mirror of
https://github.com/modelscope/DiffSynth-Studio.git
synced 2026-03-19 06:39:43 +00:00
47 lines
1.4 KiB
Python
47 lines
1.4 KiB
Python
from .base_prompter import BasePrompter
|
|
from ..models.flux_text_encoder import FluxTextEncoder2
|
|
from transformers import T5TokenizerFast
|
|
import os
|
|
|
|
|
|
class CogPrompter(BasePrompter):
|
|
def __init__(
|
|
self,
|
|
tokenizer_path=None
|
|
):
|
|
if tokenizer_path is None:
|
|
base_path = os.path.dirname(os.path.dirname(__file__))
|
|
tokenizer_path = os.path.join(base_path, "tokenizer_configs/cog/tokenizer")
|
|
super().__init__()
|
|
self.tokenizer = T5TokenizerFast.from_pretrained(tokenizer_path)
|
|
self.text_encoder: FluxTextEncoder2 = None
|
|
|
|
|
|
def fetch_models(self, text_encoder: FluxTextEncoder2 = None):
|
|
self.text_encoder = text_encoder
|
|
|
|
|
|
def encode_prompt_using_t5(self, prompt, text_encoder, tokenizer, max_length, device):
|
|
input_ids = tokenizer(
|
|
prompt,
|
|
return_tensors="pt",
|
|
padding="max_length",
|
|
max_length=max_length,
|
|
truncation=True,
|
|
).input_ids.to(device)
|
|
prompt_emb = text_encoder(input_ids)
|
|
prompt_emb = prompt_emb.reshape((1, prompt_emb.shape[0]*prompt_emb.shape[1], -1))
|
|
|
|
return prompt_emb
|
|
|
|
|
|
def encode_prompt(
|
|
self,
|
|
prompt,
|
|
positive=True,
|
|
device="cuda"
|
|
):
|
|
prompt = self.process_prompt(prompt, positive=positive)
|
|
prompt_emb = self.encode_prompt_using_t5(prompt, self.text_encoder, self.tokenizer, 226, device)
|
|
return prompt_emb
|