,mul_N512_dtypetorch.qint8_contigFalse,short,FALSE,504309.8453 +PyTorch,mul,mul_N512_dtypetorch.qint8_contigTrue,short,FALSE,481425.6449 +PyTorch,mul,mul_N512_dtypetorch.qint32_contigFalse,short,FALSE,178.6525 +PyTorch,mul,mul_N512_dtypetorch.qint32_contigTrue,short,FALSE,25.5806 +PyTorch,add_scalar,add_scalar_N2_dtypetorch.quint8_contigFalse,short,FALSE,10.0295 +PyTorch,add_scalar,add_scalar_N2_dtypetorch.quint8_contigTrue,short,FALSE,9.7243 +PyTorch,add_scalar,add_scalar_N2_dtypetorch.qint8_contigFalse,short,FALSE,10.4763 +PyTorch,add_scalar,add_scalar_N2_dtypetorch.qint8_contigTrue,short,FALSE,10.115 +PyTorch,add_scalar,add_scalar_N2_dtypetorch.qint32_contigFalse,short,FALSE,10.4586 +PyTorch,add_scalar,add_scalar_N2_dtypetorch.qint32_contigTrue,short,FALSE,10.1003 +PyTorch,add_scalar,add_scalar_N8_dtypetorch.quint8_contigFalse,short,FALSE,10.2296 +PyTorch,add_scalar,add_scalar_N8_dtypetorch.quint8_contigTrue,short,FALSE,9.6789 +PyTorch,add_scalar,add_scalar_N8_dtypetorch.qint8_contigFalse,short,FALSE,10.5389 +PyTorch,add_scalar,add_scalar_N8_dtypetorch.qint8_contigTrue,short,FALSE,10.1111 +PyTorch,add_scalar,add_scalar_N8_dtypetorch.qint32_contigFalse,short,FALSE,10.5022 +PyTorch,add_scalar,add_scalar_N8_dtypetorch.qint32_contigTrue,short,FALSE,10.1294 +PyTorch,add_scalar,add_scalar_N64_dtypetorch.quint8_contigFalse,short,FALSE,23.0866 +PyTorch,add_scalar,add_scalar_N64_dtypetorch.quint8_contigTrue,short,FALSE,10.6124 +PyTorch,add_scalar,add_scalar_N64_dtypetorch.qint8_contigFalse,short,FALSE,12.5145 +PyTorch,add_scalar,add_scalar_N64_dtypetorch.qint8_contigTrue,short,FALSE,10.2379 +PyTorch,add_scalar,add_scalar_N64_dtypetorch.qint32_contigFalse,short,FALSE,13.7177 +PyTorch,add_scalar,add_scalar_N64_dtypetorch.qint32_contigTrue,short,FALSE,11.5271 +PyTorch,add_scalar,add_scalar_N512_dtypetorch.quint8_contigFalse,short,FALSE,121.8917 +PyTorch,add_scalar,add_scalar_N512_dtypetorch.quint8_contigTrue,short,FALSE,21.5355 +PyTorch,add_scalar,add_scalar_N512_dtypetorch.qint8_contigFalse,short,FALSE,36.5354 +PyTorch,add_scalar,add_scalar_N512_dtypetorch.qint8_contigTrue,short,FALSE,15.6843 +PyTorch,add_scalar,add_scalar_N512_dtypetorch.qint32_contigFalse,short,FALSE,49.8448 +PyTorch,add_scalar,add_scalar_N512_dtypetorch.qint32_contigTrue,short,FALSE,17.9149 +PyTorch,mul_scalar,mul_scalar_N2_dtypetorch.quint8_contigFalse,short,FALSE,10.36 +PyTorch,mul_scalar,mul_scalar_N2_dtypetorch.quint8_contigTrue,short,FALSE,9.9088 +PyTorch,mul_scalar,mul_scalar_N2_dtypetorch.qint8_contigFalse,short,FALSE,10.2901 +PyTorch,mul_scalar,mul_scalar_N2_dtypetorch.qint8_contigTrue,short,FALSE,9.9043 +PyTorch,mul_scalar,mul_scalar_N2_dtypetorch.qint32_contigFalse,short,FALSE,10.2596 +PyTorch,mul_scalar,mul_scalar_N2_dtypetorch.qint32_contigTrue,short,FALSE,9.9143 +PyTorch,mul_scalar,mul_scalar_N8_dtypetorch.quint8_contigFalse,short,FALSE,10.3176 +PyTorch,mul_scalar,mul_scalar_N8_dtypetorch.quint8_contigTrue,short,FALSE,9.9044 +PyTorch,mul_scalar,mul_scalar_N8_dtypetorch.qint8_contigFalse,short,FALSE,10.3161 +PyTorch,mul_scalar,mul_scalar_N8_dtypetorch.qint8_contigTrue,short,FALSE,9.8889 +PyTorch,mul_scalar,mul_scalar_N8_dtypetorch.qint32_contigFalse,short,FALSE,10.3277 +PyTorch,mul_scalar,mul_scalar_N8_dtypetorch.qint32_contigTrue,short,FALSE,9.9309 +PyTorch,mul_scalar,mul_scalar_N64_dtypetorch.quint8_contigFalse,short,FALSE,12.3152 +PyTorch,mul_scalar,mul_scalar_N64_dtypetorch.quint8_contigTrue,short,FALSE,10.0833 +PyTorch,mul_scalar,mul_scalar_N64_dtypetorch.qint8_contigFalse,short,FALSE,12.3086 +PyTorch,mul_scalar,mul_scalar_N64_dtypetorch.qint8_contigTrue,short,FALSE,10.0465 +PyTorch,mul_scalar,mul_scalar_N64_dtypetorch.qint32_contigFalse,short,FALSE,13.4912 +PyTorch,mul_scalar,mul_scalar_N64_dtypetorch.qint32_contigTrue,short,FALSE,11.3329 +PyTorch,mul_scalar,mul_scalar_N512_dtypetorch.quint8_contigFalse,short,FALSE,36.5829 +PyTorch,mul_scalar,mul_scalar_N512_dtypetorch.quint8_contigTrue,short,FALSE,15.4998 +PyTorch,mul_scalar,mul_scalar_N512_dtypetorch.qint8_contigFalse,short,FALSE,36.4562 +PyTorch,mul_scalar,mul_scalar_N512_dtypetorch.qint8_contigTrue,short,FALSE,15.3596 +PyTorch,mul_scalar,mul_scalar_N512_dtypetorch.qint32_contigFalse,short,FALSE,49.8853 +PyTorch,mul_scalar,mul_scalar_N512_dtypetorch.qint32_contigTrue,short,FALSE,17.6362 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,355.0354 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,368.8042 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,360.2546 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,381.3022 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,374.2793 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,390.4843 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,430.9984 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,445.2845 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,433.8101 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,451.0111 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,438.0377 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,453.5154 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,709.668 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,721.3673 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,710.1981 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,721.1726 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,710.6612 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,723.0125 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,713.4903 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,724.1643 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,712.8765 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,724.4497 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,714.4023 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,726.1041 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,109.969 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,110.0344 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,110.761 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,110.5783 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,114.0811 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,114.2354 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,132.7028 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,132.6065 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,133.5545 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,133.6274 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,137.6377 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,137.5045 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,221.7957 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,221.4864 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,222.7084 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,222.4641 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,229.7689 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,229.6451 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,216.9278 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,216.7878 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,218.8793 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,217.6596 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu_BACKWARD,short,TRUE,227.4012 +PyTorch,qatEmbeddingBag,qatEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu_BACKWARD,short,TRUE,226.5648 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings10_embedding_dim64_input_size8_cpu,short,FALSE,290.8695 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings10_embedding_dim64_input_size16_cpu,short,FALSE,290.7763 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings10_embedding_dim64_input_size64_cpu,short,FALSE,297.0648 