f florianleibert Joined 15 years ago Karma 318 Submissions 105 Member since Apr 11 View on Hacker News
5.6x throughput on Kimi K2.6 by speculating less h huggingface.co 11 florianleibert 6 months ago 2 comments
Uber Drives Distributed Deep Learning Forward m mesosphere.com 1 florianleibert 9 years ago 0 comments
Training Your Custom Model with the DC/OS TensorFlow Package m mesosphere.com 1 florianleibert 9 years ago 0 comments
Distributed TensorFlow on Any Infrastructure with Mesosphere DC/OS m mesosphere.com 2 florianleibert 9 years ago 0 comments
Mesosphere DC/OS Brings Large-Scale Real-Time Processing to Geospatial Data t thenewstack.io 3 florianleibert 9 years ago 0 comments
Distributed TensorFlow on Any Infrastructure with Mesosphere DC/OS m mesosphere.com 8 florianleibert 9 years ago 1 comments
Containers and Distributed Systems: Where They Came from and Where They’re Going m mesosphere.com 288 florianleibert 9 years ago 41 comments
Continuous Delivery for DC/OS with Spinnaker e engineering.cerner.com 68 florianleibert 9 years ago 4 comments
Scaling is linear @15.8 tok/s per slot latency is constant. REpo has command launcher, Dockerfile, benchmark tool. Known limitations: BF16 KV only (FP8 crashes due to an AITER 384-expert constraint)