python ingest.py gets stuck on "load INSTRUCTOR_Transformer"

[mindwellsolutions]

After running pip install -r requirements.txt and putting a new pdf into the source_documents folder. I ran python ingest.py . The first time it installed additional reequipments but got stuck at "load INSTRUCTOR_Transformer"

Here is a full screen shot.

[PromtEngineer]

How big is the document? If you are running the CPU version, then it will take a while.

[mindwellsolutions]

Thanks for the response. Really appreciate your YouTube channel. The PDF is only 200kb, didn't realize it would still take that long on lower end Nvidia Quadro M2000 GPU.

I ran the regular prompt without "-device_type cpu" so it likely was using my GPU, which is much lower end than on my gaming PC. I'll try it on my RTX 3080 machine.

On Mon, May 29, 2023, 11:04 PM PromptEngineer @.***> wrote:

How big is the document? If you are running the CPU version, then it will take a while.

— Reply to this email directly, view it on GitHub https://github.com/PromtEngineer/localGPT/issues/17#issuecomment-1567698378, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHKCXZB6VTN3MZG7TIYJBDXIVPVXANCNFSM6AAAAAAYTJJTVM . You are receiving this because you authored the thread.Message ID: @.***>

[mindwellsolutions]

Thank you, your solution worked. Once I ran it on a machine with an RTX 3080 it got past that step finally. However now I get a "Killed" error.

load INSTRUCTOR_Transformer max_seq_length 512 Using embedded DuckDB with persistence: data will be stored in: /mnt/6903a017-f604-4f90-9652-324e10b3e675/work/h2oai/localgpt/localGPT Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s] Killed