mirror of
https://github.com/hiyouga/LLaMA-Factory.git
synced 2025-08-28 08:42:51 +08:00
48 lines
1.7 KiB
Python
48 lines
1.7 KiB
Python
import math
|
|
from typing import TYPE_CHECKING
|
|
|
|
from ...extras.logging import get_logger
|
|
|
|
|
|
if TYPE_CHECKING:
|
|
from transformers import PretrainedConfig
|
|
|
|
from ...hparams import ModelArguments
|
|
|
|
|
|
logger = get_logger(__name__)
|
|
|
|
|
|
def configure_rope(config: "PretrainedConfig", model_args: "ModelArguments", is_trainable: bool) -> None:
|
|
if model_args.rope_scaling is None:
|
|
return
|
|
|
|
if not hasattr(config, "rope_scaling"):
|
|
logger.warning("Current model does not support RoPE scaling.")
|
|
return
|
|
|
|
if is_trainable:
|
|
if model_args.rope_scaling == "dynamic":
|
|
logger.warning(
|
|
"Dynamic NTK scaling may not work well with fine-tuning. "
|
|
"See: https://github.com/huggingface/transformers/pull/24653"
|
|
)
|
|
|
|
current_max_length = getattr(config, "max_position_embeddings", None)
|
|
if current_max_length and model_args.model_max_length > current_max_length:
|
|
logger.info(
|
|
"Enlarge max model length from {} to {}.".format(current_max_length, model_args.model_max_length)
|
|
)
|
|
setattr(config, "max_position_embeddings", model_args.model_max_length)
|
|
scaling_factor = float(math.ceil(model_args.model_max_length / current_max_length))
|
|
else:
|
|
logger.warning("Input length is smaller than max length. Consider increase input length.")
|
|
scaling_factor = 1.0
|
|
else:
|
|
scaling_factor = 2.0
|
|
|
|
setattr(config, "rope_scaling", {"type": model_args.rope_scaling, "factor": scaling_factor})
|
|
logger.info(
|
|
"Using {} scaling strategy and setting scaling factor to {}".format(model_args.rope_scaling, scaling_factor)
|
|
)
|