-
### Problem description
Non-offset GHG reductions, such as those resulting from corporate energy conservation or efficiency initiatives, are currently not supported in Guardian. These are generally me…
-
启动脚本:
output_model=ppo_lora_output
if [ ! -d ${output_model} ];then
mkdir ${output_model}
fi
export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7
accelerate launch src/train_bash.py \
--stage …
-
### Proposal
While the current lobby has no issues, it is lacking two things. Engagement, Life, and Space. So why not redo it? We could redo the lobby to include a lot more open space, environment, l…
-
I will maintain in this OP a summary of current proposed syntax as I understand it to be. Note this is not authoritative, subject to change, and it may be inaccurate. Please make comments to discuss.
…
-
Hi Costa,
Thanks for sharing the awesome implementations! It is tremendously helpful for my own work.
I noticed a few uses of the function [left_padding_to_right_padding](https://github.com/vwx…
-
I am training alpaca-7B on 4 * A100 80G
I am using the provided Deepspeed-zero2 yaml file as the configuration file in the repository and running it. Even when I set the device-map of the model to No…
-
When I played around with the navigation domain (I appended the `domain.rddl` and `instance.rddl`) and displayed the DBN of single states and ground fluents, I could not interpret the results correctl…
-
- Feature Name: Central Management of SVD to CSP process
- Start Date: 2018-05-xx
- RFC PR:
- Rust Issue:
# Summary
[summary]: #summary
I need to make this a paragraph, but my main points a…
-
### Describe the bug
Please provide a clear and concise description of what the bug is. If applicable, add screenshots to help explain your problem, especially for visualization related problems.
…
-
Have reviewed the wandb training curves provided, and I have a question: why do prob_eval(train)/chosen and rewards_eval(train)/chosen gradually decrease? I originally thought that these two metrics w…