Use interactive_embedded_interpreter to dynamicly loading various third-party libraries #66512

shunting314 · 2021-10-12T21:42:53Z

Stack from ghstack:

-> Use interactive_embedded_interpreter to dynamicly loading various third-party libraries #66512

TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose.

A couple of changes need to be done for the interactive_embedded_interpreter:
1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader
2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages.
3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries.

Differential Revision: D31587278

NOTE FOR REVIEWERS: This PR has internal Facebook specific changes or comments, please review them on Phabricator!

…rd-party libraries TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)! [ghstack-poisoned]

pytorch-probot · 2021-10-12T21:42:56Z

CI Flow Status

⚛️ CI Flow

Ruleset - Version: v1
Ruleset - File: https://github.com/pytorch/pytorch/blob/ed9f31168b515c1461cc398a97c81588a9dd8f76/.github/generated-ciflow-ruleset.json
PR ciflow labels: ciflow/default

Workflows	Labels (bold enabled)	Status
Triggered Workflows
linux-bionic-py3.6-clang9	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/noarch`, `ciflow/xla`	✅ triggered
linux-vulkan-bionic-py3.6-clang9	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/vulkan`	✅ triggered
linux-xenial-cuda11.3-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/default`, `ciflow/linux`	✅ triggered
linux-xenial-py3.6-clang7-asan	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/sanitizers`	✅ triggered
linux-xenial-py3.6-clang7-onnx	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/onnx`	✅ triggered
linux-xenial-py3.6-gcc5.4	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`	✅ triggered
linux-xenial-py3.6-gcc7-bazel-test	`ciflow/all`, `ciflow/bazel`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`	✅ triggered
win-vs2019-cpu-py3	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/win`	✅ triggered
win-vs2019-cuda11.3-py3	`ciflow/all`, `ciflow/cuda`, `ciflow/default`, `ciflow/win`	✅ triggered
Skipped Workflows
libtorch-linux-xenial-cuda10.2-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/libtorch`, `ciflow/linux`	🚫 skipped
libtorch-linux-xenial-cuda11.3-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/libtorch`, `ciflow/linux`	🚫 skipped
linux-bionic-cuda10.2-py3.9-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/slow`	🚫 skipped
linux-xenial-cuda10.2-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/slow`	🚫 skipped
parallelnative-linux-xenial-py3.6-gcc5.4	`ciflow/all`, `ciflow/cpu`, `ciflow/linux`	🚫 skipped
periodic-libtorch-linux-xenial-cuda11.1-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/libtorch`, `ciflow/linux`, `ciflow/scheduled`	🚫 skipped
periodic-linux-xenial-cuda10.2-py3-gcc7-slow-gradcheck	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/scheduled`, `ciflow/slow`, `ciflow/slow-gradcheck`	🚫 skipped
periodic-linux-xenial-cuda11.1-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/scheduled`	🚫 skipped
periodic-win-vs2019-cuda11.1-py3	`ciflow/all`, `ciflow/cuda`, `ciflow/scheduled`, `ciflow/win`	🚫 skipped
puretorch-linux-xenial-py3.6-gcc5.4	`ciflow/all`, `ciflow/cpu`, `ciflow/linux`	🚫 skipped

You can add a comment to the PR and tag @pytorchbot with the following commands:

# ciflow rerun, "ciflow/default" will always be added automatically
@pytorchbot ciflow rerun

# ciflow rerun with additional labels "-l <ciflow/label_name>", which is equivalent to adding these labels manually and trigger the rerun
@pytorchbot ciflow rerun -l ciflow/scheduled -l ciflow/slow

For more information, please take a look at the CI Flow Wiki.

facebook-github-bot · 2021-10-12T21:42:57Z

🔗 Helpful links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/66512
📄 Preview docs built from this PR
📄 Preview C++ docs built from this PR
↩️ [fb-only] Re-run with SSH instructions
🔧 Opt-in to CIFlow to control what jobs run on your PRs

💊 CI failures summary and remediations

As of commit ed9f311 (more details on the Dr. CI page):

