add dockerfile.bitModel to download bit models #115
Assignees
Labels
Comments
9 tasks
vli11
added a commit
that referenced
this issue
Jun 20, 2023
* feat: Added dockerfile.bitModel to download bit models closes: #115 Signed-off-by: Valina Li <valina.li@intel.com>
jim-wang-intel
pushed a commit
to jim-wang-intel/vision-self-checkout
that referenced
this issue
Jun 30, 2023
) * feat: Added dockerfile.bitModel to download bit models closes: intel-retail#115 Signed-off-by: Valina Li <valina.li@intel.com>
jim-wang-intel
added a commit
that referenced
this issue
Jun 30, 2023
* Ovms integration (#59) * feat: Renamed to Automated self checkout (#62) Closes: issue #61 Signed-off-by: Valina Li <valina.li@intel.com> * Benchmark docker (#26) * feat: Dockerfile for benchmark scripts * Updated to support --workload option Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Initial commit for supporting opencv-ovms workload Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Integration for ovms Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Updated docker image to openvino/model_server:2022.3.0.1-gpu Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Update configs/opencv-ovms/scripts/grpc_infer_binary_bit.py Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> * updated client image Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Add script for downloading the sample image Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Name fix for container Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Updated copyright and exit code Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> --------- Signed-off-by: Valina Li <valina.li@intel.com> Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> Co-authored-by: Valina Li <valina.li@intel.com> Co-authored-by: brian-intel <brian.mcginn@intel.com> Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> * feat: Enable pipeline failure log (#56) * feat: Enable pipeline failure log Closes: issue #41 Signed-off-by: Valina Li <valina.li@intel.com> * feat: enhance stream density function (#54) * feat: enhance stream density function Now the increment of pipeline numbers can be larger if the system is high power and performed well: The increment number can be taken from user's input or can be dynamically adjusted based the actual average FPS per stream and the target FPS. The searching policy is greedy to attempt to get to the maximum pipelines first and then decrement by 1 at a time to find the maximum number of pipelines reaching the target FPS. Closes: #51 * feat: update docker-run-dlstreamer.sh for stream density increments env * feat: update the increments logic and also add logic to cleanup the child processes of pipeline parent process * fix: resolve the hung issue caused by --rm as exited containers should be removed first before checking --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * feat: Add github actions (#70) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * feat: enhance the benchmark script to handle --workload option in any argument position (#76) Now if --workload is placed in front of pipeline options, it will work. Also works if it is placed in-between or after the pipeline options. Closes: #67 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * fix: Update shecllcheck minimum severity to warning (#82) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: Remove actoins/checkout from reviewdog (#83) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: fix shellcheck script issues (#81) * fix: fix shellcheck SC2167 * wipfix: use mapfile to better deal with multiple line output in commnad line * fix: fix shellcheck issue for benchmark.sh * fix: fix shellcheck issues for camera-simulator.sh and cleanup_gpu_metrics.sh * fix: fix shellcheck issues for copy-platform-metrics.sh * fix: fix shellcheck issues for format_avc_mp4.sh * fix: fix shellcheck issues for get-gpu-info.sh * fix: remove unused run.sh and run_server.sh under benchmark-scripts * fix: fix shellcheck issues for benchmark-scripts/stop_server.sh * fix: fix shellcheck issues for benchmark-scripts/stream_density.sh * fix: fix shellcheck issues for benchmark-scripts/test_format_avc_mp4.sh * fix: fix shellcheck issues for camera-simulator/camera-simulator.sh * fix: fix some boundary case when log_time_monitor_pid may not be there any more * fix: reoslve the issue of stop_server.sh when it is meant to be re-splitting for docker rm to be working * fix: use force kill to clean up its processes as sometimes it stuck with pkill --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * fix: fixed docker-run.sh with wrong message when no --workload option (#87) * fix: fixed docker-run.sh with wrong message when no --workload option Signed-off-by: Valina Li <valina.li@intel.com> * feat: Add model download top level script and add unit test for ovms … (#86) * feat: Add model download top level script and add unit test for ovms model download Signed-off-by: Valina Li <valina.li@intel.com> * Update reviewdog.yaml added new env token * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Feat: adding core smoke test (#93) * Fix: update collect platform metrics to handle multiple gpu devices (#94) * Update reviewdog.yaml * Update .reviewdog.yml * Fix: invert the format avc mp4 script tag logic to match our expected tag (#95) * Docs: updates for release 1.0.1 (#97) * fix: resolve