The response_agent in the “email assistant with semantic memory” is using the “anthropic:claude-3-5-sonnet-latest” model. Since I don’t have an anthropic api key, I’m using gpt-4o. But I’m not seeing that the memory tool is being used after sending an email to Alice. Is this a gpt-4o issue or is something else going on (I downloaded the code directly from the course).
I am having the same issue. I suspect there is an issue with the OpenAI GPT-4O Model. I will try using some open source models using Ollama and will update here. I suspect the creator also faced the same issue and switched to Anthropic’s model