2/3 failures introduced in this PR
1/3 tentatively recognized as flaky ❄️
- Click here to rerun these jobs

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

Lint / clang-tidy (1/1)

Step: "Check for warnings" (full log | diagnosis details | 🔁 rerun)

2021-10-13T07:31:58.5987131Z /__w/pytorch/pytor...logging.h' file not found [clang-diagnostic-error]

2021-10-13T07:31:58.5976774Z Processing 2 clang-tidy jobs
2021-10-13T07:31:58.5978497Z /__w/pytorch/pytorch/torch/csrc/deploy/interactive_embedded_interpreter.cpp:6:10: error: 'glog/logging.h' file not found [clang-diagnostic-error]
2021-10-13T07:31:58.5979602Z #include <glog/logging.h>
2021-10-13T07:31:58.5980252Z          ^
2021-10-13T07:31:58.5980734Z Warnings detected!
2021-10-13T07:31:58.5981247Z Summary:
2021-10-13T07:31:58.5982268Z [clang-diagnostic-error] occurred 1 times
2021-10-13T07:31:58.5983149Z     /__w/pytorch/pytorch/torch/csrc/deploy/interactive_embedded_interpreter.cpp:6
2021-10-13T07:31:58.5983855Z 
2021-10-13T07:31:58.5984604Z Processing 38 clang-tidy jobs
2021-10-13T07:31:58.5987131Z /__w/pytorch/pytorch/torch/csrc/deploy/interactive_embedded_interpreter.cpp:6:10: error: 'glog/logging.h' file not found [clang-diagnostic-error]
2021-10-13T07:31:58.5988165Z #include <glog/logging.h>
2021-10-13T07:31:58.5989819Z          ^
2021-10-13T07:31:58.5990413Z Warnings detected!
2021-10-13T07:31:58.5990956Z Summary:
2021-10-13T07:31:58.5992073Z [clang-diagnostic-error] occurred 1 times
2021-10-13T07:31:58.5993178Z     /__w/pytorch/pytorch/torch/csrc/deploy/interactive_embedded_interpreter.cpp:6
2021-10-13T07:31:58.5993900Z 
2021-10-13T07:31:58.5994643Z Please fix the above clang-tidy warnings.
2021-10-13T07:31:58.6025078Z ##[error]Process completed with exit code 1.
2021-10-13T07:31:58.6099299Z ##[group]Run pytorch/add-annotations-github-action@master

1 failure not recognized by patterns:

Job	Step	Action
^{pytorch_linux_xenial_py3_clang5_mobile_build}	^Build	🔁 rerun

❄️ 1 failure tentatively classified as flaky

but reruns have not yet been triggered to confirm:

pytorch_ios_12_5_1_x86_64_build (1/1)

Step: "Build" (full log | diagnosis details | 🔁 rerun) ❄️

Failed to recurse into submodule path 'third_party/onnx-tensorrt'

