DeepSpeed
Workflow for AutoTP
#4961
Open
Overview
Commits
36
Changes
View On GitHub
Commits
  • update trigger condition
    delock committed 2 years ago
  • change data type to bfloat16
    delock committed 2 years ago
  • add probe of runner memory capacity
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 2 years ago
  • Merge branch 'up-master' into gma/add_autotp_workflow
    delock committed 2 years ago
  • skip numa membind when this capability is not available
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    tjruwase committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    delock committed 2 years ago
  • remove TRANSFORMERS_CACHE_OVERRIDING
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    tjruwase committed 2 years ago
  • Use official DeepSpeedExamples
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 2 years ago
  • add new models in autotp workflow and enable gptj, baichuan
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 2 years ago
  • merge autotp and cpu-inference workflow
    delock committed 2 years ago
  • more strict /blob maching
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 2 years ago
  • Enable autotp test for Baichuan
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    delock committed 2 years ago
  • put oneCCL binding for PyTorch into CPU-inference workflow
    delock committed 2 years ago
  • trust remote code and don't use meta tensor
    delock committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    tjruwase committed 2 years ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
  • remove unit test from cpu_inference workflow
    delock committed 1 year ago
  • Use stock PyTorch to test AutoTP
    delock committed 1 year ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
  • add env HF_HOME to point to /blob/hf_home
    delock committed 1 year ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
  • unset TRANSFORMERS_CACHE and peek tree structure
    delock committed 1 year ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
  • disable baichuan autotp test
    delock committed 1 year ago
  • Merge branch 'master' into gma/add_autotp_workflow
    loadams committed 1 year ago
FAQ Terms Privacy Refunds Impressum
Loading