tarekziade HF Staff
Publish MIC usage, slim fixture and reviewed Transformers expectations
8895415 verified | # Swinv2: image classification with Transformers | |
| This page is the Transformers contract of the Swinv2 MIC and also a usage example. [doc-builder](https://github.com/huggingface/doc-builder/blob/main/docs/runnable-code-blocks.md) renders the `runnable` blocks below as documentation, and its pytest plugin runs them as written against `microsoft/swinv2-tiny-patch4-window16-256`. The MIC runner runs the same blocks against each slim fixture and each checkpoint in [`integration.json`](../../integration.json). Lines marked `# doc-builder: hide` only matter there: they swap in the pinned checkpoint or variant and print the result record. | |
| ## Classify synthetic images | |
| ```py runnable:classify | |
| import torch | |
| from transformers import AutoImageProcessor, AutoModelForImageClassification, set_seed | |
| import json # doc-builder: hide | |
| import os # doc-builder: hide | |
| model_id = "microsoft/swinv2-tiny-patch4-window16-256" | |
| model_id = os.environ.get("MIC_MODEL", model_id) # doc-builder: hide | |
| variant = json.loads(os.environ.get("MIC_MODEL_KWARGS", "{}")) # doc-builder: hide | |
| device = "cuda" if torch.cuda.is_available() else "cpu" | |
| set_seed(0) # seeds torch, NumPy, and Python: loading, dropout, and data augmentation draw random numbers | |
| processor = AutoImageProcessor.from_pretrained(model_id) | |
| model = AutoModelForImageClassification.from_pretrained(model_id).to(device) | |
| if variant: # doc-builder: hide | |
| del model # doc-builder: hide | |
| model = AutoModelForImageClassification.from_pretrained(model_id, **variant).to(device) # doc-builder: hide | |
| # Two textured images with different colours: global pooling averages a texture | |
| # away, so random noise alone looks the same to the classifier; colour does not. | |
| tint = torch.tensor([[0.9, 0.2, 0.1], [0.1, 0.3, 0.9]]).view(2, 3, 1, 1) | |
| # uint8 images: the processor rescales by 1/255, which would flatten floats in [0, 1]. | |
| images = list(((0.7 * tint + 0.3 * torch.rand(2, 3, 256, 256)).clamp(0, 1) * 255).byte()) | |
| inputs = processor(images=images, return_tensors="pt") | |
| inputs = inputs.to(device, dtype=model.dtype) | |
| with torch.no_grad(): | |
| logits = model(**inputs).logits | |
| predicted = logits.argmax(dim=-1) | |
| print([model.config.id2label[i] for i in predicted.tolist()]) | |
| ``` | |
| ## One training step | |
| ```py runnable:classify:2 | |
| labels = torch.tensor([0, 1], device=device) | |
| model.train() | |
| optimizer = torch.optim.SGD(model.parameters(), lr=0.01) | |
| loss = model(**inputs, labels=labels).loss | |
| loss.backward() | |
| before = {n: p.detach().cpu().clone() for n, p in model.named_parameters() if p.requires_grad} # doc-builder: hide | |
| optimizer.step() | |
| print(f"loss: {loss.item():.4f}") | |
| if os.environ.get("MIC_RUNNER"): # doc-builder: hide | |
| named = dict(model.named_parameters()) | |
| checked = ("classifier.weight", "swinv2.encoder.layers.0.blocks.0.attention.self.query.weight") | |
| result = { | |
| "runtime": { | |
| "device": model.device.type, | |
| "dtype": str(model.dtype).removeprefix("torch."), | |
| "attn_implementation": model.config._attn_implementation, | |
| }, | |
| "predicted_classes": predicted.tolist(), | |
| "logits": {"probe": logits[:, :5].float().tolist()}, | |
| "training": { | |
| "loss": loss.item(), | |
| "gradients": {n: bool(named[n].grad is not None and torch.count_nonzero(named[n].grad) > 0 and torch.isfinite(named[n].grad).all()) for n in checked}, | |
| "updated": any(not torch.equal(p.detach().cpu(), before[n]) for n, p in model.named_parameters() if n in before), | |
| }, | |
| } | |
| print(json.dumps(result)) | |
| ``` | |