AI/diffusers
1
0
Fork 0
mirror of https://github.com/huggingface/diffusers.git synced 2026-01-27 17:22:53 +03:00
Code Activity

enable 2 test cases on XPU (#11332)

Browse Source 
* enable 2 test cases on XPU

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* Apply style fixes

---------

Signed-off-by: YAO Matrix <matrix.yao@intel.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
This commit is contained in:
Yao Matrix
2025-04-18 07:27:41 +08:00
committed by GitHub
parent ee6ad51d96
commit eef3d65954
2 changed files with 13 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/quantization/bnb/test_mixed_int8.py
View File

@@ -523,13 +523,15 @@ class SlowBnb8bitTests(Base8bitTests):
torch_dtype=torch.float16,
device_map=torch_device,
)
# CUDA device placement works.
device = torch_device if torch_device != "rocm" else "cuda"
pipeline_8bit = DiffusionPipeline.from_pretrained(
self.model_name,
transformer=transformer_8bit,
text_encoder_3=text_encoder_3_8bit,
torch_dtype=torch.float16,
).to("cuda")
).to(device)
# Check if inference works.
_ = pipeline_8bit("table", max_sequence_length=20, num_inference_steps=2)

14
tests/quantization/utils.py
View File

@@ -1,4 +1,10 @@
from diffusers.utils import is_torch_available
from diffusers.utils.testing_utils import (
backend_empty_cache,
backend_max_memory_allocated,
backend_reset_peak_memory_stats,
torch_device,
)
if is_torch_available():
@@ -30,9 +36,9 @@ if is_torch_available():
@torch.no_grad()
@torch.inference_mode()
def get_memory_consumption_stat(model, inputs):
torch.cuda.reset_peak_memory_stats()
torch.cuda.empty_cache()
backend_reset_peak_memory_stats(torch_device)
backend_empty_cache(torch_device)
model(**inputs)
max_memory_mem_allocated = torch.cuda.max_memory_allocated()
return max_memory_mem_allocated
max_mem_allocated = backend_max_memory_allocated(torch_device)
return max_mem_allocated