google-deepmind / google-deepmind/acme

Loading checkpoint does not work

Open
#119 7 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
4.1k
Forks
553
PR merge metrics
No merged PRs in 30d

Description

The add_uid variable is not passed to the agent and learner, and is set to default to true in the Snapshotter and Checkpointer.

Therefore when passing to the agent the checkpoint path say:
```python
checkpoint_subpath = "/log/fed68b00-b8cb-11eb-812c-bba6f8f69ddd"
```
a new folder will be created inside it with a new identifier and the previous checkpoints are not loaded.

To reproduce run test_dqn.py twice, the second time set checkpoint_subpath pointing to the previous log path and pass it to the agent

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.