Skip to content

fix off-by-one in steady-state MFU calculation #93

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Khadka-Bishal wants to merge 1 commit into from
+1 −1
Closed

fix off-by-one in steady-state MFU calculation #93

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -614,7 +614,7 @@ def get_weight_decay(progress):
# Final summary
t_end = time.time()
startup_time = t_start_training - t_start
steady_state_mfu = 100 * num_flops_per_token * TOTAL_BATCH_SIZE * (step - 10) / total_training_time / H100_BF16_PEAK_FLOPS if total_training_time > 0 else 0
steady_state_mfu = 100 * num_flops_per_token * TOTAL_BATCH_SIZE * (step - 11) / total_training_time / H100_BF16_PEAK_FLOPS if total_training_time > 0 else 0
peak_vram_mb = torch.cuda.max_memory_allocated() / 1024 / 1024

print("---")
Expand Down