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings120_embedding_dim64_input_size8_cpu,short,FALSE,403.8592 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings120_embedding_dim64_input_size16_cpu,short,FALSE,403.8437 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings120_embedding_dim64_input_size64_cpu,short,FALSE,409.9157 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings1000_embedding_dim64_input_size8_cpu,short,FALSE,671.7752 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings1000_embedding_dim64_input_size16_cpu,short,FALSE,672.3172 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings1000_embedding_dim64_input_size64_cpu,short,FALSE,676.8372 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings2300_embedding_dim64_input_size8_cpu,short,FALSE,674.6064 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings2300_embedding_dim64_input_size16_cpu,short,FALSE,675.0676 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings2300_embedding_dim64_input_size64_cpu,short,FALSE,680.827 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings10_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,79.0452 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings10_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,80.1169 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings10_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,88.6495 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings120_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,96.7162 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings120_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,97.7005 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings120_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,106.7757 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings1000_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,175.1056 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings1000_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,176.8257 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings1000_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,190.8659 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings2300_embedding_dim64_input_size8_cpu_BACKWARD,short,TRUE,171.2972 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings2300_embedding_dim64_input_size16_cpu_BACKWARD,short,TRUE,172.8735 +PyTorch,qatEmbedding,qatEmbedding_num_embeddings2300_embedding_dim64_input_size64_cpu_BACKWARD,short,TRUE,183.3709 +PyTorch,QBatchNorm1d,QBatchNorm1d_M1_N256_K3136_cpu_dtypetorch.qint8,short,FALSE,153.9832 +PyTorch,QBatchNorm2d,QBatchNorm2d_M1_N256_K3136_cpu_dtypetorch.qint8,short,FALSE,61.4478 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contigall_dtypetorch.quint8,short,FALSE,104.9221 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contigall_dtypetorch.qint8,short,FALSE,104.7472 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contigall_dtypetorch.qint32,short,FALSE,128.8396 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contigone_dtypetorch.quint8,short,FALSE,155.9432 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contigone_dtypetorch.qint8,short,FALSE,155.3885 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contigone_dtypetorch.qint32,short,FALSE,188.1577 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contignone_dtypetorch.quint8,short,FALSE,190.7007 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contignone_dtypetorch.qint8,short,FALSE,192.4179 +PyTorch,qcat,qcat_M256_N512_K1_L2_dim0_contignone_dtypetorch.qint32,short,FALSE,231.5225 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contigall_dtypetorch.quint8,short,FALSE,129.3093 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contigall_dtypetorch.qint8,short,FALSE,128.2626 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contigall_dtypetorch.qint32,short,FALSE,271.3915 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contigone_dtypetorch.quint8,short,FALSE,191.3597 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contigone_dtypetorch.qint8,short,FALSE,191.134 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contigone_dtypetorch.qint32,short,FALSE,397.2508 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contignone_dtypetorch.quint8,short,FALSE,237.8599 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contignone_dtypetorch.qint8,short,FALSE,240.6118 +PyTorch,qcat,qcat_M512_N512_K2_L1_dim1_contignone_dtypetorch.qint32,short,FALSE,506.6646 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.4297 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.0997 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.7053 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.9015 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0206 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.1004 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6579 +PyTorch,eq,eq_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.3904 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.3498 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.0453 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.6204 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.9011 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0282 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.1237 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.5881 +PyTorch,eq,eq_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.3807 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.5953 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.3007 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.6519 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.9558 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0615 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.2629 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6217 +PyTorch,eq,eq_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.4081 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,32.5281 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,46.3943 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,32.9841 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,45.4013 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,25.19 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,37.6235 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,24.520375 +PyTorch,eq,eq_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,37.521875 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,33.144625 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,47.185375 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,33.52025 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,46.367375 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,25.179625 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,38.30575 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,24.608625 +PyTorch,eq,eq_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,36.7722 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,34.2642 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,48.2449 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,34.0771 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,46.5628 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,24.3846 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,37.5572 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,24.827125 +PyTorch,eq,eq_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,36.9891 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.5395 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.1835 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.682 