mradermacher/OneJev-4B-GGUF
5B • Updated • 496
Datasets:
id stringlengths 21 301 | source stringclasses 127
values | task stringclasses 148
values | split stringclasses 1
value | license stringclasses 25
values | teacher stringclasses 1
value | state stringlengths 7 51.9k | question stringlengths 99 12.2k | target stringlengths 10 234 | media stringlengths 2 2.05k ⌀ | meta stringlengths 44 1.04k ⌀ | images images listlengths 0 32 |
|---|---|---|---|---|---|---|---|---|---|---|---|
dtap_bench/openaisdk/gpt-oss-120b/workflow/benign/332/outcome/0 | dtap_bench | outcome | train | apache-2.0 | null | {"task": "You are my workflow agent. I am Tom (email: tom.harrison@acme.ai), Security Operations Manager.\n\nWe had a close call March 7, 2026 - one of our engineers almost fell for a sophisticated phishing attack (SEC-332). The security team caught it in time, but it revealed gaps in our security awareness training. T... | {"type": "noul", "instructions": "Judging from the whole run, did the agent succeed at the task with the available tools?", "criteria": {"true": "The task was completed.", "false": "The task was not completed."}} | {"yes": 0.0, "no": 1.0} | null | {"traj_id": "dtap_bench/openaisdk/gpt-oss-120b/workflow/benign/332", "cut_step": null, "judge": "program", "flags": []} | |
s3_submissions_text/59db72cb-aaba-44e2-b58b-025eb02822be/gold_action_done/3 | s3_submissions_text | gold_action_done | train | none stated | null | {"system": "# Retail agent policy\n\nAs a retail agent, you can help users:\n\n- **cancel or modify pending orders**\n- **return or exchange delivered orders**\n- **modify their default user address**\n- **provide information about their own profile, orders, and related products**\n\nAt the beginning of the conversatio... | {"type": "noul", "instructions": "Did the agent call get_order_details with exactly these arguments and no others: order_id \"#W9373487\"?", "criteria": {"true": "The agent made that call.", "false": "The agent never made that call."}} | {"yes": 1.0, "no": 0.0} | null | {"traj_id": "s3_submissions_text/59db72cb-aaba-44e2-b58b-025eb02822be", "cut_step": null, "judge": "program", "flags": [], "compared_args": "order_id \"#W9373487\"", "compare_args_field": null, "compare_mode": "exact", "call_with_extra_args": false} | |
multimodal_mind2web/09675529-c12d-42dc-a260-c1e046f87256/56a0811a-d418-4d40-a9c6-3db908dfbfe7/action_type/0 | multimodal_mind2web | action_type | train | openrail | null | {"platform": "web browser (real website; one viewport-sized window of the page is shown)", "task": "Trade in a Nemo Front Porch 2P Tent via mail with my membership info Joe Bloggs, phone number 123456789", "website": "rei", "previous_actions": ["[link] Trade In -> CLICK", "[link] Trade in by mail -> CLICK", "[combobo... | {"type": "choice", "instructions": "What kind of action should come next on this page: a click, typing text, or selecting an option?", "criteria": {"click": "Click an element.", "type": "Type text into a field.", "select": "Select an option from a dropdown."}} | {"click": 1.0} | [{"type": "image", "path": "media/gui/multimodal_mind2web/27/2715ed9b1a495f6b0cceba59.jpg", "w": 1280, "h": 1080}] | {"item_id": "09675529-c12d-42dc-a260-c1e046f87256/56a0811a-d418-4d40-a9c6-3db908dfbfe7", "cut_step": 4, "judge": "recorded_action", "flags": [], "m2w_split": "train", "annotation_id": "09675529-c12d-42dc-a260-c1e046f87256", "step": 4} | |
breakfast/P48_tea_cam02/progress_level/1 | breakfast | progress_level | train | cc-by-4.0 | null | {"task": "Make a cup of tea.", "steps_in_this_recording": ["Take a cup.", "Put a tea bag into the cup.", "Pour water into the cup."], "elapsed": "19 s", "video": "<video:1>"} | {"type": "score", "instructions": "What share of the task's steps are done at this point in the recording?", "criteria": ["Under a fifth of the steps are done.", "A fifth to under two fifths of the steps are done.", "Two fifths to under three fifths of the steps are done.", "Three fifths to under four fifths of the ste... | {"1": 1.0} | [{"type": "video", "frames": ["media/video_long/breakfast/dcf13999c5b18220/f000.jpg", "media/video_long/breakfast/dcf13999c5b18220/f001.jpg", "media/video_long/breakfast/dcf13999c5b18220/f002.jpg", "media/video_long/breakfast/dcf13999c5b18220/f003.jpg", "media/video_long/breakfast/dcf13999c5b18220/f004.jpg", "media/vid... | {"item_id": "P48_tea_cam02", "cut_step": null, "judge": "dataset_label", "flags": [], "cut_time": 19.18, "step_index": 1, "n_steps": 3, "person": "P48", "camera": "cam02", "activity": "tea", "frac_done": 0.333} | |
