Instructions to use WaveCut/Cosmos3-Super-Text2Image-Quanto-FP8-Transformer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use WaveCut/Cosmos3-Super-Text2Image-Quanto-FP8-Transformer with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("WaveCut/Cosmos3-Super-Text2Image-Quanto-FP8-Transformer", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
Add FP8 stress example gallery
Browse files- .gitattributes +8 -0
- README.md +17 -0
- examples/02_arctic_greenhouse_night_shift_fp8.png +3 -0
- examples/03_control_room_restoration_fp8.png +3 -0
- examples/04_rain_market_cross_section_fp8.png +3 -0
- examples/06_robotic_assembly_line_signage_fp8.png +3 -0
- examples/07_kitchen_storm_chess_table_fp8.png +3 -0
- examples/08_orbital_cockpit_cyrillic_ui_fp8.png +3 -0
- examples/09_flood_command_center_fp8.png +3 -0
- examples/10_cyrillic_newspaper_press_fp8.png +3 -0
.gitattributes
CHANGED
|
@@ -37,3 +37,11 @@ examples/01_metro_archive_reading_room_fp8.png filter=lfs diff=lfs merge=lfs -te
|
|
| 37 |
examples/05_manuscript_restoration_table_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 38 |
examples/nvidia_example_caption_bf16.png filter=lfs diff=lfs merge=lfs -text
|
| 39 |
examples/nvidia_example_caption_fp8.png filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
examples/05_manuscript_restoration_table_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 38 |
examples/nvidia_example_caption_bf16.png filter=lfs diff=lfs merge=lfs -text
|
| 39 |
examples/nvidia_example_caption_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
examples/02_arctic_greenhouse_night_shift_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
examples/03_control_room_restoration_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
examples/04_rain_market_cross_section_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
examples/06_robotic_assembly_line_signage_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 44 |
+
examples/07_kitchen_storm_chess_table_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
examples/08_orbital_cockpit_cyrillic_ui_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
examples/09_flood_command_center_fp8.png filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
examples/10_cyrillic_newspaper_press_fp8.png filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -108,6 +108,23 @@ FP8 transformer output:
|
|
| 108 |
|
| 109 |

|
| 110 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 111 |
## Notes
|
| 112 |
|
| 113 |
- The upstream card documents BF16 as the tested precision. Treat this FP8 transformer as experimental.
|
|
|
|
| 108 |
|
| 109 |

|
| 110 |
|
| 111 |
+
## Stress Prompt Outputs
|
| 112 |
+
|
| 113 |
+
These are the ten FP8 outputs from the handwritten JSON-caption stress prompt set used in the benchmark table above. The set stresses Cyrillic signage, exact text placement, reflections, small-object consistency, multi-plane composition, UI panels, and human anatomy.
|
| 114 |
+
|
| 115 |
+
| # | Stress focus | FP8 output |
|
| 116 |
+
| --- | --- | --- |
|
| 117 |
+
| 01 | Metro archive reading room |  |
|
| 118 |
+
| 02 | Arctic greenhouse night shift |  |
|
| 119 |
+
| 03 | Control room restoration |  |
|
| 120 |
+
| 04 | Rain market cross section |  |
|
| 121 |
+
| 05 | Manuscript restoration table |  |
|
| 122 |
+
| 06 | Robotic assembly line signage |  |
|
| 123 |
+
| 07 | Kitchen storm chess table |  |
|
| 124 |
+
| 08 | Orbital cockpit Cyrillic UI |  |
|
| 125 |
+
| 09 | Flood command center |  |
|
| 126 |
+
| 10 | Cyrillic newspaper press |  |
|
| 127 |
+
|
| 128 |
## Notes
|
| 129 |
|
| 130 |
- The upstream card documents BF16 as the tested precision. Treat this FP8 transformer as experimental.
|
examples/02_arctic_greenhouse_night_shift_fp8.png
ADDED
|
Git LFS Details
|
examples/03_control_room_restoration_fp8.png
ADDED
|
Git LFS Details
|
examples/04_rain_market_cross_section_fp8.png
ADDED
|
Git LFS Details
|
examples/06_robotic_assembly_line_signage_fp8.png
ADDED
|
Git LFS Details
|
examples/07_kitchen_storm_chess_table_fp8.png
ADDED
|
Git LFS Details
|
examples/08_orbital_cockpit_cyrillic_ui_fp8.png
ADDED
|
Git LFS Details
|
examples/09_flood_command_center_fp8.png
ADDED
|
Git LFS Details
|
examples/10_cyrillic_newspaper_press_fp8.png
ADDED
|
Git LFS Details
|