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7864 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.1007 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.0939 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.644 +PyTorch,ne,ne_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2399 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.4854 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.09 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5728 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7568 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0606 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.0656 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.5755 +PyTorch,ne,ne_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.186 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.6756 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.2891 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5692 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.8301 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.193 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.2332 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6213 +PyTorch,ne,ne_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2713 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,30.6881 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,44.4952 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.8194 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,43.33 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,22.112 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.6978 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.6752 +PyTorch,ne,ne_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.7569 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,30.6535 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,44.3514 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.8102 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.2606 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,22.7416 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.675 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.4435 +PyTorch,ne,ne_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.6755 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,32.4963 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,45.9445 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,32.1904 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,41.2365 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.960125 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.6699 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.582625 +PyTorch,ne,ne_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.653 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.5126 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.2293 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5979 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.9025 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0575 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.2111 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6034 +PyTorch,lt,lt_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.3414 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.4161 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.1066 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5909 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.8661 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0769 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.1281 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.5435 +PyTorch,lt,lt_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.3255 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.6525 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.3766 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5673 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.9844 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.1277 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.2652 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6394 +PyTorch,lt,lt_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.3441 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,30.1368 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.9901 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.6197 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.5089 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,22.0542 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.2108 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.4296 +PyTorch,lt,lt_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.3959 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,29.9856 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.9188 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.2664 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.332 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,22.0548 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.1588 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.4009 +PyTorch,lt,lt_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.1072 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,31.7829 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,45.85 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,31.5806 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,41.043 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.9851 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.1132 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.588 +PyTorch,lt,lt_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.695 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.5106 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.1371 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.6594 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7381 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0487 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.078 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6736 +PyTorch,gt,gt_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.1977 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.3725 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,29.9698 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.6123 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.6552 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0132 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.036 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6138 +PyTorch,gt,gt_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.1466 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.6191 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.2911 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.567 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.8132 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0952 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.1532 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6421 +PyTorch,gt,gt_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2205 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,30.0702 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.8646 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.6099 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.0542 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,22.0923 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.0654 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.6026 +PyTorch,gt,gt_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,33.8391 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,29.8884 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.8855 