Draft: Resnet support added #246
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
joein
reviewed
May 22, 2024
| }, | ||
| { | ||
| "model": "AndrewOgn/resnet_onnx", | ||
| "dim": 2048, |
There was a problem hiding this comment.
just as a reminder: we might want to inspect other resnet models to have lower dimensionality
Shapes matching for Resnet50-onnx Example of Resnet50 to onnx conversion (basic)
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
…de default Fixed test accordingly
joein
reviewed
May 23, 2024
joein
reviewed
May 23, 2024
| @@ -59,6 +59,9 @@ def __init__(self, scale: float = 1 / 255): | |||
| def __call__(self, images: List[np.ndarray]) -> List[np.ndarray]: | |||
| return [rescale(image, scale=self.scale) for image in images] | |||
|
|
|||
| class PILtoNDarray: | |||
| def __call__(self, images: List[Union[Image.Image, np.ndarray]]) -> List[np.ndarray]: | |||
| return [np.asarray(image).swapaxes(2, 0) if isinstance(image, Image.Image) else image for image in images] | |||
There was a problem hiding this comment.
it seems to be
(H, W, C) -> (C, W, H)
but should be (H, W, C) -> (C, H, W)
so we need to use transpose((2, 0, 1)) instead of swapaxes, should not we?
There was a problem hiding this comment.
But it's literally the same thing, isn't it?
There was a problem hiding this comment.
a = np.random.random((3,4,5))
b = a.swapaxes(2, 0)
c = a.transpose((2, 0, 1))
print(a.shape, b.shape, c.shape)
>>> ((3, 4, 5), (5, 4, 3), (5, 3, 4))
joein
reviewed
May 23, 2024
added 3 commits
May 23, 2024 15:10
Resize logic
# Conflicts: # fastembed/image/onnx_image_model.py # fastembed/image/transform/operators.py
joein
reviewed
May 31, 2024
joein
reviewed
May 31, 2024
joein
reviewed
May 31, 2024
joein
reviewed
May 31, 2024
Comment on lines
99
to
134
| transforms = [] | ||
| cls._get_convert_to_rgb(transforms, config) | ||
| cls._get_resize(transforms, config) | ||
| cls._get_center_crop(transforms, config) | ||
| cls._get_pil2ndarray(transforms, config) | ||
| cls._get_rescale(transforms, config) | ||
| cls._get_normalize(transforms, config) | ||
| return cls(transforms=transforms) | ||
|
|
||
| @staticmethod | ||
| def _get_convert_to_rgb(transforms: List['Transform'], config: Dict[str, Any]): | ||
| transforms.append(ConvertToRGB()) | ||
|
|
||
| @staticmethod | ||
| def _get_resize(transforms: List['Transform'], config: Dict[str, Any]): | ||
| mode = config.get('image_processor_type', 'CLIPImageProcessor') | ||
| if mode == 'CLIPImageProcessor': | ||
| if config.get("do_resize", False): | ||
| size = config["size"] | ||
| if "shortest_edge" in size: | ||
| size = size["shortest_edge"] | ||
| elif "height" in size and "width" in size: | ||
| size = (size["height"], size["width"]) | ||
| else: | ||
| raise ValueError("Size must contain either 'shortest_edge' or 'height' and 'width'.") | ||
| transforms.append(ClipResize(size=size, resample=config.get("resample", Image.Resampling.BICUBIC))) | ||
| elif mode == 'ConvNextFeatureExtractor': | ||
| if 'size' in config and "shortest_edge" not in config['size']: | ||
| raise ValueError(f"Size dictionary must contain 'shortest_edge' key. Got {config['size'].keys()}") | ||
| shortest_edge = config['size']["shortest_edge"] | ||
| crop_pct = config.get("crop_pct", 0.875) | ||
| if shortest_edge < 384: | ||
| # maintain same ratio, resizing shortest edge to shortest_edge/crop_pct | ||
| resize_shortest_edge = int(shortest_edge / crop_pct) | ||
| transforms.append(ClipResize(size=resize_shortest_edge, resample=config.get("resample", Image.Resampling.BICUBIC))) | ||
| transforms.append(CenterCrop(size=(shortest_edge, shortest_edge))) |
There was a problem hiding this comment.
it's a temporary design, we would need to come up with something better, but since it is hidden from the users, it's okay to have it this way for now
|
The last thing to do at the moment is to push the model to Qdrant's HF hub and change the corresponding name in image/onnx_embedding.py |
joein
approved these changes
May 31, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Added support of resnet-50, classical CNN