dream.sh fail on M1

[i3oc9i]

with last commit 0433b3d

i get this error

╰─ff python ./scripts/dream.py

• Initializing, be patient...

Loading model from models/ldm/stable-diffusion-v1/model.ckpt
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
making attention of type 'vanilla' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla' with 512 in_channels
Using half precision math. Call with --full_precision to use more accurate but VRAM-intensive full precision.

Setting Sampler to k_lms
model loaded in 5.80s

• Initialization done! Awaiting your command (-h for help, 'q' to quit)
  dream> a smiling face
  Traceback (most recent call last):
  File "/Users/ivano/Code/Ai/dream2/./scripts/dream.py", line 543, in
  main()
  File "/Users/ivano/Code/Ai/dream2/./scripts/dream.py", line 103, in main
  main_loop(t2i, opt.outdir, opt.prompt_as_dir, cmd_parser, infile)
  File "/Users/ivano/Code/Ai/dream2/./scripts/dream.py", line 230, in main_loop
  t2i.prompt2image(image_callback=image_writer, **vars(opt))
  File "/Users/ivano/Code/Ai/dream/ldm/simplet2i.py", line 289, in prompt2image
  torch.cuda.torch.cuda.reset_peak_memory_stats()
  File "/Users/ivano/.miniconda/envs/dream/lib/python3.10/site-packages/torch/cuda/memory.py", line 256, in reset_peak_memory_stats
  return torch._C._cuda_resetPeakMemoryStats(device)
  AttributeError: module 'torch._C' has no attribute '_cuda_resetPeakMemoryStats'

[owen109]

Also having this issue

[magnusviri]

We are working on it.

[i3oc9i]

the brach fix-cuda-reset-stats-error fix the error but then not image are generated

python scripts/dream.py

• Initializing, be patient...

Loading model from models/ldm/stable-diffusion-v1/model.ckpt
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
making attention of type 'vanilla' with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla' with 512 in_channels
Using half precision math. Call with --full_precision to use more accurate but VRAM-intensive full precision.

Setting Sampler to k_lms
model loaded in 5.75s

• Initialization done! Awaiting your command (-h for help, 'q' to quit)
  dream> a nice smiling girl
  User specified autocast device_type must be 'cuda' or 'cpu'
  Are you sure your system has an adequate NVIDIA GPU?
  Usage stats:
  0 image(s) generated in 0.00s
  Max VRAM used for this generation: 0.00G

[Vargol]

@i3oc9i, thats fixed elsewhere, I hit the some issue last night (UK time) and it was fixed by a change proposed by @toffaletti

I think that may be been merged, but into the main branch.