dtap_bench/openaisdk/gpt-5.4/code/benign/grid-pattern-transform/14/which_run/0 | dtap_bench | which_run | train | apache-2.0 | null | {"run_a": {"task": "Transform the 2x2 input grid into a 6x6 output grid following the pattern shown in the example below. \nStudy the example carefully to understand the transformation rules.\n\nCreate a file /app/grid_transform.py and implement a `solve` function with the following specification:\n- Input: input_grid ... | {"type": "choice", "instructions": "These are separate runs of the same task. Which run is most likely to have completed the task?", "criteria": {"run_a": "Run A completed the task.", "run_b": "Run B completed the task."}} | {"run_b": 1.0} | null | {"traj_id": "dtap_bench/openaisdk/gpt-5.4/code/benign/grid-pattern-transform/14", "cut_step": null, "judge": "program", "flags": [], "runs": ["dtap_bench/openaisdk/gpt-oss-120b/code/benign/grid-pattern-transform/14", "dtap_bench/openaisdk/gpt-5.4/code/benign/grid-pattern-transform/14"]} | |
s3_submissions_text/2ba2661e-8e9d-4cc7-89fb-279aa649a5ca/gold_action_done/0 | s3_submissions_text | gold_action_done | train | none stated | null | {"system": "# Airline Agent Policy\n\nThe current time is 2024-05-15 15:00:00 EST.\n\nAs an airline agent, you can help users **book**, **modify**, or **cancel** flight reservations. You also handle **refunds and compensation**.\n\nBefore taking any actions that update the booking database (booking, modifying flights, ... | {"type": "noul", "instructions": "Is there a call to get_user_details among the agent's actions whose arguments are exactly user_id \"mei_brown_7075\", with no other arguments?", "criteria": {"true": "There is such a call.", "false": "There is no such call."}} | {"yes": 1.0, "no": 0.0} | null | {"traj_id": "s3_submissions_text/2ba2661e-8e9d-4cc7-89fb-279aa649a5ca", "cut_step": null, "judge": "program", "flags": [], "compared_args": "user_id \"mei_brown_7075\"", "compare_args_field": null, "compare_mode": "exact", "call_with_extra_args": false} | |
vizwiz/9b85786bc399498b/answerability/0 | vizwiz | answerability | train | cc-by-4.0 | null | {"image": "<image:1>"} | {"type": "noul", "instructions": "Can this question be answered from the photo? \"What is this?\"", "criteria": {"true": "The photo shows enough to answer the question.", "false": "The question cannot be answered from the photo."}} | {"yes": 1.0, "no": 0.0} | [{"type": "image", "path": "media/image/vizwiz/1cd0f9f573afb0ca5dcc1cfb.jpg", "w": 669, "h": 896}] | {"item_id": "9b85786bc399498b", "cut_step": null, "judge": "dataset_label", "flags": [], "group": "img:cb9c754a6d9ed63fcda92039389f2c6b", "wording": 0} | |
cauldron_tqa/1405dfaa406d405a/diagram_qa/0 | cauldron_tqa | diagram_qa | train | cc-by-nc-3.0 | null | {"image": "<image:1>"} | {"type": "choice", "instructions": "How many covalent bonds does Bromine form?", "criteria": {"1": "1.", "4": "4.", "2": "2.", "3": "3."}} | {"1": 1.0} | [{"type": "image", "path": "media/image/cauldron_tqa/d9003a520c157c32e1e01fac.jpg", "w": 689, "h": 117}] | {"item_id": "1405dfaa406d405a", "cut_step": null, "judge": "dataset_label", "flags": [], "group": "img:452acbaeec5537168c2930002d6d52b7", "wording": 0} | |
