Comments (2)
breizhn
commented on June 19, 2024
The model can not be converted to TF lite yet.
One of the reasons is that TF lite's exp() does not support complex64 so it can not be converted and LSTMs are still a tricky thing in TensorFlow Lite. The TensorFlow team is working on support for standard TensorFlow operations, but this is still under development and there isn't a python runtime for this yet. See this guide for more informations.
And actually you tried to convert a .h5 weight file to a tf lite model. The .h5 file in this repository contains just the the weights and no model description. You have to build the model first to convert it. See the run_evaluation.py for building the model and loading the weights.
But if you are looking for fast inference, maybe use the SavedModel format. This one is already running fast and is portable.
from dtln.
breizhn
commented on June 19, 2024
Tf lite conversion is no added.
from dtln.
Related Issues (20)
- Train stateful DTLN model HOT 6
- Pesq score and data duration HOT 1
- TF_DETERMINISTIC_OPS HOT 2
- ValueError: trying to load 500h file h5 with 10 layers but model architecture has 9 layers HOT 5
- Tf_lite HOT 9
- Question about real-time implementation using tensorflow lite HOT 5
- Increasing Loss and Validation Loss HOT 1
- n times of 10s data inference, the result from the first 10 seconds will be different from the next ten seconds
- Dimension mismatch error for real_time_dtln_audio.py HOT 1
- PESQ and STOI calculator HOT 1
- Can DTLN be combined with AEC in real time communication? HOT 1
- How to modify DTLN_model.py files to support the latest tensorflow 2.9
- ValueError: Unexpected result of `train_function` (Empty logs). HOT 1
- Validation loss is too high HOT 1
- Creating integer only models HOT 11
- About one model HOT 1
- Using convert_weights_to_tf_lite.py did not produce the same results as the pre-training model. HOT 3
- Are my val_loss values valid? HOT 2
- Is it necessary to apply a window after fft? HOT 1
- The length of the audio after noise reduction is shortened HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from dtln.