Skip to content
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.

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

update cpm 1x1 running stats #238

Merged
merged 1 commit into from
Sep 14, 2023
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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 training/nvidia/cpm-pytorch/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -48,5 +48,5 @@
| ------------------- | --------- | ---------------- | -------- | ------- | ------- | ------ | ----- | --------- |
| A100单机8卡(1x8) | fp16 | / | 1641 | 587 | 835 | 1059 | 0.92 | 12.9/40.0 |
| A100单机8卡(1x8) | fp16 | bs=128,lr=0.002 | 5469 | 771 | 1090 | 1292 | 0.918 | 23.1/40.0 |
| A100单机单卡(1x1) | fp16 | bs=192,lr=0.0005 | | 78.4 | 111.9 | 127.2 | | 34.8/40.0 |
| A100单机单卡(1x1) | fp16 | bs=192,lr=0.0005 | | 98.8 | 143.8 | 168.8 | | 39.5/40.0 |
| A100两机8卡(2x8) | fp16 | bs=192,lr=0.0005 | | 1583 | 2221 | 2583.8 | | 29.9/40.0 |
4 changes: 2 additions & 2 deletions training/nvidia/cpm-pytorch/config/config_A100x1x1.py
Original file line number Diff line number Diff line change
Expand Up @@ -3,13 +3,13 @@

fp16 = True
dist_backend = "nccl"
target_embedding_average = 0.8
target_embedding_average = 0.92

gradient_accumulation_steps = 1

train_batch_size = 32
eval_batch_size = train_batch_size
max_steps = 60000
max_steps = 3000
max_samples_termination = 439126000

warmup = 0.2
Expand Down