null error, not reading active note (~890k words)

[silasdao]

The selected text contains 886689 words and 4417275 tokens.

I've edited the frontmatter; just removed identifiable details.

Using OpenAI GPT-4, and embeddings. It's paid.

I am still getting standard 'text assistant' type responses with no contextual relevance.

Beyond the frontmatter, the plugin config is temp 0.8, token limit 800, convo-context is 20.

I was getting the context error before with the langchain agent, but nothing now. Just the standard responses.

Can you add a vector database api option, I'd definitely use pinecone or better yet, weaviate :)

---

model: gpt-4
max_tokens: 4096
temperature: 0.80
top_p: 0.8
frequency_penalty: 0.0
presence_penalty: 0.0
stream: true
stop: null
system_commands: ['exit']
n: 1
role: Natural Language Data Analyst
tasks:
- task: Analyse Text Data
details: The analysis should focus on the user "" and should leverage the masterdoc file.
- task: Generate Output Documents
details: Produce four plain text documents using the analysed text data. These include a detailed event timeline for "" etc