[Inductor] Flex attention supports dynamic shape #125994
Dr.CI classification results
{"FAILED":[{"workflowId":9086945112,"workflowUniqueId":16521569,"id":24974748384,"runnerName":"i-0fea5b265758c3a4d","authorEmail":"ybliang8@gmail.com","name":"pull / linux-focal-cuda12.1-py3.10-gcc9-sm86 / test (default, 5, 5, linux.g5.4xlarge.nvidia.gpu)","jobName":"linux-focal-cuda12.1-py3.10-gcc9-sm86 / test (default, 5, 5, linux.g5.4xlarge.nvidia.gpu)","conclusion":"failure","completed_at":"2024-05-14T23:52:26Z","html_url":"https://github.com/pytorch/pytorch/actions/runs/9086945112/job/24974748384","head_branch":"flex-dyn","pr_number":125994,"head_sha":"501454356d00aa6fe46c7298c1ae6ec396102cc1","head_sha_timestamp":"2024-05-14T15:13:53-07:00","failure_captures":["No space left on device : '/home/ec2-user/actions-runner/_work/_temp/b176d4fc-6ef7-4515-aa8c-b260bcf36613.sh'"],"failure_lines":["##[error]No space left on device : '/home/ec2-user/actions-runner/_work/_temp/b176d4fc-6ef7-4515-aa8c-b260bcf36613.sh'"],"failure_context":["+ python3 .github/scripts/get_workflow_job_id.py 9086945112 i-0fea5b265758c3a4d","+78955f5f9ddb3601566884ce217d91c132f5edc1 third_party/zstd (v1.4.7-2183-g78955f5f)","+bb1473a4b38b18268e8693044afdb8635bc8351b third_party/tensorpipe (heads/main)","+54353f7714b82c854e4737e35aefd406b8d14f2f third_party/tbb (v2021.11.0-src-rc1-88-g54353f77)","+492a2fa49b44f852ee870655c5ecdc4706f97aab third_party/sleef (3.6-16-g492a2fa)","+913d74c35a6b1d330e90bfc055208ce5b06b35a0 third_party/python-peachpy (heads/master)","+aa98d95717d1365b5c32d3942abd0292fbd19021 third_party/pybind11 (v2.11.0-102-gaa98d957)","+178e3e0646cc671708bf78e77c273940130ac637 third_party/pthreadpool (0.1-146-g178e3e0)","+df630fa6f15c4b6581d6e8c24b76be420c8367ef third_party/protobuf (v27-dev-185-gdf630fa6f)","+b557a3519ccc1e36b74dc0901a073dd7872c0af2 third_party/pocketfft (release_for_eigen-26-gb557a35)","+ce14bf6520a059ff0bca81df45603722d965ad7c third_party/opentelemetry-cpp (v1.14.2-42-gce14bf65)","+7ecb49a435bd881b9ac4011450315192885e5cc3 third_party/onnx-tensorrt (release/8.6-GA-9-g7ecb49a)"],"time":"2024-05-15T00:01:17.697360Z"}],"FLAKY":[{"workflowId":9086947810,"workflowUniqueId":37481293,"id":24974820236,"runnerName":"i-0eeadcda5077d0b5d","authorEmail":"ybliang8@gmail.com","name":"inductor / cuda12.1-py3.10-gcc9-sm86 / test (dynamic_inductor_timm, 2, 2, linux.g5.4xlarge.nvidia.gpu)","jobName":"cuda12.1-py3.10-gcc9-sm86 / test (dynamic_inductor_timm, 2, 2, linux.g5.4xlarge.nvidia.gpu)","conclusion":"failure","completed_at":"2024-05-15T01:42:40Z","html_url":"https://github.com/pytorch/pytorch/actions/runs/9086947810/job/24974820236","head_branch":"ciflow/inductor/125994","pr_number":125994,"head_sha":"501454356d00aa6fe46c7298c1ae6ec396102cc1","head_sha_timestamp":"2024-05-14T15:13:53-07:00","failure_captures":["sebotnet33ts_256","FAIL: accuracy=fail_accuracy, expected=pass"],"failure_lines":["sebotnet33ts_256 FAIL: accuracy=fail_accuracy, expected=pass"],"failure_context":["+ python benchmarks/dynamo/check_accuracy.py --actual /var/lib/jenkins/workspace/test/test-reports/training_timm_models.csv --expected benchmarks/dynamo/ci_expected_accuracy/dynamic_inductor_timm_training.csv","+ python benchmarks/dynamo/timm_models.py --ci --accuracy --timing --explain --inductor --dynamic-shapes --dynamic-batch-only --device cuda --training --amp --total-partitions 2 --partition-id 1 --output /var/lib/jenkins/workspace/test/test-reports/training_timm_models.csv","+ [[ dynamic_inductor_timm == aot_inductor ]]","+ [[ dynamic_inductor_timm == perf ]]","+ [[ dynamic_inductor_timm == perf_compare ]]","+ partition_flags=(--total-partitions "$NUM_TEST_SHARDS" --partition-id "$shard_id")","+ [[ -n 1 ]]","+ [[ -n 2 ]]","+ local partition_flags","+ partition_flags=()","+ shift","+ local shard_id=1"],"time":"2024-05-15T01:42:43.155697Z"}],"BROKEN_TRUNK":[],"UNSTABLE":[]}