add: pytest log installation

[sayakpaul]

What does this PR do?

Add the pytest-reportlog dependency. Otherwise, https://github.com/huggingface/diffusers/actions/runs/8273408523/job/22637123355 happens.

What did I test?

Created a simple yet reasonable test suite:

# coding=utf-8
# Copyright 2024 HuggingFace Inc.
#
# 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.
import gc
import unittest

import numpy as np
import torch

from diffusers import (
    DiffusionPipeline,
    LCMScheduler,
)
from diffusers.utils.import_utils import is_accelerate_available, is_peft_available
from diffusers.utils.testing_utils import (
    load_image,
    numpy_cosine_similarity_distance,
    require_torch_gpu,
    slow,
)


if is_accelerate_available():
    from accelerate.utils import release_memory

if is_peft_available():
    pass


@slow
@require_torch_gpu
class LoraSDXLIntegrationTests(unittest.TestCase):
    def tearDown(self):
        super().tearDown()
        gc.collect()
        torch.cuda.empty_cache()

    def test_sdxl_1_0_lora(self):
        generator = torch.Generator("cpu").manual_seed(0)

        pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0")
        pipe.enable_model_cpu_offload()
        lora_model_id = "hf-internal-testing/sdxl-1.0-lora"
        lora_filename = "sd_xl_offset_example-lora_1.0.safetensors"
        pipe.load_lora_weights(lora_model_id, weight_name=lora_filename)

        images = pipe(
            "masterpiece, best quality, mountain", output_type="np", generator=generator, num_inference_steps=2
        ).images

        images = images[0, -3:, -3:, -1].flatten()
        expected = np.array([0.4468, 0.4087, 0.4134, 0.366, 0.3202, 0.3505, 0.3786, 0.387, 0.3535])

        max_diff = numpy_cosine_similarity_distance(images, expected)
        assert max_diff < 1e-4  
        
        pipe.unload_lora_weights()

        release_memory(pipe)

    def test_sdxl_lcm_lora(self):
        pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16)
        pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
        pipe.enable_model_cpu_offload()

        generator = torch.Generator("cpu").manual_seed(0)

        lora_model_id = "latent-consistency/lcm-lora-sdxl"

        pipe.load_lora_weights(lora_model_id)

        image = pipe(
            "masterpiece, best quality, mountain", generator=generator, num_inference_steps=4, guidance_scale=0.5
        ).images[0]

        expected_image = load_image(
            "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/lcm_lora/sdxl_lcm_lora.png"
        )

        image_np = pipe.image_processor.pil_to_numpy(image)
        expected_image_np = pipe.image_processor.pil_to_numpy(expected_image)

        max_diff = numpy_cosine_similarity_distance(image_np.flatten(), expected_image_np.flatten())
        assert np.allclose(image_np.flatten(), expected_image_np.flatten())  # Purposefully making it fail.

        pipe.unload_lora_weights()

        release_memory(pipe)

Ran the test (from the DGX):

RUN_SLOW=1 CUDA_VISIBLE_DEVICES=1 python -m pytest -n 1 --max-worker-restart=0 --dist=loadfile \
    --make-reports=tests_torch_cuda \
    --report-log=torch_cuda.log \
    tests/lora/test_lora_layers_peft_shorter.py::LoraSDXLIntegrationTests

Exported GITHUB_RUN_ID to an arbitrary value so that scripts/log_reports.py can run without fail.

Then ran python scripts/log_reports.py. It prints:

### *torch_cuda.log: 1 failed test*


+----+-----------------+--------------+--------------------+
|    | Test Location   | Test Case    | Test Name          |
+====+=================+==============+====================+
|  0 | tests/lora/t    | LoraSDXLInte | test_sdxl_lcm_lora |
|    | est_lora_lay    | grationTests |                    |
|    | ers_peft_sho    |              |                    |
|    | rter.py         |              |                    |
+----+-----------------+--------------+--------------------+

[{'type': 'header', 'text': {'type': 'plain_text', 'text': '🤗 Results of the Diffusers scheduled nightly tests.'}}, {'type': 'section', 'text': {'type': 'mrkdwn', 'text': '*torch_cuda.log: 1 failed test*\n\n```\n+----+-----------------+--------------+--------------------+\n|    | Test Location   | Test Case    | Test Name          |\n+====+=================+==============+====================+\n|  0 | tests/lora/t    | LoraSDXLInte | test_sdxl_lcm_lora |\n|    | est_lora_lay    | grationTests |                    |\n|    | ers_peft_sho    |              |                    |\n|    | rter.py         |              |                    |\n+----+-----------------+--------------+--------------------+\n```'}}, {'type': 'section', 'text': {'type': 'mrkdwn', 'text': '*For more details:*'}, 'accessory': {'type': 'button', 'text': {'type': 'plain_text', 'text': 'Check Action results', 'emoji': True}, 'url': 'https://github.com/huggingface/diffusers/actions/runs/sayak'}}, {'type': 'context', 'elements': [{'type': 'plain_text', 'text': 'Nightly test results for 2024-03-14'}]}]

So, that confirms things.

[sayakpaul]

Merging because the PR is just about adding a dependency to a workflow file.