Weixin-Liang / LLM-scientific-feedback

Can large language models provide useful feedback on research papers? A large-scale empirical analysis.
Creative Commons Attribution 4.0 International
489 stars 49 forks source link

Conda installation does not work. #1

Closed abhiksingla closed 10 months ago

abhiksingla commented 11 months ago

Getting the following error.

Solving environment: failed

ResolvePackageNotFound: 
  - libffi==3.4.2=h7f98852_5
  - python==3.7.12=hf930737_100_cpython
  - libstdcxx-ng==12.2.0=h46fd767_19
  - libgcc-ng==12.2.0=h65d4601_19
  - libsqlite==3.40.0=h753d276_0
  - _openmp_mutex==4.5=2_gnu
  - ca-certificates==2022.12.7=ha878542_0
  - readline==8.2=h8228510_1
  - ncurses==6.3=h27087fc_1
  - xz==5.2.6=h166bdaf_0
  - libzlib==1.2.13=h166bdaf_4
  - openssl==3.1.0=h0b41bf4_0
  - tk==8.6.12=h27826a3_0
  - libnsl==2.0.0=h7f98852_0
  - _libgcc_mutex==0.1=conda_forge
  - libgomp==12.2.0=h65d4601_19
  - sqlite==3.40.0=h4ff8645_0
  - ld_impl_linux-64==2.40=h41732ed_0

Even after removing =h*** suffix from the packages, some of the packages are not found.

hiyin commented 11 months ago

I am a user who was able to implement this successfully. You can't actually install this in Mac (which I believe you did). You need somewhat Linux machine like CentOS or other distributions to be able to make this work.

Faultiness commented 11 months ago

@hiyin Thanks for your suggestion. I finally managed to run the whole pipeline on Ubuntu 22.04 in x64 Win10 VMware.

Any suggestions to run the entire pipeline natively on Mac or Windows? Just for reference:

Boltzmachine commented 11 months ago

docker on Mac also failed

Faultiness commented 11 months ago

ScienceBeam Parser only officially supports Linux due to the binaries used. Compiling is necessary for running on the Win or Mac platform. Related Link

yuhui-zh15 commented 10 months ago

Thanks for your interest in our system. Unfortunately ScienceBeam only supports Linux. Let us know if you find a way to install ScienceBeam on other systems.

yuhui-zh15 commented 10 months ago

We now added a new system that supports texts as input:

Usage:

python main_from_text.py  # If you are using non-linux operating systems or want to generate feedback from the parsed paper in text format