-
### System Info
- `transformers` version: 4.21.2
- Platform: Linux-5.10.135-122.509.amzn2.x86_64-x86_64-with-glibc2.2.5
- Python version: 3.8.5
- Huggingface_hub version: 0.10.0
- PyTorch versi…
-
**Submitting author:** @ayuei (Vincent Nguyen)
**Repository:** https://github.com/Ayuei/DeBEIR
**Branch with paper.md** (empty if default branch): paper
**Version:** v0.0.1
**Editor:** @arfon
**Review…
-
**Project description**
Ray is a fast and simple framework for building and running distributed applications.
It is packaged with RLlib, a scalable reinforcement learning library.
The project p…
-
Hello
Thank you for sharing your materials!
And I am very happy with your modified Flow. In the past, I tried to install Flow from the official repo but it always had errors. With your repo, it is e…
-
## ❓ Questions and Help
I am new to Pytorch and distributed learning. I am using mlagents to do deep reinforcement learning. Their source code does not support training with multiple GPUs. Therefor…
-
I've got a hypothesis about the tendancy of ChatGPT to agree with marketing lies over user opinion. I think it's because rude, crass, brutally honest opinions are marked as "toxic" and are therefore f…
-
I executed the run_continuous.py file for the continuous agent and found that the policy loss increased approximately linearly with training episodes until it stabilized. Why is the policy loss not re…
-
This is very painful. I wish we would stop committing images here.
```
$ git pull
remote: Enumerating objects: 1347, done.
remote: Counting objects: 100% (1347/1347), done.
remote: Compressin…
-
I am working on a reinforcement learning project using Flux.jl and CUDA.jl. When running one of my experiments, after several million steps `NaN`s pop up and propagate everywhere. I tracked down the i…
-
Example: https://mila.quebec/en/publications/
It would be nice to reuse the same code as in the Mila website. Not sure if that's 'easily' possible via RTD