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.344 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.4249 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,22.0377 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,35.0056 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.6264 +PyTorch,gt,gt_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.0507 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,31.7842 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,45.8158 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,31.5994 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,41.2045 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.956 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.9384 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.7229 +PyTorch,gt,gt_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.1293 +PyTorch,le,le_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.4409 +PyTorch,le,le_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.0634 +PyTorch,le,le_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.602 +PyTorch,le,le_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7537 +PyTorch,le,le_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0882 +PyTorch,le,le_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.0023 +PyTorch,le,le_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6048 +PyTorch,le,le_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2668 +PyTorch,le,le_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.4064 +PyTorch,le,le_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,29.96 +PyTorch,le,le_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.599 +PyTorch,le,le_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.6939 +PyTorch,le,le_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.1344 +PyTorch,le,le_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,24.9514 +PyTorch,le,le_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.5767 +PyTorch,le,le_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2132 +PyTorch,le,le_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.6484 +PyTorch,le,le_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.2545 +PyTorch,le,le_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5295 +PyTorch,le,le_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7238 +PyTorch,le,le_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.2001 +PyTorch,le,le_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.0923 +PyTorch,le,le_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6029 +PyTorch,le,le_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2461 +PyTorch,le,le_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,29.6869 +PyTorch,le,le_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.5843 +PyTorch,le,le_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,30.3092 +PyTorch,le,le_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.2157 +PyTorch,le,le_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.7831 +PyTorch,le,le_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.6798 +PyTorch,le,le_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.1468 +PyTorch,le,le_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,33.912 +PyTorch,le,le_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,29.6168 +PyTorch,le,le_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.5096 +PyTorch,le,le_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,29.8273 +PyTorch,le,le_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,42.5003 +PyTorch,le,le_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.7367 +PyTorch,le,le_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.6689 +PyTorch,le,le_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.1267 +PyTorch,le,le_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,33.7905 +PyTorch,le,le_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,31.4938 +PyTorch,le,le_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,45.118 +PyTorch,le,le_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,31.0123 +PyTorch,le,le_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,43.125 +PyTorch,le,le_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.6769 +PyTorch,le,le_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.624 +PyTorch,le,le_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.1879 +PyTorch,le,le_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,33.5701 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.4122 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.1641 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.62 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7366 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0367 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.1452 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6244 +PyTorch,ge,ge_N8_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.1445 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.3426 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.0687 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.599 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7432 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.0121 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.112 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.5655 +PyTorch,ge,ge_N8_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.1537 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,18.5808 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,30.3334 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,18.5502 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,30.7833 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,14.1244 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,25.2665 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,14.6322 +PyTorch,ge,ge_N8_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,26.2824 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,29.6852 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.5176 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,29.5751 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,41.6508 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.7045 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.7853 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.3255 +PyTorch,ge,ge_N64_dtypetorch.quint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,35.008 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,29.6213 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,43.4305 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,29.9976 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,41.8899 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.6829 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.7859 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.1738 +PyTorch,ge,ge_N64_dtypetorch.qint8_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,33.846 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantFalse,short,FALSE,31.5165 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigFalse_other_scalarFalse_out_variantTrue,short,FALSE,45.5134 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantFalse,short,FALSE,31.2167 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigFalse_other_scalarTrue_out_variantTrue,short,FALSE,43.2783 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantFalse,short,FALSE,21.601 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigTrue_other_scalarFalse_out_variantTrue,short,FALSE,34.7351 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantFalse,short,FALSE,21.2838 +PyTorch,ge,ge_N64_dtypetorch.qint32_contigTrue_other_scalarTrue_out_variantTrue,short,FALSE,34.3512 +PyTorch,QConv1d,QConv1d_IC128_OC256_kernel3_stride1_N1_L64_cpu,short,FALSE,121.0661 +PyTorch,QConv1d,QConv1d_IC256_OC256_kernel3_stride2_N4_L64_cpu,short,FALSE,150.2673 +PyTorch,QConv2d,QConv2d_IC256_OC256_kernel3_stride1_N1_H16_W16_G1_pad0_cpu,short,FALSE,170.6436 +PyTorch,qembeddingbag_byte_prepack,qembeddingbag_byte_prepack_num_embeddings80_embedding_dim128,short,FALSE,9.9914 +PyTorch,qembeddingbag_byte_prepack,qembeddingbag_byte_prepack_num_embeddings80_embedding_dim256,short,FALSE,11.3223 +PyTorch,qembeddingbag_byte_prepack,qembeddingbag_byte_prepack_num_embeddings80_embedding_dim512,short,FALSE,11.8468 +PyTorch,qembeddingbag_4bit_prepack,qembeddingbag_4bit_prepack_num_embeddings80_embedding_dim128,short,FALSE,11.0042 +PyTorch,qembeddingbag_4bit_prepack,qembeddingbag_4bit_prepack_num_embeddings80_embedding_dim256,short,FALSE,11.1633 +PyTorch,qembeddingbag_4bit_prepack,qembeddingbag_4bit_prepack_num_embeddings80_embedding_dim512,short,FALSE,12.6725 +PyTorch,qembeddingbag_2bit_prepack,qembeddingbag_2bit_prepack_num_embeddings80_embedding_dim128,short,FALSE,11.0473 +PyTorch,qembeddingbag_2bit_prepack,qembeddingbag_2bit_prepack_num_embeddings80_embedding_dim256,short,FALSE,11.2128 +PyTorch,qembeddingbag_2bit_prepack,qembeddingbag_2bit_prepack_num_embeddings80_embedding_dim512,short,FALSE,11.5582 +PyTorch,qembeddingbag_byte_unpack,qembeddingbag_byte_unpack_num_embeddings80_embedding_dim128,short,FALSE,14.0764 +PyTorch,qembeddingbag_byte_unpack,qembeddingbag_byte_unpack_num_embeddings80_embedding_dim256,short,FALSE,16.5054 +PyTorch,qembeddingbag_byte_unpack,qembeddingbag_byte_unpack_num_embeddings80_embedding_dim512,short,FALSE,20.8127 +PyTorch,qembeddingbag_4bit_unpack,qembeddingbag_4bit_unpack_num_embeddings80_embedding_dim128,short,FALSE,11.9969 +PyTorch,qembeddingbag_4bit_unpack,qembeddingbag_4bit_unpack_num_embeddings80_embedding_dim256,short,FALSE,12.1914 +PyTorch,qembeddingbag_4bit_unpack,qembeddingbag_4bit_unpack_num_embeddings80_embedding_dim512,short,FALSE,12.4254 +PyTorch,qembeddingbag_2bit_unpack,qembeddingbag_2bit_unpack_num_embeddings80_embedding_dim128,short,FALSE,12.1991 +PyTorch,qembeddingbag_2bit_unpack,qembeddingbag_2bit_unpack_num_embeddings80_embedding_dim256,short,FALSE,12.3906 +PyTorch,qembeddingbag_2bit_unpack,qembeddingbag_2bit_unpack_num_embeddings80_embedding_dim512,short,FALSE,13.1629 +PyTorch,qembeddingbag_byte_prepack,qembeddingbag_byte_prepack_num_embeddings80_embedding_dim128_batch_size10,short,FALSE,12.8605 +PyTorch,qembeddingbag_byte_prepack,qembeddingbag_byte_prepack_num_embeddings80_embedding_dim256_batch_size10,short,FALSE,14.0776 +PyTorch,qembeddingbag_byte_prepack,qembeddingbag_byte_prepack_num_embeddings80_embedding_dim512_batch_size10,short,FALSE,16.2496 +PyTorch,qembeddingbag_4bit_prepack,qembeddingbag_4bit_prepack_num_embeddings80_embedding_dim128_batch_size10,short,FALSE,9.6159 +PyTorch,qembeddingbag_4bit_prepack,qembeddingbag_4bit_prepack_num_embeddings80_embedding_dim256_batch_size10,short,FALSE,9.5633 +PyTorch,qembeddingbag_4bit_prepack,qembeddingbag_4bit_prepack_num_embeddings80_embedding_dim512_batch_size10,short,FALSE,9.5626 +PyTorch,qembeddingbag_2bit_prepack,qembeddingbag_2bit_prepack_num_embeddings80_embedding_dim128_batch_size10,short,FALSE,9.6479 +PyTorch,qembeddingbag_2bit_prepack,qembeddingbag_2bit_prepack_num_embeddings80_embedding_dim256_batch_size10,short,FALSE,9.67 +PyTorch,qembeddingbag_2bit_prepack,qembeddingbag_2bit_prepack_num_embeddings80_embedding_dim512_batch_size10,short,FALSE,9.6473 +PyTorch,qembeddingbag_byte_unpack,qembeddingbag_byte_unpack_num_embeddings80_embedding_dim128_batch_size10,short,FALSE,27.9084 +PyTorch,qembeddingbag_byte_unpack,qembeddingbag_byte_unpack_num_embeddings80_embedding_dim256_batch_size10,short,FALSE,45.0191 +PyTorch,qembeddingbag_byte_unpack,qembeddingbag_byte_unpack_num_embeddings80_embedding_dim512_batch_size10,short,FALSE,75.4599 +PyTorch,qembeddingbag_4bit_unpack,qembeddingbag_4bit_unpack_num_embeddings80_embedding_dim128_batch_size10,short,FALSE,8.9655 +PyTorch,qembeddingbag_4bit_unpack,qembeddingbag_4bit_unpack_num_embeddings80_embedding_dim256_batch_size10,short,FALSE,8.9751 +PyTorch,qembeddingbag_4bit_unpack,qembeddingbag_4bit_unpack_num_embeddings80_embedding_dim512_batch_size10,short,FALSE,8.9735 +PyTorch,qembeddingbag_2bit_unpack,qembeddingbag_2bit_unpack_num_embeddings80_embedding_dim128_batch_size10,short,FALSE,9.1326 +PyTorch,qembeddingbag_2bit_unpack,qembeddingbag_2bit_unpack_num_embeddings80_embedding_dim256_batch_size10,short,FALSE,9.1411 +PyTorch,qembeddingbag_2bit_unpack,qembeddingbag_2bit_unpack_num_embeddings80_embedding_dim512_batch_size10,short,FALSE,9.1444 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.7564 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.3127 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.7029 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.2567 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.7698 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.3691 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.8276 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.3425 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.963 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.6502 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.9717 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags10_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.5513 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.8327 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.3891 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.836 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.3409 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.7967 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.3416 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.8597 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.4254 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,49.1124 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.4759 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,49.1267 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags120_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.6559 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.9441 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.486 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.8731 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.4419 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.8811 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.4126 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.8853 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.4946 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,49.1347 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.6574 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,49.0424 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags1000_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.6891 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,48.8887 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.3275 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.7541 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size8_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.4383 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,49.0772 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.4852 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,48.8794 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size16_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.4925 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetTrue_cpu,short,FALSE,49.0408 