We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Successfully optimizing LSTM and CONV networks for execution on the CPU with reasonable scalar & SIMD performance.