Neural Network Compression: Difference between revisions

Neural Network Compression (view source)

1 byte removed , 3 February 2021

5,351

edits

Revision as of 22:09, 3 February 2021 (view source) David (talk \| contribs) No edit summary ← Older edit		Revision as of 22:15, 3 February 2021 (view source) David (talk \| contribs) (→‎Quantization) Newer edit →
Line 32:		Line 32:

	* Google uses [https://cloud.google.com/blog/products/ai-machine-learning/bfloat16-the-secret-to-high-performance-on-cloud-tpus bfloat16] for training on TPUs.		* Google uses [https://cloud.google.com/blog/products/ai-machine-learning/bfloat16-the-secret-to-high-performance-on-cloud-tpus bfloat16] for training on TPUs.


	==Factorization==		==Factorization==