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseTrue_include_last_offsetFalse_cpu,short,FALSE,64.7088 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetTrue_cpu,short,FALSE,49.0971 +PyTorch,qEmbeddingBag,qEmbeddingBag_embeddingbags2300_dim64_modesum_input_size64_offset0_sparseFalse_include_last_offsetFalse_cpu,short,FALSE,64.7124 +PyTorch,QGroupNormBenchmark,"QGroupNormBenchmark_dims(32,8,16)_num_groups2_dtypetorch.qint8",short,FALSE,14.2907 +PyTorch,QGroupNormBenchmark,"QGroupNormBenchmark_dims(32,8,16)_num_groups4_dtypetorch.qint8",short,FALSE,14.4174 +PyTorch,QGroupNormBenchmark,"QGroupNormBenchmark_dims(32,8,56,56)_num_groups2_dtypetorch.qint8",short,FALSE,35.2416 +PyTorch,QGroupNormBenchmark,"QGroupNormBenchmark_dims(32,8,56,56)_num_groups4_dtypetorch.qint8",short,FALSE,35.5288 +PyTorch,QInstanceNormBenchmark,"QInstanceNormBenchmark_dims(32,8,16)_dtypetorch.qint8",short,FALSE,14.4557 +PyTorch,QInstanceNormBenchmark,"QInstanceNormBenchmark_dims(32,8,56,56)_dtypetorch.qint8",short,FALSE,33.8378 +PyTorch,q_interpolate,q_interpolate_M32_N32_K32_dtypetorch.quint8_modenearest_scale0.5_contigTrue,short,FALSE,7.8927 +PyTorch,q_interpolate,q_interpolate_M32_N32_K32_dtypetorch.quint8_modebilinear_scale0.5_contigTrue,short,FALSE,12.1855 +PyTorch,q_interpolate,q_interpolate_M32_N32_K32_dtypetorch.quint8_modenearest_scale2.0_contigTrue,short,FALSE,7.876 +PyTorch,q_interpolate,q_interpolate_M32_N32_K32_dtypetorch.quint8_modebilinear_scale2.0_contigTrue,short,FALSE,12.1806 +PyTorch,q_interpolate,q_interpolate_M3_N720_K1280_dtypetorch.quint8_modebilinear_scale0.83333_contigTrue,short,FALSE,31.3689 +PyTorch,QLayerNormBenchmark,"QLayerNormBenchmark_dims(1,8,16)_dtypetorch.qint8",short,FALSE,14.4236 +PyTorch,QLayerNormBenchmark,"QLayerNormBenchmark_dims(8,8,16)_dtypetorch.qint8",short,FALSE,19.3001 +PyTorch,QLayerNormBenchmark,"QLayerNormBenchmark_dims(32,8,16)_dtypetorch.qint8",short,FALSE,19.5875 +PyTorch,QLayerNormBenchmark,"QLayerNormBenchmark_dims(64,128,56,56)_dtypetorch.qint8",short,FALSE,1529.4168 +PyTorch,QLinear,QLinear_N1_IN1_OUT1_cpu,short,FALSE,61.9098 +PyTorch,QLinear,QLinear_N4_IN256_OUT128_cpu,short,FALSE,71.4256 +PyTorch,QLinear,QLinear_N16_IN512_OUT256_cpu,short,FALSE,76.5413 +PyTorch,QDynamicLinear,QDynamicLinear_N1_IN1_OUT1_cpu,short,FALSE,58.6741 +PyTorch,QDynamicLinear,QDynamicLinear_N4_IN256_OUT128_cpu,short,FALSE,68.1178 +PyTorch,QDynamicLinear,QDynamicLinear_N16_IN512_OUT256_cpu,short,FALSE,76.5539 +PyTorch,MinMaxObserver,MinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_affine,short,FALSE,165.7528 +PyTorch,MinMaxObserver,MinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_symmetric,short,FALSE,152.155 +PyTorch,MovingAverageMinMaxObserver,MovingAverageMinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_affine,short,FALSE,207.921 +PyTorch,MovingAverageMinMaxObserver,MovingAverageMinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_symmetric,short,FALSE,196.0801 +PyTorch,PerChannelMinMaxObserver,PerChannelMinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_channel_affine,short,FALSE,660.8751 +PyTorch,PerChannelMinMaxObserver,PerChannelMinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_channel_symmetric,short,FALSE,631.7678 +PyTorch,MovingAveragePerChannelMinMaxObserver,MovingAveragePerChannelMinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_channel_affine,short,FALSE,689.3469 +PyTorch,MovingAveragePerChannelMinMaxObserver,MovingAveragePerChannelMinMaxObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_channel_symmetric,short,FALSE,665.8384 +PyTorch,HistogramObserver,HistogramObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_affine,short,FALSE,2269.4954 +PyTorch,HistogramObserver,HistogramObserver_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_symmetric,short,FALSE,2284.8399 +PyTorch,HistogramObserverCalculateQparams,HistogramObserverCalculateQparams_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_affine,short,FALSE,2275.4481 +PyTorch,HistogramObserverCalculateQparams,HistogramObserverCalculateQparams_C3_M512_N512_dtypetorch.quint8_cpu_qschemetorch.per_tensor_symmetric,short,FALSE,2225.9528 +PyTorch,QAdaptiveAvgPool2dBenchmark,"QAdaptiveAvgPool2dBenchmark_N4_C3_input_size(224,224)_output_size(112,112)_contigTrue_dtypetorch.qint32",short,FALSE,219.4663 +PyTorch,QAdaptiveAvgPool2dBenchmark,"QAdaptiveAvgPool2dBenchmark_N4_C3_input_size(224,224)_output_size(112,112)_contigTrue_dtypetorch.qint8",short,FALSE,219.0055 +PyTorch,QAdaptiveAvgPool2dBenchmark,"QAdaptiveAvgPool2dBenchmark_N4_C3_input_size(224,224)_output_size(112,112)_contigTrue_dtypetorch.quint8",short,FALSE,218.8345 +PyTorch,QAvgPool2dBenchmark,"QAvgPool2dBenchmark_C1_H3_W3_k(3,3)_s(1,1)_p(0,0)_N2_contigTrue_dtypetorch.qint32",short,FALSE,12.7215 +PyTorch,QAvgPool2dBenchmark,"QAvgPool2dBenchmark_C1_H3_W3_k(3,3)_s(1,1)_p(0,0)_N2_contigTrue_dtypetorch.qint8",short,FALSE,14.6691 +PyTorch,QAvgPool2dBenchmark,"QAvgPool2dBenchmark_C1_H3_W3_k(3,3)_s(1,1)_p(0,0)_N2_contigTrue_dtypetorch.quint8",short,FALSE,14.6109 +PyTorch,QMaxPool2dBenchmark,"QMaxPool2dBenchmark_C1_H3_W3_k(3,3)_s(1,1)_p(0,0)_N2_contigTrue_dtypetorch.qint32",short,FALSE,15.4468 +PyTorch,QMaxPool2dBenchmark,"QMaxPool2dBenchmark_C1_H3_W3_k(3,3)_s(1,1)_p(0,0)_N2_contigTrue_dtypetorch.qint8",short,FALSE,15.4546 +PyTorch,QMaxPool2dBenchmark,"QMaxPool2dBenchmark_C1_H3_W3_k(3,3)_s(1,1)_p(0,0)_N2_contigTrue_dtypetorch.quint8",short,FALSE,15.4778 +PyTorch,QLSTM,QLSTM_I1_H3_NL1_BTrue_DFalse_dtypetorch.qint8,short,FALSE,5217.2835 +PyTorch,QLSTM,QLSTM_I1_H3_NL1_BTrue_DTrue_dtypetorch.qint8,short,FALSE,10347.2789 +PyTorch,QLSTM,QLSTM_I5_H7_NL4_BTrue_DFalse_dtypetorch.qint8,short,FALSE,24447.7776 +PyTorch,QLSTM,QLSTM_I5_H7_NL4_BTrue_DTrue_dtypetorch.qint8,short,FALSE,48972.4577 +PyTorch,QMethodTensorInputCopyBenchmark,QMethodTensorInputCopyBenchmark_M32_N32_dtypetorch.quint8_contigFalse,short,FALSE,1.041 +PyTorch,QMethodTensorInputCopyBenchmark,QMethodTensorInputCopyBenchmark_M32_N32_dtypetorch.quint8_contigTrue,short,FALSE,1.022 +PyTorch,QuantizePerTensor,QuantizePerTensor_C3_M512_N512_dtypetorch.quint8_modeQ,short,FALSE,25.1494 +PyTorch,DequantizePerTensor,DequantizePerTensor_C3_M512_N512_dtypetorch.quint8_modeD,short,FALSE,17.1969 +PyTorch,QuantizePerChannel,QuantizePerChannel_C3_M512_N512_dtypetorch.quint8_modeQ_axis0,short,FALSE,3162.223 +PyTorch,DequantizePerChannel,DequantizePerChannel_C3_M512_N512_dtypetorch.quint8_modeD_axis0,short,FALSE,185.0221 +PyTorch,FakeQuantize,FakeQuantize_N1_C3_H512_W512_zero_point_dtypetorch.int32_cpu,short,FALSE,546.7611 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu,short,FALSE,198.838 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu,short,FALSE,208.628 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu,short,FALSE,200.6107 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu,short,FALSE,211.8795 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwdall_BACKWARD,short,TRUE,415.3363 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd1_BACKWARD,short,TRUE,416.6851 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd2_BACKWARD,short,TRUE,416.4106 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd3_BACKWARD,short,TRUE,416.1216 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwdall_BACKWARD,short,TRUE,417.0081 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd1_BACKWARD,short,TRUE,416.5927 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd2_BACKWARD,short,TRUE,417.6604 +PyTorch,learnable_kernel_tensor,learnable_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd3_BACKWARD,short,TRUE,416.0931 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwdall_BACKWARD,short,TRUE,183.1625 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd1_BACKWARD,short,TRUE,183.596 