Hi all, I use .get_local_copy() to get a local copy for each of my artifacts logged in a task. I currently have 160 files which I want to get a local copy. Each artifact is a numpy array (.npz file) uploaded using .upload_artifact() before. When I run .get_local_copy() for each artifact, it seems they are downloaded to my cache/storage_manager/... according to printed log information. But after the downloading when I check the storage place, only the latest 96 of these files appear, first subset got lost. Is there any parameter that adjusts the "number of files that can be stored in the cache"? I am using clearml python version 1.0.3 to upload artifacts and get the artifacts back from a task.
Hi ScatteredClams84
Is there any parameter that adjusts the "number of files that can be stored in the cache"? I am using clearml python version 1.0.3 to upload artifacts and get the artifacts back from a task. (edited)
Yes you are correct, the default value is 100 entries.
You can configure it in the clearml.conf, just add:sdk.storage.cache.default_cache_manager_size = 1000or from code:from clearml.storage.cache import CacheManager CacheManager.get_cache_manager(cache_file_limit=1000)
