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.

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

add |---| to render tables correctly #686

Merged
merged 1 commit into from
Oct 20, 2023
Merged

add |---| to render tables correctly #686

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
4 changes: 2 additions & 2 deletions scripts/train/benchmarking/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -139,7 +139,7 @@ Our microbatching engine enables microbatch sizes that do not divde Global Batch
## A100 80GB with 1600 Gbps node-node interconnect (RoCE)

| Model | SeqLen (T) | # GPUs | GPU | MFU | HFU | Model TFLOP | MicroBatchSize | GradAccum | GlobalBatchSize | Throughput (S/s) | Throughput (T/s) | Throughput (T/s/GPU) | GlobalBatchSize (T) | Precision | MP Mode | Sharding Strategy | Activation Checkpointing | Activation CPUOffload | NumParams |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 70b | 2048 | 64 | a100_80gb | 53.33 | 71.1 | 166 | 8 | 4 | 2048 | 12 | 26274 | 410 | 4194304 | bf16 | PURE | FULL_SHARD | True | False | 64862437376 |
| 70b | 2048 | 32 | a100_80gb | 48.56 | 64.75 | 151 | 2 | 16 | 1024 | 5 | 11962 | 373 | 2097152 | bf16 | PURE | FULL_SHARD | True | False | 64862437376 |
| 30b | 8192 | 8 | a100_80gb | 39.38 | 52.5 | 122 | 1 | 21 | 168 | 0 | 4594 | 574 | 1376256 | amp_bf16 | DEFAULT | FULL_SHARD | True | False | 30019254272 |
Expand Down Expand Up @@ -205,7 +205,7 @@ Our microbatching engine enables microbatch sizes that do not divde Global Batch
## A100 40GB with 1600 Gbps node-node interconnect (RoCE)

| Model | SeqLen (T) | # GPUs | GPU | MFU | HFU | Model TFLOP| MicroBatchSize | GradAccum | GlobalBatchSize | Throughput (S/s) | Throughput (T/s) | Throughput (T/s/GPU) | GlobalBatchSize (T) | Precision | MP Mode | Sharding Strategy | Activation Checkpointing | Activation CPUOffload | NumParams |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 70b | 2048 | 128 | a100_40gb | 48.91 | 65.21 | 152 | 4 | 1 | 512 | 23 | 48194 | 376 | 1048576 | bf16 | PURE | FULL_SHARD | True | False | 64862437376 |
| 70b | 2048 | 64 | a100_40gb | 35.87 | 47.82 | 111 | 2 | 1 | 128 | 8 | 17672 | 276 | 262144 | bf16 | PURE | FULL_SHARD | True | False | 64862437376 |
| 30b | 2048 | 128 | a100_40gb | 52.25 | 69.66 | 163 | 6 | 1 | 768 | 54 | 110803 | 865 | 1572864 | bf16 | PURE | FULL_SHARD | True | False | 29975214080 |
Expand Down