Picture for Rangharajan Venkatesan

Rangharajan Venkatesan

Enabling and Accelerating Dynamic Vision Transformer Inference for Real-Time Applications

Add code
Dec 06, 2022
Figure 1 for Enabling and Accelerating Dynamic Vision Transformer Inference for Real-Time Applications
Figure 2 for Enabling and Accelerating Dynamic Vision Transformer Inference for Real-Time Applications
Figure 3 for Enabling and Accelerating Dynamic Vision Transformer Inference for Real-Time Applications
Figure 4 for Enabling and Accelerating Dynamic Vision Transformer Inference for Real-Time Applications
Viaarxiv icon

Optimal Clipping and Magnitude-aware Differentiation for Improved Quantization-aware Training

Add code
Jun 13, 2022
Figure 1 for Optimal Clipping and Magnitude-aware Differentiation for Improved Quantization-aware Training
Figure 2 for Optimal Clipping and Magnitude-aware Differentiation for Improved Quantization-aware Training
Figure 3 for Optimal Clipping and Magnitude-aware Differentiation for Improved Quantization-aware Training
Figure 4 for Optimal Clipping and Magnitude-aware Differentiation for Improved Quantization-aware Training
Viaarxiv icon

Low-Precision Training in Logarithmic Number System using Multiplicative Weight Update

Add code
Jun 26, 2021
Figure 1 for Low-Precision Training in Logarithmic Number System using Multiplicative Weight Update
Figure 2 for Low-Precision Training in Logarithmic Number System using Multiplicative Weight Update
Figure 3 for Low-Precision Training in Logarithmic Number System using Multiplicative Weight Update
Figure 4 for Low-Precision Training in Logarithmic Number System using Multiplicative Weight Update
Viaarxiv icon

VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference

Add code
Feb 08, 2021
Figure 1 for VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
Figure 2 for VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
Figure 3 for VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
Figure 4 for VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference
Viaarxiv icon

SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networks

Add code
May 23, 2017
Figure 1 for SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networks
Figure 2 for SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networks
Figure 3 for SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networks
Figure 4 for SCNN: An Accelerator for Compressed-sparse Convolutional Neural Networks
Viaarxiv icon