Multi-Model Passive Storage Archive Compilation
This public repository serves as a centralized passive remote storage structure hosting a collection of isolated state-of-the-art weights blocks. Every architecture system is deployed under dedicated sub-directories to maintain structural data sorting.
Architecture Folders & Author Credits:
faster-whisper-turbo/- Base Model:
deepdml/faster-whisper-large-v3-turbo-ct2 - Type: CTranslate2 ASR Engine.
- Credits: OpenAI / deepdml.
- Base Model:
LTX-2.3-fp8/- Base Model:
Lightricks/LTX-2.3-fp8 - Type: Compressed 8-bit Multimodal Video Diffusion.
- Credits: Lightricks (LTX.io).
- Base Model:
LTX-2.3/- Base Model:
Lightricks/LTX-2.3 - Type: High-Fidelity 22B Video-Audio Generative Base.
- Credits: Lightricks (LTX.io).
- Base Model:
DeepSeek-R1-FP8/- Base Model:
deepseek-ai/DeepSeek-R1 - Type: Mixture-of-Experts FP8 Native 642B Reasoner.
- Credits: deepseek-ai.
- Base Model:
Passive Compliance Terms:
- Intended Use: Strictly allocated for long-term remote personal mirror backup operations.
- Intellectual Property Rights: All architecture algorithms, weights optimizations, benchmarks, and data formats belong exclusively to their respective training organizations. This collection complies with non-commercial mirror distributions.
Install from pip and serve model
# Install vLLM from pip: pip install vllm# Start the vLLM server: vllm serve "riytdxc43/vediolargemodels"# Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "riytdxc43/vediolargemodels", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'