search

bytedance / ByteMLPerf

AI Accelerator Benchmark focuses on evaluating AI Accelerators from a practical production perspective, including the ease of use and versatility of software and hardware.
https://bytemlperf.ai/
Apache License 2.0
188 stars 50 forks source link

Fix several bugs: #45

Closed MikuGhoul closed 6 months ago

MikuGhoul commented 6 months ago
  1. rename llama-2 to llama2
  2. fix llama2 and chatglm2 new_token_ids with generate_ids
  3. fix setup pass pad_token_id to tokenizer
  4. fix config model_zoo config key from model to model_name
  5. fix reporter start_time continue incremental, reset at perf start