Comments (2)
SuryanarayanaY
commented on June 10, 2024
Hi @wzzll123 ,
I have tested the code with tf-nightly and replicated the precisional differences wrt float32 and bfloat16. Attached gist for reference. There seems to be an issue.
from tensorflow.
wzzll123
commented on June 10, 2024
Hi @SuryanarayanaY , thanks for replying.
The precision discrepancies may stem from the fact that activation operators' computations don't incorporate precision improvement. Similar issues have arisen in PyTorch, which were fixed by using precision improvement. Perhaps TensorFlow could address this issue similarly?
from tensorflow.
Related Issues (20)
- Efficientdet models do not work on hexagon delegate
- Tensorflow lite conversion fails for keras model in TensorFlow version: 2.16.1 HOT 1
- Unable to build TensorFlowLite GPU Delegate for Android
- Build from source C doesn't produce .tar.gz archive
- Aborted (core dumped) in `tf.raw_ops.ResourceSparseApplyMomentum` HOT 3
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyFtrl/tf.raw_ops.ResourceApplyFtrlV2/tf.raw_ops.ResourceSparseApplyFtrl/tf.raw_ops.ResourceSparseApplyFtrlV2`
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyKerasMomentum/tf.raw_ops.ResourceSparseApplyKerasMomentum` HOT 1
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyPowerSign` HOT 3
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyRMSProp/tf.raw_ops.ResourceSparseApplyRMSProp`
- Aborted (core dumped) in `tf.raw_ops.ResourceSparseApplyAdadelta/tf.raw_ops.ResourceApplyAdadelta` HOT 1
- Aborted (core dumped) in `tf.raw_ops.ResourceSparseApplyAdagrad/tf.raw_ops.ResourceSparseApplyAdagradDA/tf.raw_ops.ResourceSparseApplyAdagradV2` HOT 1
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyAdagrad/tf.raw_ops.ResourceApplyAdagradDA/tf.raw_ops.ResourceApplyAdagradV2`
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyCenteredRMSProp/tf.raw_ops.ResourceSparseApplyCenteredRMSProp` HOT 1
- Aborted (core dumped) in `tf.raw_ops.ResourceSparseApplyProximalAdagrad` HOT 3
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyAdaMax/tf.raw_ops.ResourceApplyAdam/tf.raw_ops.ResourceApplyAdamWithAmsgrad`
- Aborted (core dumped) in `tf.raw_ops.ResourceApplyAddSign` HOT 3
- Failed to build TF Lite benchmark tool on Windows
- The signatures in SavedModel do not contain serving_default when a subclass of the keras model has multiple inputs
- MobileNetV3 quantization HOT 2
- Crash in `tf.raw_ops.ResizeNearestNeighbor/ResizeNearestNeighborGrad/ResizeArea/ResizeBicubic/ResizeBilinear` 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 tensorflow.