v0.5.0: Better Unicode Character Handling, Better AutoClass

What's Changed

• Update autoclass to remove the transformer version constraints by @lolipopshock in #27
• End-to-end paper parsing pipeline by @lolipopshock in #28
• add fix to dropped unicode by @kyleclo in #29

Full Changelog: v0.4.2...v0.5.0

v0.4.2: Better Bounding Box Scaling

What's Changed

• Check all characters in replace_unicode_tokens by @lolipopshock in #25
• Fix incorrect bbox scaling by @lolipopshock in #24

Full Changelog: v0.4.1...v0.4.2

v0.4.1: More robust handling of "empty" characters

What's Changed

• Replace Certain Unicode Characters for the input by @lolipopshock in #23

Full Changelog: v0.4.0...v0.4.1

v0.4.0: Better prediction APIs

What's Changed

• Allow batching in the predict function by @lolipopshock in #19
• Fix typo for separation in code by @lolipopshock in #20
• Better predict api design by @lolipopshock in #21

Full Changelog: v0.3.0...v0.4.0

v0.3.0: More robust Parsing for Large PDFs

What's Changed in VILA repo

• Increase the robustness for "large" PDFs by @lolipopshock in #16

Added Datasets Information for S2-VL

• Add Datasets by @lolipopshock in #11

Full Changelog: v0.2.1...v0.3.0

v0.2.1: Patch Release: Improve robustness on CPU machines

Patch Release

v0.2.0: Better Model Config Parsing, HF Models

Release v0.2.0

v0.1.1: Fix version name, allow specifying dpi

Fix version name, allow specifying dpi

v0.1.0: Initial VILA models, PDF loaders, and prediction postprocess

v0.1.0