Instructions to use easylearning/FutureWorlds with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use easylearning/FutureWorlds with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("easylearning/FutureWorlds", 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
File size: 805 Bytes
64b9802 | 1 2 3 4 5 6 7 8 9 10 | # Third-party provenance
This is a local research-code consolidation, not a public release.
- `objective.py` preserves the original ByteDance/Hugging Face Apache-2.0 header from the audited RLVR-World/verl-derived implementation.
- `reward_core.py` preserves its original upstream header. The audited upstream RLVR-World reference is commit `e1b8b6f40ca0696919ce9b8dbe965c4153bcf5f1`.
- Tokenizer and backbone loading in the archived trainers rely on iVideoGPT/RLVR-World components, Hugging Face Transformers and PyTorch. Those model assets and datasets are not included here.
- The local source manifest records file-level provenance. Before public distribution, consolidate the applicable full upstream licenses and choose the license for original additions; no blanket relicensing is implied here.
|