DeepSpeed
Increasing the token-length based on available memory for GPT models
#2280
Merged
Overview
Commits
14
Changes
View On GitHub
Commits
  • increasing the token-length based on available memory & reduce memory alloc
    Reza Yazdani committed 3 years ago
  • Merge branch 'master' of github.com:microsoft/DeepSpeed into ds-inference/support-large-token-length
    Reza Yazdani committed 3 years ago
  • merging
    Reza Yazdani committed 3 years ago
  • Merge branch 'cholmes/fix-long-seq-len-inference' of github.com:microsoft/DeepSpeed into ds-inference/support-large-token-length
    Reza Yazdani committed 3 years ago
  • formating
    Reza Yazdani committed 3 years ago
  • fix compile issue
    Reza Yazdani committed 3 years ago
  • Merge branch 'cholmes/fix-long-seq-len-inference' into ds-inference/support-large-token-length
    RezaYazdaniAminabadi committed 3 years ago
  • Merge branch 'cholmes/fix-long-seq-len-inference' into ds-inference/support-large-token-length
    RezaYazdaniAminabadi committed 3 years ago
  • fix the max_out_tokens to use a dynamic range based on available memory
    Reza Yazdani committed 3 years ago
  • fix the issue with empty prompt
    Reza Yazdani committed 3 years ago
  • Merge branch 'cholmes/fix-long-seq-len-inference' into ds-inference/support-large-token-length
    RezaYazdaniAminabadi committed 3 years ago
  • fix residual-add
    Reza Yazdani committed 3 years ago
  • fix some issues with unit tests
    Reza Yazdani committed 3 years ago
  • fix formatting
    Reza Yazdani committed 3 years ago
FAQ Terms Privacy Refunds Impressum
Loading