Merge pull request #8 from dogewanwan/master

Support for using the script with textual inversion repo, should be safe for use in normal repositories too.
This commit is contained in:
AUTOMATIC1111 2022-08-24 23:00:45 +03:00 committed by GitHub
commit ec8a252260
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -58,6 +58,7 @@ parser.add_argument("--grid-extended-filename", action='store_true', help="save
parser.add_argument("--jpeg-quality", type=int, default=80, help="quality for saved jpeg images")
parser.add_argument("--disable-pnginfo", action='store_true', help="disable saving text information about generation parameters as chunks to png files")
parser.add_argument("--inversion", action='store_true', help="switch to stable inversion version; allows for uploading embeddings; this option should be used only with textual inversion repo")
opt = parser.parse_args()
GFPGAN_dir = opt.gfpgan_dir
@ -189,8 +190,8 @@ if os.path.exists(GFPGAN_dir):
print("Error loading GFPGAN:", file=sys.stderr)
print(traceback.format_exc(), file=sys.stderr)
config = OmegaConf.load("configs/stable-diffusion/v1-inference.yaml")
model = load_model_from_config(config, "models/ldm/stable-diffusion-v1/model.ckpt")
config = OmegaConf.load(opt.config)
model = load_model_from_config(config, opt.ckpt)
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
model = (model if opt.no_half else model.half()).to(device)
@ -467,9 +468,17 @@ def process_images(outpath, func_init, func_sample, prompt, seed, sampler_name,
return output_images, seed, info
def txt2img(prompt: str, ddim_steps: int, sampler_name: str, use_GFPGAN: bool, prompt_matrix: bool, ddim_eta: float, n_iter: int, batch_size: int, cfg_scale: float, seed: int, height: int, width: int):
def load_embeddings(fp):
if fp is not None and hasattr(model, "embedding_manager"):
# load the file
model.embedding_manager.load(fp.name)
def txt2img(prompt: str, ddim_steps: int, sampler_name: str, use_GFPGAN: bool, prompt_matrix: bool, ddim_eta: float, n_iter: int, batch_size: int, cfg_scale: float, seed: int, height: int, width: int, embeddings_fp):
outpath = opt.outdir or "outputs/txt2img-samples"
load_embeddings(embeddings_fp)
if sampler_name == 'PLMS':
sampler = PLMSSampler(model)
elif sampler_name == 'DDIM':
@ -564,6 +573,7 @@ txt2img_interface = gr.Interface(
gr.Number(label='Seed', value=-1),
gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512),
gr.Slider(minimum=64, maximum=2048, step=64, label="Width", value=512),
gr.File(label = "Embeddings file for textual inversion", visible=opt.inversion)
],
outputs=[
gr.Gallery(label="Images"),
@ -576,9 +586,11 @@ txt2img_interface = gr.Interface(
)
def img2img(prompt: str, init_img, ddim_steps: int, use_GFPGAN: bool, prompt_matrix, loopback: bool, n_iter: int, batch_size: int, cfg_scale: float, denoising_strength: float, seed: int, height: int, width: int, resize_mode: int):
def img2img(prompt: str, init_img, ddim_steps: int, use_GFPGAN: bool, prompt_matrix, loopback: bool, n_iter: int, batch_size: int, cfg_scale: float, denoising_strength: float, seed: int, height: int, width: int, resize_mode: int, embeddings_fp):
outpath = opt.outdir or "outputs/img2img-samples"
load_embeddings(embeddings_fp)
sampler = KDiffusionSampler(model)
assert 0. <= denoising_strength <= 1., 'can only work with strength in [0.0, 1.0]'
@ -693,7 +705,8 @@ img2img_interface = gr.Interface(
gr.Number(label='Seed', value=-1),
gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512),
gr.Slider(minimum=64, maximum=2048, step=64, label="Width", value=512),
gr.Radio(label="Resize mode", choices=["Just resize", "Crop and resize", "Resize and fill"], type="index", value="Just resize")
gr.Radio(label="Resize mode", choices=["Just resize", "Crop and resize", "Resize and fill"], type="index", value="Just resize"),
gr.File(label = "Embeddings file for textual inversion", visible=opt.inversion)
],
outputs=[
gr.Gallery(),
@ -739,6 +752,7 @@ if GFPGAN is not None:
allow_flagging="never",
), "GFPGAN"))
demo = gr.TabbedInterface(
interface_list=[x[0] for x in interfaces],
tab_names=[x[1] for x in interfaces],