nightingale01

.Jsonl metadata not detected

5

### Describe the bug Hi I have the following directory structure: |--dataset | |-- images | |-- metadata1000.csv | |-- metadata1000.jsonl | |-- padded_images Example of metadata1000.jsonl file {"caption": "a...

Chunk Size for Einsum and Einsumv2 attention vs Resolution

2

Hi the Chunk size in python_coreml_stable_diffusion/attention.py is 512 only for 512x 512? How to change it for arbitrary resolution? I know its based on the sequence length but any reason...

@LowerComplex.register_lower_func(op_type="complex_real") def _lower_complex_real(op: Operation): complex_input: ComplexVar = op.data # Use an identity op to avoid the block's input name inconsistency issue. If we directly use # complex_input.real, the var's name...

unsqueeze or expand dims not supported for complex64,

2

@register_torch_op(torch_alias=["unsqueeze_copy"]) def unsqueeze(context, node): inputs = _get_inputs(context, node, expected=2) unsqueeze = mb.expand_dims(x=inputs[0], axes=[inputs[1].val], name=node.name) context.add(unsqueeze) also ValueError: Op "1298" (op_type: expand_dims) Input x="freqs_cis.1" expects tensor or scalar of dtype from...

RoPE implemetation, Torch.polar op not available in MIL requesting support

2

## 🌱 Describe your Feature Request - having support for converting torch.polar to MIL will allow using RoPE embeddings which is fairly common now days currently get this issue:RuntimeError: PyTorch...

Batched Text to image generation really slow with pipeline

7

### Describe the bug batched diffusers pipeline inference is really slow ### Reproduction ``` class prompt_data(Dataset): def __init__(self,prompt_file): self.txt_prompts = [] with open(prompt_file,'r') as file:self.txt_prompts = file.readlines() def __len__(self): return...

chunk unet does not work with coremltool>7.1

1

chunk unet does not work with coremltool>7.1 due to missing CacheDoublyLinkedList attribute index

Quantized Models Chunking into unequal sizes

4

## 🐞Describing the bug with reference to this issue https://github.com/apple/ml-stable-diffusion/issues/353, i used the bisect_model() function to split a quantized model into 2 chunks, i tried with 7.1 and 7.0 with...

torch.nn.functional.grid_sample for volumetric input/ Rank 5 input and grid

3

## 🌱 Describe your Feature Request support torch.nn.function.grid_sample for 5d Volumetric input by supporting 5d input for `mb.affine and mb.resample` for rank >=4 :https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html ## How can this feature be...

unet Chunking not supported in coremltools>7.1 and <7.1 does not support dynamic reshape

1

## 🐞Describing the bug using the torch2coreml script with the --chunk-unet option would result in the following error: AttributeError: 'CacheDoublyLinkedList' object has no attribute 'index'. This error occurred because the...