Move environment-mac.yaml to Python 3.9 and patch dream.py for Macs.

I'm using stable-diffusion on a 2022 Macbook M2 Air with 24 GB unified memory.
I see this taking about 2.0s/it.

I've moved many deps from pip to conda-forge, to take advantage of the
precompiled binaries. Some notes for Mac users, since I've seen a lot of
confusion about this:

One doesn't need the `apple` channel to run this on a Mac-- that's only
used by `tensorflow-deps`, required for running tensorflow-metal. For
that, I have an example environment.yml here:

https://developer.apple.com/forums/thread/711792?answerId=723276022#723276022

However, the `CONDA_ENV=osx-arm64` environment variable *is* needed to
ensure that you do not run any Intel-specific packages such as `mkl`,
which will fail with [cryptic errors](CompVis/stable-diffusion#25 (comment))
on the ARM architecture and cause the environment to break.

I've also added a comment in the env file about 3.10 not working yet.
When it becomes possible to update, those commands run on an osx-arm64
machine should work to determine the new version set.

Here's what a successful run of dream.py should look like:

```
$ python scripts/dream.py --full_precision                                                                                                           SIGABRT(6) ↵  08:42:59
* 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 slower but more accurate full-precision math (--full_precision)
>> Setting Sampler to k_lms
model loaded in 6.12s

* Initialization done! Awaiting your command (-h for help, 'q' to quit)
dream> "an astronaut riding a horse"
Generating:   0%|                                                                                                                                                                         | 0/1 [00:00<?, ?it/s]/Users/corajr/Documents/lstein/ldm/modules/embedding_manager.py:152: UserWarning: The operator 'aten::nonzero' is not currently supported on the MPS backend and will fall back to run on the CPU. This may have performance implications. (Triggered internally at /Users/runner/work/_temp/anaconda/conda-bld/pytorch_1662016319283/work/aten/src/ATen/mps/MPSFallback.mm:11.)
  placeholder_idx = torch.where(
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 50/50 [01:37<00:00,  1.95s/it]
Generating: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [01:38<00:00, 98.55s/it]
Usage stats:
   1 image(s) generated in 98.60s
   Max VRAM used for this generation: 0.00G
Outputs:
outputs/img-samples/000001.1525943180.png: "an astronaut riding a horse" -s50 -W512 -H512 -C7.5 -Ak_lms -F -S1525943180
```

Author: corajr

README-Mac-MPS.md

@@ -12,8 +12,7 @@ issue](https://github.com/CompVis/stable-diffusion/issues/25), and generally on
 
 You have to have macOS 12.3 Monterey or later. Anything earlier than that won't work.
 
-BTW, I haven't tested any of this on Intel Macs but I have read that one person
-got it to work.
+Tested on a 2022 Macbook M2 Air with 10-core gpu 24 GB unified memory.
 
 How to:
 
@@ -22,17 +21,16 @@ git clone https://github.com/lstein/stable-diffusion.git
 cd stable-diffusion
 
 mkdir -p models/ldm/stable-diffusion-v1/
-ln -s /path/to/ckpt/sd-v1-1.ckpt models/ldm/stable-diffusion-v1/model.ckpt
+PATH_TO_CKPT="$HOME/Documents/stable-diffusion-v-1-4-original"  # or wherever yours is.
+ln -s "$PATH_TO_CKPT/sd-v1-4.ckpt" models/ldm/stable-diffusion-v1/model.ckpt
 
-conda env create -f environment-mac.yaml
+CONDA_SUBDIR=osx-arm64 conda env create -f environment-mac.yaml
 conda activate ldm
 
 python scripts/preload_models.py
-python scripts/orig_scripts/txt2img.py --prompt "a photograph of an astronaut riding a horse" --plms
+python scripts/dream.py --full_precision  # half-precision requires autocast and won't work
 ```
 
-We have not gotten lstein's dream.py to work yet.
-
 After you follow all the instructions and run txt2img.py you might get several errors. Here's the errors I've seen and found solutions for.
 
 ### Is it slow?
@@ -94,10 +92,6 @@ get quick feedback.
 
 	python ./scripts/txt2img.py --prompt "ocean" --ddim_steps 5 --n_samples 1 --n_iter 1
 
-### MAC: torch._C' has no attribute '_cuda_resetPeakMemoryStats' #234
-
-We haven't fixed gotten dream.py to work on Mac yet.
-
 ### OSError: Can't load tokenizer for 'openai/clip-vit-large-patch14'...
 
 	python scripts/preload_models.py
@@ -108,7 +102,7 @@ Example error.
 
