ampere_model_library icon indicating copy to clipboard operation
ampere_model_library copied to clipboard

Published 20 hours ago • verified publisher icon AmpereComputingAI

add gpt2 model

Open MarcelWilnicki opened this issue 1 year ago • 1 comments

TORCH_COMPILE=1 AIO_NUM_THREADS=80 numactl --membind=0 --cpubind=0 python natural_language_processing/text_generation/gpt/run.py -m gpt2 --lambada_path /ampere/aml/lambada_test_plain_text.txt --num_runs 10

MarcelWilnicki avatar Dec 18 '23 19:12 MarcelWilnicki

AIO is not being used

For the AIO to work it needs to be the latest AIO built from main. I believe the most recent release does not include these fixes yet.

MarcelWilnicki avatar Jan 29 '24 15:01 MarcelWilnicki