vij_wan

Author: mailvijayasingh

lat.npy

@@ -0,0 +1,61 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import Sequence
+import jax
+import time
+from maxdiffusion.pipelines.wan.wan_pipeline import WanPipeline
+from maxdiffusion import pyconfig
+from absl import app
+from maxdiffusion.utils import export_to_video
+
+def run(config):
+  pipeline = WanPipeline.from_pretrained(config)
+  s0 = time.perf_counter()
+  videos = pipeline(
+    prompt=[config.prompt]*jax.device_count(),
+    negative_prompt=[config.negative_prompt]*jax.device_count(),
+    height=config.height,
+    width=config.width,
+    num_frames=config.num_frames,
+    num_inference_steps=config.num_inference_steps,
+    guidance_scale=config.guidance_scale,
+  )
+
+  print("compile time: ", (time.perf_counter() - s0))
+  for i in range(len(videos)):
+    export_to_video(videos[i], f"wan_output_{i}.mp4", fps=16)
+  s0 = time.perf_counter()
+  with jax.profiler.trace("/tmp/trace/"):
+    videos = pipeline(
+      prompt=[config.prompt]*jax.device_count(),
+      negative_prompt=[config.negative_prompt]*jax.device_count(),
+      height=config.height,
+      width=config.width,
+      num_frames=config.num_frames,
+      num_inference_steps=config.num_inference_steps,
+      guidance_scale=config.guidance_scale,
+    )
+  print("generation time: ", (time.perf_counter() - s0))
+  for i in range(len(videos)):
+    export_to_video(videos[i], f"wan_output_{i}.mp4", fps=16)
+
+
+def main(argv: Sequence[str]) -> None:
+  pyconfig.initialize(argv)
+  run(pyconfig.config)
+
+
+if __name__ == "__main__":
+  app.run(main)

latents.npy

@@ -0,0 +1,32 @@
+import torch
+import torch
+from diffusers.utils import export_to_video
+from diffusers import AutoencoderKLWan, WanPipeline
+from diffusers.schedulers.scheduling_unipc_multistep import UniPCMultistepScheduler
+# import torch_xla.core.xla_model as xm
+# import torch_xla.runtime as xr
+
+import numpy as np
+model_id = "Wan-AI/Wan2.1-T2V-14B-Diffusers"
+vae = AutoencoderKLWan.from_pretrained(model_id, subfolder="vae", torch_dtype=torch.float32)
+pipe = WanPipeline.from_pretrained(model_id, vae=vae, torch_dtype=torch.bfloat16)
+flow_shift = 5.0  # 5.0 for 720P, 3.0 for 480P
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config, flow_shift=flow_shift)
+
+
+prompt = "A cat and a dog baking a cake together in a kitchen. The cat is carefully measuring flour, while the dog is stirring the batter with a wooden spoon. The kitchen is cozy, with sunlight streaming through the window."
+negative_prompt = "Bright tones, overexposed, static, blurred details, subtitles, style, works, paintings, images, static, overall gray, worst quality, low quality, JPEG compression residue, ugly, incomplete, extra fingers, poorly drawn hands, poorly drawn faces, deformed, disfigured, misshapen limbs, fused fingers, still picture, messy background, three legs, many people in the background, walking backwards"
+
+
+output = pipe(
+   prompt=prompt,
+   negative_prompt=negative_prompt,
+   height=480,
+   width=720,
+   num_frames=21,
+   guidance_scale=5.0,
+   num_inference_steps=10,
+).frames[0]
+
+
+export_to_video(output, "output.mp4", fps=16)

src/maxdiffusion/generate_wan_bsgt.py

@@ -404,6 +404,9 @@ def __call__(
           num_frames=num_frames,
           num_channels_latents=num_channel_latents
         )
+      
+      # import pdb
+      # pdb.set_trace()
 
       data_sharding = PositionalSharding(self.devices_array).replicate()
       if len(prompt) % jax.device_count() == 0: