github/AMD-SHARK-Studio
1
1
Fork 0
mirror of https://github.com/nod-ai/AMD-SHARK-Studio.git synced 2026-04-03 03:00:17 -04:00
Code Issues Packages Projects Releases Wiki Activity
1,743 Commits 66 Branches 705 Tags
diffusers-version
Commit Graph

1743 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Ean Garvey
6aeda3670f Split nightly workflow by backend (IREE / SHARK) (#313) 
* Fix validation for nightly builds.

* Add option to generate shark_tank inside SHARK project
Add shark_arg for updating tank on mismatched hash (downloader)

* Fixup CI tank dir option.

* Fixup work directory variable
 2022-09-09 22:51:30 +05:30
Prashant Kumar
bb52b224d0 Add sparse architecture and test with torchrec SparseArch. 
Features that don't work with current implementation:
    -- embeddingbag config with multiple features.
 2022-09-09 21:49:30 +05:30
Stanley Winata
95ec3d7216 [tank][v-diffusion] Polish up v-diffusion UX (#315) 2022-09-08 12:55:51 -07:00
powderluv
18872222d3 Update README.md 2022-09-07 01:14:30 -07:00
Ean Garvey
d453f2e49d Enable CPU benchmarks on test-models workflows. (#299) 
* Update test-models.yml

* Update README.md
 2022-09-07 01:22:58 -05:00
Ean Garvey
3824d37d27 Add metadata to benchmark results. (#297) 2022-09-06 13:03:48 -05:00
Ean Garvey
d946287723 Update xfails for torchvision models. (#310) 2022-09-01 13:06:12 -05:00
Prashant Kumar
885b0969f5 [WEB] Cache the compiled module. 
-- Don't compile the module again and again.
 2022-09-01 23:08:08 +05:30
Gaurav Shukla
a886cba655 [WEB] Add v_diffusion model in the shark web (#306) 
This commit adds adds `v_diffusion` model web visualization as a part of
shark web.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>

Signed-off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-09-01 06:34:51 -07:00
Vivek Khandelwal
4afe2e3adb Add func to save intermediate images in v_diffusion_pytorch 2022-09-01 18:36:58 +05:30
Gaurav Shukla
fe080eaee6 [WEB] Introduce web interface for the SHARK models (#298) 
This commit introduces web application for SHARK using gradio platform.
This adds web visualization of `Resnet50` and `Albert_Maskfill` models
as a start.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>

Signed-off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-08-31 23:17:52 -07:00
Quinn Dawkins
3703f014d9 Add scripts for generating images on ats-m (#305) 2022-08-31 23:07:02 -07:00
Daniel Garvey
d45a496030 adds a flag to enable directory choice (#303) 
individual tests will require implementation of the flag
alternatively, simply passing shark_default_sha in your
individual app's download function will allow for this behavior
 2022-08-31 22:17:40 -07:00
powderluv
4ee164c66f remove a100 cpu 2022-08-31 12:59:47 -07:00
powderluv
bf84c033bb add icelake 2022-08-31 12:58:40 -07:00
Prashant Kumar
5105f62551 Add the dlrm_model in shark example. (#301) 
-- DLRM model is added in the shark example.
-- The model is verified on cpu, gpu and vulkan.

Co-authored-by: powderluv <powderluv@users.noreply.github.com>
 2022-08-31 12:54:21 -07:00
Quinn Dawkins
99be837d84 Add lockstep tracer based on TorchMLIR eager mode + examples (#243) 2022-08-31 15:50:24 -04:00
Quinn Dawkins
b7766898ee Add cfg sampling from tank model for v-diffusion and move compilation outside of the sampling loop (#302) 2022-08-31 11:35:04 -07:00
powderluv
57f73dfbc9 Update nightly.yml 2022-08-28 23:59:03 -07:00
powderluv
50b2b9638d Update nightly.yml 2022-08-28 23:43:32 -07:00
Daniel Garvey
1bfd00e2f8 fixes an install issue (#295) 2022-08-25 18:52:00 -05:00
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
Phaneesh Barwaria
02d857260c Update ReadMe 
-Add gsutil installation for resnet50 example
 2022-08-25 20:28:50 +05:30
Phaneesh Barwaria
1322ec5935 Simplified Testing Interface (#289) 2022-08-24 23:54:56 -05:00
Daniel Garvey
48e9818f7e adds support to default to tuned model (#287) 
currently setup for tf bert/resnet50
going to refactor test class to avoid having to
add an argument to 50+ files
 2022-08-24 16:30:02 -05:00
Ean Garvey
14857770dc Fix local artifact recognition and usage by SHARK downloader. (#286) 
* Fix local artifact recognition and usage by SHARK downloader.

