quark.torch.quantization.graph.optimization.model_optimization

quark.torch.quantization.graph.optimization.model_optimization#

Module Contents#

Functions#

trans_opsfunc_2_quant_module(→ torch.fx.GraphModule)

optimize the pure torch.ops.aten.*** functional model,

apply_pre_hw_constrain_passes(→ torch.fx.GraphModule)

apply_post_hw_constrain_passes(→ torch.fx.GraphModule)

quark.torch.quantization.graph.optimization.model_optimization.trans_opsfunc_2_quant_module(model: torch.fx.GraphModule) → torch.fx.GraphModule#
optimize the pure torch.ops.aten.*** functional model,

replace the

Contents