On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,181 papers · 148 categories

Trend · papers per month

count % of corpus

3 results for “TensorQuant”

TensorQuant - A Simulation Toolbox for Deep Neural Network Quantizationcs.CV

TensorQuant simulates quantization of DNNs for efficient computation.

problem Optimal quantization method and bit number for DNNs are topology-dependent.

method TensorQuant simulates quantization of existing DNN topologies.

result TensorQuant supports experimental evaluation of quantization impact.

quantization deep neural networks TensorFlow fixed-point simulation

2017-10-13abs ↗pdf ↗

Sparsity in Deep Neural Networks - An Empirical Investigation with TensorQuantcs.LG

The paper investigates sparsity in deep neural networks, especially in larger models.

problem Reducing resource demands in deep neural networks, particularly in larger models.

method Extended TensorQuant toolbox to investigate sparsity in deeper models and various classification problem sizes.

result Promoted sparsity in deeper models, showing differences in sparsity for activations, weights, and gradients.

sparsity deep neural networks TensorQuant resource reduction embedded systems

2018-08-27abs ↗pdf ↗

High-Accuracy Low-Precision Trainingcs.LG

A new method reduces training time and improves accuracy in machine learning models.

problem The tradeoff between training accuracy and precision in low-precision machine learning.

method A low-precision stochastic gradient descent variant called HALP, using SVRG to reduce gradient variance and bit centering to reduce quantization error.

result HALP converges at the same theoretical rate as full-precision algorithms despite quantization noise.

machine learning low precision training SVRG quantization

2018-03-09abs ↗pdf ↗