Features
Stacks, adaptive truncation, precision, GPUs, large problems and logging.
Stacks
Contract and compress several trains in a single sweep.
Adaptive truncation
Trim bonds below chi_out with a relative singular-value cutoff.
Precision and reproducibility
Data types of sketches and results, and seeding.
GPU execution
Run the sweep on NVIDIA or AMD GPUs through CuPy.
Large problems
Run stacks that fit neither on the GPU nor in host memory.
Logging
Turn on the library's progress and timing messages.