-
## 🚀 Feature
Create new `CheckpointIO` classes that allow:
- Splitting a checkpoint by keys, enabling #5339
- Splitting a checkpoint by size, relaxing the memory and storage constraints
### M…
-
### You know, streamlined crypto is the kind of Agile ICO that we need. Why don't we parse the existing Interne…
-
## **Overview**
We are excited to expand the capabilities of the Livepeer [AI Network](https://docs.livepeer.org/ai/pipelines/overview#models-on-the-ai-subnet/) by developing a robust `object detec…
-
Seems like no releases or PRs being merged for sometime now. In case this project is deprecated, what are suggested pathways for a user to enable distributed learning in the cloud please?
-
## 🚀 Feature
It is often beneficial to have an explicit guarantee that the callbacks will be called in the order that they were added to the callback list. If I understand it right, the order r…
z-a-f updated
5 months ago
-
# Is it a Highly scalable realtime framework ?
Yes! for NodeJS, GraalJS, embedded IoT, and the Browsers
# is it an ENTERPRISE-CLASS, MULTICLOUD TO EDGE NOSQL DATABASE?
Yes! Develop with agility. …
-
## 🚀 Feature
I propose adding instead of batch size a dictionary with batch size per GPU, for example {"cuda0":4, "cuda1", 6}
### Motivation
I have gv100 (32gb) and 3090 (24gb). using the cur…
-
### Description
Hello Ray maintainers and community,
we've been using Ray for our works and find it to be a valuable tool for scalable and distributed machine learning. I believe it would be ben…
-
## 🚀 Feature
I am wondering if it is possible to include the asynchronous evaluation during the training process
### Motivation
For RL projects (or imitation training + online rollout evaluat…
-
## 🚀 Feature
### Motivation
DeepSpeed provides its own communication to improve DeepSpeed speed.
Here is an example: https://github.com/microsoft/Megatron-DeepSpeed/pull/50/files
#…