tracel-ai/burn · error

Can load learning agent checkpoint.

Error message

Can load learning agent checkpoint.

What it means

Guard in `RLCheckpointer::load_checkpoint`: restoring the learning agent record for the requested epoch failed (missing or unreadable checkpoint), so loading panics rather than returning an agent without its state.

Source

Thrown at crates/burn-train/src/learner/rl/checkpointer.rs:71

        }
    }

    /// Load a training checkpoint.
    pub fn load_checkpoint(
        &self,
        learning_agent: RLC::LearningAgent,
        epoch: usize,
    ) -> RLC::LearningAgent {
        let record = self
            .policy
            .restore(epoch)
            .expect("Can load model checkpoint.");
        let policy = learning_agent.policy().load_record(record);

        let record = self
            .learning_agent
            .restore(epoch)
            .expect("Can load learning agent checkpoint.");
        let mut learning_agent = learning_agent.load_record(record);
        learning_agent.update_policy(policy);

        learning_agent
    }
}

View on GitHub (pinned to d16f7ba2ed)

Solutions

1. Confirm the agent checkpoint for the epoch was saved (save and load use the same strategy)
2. Check disk/permissions on the checkpoint directory
3. Use a valid epoch number present in the checkpoint storage

When it happens

Trigger: Thrown at crates/burn-train/src/learner/rl/checkpointer.rs:71 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of tracel-ai/burn@d16f7ba2ed (2026-09-05). Data as JSON: /api/errors/2c6b3b41e9246d3b. Report an issue: GitHub.