Comments (1)
The option is set in the same way for XLA GPU (both TPU and GPU use the default value for this flag). So it is not so easy to fix it like that, and there was some doubt whether it can be fixed apart from avoiding the reshape bitcast (which might be added on model side).
Quoting from the chat: "A reshape of a dus is not longer the same tensor. A copy might be needed. Not always though"
from paxml.
Related Issues (15)
- ERROR: error loading package 'paxml' HOT 7
- Error running Common Crawl example HOT 2
- Installing paxml from source failed due to dependency problem HOT 5
- What does USE_REPEATED_LAYER do? HOT 1
- ARM64 Build HOT 4
- Int8 checkpoint
- How to continue training from a checkpoint?
- Pipeline Parallelism: F external/org_tensorflow/tensorflow/compiler/xla/array.h:446] Check failed: n < sizes_size Fatal Python error: Aborted HOT 3
- Pipeline Parallelism: USE_REPEATED_LAYERS bug HOT 3
- Remaining tutorials
- [Feature Request] Need ZeRo-1/2 to cooperate with PP+TP+DP. Which may more faster than FSDP sometimes.
- [Question] Very low MFU(30%~35%) when train bf16 Llama2 and GPT model with single SXM4 A100 machine.
- Use bfloat16 for eval HOT 1
- Jax + tpu and AQT int8 train model loss is abnormal
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 paxml.