On November 8, according to media reports, Amazon is spending millions of dollars to train its own large language model (LLM), internally codenamed "Olympus," which could be launched as early as the end of December.。