Test record

[samuraieng]

Try with several models before PR
Must run with both llama-cli and llama-server

LLM
[ ] - Qwen2
[ ] - Qwen2.5
[ ] - Qwen3
[ ] - DeepSeek
[ ] - llama
[ ] - gpt-oss
[ ] - sarashina2.,2
[ ] - LLM-jp

VLM
[ ] - Qwen 2
[ ] - Qwen 2.5
[ ] - Qwen 3
[ ] - DeepSeek
[ ] - llama
[ ] - sarashina2.,2

[samuraieng]

Skip this work. It seems it is not required to submit some logs.