Skip downloading litert models if they already exist. #63
Conversation
This file contains hidden or 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
Implement a caching mechanism using a JSON file to track the state of downloaded models. This avoids redundant downloads during prebuild if the model configuration has not changed. A --force flag is also added to allow users to manually bypass the cache and re-download all models when necessary.
rmahdav
reviewed
Jan 22, 2026
| fs.mkdirSync(MODEL_DIR, { recursive: true }); | ||
| } | ||
|
|
||
| if (!forceDownload && fs.existsSync(CACHE_FILE)) { |
Collaborator
There was a problem hiding this comment.
This PR is a great improvement to the benchmark, thanks Brendan!
However, this looks like all-or-nothing situation now. Meaning, if just one of the models has changed, all the models will be downloaded again. I think it is better if we check the model changes and skip downloading them one by one. Something like:
for (const modelInfo of MODELS_TO_DOWNLOAD) {
...
const cachedEntry = cachedModels.find(m => m.filename === filename && m.repo === repo);
const configMatches = cachedEntry && JSON.stringify(cachedEntry) === JSON.stringify(modelInfo);
const fileExists = fs.existsSync(outputPath);
if (!forceDownload && configMatches && fileExists) {
console.log(`Model **${filename}** is up to date. Skipping.`);
newCache.push(modelInfo);
continue;
}
// Download if missing or changed
...
}
(The code snippet is Gemini suggested to my change request, so take it with a grain of salt)
Collaborator
Author
There was a problem hiding this comment.
Oh yeah, I was trying to do something quick for testing. It's pretty easy to do per file though, so I'll change it. I added a little helper class in this other PR. Once that lands I'll update this one do the same thing.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Implement a caching mechanism using a JSON file to track the state of downloaded models. This avoids redundant downloads during prebuild if the model configuration has not changed.
A --force flag is also added to allow users to manually bypass the cache and re-download all models when necessary.