Loading scripts/optuna_gpt.py +7 −0 Original line number Diff line number Diff line Loading @@ -64,6 +64,13 @@ def objective(trial: optuna.Trial) -> float: raise optuna.TrialPruned(f"OOM error: {e}") else: raise e except AssertionError as e: # Cache dimension and n_head mismatch if "embed_dim must be divisible by num_heads" in str(e): logger.error("embed_dim must be divisible by num_heads") raise optuna.TrialPruned(f"embed_dim must be divisible by num_heads: {e}") else: raise e except Exception as e: loss = 1e6 raise e Loading Loading
scripts/optuna_gpt.py +7 −0 Original line number Diff line number Diff line Loading @@ -64,6 +64,13 @@ def objective(trial: optuna.Trial) -> float: raise optuna.TrialPruned(f"OOM error: {e}") else: raise e except AssertionError as e: # Cache dimension and n_head mismatch if "embed_dim must be divisible by num_heads" in str(e): logger.error("embed_dim must be divisible by num_heads") raise optuna.TrialPruned(f"embed_dim must be divisible by num_heads: {e}") else: raise e except Exception as e: loss = 1e6 raise e Loading