test: add llama_3.2_1B model and fix for test lora script issue#4139

Merged

LarryXFly merged 2 commits intoNVIDIA:mainfrom

ruodil:user/ruodil/test_perf

May 12, 2025

Commits on May 12, 2025

test: add llama_v3.1_8b_fp8 model, llama_v3.1_405b model and llama_nemotron_49b model in perf test, and modify original llama models dtype from float16 to bfloat16 according to README.md
ruodil
committed
add llama_3.2_1B model and fix for lora script issue
ruodil
committed