github/SHARK-Studio
1
1
Fork 0
mirror of https://github.com/nod-ai/SHARK-Studio.git synced 2026-01-09 22:07:55 -05:00
Code Issues Packages Projects Releases Wiki Activity
1,454 Commits 66 Branches 705 Tags
3662224c04fab315b87ddc657f6e108bba5b434b
Commit Graph

59 Commits

Author SHA1 Message Date
jinchen62
3662224c04 Update brevitas requirement (#1677) 
also clean up useless args

Co-authored-by: powderluv <powderluv@users.noreply.github.com>
 2023-07-19 22:03:32 -07:00
AyaanShah2204
536aba1424 unpinned torch_mlir (#1668) 
Co-authored-by: powderluv <powderluv@users.noreply.github.com>
 2023-07-19 06:28:00 -07:00
jinchen62
bc6fee1a0c Add int4/int8 vicuna (#1598) 2023-07-05 07:01:51 -07:00
Ean Garvey
6e6c90f62b Pin torch-mlir and use local-task in OPT. (#1592) 2023-06-23 19:17:05 -07:00
Ranvir Singh Virk
18daec78c8 Added check for python version (#1570) 
* Added check for python version

* Update for PYTHON_VERSION_X_Y
 2023-06-21 18:56:47 -07:00
Boian Petkantchin
9acf519078 Add option to skip venv creation in setup script 2023-05-16 09:54:07 -07:00
Ean Garvey
96b1560da5 Make batch size configurable via pytest and fix sharktank generation. (#1227) 
* Fix sharktank generation and add batch_size pytest option for torch.

* Disable torch dynamo until py3.11 supported

* Compile torchmodel without dynamo if torch.compile fails

* Use release versions of TF/Keras for importer.

* Pin torchvision and remove debug prints.

* Remove duplicates from torch model list.

* Update generate_sharktank.py

* xfail a few models that fail sharktank generation/ numerics
 2023-03-28 14:33:39 -05:00
Anush Elangovan
c561ebf43c Drop the torch-mlir pin 
Seems to work now with top of master
 2023-03-10 15:39:04 -08:00
mariecwhite
a52331d4aa Install IREE pre-releases (#1139) 2023-03-02 23:17:56 -06:00
yzhang93
ad821a1fc8 Use old torch-mlir package to avoid crash on rdna2 (#1137) 2023-03-02 18:16:58 -08:00
powderluv
65775046d8 update IREE pip links 2023-02-21 19:31:23 -08:00
Ean Garvey
2ae047f1a8 Update importer/benchmark setup for python3.11 (#1043) 2023-02-20 11:29:00 -06:00
powderluv
6d54eb68dc update to support 3.11 2023-02-11 15:23:18 -08:00
Ean Garvey
7f27b61c98 Update setup_venv.sh to install triton if BENCHMARK=1 2023-01-19 00:26:46 -06:00
Ean Garvey
72f29b67d5 Add Resnet50 fp16 variant to pytests. (#760) 2023-01-10 16:31:11 -08:00
powderluv
d946cffabc Revert "Move most xfails to entries in tank/all_models.csv and temporarily remove multiprocessing and TF gpu support. (#602)" (#622) 
This reverts commit fe618811ee.
 2022-12-13 21:49:46 -08:00
Ean Garvey
fe618811ee Move most xfails to entries in tank/all_models.csv and temporarily remove multiprocessing and TF gpu support. (#602) 
* Move most xfails to entries in tank/all_models.csv

* enable usage of pytest without specifying tank/test_models.py

* add dict_configs.py to gitignore.

* Pin versions for runtimes and torch-mlir for setup.
 2022-12-13 18:11:17 -08:00
Ean Garvey
1699db79b5 Disable SHARK-Runtime flags if USE_IREE=1 specified during setup. (#588) 
* Disable SHARK-Runtime flags if USE_IREE=1 specified during setup.

* Update setup_venv.sh

* Autodetect cpu count for runtime flags.
 2022-12-08 02:31:31 -06:00
Anush Elangovan
1106456651 Update cuda 11.7 nightly URL and add index.spec 2022-12-01 22:49:23 -08:00
Ean Garvey
45677c1e23 Install torch version required by torch-mlir when setting up importer venv. (#486) 2022-11-14 14:01:01 +05:30
PhaneeshB
64d6da75c7 Resolve Mac torch-mlir torch setup dependency. Enable MacOS CI 2022-11-07 15:38:37 +05:30
Ean Garvey
1e95e4b502 Change dependency installation order in venv setup script. (#470) 2022-11-04 20:53:54 -05:00
Ean Garvey
9956099516 Add pytest option for updating tank and fix save_mlir function. (#413) 
* Use IREE tf tools to save .mlir modules when generating shark_tank.

* Add option to pytest for enabling auto-updates to local shark tank.

