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
976 Commits 66 Branches 705 Tags
enable_caching_alloc
Commit Graph

459 Commits

Author SHA1 Message Date
Phaneesh Barwaria
831f206cd0 Revert "Add target triple selection for multiple cards" (#655) 
This reverts commit acb905f0cc.
 2022-12-16 15:01:45 -08:00
Gaurav Shukla
72648aa9f2 Revert "[SD][WEB] Deduce vulkan-target-triple in the presence of multiple cards" 
This reverts commit 35e623deaf.
 2022-12-17 04:28:18 +05:30
Gaurav Shukla
35e623deaf [SD][WEB] Deduce vulkan-target-triple in the presence of multiple cards 
1. Get the correct vulkan-target-triple for a specified device in the
   presence of multiple cards.
2. Use tuned unet model for rdna3 cards.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-12-17 03:04:47 +05:30
Anush Elangovan
6263636738 Fix more lints 2022-12-16 13:26:15 -08:00
yzhang93
c73eed2e51 Add VAE winograd tuned model (#647) 2022-12-16 13:01:45 -08:00
Anush Elangovan
30fdc99f37 Set to enable llpc 
Use an env var to enable llpc
 2022-12-16 12:57:30 -08:00
PhaneeshB
acb905f0cc Add target triple selection for multiple cards 2022-12-17 02:24:37 +05:30
Gaurav Shukla
bba06d0142 [SD][WEB] Avoid passing args to utils APIs 
Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-12-17 01:41:33 +05:30
Ean Garvey
a14a47af12 Move most xfails to entries in tank/all_models.csv and temporarily remove multiprocessing and TF gpu support. (#646) 
-Adds date variable back to nightly.yml so shark_tank uploads are dated again
-added specification for nightly pytests to not run tests on metal (vulkan is sufficient)
-added some paths/filetypes to be ignored when triggering workflow runs. (no test-models on changes to .md files or anything in the shark/examples/ directory or its subdirectories.
-pytest only picks up tank/test_models.py, so no need to specify which file to run when running pytest from SHARK base directory.
-Cleaned up xfails so that they can be added to models as csv entries. Columns 7-9 in all_models.csv trigger xfails with cpu, cuda, vulkan, respectively, and row 10 can be populated with a reason for the xfails.
-Fixed a few defaults for shark_args and pytest args (defined in conftest.py)
-Fixes --update_tank option in shark_downloader
removes some multiprocessing in pytest / TF+CUDA support because it breaks pytest and false passes, leaving regressions at large.
-Adds xfails for and removes albert torch from gen_sharktank list (tank/torch_model_list.csv).
-Cleans up xfails for cpu, cuda, vulkan (removing old ones)
 2022-12-16 12:56:32 +05:30
Phaneesh Barwaria
73457336bc add flag for toggling vulkan validation layers (#624) 
* add vulkan_validation_layers flag

* categorize SD flags

* stringify true and false for flag
 2022-12-15 20:40:59 -06:00
nirvedhmeshram
2928179331 Add more NVIDIA targets (#640) 2022-12-15 11:24:38 -06:00
Stanley Winata
24a16a4cfe [Stable Diffusion] Disable binding fusion to work with moltenVK on mac. (#639) 
Co-authored-by: Stanley <stanley@MacStudio.lan>
 2022-12-16 00:22:49 +07:00
yzhang93
6508e3fcc9 Update tuned model SD v2.1base (#634) 2022-12-14 16:02:35 -05:00
Prashant Kumar
898bc9e009 Add the stable diffusion v2.1 version. 2022-12-14 20:19:41 +05:30
Gaurav Shukla
e67ea31ee2 [SHARK][SD] Add --local_tank_cache flag in the stable diffusion 
This flag can be used to set local shark_tank cache directory.

Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-12-14 20:00:25 +05:30
Gaurav Shukla
986c126a5c [SHARK][SD] Add support for negative prompts 
Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-12-14 18:20:09 +05:30
powderluv
5ddce749b8 lint fix 2022-12-13 22:02:32 -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
powderluv
09c45bfb80 clean up cache printf 2022-12-13 14:11:14 -08:00
Boian Petkantchin
e9e9ccd379 Add stress test 2022-12-13 13:21:51 -08:00
Boian Petkantchin
a9b27c78a3 Return dynamic model if specified when downloading from the tank 2022-12-13 13:21:51 -08:00
Boian Petkantchin
bc17c29b2e In get_iree_runtime_config get the specific device instead of the default 2022-12-13 13:21:51 -08:00
Boian Petkantchin
aaf60bdee6 Simplify iree_device_map 2022-12-13 13:21:51 -08:00
Gaurav Shukla
d913453e57 [WEB] Update models to 8dec and also default values (#620) 
1. Update the models to 8 dec.
2. precision is default to `fp16` in CLI.
3. version is default to `v2.1base` in CLI as well as web.
4. The default scheduler is set to `EulerDiscrete` now.

