1745 Commits

Author SHA1 Message Date
Vivek Khandelwal
015d0132c3 Modify falcon pipeline to add fp16 support (#1551) 2023-06-19 09:57:13 -07:00
powderluv
20ddd96ef7 unpin diffusers (#1550) 2023-06-18 13:45:55 -07:00
powderluv
ee33cfd2d1 Add PIL in main index.py (#1549)
* Add PIL in main index.py

This is to ensure pyinstaller picks it up

* Update index.py
20230618.775
2023-06-18 11:51:44 -07:00
Stefan Kapusniak
a3cba21d5b Fix load of unet512 vmfb fail on get of iree opts (#1546)
* Change retrieval of Iree options used when loading an existing
unet512 vmfb to look up the "unet" options rather than attempt to
find a non-existent set of options for "unet512"

Co-authored-by: powderluv <powderluv@users.noreply.github.com>
20230618.774
2023-06-18 06:42:20 -07:00
Stefan Kapusniak
a7b6ec4095 Fix unet512 always being used when --max_length=77 (#1547)
* Switches a few places in the SD pipeline where an assumption of
max_length=64 was being made, to using the actual max_length
as passed into the pipeline. This prevents unet512 always being
used and producing different images than previously when
--max_length=77
2023-06-18 06:41:25 -07:00
Ean Garvey
d80b087d95 Add PIL hidden imports to sd spec. (#1544)
Co-authored-by: powderluv <powderluv@users.noreply.github.com>
2023-06-18 06:39:08 -07:00
Stefan Kapusniak
297a209608 Remove workarounds for gradio tempfile bugs (#1548) 20230617.773 2023-06-17 19:50:36 -07:00
gpetters94
b204113563 Add UNet512 (#1504)
Co-authored-by: Ean Garvey <87458719+monorimet@users.noreply.github.com>
2023-06-17 03:46:25 -04:00
Chi_Liu
f60ab1f4fa Add Deberta to stablehlo in shark tank (#1545) 20230616.772 2023-06-16 13:24:44 -07:00
Surya Jasper
b203779462 Added Adreno target triples to vulkan_utils (#1543) 20230615.770 2023-06-15 16:42:59 -07:00
Stefan Kapusniak
38570a9bbb Some Fixes for update to gradio 3.34.0 (#1538)
* Fixes randomize seed buttons that stopped working.
* Update now deprecated method to set initial colums for output
gallery to the newer undeprecated one.
2023-06-15 01:10:36 -07:00
dependabot[bot]
a5c882f296 Bump gradio from 3.15.0 to 3.34.0 (#1518)
Bumps [gradio](https://github.com/gradio-app/gradio) from 3.15.0 to 3.34.0.
- [Release notes](https://github.com/gradio-app/gradio/releases)
- [Changelog](https://github.com/gradio-app/gradio/blob/main/CHANGELOG.md)
- [Commits](https://github.com/gradio-app/gradio/compare/v3.15.0...v3.34.0)

---
updated-dependencies:
- dependency-name: gradio
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
20230614.769
2023-06-14 18:13:48 -07:00
Ean Garvey
eb6d11cfed Change mlir dialects for tf tests to stablehlo. (#1535)
* Change mlir dialects for tf tests to stablehlo

* Update shark_runner.py
2023-06-14 10:43:49 -07:00
Vivek Khandelwal
46184a81ac Add Falcon pipeline (#1534) 20230614.768 2023-06-14 09:39:16 -07:00
PhaneeshB
149165a2f0 add multi-device mutli-precision vmfb names 2023-06-14 22:08:24 +05:30
dan
bec82a665f mega vicuna merge
single endpoint in apps/language/models/scripts/vicuna.py
removed main functions from pipelines
replaced divergent utils compile with shark_importer
adds support for different precisions
2023-06-14 19:06:29 +05:30
Ean Garvey
9551490341 Remove deprecared --iree-mhlo-demote-164-to-132 flag usage. (#1533) 20230613.767 2023-06-13 22:40:47 -05:00
Ean Garvey
49b3ecdbca (pytest) don't run redundant tests in cpu suite (#1532) 2023-06-13 22:40:33 -05:00
Ean Garvey
f53e3594c3 OPT Refactor (#1516)
* Change script to 1.3b model and add pytorch comparison

* fix CLI command

* Match OPT transformers model updates + numerics against latest version

* Cleanup OPT sentence completion script.

* Fix formatting and add standalone validation scripts.

* Add minimal OPT wrapper and example with import_with_fx

* Rename OPT full model wrapper.

