Update README.md
Browse files
README.md
CHANGED
|
@@ -90,7 +90,37 @@ dataset_info:
|
|
| 90 |
- name: question_type
|
| 91 |
dtype: string
|
| 92 |
- name: target_options
|
| 93 |
-
dtype: string
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 94 |
configs:
|
| 95 |
- config_name: maze
|
| 96 |
data_files:
|
|
@@ -115,7 +145,15 @@ configs:
|
|
| 115 |
- config_name: spatial_map_text_only
|
| 116 |
data_files:
|
| 117 |
- split: val
|
| 118 |
-
path: spatial_map/spatial_map_text_only_val.parquet
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 119 |
---
|
| 120 |
|
| 121 |
A key question for understanding multimodal vs. language capabilities of models is what is
|
|
|
|
| 90 |
- name: question_type
|
| 91 |
dtype: string
|
| 92 |
- name: target_options
|
| 93 |
+
dtype: string
|
| 94 |
+
- config_name: spatial_map_noinstruction
|
| 95 |
+
features:
|
| 96 |
+
- name: id
|
| 97 |
+
dtype: int32
|
| 98 |
+
- name: image
|
| 99 |
+
dtype: image
|
| 100 |
+
- name: prompt
|
| 101 |
+
dtype: string
|
| 102 |
+
- name: ground_truth
|
| 103 |
+
dtype: string
|
| 104 |
+
- name: task
|
| 105 |
+
dtype: string
|
| 106 |
+
- name: question_type
|
| 107 |
+
dtype: string
|
| 108 |
+
- name: target_options
|
| 109 |
+
dtype: string
|
| 110 |
+
- config_name: spatial_map_text_only_noinstruction
|
| 111 |
+
features:
|
| 112 |
+
- name: id
|
| 113 |
+
dtype: int32
|
| 114 |
+
- name: prompt
|
| 115 |
+
dtype: string
|
| 116 |
+
- name: ground_truth
|
| 117 |
+
dtype: string
|
| 118 |
+
- name: task
|
| 119 |
+
dtype: string
|
| 120 |
+
- name: question_type
|
| 121 |
+
dtype: string
|
| 122 |
+
- name: target_options
|
| 123 |
+
dtype: string
|
| 124 |
configs:
|
| 125 |
- config_name: maze
|
| 126 |
data_files:
|
|
|
|
| 145 |
- config_name: spatial_map_text_only
|
| 146 |
data_files:
|
| 147 |
- split: val
|
| 148 |
+
path: spatial_map/spatial_map_text_only_val.parquet
|
| 149 |
+
- config_name: spatial_map_noinstruction
|
| 150 |
+
data_files:
|
| 151 |
+
- split: val
|
| 152 |
+
path: spatial_map/spatial_map_val_noinstruction.parquet
|
| 153 |
+
- config_name: spatial_map_text_only_noinstruction
|
| 154 |
+
data_files:
|
| 155 |
+
- split: val
|
| 156 |
+
path: spatial_map/spatial_map_text_only_val_noinstruction.parquet
|
| 157 |
---
|
| 158 |
|
| 159 |
A key question for understanding multimodal vs. language capabilities of models is what is
|