llama.cpp
sycl : implementation of reordered Q4_0 MMVQ for Intel GPUs
#12858
Merged
Overview
Commits
17
Changes
View On GitHub
Commits
  • sycl : Implemented reorder Q4_0 mmvq
    Alberto Cabrera committed 1 year ago
  • sycl : Fixed mmvq being called when reorder is disabled
    Alberto Cabrera committed 1 year ago
  • sycl : Improved comments in the quants header
    Alberto Cabrera committed 1 year ago
  • Use static_assert
    Rbiessy committed 1 year ago
  • safe_div -> ceil_div
    Rbiessy committed 1 year ago
  • Clarify qi comment
    Rbiessy committed 1 year ago
  • change the reorder tensor from init to execute OP
    NeoZhangJianyu committed 1 year ago
  • dbg
    Rbiessy committed 1 year ago
  • Merge remote-tracking branch 'neozhang/jianyuzh_fix_reorder' into Alcpz/mmvq_q4_0_reorder
    Alberto Cabrera committed 1 year ago
  • Undo changes to test-backend-ops
    Alberto Cabrera committed 1 year ago
  • Refactor changes on top of q4_0 reorder fix
    Alberto Cabrera committed 1 year ago
  • Merge remote-tracking branch 'upstream/master' into Alcpz/mmvq_q4_0_reorder
    Alberto Cabrera committed 1 year ago
  • Missing Reverts
    Alberto Cabrera committed 1 year ago
  • Refactored opt_for_reorder logic to simplify code path
    Alberto Cabrera committed 1 year ago
  • Merge branch 'master' into Alcpz/mmvq_q4_0_reorder
    GitHub committed 1 year ago
  • Explicit inlining and unroll
    Alberto Cabrera committed 1 year ago
  • Renamed mul_mat_algo enum for consistency
    Alberto Cabrera committed 1 year ago
FAQ Terms Privacy Refunds Impressum
Loading