alondmnt / joplin-plugin-jarvis Goto Github PK

Thank you for building and sharing this software!

As a potential user of this plugin I feel nervous about having my entire second brain uploaded to a third party service. I like the idea of doing simple chats though. What I would like to see in the documentation is a separate, clear, standardised indication of which data is uploaded for each of the types of queries listed.

For example:

• Chat:
• Start a new note, or continue an existing conversation in a saved note. Place the cursor after your prompt and run the command Chat with Jarvis (from the toolbar or Tools/Jarvis menu). Each time you run the command Jarvis will append its response to the note at the current cursor position (given the previous content that both of you created). If you don't like the response, run the command again to replace it with a new one.
• Data shared: Only the current note you are in.

Not sure if the above 'Data shared' annotation is actually correct, which is why I think it should be clearer to the user. I'm happy to submit a PR implementing this if you can let me know the details of the data shared. Thank you!

Go you have any plans to add support for ChatGPT 4o mini? It's cheaper and better than ChatGPT 3.5

Can we have some feedback on what's happening after a request is sent? An indicator that we're awaiting a response, and an error message if there is one?

For instance, I might press Ctrl+Shift+C and the border of the textarea starts glowing for the duration of whatever timeout period we're waiting. It gets an error message so the textarea stops glowing and the message is displayed to the user. If there's no response, once the timeout is reached the textarea stops glowing and the error "Request timed out. Check your connection." is displayed.

When Jarvis is enabled, it keeps throwing this error in the Joplin dev tools:

models/Setting: Could not set setting on the keychain. Will be saved to database instead: plugin-joplin.plugin.alondmnt.jarvis.springer_api_key: Error: Password is required.

Jarvis is working fine with the API key for GPT - so, I don't know why it's throwing this error over and over

Joplin 2.14.22 (prod, win32)
Client ID: 19f20674be77445bbd5535190049d93c
Sync Version: 3
Profile Version: 46
Keychain Supported: Yes
Revision: e579eb9

it's OpenAI models' tokenizer, allowing a precise split
the pure js version is not small, its unpack size is 10M https://www.npmjs.com/package/js-tiktoken

https://github.com/logancyang/obsidian-copilot

Description:

This feature request proposes the addition of a dedicated chatbox interface within Joplin, enhancing the user experience for those utilizing AI-driven conversation features. Inspired by similar functionalities in other note-taking applications, this feature aims to provide a focused and interactive environment for users engaging with AI in Joplin.

Key Features of the Proposed Chatbox:

1. Dedicated Chat Interface: Implementing a separate chatbox window in the Joplin interface, dedicated to AI interactions. This design will allow for a more immersive and focused conversation experience.

2. Integration with Notes: Enabling the chatbox to use the currently open note as context, thereby improving the relevance and accuracy of AI responses within the chat.

3. One-Click Functionality:

   • Save Conversation: Allowing users to save the entire chat conversation into a new or existing note with a single click.
   • Copy to Note: Facilitating the copying of selected chat portions directly into a note with ease.

Rationale for the Request:

The addition of a dedicated chatbox is intended to streamline the AI interaction experience within Joplin, making it more accessible, efficient, and user-friendly. Such a feature would provide a distinct space for AI conversations, separate from the main note-editing area, thus minimizing distractions and improving overall focus. The capability to seamlessly save and incorporate chat contents into notes would greatly enhance the practicality of AI features for a variety of note-taking and research activities.

Anticipated Impact:

The implementation of this feature is expected to substantially increase the productivity and efficiency of users who rely on AI assistance for various purposes, including note-taking and brainstorming. By introducing this advanced AI integration feature, Joplin could attract new users and align more closely with the evolving needs and expectations of its user base.

Your consideration of this feature request is greatly appreciated. A dedicated chatbox would mark a significant enhancement in the functionality and user experience of AI integration within Joplin.

Hello, could you please add an option to customize the OpenAI proxy in the settings?

Hi,

I want to try connecting Jarvis using oobabooga/text-generation-webui as a backend instead of openai for private and offline solution. Is that possible? Thanks!

