-
Hi all,
today (30th Nov 2023), we have proceeded to rename the "master" branch of this repository to "main".
You can find more details about the migration @ https://tracker.moodle.org/browse/MDL…
-
Thank you for [your code](https://github.com/AscendNTNU/msp_flightcontroller_interface) and contribution in this area of MSP! Given the dearth of documentation resources, I can imagine that contributi…
-
It’s probably worth having separate issues for subtasks of the main issue. I’ll transfer info from the main issue here.
-
"Enables the command extensions... " For real? How about explaining what command extensions are or even a link (anchor) to what command extensions are. This is beyond absurd. Whoever wrote this mus…
-
Why i have this error when i try to run llama-7b on windows (CPU: i5-7300HQ @2.50GHZ , memory:24576MB RAM):
>torchrun --nproc_per_node 1 example_chat_completion.py --ckpt_dir llama-2-7b --tokenizer_p…
-
I have been trying to extract data (title, question answered, entities, summary) from documents chunks.
I believed typed predictors would be good for this, but I keep running into "Too many retrie…
-
Perform prompt engineering to ensure accuracy, minimize hallucinations, watch out for unnecessary jargon, adjust tone and level of depth to match user’s.
-
https://www.microsoft.com/en-us/research/uploads/prod/2021/06/ACL2021_PENS_Camera_Ready_1862_Paper.pdf
-
I feel that the current prompt/answer rating system can be too subjective and unclear at times, which may be affecting RLHF quality. So, from current experiences, here's my proposal on a modified rati…
-
你好,感谢分享IFT部分的代码,这边做了一些实验,有一些疑问。
1. 首先我只用了embedding fusion那块,然后发现gsm8k和truthfulqa的效果有提升,其他的基本差不多
2. 然后我加上了dynamic relation propagation,发现有些指标有提升,但是gsm8k和mmlu都不太好
3. 我发现论文中学习率是5e-7,而我之前设置的是2e-5,进行了调整…