Text Generation
Transformers
Safetensors
JAX
PyTorch
English
transformer
deepspeed
tensorflow
Mixture of Experts
xai
hipl
rlhf
Instructions to use ZeppelinCorp/Okamela with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ZeppelinCorp/Okamela with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="ZeppelinCorp/Okamela")# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("ZeppelinCorp/Okamela", dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use ZeppelinCorp/Okamela with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "ZeppelinCorp/Okamela" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ZeppelinCorp/Okamela", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/ZeppelinCorp/Okamela
- SGLang
How to use ZeppelinCorp/Okamela with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "ZeppelinCorp/Okamela" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ZeppelinCorp/Okamela", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "ZeppelinCorp/Okamela" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ZeppelinCorp/Okamela", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use ZeppelinCorp/Okamela with Docker Model Runner:
docker model run hf.co/ZeppelinCorp/Okamela
πΈ Okamela AI: The Future of Brain-Scale Intelligence π
π― Overview
Okamela AI is a 9.223 quintillion parameter AI model developed by Chatflare Corporation or Zeppelin Corporation, surpassing all existing AI models by leveraging the power of multimodal, multilingual, and cybersecurity-focused intelligence.
Okamela AI is designed for cutting-edge brain-scale applications and outperforms models like GPT-4, DeepSeek, and Hunyuan Large in both speed and capability.
π‘ Key Features
- π Multimodal Understanding β Processes text, images, audio, tabular data, and more.
- π Multilingual Support β Seamlessly understands 200+ languages with high accuracy.
- π§ Advanced Reasoning β Combines MoE (Mixture of Experts) and Transformer architecture.
- π‘οΈ Enhanced Security β Integrated with cybersecurity capabilities for threat detection.
- π Ultra-Fast Inference β Optimized with Fugaku, NVIDIA DGX H100, and Cerebras CS-2 hardware.
π οΈ Technical Specifications
- Parameters: 9.223 quintillion (9,223,372,036,854,775,807 parameters)
- Architecture: Mixture of Experts (MoE) + Transformer
- Training Data: Eclipse Corpuz Dataset (Multimodal and Multilingual)
- Libraries: PyTorch, DeepSpeed, and Hugging Face Transformers
- Hardware: Fugaku + NVIDIA DGX H100 + Cerebras CS-2
- Downloads last month
- -
