search

aida-ugent / SkillGPT

Other
56 stars 11 forks source link

Cannot init redis esco embeddings #1

Open Bachdz opened 1 year ago

Bachdz commented 1 year ago

When trying to initialize Redis, this error occurred: FileNotFoundError: [Errno 2] No such file or directory: './esco_embeddings/df_skillGroups_en.parquet'.

This means that the ESCO entries (.parquet file) are missing. Can you provide the converted entries for the initialization of the Redis database?

Further, the worker_addr in the Initialize ESCO embeddings section is incorrect. The variable name should be API_HOST and API_PORT instead of REDIS_HOST and REDIS_PORT.

Astriel commented 6 months ago

Hello, Indeed I'm facing the same problem. Did you find a solution or received the files ? I've got the whole ESCO normalization extracted from the ESCO website, but can't find parquet files.

Bachdz commented 6 months ago

The parquet files are not included in the repository. You can embed the ESCO data in vectors to create the parquet files, or you can contact the authors directly and ask for permission for the data. This is what I have done.

Astriel commented 6 months ago

I'm going to contact the author to be able to try it out. Not sure I would be able to embed the ESCO data in vectors myself ! Thank you for the tip.