Skip to content

Internal. #580

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
wants to merge 1 commit into from
Closed

Internal. #580

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
21 changes: 6 additions & 15 deletions jax_tpu_embedding/sparsecore/lib/nn/tests/preprocess_input_benchmarks.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,17 +26,14 @@

The --benchmark_filter flag uses a regex to select benchmarks. For parameterized
benchmarks, the name is typically formatted as:
`[benchmark_name]/[param1]:[value1]/[param2]:[value2]`.
Boolean parameters are often represented as 0 for False and 1 for True.
`[benchmark_name]/[param1]:[value1]`.

For example, to run only the `sparse_coo` benchmarks:
`--benchmark_filter=preprocess_input_benchmark_sparse_coo`

To run only the `sparse_coo` benchmark where `has_leading_dimension` is `False`:
`--benchmark_filter='preprocess_input_benchmark_sparse_coo/has_leading_dimension:0'`

To run all benchmarks across all suites where `has_leading_dimension` is `False`:
`--benchmark_filter='/has_leading_dimension:0'`
To run only the ragged benchmark with ragged=True:
`--benchmark_filter='preprocess_input_benchmark/ragged:1'`

To upload the profile to pprof:
pprof -flame /tmp/preprocess.prof
Expand Down Expand Up @@ -220,12 +217,12 @@ def apply_fdo_stats(

@google_benchmark.register
@google_benchmark.option.unit(google_benchmark.kMillisecond)
@google_benchmark.option.arg_names(["ragged", "has_leading_dimension"])
@google_benchmark.option.args_product([[False, True], [False, True]])
@google_benchmark.option.arg_names(["ragged"])
@google_benchmark.option.args_product([[False, True]])
@google_benchmark.option.iterations(100)
def preprocess_input_benchmark(state: google_benchmark.State):
"""Benchmark for preprocessing input for sparse-dense matmul."""
ragged, has_leading_dimension = state.range(0), state.range(1)
ragged = state.range(0)
if ragged:
features, feature_weights = _GLOBAL_RAGGED_FEATURES, _GLOBAL_RAGGED_WEIGHTS
else:
Expand All @@ -241,7 +238,6 @@ def preprocess_input_benchmark(state: google_benchmark.State):
local_device_count=4,
global_device_count=16,
num_sc_per_device=4,
has_leading_dimension=has_leading_dimension,
batch_number=batch_num,
allow_id_dropping=batch_num == 0,
)
Expand All @@ -253,12 +249,9 @@ def preprocess_input_benchmark(state: google_benchmark.State):

@google_benchmark.register
@google_benchmark.option.unit(google_benchmark.kMillisecond)
@google_benchmark.option.arg_name("has_leading_dimension")
@google_benchmark.option.args_product([[False, True]])
@google_benchmark.option.iterations(100)
def preprocess_input_benchmark_sparse_coo(state: google_benchmark.State):
"""Benchmark for preprocessing input for sparse-dense matmul."""
has_leading_dimension = state.range(0)
batch_num = 0
while state:
if batch_num == 0:
Expand All @@ -272,7 +265,6 @@ def preprocess_input_benchmark_sparse_coo(state: google_benchmark.State):
local_device_count=4,
global_device_count=16,
num_sc_per_device=4,
has_leading_dimension=has_leading_dimension,
batch_number=batch_num,
allow_id_dropping=batch_num == 0,
)
Expand Down Expand Up @@ -312,7 +304,6 @@ def worker(host_id: int, batch_number: int):
local_device_count=4,
global_device_count=16,
num_sc_per_device=4,
has_leading_dimension=False,
enable_minibatching=True,
all_reduce_interface=all_reduce_interfaces[host_id],
batch_number=batch_number,
Expand Down
Loading