[KNOWN BUG] AOTI Inference via cpp Runner bug

[Jack-Khuu]

🐛 Describe the bug

As of #1426, torchchat/main is failing 2 jobs related to AOTI inference via c++ runners.

Versions

fd1857a

[mikekgfb]

Seems to be a problem in sentencepiece

torchchat/tokenizer/sentencepiece.cpp

Lines 39 to 43 in cc0ffce

	const auto status = _processor->Load(tokenizer_path);
	if (!status.ok()) {
	fprintf(stderr, "couldn't load %s\n. If this tokenizer artifact is for llama3, please pass `-l 3`.", tokenizer_path.c_str());
	exit(EXIT_FAILURE);
	}