You are working on a niche product in the image recognition domain. Your team has developed a model that is dominated by custom C++ TensorFlow ops your team has implemented. These ops are used inside your main training loop and are performing bulky matrix multiplications. It currently takes up to several days to train a model. You want to decrease this time significantly and keep the cost low by using an accelerator on Google Cloud. What should you do?
dhs227
Highly Voted 4 years agoffggrre
6 months agoHelinia
3 months, 3 weeks agoaiguy
Highly Voted 4 years agogopinath_k
3 years, 1 month agotavva_prudhvi
2 years agocetanx
10 months, 4 weeks agoFP77
8 months, 1 week agosquishy_fishy
5 months, 4 weeks agoCGS22
Most Recent 2 weeks, 5 days agoPreetmehta1234
1 month, 4 weeks agoMatt_108
3 months, 1 week agoKimich
4 months, 3 weeks agoemmylou
5 months agoDataFrame
5 months, 1 week agoNirca
6 months, 2 weeks agokumarts
6 months, 3 weeks agoIrisXia
8 months, 2 weeks agoKC_go_reply
9 months, 1 week agoZZHZZH
9 months, 2 weeks agoQix
9 months, 2 weeks agolucaluca1982
1 year agolucaluca1982
1 year agomidgoo
1 year, 1 month ago