Граф коммитов

3451 Коммитов

Автор SHA1 Сообщение Дата
Ethan Winters 95ea1b81fe
add handling of optional param (#3608) 2024-11-20 09:50:37 -08:00
utkshukla ebebef47cd
3617181 (#3611) 2024-11-20 14:35:22 +05:30
Sai Kothinti 187c551b97
remove benchmark tag for stabilityai-stable-diffusion-2-1 (#3609) 2024-11-20 14:04:27 +05:30
abhishekMS2024 725fc48431
Updated the model card - sample input for facebook-dinov2-base-imagenet1k-1-layer model (#3606) 2024-11-19 18:55:08 +05:30
utkshukla 708876c964
Llama-2 Model Release with latest FMI Version (#3601)
* 3603071

* removed llama-2 chat model updates

* 3617181
2024-11-19 12:36:51 +05:30
Arun 68598bfb03
Add mlflow test to verify mlflow score script (#3604)
* Add mlflow unit test

* Pass model folder to test

* Fix flake errors

* Fix test conda env

* Update test

* Update input args

* Print logs

* Update print

* fix space

* fix doc string
2024-11-18 09:30:12 -08:00
Kelly 302cc6baef
Improve logging for environment release tag already found in copy_assets (#3605)
Co-authored-by: lykelly19 <kellyly@microsoft.com>
2024-11-18 07:58:50 -08:00
saoh f893012484
Updated llama-2 model cards (#3602)
Co-authored-by: Kelly <40868256+lykelly19@users.noreply.github.com>
2024-11-15 14:22:37 -08:00
utkshukla a34b1cc422
FMI Image 58 for Allam-2-7b-instruct and phi-3.5-min-instruct (#3603) 2024-11-15 15:34:16 +05:30
changliu2 e8b5c14f1c
Ignite2024: updated quality metric description on Evaluator Library (#3599)
* Update groundedness pro description.md

* Update groundedness description.md

* Update g.pro description.md

* Update retrieval description.md

* Update retrieval description.md

* Update description.md

* Update relevance description.md

* Update coherence description.md

* Update relevance description.md

* Update fluency description.md

* Update similarity description.md

* Update description.md

* Update f1 description.md

* Update bleu description.md

* Update rouge description.md

* Update gleu description.md

* Update meteor description.md

---------

Co-authored-by: Diondra <16376603+diondrapeck@users.noreply.github.com>
2024-11-13 22:18:31 -05:00
Shu Peng 2a1f10cbe2
Update scripts (#3595) 2024-11-13 19:25:49 +00:00
Sai Kothinti 7c78b01413
Online Evaluation: handle empty data when queried (#3593)
* handle empty data when queried

* fix incorrect span_id, trace_id
2024-11-13 00:57:12 +05:30
Janhavi Mahajan 9ffe6988d6
Create data_mapping for every evaluator passed in evaluator_configuration (#3591)
* Always use new data_mapping dictionary

* Remove dead code
2024-11-12 16:01:01 +00:00
visahan-24 0dc2d60d3e
Distillation batch scoring (#3562)
* Code changes for moving to Batch Scoring for Teacher Model

* Addition of copy right header

* Lengthy line formating fix

* Fix for unterminated f string

* Remove unwanted imports

* File formatting fixes

* File formatting fixes

* Fix for noqa

* Doc String for copy file contents function

* Fix for closing the connection in post processing

* Making changes for fixing pipeline changes

* Version changes

* Merge conflict fixes

* Environmental version fixes

* Adding new line

---------

Co-authored-by: Kelly <40868256+lykelly19@users.noreply.github.com>
2024-11-12 16:15:20 +05:30
Yeshwanth N 863be53185
[BugFix]:Text Classification pipeline,ValueError-trying to set a tensor of incorrect shape in weight (#3592) 2024-11-12 15:57:04 +05:30
cassieesvelt 92b77d33e5
Add gigapath model assets (#3590)
* gigapath don't sync

* add gigapath assets

* remove dummy model
2024-11-11 16:52:56 -08:00
rdondera-microsoft a390ade08f
Enable unit testing for model evaluation component (#3575)
* Workflow for model eval unit tests.

* Make conda verbose.

* Force CPU installation for torch related packages.

* Add pytest as dependent package.

* Install pytest-xdist.

* Modify conda dependencies in workflow configuration.

* Escape slashes in sed command.

* Properly append two lines to conda env file.

