triton-inference-server/hugectr_backend

Issues

• Unable to launch Triton Server with hps backend using latest HugeCTR and hugectr_backend repos

  #58 opened a year ago by sezhiyanhari
  4

• HugeCTR model on sagemaker endpoint with Hierarchical Parameter Serve

  #55 opened a year ago by andompesta
  1

• how to update model weight (triton_server) after continue training!!

  #44 opened 2 years ago by kimharo
  1

• How to turn off only the request log? (The request log is too verbose, resulting in poor throughput)

  #43 opened 2 years ago by kisow
  3

• [BUG] HugeCTR backend 3.4.1 has crashed

  #40 opened 2 years ago by kisow
  3

• hugectr memory pool is empty

  #38 opened 2 years ago by kimharo
  6

• how to put all embedding volume into gpu embedding cache?

  #39 opened 2 years ago by kimharo
  5

• [BUG] HugeCTR Backend server dies with SIGBUS during server startup (22.04)

  #42 opened 2 years ago by kisow
  1

• [BUG] big CATCOLUMN, ROWINDEX server crash

  #35 opened 3 years ago by kimharo
  3

• Inconsistency between hugeCTR v3.3.1 and hugectr_backend v3.3.1

  #31 opened 3 years ago by dfisk
  5

• [BUG]Load criteo model failed in triton inference

  #23 opened 3 years ago by lunachy
  3

• [BUG] HugeCTR backend crashed when I make pressure test with 100 concurrents，and the coredump file not generated.

  #27 opened 3 years ago by offbye
  5

• [BUG] support google cloud storage path in hugeCTR backend

  #21 opened 3 years ago by mengdong
  2

• The max_batch_size configuration not work in batching inference?

  #26 opened 3 years ago by offbye
  2

• inference error use triton http client.

  #24 opened 3 years ago by offbye
  1