The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 14 new columns ({'applications', 'release', 'domains', 'action_entropy', 'screen_height', 'complexity', 'action_frequency', 'task_description_alignment', 'step_num', 'websites', 'system', 'task_description_ambiguity', 'screen_width', 'verify_feedback'}) and 9 missing columns ({'task_difficulty', 'efficiency_score', 'natural_language_task', 'domain', 'alignment_score', 'traj', 'task_completed', 'actual_task', 'reason'}).

This happened while the json dataset builder was generating data using

hf://datasets/KrishPatel0111/AgentNet/meta_data_merged.jsonl (at revision 7e694d29fbfee27a7e86eca899806e5532002dc9)

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1831, in _prepare_split_single
                  writer.write_table(table)
                File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 714, in write_table
                  pa_table = table_cast(pa_table, self._schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              task_id: string
              instruction: string
              verify_feedback: struct<quality: string, feedback: string>
                child 0, quality: string
                child 1, feedback: string
              domains: string
              system: string
              release: string
              applications: list<item: string>
                child 0, item: string
              websites: list<item: string>
                child 0, item: string
              step_num: int64
              screen_width: int64
              screen_height: int64
              task_description_alignment: string
              task_description_ambiguity: string
              complexity: struct<multi_application_or_website: bool, contains_professional_knowledge: bool, involves_uncommon_ (... 119 chars omitted)
                child 0, multi_application_or_website: bool
                child 1, contains_professional_knowledge: bool
                child 2, involves_uncommon_features: bool
                child 3, involves_repetitive_simple_subtasks: bool
                child 4, logically_coherent_long_task: bool
                child 5, complexity_score: int64
              action_entropy: double
              action_frequency: struct<moveTo: int64, dragTo: int64, click: int64, doubleClick: int64, write: int64, press: int64, t (... 121 chars omitted)
                child 0, moveTo: int64
                child 1, dragTo: int64
                child 2, click: int64
                child 3, doubleClick: int64
                child 4, write: int64
                child 5, press: int64
                child 6, terminate: int64
                child 7, hotkey: int64
                child 8, hscroll: int64
                child 9, scroll: int64
                child 10, rightClick: int64
                child 11, tripleClick: int64
                child 12, middleClick: int64
              to
              {'task_id': Value('string'), 'instruction': Value('string'), 'traj': List({'index': Value('int64'), 'step_id': Value('string'), 'image': Value('string'), 'value': {'observation': Value('string'), 'thought': Value('string'), 'action': Value('string'), 'code': Value('string'), 'last_step_correct': Value('bool'), 'last_step_redundant': Value('bool'), 'reflection': Value('string')}, 'marks': List(Value('string'))}), 'domain': Value('string'), 'task_completed': Value('bool'), 'alignment_score': Value('int64'), 'efficiency_score': Value('int64'), 'task_difficulty': Value('int64'), 'reason': Value('string'), 'natural_language_task': Value('string'), 'actual_task': Value('string')}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1334, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 911, in stream_convert_to_parquet
                  builder._prepare_split(
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1702, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1833, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 14 new columns ({'applications', 'release', 'domains', 'action_entropy', 'screen_height', 'complexity', 'action_frequency', 'task_description_alignment', 'step_num', 'websites', 'system', 'task_description_ambiguity', 'screen_width', 'verify_feedback'}) and 9 missing columns ({'task_difficulty', 'efficiency_score', 'natural_language_task', 'domain', 'alignment_score', 'traj', 'task_completed', 'actual_task', 'reason'}).
              
              This happened while the json dataset builder was generating data using
              
              hf://datasets/KrishPatel0111/AgentNet/meta_data_merged.jsonl (at revision 7e694d29fbfee27a7e86eca899806e5532002dc9)
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