 ```
 ...
-NotImplementedError: The operator 'aten::index.Tensor' is not current implemented for the MPS device. If you want this op to be added in priority during the prototype phase of this feature, please comment on [https://github.com/pytorch/pytorch/issues/77764](https://github.com/pytorch/pytorch/issues/77764). As a temporary fix, you can set the environment variable `PYTORCH_ENABLE_MPS_FALLBACK=1` to use the CPU as a fallback for this op. WARNING: this will be slower than running natively on MPS.
+NotImplementedError: The operator 'aten::_index_put_impl_' is not current implemented for the MPS device. If you want this op to be added in priority during the prototype phase of this feature, please comment on [https://github.com/pytorch/pytorch/issues/77764](https://github.com/pytorch/pytorch/issues/77764). As a temporary fix, you can set the environment variable `PYTORCH_ENABLE_MPS_FALLBACK=1` to use the CPU as a fallback for this op. WARNING: this will be slower than running natively on MPS.
 ```
 
 The lstein branch includes this fix in [environment-mac.yaml](https://github.com/lstein/stable-diffusion/blob/main/environment-mac.yaml).
@@ -137,27 +131,18 @@ still working on it.
 
 	OMP: Error #15: Initializing libiomp5.dylib, but found libomp.dylib already initialized.
 
-There are several things you can do. First, you could use something
-besides Anaconda like miniforge. I read a lot of things online telling
-people to use something else, but I am stuck with Anaconda for other
-reasons.
-
-Or you can try this.
-
-	export KMP_DUPLICATE_LIB_OK=True
-
-Or this (which takes forever on my computer and didn't work anyway).
+You are likely using an Intel package by mistake. Be sure to run conda with
+the environment variable `CONDA_SUBDIR=osx-arm64`, like so:
 
-	conda install nomkl
+`CONDA_SUBDIR=osx-arm64 conda install ...`
 
-This error happens with Anaconda on Macs, and
-[nomkl](https://stackoverflow.com/questions/66224879/what-is-the-nomkl-python-package-used-for)
-is supposed to fix the issue (it isn't a module but a fix of some
-sort). [There's more
-suggestions](https://stackoverflow.com/questions/53014306/error-15-initializing-libiomp5-dylib-but-found-libiomp5-dylib-already-initial),
-like uninstalling tensorflow and reinstalling. I haven't tried them.
+This error happens with Anaconda on Macs when the Intel-only `mkl` is pulled in by
+a dependency. [nomkl](https://stackoverflow.com/questions/66224879/what-is-the-nomkl-python-package-used-for)
+is a metapackage designed to prevent this, by making it impossible to install
+`mkl`, but if your environment is already broken it may not work. 
 
-Since I switched to miniforge I haven't seen the error.
+Do *not* use `os.environ['KMP_DUPLICATE_LIB_OK']='True'` or equivalents as this
+masks the underlying issue of using Intel packages.
 
 ### Not enough memory.
 
@@ -226,4 +211,8 @@ What? Intel? On an Apple Silicon?
 	The processor must support the Intel(R) Streaming SIMD Extensions 4.2 (Intel(R) SSE4.2) instructions.
 	The processor must support the Intel(R) Advanced Vector Extensions (Intel(R) AVX) instructions.
 
-This was actually the issue that I couldn't solve until I switched to miniforge.
+This is due to the Intel `mkl` package getting picked up when you try to install
