Sep 01 08:31:362023-09-01 06:31:36,514 INFO success: machine-learning entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
Sep 01 08:31:39Traceback (most recent call last):
Sep 01 08:31:39File "<frozen runpy>", line 198, in _run_module_as_main
Sep 01 08:31:39File "<frozen runpy>", line 88, in _run_code
Sep 01 08:31:39File "/app/code/machine-learning/app/main.py", line 13, in <module>
Sep 01 08:31:39from app.models.base import InferenceModel
Sep 01 08:31:39File "/app/code/machine-learning/app/models/__init__.py", line 1, in <module>
Sep 01 08:31:39from .clip import CLIPEncoder
Sep 01 08:31:39File "/app/code/machine-learning/app/models/clip.py", line 8, in <module>
Sep 01 08:31:39from clip_server.model.clip import BICUBIC, _convert_image_to_rgb
Sep 01 08:31:39ModuleNotFoundError: No module named 'clip_server'
Sep 01 08:31:40172.18.0.1 - - [01/Sep/2023:06:31:40 +0000] "GET / HTTP/1.1" 302 5 "-" "Mozilla (CloudronHealth)"
Sep 01 08:31:402023-09-01 06:31:40,088 INFO exited: machine-learning (exit status 1; not expected)
Sep 01 08:31:412023-09-01 06:31:41,090 INFO spawned: 'machine-learning' with pid 77776
Sep 01 08:31:422023-09-01 06:31:42,092 INFO success: machine-learning entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
Sep 01 08:31:42I20230901 06:31:42.149549 224 batched_indexer.cpp:279] Running GC for aborted requests, req map size: 0
Sep 01 08:31:43I20230901 06:31:43.207621 223 raft_server.cpp:545] Term: 59, last_index index: 4264, committed_index: 4264, known_applied_index: 4264, applying_index: 0, queued_writes: 0, pending_queue_size: 0, local_sequence: 16159
Sep 01 08:31:43I20230901 06:31:43.207686 265 raft_server.h:60] Peer refresh succeeded!
Sep 01 08:31:45Traceback (most recent call last):
Sep 01 08:31:45File "<frozen runpy>", line 198, in _run_module_as_main
Sep 01 08:31:45File "<frozen runpy>", line 88, in _run_code
Sep 01 08:31:45File "/app/code/machine-learning/app/main.py", line 13, in <module>
Sep 01 08:31:45from app.models.base import InferenceModel
Sep 01 08:31:45File "/app/code/machine-learning/app/models/__init__.py", line 1, in <module>
Sep 01 08:31:45from .clip import CLIPEncoder
Sep 01 08:31:45File "/app/code/machine-learning/app/models/clip.py", line 8, in <module>
Sep 01 08:31:45from clip_server.model.clip import BICUBIC, _convert_image_to_rgb
Sep 01 08:31:45ModuleNotFoundError: No module named 'clip_server'
Sep 01 08:31:452023-09-01 06:31:45,357 INFO exited: machine-learning (exit status 1; not expected)
Sep 01 08:31:462023-09-01 06:31:46,360 INFO spawned: 'machine-learning' with pid 77780
Sep 01 08:31:472023-09-01 06:31:47,361 INFO success: machine-learning entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
Sep 01 08:31:50172.18.0.1 - - [01/Sep/2023:06:31:50 +0000] "GET / HTTP/1.1" 302 5 "-" "Mozilla (CloudronHealth)"
Sep 01 08:31:50Traceback (most recent call last):

I've restored each update to see when it started and it seems that it got introduced with v1.37.0, v1.36.0 is working fine, after that the restarts begin.

I also installed a fresh one to check if it's happening there as well and does, so something needs fixing

Edit: Corresponding update is this one https://github.com/immich-app/immich/releases/tag/v1.75.0