File size: 401 Bytes
72268ee |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
from ._base import *
class QwenGPTQForCausalLM(BaseGPTQForCausalLM):
layer_type = "QWenBlock"
layers_block_name = "transformer.h"
outside_layer_modules = ["transformer.wte", "transformer.wpe", "transformer.ln_f"]
inside_layer_modules = [
["attn.c_attn"],
["attn.c_proj"],
["mlp.w1", "mlp.w2"],
["mlp.c_proj"]
]
__all__ = ["QwenGPTQForCausalLM"]
|