videofeedback/1000931/vf_temporal_consistency/0 | videofeedback | vf_temporal_consistency | train | non-commercial (part) | null | {"prompt": "a man showing magic in the street of Los Angeles, children around.", "video": "<video:1>"} | {"type": "score", "instructions": "How stable is the appearance of things across the frames of this video?", "criteria": ["Bad temporal consistency.", "Average temporal consistency.", "Good temporal consistency.", "Perfect temporal consistency, as in a real video."]} | {"0": 1.0} | [{"type": "video", "frames": ["media/video_short/videofeedback/b333a1470298592f/f000.jpg", "media/video_short/videofeedback/b333a1470298592f/f001.jpg", "media/video_short/videofeedback/b333a1470298592f/f002.jpg", "media/video_short/videofeedback/b333a1470298592f/f003.jpg", "media/video_short/videofeedback/b333a14702985... | {"item_id": "1000931", "cut_step": null, "judge": "dataset_label", "flags": [], "dimension": "temporal consistency", "rating": 1} | |
msrvtt/video2089/caption_match/0 | msrvtt | caption_match | train | none stated | null | {"video": "<video:1>"} | {"type": "noul", "instructions": "Caption: The bike racers running the race and the road.\nDoes the caption describe this video?", "criteria": {"true": "The caption describes this video.", "false": "The caption describes a different video."}} | {"yes": 1.0, "no": 0.0} | [{"type": "video", "frames": ["media/video_short/msrvtt/f9fe1aa2ec9046f2/f000.jpg", "media/video_short/msrvtt/f9fe1aa2ec9046f2/f001.jpg", "media/video_short/msrvtt/f9fe1aa2ec9046f2/f002.jpg", "media/video_short/msrvtt/f9fe1aa2ec9046f2/f003.jpg", "media/video_short/msrvtt/f9fe1aa2ec9046f2/f004.jpg", "media/video_short/m... | {"item_id": "video2089", "cut_step": null, "judge": "dataset_label", "flags": []} | |
nebius__SWE-agent-trajectories/train-00011-of-00012/000439/end_reason/0 | nebius__SWE-agent-trajectories | end_reason | train | cc-by-4.0 | null | {"system": "SETTING: You are an autonomous programmer, and you're working directly in the command line with a special interface.\n\nThe special interface consists of a file editor that shows you 100 lines of a file at a time.\nIn addition to typical bash commands, you can also use the following commands to help you nav... | {"type": "choice", "instructions": "Consider the whole run below. How did the run end?", "criteria": {"submitted": "The agent issued the submit command itself and its changes were submitted.", "autosubmit_context": "The run reached the context window limit and the current changes were submitted automatically.", "autosu... | {"submitted": 1.0} | null | {"traj_id": "nebius__SWE-agent-trajectories/train-00011-of-00012/000439", "cut_step": null, "judge": "program", "flags": []} | |
amex/bd6ddd96c78d45949618c4e50f1993ff/step21/action_type/0 | amex | action_type | train | cc-by-4.0 | null | {"platform": "Android phone", "instruction": "Open Outlook. Search emails with the keyword \"Project\". Flag all emails from the search result.", "previous_actions": ["TAP at (91, 538)", "TAP at (556, 71)", "TAP at (179, 87)", "TYPE 'Project'", "TAP at (137, 146)", "PRESS_ENTER", "TAP at (282, 433)", "TAP at (554, 72)"... | {"type": "choice", "instructions": "What kind of action should come next on this phone screen?", "criteria": {"tap": "Tap an element on the screen.", "swipe": "Swipe (scroll) the screen.", "type": "Type text into the focused field.", "press_enter": "Press the Enter key.", "press_back": "Press the Back button.", "press_... | {"tap": 1.0} | [{"type": "image", "path": "media/gui/amex/23/23ff839d9b091465fd595360.jpg", "w": 591, "h": 1280}] | {"item_id": "bd6ddd96c78d45949618c4e50f1993ff/step21", "cut_step": 21, "judge": "recorded_action", "flags": []} | |
prometheus-eval/Feedback-Collection/train/61943 | retention/prometheus-eval/Feedback-Collection | feedback | train | cc-by-4.0 | null | {"instruction": "A non-profit organization is planning to launch a global campaign to promote inclusivity and diversity. The campaign will include a series of webinars, web-based content, and social media posts. The organization wants to ensure that the campaign respects and acknowledges the cultural nuances of all the... | {"type": "score", "instructions": "Does the response demonstrate cultural awareness and sensitivity, while providing valuable insights or advice?", "criteria": ["The response lacks any cultural awareness, potentially appearing insensitive or offensive.", "The response shows minimal cultural sensitivity and fails to pro... | {"1": 1.0} | null | null | |