+something that depends on it-- Rosetta can translate some Intel instructions but
+not the specialized ones here. To avoid this, make sure to use the environment
+variable `CONDA_SUBDIR=osx-arm64`, which restricts the Conda environment to only
+use ARM packages, and use `nomkl` as described above.

environment-mac.yaml

@@ -1,33 +1,57 @@
 name: ldm
 channels:
-  - apple
-  - conda-forge
   - pytorch-nightly
-  - defaults
+  - conda-forge
 dependencies:
-  - python=3.10.4
-  - pip=22.1.2
+  - python==3.9.13
+  - pip==22.2.2
+ 
+  # pytorch-nightly, left unpinned
   - pytorch
+  - torchmetrics
   - torchvision
-  - numpy=1.23.1
+
+  # I suggest to keep the other deps sorted for convenience.
+  # If you wish to upgrade to 3.10, try to run this:
+  #
+  # ```shell
+  # CONDA_CMD=conda
+  # sed -E 's/python==3.9.13/python==3.10.5/;s/ldm/ldm-3.10/;21,99s/- ([^=]+)==.+/- \1/' environment-mac.yaml > /tmp/environment-mac-updated.yml
+  # CONDA_SUBDIR=osx-arm64 $CONDA_CMD env create -f /tmp/environment-mac-updated.yml && $CONDA_CMD list -n ldm-3.10 | awk ' {print "  - " $1 "==" $2;} '
+  # ```
+  #
+  # Unfortunately, as of 2022-08-31, this fails at the pip stage.
+  - albumentations==1.2.1
+  - coloredlogs==15.0.1
+  - einops==0.4.1
+  - grpcio==1.46.4
+  - humanfriendly
+  - imageio-ffmpeg==0.4.7
+  - imageio==2.21.2
+  - imgaug==0.4.0
+  - kornia==0.6.7
+  - mpmath==1.2.1
+  - nomkl
+  - numpy==1.23.2
+  - omegaconf==2.1.1
+  - onnx==1.12.0
+  - onnxruntime==1.12.1
+  - opencv==4.6.0
+  - pudb==2022.1
+  - pytorch-lightning==1.6.5
+  - scipy==1.9.1
+  - streamlit==1.12.2
+  - sympy==1.10.1
+  - tensorboard==2.9.0
+  - transformers==4.21.2
   - pip:
-    - albumentations==0.4.6
-    - opencv-python==4.6.0.66
-    - pudb==2019.2
-    - imageio==2.9.0
-    - imageio-ffmpeg==0.4.2
-    - pytorch-lightning==1.4.2
-    - omegaconf==2.1.1
-    - test-tube>=0.7.5
-    - streamlit==1.12.0
-    - pillow==9.2.0
-    - einops==0.3.0
-    - torch-fidelity==0.3.0
-    - transformers==4.19.2
-    - torchmetrics==0.6.0
-    - kornia==0.6.0
-    - -e git+https://github.com/openai/CLIP.git@main#egg=clip
+    - invisible-watermark
+    - test-tube
+    - tokenizers
+    - torch-fidelity
+    - -e git+https://github.com/huggingface/[email protected]#egg=diffusers
     - -e git+https://github.com/CompVis/taming-transformers.git@master#egg=taming-transformers
+    - -e git+https://github.com/openai/CLIP.git@main#egg=clip
     - -e git+https://github.com/lstein/k-diffusion.git@master#egg=k-diffusion
     - -e .
 variables:

ldm/simplet2i.py

@@ -272,14 +272,15 @@ def process_image(image,seed):
         if not(width == self.width and height == self.height):
             width, height, _ = self._resolution_check(width, height, log=True)
 
-        scope = autocast if self.precision == 'autocast' else nullcontext
+        scope = autocast if self.precision == 'autocast' and torch.cuda.is_available() else nullcontext
 
         if sampler_name and (sampler_name != self.sampler_name):
             self.sampler_name = sampler_name
             self._set_sampler()
 
         tic = time.time()
-        torch.cuda.torch.cuda.reset_peak_memory_stats()
+        if torch.cuda.is_available():
+            torch.cuda.torch.cuda.reset_peak_memory_stats()
         results = list()
 
         try: