FastChat icon indicating copy to clipboard operation
FastChat copied to clipboard

ConnectionError when launch the model worker(s)

Open sz2three opened this issue 2 years ago • 1 comments

when launch the model worker(s): python3 -m fastchat.serve.model_worker --model-name 'vicuna-7b-v1.1' --model-path /path/to/vicuna/weights it has error as follows: 2023-05-30 17:19:25 | ERROR | stderr | ConnectionError: HTTPConnectionPool(host='localhost', port=21001): Max retries exceeded with url: /register_worker 2023-05-30 17:19:25 | ERROR | stderr | (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fed5f8683a0>: Failed to establish a new 2023-05-30 17:19:25 | ERROR | stderr | connection: [Errno 111] Connection refused'))

sz2three avatar May 30 '23 08:05 sz2three

Same Issue here, any solution yet?

shad00m avatar May 31 '23 07:05 shad00m

solution?

David-Lee-1990 avatar Oct 27 '23 08:10 David-Lee-1990

same issue

npuichigo avatar Oct 27 '23 09:10 npuichigo

How do you run the controller?

surak avatar Oct 27 '23 11:10 surak

In my side, when the request traffic increases, I god 502 error and find the same error in the log of model worker

npuichigo avatar Oct 27 '23 12:10 npuichigo

I got exactly the same issue

ruifengma avatar Jan 03 '24 08:01 ruifengma

How do you run the controller?

for me, I just simply run python3 -m fastchat.server.controller --host 0.0.0.0 --port 30001

ruifengma avatar Jan 03 '24 08:01 ruifengma

Before this, you run

python3 -m fastchat.serve.controller --port 21001

to Launch a controller

Then, when run python3 -m fastchat.serve.model_worker --model-name 'vicuna-7b-v1.1' --model-path /path/to/vicuna/weights to load a model, it finds the controller by default at --port 21001

So, if your controller was not run at 21001, you should specify the fastchat.serve.model_worker where it is

qroam avatar Mar 20 '24 08:03 qroam