remote: Total 0 (delta 0), reused 0 (delta 0), pack-reused 0        
remote: Enumerating objects: 610, done.        
remote: Counting objects:   0% (1/610)        
remote: Counting objects:   1% (7/610)        
remote: Counting objects:   2% (13/610)        
remote: Counting objects:   3% (19/610)        
remote: Counting objects:   4% (25/610)        
remote: Counting objects:   5% (31/610)        
remote: Counting objects:   6% (37/610)        
remote: Counting objects:   7% (43/610)        
remote: Counting objects:   8% (49/610)        
remote: Counting objects:   9% (55/610)        
remote: Counting objects:  10% (61/610)        
remote: Counting objects:  11% (68/610)        
remote: Counting objects:  12% (74/610)        
remote: Counting objects:  13% (80/610)        
remote: Counting objects:  14% (86/610)        
remote: Counting objects:  15% (92/610)        
remote: Counting objects:  16% (98/610)        
remote: Counting objects:  17% (104/610)        
remote: Counting objects:  18% (110/610)        
remote: Counting objects:  19% (116/610)        
remote: Counting objects:  20% (122/610)        
remote: Counting objects:  21% (129/610)        
remote: Counting objects:  22% (135/610)        
remote: Counting objects:  23% (141/610)        
remote: Counting objects:  24% (147/610)        
remote: Counting objects:  25% (153/610)        
remote: Counting objects:  26% (159/610)        
remote: Counting objects:  27% (165/610)        
remote: Counting objects:  28% (171/610)        
remote: Counting objects:  29% (177/610)        
remote: Counting objects:  30% (183/610)        
remote: Counting objects:  31% (190/610)        
remote: Counting objects:  32% (196/610)        
remote: Counting objects:  33% (202/610)        
remote: Counting objects:  34% (208/610)        
remote: Counting objects:  35% (214/610)        
remote: Counting objects:  36% (220/610)        
remote: Counting objects:  37% (226/610)        
remote: Counting objects:  38% (232/610)        
remote: Counting objects:  39% (238/610)        
remote: Counting objects:  40% (244/610)        
remote: Counting objects:  41% (251/610)        
remote: Counting objects:  42% (257/610)        
remote: Counting objects:  43% (263/610)        
remote: Counting objects:  44% (269/610)        
remote: Counting objects:  45% (275/610)        
remote: Counting objects:  46% (281/610)        
remote: Counting objects:  47% (287/610)        
remote: Counting objects:  48% (293/610)        
remote: Counting objects:  49% (299/610)        
remote: Counting objects:  50% (305/610)        
remote: Counting objects:  51% (312/610)        
remote: Counting objects:  52% (318/610)        
remote: Counting objects:  53% (324/610)        
remote: Counting objects:  54% (330/610)        
remote: Counting objects:  55% (336/610)        
remote: Counting objects:  56% (342/610)        
remote: Counting objects:  57% (348/610)        
remote: Counting objects:  58% (354/610)        
remote: Counting objects:  59% (360/610)        
remote: Counting objects:  60% (366/610)        
remote: Counting objects:  61% (373/610)        
remote: Counting objects:  62% (379/610)        
remote: Counting objects:  63% (385/610)        
remote: Counting objects:  64% (391/610)        
remote: Counting objects:  65% (397/610)        
remote: Counting objects:  66% (403/610)        
remote: Counting objects:  67% (409/610)        
remote: Counting objects:  68% (415/610)        
remote: Counting objects:  69% (421/610)        
remote: Counting objects:  70% (427/610)        
remote: Counting objects:  71% (434/610)        
remote: Counting objects:  72% (440/610)        
remote: Counting objects:  73% (446/610)        
remote: Counting objects:  74% (452/610)        
remote: Counting objects:  75% (458/610)        
remote: Counting objects:  76% (464/610)        
remote: Counting objects:  77% (470/610)        
remote: Counting objects:  78% (476/610)        
remote: Counting objects:  79% (482/610)        
remote: Counting objects:  80% (488/610)        
remote: Counting objects:  81% (495/610)        
remote: Counting objects:  82% (501/610)        
remote: Counting objects:  83% (507/610)        
remote: Counting objects:  84% (513/610)        
remote: Counting objects:  85% (519/610)        
remote: Counting objects:  86% (525/610)        
remote: Counting objects:  87% (531/610)        
remote: Counting objects:  88% (537/610)        
remote: Counting objects:  89% (543/610)        
remote: Counting objects:  90% (549/610)        
remote: Counting objects:  91% (556/610)        
remote: Counting objects:  92% (562/610)        
remote: Counting objects:  93% (568/610)        
remote: Counting objects:  94% (574/610)        
remote: Counting objects:  95% (580/610)        
remote: Counting objects:  96% (586/610)        
remote: Counting objects:  97% (592/610)        
remote: Counting objects:  98% (598/610)        
remote: Counting objects:  99% (604/610)        
remote: Counting objects: 100% (610/610)        
remote: Counting objects: 100% (610/610), done.        
remote: Compressing objects:   0% (1/323)        
remote: Compressing objects:   1% (4/323)        
remote: Compressing objects:   2% (7/323)        
remote: Compressing objects:   3% (10/323)        
remote: Compressing objects:   4% (13/323)        
remote: Compressing objects:   5% (17/323)        
remote: Compressing objects:   6% (20/323)        
remote: Compressing objects:   7% (23/323)        
remote: Compressing objects:   8% (26/323)        
remote: Compressing objects:   9% (30/323)        
remote: Compressing objects:  10% (33/323)        
remote: Compressing objects:  11% (36/323)        
remote: Compressing objects:  12% (39/323)        
remote: Compressing objects:  13% (42/323)        
remote: Compressing objects:  14% (46/323)        
remote: Compressing objects:  15% (49/323)        
remote: Compressing objects:  16% (52/323)        
remote: Compressing objects:  17% (55/323)        
remote: Compressing objects:  18% (59/323)        
remote: Compressing objects:  19% (62/323)        
remote: Compressing objects:  20% (65/323)        
remote: Compressing objects:  21% (68/323)        
remote: Compressing objects:  22% (72/323)        
remote: Compressing objects:  23% (75/323)        
remote: Compressing objects:  24% (78/323)        
remote: Compressing objects:  25% (81/323)        
remote: Compressing objects:  26% (84/323)        
remote: Compressing objects:  27% (88/323)        
remote: Compressing objects:  28% (91/323)        
remote: Compressing objects:  29% (94/323)        
remote: Compressing objects:  30% (97/323)        
remote: Compressing objects:  31% (101/323)        
remote: Compressing objects:  32% (104/323)        
remote: Compressing objects:  33% (107/323)        
remote: Compressing objects:  34% (110/323)        
remote: Compressing objects:  35% (114/323)        
remote: Compressing objects:  36% (117/323)        
remote: Compressing objects:  37% (120/323)        
remote: Compressing objects:  38% (123/323)        
remote: Compressing objects:  39% (126/323)        
remote: Compressing objects:  40% (130/323)        
remote: Compressing objects:  41% (133/323)        
remote: Compressing objects:  42% (136/323)        
remote: Compressing objects:  43% (139/323)        
remote: Compressing objects:  44% (143/323)        
remote: Compressing objects:  45% (146/323)        
remote: Compressing objects:  46% (149/323)        
remote: Compressing objects:  47% (152/323)        
remote: Compressing objects:  48% (156/323)        
remote: Compressing objects:  49% (159/323)        
remote: Compressing objects:  50% (162/323)        
remote: Compressing objects:  51% (165/323)        
remote: Compressing objects:  52% (168/323)        
remote: Compressing objects:  53% (172/323)        
remote: Compressing objects:  54% (175/323)        
remote: Compressing objects:  55% (178/323)        
remote: Compressing objects:  56% (181/323)        
remote: Compressing objects:  57% (185/323)        
remote: Compressing objects:  58% (188/323)        
remote: Compressing objects:  59% (191/323)        
remote: Compressing objects:  60% (194/323)        
remote: Compressing objects:  61% (198/323)        
remote: Compressing objects:  62% (201/323)        
remote: Compressing objects:  63% (204/323)        
remote: Compressing objects:  64% (207/323)        
remote: Compressing objects:  65% (210/323)        
remote: Compressing objects:  66% (214/323)        
remote: Compressing objects:  67% (217/323)        
remote: Compressing objects:  68% (220/323)        
remote: Compressing objects:  69% (223/323)        
remote: Compressing objects:  70% (227/323)        
remote: Compressing objects:  71% (230/323)        
remote: Compressing objects:  72% (233/323)        
remote: Compressing objects:  73% (236/323)        
remote: Compressing objects:  74% (240/323)        
remote: Compressing objects:  75% (243/323)        
remote: Compressing objects:  76% (246/323)        
remote: Compressing objects:  77% (249/323)        
remote: Compressing objects:  78% (252/323)        
remote: Compressing objects:  79% (256/323)        
remote: Compressing objects:  80% (259/323)        
remote: Compressing objects:  81% (262/323)        
remote: Compressing objects:  82% (265/323)        
remote: Compressing objects:  83% (269/323)        
remote: Compressing objects:  84% (272/323)        
remote: Compressing objects:  85% (275/323)        
remote: Compressing objects:  86% (278/323)        
remote: Compressing objects:  87% (282/323)        
remote: Compressing objects:  88% (285/323)        
remote: Compressing objects:  89% (288/323)        
remote: Compressing objects:  90% (291/323)        
remote: Compressing objects:  91% (294/323)        
remote: Compressing objects:  92% (298/323)        
remote: Compressing objects:  93% (301/323)        
remote: Compressing objects:  94% (304/323)        
remote: Compressing objects:  95% (307/323)        
remote: Compressing objects:  96% (311/323
Receiving objects:   0% (1/370)
Receiving objects:   1% (4/370)
Receiving objects:   2% (8/370)
Receiving objects:   3% (12/370)
Receiving objects:   4% (15/370)
Receiving objects:   5% (19/370)
Receiving objects:   6% (23/370)
Receiving objects:   7% (26/370)
Receiving objects:   8% (30/370)
Receiving objects:   9% (34/370)
Receiving objects:  10% (37/370)
Receiving objects:  11% (41/370)
Receiving objects:  12% (45/370)
Receiving objects:  13% (49/370)
Receiving objects:  14% (52/370)
Receiving objects:  15% (56/370)
Receiving objects:  16% (60/370)
Receiving objects:  17% (63/370)
Receiving objects:  18% (67/370)
Receiving objects:  19% (71/370)
Receiving objects:  20% (74/370)
Receiving objects:  21% (78/370)
Receiving objects:  22% (82/370)
Receiving objects:  23% (86/370)
Receiving objects:  24% (89/370)
Receiving objects:  25% (93/370)
Receiving objects:  26% (97/370)
Receiving objects:  27% (100/370)
Receiving objects:  28% (104/370)
Receiving objects:  29% (108/370)
Receiving objects:  30% (111/370)
Receiving objects:  31% (115/370)
Receiving objects:  32% (119/370)
Receiving objects:  33% (123/370)
Receiving objects:  34% (126/370)
Receiving objects:  35% (130/370)
Receiving objects:  36% (134/370)
Receiving objects:  37% (137/370)
Receiving objects:  38% (141/370)
Receiving objects:  39% (145/370)
Receiving objects:  40% (148/370)
Receiving objects:  41% (152/370)
Receiving objects:  42% (156/370)
Receiving objects:  43% (160/370)
Receiving objects:  44% (163/370)
Receiving objects:  45% (167/370)
Receiving objects:  46% (171/370)
Receiving objects:  47% (174/370)
Receiving objects:  48% (178/370)
Receiving objects:  49% (182/370)
Receiving objects:  50% (185/370)
Receiving objects:  51% (189/370)
Receiving objects:  52% (193/370)
Receiving objects:  53% (197/370)
Receiving objects:  54% (200/370)
Receiving objects:  55% (204/370)
Receiving objects:  56% (208/370)
Receiving objects:  57% (211/370)
Receiving objects:  58% (215/370)
Receiving objects:  59% (219/370)
Receiving objects:  60% (222/370)
Receiving objects:  61% (226/370)
Receiving objects:  62% (230/370)
Receiving objects:  63% (234/370)
Receiving objects:  64% (237/370)
Receiving objects:  65% (241/370)
Receiving objects:  66% (245/370)
Receiving objects:  67% (248/370)
Receiving objects:  68% (252/370)
Receiving objects:  69% (256/370)
Receiving objects:  70% (259/370)
Receiving objects:  71% (263/370)
Receiving objects:  72% (267/370)
Receiving objects:  73% (271/370)
Receiving objects:  74% (274/370)
Receiving objects:  75% (278/370)
Receiving objects:  76% (282/370)
Receiving objects:  77% (285/370)
Receiving objects:  78% (289/370)
Receiving objects:  79% (293/370)
Receiving objects:  80% (296/370)
Receiving objects:  81% (300/370)
Receiving objects:  82% (304/370)
Receiving objects:  83% (308/370)
Receiving objects:  84% (311/370)
