I will quantize and compress your pytorch model for faster inference

C
contib5
C
contib5
Sarthak S

About this gig

I make deep learning models smaller and faster. Send me your PyTorch model and I will quantize it (INT8 / low-bit), prune or fine-tune it to keep accuracy, and give you a clear report of size, speed and accuracy before and after, plus clean reproducible code. Good for edge devices, cheaper GPU inference, and FPGA deployment. Please message me before ordering with your model type, framework version, and target hardware.

Get to know Sarthak S

Sarthak S

ML Engineer, model compression and FPGA inference

  • FromSingapore
  • Member sinceNov 2026
  • Languages

    English
I build fast, efficient deep learning systems. I quantize and compress PyTorch models, turn neural networks into FPGA designs with Vivado, and run and debug large GPU training jobs. I can speed up your model, shrink it for edge devices, or get it running on hardware, with clear reports and reproducible code.