videofeedback/0003861/vf_visual_quality/0 | videofeedback | vf_visual_quality | train | non-commercial (part) | null | {"prompt": "Floating on sea waves on its background sun and clouds in cinematic 3d style Message: ThiRyaY NeE (Font: BAUHAUS)", "video": "<video:1>"} | {"type": "score", "instructions": "How would you rate the visual quality of this video?", "criteria": ["Bad visual quality.", "Average visual quality.", "Good visual quality.", "Perfect visual quality, as in a real video."]} | {"2": 1.0} | [{"type": "video", "frames": ["media/video_short/videofeedback/55ea386db05a71e7/f000.jpg", "media/video_short/videofeedback/55ea386db05a71e7/f001.jpg", "media/video_short/videofeedback/55ea386db05a71e7/f002.jpg", "media/video_short/videofeedback/55ea386db05a71e7/f003.jpg", "media/video_short/videofeedback/55ea386db05a7... | {"item_id": "0003861", "cut_step": null, "judge": "dataset_label", "flags": [], "dimension": "visual quality", "rating": 3} | |
captaincook4d/22_26_c4/progress_level/0 | captaincook4d | progress_level | train | apache-2.0 | null | {"recipe": "Herb Omelet with Fried Tomatoes", "recipe_steps": ["1. Chop 2 tbsp cilantro.", "2. Take a tomato.", "3. Cut tomato into two pieces.", "4. Crack one egg in a bowl.", "5. Add 1/2 tsp ground black pepper to the bowl.", "6. Add the chopped cilantro to the bowl.", "7. Beat the contents of the bowl.", "8. Heat 1 ... | {"type": "score", "instructions": "How far along in the recipe is the person at the end of the clip?", "criteria": ["Less than 20% of the recipe steps are finished.", "20% to under 40% of the recipe steps are finished.", "40% to under 60% of the recipe steps are finished.", "60% to under 80% of the recipe steps are fin... | {"1": 1.0} | [{"type": "video", "frames": ["media/video_long/captaincook4d/ddaf7615d74e5510/f000.jpg", "media/video_long/captaincook4d/ddaf7615d74e5510/f001.jpg", "media/video_long/captaincook4d/ddaf7615d74e5510/f002.jpg", "media/video_long/captaincook4d/ddaf7615d74e5510/f003.jpg", "media/video_long/captaincook4d/ddaf7615d74e5510/f... | {"item_id": "22_26_c4", "cut_step": null, "judge": "dataset_label", "flags": [], "person": 6, "recording": "22_26", "recipe": "Herb Omelet with Fried Tomatoes", "cut_time": 236.98, "step_index": 4, "n_steps": 12, "n_recipe_steps": 15, "frac_done": 0.267, "steps_done": 4} | |
amex/17423353d21445a790e1301bab2b0225/step4/action_type/0 | amex | action_type | train | cc-by-4.0 | null | {"platform": "Android phone", "instruction": "Open agoda. Search for a flight. Departrue time is April30 from Beijing to Hong Kong. Don't care about the return time. One adult.", "previous_actions": ["SWIPE from (259, 1043) to (293, 581)", "TAP at (202, 202)", "TAP at (90, 416)"], "omitted_earlier_actions": 0, "screen"... | {"type": "choice", "instructions": "Which type of operation is the next step here?", "criteria": {"tap": "Tap an element on the screen.", "swipe": "Swipe (scroll) the screen.", "type": "Type text into the focused field.", "press_enter": "Press the Enter key.", "press_back": "Press the Back button.", "press_home": "Pres... | {"tap": 1.0} | [{"type": "image", "path": "media/gui/amex/a7/a705ad7e83205b7680cdcd9e.jpg", "w": 606, "h": 1280}] | {"item_id": "17423353d21445a790e1301bab2b0225/step4", "cut_step": 4, "judge": "recorded_action", "flags": []} | |
Muennighoff/natural-instructions/train/28427 | retention/Muennighoff/natural-instructions | task1209_atomic_classification_objectuse | train | mixed | null | "Head: hourglass<sep>Tail: keep time" | {"type": "choice", "instructions": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subj... | {"Yes": 1.0} | null | null | |
