Skip to content

[idefics] idefics-9b test use 4bit quant #25734

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Aug 24, 2023
Merged

[idefics] idefics-9b test use 4bit quant #25734

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 18 additions & 3 deletions tests/models/idefics/test_modeling_idefics.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,8 +16,15 @@

import unittest

from transformers import IdeficsConfig, is_torch_available, is_vision_available
from transformers.testing_utils import TestCasePlus, require_torch, require_vision, slow, torch_device
from transformers import BitsAndBytesConfig, IdeficsConfig, is_torch_available, is_vision_available
from transformers.testing_utils import (
TestCasePlus,
require_bitsandbytes,
require_torch,
require_vision,
slow,
torch_device,
)
from transformers.utils import cached_property

from ...test_configuration_common import ConfigTester
Expand Down Expand Up @@ -434,6 +441,7 @@ class IdeficsModelIntegrationTest(TestCasePlus):
def default_processor(self):
return IdeficsProcessor.from_pretrained("HuggingFaceM4/idefics-9b") if is_vision_available() else None

@require_bitsandbytes
@slow
def test_inference_natural_language_visual_reasoning(self):
cat_image_path = self.tests_dir / "fixtures/tests_samples/COCO/000000039769.png"
Expand All @@ -459,7 +467,14 @@ def test_inference_natural_language_visual_reasoning(self):
],
]

model = IdeficsForVisionText2Text.from_pretrained("HuggingFaceM4/idefics-9b").to(torch_device)
# the CI gpu is small so using quantization to fit
quantization_config = BitsAndBytesConfig(
load_in_4bit=True,
bnb_4bit_compute_dtype="float16",
)
model = IdeficsForVisionText2Text.from_pretrained(
"HuggingFaceM4/idefics-9b", quantization_config=quantization_config, device_map="auto"
)
processor = self.default_processor
inputs = processor(prompts, return_tensors="pt").to(torch_device)
generated_ids = model.generate(**inputs, max_length=100)
Expand Down