Blog post on how to use olive to find out if quantizing before
fine-tuning provides better model quality.
---------
Co-authored-by: Jambay Kinley <jambaykinley@microsoft.com>
### Description
Update TensorRT Documentation
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
Addresses #21638
Changed install to genai, moved genai section up, accessibility changes
in docs/install.
Still todo (another PR, soon to come):
- Change theme for accessibility
- Change videos linked on homepage
- Update homepage content to more closely reflect current ORT status.
Preview at : https://maanavd.github.io/onnxruntime/
### Description
This PR updates the Phi-3 vision tutorial to work with the latest ONNX
models uploaded and replaces Phi-3 vision with Phi-3.5 vision as the
example model.
### Motivation and Context
This helps document how to use the new uploaded ONNX models with ONNX
Runtime GenAI v0.5.1.
### Description
<!-- Describe your changes. -->
The Getting Started support matrix and ORT Release Roadmap have changed
since the ORT 1.20.0 release.
Updated w/ info about new pkgs + CUDA default version and added 1.20.1
patch release plans.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
Add Set runtime option for C API in GenIA
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
<!-- Describe your changes. -->
Included new version info for CUDA, TensorRT, and DirectML execution
providers following ORT 1.20 release publish.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
<!-- Describe your changes. -->
Updating links in the operator docs to reference releases 1.19 + 1.20.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
<!-- Describe your changes. -->
Updating ORT Release Roadmap to remove features no longer planned, add a
link to RC1 on GitHub, and change GenAI --> generate() API.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
<!-- Describe your changes. -->
Updating release roadmap page to include new announcement, remove
features that are no longer planned, and add Olive section to feature
list dropdown.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
<!-- Describe your changes. -->
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
Update eps.md about Python wheels for MIGraphX and ROCm
### Motivation and Context
There was no instructions to show how to build python wheels for
MIGraphX and ROCm backends. It is neccesary to add these command
examples.
At some point will need to merge cards + testimonials objects to be one.
As always, preview at: https://maanavd.github.io/onnxruntime/
---------
Co-authored-by: Sophie Schoenmeyer <107952697+sophies927@users.noreply.github.com>
### Description
Updated Announcements section to separate 2 announcements into list +
bold them. These changes will make Python announcement more visible to
users.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
Correct variable name from `test_data` to `feed_dict` to fix example
code in mixed precision example docs.
### Motivation and Context
Fixes#21822
### Description
This PR updates the documentation regarding the ACL execution provider
to reflect the changes made in #22032.
It also updates various references to Arm's trademarks in accordance
with the Arm trademark use guidelines.
### Motivation and Context
Ensures documentation is accurate following recent changes.
---------
Signed-off-by: Michael Tyler <michael.tyler@arm.com>
### Description
This PR contains the steps of building onnxruntime in AIX OS.
### Motivation and Context
AIX OS users can build the onnxruntime in their local env.
### Description
Add download links to previous versions of onnxruntime-gpu.
### Motivation and Context
There was a change in feeds recently, which make it hard to find the
download links for previous versions of onnxruntime-gpu.
### Description
* Add CUDA/CuDNN requirements for 1.19.x
* Update installation guide for cuda 11 package and nightly.
* Remove a sentence for ROCm: `Alternatively, each major ORT release has
a corresponding C/C++ ROCm package, found
[here](https://github.com/microsoft/onnxruntime/releases/).` since
onnxruntime_rocm package is not available in recent releases.
### Motivation and Context
Doc is not up to date.
Likely will need to change HLJS theme for the rest.
Should be good to instantly approve/merge, but feel free to review as
necessary. Not currently deploying a preview as another more significant
PR is being reviewed, but I can do it if requested :)
Update according to the change request from NimbleEdge
### Description
<!-- Describe your changes. -->
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
