StayThePath Posted January 11 Share Posted January 11 (edited) Hi there, you can read the whole write up on the GitHub or DockerHub, but to try to keep it short it's an OpenAI based chatbot that allows you to click any movie title mentioned to download it to Radarr. Here is a short video. I find it really useful and fun to talk with, but there is a good chance that's just because it's something I built. I'd love to hear how this could be more useful to you. Anyway, this is for guide, so here is the relevant information on how to add this on Unraid (and more). Please note: OpenAI charges for API usage. It's recommended to monitor your usage closely, especially when first using this application, to understand potential costs. That said, using GPT-3.5 makes it relatively cheap, it seems more than sufficient to me, and it's much faster. I haven't really seen a lot of stuff GPT-4 does that is much better than GPT-3.5, but you can always give it a shot. _______________________________________________________________________________________________ Installation Unraid I don't have this in the Community Applications yet, but I hope to get it there at some point. However, you can copy the chatarr.xml from the root of the repo directory to: /boot/config/plugins/dockerMan/templates-user/ Then you can simply go to your docker tab and you should have a Chatarr template you can choose in the template dropdown at the very top and you should be able to just add it and be good to go. If you want to just paste the XML straight into that folder, here it is: <?xml version="1.0"?> <Container version="2"> <Name>chatarr</Name> <Repository>staythepath/chatarr:latest</Repository> <Registry>https://hub.docker.com/r/staythepath/chatarr</Registry> <Network>bridge</Network> <MyIP/> <Shell>sh</Shell> <Privileged>false</Privileged> <Support>https://github.com/staythepath/chatarr/issues</Support> <Project/> <Overview>An OpenAI based chatbot with radarr integration. You can talk to the bot about movies and any time it mentions a movie you can click the movie title to add the movie to radarr. If you hover over the title a popover comes up with details about the movie. If you click a persons name a popover comes up with details about the person. The discord bot is sort of functional, but you have to restart the container to get it to stop and you have to save the configuration with the box checked to get it to start. </Overview> <Category>MediaApp:Other Other:</Category> <WebUI>http://[IP]:[PORT:1138]/</WebUI> <TemplateURL>https://github.com/staythepath/chatarr/template.xml</TemplateURL> <Icon>https://github.com/staythepath/chatarr/blob/main/resources/icon128.png?raw=true</Icon> <ExtraParams/> <PostArgs/> <CPUset/> <DateInstalled>0</DateInstalled> <DonateText/> <DonateLink/> <Requires/> <Config Name="WebUI" Target="1138" Default="1138" Mode="tcp" Description="WebUI port." Type="Port" Display="always" Required="true" Mask="false">1138</Config> </Container> _______________________________________________________________________________________________ Docker It's also easy to just use docker: docker pull staythepath/chatarr:latest Pick a host port and run it. The container port needs to be 1138. I think the container port is done automatically, but I'm not sure because I'm dumb. _______________________________________________________________________________________________ Linux Alternatively you can clone the repo: git clone https://github.com/staythepath/chatarr.git Then go to the folder: cd chatarr Install the requirements(I would start a new environment for this): pip3 install -r requirements.txt Then you should be able to start it with: gunicorn --bind 0.0.0.0:1138 --timeout 60 app:app _______________________________________________________________________________________________ Setup OpenAI THIS COSTS MONEY!!!! IF YOU ARE USING IT A LOT ON GPT-4 IT MIGHT ADD UP!!!! CHECK YOUR USAGE TO KNOW WHAT YOU ARE SPENDING WHEN USING THIS!!! However using gpt-3.5 seems perfectly sufficient to me and it's really cheap. I spent well under 5 dollars over a week of hammering the GPT-3.5 API constantly while I was developing this. Under regular use(whatever that is) it shouldn't cost too much as long as you stick to GPT-3.5. Sign up with OpenAI and grab an API key. Radarr API Key, URL, Quality, Root folder If you're reading this, I'm sure you can figure this stuff out. You can find all this stuff is in Radarr itself. You do need to specify the exact quality you want. I have a quality profile called, "this" and I just put in "this" (no quotes) and it uses that profile when downloading. Making the quality input a dropdown is on the (long) list of feature I want to add. The root folder should auto populate, and I think most people won't need to change it from what it loads with, but you should definitely check to make sure you are using the right one. All this stuff is necessary to send the movie to Radarr. TMDb key Just go to https://www.themoviedb.org and sign up and grab an API key. This is where I grab all(well most) of the movie data from. There is a cache I'm working on to prevent having to make API calls for every popover here, but if you are even kind of into movies, you will run across stuff not in the cache. It's a work in progress. Model You're just selecting which OpenAI model to use. As I've explained, GPT-3.5 feels sufficient to me in most cases and it's also faster, but GPT-4 definitely tends to give more in depth and accurate answers. If you feel the need to bump it to GPT-4, just watch your usage. ___________ So that's it. I'm lowkey dying for some feedback, so please even if Edited January 11 by StayThePath Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.