remove hardcode

This commit is contained in:
Dan Saunders
2025-08-19 15:41:32 +00:00
parent 9a6e9d8d15
commit c3db6dd307

View File

@@ -408,8 +408,7 @@ def calculate_total_num_steps(cfg, train_dataset, update=True):
if update:
cfg.total_num_tokens = total_num_tokens
skip_estimates = cfg.model_config_type == "mamba"
skip_estimates = True
skip_estimates = cfg.sequence_len is None or cfg.model_config_type == "mamba"
if (
not skip_estimates