Receiving objects:  85% (315/370)
Receiving objects:  86% (319/370)
Receiving objects:  87% (322/370)
Receiving objects:  88% (326/370)
Receiving objects:  89% (330/370)
Receiving objects:  90% (333/370)
Receiving objects:  91% (337/370)
Receiving objects:  92% (341/370)
Receiving objects:  93% (345/370)
Receiving objects:  94% (348/370)
remote: Total 370 (delta 199), reused 107 (delta 23), pack-reused 0        
Receiving objects:  95% (352/370)
Receiving objects:  96% (356/370)
Receiving objects:  97% (359/370)
Receiving objects:  98% (363/370)
Receiving objects:  99% (367/370)
Receiving objects: 100% (370/370)
Receiving objects: 100% (370/370), 1.19 MiB | 13.05 MiB/s, done.
Resolving deltas:   0% (0/199)
Resolving deltas:   1% (2/199)
Resolving deltas:   2% (4/199)
Resolving deltas:   3% (6/199)
Resolving deltas:   4% (8/199)
Resolving deltas:   5% (10/199)
Resolving deltas:   6% (12/199)
Resolving deltas:   7% (14/199)
Resolving deltas:   8% (16/199)
Resolving deltas:   9% (18/199)
Resolving deltas:  10% (20/199)
Resolving deltas:  11% (22/199)
Resolving deltas:  12% (24/199)
Resolving deltas:  13% (26/199)
Resolving deltas:  14% (28/199)
Resolving deltas:  15% (30/199)
Resolving deltas:  16% (32/199)
Resolving deltas:  17% (34/199)
Resolving deltas:  18% (36/199)
Resolving deltas:  19% (38/199)
Resolving deltas:  20% (40/199)
Resolving deltas:  21% (42/199)
Resolving deltas:  22% (44/199)
Resolving deltas:  23% (46/199)
Resolving deltas:  24% (48/199)
Resolving deltas:  25% (50/199)
Resolving deltas:  26% (52/199)
Resolving deltas:  27% (54/199)
Resolving deltas:  28% (56/199)
Resolving deltas:  29% (58/199)
Resolving deltas:  30% (60/199)
Resolving deltas:  31% (62/199)
Resolving deltas:  32% (64/199)
Resolving deltas:  33% (66/199)
Resolving deltas:  34% (68/199)
Resolving deltas:  35% (70/199)
Resolving deltas:  36% (72/199)
Resolving deltas:  37% (74/199)
Resolving deltas:  38% (76/199)
Resolving deltas:  39% (78/199)
Resolving deltas:  40% (80/199)
Resolving deltas:  41% (82/199)
Resolving deltas:  42% (84/199)
Resolving deltas:  43% (86/199)
Resolving deltas:  44% (88/199)
Resolving deltas:  45% (90/199)
Resolving deltas:  46% (92/199)
Resolving deltas:  47% (94/199)
Resolving deltas:  48% (96/199)
Resolving deltas:  49% (98/199)
Resolving deltas:  50% (100/199)
Resolving deltas:  51% (102/199)
Resolving deltas:  52% (104/199)
Resolving deltas:  53% (106/199)
Resolving deltas:  54% (108/199)
Resolving deltas:  55% (110/199)
Resolving deltas:  56% (112/199)
Resolving deltas:  57% (114/199)
Resolving deltas:  58% (116/199)
Resolving deltas:  59% (118/199)
Resolving deltas:  60% (120/199)
Resolving deltas:  61% (122/199)
Resolving deltas:  62% (124/199)
Resolving deltas:  63% (126/199)
Resolving deltas:  64% (128/199)
Resolving deltas:  65% (130/199)
Resolving deltas:  66% (132/199)
Resolving deltas:  67% (134/199)
Resolving deltas:  68% (136/199)
Resolving deltas:  69% (138/199)
Resolving deltas:  70% (140/199)
Resolving deltas:  71% (142/199)
Resolving deltas:  72% (144/199)
Resolving deltas:  73% (146/199)
Resolving deltas:  74% (148/199)
Resolving deltas:  75% (150/199)
Resolving deltas:  76% (152/199)
Resolving deltas:  77% (154/199)
Resolving deltas:  78% (156/199)
Resolving deltas:  79% (158/199)
Resolving deltas:  80% (160/199)
Resolving deltas:  81% (162/199)
Resolving deltas:  82% (164/199)
Resolving deltas:  83% (166/199)
Resolving deltas:  84% (168/199)
Resolving deltas:  85% (170/199)
Resolving deltas:  86% (172/199)
Resolving deltas:  87% (174/199)
Resolving deltas:  88% (176/199)
Resolving deltas:  89% (178/199)
Resolving deltas:  90% (180/199)
Resolving deltas:  91% (182/199)
Resolving deltas:  92% (184/199)
Resolving deltas:  93% (186/199)
Resolving deltas:  94% (188/199)
Resolving deltas:  95% (190/199)
Resolving deltas:  96% (192/199)
Resolving deltas:  97% (194/199)
Resolving deltas:  98% (196/199)
Resolving deltas:  99% (198/199)
Resolving deltas: 100% (199/199)
Resolving deltas: 100% (199/199), completed with 151 local objects.
From ssh://github.com/facebook/zstd
 * branch            aec56a52fbab207fc639a1937d1e708a282edca8 -> FETCH_HEAD
Submodule path 'third_party/zstd': checked out 'aec56a52fbab207fc639a1937d1e708a282edca8'
Failed to recurse into submodule path 'third_party/onnx-tensorrt'