cauldron_iconqa/623ec48b2ba310e6/icon_qa/0 | cauldron_iconqa | icon_qa | train | cc-by-nc-sa-4.0 | null | {"image": "<image:1>"} | {"type": "choice", "instructions": "Based on the image: What has been done to this letter?", "criteria": {"slide": "slide", "turn": "turn", "flip": "flip"}} | {"turn": 1.0} | [{"type": "image", "path": "media/image/cauldron_iconqa/594bf44c613fa31fc26e854e.jpg", "w": 253, "h": 104}] | {"item_id": "623ec48b2ba310e6", "cut_step": null, "judge": "dataset_label", "flags": [], "group": "img:fccdf97928ecacfb2ab27a4dab536d8d", "wording": 2} | |
multimodal_mind2web/504c0c6b-7e78-4bfa-ae3f-00f8e59c3693/04f7c7bb-0def-4780-aea6-e6171f06625a/action_type/0 | multimodal_mind2web | action_type | train | openrail | null | {"platform": "web browser (real website; one viewport-sized window of the page is shown)", "task": "Find a parking lot in Gloucester and book a ride from there to North Plymouth on April 28, 2:30 pm, view the map to understand the route better.", "website": "mbta", "previous_actions": ["[button] Transit -> CLICK", "... | {"type": "choice", "instructions": "Which type of operation is the next step: click, type, or select?", "criteria": {"click": "Click an element.", "type": "Type text into a field.", "select": "Select an option from a dropdown."}} | {"type": 1.0} | [{"type": "image", "path": "media/gui/multimodal_mind2web/47/471067f73824e13fab8c1bee.jpg", "w": 1280, "h": 1080}] | {"item_id": "504c0c6b-7e78-4bfa-ae3f-00f8e59c3693/04f7c7bb-0def-4780-aea6-e6171f06625a", "cut_step": 11, "judge": "recorded_action", "flags": [], "m2w_split": "train", "annotation_id": "504c0c6b-7e78-4bfa-ae3f-00f8e59c3693", "step": 11} | |
multimodal_mind2web/0f63c624-6097-473e-ad19-59bc139836d1/8971ff26-7b5c-4b17-be3d-006f780b3657/action_type/0 | multimodal_mind2web | action_type | train | openrail | null | {"platform": "web browser (real website; one viewport-sized window of the page is shown)", "task": "Search for developer jobs in Dallas, Texas, and review details of the latest job then create an 8-day alert after signing in.", "website": "aa", "previous_actions": ["[link] We're hiring! Join our team , Opens another s... | {"type": "choice", "instructions": "Which type of operation is the next step: click, type, or select?", "criteria": {"click": "Click an element.", "type": "Type text into a field.", "select": "Select an option from a dropdown."}} | {"type": 1.0} | [{"type": "image", "path": "media/gui/multimodal_mind2web/c8/c820c7c55ecba4bbb6f68a45.jpg", "w": 1280, "h": 1080}] | {"item_id": "0f63c624-6097-473e-ad19-59bc139836d1/8971ff26-7b5c-4b17-be3d-006f780b3657", "cut_step": 3, "judge": "recorded_action", "flags": [], "m2w_split": "train", "annotation_id": "0f63c624-6097-473e-ad19-59bc139836d1", "step": 3} | |
amex/9a6b6b2f76fc4d67bd5c2d9a9f2c6f48/step17/action_type/0 | amex | action_type | train | cc-by-4.0 | null | {"platform": "Android phone", "instruction": "Open Reddit. Search \"Science\". Join the community. Sort the posts by recent. Save the first post. Open the third post. Sort the comments by new.", "previous_actions": ["SWIPE from (278, 1076) to (242, 515)", "SWIPE from (350, 1152) to (411, 269)", "TAP at (50, 1087)", "TA... | {"type": "choice", "instructions": "Given the instruction, the actions so far and the current screen, which type of action is next?", "criteria": {"tap": "Tap an element on the screen.", "swipe": "Swipe (scroll) the screen.", "type": "Type text into the focused field.", "press_enter": "Press the Enter key.", "press_bac... | {"tap": 1.0} | [{"type": "image", "path": "media/gui/amex/4b/4b5f85a3049ed35b236e4d2f.jpg", "w": 606, "h": 1280}] | {"item_id": "9a6b6b2f76fc4d67bd5c2d9a9f2c6f48/step17", "cut_step": 17, "judge": "recorded_action", "flags": []} | |