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd2_BACKWARD,short,TRUE,183.0808 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd3_BACKWARD,short,TRUE,182.9406 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwdall_BACKWARD,short,TRUE,183.9915 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd1_BACKWARD,short,TRUE,183.407 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd2_BACKWARD,short,TRUE,182.8545 +PyTorch,original_kernel_tensor,original_kernel_tensor_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd3_BACKWARD,short,TRUE,183.1087 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu,short,FALSE,309.1467 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu,short,FALSE,312.9401 +PyTorch,original_kernel_channel,original_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu,short,FALSE,300.9107 +PyTorch,original_kernel_channel,original_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu,short,FALSE,308.721 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwdall_BACKWARD,short,TRUE,542.2402 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd1_BACKWARD,short,TRUE,544.185 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd2_BACKWARD,short,TRUE,542.8632 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd3_BACKWARD,short,TRUE,543.9898 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwdall_BACKWARD,short,TRUE,544.0337 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd1_BACKWARD,short,TRUE,544.0846 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd2_BACKWARD,short,TRUE,543.6945 +PyTorch,learnable_kernel_channel,learnable_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd3_BACKWARD,short,TRUE,542.707 +PyTorch,original_kernel_channel,original_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwdall_BACKWARD,short,TRUE,182.4091 +PyTorch,original_kernel_channel,original_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits4_cpu_bwd1_BACKWARD,short,TRUE,183.0807 +PyTorch,original_kernel_channel,original_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwdall_BACKWARD,short,TRUE,183.1147 +PyTorch,original_kernel_channel,original_kernel_channel_N1_C3_H512_W512_zero_point_dtypetorch.int32_nbits8_cpu_bwd1_BACKWARD,short,TRUE,182.72 +PyTorch,q_argsort,q_argsort_M512_N512_dtypetorch.quint8,short,FALSE,446.4263 +PyTorch,q_clone,q_clone_M512_N512_dtypetorch.quint8,short,FALSE,10.9374 +PyTorch,q_mean,q_mean_M512_N512_dtypetorch.quint8,short,FALSE,10.2288 +PyTorch,q_relu,q_relu_M512_N512_dtypetorch.quint8,short,FALSE,10.3366 +PyTorch,q_relu_,q_relu__M512_N512_dtypetorch.quint8,short,FALSE,25.3594 +PyTorch,q_sort,q_sort_M512_N512_dtypetorch.quint8,short,FALSE,447.1303 +PyTorch,qtopk,qtopk_M512_N512_k5_dtypetorch.quint8,short,FALSE,64.856 +PyTorch,abs,abs_M512_N512_cpu,short,FALSE,12.3046 +PyTorch,abs_,abs__M512_N512_cpu,short,FALSE,7.638213467 +PyTorch,acos,acos_M512_N512_cpu,short,FALSE,18.7028 +PyTorch,acos_,acos__M512_N512_cpu,short,FALSE,65.8008 +PyTorch,argsort,argsort_M512_N512_cpu,short,FALSE,1424.792 +PyTorch,asin,asin_M512_N512_cpu,short,FALSE,17.6292 +PyTorch,asin_,asin__M512_N512_cpu,short,FALSE,13.7757 +PyTorch,atan,atan_M512_N512_cpu,short,FALSE,17.7172 +PyTorch,atan_,atan__M512_N512_cpu,short,FALSE,13.9644 +PyTorch,ceil,ceil_M512_N512_cpu,short,FALSE,11.1606 +PyTorch,ceil_,ceil__M512_N512_cpu,short,FALSE,7.4759 +PyTorch,clone,clone_M512_N512_cpu,short,FALSE,12.2572 +PyTorch,cos,cos_M512_N512_cpu,short,FALSE,18.5237 +PyTorch,cos_,cos__M512_N512_cpu,short,FALSE,14.7932 +PyTorch,cosh,cosh_M512_N512_cpu,short,FALSE,80.2281 +PyTorch,digamma,digamma_M512_N512_cpu,short,FALSE,672.0418 +PyTorch,erf,erf_M512_N512_cpu,short,FALSE,20.1027 +PyTorch,erf_,erf__M512_N512_cpu,short,FALSE,16.4605 +PyTorch,erfc,erfc_M512_N512_cpu,short,FALSE,21.6993 +PyTorch,erfc_,erfc__M512_N512_cpu,short,FALSE,17.3411 +PyTorch,erfinv,erfinv_M512_N512_cpu,short,FALSE,20.9871 +PyTorch,exp,exp_M512_N512_cpu,short,FALSE,14.1471 +PyTorch,exp_,exp__M512_N512_cpu,short,FALSE,52.1716 +PyTorch,expm1,expm1_M512_N512_cpu,short,FALSE,32.8849 +PyTorch,expm1_,expm1__M512_N512_cpu,short,FALSE,28.8886 +PyTorch,floor,floor_M512_N512_cpu,short,FALSE,11.1898 +PyTorch,floor_,floor__M512_N512_cpu,short,FALSE,7.2972 +PyTorch,frac,frac_M512_N512_cpu,short,FALSE,12.9296 +PyTorch,frac_,frac__M512_N512_cpu,short,FALSE,9.1193 +PyTorch,hardshrink,hardshrink_M512_N512_cpu,short,FALSE,14.6546 +PyTorch,lgamma,lgamma_M512_N512_cpu,short,FALSE,110.4636 +PyTorch,log,log_M512_N512_cpu,short,FALSE,14.8297 +PyTorch,log10,log10_M512_N512_cpu,short,FALSE,15.6434 +PyTorch,log10_,log10__M512_N512_cpu,short,FALSE,11.8294 +PyTorch,log1p,log1p_M512_N512_cpu,short,FALSE,27.0109 +PyTorch,log1p_,log1p__M512_N512_cpu,short,FALSE,23.1485 +PyTorch,log2,log2_M512_N512_cpu,short,FALSE,15.3609 +PyTorch,log2_,log2__M512_N512_cpu,short,FALSE,11.5224 +PyTorch,log_,log__M512_N512_cpu,short,FALSE,81.0499 +PyTorch,logit,logit_M512_N512_cpu,short,FALSE,18.1755 +PyTorch,logit_,logit__M512_N512_cpu,short,FALSE,83.57725 +PyTorch,neg,neg_M512_N512_cpu,short,FALSE,11.1491 +PyTorch,neg_,neg__M512_N512_cpu,short,FALSE,7.4216 +PyTorch,reciprocal,reciprocal_M512_N512_cpu,short,FALSE,16.1436 +PyTorch,reciprocal_,reciprocal__M512_N512_cpu,short,FALSE,12.59 +PyTorch,relu,relu_M512_N512_cpu,short,FALSE,13.1991 +PyTorch,relu_,relu__M512_N512_cpu,short,FALSE,8.7194 +PyTorch,round,round_M512_N512_cpu,short,FALSE,11.1888 +PyTorch,round_,round__M512_N512_cpu,short,FALSE,7.4217 +PyTorch,rsqrt,rsqrt_M512_N512_cpu,short,FALSE,20.9455 +PyTorch,rsqrt_,rsqrt__M512_N512_cpu,short,FALSE,17.8695 +PyTorch,sigmoid,sigmoid_M512_N512_cpu,short,FALSE,32.1797 +PyTorch,sigmoid_,sigmoid__M512_N512_cpu,short,FALSE,28.0707 +PyTorch,sign,sign_M512_N512_cpu,short,FALSE,13.2475 +PyTorch,sgn,sgn_M512_N512_cpu,short,FALSE,13.1844 +PyTorch,sin,sin_M512_N512_cpu,short,FALSE,18.4476 +PyTorch,sin_,sin__M512_N512_cpu,short,FALSE,14.5837 +PyTorch,sinh,sinh_M512_N512_cpu,short,FALSE,81.4856 +PyTorch,sqrt,sqrt_M512_N512_cpu,short,FALSE,12.4782 +PyTorch,sqrt_,sqrt__M512_N512_cpu,short,FALSE,9.536 +PyTorch,square,square_M512_N512_cpu,short,FALSE,15.1528 +PyTorch,square_,square__M512_N512_cpu,short,FALSE,10.2758 +PyTorch,tan,tan_M512_N512_cpu,short,FALSE,20.1579 +PyTorch,tan_,tan__M512_N512_cpu,short,FALSE,16.0497 +PyTorch,tanh,tanh_M512_N512_cpu,short,FALSE,14.9006 +PyTorch,tanh_,tanh__M512_N512_cpu,short,FALSE,11.5883 +PyTorch,trunc,trunc_M512_N512_cpu,short,FALSE,12.7292 +PyTorch,trunc_,trunc__M512_N512_cpu,short,FALSE,8.5329 +PyTorch,unique,unique_M512_N512_cpu,short,FALSE,21486.0475 +PyTorch,zero_,zero__M512_N512_cpu,short,FALSE,7.4783 +PyTorch,bernoulli_,bernoulli__M512_N512_cpu,short,FALSE,3986.396 +PyTorch,cauchy_,cauchy__M512_N512_cpu,short,FALSE,7504.3189 +PyTorch,digamma_,digamma__M512_N512_cpu,short,FALSE,1276.7406 +PyTorch,exponential_,exponential__M512_N512_cpu,short,FALSE,6277.716 +PyTorch,normal_,normal__M512_N512_cpu,short,FALSE,1334.7115 +PyTorch,random_,random__M512_N512_cpu,short,FALSE,1209.5657 +PyTorch,sign_,sign__M512_N512_cpu,short,FALSE,9.3767 +PyTorch,uniform_,uniform__M512_N512_cpu,short,FALSE,1248.3519 +PyTorch,half,half_M512_N512_cpu,short,FALSE,14.2058 +PyTorch,long,long_M512_N512_cpu,short,FALSE,28.371 +PyTorch,arange,arange_start0_end1000_step2.5_cpu_dtypetorch.float32,short,FALSE,7.791 +PyTorch,arange,arange_start-1024_end2048_step1_cpu_dtypetorch.float32,short,FALSE,9.501666667 +PyTorch,add_,add__M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,2.428333333 +PyTorch,add_,add__M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,10.516 +PyTorch,add_,add__M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,10.43566667 +PyTorch,sub_,sub__M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,2.567 +PyTorch,sub_,sub__M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,10.36466667 +PyTorch,sub_,sub__M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,10.44266667 +PyTorch,mul_,mul__M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,2.462 +PyTorch,mul_,mul__M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,9.666666667 +PyTorch,mul_,mul__M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,9.804666667 +PyTorch,div_,div__M1_N1_K1_cpu_dtype_onetorch.float32_dtype_twotorch.float32,short,FALSE,2.431333333 +PyTorch,div_,div__M64_N64_K64_cpu_dtype_onetorch.float32_dtype_twotorch.float32,short,FALSE,192.4236667 +PyTorch,div_,div__M64_N64_K128_cpu_dtype_onetorch.float32_dtype_twotorch.float32,short,FALSE,194.4426667 +PyTorch,sub,"sub_in_one[64,1,64]_in_two[1,64,1]_cpu_dtypetorch.float32",short,FALSE,14.90066667 +PyTorch,div,"div_in_one[64,1,64]_in_two[1,64,1]_cpu_dtypetorch.float32",short,FALSE,14.753 +PyTorch,mul,"mul_in_one[64,1,64]_in_two[1,64,1]_cpu_dtypetorch.float32",short,FALSE,14.34233333 +PyTorch,sub,sub_M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,4.091333333 +PyTorch,sub,sub_M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,14.359 +PyTorch,sub,sub_M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,14.472 +PyTorch,div,div_M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,9.372333333 +PyTorch,div,div_M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,38.91933333 +PyTorch,div,div_M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,39.602 +PyTorch,mul,mul_M1_N1_K1_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,3.932 +PyTorch,mul,mul_M64_N64_K64_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,13.98933333 +PyTorch,mul,mul_M64_N64_K128_cpu_dtype_onetorch.int32_dtype_twotorch.int32,short,FALSE,14.05733333 +PyTorch,logical_and,"logical_and_in_one[64,1,64]_in_two[1,64,1]_cpu_dtypetorch.bool",short,FALSE,41.10566667 +PyTorch,logical_and,logical_and_M1_N1_K1_cpu_dtype_onetorch.bool_dtype_twotorch.bool,short,FALSE,4.987666667 +PyTorch,logical_and,logical_and_M64_N64_K64_cpu_dtype_onetorch.bool_dtype_twotorch.bool,short,FALSE,10.88633333 +PyTorch,logical_and,logical_and_M64_N64_K128_cpu_dtype_onetorch.bool_dtype_twotorch.bool,short,FALSE,10.844 +PyTorch,bmm,bmm_B2_M1_N8_K2_cpu_dtypetorch.float32,short,FALSE,4.132666667 +PyTorch,bmm,bmm_B2_M1_N8_K2_cpu_dtypetorch.bfloat16,short,FALSE,4.138 +PyTorch,bmm,bmm_B128_M64_N32_K64_cpu_dtypetorch.float32,short,FALSE,26.016 +PyTorch,bmm,bmm_B128_M64_N32_K64_cpu_dtypetorch.bfloat16,short,FALSE,122.5856667 +PyTorch,baddbmm,baddbmm_B2_M1_N8_K2_cpu_dtypetorch.float32,short,FALSE,6.079333333 +PyTorch,baddbmm,baddbmm_B2_M1_N8_K2_cpu_dtypetorch.bfloat16,short,FALSE,6.154666667 +PyTorch,baddbmm,baddbmm_B128_M64_N32_K64_cpu_dtypetorch.float32,short,FALSE,46.773 +PyTorch,baddbmm,baddbmm_B128_M64_N32_K64_cpu_dtypetorch.bfloat16,short,FALSE,138.773 +PyTorch,index_add_,index_add__M8_N32_K1_dim0_cpu_dtypetorch.float32,short,FALSE,40.94966667 +PyTorch,index_add_,index_add__M512_N512_K1_dim2_cpu_dtypetorch.float32,short,FALSE,124.4306667 +PyTorch,index_select,index_select_M8_N8_K1_dim1_cpu,short,FALSE,4.461 +PyTorch,index_select,index_select_M256_N512_K1_dim1_cpu,short,FALSE,79.63533333 +PyTorch,index_select,index_select_M512_N512_K1_dim1_cpu,short,FALSE,167.888 +PyTorch,index_select,index_select_M8_N8_K2_dim1_cpu,short,FALSE,4.544 +PyTorch,index_select,index_select_M256_N512_K2_dim1_cpu,short,FALSE,305.2126667 +PyTorch,index_select,index_select_M512_N512_K2_dim1_cpu,short,FALSE,603.7596667 +PyTorch,mm,mm_M1_N1_K1_cpu_dtypetorch.float32,short,FALSE,4.919 +PyTorch,mm,mm_M64_N64_K64_cpu_dtypetorch.float32,short,FALSE,13.90066667 +PyTorch,mm,mm_M64_N64_K128_cpu_dtypetorch.float32,short,FALSE,13.20866667 +PyTorch,stack,"stack_sizes(1,1,1)_N2_cpu_dim0",short,FALSE,5.519 +PyTorch,stack,"stack_sizes(1,1,1)_N2_cpu_dim1",short,FALSE,5.738333333 +PyTorch,stack,"stack_sizes(1,1,1)_N2_cpu_dim2",short,FALSE,5.746666667 +PyTorch,stack,"stack_sizes(1,1,1)_N2_cpu_dim3",short,FALSE,6.127 +PyTorch,stack,"stack_sizes(512,512,2)_N2_cpu_dim0",short,FALSE,19.943 +PyTorch,stack,"stack_sizes(512,512,2)_N2_cpu_dim1",short,FALSE,20.89633333 +PyTorch,stack,"stack_sizes(512,512,2)_N2_cpu_dim2",short,FALSE,116.437 +PyTorch,stack,"stack_sizes(512,512,2)_N2_cpu_dim3",short,FALSE,42.9705 +PyTorch,stack,"stack_sizes(128,1024,2)_N2_cpu_dim0",short,FALSE,19.77333333 +PyTorch,stack,"stack_sizes(128,1024,2)_N2_cpu_dim1",short,FALSE,20.22633333 +PyTorch,stack,"stack_sizes(128,1024,2)_N2_cpu_dim2",short,FALSE,116.23 +PyTorch,stack,"stack_sizes(128,1024,2)_N2_cpu_dim3",short,FALSE,42.9005 +PyTorch,addcmul,addcmul_M1_N2_cpu_dtypetorch.float32,short,FALSE,4.179 +PyTorch,addcmul,addcmul_M1_N2_cpu_dtypetorch.bfloat16,short,FALSE,4.164666667 +PyTorch,addcmul,addcmul_M32_N64_cpu_dtypetorch.float32,short,FALSE,4.956333333 +PyTorch,addcmul,addcmul_M32_N64_cpu_dtypetorch.bfloat16,short,FALSE,4.974666667 +PyTorch,addcdiv,addcdiv_M1_N2_cpu_dtypetorch.float32,short,FALSE,4.309333333 +PyTorch,addcdiv,addcdiv_M1_N2_cpu_dtypetorch.bfloat16,short,FALSE,4.338 +PyTorch,addcdiv,addcdiv_M32_N64_cpu_dtypetorch.float32,short,FALSE,5.157666667 +PyTorch,addcdiv,addcdiv_M32_N64_cpu_dtypetorch.bfloat16,short,FALSE,5.259333333 +PyTorch,topk,"topk_shape(16,4)_k4_dim1_cpu_dtypetorch.float32",short,FALSE,7.123333333 +PyTorch,topk,"topk_shape(1048576,)_k16_dim0_cpu_dtypetorch.float32",short,FALSE,2569.296333 +PyTorch,where,"where_cond_shape(8,16,1)_input_shape(1,)_other_shape(1,)_cpu_dtypetorch.float32",short,FALSE,5.682333333 +PyTorch,where,"where_cond_shape(8,16,1)_input_shape(16,1)_other_shape(8,16,1)_cpu_dtypetorch.float32",short,FALSE,5.763 +PyTorch,where,"where_cond_shape(8,16,1)_input_shape(8,1,1)_other_shape(1,)_cpu_dtypetorch.float32",short,FALSE,5.744666667 +PyTorch,clamp,clamp_M512_N512_cpu,short,FALSE,15.26233333 +PyTorch,gelu,gelu_M512_N512_cpu,short,FALSE,31.33166667 \ No newline at end of file diff --git a/benchmarks/operator_benchmark/pt/nan_to_num_test.py b/benchmarks/operator_benchmark/pt/nan_to_num_test.py index 77aa92464496..8b848fd85928 100644 --- a/benchmarks/operator_benchmark/pt/nan_to_num_test.py +++ b/benchmarks/operator_benchmark/pt/nan_to_num_test.py @@ -43,6 +43,10 @@ class ReplaceNaNBenchmark(op_bench.TorchBenchmarkBase): self.op_func = op_func self.set_module_name("nan_to_num") + # To make casename unique as nan_to_num and nan_to_num_ are two different functions. + if op_func is torch.nan_to_num_: + self.set_module_name("nan_to_num_") + def forward(self, input, replace_inf: bool): # compare inplace if replace_inf: diff --git a/benchmarks/operator_benchmark/pt/quantization_test.py b/benchmarks/operator_benchmark/pt/quantization_test.py index e0d3483963af..10515864c4a7 100644 --- a/benchmarks/operator_benchmark/pt/quantization_test.py +++ b/benchmarks/operator_benchmark/pt/quantization_test.py @@ -193,8 +193,8 @@ def fakeQuantizePerTensorOriginalKernel( fake_quantize_per_tensor_ops = op_bench.op_list( attrs=( - ("learnable_kernel", fakeQuantizePerTensorLearnableKernel), - ("original_kernel", fakeQuantizePerTensorOriginalKernel), + ("learnable_kernel_tensor", fakeQuantizePerTensorLearnableKernel), + ("original_kernel_tensor", fakeQuantizePerTensorOriginalKernel), ), attr_names=("op_name", "op_func"), ) @@ -297,8 +297,8 @@ def fakeQuantizePerChannelOriginalKernel( fake_quantize_per_channel_ops = op_bench.op_list( attrs=( - ("learnable_kernel", fakeQuantizePerChannelLearnableKernel), - ("original_kernel", fakeQuantizePerChannelOriginalKernel), + ("learnable_kernel_channel", fakeQuantizePerChannelLearnableKernel), + ("original_kernel_channel", fakeQuantizePerChannelOriginalKernel), ), attr_names=("op_name", "op_func"), )