* Update generate_sharktank.py

* Update generate_sharktank.py
 2022-08-24 14:37:16 -05:00
Vivek Khandelwal
f79a6bf5aa Update setup_v_diffusion_pytorch.sh (#291) 
Fix minor issue with v-diffusion PyTorch version
 2022-08-24 22:00:02 +05:30
Prashant Kumar
7dc27a7477 Don't remove the latest .whl package from CI. (#290) 
Previously, the CI was removing the latest package and pointing to the
stale package.
 2022-08-24 09:03:48 -07:00
Chi_Liu
17dba601c8 Add huggingface top5 image classification automodel (#268) 2022-08-22 15:05:38 -07:00
Chi_Liu
064aa3b1f4 Fix tmp dir bug (#285) 2022-08-22 15:00:35 -07:00
Ean Garvey
4960efc686 Update requirements-importer.txt (#284) 2022-08-19 23:21:41 -05:00
Ean Garvey
a3654f33da Fix sourcing for canonical MiniLM shark_tank model artifacts. (#278) 
* Fix generation of MiniLM artifacts.

* Fix miniLM output for validation. Xfail numerics failure on mpnet.

* Update distilbert-base-uncased_tf_test.py

* try-except for transition of minilm model
 2022-08-17 23:03:47 -05:00
Daniel Garvey
82c541dfb8 fix missing model download path (#281) 2022-08-17 23:02:50 -05:00
Stanley Winata
55bcb2eb3c Level Zero Backend (#280) 2022-08-17 19:19:27 -07:00
Daniel Garvey
1a85550879 fix nightly upload check (#277) 2022-08-17 14:31:15 -05:00
Ean Garvey
334f2f76c4 Update README.md (#273) 
* Update README.md

* Update README.md

* Update README.md

* Update README.md

Co-authored-by: powderluv <powderluv@users.noreply.github.com>
 2022-08-17 10:38:27 -07:00
Vivek Khandelwal
03601ccdd6 Add v_diffusion_pytorch model in shark/tank (#271) 2022-08-17 22:53:31 +05:30
Prashant Kumar
88b0dec0ee Update unet_model to run on shark. 
-- Verified unet_model runs on the cpu/gpu/vulkan backend.
 2022-08-17 13:16:02 +05:30
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
a8b021dc8d Add benchmarks to MHLO miniLM and resnet50 and add dialect, num_iterations (#264) 2022-08-16 13:55:40 -05:00
Daniel Garvey
5e931debd5 Sharktank-ci (#262) 2022-08-15 13:32:24 -05:00
Ean Garvey
22ff92c48b Add config.VmModule argument to from_flatbuffer call. (#266) 2022-08-14 15:11:19 -07:00
powderluv
7f5aaa3477 Update nightly.yml 2022-08-14 12:22:50 -07:00
powderluv
904e0e1444 Update nightly.yml 2022-08-14 09:57:10 -07:00
powderluv
db6e2207ed Update _common.py 2022-08-13 13:49:01 -07:00
Daniel Garvey
7975087ee2 change backend name (#265) 2022-08-13 12:01:12 -07:00
Daniel Garvey
e8482d47f5 split nightly pytest commands (#259) 
prevents oom
 2022-08-12 16:11:46 -07:00
Ean Garvey
3e900d2b25 Change Resnet50 directory names. (#263) 2022-08-12 16:10:59 -07:00
Ean Garvey
4b5d09fc6c Add TF ResNet50 to tank tests. (#261) 
* Add TensorFlow Resnet50 test to shark tank.
 2022-08-12 09:20:43 -07:00
Prashant Kumar
02b1e7ac36 Update torch_mlir.compile API. 
torch_mlir.compile API is updated and verified by compiling all the
torch models via generate_sharktank script.
 2022-08-10 22:50:15 +05:30