Hugging Face's logo

Datasets:
remyxai
/
vqasynth_spacellava

Tasks:
Visual Question Answering
Modalities:
Image
Text
Formats:
parquet
Size:
10K - 100K
Tags:
remyx
Synthetic
Libraries:
Datasets
Dask
License:
Dataset card Viewer Files Community
1
salma-remyx commited on
Commit
c642c8e
•
1 Parent(s): ee16ba1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -1
README.md CHANGED
@@ -40,4 +40,12 @@ task_categories:
40
  pretty_name: OpenSpaces
41
  size_categories:
42
  - 10K<n<100K
43
- ---
 
 
 
 
 
 
 
 
 
40
  pretty_name: OpenSpaces
41
  size_categories:
42
  - 10K<n<100K
43
+ ---
44
+
45
+
46
+
47
+
48
+ # VQASynth_spacellava
49
+ ![image/png](https://cdn-uploads.huggingface.co/production/uploads/647777304ae93470ffc28913/t_hxwRDUgf-HywL89rJX4.png)
50
+
51
+ Uses the [VQASynth](https://github.com/remyxai/VQASynth) pipeline to synthesize spatialVQA samples, mixed with general VQA samples used to fine-tune LLaVA-v1.5-13b.