moniker
moniker
PostsComments
IImmich
Created by moniker on 8/25/2023 in #help-desk-support
connect ECONNREFUSED 172.23.0.6:3003 - ML not starting
It seems like the machine_learning container is not starting. Here is my docker-compose block for that service (it previously worked before updating): 
  immich-machine-learning:
    image: ghcr.io/immich-app/immich-machine-learning:release
    volumes:
      - ${UPLOAD_LOCATION}:/usr/src/app/upload
      #- ./model-cache:/cache
      - ./model-cache/yolos-tiny:/usr/src/app/hustvl/yolos-tiny
      - ./model-cache/resnet-50:/usr/src/app/microsoft/resnet-50
      - ./model-cache/clip-ViT-B-32:/usr/src/app/clip-ViT-B-32
    env_file:
      - .env
    environment:
      - NODE_ENV=production
    depends_on:
      - database
    restart: always
  immich-machine-learning:
    image: ghcr.io/immich-app/immich-machine-learning:release
    volumes:
      - ${UPLOAD_LOCATION}:/usr/src/app/upload
      #- ./model-cache:/cache
      - ./model-cache/yolos-tiny:/usr/src/app/hustvl/yolos-tiny
      - ./model-cache/resnet-50:/usr/src/app/microsoft/resnet-50
      - ./model-cache/clip-ViT-B-32:/usr/src/app/clip-ViT-B-32
    env_file:
      - .env
    environment:
      - NODE_ENV=production
    depends_on:
      - database
    restart: always
Here are the logs from the container: 
docker logs immich-immich-machine-learning
INFO:     Started server process [7]
INFO:     Waiting for application startup.
docker logs immich-immich-machine-learning
INFO:     Started server process [7]
INFO:     Waiting for application startup.
Those logs aren't super helpful. Is there a way to get more verbose logs, and has anyone run into this? Previously I had an issue with the container not being allowed to access the internet and having to bind mount the models within it.
24 replies
IImmich
Created by moniker on 4/29/2023 in #help-desk-support
machine learning offline cache
Hi, I'm trying to run the machine learning container offline. There are errors related to files not being found. I'm assuming they would end up in the model_cache folder. I'e attempted to clone the models from huggingface so they are loaded from disk. It doesn't seem to work though. Here is my mode_cache folder (note that I moved the models into the models--.. folders as they were created on startup). 
-rw-r--r-- 1 moniker root    1 Apr 29 20:33 version.txt

models--microsoft--resnet-50:
total 300653
drwxr-xr-x 3 moniker root         9 Apr 29 21:00 .
drwxr-xr-x 3 moniker root         4 Apr 29 21:00 resnet-50
drwxr-xr-x 5 moniker root         6 Apr 29 20:49 ..
-rw-r--r-- 1 moniker root 102567489 Apr 29 20:40 pytorch_model.bin
-rw-r--r-- 1 moniker root 102450913 Apr 29 20:40 flax_model.msgpack
-rw-r--r-- 1 moniker root 102753944 Apr 29 20:40 tf_model.h5
-rw-r--r-- 1 moniker root       266 Apr 29 20:40 preprocessor_config.json
-rw-r--r-- 1 moniker root     69556 Apr 29 20:40 config.json
-rw-r--r-- 1 moniker root      2642 Apr 29 20:40 README.md

models--hustvl--yolos-tiny:
total 25432
drwxr-xr-x 3 moniker root           7 Apr 29 21:00 .
drwxrwxr-x 3 moniker moniker        4 Apr 29 21:00 yolos-tiny
drwxr-xr-x 5 moniker root           6 Apr 29 20:49 ..
-rw-rw-r-- 1 moniker moniker 26021147 Apr 29 20:41 pytorch_model.bin
-rw-rw-r-- 1 moniker moniker     4133 Apr 29 20:41 config.json
-rw-rw-r-- 1 moniker moniker      275 Apr 29 20:41 preprocessor_config.json
-rw-rw-r-- 1 moniker moniker     4159 Apr 29 20:41 README.md

