Hugging Face's logo

sbintuitions
/
sarashina2.2-vision-3b

Image-to-Text
Transformers
Safetensors
Japanese
English
sarashina2_vision
text-generation
multimodal
vision-language
custom_code
Model card Files
xet
Community
toshi-456 commited on
Commit
147b6cc
·
verified ·
1 Parent(s): 61ac967

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -6
README.md CHANGED
@@ -22,24 +22,24 @@ This model is based on [Sarashina2.2-3B-Instruct](https://huggingface.co/sbintui
22
 
23
  |Model|Params(B)|[BussinessSlide VQA](https://github.com/stockmarkteam/business-slide-questions)<sup>*1</sup>|[Heron-Bench](https://arxiv.org/abs/2404.07824)<sup>*1</sup>|[JDocQA](https://arxiv.org/abs/2403.19454)<sup>*1</sup>|[JMMMU](https://arxiv.org/abs/2410.17250)|
24
  |-|-|-|-|-|-|
25
- |[Sarashina2.2-Vision-3B](https://huggingface.co/sbintuitions/sarashina2.2-vision-3b)|3.8|3.932|**3.214**|<u>3.327</u>|<u>0.486</u>|
26
  |[Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct)|3.8|3.516|2.000|3.019|0.450|
27
- |[Qwen3-VL-4B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-4B-Instruct)|4.4|**4.105**|2.330|**3.596**|**0.493**|
28
  |[InternVL3_5-4B](https://huggingface.co/OpenGVLab/InternVL3_5-4B)|4.7|3.311|1.893|2.626|0.437|
29
  |[Sarashina2-Vision-14B](https://huggingface.co/sbintuitions/sarashina2-vision-14b)|14.4|3.110|2.184|-<sup>*2</sup>|0.432|
30
  |[Stockmark-2-VL-100B-beta](https://huggingface.co/stockmark/Stockmark-2-VL-100B-beta)|96.5|<u>3.973</u>|<u>2.563</u>|3.168|-<sup>*2</sup>|
31
 
32
  *1. [gpt-oss-120b](https://huggingface.co/openai/gpt-oss-120b) was used for LLM-as-a-Judge.
33
 
34
- *2. Score cannot be measured because some input data exceeds the model's `max_position_embeddings`.
35
 
36
  ### English Performance
37
 
38
  |Model|Params(B)|[DocVQA](https://arxiv.org/abs/2007.00398)|[InfoVQA](https://arxiv.org/abs/2104.12756)|[RealworldQA](https://huggingface.co/datasets/xai-org/RealworldQA)
39
  |-|-|-|-|-|
40
- |[Sarashina2.2-Vision-3B](https://huggingface.co/sbintuitions/sarashina2.2-vision-3b)|3.8|0.831|0.567|<u>0.625</u>|
41
- |[Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct)|3.8|<u>0.924</u>|<u>0.750</u>|0.586|
42
- |[Qwen3-VL-4B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-4B-Instruct)|4.4|**0.948**|**0.798**|**0.712**|
43
  |[InternVL3_5-4B](https://huggingface.co/OpenGVLab/InternVL3_5-4B)|4.7|0.823|0.541|0.553|
44
  |[Sarashina2-Vision-14B](https://huggingface.co/sbintuitions/sarashina2-vision-14b)|14.4|0.729|0.490|0.519||
45
 
 
22
 
23
  |Model|Params(B)|[BussinessSlide VQA](https://github.com/stockmarkteam/business-slide-questions)<sup>*1</sup>|[Heron-Bench](https://arxiv.org/abs/2404.07824)<sup>*1</sup>|[JDocQA](https://arxiv.org/abs/2403.19454)<sup>*1</sup>|[JMMMU](https://arxiv.org/abs/2410.17250)|
24
  |-|-|-|-|-|-|
25
+ |[Sarashina2.2-Vision-3B](https://huggingface.co/sbintuitions/sarashina2.2-vision-3b)|3.8|3.932|3.214|3.327|0.486|
26
  |[Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct)|3.8|3.516|2.000|3.019|0.450|
27
+ |[Qwen3-VL-4B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-4B-Instruct)|4.4|4.105|2.330|3.596|0.493|
28
  |[InternVL3_5-4B](https://huggingface.co/OpenGVLab/InternVL3_5-4B)|4.7|3.311|1.893|2.626|0.437|
29
  |[Sarashina2-Vision-14B](https://huggingface.co/sbintuitions/sarashina2-vision-14b)|14.4|3.110|2.184|-<sup>*2</sup>|0.432|
30
  |[Stockmark-2-VL-100B-beta](https://huggingface.co/stockmark/Stockmark-2-VL-100B-beta)|96.5|<u>3.973</u>|<u>2.563</u>|3.168|-<sup>*2</sup>|
31
 
32
  *1. [gpt-oss-120b](https://huggingface.co/openai/gpt-oss-120b) was used for LLM-as-a-Judge.
33
 
34
+ *2. These scores cannot be measured because some input data exceeds the model's `max_position_embeddings`.
35
 
36
  ### English Performance
37
 
38
  |Model|Params(B)|[DocVQA](https://arxiv.org/abs/2007.00398)|[InfoVQA](https://arxiv.org/abs/2104.12756)|[RealworldQA](https://huggingface.co/datasets/xai-org/RealworldQA)
39
  |-|-|-|-|-|
40
+ |[Sarashina2.2-Vision-3B](https://huggingface.co/sbintuitions/sarashina2.2-vision-3b)|3.8|0.831|0.567|0.625|
41
+ |[Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct)|3.8|0.924|0.750|0.586|
42
+ |[Qwen3-VL-4B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-4B-Instruct)|4.4|0.948|0.798|0.712|
43
  |[InternVL3_5-4B](https://huggingface.co/OpenGVLab/InternVL3_5-4B)|4.7|0.823|0.541|0.553|
44
  |[Sarashina2-Vision-14B](https://huggingface.co/sbintuitions/sarashina2-vision-14b)|14.4|0.729|0.490|0.519||
45