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

Signed-off-by: Gaurav Shukla <gaurav@nod-labs.com>
 2022-12-13 13:08:33 -08:00
powderluv
08e373aef4 Update stable_diffusion_amd.md 2022-12-13 11:47:29 -08:00
Prashant Kumar
4cb50a3d06 Update the models to 8th Dec version. 2022-12-14 00:01:46 +05:30
Prashant Kumar
8ae76d18b5 Add euler scheduler. Also, make it default for sd2.1. 2022-12-13 00:03:45 +05:30
Prashant Kumar
e5be1790e5 Enable the v2.1 base version with --version="v2.1base". (#611) 2022-12-12 07:02:01 -08:00
Ean Garvey
616ee9b824 Don't include baseline benchmarks if setup without IMPORTER=1. (#607) 2022-12-10 14:58:29 -06:00
Stanley Winata
57c94f8f80 [vulkan] Add "radeon" check to the default AMD triple (#604) 2022-12-10 09:05:48 -08:00
powderluv
2a59c4f670 Update stable_diffusion_amd.md 2022-12-09 16:54:47 -08:00
Ean Garvey
0225292a44 Remove print statements from compile utils (#593) 2022-12-08 13:40:47 -08:00
Ean Garvey
589a7ed02f Print a message when a model is downloaded via shark_downloader. (#595) 2022-12-08 15:27:58 -06:00
Quinn Dawkins
b3a42cd0b1 Don't do nchw-to-nhwc transpose for stable diffusion models (#592) 2022-12-08 12:19:23 -05:00
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
Ean Garvey
40eea21863 Enable conv nchw-to-nhwc flag by default for most models + minor fixes (#584) 2022-12-07 16:24:02 -08:00
Ean Garvey
d2475ec169 Add mnasnet to torch models and minor fixes. (#577) 
* Minor fixes to benchmark runner

* Add Mnasnet to tank.
 2022-12-07 22:30:58 +05:30
Stanley Winata
6049f86bc4 [Vulkan][Utils] Automatic platform/OS detection (#569) 
To enable AMD gpus on macOS, we need this detection to let the compiler know that we would be needing moltenVK to use this GPU.
 2022-12-07 12:05:00 +07:00
Phaneesh Barwaria
1096936a15 Enable f32 path for SD (#567) 2022-12-06 19:29:12 +05:30
Stanley Winata
c4444ff695 [vulkan][utils] Add rdna3 detection (#565) 2022-12-05 23:56:06 -08:00
powderluv
2b8d784660 update latest sd build 2022-12-05 22:16:13 -08:00
Daniel Garvey
18f447d8d8 fix hash comparison (#563) 
Co-authored-by: dan <dan@nod-labs.com>
 2022-12-05 21:43:05 -08:00
Daniel Garvey
d7e1078d68 remove nodcloud from client (#562) 
Co-authored-by: dan <dan@nod-labs.com>
 2022-12-05 23:13:19 -06:00
Daniel Garvey
6be592653f remove gsutil_flags and fix download (#559) 2022-12-05 20:29:00 -08:00
Daniel Garvey
8859853b41 Revert "Revert "find gsutil on linux (#557)" (#560)" (#561) 
This reverts commit 3c46021102.
 2022-12-05 20:27:43 -08:00
Daniel Garvey
3c46021102 Revert "find gsutil on linux (#557)" (#560) 
This reverts commit bba8646669.
 2022-12-05 21:53:47 -06:00
Daniel Garvey
bba8646669 find gsutil on linux (#557) 
* find gsutil on linux

* cleaned up downloader and ditched gsutil

Co-authored-by: dan <dan@nod-labs.com>
 2022-12-05 19:03:48 -08:00
Daniel Garvey
b0dc19a910 revert parallel downloads to 1 (#555) 
Co-authored-by: dan <dan@nod-labs.com>
 2022-12-05 15:42:42 -08:00
Daniel Garvey
df79ebd0f2 replace gsutil with variable path for pyinstaller (#541) 
Co-authored-by: dan <dan@nod-labs.com>
 2022-12-05 15:08:57 -08:00