.:
total 43
drwxr-xr-x 3 moniker root    9 Apr 29 21:00 models--microsoft--resnet-50
drwxr-xr-x 3 moniker root    7 Apr 29 21:00 models--hustvl--yolos-tiny
drwxr-xr-x 5 moniker root    6 Apr 29 20:49 .
drwxrwxr-x 4 moniker moniker 8 Apr 29 20:41 clip-ViT-B-32
-rw-r--r-- 1 moniker root    1 Apr 29 20:33 version.txt
drwxr-xr-x 7 moniker root    9 Apr 29 20:33 ..
-rw-r--r-- 1 moniker root    1 Apr 29 20:33 version.txt

models--microsoft--resnet-50:
total 300653
drwxr-xr-x 3 moniker root         9 Apr 29 21:00 .
drwxr-xr-x 3 moniker root         4 Apr 29 21:00 resnet-50
drwxr-xr-x 5 moniker root         6 Apr 29 20:49 ..
-rw-r--r-- 1 moniker root 102567489 Apr 29 20:40 pytorch_model.bin
-rw-r--r-- 1 moniker root 102450913 Apr 29 20:40 flax_model.msgpack
-rw-r--r-- 1 moniker root 102753944 Apr 29 20:40 tf_model.h5
-rw-r--r-- 1 moniker root       266 Apr 29 20:40 preprocessor_config.json
-rw-r--r-- 1 moniker root     69556 Apr 29 20:40 config.json
-rw-r--r-- 1 moniker root      2642 Apr 29 20:40 README.md

models--hustvl--yolos-tiny:
total 25432
drwxr-xr-x 3 moniker root           7 Apr 29 21:00 .
drwxrwxr-x 3 moniker moniker        4 Apr 29 21:00 yolos-tiny
drwxr-xr-x 5 moniker root           6 Apr 29 20:49 ..
-rw-rw-r-- 1 moniker moniker 26021147 Apr 29 20:41 pytorch_model.bin
-rw-rw-r-- 1 moniker moniker     4133 Apr 29 20:41 config.json
-rw-rw-r-- 1 moniker moniker      275 Apr 29 20:41 preprocessor_config.json
-rw-rw-r-- 1 moniker moniker     4159 Apr 29 20:41 README.md

.:
total 43
drwxr-xr-x 3 moniker root    9 Apr 29 21:00 models--microsoft--resnet-50
drwxr-xr-x 3 moniker root    7 Apr 29 21:00 models--hustvl--yolos-tiny
drwxr-xr-x 5 moniker root    6 Apr 29 20:49 .
drwxrwxr-x 4 moniker moniker 8 Apr 29 20:41 clip-ViT-B-32
-rw-r--r-- 1 moniker root    1 Apr 29 20:33 version.txt
drwxr-xr-x 7 moniker root    9 Apr 29 20:33 ..
10 replies
IImmich
Created by moniker on 12/11/2022 in #help-desk-support
Can't connect to api endpoint after docker-compose setup
I've attempted to set up Immich using the docker-compose setup doc (https://immich.app/docs/installation/recommended-installation, https://immich.app/docs/usage/post-installation). I am unable to log in on the mobile app but am able to use the app from a web browser on desktop. The only changes I've made to my .env are setting the upload, jwt_secret, and setting the log_level to verbose in an attempt to get more information on anything that I'm seeing: 
UPLOAD_LOCATION=./upload

###################################################################################
# Log message level - [simple|verbose]
###################################################################################

LOG_LEVEL=verbose

###################################################################################
# JWT SECRET
#
# This JWT_SECRET is used to sign the authentication keys for user login
# You should set it to a long randomly generated value
# You can use this command to generate one: openssl rand -base64 128
###################################################################################

JWT_SECRET=redacted
UPLOAD_LOCATION=./upload

###################################################################################
# Log message level - [simple|verbose]
###################################################################################

LOG_LEVEL=verbose

###################################################################################
# JWT SECRET
#
# This JWT_SECRET is used to sign the authentication keys for user login
# You should set it to a long randomly generated value
# You can use this command to generate one: openssl rand -base64 128
###################################################################################

JWT_SECRET=redacted
When I attempt to curl the api endpoint from localhost I get the following: 
$ curl http://localhost:2283/api
{"statusCode":404,"message":"Cannot GET /api","error":"Not Found"}
$ curl http://localhost:2283/api
{"statusCode":404,"message":"Cannot GET /api","error":"Not Found"}
I've attempted to watch this live while using docker-compose up instead of with -d, but there doesn't seem to be any errors logging, even with LOG_LEVEL set to verbose. Does anyone have some hints at what to do here to debug this (probably user error)? Of particular interest would be if there are any other curl requests that reflect mine being wrong that I could use to debug.
60 replies