Exited with code exit status 1

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

suo

nice! Some small nits but otherwise looks good

suo · 2021-10-12T22:07:27Z

torch/csrc/deploy/deploy.cpp

-InterpreterManager::InterpreterManager(size_t nInterp) : resources_(nInterp) {
+InterpreterManager::InterpreterManager(
+    size_t nInterp,
+    const std::string& pylibRoot)


this hsould probably be called pythonPath or something, which is a more conventional way of describing what this string is for.

suo · 2021-10-12T22:08:14Z

torch/csrc/deploy/deploy.h

-  explicit InterpreterManager(size_t nInterp = 2);
+  explicit InterpreterManager(
+      size_t nInterp = 2,
+      const std::string& pylibRoot = "");


c10::optional<string> I think is more explicit than using an empty string as the null value.

…various third-party libraries" TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)! [ghstack-poisoned]

…rd-party libraries Pull Request resolved: #66512 TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. ghstack-source-id: 140415316 Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)!

…various third-party libraries" TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)! [ghstack-poisoned]

…rd-party libraries Pull Request resolved: #66512 TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. ghstack-source-id: 140417244 Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)!

shunting314 · 2021-10-12T23:12:18Z

Clang-tidy job fail on torch/csrc/deploy/interactive_embedded_interpreter.cpp , complaining <glog/logging.h> not found. I guess we need add interactive_embedded_interpreter.cpp to the OSS build to fix that. I have a separate PR for that.

…various third-party libraries" TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)! [ghstack-poisoned]

…rd-party libraries Pull Request resolved: #66512 TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. ghstack-source-id: 140453213 Differential Revision: [D31587278](https://our.internmc.facebook.com/intern/diff/D31587278/) **NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D31587278/)!

