Comments (12)
Just to confirm, you are getting this error when trying to use the youtube transcript data connector, correct?
If so, this is most likely something wrong with the youtube link you are providing. If you share the link here with me I can take a look at it for you.
from anything-llm.
This video for example: https://youtu.be/XwL_cRuXM2E?si=UdD6Atjk5HBRjMEg
Also when I want to use the agent in the chat it says that the agent session starter and then after a while thet it ended but I get no response. When I ask the same question again the same thing happens and when I refresh the site the output is there but only from the first question.
from anything-llm.
You are using a shortened version of the youtube video link. It needs to be in this format: https://www.youtube.com/watch?v=XwL_cRuXM2E&ab_channel=bycloud. You can get that link by just visiting the shortened link in your browser and it will expand to the full link.
Also, try using another LLM provider as your agent provider to see if it fixes the issue. It is highly likely the agent is getting stuck because of the quality of model you are using.
from anything-llm.
That doesn't fix it I get the same error with the long URL.
from anything-llm.
@shatfield4 The Youtube link collector works with both types of YT URLS
anything-llm/collector/utils/extensions/YoutubeTranscript/index.js
Lines 16 to 20 in 9a38b32
If @shatfield4 can scrape the URL and @TheSirTigerr cannot then something else is going on. Additionally, is the docker container running with --cap-add SYS_ADMIN
? Most docker container hosting services (if applicable) do not allow this permission due to bot-scrapers
from anything-llm.
I have a Ubuntu VM on digitalocean. How can I check if I have --cap-add SYS_ADMIN
on?
from anything-llm.
When you started the docker container in the ubuntu VM you would have to run some kind of docker command like docker run -p 3001:3001 mintplexlabs/anythingllm
or something along those lines. The command needs to match what we have in our docs to get the full feature set due to some docker engine defaults.
from anything-llm.
I'm seeing a similar issue on Windows 10. (No docker containers.)
I can get transcripts for some channels, such as https://www.youtube.com/watch?v=O5GY7_aVBtk
But other channels fail, such as https://www.youtube.com/watch?v=V9KJ7nvhRWk
This results in a zero-byte file in the AppData\Roaming\anythingllm-desktop\storage\documents[channelname] folder.
I did confirm that Youtube shows a transcript for that video. This seems to be across all videos in a channel.
from anything-llm.
So the transcript is fixed I used --cap-add SYS_ADMIN
that fixed it. But web searching/agent still doesn't work in chat see picture.
from anything-llm.
Do you know if the docker container is using a proxy or anything to reach your container? Some providers will do this and it makes using websockets (which is how agents work) unusable until worked around. I think that may be what is happening here?
You can also check to see if in the frontend network requests if the websocket connection is attempting to reach ws
or wss
when starting an agent chat.
from anything-llm.
It sends the wss
i use Nginx Proxy Manager
from anything-llm.
When you have the Nginx proxy running - do you have an allowance for WSS to connect?
# Default server configuration
# Example config for regular setup + SSL + Websockets.
server {
listen 80;
server_name mysite.com;
return 301 https://mysite.com$request_uri;
}
server {
listen 443 ssl;
ssl on;
server_name mysite.useanything.com;
ssl_certificate /etc/letsencrypt/live/mysite.com/fullchain.pem;
ssl_certificate_key /etc/letsencrypt/live/mysite.com/privkey.pem;
# Enable websocket connections for agent protocol.
location ~* ^/api/agent-invocation/(.*) {
proxy_pass http://localhost:3001;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "Upgrade";
}
location / {
proxy_connect_timeout 605;
proxy_send_timeout 605;
proxy_read_timeout 605;
send_timeout 605;
keepalive_timeout 605;
proxy_buffering off;
proxy_cache off;
proxy_pass http://ip-of-server:3001$request_uri;
}
}
from anything-llm.
Related Issues (20)
- [BUG]: LM Studio does not allow for different model selection HOT 2
- Does running @agWhy does the agent have to configure the model separately, and will it eventually hand it over to a system-configured model?ent AnythingLLM call only the model in the ai agent session and not the model configured by the system? HOT 1
- [BUG]: AnythingLLM Desktop not loading in my Windows 11 (Its been 5 to 7 hours) HOT 4
- [CHORE]: Increase Ollama timeout HOT 3
- [BUG]: Error after embedding document ""Could not respond to message" HOT 11
- [FEAT]: zscaler block connection to internet please add a configuration for certificate HOT 2
- [BUG]: When querying Ollama through AnythingLLM, the call is resetting the time the model is in Memory from unlimited, to 5 minutes HOT 1
- [BUG]: Can't load local documents or fetch website HOT 3
- General Feedback for AnythingLLM HOT 1
- [DOCS]: "AnythingLLM Debug mode on MacOS" instructions HOT 1
- [BUG]: uninstall issue v 1.4..4 HOT 3
- [BUG]: Youtube transcript retrieves an empty file o some channels HOT 1
- [FEAT]: Add ability to retrieve documents from paperless-ngx
- [BUG]: Unable to get "Live web search and browsing" agent working using google custom search engine. Error: getaddrinfo ENOTFOUND - http errno: -3008 HOT 6
- [FEAT]: Add support for TSX files as text files. HOT 1
- [BUG]: OpenAI::CreateModeration failed with: Request failed with status code 403 HOT 1
- [BUG]: UI bug fixes
- [FEAT]: Add speech-to-text prompting possibility activated with a key combination
- [BUG]: Failed to vectorize file HOT 1
- Is there a limit to the content length of the application's ability to crawl the site? Is the crawling content accurate and comprehensive? Because I need to extract the information from the text accurately and that's very important HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from anything-llm.