# Error when using transform function of pixel\_values

**URL:** <https://discuss.huggingface.co/t/error-when-using-transform-function-of-pixel-values/45263>\
**Category:** Intermediate\
**Created:** [July 1, 2023, 3:23pm UTC](https://discuss.huggingface.co/t/error-when-using-transform-function-of-pixel-values/45263 "2023-07-01T15:23:13Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Andyrasika](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/andyrasika/32/31603_2.png) [@Andyrasika](https://discuss.huggingface.co/u/Andyrasika)\
**Post date:** [July 1, 2023, 3:23pm UTC](https://discuss.huggingface.co/t/error-when-using-transform-function-of-pixel-values/45263/1 "2023-07-01T15:23:13Z")

</div>

Hi,  
I am working on [Flickr 8k Dataset | Kaggle](https://www.kaggle.com/datasets/adityajn105/flickr8k) and  
have the following dataset:

```auto
image_dataset:
DatasetDict({
    train: Dataset({
        features: ['pixel_values', 'labels'],
        num_rows: 7281
    })
    test: Dataset({
        features: ['pixel_values', 'labels'],
        num_rows: 810
    })
})

normalize = Normalize(
    mean=feature_extractor.image_mean, 
    std=feature_extractor.image_std
)

vit_model = ViTModel.from_pretrained(config.ENCODER)
feature_extractor = AutoImageProcessor.from_pretrained(config.ENCODER)
tokenizer = AutoTokenizer.from_pretrained(config.DECODER)
tokenizer.pad_token = tokenizer.unk_token

_transforms = Compose([
    RandomResizedCrop(feature_extractor.size, scale=[0.8, 1]),
    RandomHorizontalFlip(),
    ToTensor(),
    normalize
])

```

when running the following command:  
plt.imshow(\_transforms(image\_dataset[‘train’][0][‘pixel\_values’]).detach().numpy().transpose(1, 2, 0))

```auto
i get error: 
 Traceback (most recent call last) ────────────────────────────────╮
│ /tmp/ipykernel_29/84940155.py:1 in <module> │
│ │
│ [Errno 2] No such file or directory: '/tmp/ipykernel_29/84940155.py' │
│ │
│ /opt/conda/lib/python3.10/site-packages/torchvision/transforms/transforms.py:95 in __call__ │
│ │
│ 92 │ │
│ 93 │ def __call__ (self, img): │
│ 94 │ │ for t in self.transforms: │
│ ❱ 95 │ │ │ img = t(img) │
│ 96 │ │ return img │
│ 97 │ │
│ 98 │ def __repr__ (self) -> str: │
│ │
│ /opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py:1501 in _call_impl │
│ │
│ 1498 │ │ if not (self._backward_hooks or self._backward_pre_hooks or self._forward_hooks │
│ 1499 │ │ │ │ or _global_backward_pre_hooks or _global_backward_hooks │
│ 1500 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│ ❱ 1501 │ │ │ return forward_call(*args, **kwargs) │
│ 1502 │ │ # Do not call functions when jit is used │
│ 1503 │ │ full_backward_hooks, non_full_backward_hooks = [], [] │
│ 1504 │ │ backward_pre_hooks = [] │
│ │
│ /opt/conda/lib/python3.10/site-packages/torchvision/transforms/transforms.py:980 in forward │
│ │
│ 977 │ │ │ PIL Image or Tensor: Randomly cropped and resized image. │
│ 978 │ │ """ │
│ 979 │ │ i, j, h, w = self.get_params(img, self.scale, self.ratio) │
│ ❱ 980 │ │ return F.resized_crop(img, i, j, h, w, self.size, self.interpolation, antialias= │
│ 981 │ │
│ 982 │ def __repr__ (self) -> str: │
│ 983 │ │ interpolate_str = self.interpolation.value │
│ │
│ /opt/conda/lib/python3.10/site-packages/torchvision/transforms/functional.py:663 in resized_crop │
│ │
│ 660 │ if not torch.jit.is_scripting() and not torch.jit.is_tracing(): │
│ 661 │ │ _log_api_usage_once(resized_crop) │
│ 662 │ img = crop(img, top, left, height, width) │
│ ❱ 663 │ img = resize(img, size, interpolation, antialias=antialias) │
│ 664 │ return img │
│ 665 │
│ 666 │
│ │
│ /opt/conda/lib/python3.10/site-packages/torchvision/transforms/functional.py:479 in resize │
│ │
│ 476 │ _, image_height, image_width = get_dimensions(img) │
│ 477 │ if isinstance(size, int): │
│ 478 │ │ size = [size] │
│ ❱ 479 │ output_size = _compute_resized_output_size((image_height, image_width), size, max_si │
│ 480 │ │
│ 481 │ if (image_height, image_width) == output_size: │
│ 482 │ │ return img │
│ │
│ /opt/conda/lib/python3.10/site-packages/torchvision/transforms/functional.py:387 in │
│ _compute_resized_output_size │
│ │
│ 384 │ │ │
│ 385 │ │ new_w, new_h = (new_short, new_long) if w <= h else (new_long, new_short) │
│ 386 │ else: # specified both h and w │
│ ❱ 387 │ │ new_w, new_h = size[1], size[0] │
│ 388 │ return [new_h, new_w] │
│ 389 │
│ 390 │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
KeyError: 1

```

how to resolve this issue.  
i even tried: [Image classification using LoRA](https://huggingface.co/docs/peft/task_guides/image_classification_lora) with no success  
Please let me know if you need any further information  
Thanks,  
Ankush Singal