Skip to content
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

Error: Failed to create TFLite interpreter from model #65

Open
Swikriti0998 opened this issue May 13, 2024 · 8 comments
Open

Error: Failed to create TFLite interpreter from model #65

Swikriti0998 opened this issue May 13, 2024 · 8 comments

Comments

@Swikriti0998
Copy link

I am trying to load the tensor flow model but its giving me the below error, I have my pose-detection.tflite in assets/pose-detection

error: Failed to load Tensorflow Model 3! [Error: Failed to create TFLite interpreter from model "http://192.168.1.10:8081/assets/?unstable_path=.%2Fassets%2Fpose-detection/pose-detection.tflite?platform=ios&hash=361a28992b649751d6d66cd8cfb1e2a6"!]

I tried both true and false inside the app.config.ts file.

'react-native-fast-tflite',
{
enableCoreMLDelegate: false/true,
}

@vmelentev
Copy link

vmelentev commented May 13, 2024

I do believe you have to edit your metro.config.js file to configure your project to be able to interoperate .tflite files.

Try adding this code to metro.config.js:


config.resolver.assetExts.push(
    'tflite'
);

@Swikriti0998
Copy link
Author

Already edited my metro config and it's reading the tflite file otherwise it would give below error.
error - Unable to resolve "@assets/pose-detection/pose-detection.tflite" from "src/screens/home/index.tsx"

However while loading the application I am getting the below error
error - {"error": [Error: CoreML Delegate is not enabled! Set $EnableCoreMLDelegate to true in Podfile and rebuild.], "model": undefined, "state": "error"}

@mrousavy
Copy link
Owner

did you rebuild pods & rebuild your app after enabling the core ml delegate?

@juv85
Copy link

juv85 commented Aug 9, 2024

Hello, @Swikriti0998 did you solve this ? I encounter the exact same issue.

@Swikriti0998
Copy link
Author

@mrousavy, yes. Cleared cache build the app again but same error.

@juv85, no I couldn't.

@juv85
Copy link

juv85 commented Sep 20, 2024

@Swikriti0998 what framework did you use to build your model?
For my case I realized that fine tuned models trained with anything other than tensorflow model maker throw this error.

After researches I saw this is related to the model's architecture. This library doesn't seem to take different model architectures into account for the moment (I might be wrong but this is my understanding)

@lucksp
Copy link

lucksp commented Nov 20, 2024

What's interesting is I also noticed issues with newly trained models, from VertexAI exported as TFLite. My older model from Aug 26 works, but the model from Oct 3 will not load. I tried updating package to newer version.

@nzxtua
Copy link

nzxtua commented Jan 8, 2025

I also experiencing the error. Our project is currently on react-native 0.72.6 and react-native-fast-tflite 1.5 1 with delegates enabled (core-ml/nnapi). There are three .tflite files in the repo: Int8, Float16 and Float32. On Android Int8 and Float32 models work great, but Float16 won't load. On iOS all three models are throwing the error. I don't know exact details, about the files, when and how, they were converted. Just pointing the weird behavior, and will keep you updated if any updates on the topic.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

6 participants