task_id
string
instruction
string
traj
list
domain
string
task_completed
bool
alignment_score
int64
efficiency_score
int64
task_difficulty
int64
reason
string
actual_task
string
natural_language_task
string
0030dc52-2a4a-4c0e-895b-48284c200efe
Open the Pikachu picture on the desktop using GIMP, and then select the "Venetian Blinds" filter in the animation.
[ { "index": 0, "image": "30a6f01b-9daf-4107-a755-a4a602d0de8e.png", "value": { "code": "pyautogui.click(x=0.018, y=0.508)", "last_step_correct": true, "last_step_redundant": false, "observation": "The current screen shows a desktop environment with a blue background featuring a la...
infeasible
false
6
9
4
The agent successfully completed the first part of the task by efficiently opening GIMP, loading the pikachu.jpeg file from the desktop, and navigating to the Animation filters menu. All steps were logical and executed without redundancy. However, the task could not be completed because the 'Venetian Blinds' filter does not exist in GIMP's Animation category. The agent correctly identified this limitation and appropriately terminated the task rather than continuing futilely.
Launch GIMP application, open the pikachu.jpeg file from the desktop using the file dialog, navigate to Filters > Animation menu to locate the Venetian Blinds filter, and discover that this specific filter is not available in GIMP's Animation category.
Open the image editing software GIMP, load the pikachu.jpeg file from the desktop, then navigate to the Filters menu and look for a 'Venetian Blinds' filter in the Animation section to apply it to the image.
0019cfaa-b2a0-424d-a2b0-b068d7956a7e
Open "Notifications" in the settings and turn on Notification Popups in Archive Manager.
[ { "index": 0, "image": "2614ed71-c198-440e-bc4b-a6360e23465b.png", "value": { "code": "pyautogui.click(x=0.021, y=0.951)", "last_step_correct": true, "last_step_redundant": false, "observation": "The current screen shows a desktop environment with a blue background featuring a la...
infeasible
false
8
7
4
The agent successfully navigated through all the correct steps to reach the Archive Manager notification settings and located the exact 'Notification Popups' toggle that needed to be enabled. However, the final critical action of actually enabling the toggle failed due to an unresponsive interface element. The navigation was efficient and accurate, but the task could not be completed due to a technical issue beyond the agent's control.
Open the system Settings application, navigate to the Notifications section, locate Archive Manager in the Applications list, open its detailed notification settings, and enable the 'Notification Popups' toggle switch.
Please open the system settings, navigate to the Notifications section, find Archive Manager in the applications list, and enable the 'Notification Popups' option for that application.
00406a3b-9879-40b0-94d8-f7d21b7858d7
Please help me change the shortcut key for closing the window to 'Ctrl+Shift' and delete another shortcut key for closing the window.
[ { "index": 0, "image": "9ba75579-f068-4b13-8b23-7e03c0cc9ab8.png", "value": { "code": "pyautogui.click(x=0.019, y=0.24)", "last_step_correct": true, "last_step_redundant": null, "observation": "The screen displays a Linux desktop environment with a GNOME or similar interface. The...
VScode
null
null
null
null
null
null
null
0022f50b-7983-4c7a-9005-01a6d924c152
I'm editing"The Xiaomi Product Ecosystem Chain.pptx,"and I'm not very satisfied with the content of part 02 and part 04.Could you please add a strikethrough to the main titles of the chapter pages for these two sections to distinguish them,and also update the table of contents page accordingly?
[ { "index": 0, "image": "36d34fb0-99a1-4753-82ec-4452abc39fa6.png", "value": { "code": "pyautogui.click(x=0.086, y=0.611)", "last_step_correct": true, "last_step_redundant": null, "observation": "I'm looking at LibreOffice Impress with \"The Xiaomi Product Ecosystem Chain.pptx\" f...
libreoffice_impress
null
null
null
null
null
null
null
00721815-cc6b-41b2-b078-e743c9108357
Can you enable the sound field locator and then change the size to 6.0 and dry to 7.0?
[ { "index": 0, "image": "6500d241-ebba-4b93-9c93-52250365c95c.png", "value": { "code": "pyautogui.click(x=0.018, y=0.69)", "last_step_correct": null, "last_step_redundant": null, "observation": "I'm looking at a Linux desktop environment with a purple geometric jellyfish wallpaper...
VLC
null
null
null
null
null
null
null
00535dcd-6fcf-4129-aaa6-753c801bed5f
"Open VS Code and modify settings, enable \"Send Keybindings To Shell\", set \"Decorations Enabled\"(...TRUNCATED)
[{"index":0,"image":"11047698-5836-4b68-ab9b-6f0797647586.png","value":{"code":"pyautogui.click(x=0.(...TRUNCATED)
VScode
null
null
null
null
null
null
null
0064b970-52f8-4dba-affb-cf2d7e670750
"Open GIMP, open the file/Wonderland.png, draw an ellipse in the middle of the image, and then fill (...TRUNCATED)
[{"index":0,"image":"82e694cf-fc6e-4f84-81b6-f9ffaf81afc4.png","value":{"code":"pyautogui.click(x=0.(...TRUNCATED)
Gimp
null
null
null
null
null
null
null
004e0c77-696e-4cba-b673-b9ac7c48c92e
"Please help me modify the document \"Development data of online education.ods\" from the desktop in(...TRUNCATED)
[{"index":0,"image":"6203c21d-c493-45cf-a5e5-3f203619126e.png","value":{"code":"pyautogui.click(x=0.(...TRUNCATED)
MultiApp
null
null
null
null
null
null
null
0059e663-2215-43b8-ad1e-7012d1055475
"I think it would be better to replace \"applications\" in the current slide with \"utilizations\". (...TRUNCATED)
[{"index":0,"image":"356ea094-a55b-4985-867f-fcf61e363c59.png","value":{"code":"pyautogui.click(x=0.(...TRUNCATED)
libreoffice_impress
null
null
null
null
null
null
null
00af4cc2-25da-46cc-bbc9-c48f1d7dd242
"Open the terminal, navigate to the \"Desktop/pp\" directory, create a new blank \"goast.out\" file (...TRUNCATED)
[{"index":0,"image":"431d8930-e966-4604-a006-87b84c11dc17.png","value":{"code":"pyautogui.click(x=0.(...TRUNCATED)
OS
true
10
8
3
"The agent successfully completed all required steps of the task: opening the terminal, navigating t(...TRUNCATED)
"Open the terminal application, navigate to the Desktop/pp directory using cd ~/Desktop/pp, create a(...TRUNCATED)
"Could you open the terminal for me, go to the pp folder on my Desktop, create an empty file called (...TRUNCATED)
End of preview.