coolguy/coorl-lost-cities
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Only record evaluation_seconds when eval actually ran

Browse Source 
Previously the timer wrapped every call to _evaluate(), but on iterations
that skip eval (iteration % eval_every != 0) the function returns
immediately and the recorded value was just function-call overhead
(~3 µs), which made W&B show a wildly bimodal "evaluation_seconds"
metric. Now only set the key when eval_metrics is non-empty so
non-eval iterations have no data point.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
coolguy
2026-05-07 17:26:12 +09:00
co-authored by Claude Opus 4.7
parent e6b7a41dc0
commit 5aab84c15f
1 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/coolrl_lost_cities/games/classic/deep_cfr/trainer.py
+2 -1
View File
@@ -325,7 +325,8 @@ class DeepCFRTrainer:
eval_started = time.perf_counter() eval_started = time.perf_counter()
eval_metrics = self._evaluate(iteration) eval_metrics = self._evaluate(iteration)
self._runtime_metrics["evaluation_seconds"] = time.perf_counter() - eval_started if eval_metrics:
self._runtime_metrics["evaluation_seconds"] = time.perf_counter() - eval_started
self._runtime_metrics["advantage_memory_size"] = self._advantage_memory_size() self._runtime_metrics["advantage_memory_size"] = self._advantage_memory_size()
for player, memory in enumerate(self.advantage_memories): for player, memory in enumerate(self.advantage_memories):
self._runtime_metrics[f"advantage_player_{player}_memory_size"] = len(memory) self._runtime_metrics[f"advantage_player_{player}_memory_size"] = len(memory)