Oracle offers GraphPipe spec for machine learning data transmission

(Foto: Istock)

Oracle offers GraphPipe spec for machine learning data transmission

IDG NEWS: GraphPipe is intended to bring the efficiency of a binary, memory-mapped format while being simple and light on dependencies

Vil du fortsette å lese, velg et av alternativene nedenfor

  • Logg inn!

    Du har abonnement og er registrert som bruker.

  • Har abonnement!

    Du har abonnement, men ikke registrert deg.

  • Bestill abonnement!

    Digital tilgang er inkludert i alle våre abonnement.

Oracle has developed an open source specification for transmitting tensor data, which the company wants to become a standard for machine learning.

for machine and data from on being and servers also transmission. querying dependencies. a intended provides learning are the GraphPipe any and specification to framework. Called clients efficiency format network protocol for the light of models There is a deploying binary, memory-mapped simple GraphPipe, bring while

includes: It

similar deserialization. are input of output and avoiding A names. benefit names, during additional definitions. of a definitions buffers, to that protocol with Flatbuffer tensors, copy includes request Google Flatbuffers provide set message an - memory input, flatbuffer

- serving Guidelines models. for

machine from frameworks. - models serving learning Examples various of

graph. Google’s for for Python, Clients a - plugin model are through available inside a library, TensorFlow TensorFlow GraphPipe. Go, querying Client local Java. served including remote libraries for and There’s models for a

tensor provides about With a and output per one model a and message metadata name. shapes The inputs and of GraphPipe, also outputs.  model types remote request accepts returns

Oracle says GraphPipe …

IDG News Service