MrYxJ/calculate-flops.pytorch
A PyTorch-based tool that calculates theoretical FLOPs, MACs, and parameter counts for neural networks including transformers and LLMs.
Not currently ranked — collecting fresh signals.
star history
Calflops computes floating-point operations, multiply-add operations, and parameter counts for PyTorch neural network models. It supports CNNs, RNNs, GCNs, transformers, and custom layers via torch.nn.function. It also provides a HuggingFace Space interface and a HuggingFace integration for directly analyzing models by name without downloading weights.
Frequently asked
- What is MrYxJ/calculate-flops.pytorch?
- A PyTorch-based tool that calculates theoretical FLOPs, MACs, and parameter counts for neural networks including transformers and LLMs.
- Is calculate-flops.pytorch open source?
- Yes — MrYxJ/calculate-flops.pytorch is open source, released under the MIT license.
- What language is calculate-flops.pytorch written in?
- MrYxJ/calculate-flops.pytorch is primarily written in Python.
- How popular is calculate-flops.pytorch?
- MrYxJ/calculate-flops.pytorch has 944 stars on GitHub.
- Where can I find calculate-flops.pytorch?
- MrYxJ/calculate-flops.pytorch is on GitHub at https://github.com/MrYxJ/calculate-flops.pytorch.