…rd-party libraries (#66512) Summary: Pull Request resolved: #66512 TLDR, we are able to use the interactive_embedded_interpreter (basically just torch::deploy interpreter with an interactive shell) to dynamicly load various third party libraries. We use the popular libraries numpy, scipy, regex, pandas for illustration purpose. A couple of changes need to be done for the interactive_embedded_interpreter: 1, we need link with :embedded_interpreter_all rather than :embedded_interpreter so we can enable DEEPBIND and use our custom loader 2, we provide a pylibRoot path to construct the InterpreterManager. The path will be added to the embedded interpreter's sys.path. Typically we can pass in the python library root path in a conda environment so torch::deploy interpreter can find all installed packages. 3, we allow interactive_embedded_interpreter execute a script to ease recording the exploration of various python libraries. ghstack-source-id: 140453213 Test Plan: Install numpy, scipy, regex, pandas in the conda environment or on the machine directly. Suppose /home/shunting/.local/lib/python3.8/site-packages/ is the root path for the installed libraries. - buck run mode/opt :interactive_embedded_interpreter -- --pylib_root=/home/shunting/.local/lib/python3.8/site-packages/ --pyscript=~/p7/iei_examples/try_regex.py content of try_regex.py: ``` import regex print(regex) pat = r'(.+)\1' print(regex.match(pat, "abcabc")) print(regex.match(pat, "abcba")) print("bye") ``` - buck run mode/opt :interactive_embedded_interpreter -- --pylib_root=/home/shunting/.local/lib/python3.8/site-packages/ --pyscript=~/p7/iei_examples/try_numpy.py content of try_numpy.py: ``` import numpy as np print(f"numpy at {np}") a = np.random.rand(2, 3) b = np.random.rand(3, 2) print(np.matmul(a, b)) ``` - buck run mode/opt :interactive_embedded_interpreter -- --pylib_root=/home/shunting/.local/lib/python3.8/site-packages/ --pyscript=~/p7/iei_examples/try_scipy.py content of try_scipy.py: ``` import numpy as np from scipy import linalg mat_a = np.array([[1, 0, 0, 0], [1, 1, 0, 0], [1, 2, 1, 0], [1, 3, 3, 1]]) mat_b = linalg.inv(mat_a) print(mat_b) ``` - buck run mode/opt :interactive_embedded_interpreter -- --pylib_root=/home/shunting/.local/lib/python3.8/site-packages/ --pyscript=~/p7/iei_examples/try_pandas.py content of try_pandas.py: ``` import pandas as pd print(f"pandas at {pd}") df = pd.DataFrame({ "col1": [1, 2, 3, 4], "col2": [2, 4, 8, 16], }) print(df) ``` Reviewed By: suo Differential Revision: D31587278 fbshipit-source-id: c0b031c1fa71a77cdfeba1d04514f83127f79012

pytorch-probot bot added the ciflow/default label Oct 12, 2021

facebook-github-bot added the cla signed label Oct 12, 2021

shunting314 requested review from suo and wconstab October 12, 2021 21:43

suo approved these changes Oct 12, 2021

View reviewed changes

facebook-github-bot closed this in a453ebc Oct 13, 2021

suo mentioned this pull request Oct 13, 2021

[Meta] CI Revert Tracker #66178

Closed

facebook-github-bot deleted the gh/shunting314/5/head branch October 17, 2021 14:27

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Use interactive_embedded_interpreter to dynamicly loading various third-party libraries #66512

Use interactive_embedded_interpreter to dynamicly loading various third-party libraries #66512

Uh oh!

shunting314 commented Oct 12, 2021 •

edited

Loading

Uh oh!

pytorch-probot bot commented Oct 12, 2021 •

edited

Loading

⚛️ CI Flow

Uh oh!

facebook-github-bot commented Oct 12, 2021 •

edited

Loading

Uh oh!

suo left a comment

Uh oh!

suo Oct 12, 2021

Uh oh!

suo Oct 12, 2021

Uh oh!

shunting314 commented Oct 12, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Use interactive_embedded_interpreter to dynamicly loading various third-party libraries #66512

Use interactive_embedded_interpreter to dynamicly loading various third-party libraries #66512

Uh oh!

Conversation

shunting314 commented Oct 12, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-probot bot commented Oct 12, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⚛️ CI Flow

Uh oh!

facebook-github-bot commented Oct 12, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful links

💊 CI failures summary and remediations

🕵️ 1 new failure recognized by patterns

Lint / clang-tidy (1/1)

1 failure not recognized by patterns:

❄️ 1 failure tentatively classified as flaky

pytorch_ios_12_5_1_x86_64_build (1/1)

Uh oh!

suo left a comment

Choose a reason for hiding this comment

Uh oh!

suo Oct 12, 2021

Choose a reason for hiding this comment

Uh oh!

suo Oct 12, 2021

Choose a reason for hiding this comment

Uh oh!

shunting314 commented Oct 12, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

shunting314 commented Oct 12, 2021 •

edited

Loading

pytorch-probot bot commented Oct 12, 2021 •

edited

Loading

facebook-github-bot commented Oct 12, 2021 •

edited

Loading