Add Predict2.5 inferences in gallery with examples #32
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Add gallery.md with Text2World, Image2World, and Video2World examples - Include sample prompts, inputs, JSON structures, and results - Add all required asset files (images and videos)
jingyijin2
reviewed
Nov 9, 2025
Collaborator
jingyijin2
left a comment
jingyijin2
left a comment
There was a problem hiding this comment.
Would really like to see more diversity and great results. Will tag people from multiple groups in search of good examples.
|
|
||
| ## Model Details | ||
|
|
||
| | Model Name | Prompt | Input | JSON Structure | Results | |
Collaborator
There was a problem hiding this comment.
This 6-column table is difficult to navigate. Let's make a table with 2-3 columns with just:
- Prompt
- Input Media (None for t2w, the image for i2w, and the video for v2w)
- Results
Thank you!
| | Text2World | The video is shot from a pedestrian's perspective, showing a matte black van swiftly turning right at a stop sign. It is dusk, windy, and there is a cyclist approaching the intersection, waiting for the van to complete its turn. | Text prompt only | `{"inference_type": "text2world", "name": "TextOutput_3", "prompt": "The video is shot from a pedestrian's perspective, showing a matte black van swiftly turning right at a stop sign. It is dusk, windy, and there is a cyclist approaching the intersection, waiting for the van to complete its turn."}` |  | | ||
| | Text2World | The video is shot from the driver's view, showing several black SUVs steadily proceeding onto the highway. It is morning, bright and clear. A motorcycle merges into the lane ahead, and another vehicle signals to enter from an adjacent on-ramp. | Text prompt only | `{"inference_type": "text2world", "name": "TextOutput_4", "prompt": "The video is shot from the driver's view, showing several black SUVs steadily proceeding onto the highway. It is morning, bright and clear. A motorcycle merges into the lane ahead, and another vehicle signals to enter from an adjacent on-ramp."}` |  | | ||
| | Text2World | The video is shot from the driver's view, showing many red trucks steadily driving down a steep slope. It is evening, cloudy and windy. There is a car approaching from behind, and fallen leaves cover parts of the road, reducing traction and requiring careful control. | Text prompt only | `{"inference_type": "text2world", "name": "TextOutput_1", "prompt": "The video is shot from the driver's view, showing many red trucks steadily driving down a steep slope. It is evening, cloudy and windy. There is a car approaching from behind, and fallen leaves cover parts of the road, reducing traction and requiring careful control."}` |  | | ||
| | Image2World | The video is taken from the perspective of a vehicle's dashboard camera, showing the view of the road ahead. The sky is clear and blue, indicating good weather conditions. The road is lined with green trees and bushes, adding a touch of nature to the urban setting. There are multiple vehicles on the road, including cars and trucks, all moving in the same direction. The vehicles vary in color, with shades of red, blue, and white being prominent. The traffic lights are visible, showing a green light, indicating that it is safe to proceed. The video does not show any accidents or unusual events, just a typical day on the road. | Text prompt +  | `{"inference_type": "image2world", "input_path": "I1_input.jpg", "guidance": 7, "seed": 0, "prompt": "The video is taken from the perspective of a vehicle's dashboard camera, showing the view of the road ahead. The sky is clear and blue, indicating good weather conditions. The road is lined with green trees and bushes, adding a touch of nature to the urban setting. There are multiple vehicles on the road, including cars and trucks, all moving in the same direction. The vehicles vary in color, with shades of red, blue, and white being prominent. The traffic lights are visible, showing a green light, indicating that it is safe to proceed. The video does not show any accidents or unusual events, just a typical day on the road."}` | [IO1](samples/assets/IO1_output.mp4) | |
Collaborator
There was a problem hiding this comment.
Since the guidance value and seed are all the same for the examples, could we add this information before the table that those are the default values used across the examples?
|
|
||
| | Model Name | Prompt | Input | JSON Structure | Results | | ||
| |------------|--------|-------|----------------|---------| | ||
| | Text2World | The video is shot from a pedestrian's perspective, showing two departing black trucks quickly turn left at an intersection. It is afternoon, with fog reducing visibility. Nearby, another vehicle is attempting to make a right turn while a pedestrian waits to cross. | Text prompt only | `{"inference_type": "text2world", "name": "TextOutput_1", "prompt": "The video is shot from a pedestrian's perspective, showing two departing black trucks quickly turn left at an intersection. It is afternoon, with fog reducing visibility. Nearby, another vehicle is attempting to make a right turn while a pedestrian waits to cross."}` |  | |
Collaborator
There was a problem hiding this comment.
Please use these tags to play the video directly instead of including as links:
|
|
||
| Lets see some examples of above models with their details such as prompts, inputs, etc. | ||
|
|
||
| ## Model Details |
Collaborator
There was a problem hiding this comment.
Please split the examples into 3 subsections for t2w, i2w, and v2w.
added 5 commits
November 10, 2025 13:44
- Seperate the examples to view it clearly - Add JSON example files for all models (Text2World, Image2World, Video2World)
- Seperate the examples for models - Replace image syntax  with HTML5 <video> tags - Enables inline video playback on GitHub instead of broken thumbnails - Text2World: All output videos now playable (T1-T5) - Image2World: All output videos now playable (IO1-IO5) - Video2World: Both input and output videos now playable (V1-V5, VO1-VO5) - Fix table formatting: remove double pipes throughout - Update JSON example files
- Set all videos to width 4096px for high-quality display - Convert Image2World input images from markdown to HTML img tags - Standardize all media sizes across the gallery
- Restructured gallery.md to match NVIDIA Cosmos Cookbook style - Added Base Parameters sections for all three models (Text2World, Image2World, Video2World) - Added inference_type and name fields to parameter configurations
- Reorganize common parameters with proper table formatting - Standardize parameter display across Text2World, Image2World, and Video2World
mpatekarhub
commented
Dec 4, 2025
Collaborator
Author
mpatekarhub
left a comment
mpatekarhub
left a comment
There was a problem hiding this comment.
Hello
Given feedback is checked & accordinly changes are made.
I can see that its visible locally.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Brief description of the changes in this PR.
Type of Change
Changes Made
Testing
Checklist
Additional Notes
Any additional information that reviewers should know.