mindspore_gs.ptq.RoundToNearest
- class mindspore_gs.ptq.RoundToNearest(config=None)[source]
Native implementation for post training quantization based on min/max statistic values.
- Parameters
config (
mindspore_gs.ptq.PTQConfig
) – config for RoundToNearst, default isNone
.- Raises
TypeError – If config type is not PTQConfig when it’s not
None
.
Examples
>>> import mindspore_gs >>> from mindspore_gs import ptq >>> from mindspore_gs.ptq import RoundToNearest as rtn >>> from mindspore_gs.ptq import PTQConfig >>> # Define the network structure of LeNet5. Refer to >>> # https://gitee.com/mindspore/docs/blob/r2.2/docs/mindspore/code/lenet.py >>> ptq = rtn() >>> network = LeNet5() >>> fake_quant_net = ptq.apply(net_work) >>> quant_net = ptq.convert(fake_quant_net)
- apply(network: Cell)[source]
Define how to add fake quantizer to network.
- Parameters
network (Cell) – Network to be fake quantized.
- Returns
fake quantized network.
- convert(net_opt: Cell, ckpt_path='')[source]
Define how to convert a compressed network to a standard network before exporting.
- Parameters
net_opt (Cell) – Network to be converted which is transformed by RoundToNearest.apply.
ckpt_path (str) – Path to checkpoint file for net_opt. Default is
""
, which means not loading checkpoint file to net_opt.
- Returns
An instance of Cell represents converted network.
- Raises
TypeError – If net_opt is not Cell.
TypeError – If ckpt_path is not string.
ValueError – If ckpt_path is not empty and invalid.