* Cosmetics.
2024-11-11 15:52:40 -08:00
abhishekMS2024 3c893e2226
openai-whisper-large-v3 gpu inferencing error fix (#3586) 2024-11-11 22:33:27 +05:30
Yeshwanth N d232c31ac0
[VulnerabilityFix]: Upgrade Werkzeug to 3.1.3 in acpt_image_framework_selector (#3589) 2024-11-11 22:07:29 +05:30
Sai Kothinti 6337cb99ee
add score logic extraction from evaluation sdk for online evaluation feature (#3584)
* add score logic extraction from evaluation sdk

* fix indentation

* fix indentation

* fix doc

* add default trace_id, span_id if user doesn't provides

* fix flake8 reqs

* change score to float and correct logs

* resolve flake issues

* revert unnecessary file changes

* made numeric values compulsory add base condition

* flake 8 whitespace fix

---------

Co-authored-by: apeddauppari <apeddauppari@microsoft.com>
2024-11-11 20:03:49 +05:30
Chandra Sekhar Gupta f6e0a97a18
Migrate evaluationresult assets to internal repo (#3585)
* updating model evaluation environment with modelling_llama patch.

* reverting the changes for transformers patch.

* fixing quality assets and adding latest model details in tag.

* fixing the issues with latest model asset details in tags.

* updating changelog and bumping up azureml-assets version.

---------

Co-authored-by: Chandra Sekhar Gupta Aravapalli <caravapalli@microsoft.com>
2024-11-11 19:03:18 +05:30
sarthaks95 45542d6b1c
Fix vulnerability - model-management - Pin onnx==1.17.0 (#3587)
Co-authored-by: Sarthak Singhal <sarsinghal@microsoft.com>
2024-11-11 17:36:08 +05:30
abhishekMS2024 3406fdac9e
import model error fix for stable-diffusion-xl-base-1.0 (#3581) 2024-11-10 18:13:48 +05:30
Chandra Sekhar Gupta 392f40e6ca
Update the latest model asset ids for evaluationresults of mistral models (#3582)
* updating model evaluation environment with modelling_llama patch.

* reverting the changes for transformers patch.

* update latest model asset ids for mistralai models.

---------

Co-authored-by: Chandra Sekhar Gupta Aravapalli <caravapalli@microsoft.com>
Co-authored-by: vizhur <vizhur@microsoft.com>
2024-11-08 09:33:30 -05:00
Rahul Kumar 702cec7349
Update components for AOAI Evals (#3535)
* add new environment aoai-evaluation

* remove azureml-metrics from dependencies

* pin packages

* reorder logic for batch score result

* fix code style issue

* remove env

* update env name

---------

Co-authored-by: Ankush Bhatia <coolankush07@gmail.com>
2024-11-08 17:53:20 +05:30
Shu Peng cc7556dd57
Upgrade tika server to 3.0.0 (#3579)
* upgrade tika to 3.0.0

* Increase component versions.
2024-11-07 16:49:18 -08:00
Diondra 9d8a31a8ea
Upgrade to v6 (#3578) 2024-11-07 17:55:54 -05:00
apeddauppari 3fcebcb019
AOAI eval fixesOnline Evaluation Schedule AOAI Evaluators Support (#3568)
* AOAI eval fixes

* Fix for evaluator issues and logger

* Bug fixes

* Remove debug logs

* check-style fix and logger changes

* Added logger to check query results

---------

Co-authored-by: Ankush Bhatia <ankushbhatia@microsoft.com>
2024-11-07 23:04:44 +05:30
Yeshwanth N 8efee95899
[BugFix]: Phi-3.5-mini-instruct fails to fine tune with lora (#3577) 2024-11-07 11:55:53 +00:00
Aishwary Gupta 7f842dbf7e
Upgrade ubuntu base image to 22.04 (#3574) 2024-11-07 16:45:41 +05:30
XiangRao 82ec219d64
Upgrate DBCopilot component version (#3572) 2024-11-07 02:52:14 +00:00
Jeff Omhover 049a0afdc8
Add multimodal tag to phi-3.5-vision (#3543)
* Add multimodal tag to phi-3.5-vision

* Update spec.yaml

---------

Co-authored-by: Ali Soylemezoglu <acsoylemezoglu@gmail.com>
Co-authored-by: Kelly <40868256+lykelly19@users.noreply.github.com>
2024-11-06 16:51:58 -08:00
Aditi Singh 834e1262c8
added MLModel files (#3564) 2024-11-07 01:56:37 +05:30
Chandra Sekhar Gupta 30ccdefeb8
Update model asset ids to prod registries for evaluationresult quality assets of phi models (#3571)
* updating model evaluation environment with modelling_llama patch.

* reverting the changes for transformers patch.

* updating quality assets for phi models.

---------

Co-authored-by: Chandra Sekhar Gupta Aravapalli <caravapalli@microsoft.com>
2024-11-06 21:46:45 +05:30
Chandra Sekhar Gupta d3091b6b2b
Add quality benchmark tag for phi-35-MoE-instruct model (#3573)
* updating model evaluation environment with modelling_llama patch.

* reverting the changes for transformers patch.

* add quality benchmark tag for phi-35-moe-instruct model.

---------

Co-authored-by: Chandra Sekhar Gupta Aravapalli <caravapalli@microsoft.com>
2024-11-06 21:21:22 +05:30
Yeshwanth N 2442cb2caf
[VulnerabilityFix]: Uprage onnx to 1.77.0 in acft-mmtracking-video-gpu,acpt-automl-image-framework-selector-gpu,acft-multimodal-gpu (#3570) 2024-11-06 14:02:04 +05:30
Jayesh Tanna 3ae6492140
migrating component environment (#3505)
* migrating component env to assets

* adding category

* changing category

* resolving pr comments

* making comp images public

* upgrading to ubuntu22 for component version
2024-11-06 10:00:44 +05:30
abhishekMS2024 c7e3e6820d
Added support for model_framework for import_model component (#3566) 2024-11-06 09:09:49 +05:30
abhishekMS2024 a47fb18726
Update the model_flavor to OSS (#3560) 2024-11-06 09:02:04 +05:30
Diondra bd9a588a04
Remove preview tags from GA built-in evaluators (#3569)
* Remove preview tags from GA evaluators

* Fix formatting error with empty tags
2024-11-05 18:06:09 -05:00
Rahul Kumar 5fffc9244c
Add new environment aoai-evaluation (#3531)
* add new environment aoai-evaluation

* remove azureml-metrics from dependencies

* pin packages

* reorder logic for batch score result

* fix code style issue
2024-11-05 15:40:02 +00:00
Rahul Kumar 4c2a6914bc
upgrade forecasting components (#3563) 2024-11-05 10:51:15 +05:30
Chandra Sekhar Gupta 3defa2e329
fixing quality assets and adding latest model details in tag. (#3558)
* updating model evaluation environment with modelling_llama patch.

* reverting the changes for transformers patch.

* fixing quality assets and adding latest model details in tag.

* fixing the issues with latest model asset details in tags.

---------

Co-authored-by: Chandra Sekhar Gupta Aravapalli <caravapalli@microsoft.com>
2024-11-05 06:39:33 +05:30
Kelly d958a5cd0b
Revert "Code changes for moving to Batch Scoring for Teacher Model (#3507)" (#3561)
This reverts commit 04eff530e2.
2024-11-04 08:43:59 -08:00
rdondera-microsoft 13b756960b
Fix input column names for OD/IS components (#3557)
* Fix input column names for vision finetuning validation.

* Update input column names for classification, for consistency purposes.
2024-11-04 07:58:51 -08:00
visahan-24 04eff530e2
Code changes for moving to Batch Scoring for Teacher Model (#3507)
* Code changes for moving to Batch Scoring for Teacher Model

* Addition of copy right header

* Lengthy line formating fix

* Fix for unterminated f string

* Remove unwanted imports

* File formatting fixes

* File formatting fixes

* Fix for noqa

* Doc String for copy file contents function

* Fix for closing the connection in post processing

* Making changes for fixing pipeline changes

* Version changes

* Merge conflict fixes

* Environmental version fixes
2024-11-04 18:41:37 +05:30
Chandra Sekhar Gupta 4f983518a6
Fix the issue with supporting new evaluationresult asset types (#3544)
* updating model evaluation environment with modelling_llama patch.

* add cost, quality, performance tags for evaluationresult.

* reverting the changes for transformers patch.

* updating the changelog and bump up the version.

* fix the issue with supporting new evaluationresult assets.

* updated the change log file.

* updating the test cases.

* updating the test cases.

* fixing unit tests.

* fixing change log file.

---------

Co-authored-by: Chandra Sekhar Gupta Aravapalli <caravapalli@microsoft.com>
2024-11-04 08:34:59 +05:30
sarthaks95 81f80b9b91
Update model versions - phi-3-medium-4k-instruct, phi-3-medium-128k-instruct, phi-3-small-8k-instruct, phi-3-mini-4k-instruct, phi-3-mini-128k-instruct (#3545)
Co-authored-by: Sarthak Singhal <sarsinghal@microsoft.com>
2024-11-02 18:22:57 +05:30
Diondra 74d61d677e
Update groundedness pro asset description (#3556) 2024-11-01 14:41:00 -04:00
apeddauppari 4d545071e8
Add AppInsights ServiceName Implementation (#3555)
* Add service name implementation

* indentation fix

* fstring indentation fix

* Evaluator Built in image bump to v6

* indentation fix - no additional line

---------

Co-authored-by: Shruti Iyer <iyershruti@microsoft.com>
2024-11-01 22:29:37 +05:30