First thanks for the energy and effort that has been put into this plugin.

I found two additional local opensource options

I discovered https://github.com/janhq/jan when I was researching LM Studio. I came across it in a LM Studio post on reddit that expressed both praise of LMStudio and also concerns about its future as they go more and more commercial.

Someone from the jan.ai team also posted this https://twitter.com/janframework/status/1745472833579540722?t=osxIAvq8ztXuDbNAm11thA praising Ava

I just wanted to these on people's radar.

I am a little out of my element here, but I do hope to figure out a way to make this plugin work using the guide with my joplin.

Edited: corrected links

I have a lot of notes, most of which were imported from Evernote as HTML. When Jarvis starts to update the note database, it gets to a certain number of notes, usually 250, before not progressing any further. I haven't noticed any logs in DevTools from Jarvis, and the "Universal Sentence Encoder" file exists and has some data in it. And occasionally the related notes search comes up and I can actually use it, but not often.

I wonder if one of my notes is causing the issue, but since I didn't find logs, I wouldn't know which one.

https://lmstudio.ai

it can provide an open api style

I am no expert just found both your project and this tonight

here's code they suggest for python

# Example: reuse your existing OpenAI setup
import os
import openai

openai.api_base = "http://localhost:1234/v1" # point to the local server
openai.api_key = "" # no need for an API key

completion = openai.ChatCompletion.create(
  model="local-model", # this field is currently unused
  messages=[
    {"role": "system", "content": "Always answer in rhymes."},
    {"role": "user", "content": "Introduce yourself."}
  ]
)

print(completion.choices[0].message)

How to speak English?
erro:~~~
after I click ok

I create a new note, and select "Chat with your notes". I got then stuck :

Note :

• I verified that Universal Sentence Encoder.sqlite has been created
• My setup is quite small (<100 notes)
• Using Linux/Flatpak

It would be an awesome feature for Jarvis to use a local LLM via Ollama.

https://ai.google.dev/palm_docs/deprecation

PaLM API going away Aug 2024.

is it possible to include the gpt4-0314 models instead of the newest gpt4?

How about support for GPT-4?

Ref #9 where it seems like Azure OpenAI should work but maybe a bit untested.

I am very intrigued by your project, so I am trying to make it work with Azure. However, I am getting some error messages and I am not quite sure how to set it up. I am trying Jarvis: Chat for now. Are you able to give me some hints?

I've tried these two alternatives:

I have a small python script I've used to verify my Azure OpenAI endpoint ID, API key and model names.

Hello,

Would you mind adding support for Groq?

https://console.groq.com/docs/quickstart

Thank you

Thanks for the plugin, which looks very interesting! It would be useful if I could exclude an entire notebook by notebook - or conversely designate which notebooks I want included - and not only by individual note. I have a huge notebook that I wouldn't want to include. If I modify all thousands of notes in there to add the exclude tag, all the notes will show that new modification date.

Can you provide support for Azure OpenAI API? Thanks!

Is it possible to train the plugin on all the notes in the notebook so it has context?

https://medium.com/@rdcolema7/deploying-gpt-2-models-in-custom-applications-f8117e482837

split by min(max_token, tokens_of_X_paragraphs)

pro:

• splitting by paragraphs should result in more semantically self-contained embeddings than splitting by words
• thus better semantic search

cons:

• likely more storage needed
• (only once) longer init

// pseudo code anyway
embeddings = []
cur_split, cur_token = [], 0
for (p in paragraphs) {
  tokens_p = calc_tokens(p);
  if ((cur_token + tokens_p) >= max_tokens) {
    embeddings.push(calc_embeddings(cur_split));
    cur_split = [p];
    cur_token = tokens_p;
  } else {
    cur_split.push(p);
    cur_token += tokens_p;
  }
}
// and the embeddings of last split

OpenAI has a streaming API to allow for results to come token by token. It would be a nice addition to add this to both the chat and edit features of Jarvis.