21 lines
622 B
Python
21 lines
622 B
Python
from ._base import *
|
|
|
|
|
|
class LlamaGPTQForCausalLM(BaseGPTQForCausalLM):
|
|
layer_type = "LlamaDecoderLayer"
|
|
layers_block_name = "model.layers"
|
|
outside_layer_modules = ["model.embed_tokens", "model.norm"]
|
|
inside_layer_modules = [
|
|
["self_attn.k_proj", "self_attn.v_proj", "self_attn.q_proj"],
|
|
["self_attn.o_proj"],
|
|
["mlp.up_proj", "mlp.gate_proj"],
|
|
["mlp.down_proj"]
|
|
]
|
|
|
|
@staticmethod
|
|
def _resize_attention_mask(attention_mask):
|
|
attention_mask = [each.unsqueeze(1) for each in attention_mask]
|
|
return attention_mask
|
|
|
|
|
|
__all__ = ["LlamaGPTQForCausalLM"]
|