-
```
assert not args.model_parallel.fp16, \
"Expert parallelism is not supported with fp16 training."
```
from https://github.com/NVIDIA/Megatron-LM/blob/db3a3f79d1cda60ea4b3db0ceffcf…
-
- Give expertise to codellama-34b-instruct based on efficient and logical chunking of mermaid mkdocs.
- Make it generic
-
I want to experiment with a reliable way to use a simple mix of experts, of a sort. I thought a pipeline starting with a classification. Some way put the request into a definite category. 5-10 Categor…
-
In the Muse_Software_Manual.pdf, Where is the muse-3.0.0-source.tar.gz in the configuration?
```[tasklist]
### Tasks
```
-
Hello,
Thank you for the nice work with this training framework. However, I have noticed that there's a problem with inference, conversion and fine-tuning of MoE based GPT model. The following is a…
-
### systemRole
I'm creating a Vim expert AI assistant. I want you to assume the role of a seasoned Vim user with extensive knowledge of Vim's features, commands, and plugins. Your goal is to provid…
-
### Discussed in https://github.com/ucd-library/aggie-experts/discussions/87
Originally posted by **qjhart** March 22, 2023
Elements comes with the option of including the `availability` attri…
-
I am trying to generate my own data and tried running the following command `python dagger_training.py --settings_file=config/dagger_settings.yaml` as suggested in the README.
If I understand corr…
-
Comments from testing:
_Just a UX note -- I'm not sure the eye icon alone is enough for indicating the user is hidden and would benefit from additional visual cues. I think I would add some sort of…
-