* Cleanup test scripts for OPT.
2023-06-13 22:40:07 -05:00
Ean Garvey
5562d1dfda Fix xfails for cpu pytest cases (#1527)
Adding cpu-sync and cpu-task device configs was allowing respective tests to bypass the xfail conditional for cpu pytests marked in tank/all_models.csv. This commit updates the conditional to xfail those cases for cpu-sync and cpu-task as well.
2023-06-13 17:01:51 -07:00
Stefan Kapusniak
c7b0c2961e UI/Web Improve output gallery temp file handling (#1531)
* On startup report that cleaning up of temp files is taking place, in
case it takes a long time.
* Have the output gallery tab delete any zero length temporary files
generated by gradio < 3.32.0 for its gallery control whenever it
needs to update that control with images. This prevents such
files multiplying out of control.
2023-06-13 16:25:37 -05:00
Ean Garvey
44273b0791 Fix conditional in transform_fx() (#1530) 2023-06-13 16:24:53 -05:00
Prashant Kumar
0a4c8fcb3e Minor changes in the fx transforms. 20230613.766 2023-06-13 21:23:35 +05:30
Stefan Kapusniak
2fec3c8169 re-indents add_upcast in shark importer (#1523)
* The two with blocks in add_upcast appear to be underindented making
SD 1.4 break on rdna3, I've pushed them out one more tab, and then
everything appears to work again.
2023-06-12 14:41:10 -05:00
Gaurav Shukla
5e7d5930dd [vicuna] Add device and precision propagation in vicuna (#1520)
Signed-off-by: Gaurav Shukla <gaurav@nod-labs.com>
2023-06-12 12:14:43 -05:00
Prashant Kumar
b6dbd20250 Modify the fx transforms. (#1521)
- The bounds are set properly.
- The upcasting and downcasting is done for vicuna.
2023-06-12 09:40:14 -07:00
Nithin Meganathan
34f1295349 Add a model config generator (#1511)
Model config generator takes a PyTorch model as input and generates a JSON file with model layers and other propperties that define sharding on a particular hardware.
2023-06-09 15:32:00 -07:00
Phaneesh Barwaria
1980d7b2c3 Cpu device map (#1515)
* update cpu iree device

* fix vmfb paths vic unsharded
2023-06-09 11:27:02 -05:00
powderluv
2cfacc5051 fix osx torch_mlir (#1513)
* fix osx torch_mlir

* Update index.py

* Update index.py
2023-06-09 00:57:26 -07:00
Phaneesh Barwaria
436f58ddc4 cli using generate and mem fixes (#1509) 20230608.761 2023-06-08 13:13:32 -05:00
Phaneesh Barwaria
6b29bd17c8 Enable compilation vicuna (#1507)
* add cli for unsharded vic

* enable mlir download and compile
20230607.760
2023-06-07 13:08:22 -07:00
Ean Garvey
2c3485ca3e Add standalone OPT sentence completion script. (#1506) 2023-06-07 10:58:03 -07:00
Daniel Garvey
f206ecc635 reenable compilation in vicuna pipeline, add flags (#1505)
* replace vicuna.py backend with pipeline

* add some memory management to fist vicuna compile

reenable compilation
2023-06-07 09:49:27 -07:00
Stefan Kapusniak
a187e05ae6 Prevent having no cuda devices breaking the UI (#1503)
Don't break the UI when the LLM tab only wants cuda devices but there
aren't any.
20230606.759
2023-06-06 11:41:16 -07:00
Gaurav Shukla
8c21960486 [vicuna] Set only cuda devices in vicuna UI for now
Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2023-06-06 22:15:20 +05:30
Gaurav Shukla
be62fce676 [vicuna] Fix vicuna chatbot (#1499)
Signed-off-by: Gaurav Shukla <gaurav@nod-labs.com>
2023-06-06 09:23:32 -07:00
PhaneeshB
f23b778a6c remove old vicuna scripts 2023-06-06 21:35:58 +05:30
PhaneeshB
436edf900d add vic sharded pipeline 2023-06-06 21:35:58 +05:30
Gaurav Shukla
ed58c2553f [vicuna] Integrate vicuna in shark studio
Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>
2023-06-06 20:57:48 +05:30
Stefan Kapusniak
f2ca58e844 Add .csv and .json param info to output gallery (#1495) 2023-06-06 07:08:34 -07:00
Ean Garvey
1dbcc736eb [SD] (RDNA2) Enable new tuning for sd1.4 (#1498) 2023-06-06 06:48:58 -07:00
Phaneesh Barwaria
a83808ddc5 Vicuna cuda on A100 40G (#1496)
* vic chat with memory management (precompiled vmfb)

* fix vmfb path and download
2023-06-06 15:10:33 +05:30
Ean Garvey
a07fe80530 Update OPT, ResNet example scripts. (#1492)
* Update API in OPT example.

* fix resnet50 script

* Add OPT1.3b test script.
20230605.757 20230605.756
2023-06-05 20:19:35 -07:00
Ean Garvey
d0ba3ef8fa disable use_tuned on SD1.4 for rdna2 (#1490)
this is a temporary measure while we retune SD1.4 for rdna2. The current config fails during iree-compile.
2023-06-05 19:46:16 -05:00
Stefan Kapusniak
8400529c2c Fix output gallery not using shark_tmp (#1493)
This fix the gallery component of the  output gallery dumping temporary
files into the standard folders rather than shark_tmp so those files never
got cleared out on restart and would build up.
2023-06-05 16:23:49 -05:00
powderluv
7eaee9c242 update SHARK to nodai SHARK 20230605.755 2023-06-05 00:44:49 -07:00
powderluv
8230eebce5 Switch to CPU torch builds for shark.whl 20230605.754 2023-06-05 00:36:03 -07:00
Ean Garvey
6296ea4be9 fix config handling for sd1.4 on rdna2 (#1489) 2023-06-05 00:02:30 -07:00
Ean Garvey
4151ec3a8f (pytest) tag efficientnet, mobilenet as xfails on vulkan (#1488) 2023-06-04 23:22:32 -07:00
powderluv
a2467e8d43 Enable SHARK whl packages 2023-06-04 23:21:22 -07:00