the issue of the current model downloader (#98) * fix: resolve the issue of the current model downloader current getModel.sh script and other model downloading related scripts are all using relative path like . or .., which is not working when the scripts are running in different directories. This PR resolves the issue of using relative path and uses the absolute path to run via finding out the execution path on the runtime. Fixes: #96 * fix: remove check point model files and update the model downloader Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * feat: add trivy code scan (#102) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: Remove old linting workflows as reviewdog runs them (#103) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update reviewdog.yaml fix golangci-lint install * Update reviewdog.yaml fix: create new directory for golangci-lint install * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * fix reviewdog golangcit * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update .reviewdog.yml fix: Add shellcheck to reviewdog and fixed hadolint cmd call * feat: Add cleaning ovms containers to makefile (#101) feat: Add cleaning ovms containers to makefile closes: #100 Signed-off-by: Valina Li <valina.li@intel.com> * Update .reviewdog.yml * Create .golangci.yml feat: Add golangci config to .github/ * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update .reviewdog.yml * feat: Made CPU as default device for ovms pipeline (#114) closes: #112 Signed-off-by: Valina Li <valina.li@intel.com> * Fix: igt path pointing to the incorrect directory causing the igt log to not be written (#110) Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * Update gotest.yaml * Update reviewdog.yaml * fix: move workflow config files to .github directory (#121) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update codeql.yaml * fix: fixed pipelinesetup doc has incorect link to models.list.yml (#120) closes: #119 Signed-off-by: Valina Li <valina.li@intel.com> * feat: Added dockerfile.bitModel to download bit models (#116) * feat: Added dockerfile.bitModel to download bit models closes: #115 Signed-off-by: Valina Li <valina.li@intel.com> * Fix: pcm directory inconsistent between dockerfile and run script (#107) * Merge doc changes (#118) * Docs: merge from the 1.0.1 changes * only add new commands * feat: added OVMS sample image download into run script (#125) Closes: #124 Signed-off-by: Valina Li <valina.li@intel.com> * docs: Updated doc to include OVMS info (#126) * docs: Updated doc to include OVMS info Closes: #65 Signed-off-by: Valina Li <valina.li@intel.com> * Update codeql.yaml fix: there are no cpp files in the project * fix: update license to apache-2.0 (#130) * fix: update license to apache-2.0 Closes: #129 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * docs: Updated mkdoc to navigate to OVMS docs (#132) closes: #131 Signed-off-by: Valina Li <valina.li@intel.com> * feat: clean up checked-in dlstreamer models (#133) Model downloader can download those dlstreamer models, so we don't check in those model files. Closes: #99 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * feat: add docker image build to workflows (#137) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update build.yaml fix: syntax error * Update build.yaml fix: Remove OVMS server build * feat: add platform and CPU_ONLY parameters to enable smoke testing on all platform types (#136) * fix: add platform and CPU_ONLY parameters to enable smoke testing on all platform types --------- Signed-off-by: Brian McGinn <brian.mcginn@intel.com> Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> * fix: resolve hadolint issues (#140) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update reviewdog.yaml fix: Create directory for storing results for artifact * Update reviewdog.yaml fix: use absolute path * Update reviewdog.yaml * Update reviewdog.yaml * fix: Save scan results to artifact (#141) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: resolve the make build-ovms-server rebuild issue (#143) Fixes: #142 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * fix: move sample output label to correct location and fix json format (#146) Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * docs: point model list to github model file (#148) Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * fix: update the model list URL in pipelinesetup.md Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> --------- Signed-off-by: Valina Li <valina.li@intel.com> Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> Signed-off-by: Brian McGinn <brian.mcginn@intel.com> Co-authored-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> Co-authored-by: Valina Li <valina.li@intel.com> Co-authored-by: brian-intel <brian.mcginn@intel.com> Co-authored-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com>
brian-intel
added a commit
that referenced
this issue
Jul 12, 2023
