Comments (6)
@vrv Any news on fixed-point / quantized data type support?
I am designing an FPGA-based accelerator for CNNs as my master thesis in the next few months. It would be awesome to be able to train the models with fixed-point weights and activations.
Do you have any roadmap for fixed-point support?
from tensorflow.
We are working on support for operators and kernels that take quantized (fixed-point) data -- we're still working on the right APIs to expose them, and with that we'll have more documentation on why and how they are used. Stay tuned!
from tensorflow.
Awesome, this is great work 😮
Please give my sincere thanks to 🎅
from tensorflow.
Closing for now. When the quantized ops come online, they'll come with documentation.
from tensorflow.
@dgschwend have you find the answer?
from tensorflow.
@vrv could you please reopen, point to the correct doc, or confirm that is still should be considered internal (though it is exposed in many place by now)
from tensorflow.
Related Issues (20)
- tf.range miss some dtypes support HOT 1
- Custom metrics results returned by 'History' callbacks doesnt work properly HOT 2
- Inspect `tf.data.AUTOTUNE `values during runtime HOT 1
- `tf.distribute` and `keras` 3 incompatibility: ValueError: Invalid reduction dimension 0 for input with 0 dimensions. HOT 1
- error in perks
- tensorflow.python.framework.errors_impl.AbortedError: Exception encountered when calling MaxPooling1D.call() HOT 4
- TensorFlow 2.16 Training an image Autoencoder output pure black image while 2.10 is normal HOT 2
- [Question] - Extend some GPU op HOT 1
- Failed to load native TensorFlow Lite methods. Check that the correct native libraries are present HOT 3
- Failed to load native TensorFlow Lite methods. Check that the correct native libraries are present HOT 2
- Unpublished CVE-2023-33976 affecting TF<2.13.0
- Distributed Training without strategies- can't average gradients HOT 1
- Does TensorFlow2.13.0 support RISC-V HOT 1
- error: 'NPY_NTYPES' was not declared in this scope; did you mean 'NPY_TYPES'?
- A KerasTensor is symbolic: it's a placeholder for a shape an a dtype. It doesn't have any actual numerical value. You cannot convert it to a NumPy array.
- TensorFlow Tutorials not compatible with latest release.
- DLL load failed while importing _pywrap_tensorflow_internal
- TfLiteVision can't be initialised HOT 1
- Issue with tf version 2.15 (ModuleNotFoundError: No module named 'tensorflow.python.distribute.distribution_strategy_context')
- tf.distribute.MirroredStrategy error: Multiple OpKernel registrations match NodeDef at the same priority
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 tensorflow.