when I command
python edit_cli.py --steps 100 --resolution 512 --seed 1371 --cfg-text 7.5 --cfg-image 1.2 --input imgs/example.jpg --output imgs/output.jpg --edit "turn him into a cyborg"
ImportError: cannot import name 'autocast' from 'torch'
this error happend.
and I searched someone says I should change it to
from torch.cuda.amp import autocast
when I command
python edit_cli.py --steps 100 --resolution 512 --seed 1371 --cfg-text 7.5 --cfg-image 1.2 --input imgs/example.jpg --output imgs/output.jpg --edit "turn him into a cyborg"
ImportError: cannot import name 'autocast' from 'torch'
this error happend.
and I searched someone says I should change it to
from torch.cuda.amp import autocast
then it cause other error saying
C:\Users\pedro\instruct-pix2pix-main\edit_cli.py:128 in │
│ │
│ 125 │
│ 126 │
│ 127 if name == "main": │
│ ❱ 128 │ main() │
│ 129 │
│ │
│ C:\Users\pedro\instruct-pix2pix-main\edit_cli.py:98 in main │
│ │
│ 95 │ │ input_image.save(args.output) │
│ 96 │ │ return │
│ 97 │ │
│ ❱ 98 │ with torch.no_grad(), autocast("cuda"), model.ema_scope(): │
│ 99 │ │ cond = {} │
│ 100 │ │ cond["c_crossattn"] = [model.get_learned_conditioning([args.edit])] │
│ 101 │ │ input_image = 2 torch.tensor(np.array(input_image)).float() / 255 - 1 │
│ │
│ C:\Users\pedro\anaconda3\envs\aip\lib\site-packages\torch\cuda\amp\autocast_mode.py:127 in │
│ enter │
│ │
│ 124 │ │
│ 125 │ def enter(self): │
│ 126 │ │ self.prev = torch.is_autocast_enabled() │
│ ❱ 127 │ │ torch.set_autocast_enabled(self._enabled) │
│ 128 │ │ torch.autocast_increment_nesting() │
│ 129 │ │
│ 130 │ def exit(self, args): │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
TypeError: enabled must be a bool (got str)
torch version is '1.9.0+cu111' python version : 3.7.9