Datasets:
v1.1: regenerated with validation checks and answer-balance quotas
Browse files- README.md +63 -0
- data/train-00000-of-00001.parquet +2 -2
README.md
CHANGED
|
@@ -6,6 +6,69 @@ sdk: docker
|
|
| 6 |
pinned: false
|
| 7 |
license: mit
|
| 8 |
arxiv: 2401.xxxxx
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
A Python-based pipeline for CLEVR-style scenes with counterfactual variants and Blender rendering. Configurable object counts, counterfactual types, and automated questions and answers.
|
|
|
|
| 6 |
pinned: false
|
| 7 |
license: mit
|
| 8 |
arxiv: 2401.xxxxx
|
| 9 |
+
dataset_info:
|
| 10 |
+
features:
|
| 11 |
+
- name: original_image
|
| 12 |
+
dtype: image
|
| 13 |
+
- name: semantic_counterfactual_image
|
| 14 |
+
dtype: image
|
| 15 |
+
- name: negative_counterfactual_image
|
| 16 |
+
dtype: image
|
| 17 |
+
- name: semantic_counterfactual_type
|
| 18 |
+
dtype: string
|
| 19 |
+
- name: negative_counterfactual_type
|
| 20 |
+
dtype: string
|
| 21 |
+
- name: semantic_counterfactual_description
|
| 22 |
+
dtype: string
|
| 23 |
+
- name: negative_counterfactual_description
|
| 24 |
+
dtype: string
|
| 25 |
+
- name: original_question
|
| 26 |
+
dtype: string
|
| 27 |
+
- name: semantic_counterfactual_question_1
|
| 28 |
+
dtype: string
|
| 29 |
+
- name: negative_counterfactual_question_1
|
| 30 |
+
dtype: string
|
| 31 |
+
- name: original_question_answer_type
|
| 32 |
+
dtype: string
|
| 33 |
+
- name: semantic_counterfactual_question_1_answer_type
|
| 34 |
+
dtype: string
|
| 35 |
+
- name: negative_counterfactual_question_1_answer_type
|
| 36 |
+
dtype: string
|
| 37 |
+
- name: original_question_reasoning_op
|
| 38 |
+
dtype: string
|
| 39 |
+
- name: semantic_counterfactual_question_1_reasoning_op
|
| 40 |
+
dtype: string
|
| 41 |
+
- name: negative_counterfactual_question_1_reasoning_op
|
| 42 |
+
dtype: string
|
| 43 |
+
- name: original_image_answer_to_original_question
|
| 44 |
+
dtype: string
|
| 45 |
+
- name: original_image_answer_to_semantic_counterfactual_question_1
|
| 46 |
+
dtype: string
|
| 47 |
+
- name: original_image_answer_to_negative_counterfactual_question_1
|
| 48 |
+
dtype: string
|
| 49 |
+
- name: semantic_counterfactual_image_answer_to_original_question
|
| 50 |
+
dtype: string
|
| 51 |
+
- name: semantic_counterfactual_image_answer_to_semantic_counterfactual_question_1
|
| 52 |
+
dtype: string
|
| 53 |
+
- name: semantic_counterfactual_image_answer_to_negative_counterfactual_question_1
|
| 54 |
+
dtype: string
|
| 55 |
+
- name: negative_counterfactual_image_answer_to_original_question
|
| 56 |
+
dtype: string
|
| 57 |
+
- name: negative_counterfactual_image_answer_to_semantic_counterfactual_question_1
|
| 58 |
+
dtype: string
|
| 59 |
+
- name: negative_counterfactual_image_answer_to_negative_counterfactual_question_1
|
| 60 |
+
dtype: string
|
| 61 |
+
splits:
|
| 62 |
+
- name: train
|
| 63 |
+
num_bytes: 38589925
|
| 64 |
+
num_examples: 100
|
| 65 |
+
download_size: 38575846
|
| 66 |
+
dataset_size: 38589925
|
| 67 |
+
configs:
|
| 68 |
+
- config_name: default
|
| 69 |
+
data_files:
|
| 70 |
+
- split: train
|
| 71 |
+
path: data/train-*
|
| 72 |
---
|
| 73 |
|
| 74 |
A Python-based pipeline for CLEVR-style scenes with counterfactual variants and Blender rendering. Configurable object counts, counterfactual types, and automated questions and answers.
|
data/train-00000-of-00001.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:948a456ab36e3a089125d7d203be8d7464d01dddeba317d657784b38420ae958
|
| 3 |
+
size 38575846
|