qbcompiler.configs.quantization_config.BitConfig Class Reference#
|
Mobilint SDK qb Compiler v1.0
MCS002-EN
|
qbcompiler.configs.quantization_config.BitConfig Class Reference
Configuration for bit precision. More...
Inheritance diagram for qbcompiler.configs.quantization_config.BitConfig:
Public Member Functions | |
| __init__ (self, int query_act_bits, int key_act_bits, int value_act_bits, int output_act_bits, int ffn_act_bits, int head_act_bits, int query_weight_bits, int key_weight_bits, int value_weight_bits, int output_weight_bits, int ffn_weight_bits, int head_weight_bits, bool mixed_precision_apply, str save_path, str load_path, List[str] activation_16bits, List[str] weight_16bits, bool optional=optional) | |
| Initialize the BitConfig. | |
| default_config (cls) | |
| from_kwargs (cls, **kwargs) | |
| from_dict (cls, dict config_dict) | |
| Create BitConfig from dictionary (JSON structure) | |
| to_dict (self) | |
Public Attributes | |
| query_act_bits = query_act_bits | |
| key_act_bits = key_act_bits | |
| value_act_bits = value_act_bits | |
| output_act_bits = output_act_bits | |
| ffn_act_bits = ffn_act_bits | |
| head_act_bits = head_act_bits | |
| query_weight_bits = query_weight_bits | |
| key_weight_bits = key_weight_bits | |
| value_weight_bits = value_weight_bits | |
| output_weight_bits = output_weight_bits | |
| ffn_weight_bits = ffn_weight_bits | |
| head_weight_bits = head_weight_bits | |
| mixed_precision_apply = mixed_precision_apply | |
| save_path = save_path | |
| load_path = load_path | |
| activation_16bits = activation_16bits | |
| weight_16bits = weight_16bits | |
Static Public Attributes | |
| bool | optional = False |
| dict | DEFAULTS |
| dict | py_to_cpp_name |
| dict | min_max_check |
Detailed Description
Configuration for bit precision.
Defines bit-width parameterization for activations and weights used in mixed-precision quantization (e.g., attention and FFN components).
Definition at line 316 of file quantization_config.py.
The documentation for this class was generated from the following file:
Generated by