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

Improve speed in combining per-channel data #21563

Merged
merged 2 commits into from
Aug 6, 2024
Merged

Improve speed in combining per-channel data #21563

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
00fcfdb
Improve speed in combining per-channel data
duanshengliu Jul 30, 2024
4fbb11a
Merge branch 'main' into improve-combine-per-channel-data-speed
duanshengliu Aug 5, 2024
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
14 changes: 5 additions & 9 deletions onnxruntime/python/tools/quantization/base_quantizer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -418,6 +418,9 @@ def quantize_weight_per_channel_impl(
zero_point_list = []
scale_list = []
quantized_per_channel_data_list = []
weights_shape = list(weights.shape)
reshape_dims = list(weights_shape) # deep copy
reshape_dims[channel_axis] = 1 # only one per channel for reshape
for i in range(channel_count):
per_channel_data = weights.take(i, channel_axis)
channel_override_index = i if i < num_channel_overrides else 0
Expand Down Expand Up @@ -460,17 +463,10 @@ def quantize_weight_per_channel_impl(

zero_point_list.append(zero_point)
scale_list.append(scale)
quantized_per_channel_data_list.append(quantized_per_channel_data)
quantized_per_channel_data_list.append(np.asarray(quantized_per_channel_data).reshape(reshape_dims))

# combine per_channel_data into one
weights_shape = list(weights.shape)
reshape_dims = list(weights_shape) # deep copy
reshape_dims[channel_axis] = 1 # only one per channel for reshape
quantized_weights = np.asarray(quantized_per_channel_data_list[0]).reshape(reshape_dims)
for i in range(1, len(quantized_per_channel_data_list)):
channel_weights = np.asarray(quantized_per_channel_data_list[i]).reshape(reshape_dims)
quantized_weights = np.concatenate((quantized_weights, channel_weights), channel_axis)

quantized_weights = np.concatenate(quantized_per_channel_data_list, channel_axis)
q_weight_name = weight_name + TENSOR_NAME_QUANT_SUFFIX
zp_name = weight_name + "_zero_point"
scale_name = weight_name + "_scale"
Expand Down
Loading