oppo-mente-lab/subject-diffusion

Subject-Diffusion:Open Domain Personalized Text-to-Image Generation without Test-time Fine-tuning

subject-diffusion's Issues

Checkpoints and dataset?

Thank you for this very exciting project!

I see the script for generating images using pretrained checkpoints, but I don't see the checkpoints.

Can you please provide the checkpoints and dataset soon?

It seems incapable of generalizing unlike other finetuning methods

They all seem to have the same expression and don't seem like it's capable of generalizing to different facial expression and the faces don't match the styles in the other pictures.

Why model is not in cuda?

Thanks for your great work and sharing your code !!!
When i run the code , model is in cpu , not in GPU how to solve it? I mantually add the code "self..xx.cuda()" , but it is not working!
Looking forward to your reply!

About MODEL_ARGS in train.sh

Could you please tell me what versions of these models are?

"blip-image-captioning-large" or "blip2-opt-2.7b" ?

Dear Author, Thank you for your outstanding work. I have noticed that the data_process.py script uses two BLIP models, namely “blip-image-captioning-large” and “blip2-opt-2.7b”. May I ask which one you used?

where do we need the image_embeddings_cls in training_step

hi, thanks for your excellent work here!

I am reading the code and a little bit confused by the image_embeddings_cls in the training_step. The attention layer accurately takes the image_embeddings as inputs, leaving the image_embeddings_cls to be recorded by self.image_infos and then unset (Line 556~596).

image_embeddings_cls, image_embeddings = self.encode_images(
            batch["entity_images"], batch["image_token_idx_mask"], batch["bboxes"], latents.device)
...
self.image_infos["image_embedding"] = image_embeddings_cls[batch["image_token_idx_mask"]]
...
objects = image_embeddings
self.image_infos["image_embedding"] = None

TypeError: __call__() missing 1 required positional argument: 'data'

I have problem below :
File "/home/yons/SH100k/Subject-Diffusion-main/train.py", line 880, in
trainer.fit(model, datamoule)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 696, in fit
self._call_and_handle_interrupt(
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 650, in _call_and_handle_interrupt
return trainer_fn(*args, **kwargs)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 735, in _fit_impl
results = self._run(model, ckpt_path=self.ckpt_path)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 1166, in _run
results = self._run_stage()
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 1252, in _run_stage
return self._run_train()
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 1283, in _run_train
self.fit_loop.run()
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/loops/loop.py", line 200, in run
self.advance(*args, **kwargs)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/loops/fit_loop.py", line 271, in advance
self._outputs = self.epoch_loop.run(self._data_fetcher)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/loops/loop.py", line 195, in run
self.on_run_start(*args, **kwargs)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/loops/epoch/training_epoch_loop.py", line 147, in on_run_start
_ = iter(data_fetcher) # creates the iterator inside the fetcher
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/utilities/fetching.py", line 180, in iter
self.prefetching()
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/utilities/fetching.py", line 241, in prefetching
self._fetch_next_batch(iterator)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/utilities/fetching.py", line 277, in _fetch_next_batch
batch = next(iterator)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/supporters.py", line 557, in next
return self.request_next_batch(self.loader_iters)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/trainer/supporters.py", line 569, in request_next_batch
return apply_to_collection(loader_iters, Iterator, next)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/pytorch_lightning/utilities/apply_func.py", line 99, in apply_to_collection
return function(data, *args, **kwargs)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/prefetch_generator/init.py", line 116, in next
raise next_item
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/prefetch_generator/init.py", line 98, in run
for item in self.generator: self.queue.put((True , item))
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 681, in next
data = self._next_data()
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 1376, in _next_data
return self._process_data(data)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 1402, in _process_data
data.reraise()
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/torch/_utils.py", line 461, in reraise
raise exception
TypeError: Caught TypeError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/torch/utils/data/_utils/worker.py", line 302, in _worker_loop
data = fetcher.fetch(index)
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/torch/utils/data/_utils/fetch.py", line 32, in fetch
data.append(next(self.dataset_iter))
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/webdataset/pipeline.py", line 68, in iterator
for sample in self.iterator1():
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/webdataset/pipeline.py", line 60, in iterator1
source = self.invoke(self.pipeline[0])
File "/home/yons/anaconda3/envs/subject-diffusions/lib/python3.9/site-packages/webdataset/pipeline.py", line 54, in invoke
result = f(*args, **kwargs)
TypeError: call() missing 1 required positional argument: 'data'
hope someone can help me

About the pretrained checkpoints

are you planning to release the pretrained checkpoints?

Can you help revise the conda environment setup?

Hi it seems the huggingface/diffuser version in yaml need to be updated. I simply changed to huggingface-hub==0.13.2 and it worked.

The conflict is caused by:
The user requested huggingface-hub==0.11.0
diffusers 0.18.2 depends on huggingface-hub>=0.13.2

And then it also has another conflit:

The conflict is caused by:
The user requested protobuf==4.21.9
open-clip-torch 2.20.0 depends on protobuf<4

UNet2DConditionModel as UNet2DConditionModel_GLIGEN?

hello, I can't find third_party.diffusers in the code. Is there any missing in code

from third_party.diffusers import AutoencoderKL, PNDMScheduler, UNet2DConditionModel as UNet2DConditionModel_GLIGEN