forked from otcshare/chromium-src
-
Notifications
You must be signed in to change notification settings - Fork 0
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Prototype additional WebNN operator implementations via DirectML #1
Open
fdwr
wants to merge
72
commits into
dml_base
Choose a base branch
from
dml_sd
base: dml_base
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
fdwr
changed the title
Prototype additional op implementations via DirectML
Prototype additional WebNN operator implementations via DirectML
Apr 17, 2023
…ConvTranspose2d and Split
…t cannot be created
Reallocate DescriptorHeap if the count increases when executing an op
Fix duplicated graph inputs issue
The root cause is IDMLBindingTable::BindOutputs() use the references to the DML_BUFFER_BINDING objects storaged in base::flat_map<>. These references may become invalid if the flat_map reallocates the storage when adding more outputs. Fix this issue by reserve the flat_map capacity to avoid the reallocation.
Fix the issue of calculating Resample2d's scales
Fix resample2d scales calculation and mode setting issue
1. Avoid mapping shared memory for every compute 2. Copy input/output and uploading/readback buffers in one big chunk
Optimize shared memory mapping and copy for input and output buffers
Count operands index from 1
… in MojoGraph::OnGraphCreated (thanks Mingming)
Non-controversial nits.
Apply CommandQueue::WaitAsync() to not block GPU main thread
Post IDMLDevice1::CompileGraph into other threads
For NPU, use D3D feature level core, switch to command list type COMPUTE, and re-enable DXCore over DXGI
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
*Not for actual review, just for comparison, since many of the needed operators do not even exist yet in WebNN v1, and since the Chromium main branch has many differences since this was forked from Mingming's original branch.