Giters

intel / intel-npu-acceleration-library

Intel® NPU Acceleration Library

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

[Bug] Tiny Llama sample not works properly

invent00 opened this issue · comments

commented

Describe the bug
tiny llama sample code are not working on my MTL system.
MatMul sample code works properly.

To Reproduce
Steps to reproduce the behavior:

  1. install intel-npu-acceleration-library on venv
   pip install intel-npu-acceleration-library
  1. execute tiny-llama sample code
    image

Expected behavior
execute tiny-llama on NPU and show result

Error Screenshots

image

environment:

  • OS: Windows 11 23H2 22631.3447
  • Python version: 3.9.13 and 3.11.8
  • CPU: Ultra 5 125U
  • NPU driver: 32.0.100.2267
commented

Hi @invent00 , what transformers version do you have?

commented

Can you try this PR: #11 I can replicate your issue with a newer version fo the transformers library

commented

Should be fixed with #11. Closing this issue. Please let me know if you still have issues

commented

Thank you for your support.
I will try new version.

transformers version:
transformers==4.39.3