* Ovms integration (#59) * feat: Renamed to Automated self checkout (#62) Closes: issue #61 Signed-off-by: Valina Li <valina.li@intel.com> * Benchmark docker (#26) * feat: Dockerfile for benchmark scripts * Updated to support --workload option Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Initial commit for supporting opencv-ovms workload Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Integration for ovms Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Updated docker image to openvino/model_server:2022.3.0.1-gpu Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Update configs/opencv-ovms/scripts/grpc_infer_binary_bit.py Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> * updated client image Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Add script for downloading the sample image Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Name fix for container Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> * Updated copyright and exit code Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> --------- Signed-off-by: Valina Li <valina.li@intel.com> Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> Co-authored-by: Valina Li <valina.li@intel.com> Co-authored-by: brian-intel <brian.mcginn@intel.com> Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> * feat: Enable pipeline failure log (#56) * feat: Enable pipeline failure log Closes: issue #41 Signed-off-by: Valina Li <valina.li@intel.com> * feat: enhance stream density function (#54) * feat: enhance stream density function Now the increment of pipeline numbers can be larger if the system is high power and performed well: The increment number can be taken from user's input or can be dynamically adjusted based the actual average FPS per stream and the target FPS. The searching policy is greedy to attempt to get to the maximum pipelines first and then decrement by 1 at a time to find the maximum number of pipelines reaching the target FPS. Closes: #51 * feat: update docker-run-dlstreamer.sh for stream density increments env * feat: update the increments logic and also add logic to cleanup the child processes of pipeline parent process * fix: resolve the hung issue caused by --rm as exited containers should be removed first before checking --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * feat: Add github actions (#70) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * feat: enhance the benchmark script to handle --workload option in any argument position (#76) Now if --workload is placed in front of pipeline options, it will work. Also works if it is placed in-between or after the pipeline options. Closes: #67 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * fix: Update shecllcheck minimum severity to warning (#82) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: Remove actoins/checkout from reviewdog (#83) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: fix shellcheck script issues (#81) * fix: fix shellcheck SC2167 * wipfix: use mapfile to better deal with multiple line output in commnad line * fix: fix shellcheck issue for benchmark.sh * fix: fix shellcheck issues for camera-simulator.sh and cleanup_gpu_metrics.sh * fix: fix shellcheck issues for copy-platform-metrics.sh * fix: fix shellcheck issues for format_avc_mp4.sh * fix: fix shellcheck issues for get-gpu-info.sh * fix: remove unused run.sh and run_server.sh under benchmark-scripts * fix: fix shellcheck issues for benchmark-scripts/stop_server.sh * fix: fix shellcheck issues for benchmark-scripts/stream_density.sh * fix: fix shellcheck issues for benchmark-scripts/test_format_avc_mp4.sh * fix: fix shellcheck issues for camera-simulator/camera-simulator.sh * fix: fix some boundary case when log_time_monitor_pid may not be there any more * fix: reoslve the issue of stop_server.sh when it is meant to be re-splitting for docker rm to be working * fix: use force kill to clean up its processes as sometimes it stuck with pkill --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * fix: fixed docker-run.sh with wrong message when no --workload option (#87) * fix: fixed docker-run.sh with wrong message when no --workload option Signed-off-by: Valina Li <valina.li@intel.com> * feat: Add model download top level script and add unit test for ovms … (#86) * feat: Add model download top level script and add unit test for ovms model download Signed-off-by: Valina Li <valina.li@intel.com> * Update reviewdog.yaml added new env token * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Feat: adding core smoke test (#93) * Fix: update collect platform metrics to handle multiple gpu devices (#94) * Update reviewdog.yaml * Update .reviewdog.yml * Fix: invert the format avc mp4 script tag logic to match our expected tag (#95) * Docs: updates for release 1.0.1 (#97) * fix: resolve the issue of the current model downloader (#98) * fix: resolve the issue of the current model downloader current getModel.sh script and other model downloading related scripts are all using relative path like . or .., which is not working when the scripts are running in different directories. This PR resolves the issue of using relative path and uses the absolute path to run via finding out the execution path on the runtime. Fixes: #96 * fix: remove check point model files and update the model downloader Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * feat: add trivy code scan (#102) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: Remove old linting workflows as reviewdog runs them (#103) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update reviewdog.yaml fix golangci-lint install * Update reviewdog.yaml fix: create new directory for golangci-lint install * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * fix reviewdog golangcit * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update .reviewdog.yml fix: Add shellcheck to reviewdog and fixed hadolint cmd call * feat: Add cleaning ovms containers to makefile (#101) feat: Add cleaning ovms containers to makefile closes: #100 Signed-off-by: Valina Li <valina.li@intel.com> * Update .reviewdog.yml * Create .golangci.yml feat: Add golangci config to .github/ * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update reviewdog.yaml * Update .reviewdog.yml * feat: Made CPU as default device for ovms pipeline (#114) closes: #112 Signed-off-by: Valina Li <valina.li@intel.com> * Fix: igt path pointing to the incorrect directory causing the igt log to not be written (#110) Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * Update gotest.yaml * Update reviewdog.yaml * fix: move workflow config files to .github directory (#121) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update codeql.yaml * fix: fixed pipelinesetup doc has incorect link to models.list.yml (#120) closes: #119 Signed-off-by: Valina Li <valina.li@intel.com> * feat: Added dockerfile.bitModel to download bit models (#116) * feat: Added dockerfile.bitModel to download bit models closes: #115 Signed-off-by: Valina Li <valina.li@intel.com> * Fix: pcm directory inconsistent between dockerfile and run script (#107) * Merge doc changes (#118) * Docs: merge from the 1.0.1 changes * only add new commands * feat: added OVMS sample image download into run script (#125) Closes: #124 Signed-off-by: Valina Li <valina.li@intel.com> * docs: Updated doc to include OVMS info (#126) * docs: Updated doc to include OVMS info Closes: #65 Signed-off-by: Valina Li <valina.li@intel.com> * Update codeql.yaml fix: there are no cpp files in the project * fix: update license to apache-2.0 (#130) * fix: update license to apache-2.0 Closes: #129 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> --------- Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * docs: Updated mkdoc to navigate to OVMS docs (#132) closes: #131 Signed-off-by: Valina Li <valina.li@intel.com> * feat: clean up checked-in dlstreamer models (#133) Model downloader can download those dlstreamer models, so we don't check in those model files. Closes: #99 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * feat: add docker image build to workflows (#137) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update build.yaml fix: syntax error * Update build.yaml fix: Remove OVMS server build * feat: add platform and CPU_ONLY parameters to enable smoke testing on all platform types (#136) * fix: add platform and CPU_ONLY parameters to enable smoke testing on all platform types --------- Signed-off-by: Brian McGinn <brian.mcginn@intel.com> Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> * fix: resolve hadolint issues (#140) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * Update reviewdog.yaml fix: Create directory for storing results for artifact * Update reviewdog.yaml fix: use absolute path * Update reviewdog.yaml * Update reviewdog.yaml * fix: Save scan results to artifact (#141) Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> * fix: resolve the make build-ovms-server rebuild issue (#143) Fixes: #142 Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> * fix: move sample output label to correct location and fix json format (#146) Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * docs: point model list to github model file (#148) Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * feat: temp pythonStream.py example Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * feat: Add grpc python files * fix: change to rtsp stream * save to jpg then read * encode instead of readfile * move code to grpc python file Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * fix: unit tests Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * feat: separate postprocess from the main grpc python * fix: output to file * feat: handle multiple servers and pass through inputsrc Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * merge updates * fix: PR merge * fix: PR updates Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * fix: Go env * fix: Add try loop around frame read. Ignore exceptions for now Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * fix: Stream Density for OVMS Signed-off-by: Brian McGinn <brian.mcginn@intel.com> * fix: Bash command * Update configs/opencv-ovms/scripts/grpc_python.py --------- Signed-off-by: Valina Li <valina.li@intel.com> Signed-off-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> Signed-off-by: Jim Wang <yutsung.jim.wang@intel.com> Signed-off-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com> Signed-off-by: Brian McGinn <brian.mcginn@intel.com> Co-authored-by: Jubilee Steinbrink <jubilee.steinbrink@intel.com> Co-authored-by: Valina Li <valina.li@intel.com> Co-authored-by: Jim Wang @ Intel <yutsung.jim.wang@intel.com> Co-authored-by: Marc-Philippe Fuller <marc-philippe.fuller@intel.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Version
What release version was the issue found in? 1.5.0
Hardware configuration
List Hardware components used.
11th Gen Intel® Core™ i7 with Mesa Intel® Xe Graphics
Describe the bug
need to automate the way to download bit models for OVMS pipeline
To Reproduce
Steps to reproduce the behavior:
Missing dockerfile and script file.
The text was updated successfully, but these errors were encountered: