Datasets:
Upload dataset
Browse files- README.md +36 -8
- test/test-00000-of-00001.parquet +3 -0
README.md
CHANGED
|
@@ -1,4 +1,12 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
dataset_info:
|
| 3 |
- config_name: mantis_eval
|
| 4 |
features:
|
|
@@ -24,19 +32,39 @@ dataset_info:
|
|
| 24 |
num_examples: 217
|
| 25 |
download_size: 473031413
|
| 26 |
dataset_size: 479770102
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
configs:
|
| 28 |
- config_name: mantis_eval
|
| 29 |
data_files:
|
| 30 |
- split: test
|
| 31 |
path: mantis_eval/test-*
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
-
|
| 35 |
-
|
| 36 |
-
- en
|
| 37 |
-
pretty_name: Mantis-Eval
|
| 38 |
-
size_categories:
|
| 39 |
-
- n<1K
|
| 40 |
---
|
| 41 |
|
| 42 |
This is a newly curated dataset to evaluate multimodal language models' capability to reason over multiple images. More details are shown in https://tiger-ai-lab.github.io/Mantis/.
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
size_categories:
|
| 6 |
+
- n<1K
|
| 7 |
+
task_categories:
|
| 8 |
+
- question-answering
|
| 9 |
+
pretty_name: Mantis-Eval
|
| 10 |
dataset_info:
|
| 11 |
- config_name: mantis_eval
|
| 12 |
features:
|
|
|
|
| 32 |
num_examples: 217
|
| 33 |
download_size: 473031413
|
| 34 |
dataset_size: 479770102
|
| 35 |
+
- config_name: test
|
| 36 |
+
features:
|
| 37 |
+
- name: id
|
| 38 |
+
dtype: string
|
| 39 |
+
- name: question_type
|
| 40 |
+
dtype: string
|
| 41 |
+
- name: question
|
| 42 |
+
dtype: string
|
| 43 |
+
- name: images
|
| 44 |
+
sequence: image
|
| 45 |
+
- name: options
|
| 46 |
+
sequence: string
|
| 47 |
+
- name: answer
|
| 48 |
+
dtype: string
|
| 49 |
+
- name: data_source
|
| 50 |
+
dtype: string
|
| 51 |
+
- name: category
|
| 52 |
+
dtype: string
|
| 53 |
+
splits:
|
| 54 |
+
- name: test
|
| 55 |
+
num_bytes: 479770102.0
|
| 56 |
+
num_examples: 217
|
| 57 |
+
download_size: 473031413
|
| 58 |
+
dataset_size: 479770102.0
|
| 59 |
configs:
|
| 60 |
- config_name: mantis_eval
|
| 61 |
data_files:
|
| 62 |
- split: test
|
| 63 |
path: mantis_eval/test-*
|
| 64 |
+
- config_name: test
|
| 65 |
+
data_files:
|
| 66 |
+
- split: test
|
| 67 |
+
path: test/test-*
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
---
|
| 69 |
|
| 70 |
This is a newly curated dataset to evaluate multimodal language models' capability to reason over multiple images. More details are shown in https://tiger-ai-lab.github.io/Mantis/.
|
test/test-00000-of-00001.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d55c78042d07546214e15446701698daa77cc650cd8c89f4b61df4a6574bfcf
|
| 3 |
+
size 473031413
|