* xfail mobilenet torch on cpu, cuda and fix CI macos setup

* Update test-models.yml to disable macos vulkan CI.
 2022-10-25 21:29:18 +05:30
anush elangovan
a3f733490c Force update of packages 
Pickup tools from upstream IREE
 2022-10-19 05:20:53 +00:00
anush elangovan
8a11f138d1 Update SHARK-Runtime releases page 2022-10-19 05:06:36 +00:00
Ean Garvey
13bae2538a Update URL for IREE compiler/runtime install (#397) 
* Update URL for IREE compiler/runtime install

* Update gh-pages-releases.yml

* Update test_models.py

* Update assets path
 2022-10-12 15:47:11 -05:00
Daniel Garvey
d38e37bd99 seperate importer and benchmark deps (#393) 2022-10-08 23:31:20 -05:00
Ean Garvey
006adf8746 Fix issue with FASTAPI pip install. (#382) 2022-10-01 14:55:24 -05:00
Ean Garvey
56f8a0d85a Update torch-mlir releases page in setup_venv.sh (#374) 
* Update README.md

* Update setup_venv.sh
 2022-09-28 11:07:44 -07:00
Anush Elangovan
e304041574 Remove redundant {} 2022-09-13 16:12:35 -07:00
Anush Elangovan
1776c55e73 Fix torch-mlir download URL 2022-09-13 16:07:25 -07:00
Anush Elangovan
4e4c34c717 fix release downloads 2022-09-13 15:00:47 -07:00
Daniel Garvey
64424877ac No iree instal (#294) 
* adds support to default to tuned model

currently setup for tf bert/resnet50
going to refactor test class to avoid having to
add an argument to 50+ files

* adds an option to avoid installing iree

useful when building iree from source
specify env variable NO_BACKEND=1
 2022-08-25 15:02:28 -05:00
Ean Garvey
3514822cac Improvements to pytest benchmarks. (#267) 
* Add ONNX env var flags for venv setup.

* Setup arguments for ONNX benchmarking via pytest.

* Enable ONNX benchmarking on MiniLM via pytest (experimental)

* Fix sequence lengths to 128 for TF model creation and fix issue with benchmarks.

* Disable CI CPU benchmarks on A100, change some default args.

* add xfails for roberta TF model tests on GPU.
 2022-08-17 02:29:48 -05:00
Ean Garvey
d556c0d6ef Update nightly.yml to remove IMPORTER=1 in package validation. (#226) 
* Update nightly.yml to fix wheel building issues.

* Update setup_venv.sh to only install torch+cu116 if IMPORTER=1 is set.
 2022-07-31 20:39:59 -07:00
Ean Garvey
0ee515a7be Fix GPU benchmarks on PyTorch model tests. (#218) 
* Add CUDA_BENCHMARKS option to setup.venv to enable PyTorch benchmarks on CUDA.

* Fix PyTorch GPU benchmarks for tank models.
 2022-07-29 15:17:55 -07:00
Stanley Winata
e48642f8a5 Update benchmark command to ToM and Add to test (#125) 
-Update benchmark_cl to latest benchmark_module API
-tensorflow OOM issue.
 2022-06-13 23:57:05 -07:00
Anush Elangovan
e4dd179a73 Tweak requirements.txt 2022-06-12 02:01:29 -07:00
Anush Elangovan
879942d87f Fix macOS importer reqs 2022-06-12 01:45:05 -07:00
Anush Elangovan
3c26c02a5e Add OSX-importer requirements 2022-06-12 07:38:02 +00:00
Anush Elangovan
626482e46a Refactor setup_venv.sh 
IMPORTER=1 ./setup_venv.sh is required to get importer deps

Without importers venv is 1.1G, with it is 3.9G
 2022-06-11 06:26:31 +00:00
Stanley Winata
8565be9b6b ORT-HF Benchmark Integration (#101) 
-Add HF Benchmarker class.
-Add sample to benchmark HF model.

Example:
```bash
python -m benchmarks.hf_model_benchmark --num_iterations=10 --model_name="microsoft/MiniLM-L12-H384-uncased"
```
 2022-06-07 23:49:39 -07:00
Anush Elangovan
22aa0e51a7 Move macOS to 3.10 2022-06-05 10:36:31 -07:00
Stanley Winata
276dcf1441 Add benchmark for TF (#87) 
-Refactor SharkBenchmarker to run TF
-Add example and test to benchmark TF
 2022-06-03 22:10:27 -07:00
anush elangovan
af6dc91b83 Update vulkan check 2022-06-03 23:45:42 +00:00
anush elangovan
8fa5b39bd6 Add importer utils to linux build 2022-06-02 19:33:51 +00:00
Prashant Kumar
96577000a3 Remove functorch dependency. 2022-05-24 18:42:35 +05:30
anush elangovan
49bec49c53 prep for pypi 2022-05-16 05:44:18 +00:00
Stanley Winata
2117c1ab39 Install package to pip venv during setup_venv 
On fresh pip venv, this commit will help install
the packages onto the newly create venv.
 2022-05-15 15:00:17 -07:00
Anush Elangovan
c8aafae562 Update links to SHARK 2022-05-15 14:47:31 -07:00