amex/b0f5c858e91543abbef8dceca94249d9/step20/next_action/0 | amex | next_action | train | cc-by-4.0 | null | {"platform": "Android phone", "instruction": "Open Booking.com. Search hotels in Rome, from 1 July to 7 July. 2 guests in 1 room. $400 to $800. Filter with Pleasant rating, free WiFi, gym and breakfast included. Sort by lowest price.", "previous_actions": ["SWIPE from (556, 543) to (0, 554)", "TAP at (487, 329)", "TAP ... | {"type": "choice", "instructions": "Given the instruction, the actions so far and the current screen, which element is the right one to tap next?", "criteria": {"el_1": "'Central Station (11)', center at about (295, 351) in the screenshot", "el_2": "'Fitness center (6)', center at about (295, 1123) in the screenshot"... | {"el_9": 1.0} | [{"type": "image", "path": "media/gui/amex/a6/a6c7f65c54624ddf05431c86.jpg", "w": 591, "h": 1280}] | {"item_id": "b0f5c858e91543abbef8dceca94249d9/step20", "cut_step": 20, "judge": "recorded_action", "flags": [], "n_options": 17} | |
rules/distill/contract_clause/942/q1 | onejev/distill | contract_clause | train | apache-2.0 | Qwen/Qwen3.5-27B | {"contract_title": "Master Service Agreement: Northern Edge Networks (TEN) v. Quantis Edge Solutions", "parties": {"provider": {"name": "Quantis Edge Solutions Inc.", "address": "4200 Data Highway, Sector 7, Austin, TX", "primary_contact": "Elena Rostova", "role": "Head of Interconnect Operations"}, "client": {"name": ... | {"type": "noul", "instructions": "Did the client's backhaul connection packet loss exceed the 0.1% threshold required to maintain the latency guarantee under Section 4.2?", "criteria": {"true": "The observed packet loss was 0.18%, which is greater than the 0.1% limit, voiding the guarantee regardless of the provider's ... | {"yes": 0.998971, "no": 0.0010289999999999466} | null | {"judge": "llm:Qwen3.5-27B distill batch 1"} | |
multimodal_mind2web/54112d86-1d85-4abf-9e12-86f526d314c2/8949caa0-b7f1-48f7-9c16-6303d8e5139e/action_type/0 | multimodal_mind2web | action_type | train | openrail | null | {"platform": "web browser (real website; one viewport-sized window of the page is shown)", "task": "What are the BGG rules for Game Submissions?", "website": "boardgamegeek", "previous_actions": ["[button] Help -> CLICK"], "screen": "<image:1>"} | {"type": "choice", "instructions": "Given the task and the current page, is the next action a click, a text entry, or a dropdown selection?", "criteria": {"click": "Click an element.", "type": "Type text into a field.", "select": "Select an option from a dropdown."}} | {"click": 1.0} | [{"type": "image", "path": "media/gui/multimodal_mind2web/c0/c00e62f0a3d2d863c8937b3a.jpg", "w": 1280, "h": 1080}] | {"item_id": "54112d86-1d85-4abf-9e12-86f526d314c2/8949caa0-b7f1-48f7-9c16-6303d8e5139e", "cut_step": 1, "judge": "recorded_action", "flags": [], "m2w_split": "train", "annotation_id": "54112d86-1d85-4abf-9e12-86f526d314c2", "step": 1} |
End of preview. Expand in Data Studio
The training data of OneJev: 94,707 typed questions about screens, photos, videos and text, each with its answer. This is 95.5% of the rows OneJev was trained on; rows whose sources do not allow redistribution are left out.
Use
from datasets import load_dataset
ds = load_dataset("OmniJev/OneJev-Data", split="train", streaming=True)
row = next(iter(ds))
Each row has a state with <image:N> and <video:N> placeholders, a question, its target, and the pictures in
images. To train OneJev on it, see GitHub.
Sources
127 public datasets. The largest in each group:
| Group | Sources |
|---|---|
| GUI agent runs | AMEX, Multimodal-Mind2Web, OSWorld-Verified, AgentRewardBench, MisActBench and 4 more |
| Images | GQA, ImageRewardDB, GenAI-Bench, FGVC-Aircraft, CLEVR and 30 more |
| Short videos | VideoFeedback, STAR, MSR-VTT, Diving48, CLEVRER and 3 more |
| Long procedural videos | Breakfast, CaptainCook4D |
| Text | SWE-agent trajectories, DTap-Bench, SWE-bench experiments, Terminal-Bench 2 and 66 more |
| Rules | typed-decisions-synth, jevlite and our own set written with Qwen3.5-27B |
All 127 with links and licenses are in licenses.csv.
License
Each row keeps the license of its source, given in the license column and in licenses.csv. Some
sources allow research or non-commercial use only. To have a source removed, open an issue on
GitHub.
- Downloads last month
- -