|
|
import torch |
|
|
from huggingface_guess import model_list |
|
|
|
|
|
from backend import memory_management |
|
|
from backend.diffusion_engine.base import ForgeDiffusionEngine, ForgeObjects |
|
|
from backend.modules.k_prediction import PredictionDiscreteFlow |
|
|
from backend.patcher.clip import CLIP |
|
|
from backend.patcher.unet import UnetPatcher |
|
|
from backend.patcher.vae import VAE |
|
|
from backend.text_processing.gemma_engine import GemmaTextProcessingEngine |
|
|
|
|
|
|
|
|
class Lumina2(ForgeDiffusionEngine): |
|
|
matched_guesses = [model_list.Lumina2] |
|
|
|
|
|
def __init__(self, estimated_config, huggingface_components): |
|
|
super().__init__(estimated_config, huggingface_components) |
|
|
self.is_inpaint = False |
|
|
|
|
|
clip = CLIP(model_dict={"gemma2": huggingface_components["text_encoder"]}, tokenizer_dict={"gemma2": huggingface_components["tokenizer"]}) |
|
|
|
|
|
vae = VAE(model=huggingface_components["vae"]) |
|
|
|
|
|
k_predictor = PredictionDiscreteFlow(estimated_config) |
|
|
|
|
|
unet = UnetPatcher.from_model(model=huggingface_components["transformer"], diffusers_scheduler=None, k_predictor=k_predictor, config=estimated_config) |
|
|
|
|
|
self.text_processing_engine_gemma = GemmaTextProcessingEngine( |
|
|
text_encoder=clip.cond_stage_model.gemma2, |
|
|
tokenizer=clip.tokenizer.gemma2, |
|
|
) |
|
|
|
|
|
self.forge_objects = ForgeObjects(unet=unet, clip=clip, vae=vae, clipvision=None) |
|
|
self.forge_objects_original = self.forge_objects.shallow_copy() |
|
|
self.forge_objects_after_applying_lora = self.forge_objects.shallow_copy() |
|
|
|
|
|
self.use_shift = True |
|
|
self.is_flux = True |
|
|
|
|
|
@torch.inference_mode() |
|
|
def get_learned_conditioning(self, prompt: list[str]): |
|
|
memory_management.load_model_gpu(self.forge_objects.clip.patcher) |
|
|
shift = getattr(prompt, "distilled_cfg_scale", 6.0) |
|
|
self.forge_objects.unet.model.predictor.set_parameters(shift=shift) |
|
|
return self.text_processing_engine_gemma(prompt) |
|
|
|
|
|
@torch.inference_mode() |
|
|
def get_prompt_lengths_on_ui(self, prompt): |
|
|
token_count = len(self.text_processing_engine_gemma.tokenize([prompt])[0]) |
|
|
return token_count, max(999, token_count) |
|
|
|
|
|
@torch.inference_mode() |
|
|
def encode_first_stage(self, x): |
|
|
sample = self.forge_objects.vae.encode(x.movedim(1, -1) * 0.5 + 0.5) |
|
|
sample = self.forge_objects.vae.first_stage_model.process_in(sample) |
|
|
return sample.to(x) |
|
|
|
|
|
@torch.inference_mode() |
|
|
def decode_first_stage(self, x): |
|
|
sample = self.forge_objects.vae.first_stage_model.process_out(x) |
|
|
sample = self.forge_objects.vae.decode(sample).movedim(-1, 1) * 2.0 - 1.0 |
|
|
return sample.to(x) |
|
|
|
