localGPT icon indicating copy to clipboard operation
localGPT copied to clipboard

Installation with Docker image and Windows 11 is very slow

Open paulusTecnion opened this issue 1 year ago • 5 comments

image

I'm already busy for almost 24 hours! Why is this taking so long? Anything I can do about it?

paulusTecnion avatar Oct 26 '24 14:10 paulusTecnion

Now it stopped with this message:

62384.4 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 62.5/62.5 KB 2.1 MB/s eta 0:00:00 68492.9 Downloading requests-2.28.2-py3-none-any.whl (62 kB) 68492.9 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 62.8/62.8 KB 4.7 MB/s eta 0:00:00 69246.9 ERROR: Exception: 69246.9 Traceback (most recent call last): 69246.9 File "/usr/lib/python3/dist-packages/pip/_internal/cli/base_command.py", line 165, in exc_logging_wrapper 69246.9 status = run_func(*args) 69246.9 File "/usr/lib/python3/dist-packages/pip/_internal/cli/req_command.py", line 205, in wrapper 69246.9 return func(self, options, args) 69246.9 File "/usr/lib/python3/dist-packages/pip/_internal/commands/install.py", line 339, in run 69246.9 requirement_set = resolver.resolve( 69246.9 File "/usr/lib/python3/dist-packages/pip/_internal/resolution/resolvelib/resolver.py", line 94, in resolve 69246.9 result = self._result = resolver.resolve( 69246.9 File "/usr/lib/python3/dist-packages/pip/_vendor/resolvelib/resolvers.py", line 481, in resolve 69246.9 state = resolution.resolve(requirements, max_rounds=max_rounds) 69246.9 File "/usr/lib/python3/dist-packages/pip/_vendor/resolvelib/resolvers.py", line 392, in resolve 69246.9 raise ResolutionTooDeep(max_rounds) 69246.9 pip._vendor.resolvelib.resolvers.ResolutionTooDeep: 2000000

paulusTecnion avatar Oct 26 '24 15:10 paulusTecnion

Same issue

LoneDev6 avatar Oct 28 '24 14:10 LoneDev6

@paulusTecnion @LoneDev6 The problem is that it's looking for version compatibility. It looks like to me like that is the main time consumer. slow I'm still not done with my first install, so I'm not sure. But one possible simple solution that I'm thinking about is once you have installed it, assuming that you wanna install again in a similar environment would be to generate a new, more specific requirements.txt pip freeze > requirements.txt

ErnestMolner avatar Oct 29 '24 09:10 ErnestMolner

I was not on docker but I also face the same problem of a very slow package installation, when I ran pip install -r requirements.txt

TinoCoda avatar Nov 11 '24 16:11 TinoCoda

This is unusable in its current state. Any installation over and hour is no good. I'd be embarrased to release something like that, it would have been better not to bother and save wasting everyones time.

DrJaymz avatar Nov 29 '24 09:11 DrJaymz