text-generation-inference
Improve vlm support (add idefics3 support)
#2437
Merged
Overview
Commits
17
Changes
View On GitHub
Commits
  • feat: expand vlm support and add image token logic and tests
    drbh committed 1 year ago
  • fix: avoid unused perceiver config
    drbh committed 1 year ago
  • feat: integrate image tokens into inputs embeds
    drbh committed 1 year ago
  • feat: add simple idefics3 test
    drbh committed 1 year ago
  • feat: update docs, image token logic and weight names
    drbh committed 1 year ago
  • fix: improve image processing
    drbh committed 1 year ago
  • feat: improve prefix for idefics3
    drbh committed 1 year ago
  • fix: bump idefics3 tests and snapshots
    drbh committed 1 year ago
  • fix: improve text model loading
    drbh committed 1 year ago
  • feat: consolidate changes with existing vlms and add support and test for smolvlm
    drbh committed 1 year ago
  • fix: create new idefic3 file, simplify logic and adjust llama weight loading
    drbh committed 1 year ago
  • fix: lint with ruff
    drbh committed 1 year ago
  • fix: clean up idefics 3 and improve prefix handling
    drbh committed 350 days ago
  • fix: improve typing
    drbh committed 350 days ago
  • fix: improve prompt_split_image with ref to original impl
    drbh committed 350 days ago
  • fix: adjust ruff lints and small refactors
    drbh committed 350 days ago
  • fix: adjust FlashLlamaModel prefix logic
    drbh committed 349 days ago
FAQ Terms Privacy Refunds Impressum
Loading