Diego Fiori comments

Results 58 comments of


                                            Diego Fiori

Improve Documentation and Sample Code

Hi @ylassoued, examples of the config file and of dataset generation have been added to the new readme with #203 and #204. Feel free to open another issue if you...

Question: How to load the ChatLLaMA model weights from the directory for training? Also, can you provide a sample of the training data?

Hi @Alla-Abdella, thank you for contacting us! We are currently working on refining our code and we will release some training data examples very soon.

ModuleNotFoundError: No module named 'config'

Hi @MohamedAliRashad and @ming1523, thanks for pointing out the error. We are currently fixing multiple bugs, you can see the updates on #203.

ModuleNotFoundError: No module named 'config'

Closing the issue since the problem has been solved in #203.

generate_dataset.py Error

Fixed in #212. It should work now pulling the latest commit.

path_to_config_file.yaml

Closing the issue since examples have been provided in the new readme #204.

Examples for dataset/sections_dataset.json

Hi @TonyZhanghm, thanks for asking 😄 We actually worked on it and we will merge a PR with the code for downloading (and formatting) the Anthropic and SHP datasets. We...

Examples for dataset/sections_dataset.json

Closing the issue since it has been fixed in #203.

what is the training data format in chatllama actor_config/reward_config.

Hi @lonelydancer, thank you for reaching out. Today, we are gonna release a new readme with more extensive examples. Please let us know if you have any other feedback 😃

what is the training data format in chatllama actor_config/reward_config.

Data examples for reward training has been released with #203. For updates on distributed training see #242 and #229