116 lines
3.7 KiB
Python
116 lines
3.7 KiB
Python
from logging import getLogger
|
|
from typing import Dict, List, Optional, Union
|
|
|
|
import numpy as np
|
|
from diffusers import OnnxRuntimeModel
|
|
from diffusers.pipelines.onnx_utils import ORT_TO_NP_TYPE
|
|
from optimum.onnxruntime.modeling_diffusion import ORTModelUnet
|
|
|
|
from ...server import ServerContext
|
|
|
|
logger = getLogger(__name__)
|
|
|
|
|
|
class UNetWrapper(object):
|
|
input_types: Optional[Dict[str, np.dtype]] = None
|
|
prompt_embeds: Optional[List[np.ndarray]] = None
|
|
prompt_index: int = 0
|
|
server: ServerContext
|
|
wrapped: Union[OnnxRuntimeModel, ORTModelUnet]
|
|
xl: bool
|
|
|
|
def __init__(
|
|
self,
|
|
server: ServerContext,
|
|
wrapped: Union[OnnxRuntimeModel, ORTModelUnet],
|
|
xl: bool,
|
|
):
|
|
self.server = server
|
|
self.wrapped = wrapped
|
|
self.xl = xl
|
|
|
|
self.cache_input_types()
|
|
|
|
def __call__(
|
|
self,
|
|
sample: Optional[np.ndarray] = None,
|
|
timestep: Optional[np.ndarray] = None,
|
|
encoder_hidden_states: Optional[np.ndarray] = None,
|
|
**kwargs,
|
|
):
|
|
logger.trace(
|
|
"UNet parameter types: %s, %s, %s",
|
|
sample.dtype,
|
|
timestep.dtype,
|
|
encoder_hidden_states.dtype,
|
|
)
|
|
|
|
if self.prompt_embeds is not None:
|
|
step_index = self.prompt_index % len(self.prompt_embeds)
|
|
logger.trace("multiple prompt embeds found, using step: %s", step_index)
|
|
encoder_hidden_states = self.prompt_embeds[step_index]
|
|
self.prompt_index += 1
|
|
|
|
if self.input_types is None:
|
|
self.cache_input_types()
|
|
|
|
if encoder_hidden_states.dtype != self.input_types["encoder_hidden_states"]:
|
|
logger.trace("converting UNet hidden states to input dtype")
|
|
encoder_hidden_states = encoder_hidden_states.astype(
|
|
self.input_types["encoder_hidden_states"]
|
|
)
|
|
|
|
if sample.dtype != self.input_types["sample"]:
|
|
logger.trace("converting UNet sample to input dtype")
|
|
sample = sample.astype(self.input_types["sample"])
|
|
|
|
if timestep.dtype != self.input_types["timestep"]:
|
|
logger.trace("converting UNet timestep to input dtype")
|
|
timestep = timestep.astype(self.input_types["timestep"])
|
|
|
|
return self.wrapped(
|
|
sample=sample,
|
|
timestep=timestep,
|
|
encoder_hidden_states=encoder_hidden_states,
|
|
**kwargs,
|
|
)
|
|
|
|
def __getattr__(self, attr):
|
|
return getattr(self.wrapped, attr)
|
|
|
|
def cache_input_types(self):
|
|
# TODO: use server dtype as default
|
|
if isinstance(self.wrapped, ORTModelUnet):
|
|
session = self.wrapped.session
|
|
elif isinstance(self.wrapped, OnnxRuntimeModel):
|
|
session = self.wrapped.model
|
|
else:
|
|
raise ValueError()
|
|
|
|
inputs = session.get_inputs()
|
|
self.input_types = dict(
|
|
[(input.name, ORT_TO_NP_TYPE[input.type]) for input in inputs]
|
|
)
|
|
logger.debug("cached UNet input types: %s", self.input_types)
|
|
|
|
# [
|
|
# (
|
|
# input.name,
|
|
# next(
|
|
# [
|
|
# TENSOR_TYPE_TO_NP_TYPE[field[1].elem_type]
|
|
# for field in input.type.ListFields()
|
|
# ],
|
|
# np.float32,
|
|
# ),
|
|
# )
|
|
# for input in self.wrapped.model.graph.input
|
|
# ]
|
|
|
|
def set_prompts(self, prompt_embeds: List[np.ndarray]):
|
|
logger.debug(
|
|
"setting prompt embeds for UNet: %s", [p.shape for p in prompt_embeds]
|
|
)
|
|
self.prompt_embeds = prompt_embeds
|
|
self.prompt_index = 0
|