quantize-rs 0.7.0

Neural network quantization toolkit for ONNX models
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

# Example quantize-rs configuration file (TOML format)

# Global settings
bits = 8
per_channel = true

# Individual models
[[models]]
input = "models/resnet18.onnx"
output = "quantized/resnet18_int8.onnx"

[[models]]
input = "models/mobilenet.onnx"
output = "quantized/mobilenet_int8.onnx"
per_channel = false  # Override

[[models]]
input = "models/efficientnet.onnx"
output = "quantized/efficientnet_int8.onnx"
skip_existing = true

# Batch processing
[batch]
input_dir = "models/*.onnx"
output_dir = "quantized_batch/"
skip_existing = true
continue_on_error = true