Local tensor visualization beats Jupyter notebooks for quick diffusion latent inspection.
Machine learning engineers, data scientists debugging deep learning pipelines and tensor shapes
Jupyter Notebook · TensorBoard · Netron
Tensor Spy lets you quickly inspect the contents of numpy & pytorch tensors locally (your tensors are not uploaded to any servers).
This is useful to validate your deep learning data pipelines, to check which layers in your diverging model are actually going haywire, and just because it's kind of cool & a lot more convenient for one-off inspections than loading things up in python.
If you work with diffusion models, inspecting the latent space can be quite informative: you want some "noise" in there but it should probably be fairly smooth for your LDM (Latent Diffusion Model) to be able to target it well.
Also, if you haven't looked at your data, it's probably not what you think it is ;)
Basic stats are auto-computed, and any inf/nan values are both counted and rendered with contrasting colors, to help you quickly identify issue hotspots.
The site is free, and our broad intention is to keep it that way (we run a bunch of pro-bono little utility sites in addition to our commercial ones, they're all linked on the about page).
Would love to hear your thoughts, I'm sure there are some stats or utility features we missed, so please give it a spin and let us know!
---
Agentic coding is a brave new world. Three years ago, after the initial rush of ChatGPT's launch, I commented to some friends that "we're standing on the beach and the water just receded". The tsunami is really hitting now. As in: this project took about 2 weeks, and not only would we not have done it without agentic coding, it would have taken months using "traditional methods". With agentic coding, adding .pt/.pth support was basically a single request. And it just worked. Time to adapt yet again.
Takes micrograd's toy autodiff idea and scales it to real tensor operations with NumPy and optional CuPy GPU support. The repo exposes an nn-like module, common activations (ReLU/Softmax), self-attention and an MNIST MLP example, so it's excellent as a learning resource. It's explicitly flagged as WIP — expect performance and OOM rough edges, so treat it as a study tool rather than a training backbone.
AST rewriting for runtime types without code changes — why didn't this exist before.
Two MacBooks syncing gradients over Thunderbolt — slower than single-GPU but it works.
Best interactive transformer visualizer I've seen — finally explains tensor shapes clearly.
Deterministic fingerprinting for model structure without loading weights.
Pure Rust autodiff + GPU math avoids C++ FFI hell, but matmul claim needs apples-to-apples benchmarks.
