EASI Release v0.2.1

[Release] Release EASI V0.2.1 (#28)

* [Sync] sync submodule update

* [Docs] Add ERIQ && OSI-bench info

* [Docs] update readme

* [Docs] Added lmms-eval results on osibench, starebench, spatialvizbench, mmsi-video-bench

* [Docs] Added lmmseval task names to supported bench

---------

Co-authored-by: oscarqjh <oscar.jh9@gmail.com>

EASI Release v0.2.0

• New Backend Support: Integrated lmms-eval alongside VLMEvalKit, offering flexible evaluation options.
• Expanded benchmark support: Added DSR-Bench.

EASI Release v0.1.5

Added

• 1 new benchmark:
  • STI-Bench
• 2 new models:
  • SenseNova-SI-1.1-BAGEL-7B-MoT, SenseNova-SI-1.3-InternVL3-8B
• Add Benchmark Verification Info

EASI Release v0.1.4

• Expanded benchmark support
  Added 4 benchmarks: SPBench, MMSI-Video-Bench, VSI-SUPER-Recall, VSI-SUPER-Count.

EASI Release v0.1.3

• Expanded benchmark support
  Added 3 image benchmarks: ERQA, RefSpatial-Bench, RoboSpatial-Home.

• Improved environment & deployment support
  Added a generic EASI Dockerfile, as well as model-specific Dockerfiles for Cambrian-S and VLM3R, simplifying environment setup and improving reproducible evaluation.

EASI Release v0.1.2

New Features

• Added 5 Spatial Intelligence models:

  • SenseNova-SI 1.1 Series: Qwen2.5-VL-3B, Qwen2.5-VL-7B, Qwen3-VL-8B
  • SenseNova-SI 1.2 Series: InternVL3-8B
  • VLM-3R

• Added 1 unified understanding–generation model:

  • BAGEL-7B-MoT

• Added 4 image benchmarks:

  • STAR-Bench, OmniSpatial, Spatial-Visualization-Benchmark, SPAR-Bench

• LLM-based answer extraction for EASI benchmarks

  • Added optional LLM-based answer extraction for several EASI benchmarks.

EASI Release v0.1.1

New Features:

• Added 9 new Spatial Intelligence models (Total models increased from 7 → 16):
  • SenseNova-SI 1.1 Series
    • SenseNova-SI-1.1-InternVL3-8B
    • SenseNova-SI-1.1-InternVL3-2B
  • SpaceR: SpaceR-7B
  • VST Series: VST-3B-SFT, VST-7B-SFT
  • Cambrian-S series: Cambrian-S-0.5B, Cambrian-S-1.5B, Cambrian-S-3B, Cambrian-S-7B
• Added 1 new image–video benchmark (Total benchmarks increased from 6 → 7):
  • VSI-Bench-Debiased

EASI Release v0.1.0

New Features:

• Supports 7 recent Spatial Intelligence models:
  • SenseNova-SI Family: SenseNova-SI-InternVL3-8B, SenseNova-SI-InternVL3-2B
  • MindCube Family: MindCube-3B-RawQA-SFT, MindCube-3B-Aug-CGMap-FFR-Out-SFT,MindCube-3B-Plain-CGMap-FFR-Out-SFT
  • SpatialLadder: SpatialLadder-3B
  • SpatialMLLM: SpatialMLLM-4B
• Supports 6 recent Spatial Intelligence benchmarks:
  • 4 image-based benchmarks: MindCube, ViewSpatial, EmbSpatial and MMSI(no circular evaluation)
  • 2 image-and-video benchmarks: VSI-Bench and SITE-Bench
• Introduces a standardized testing protocol as outlined in EASI