Andromeda: Ultra-Fast and Ultra-Intelligent SOTA Language Model ๐๐
Welcome to Andromeda, The Fastest, Most Creative, and Reliable Language Model Ever Built, train your own verison, conduct inference, and finetune your own verison with simple plug in and play scripts get started in 10 seconds, and:
- ๐ผ Handle Ultra Long Sequences (32,000-200,000+ context lengths)
- โก Ultra Fast Processing (32,000+ tokens in under 100ms)
- ๐ Superior Reasoning Capabilities
๐ Updates
- READY FOR TRAINING, help us with the strategy!
- And, here is the WANDB link to watch Andromeda train live!
Appreciation
- All the creators in Agora, Join Agora the community of AI engineers changing the world with their creations.
- LucidRains for inspiring me to devote myself to open source AI
Hiring
We're hiring: Engineers, Researchers, Interns, And, Customer Success Professionals to work on democratizing Andromeda, email me at with your story kye@apac.ai
๐ป Usage
There are two methods to use Andromeda
pip install TheBestLLMEver
git clone https://github.com/kyegomez/Andromeda.git
For detailed instructions, refer to the Training SOP and Documentation.
Method 1
To get started:
- Clone the repository and install the required packages:
git clone https://github.com/kyegomez/Andromeda
cd Andromeda
pip3 install -r requirements.txt
cd Andromeda
python3 train.py
For further instructions, refer to the Training SOP.
๐ Training
Set the environment variables:
ENTITY_NAME
: Your wandb project nameOUTPUT_DIR
: Directory to save the weights (e.g.,./weights
)MASTER_ADDR
: For distributed trainingMASTER_PORT
For master port distributed trainingRANK
- Number of nodes servicesWORLD_SIZE
Number of gpus
Configure the training:
- Accelerate Config
- Enable Deepspeed 3
- Accelerate launch train_distributed_accelerate.py
For more information, refer to the Training SOP.
๐๏ธ Dataset Building
To preprocess a
different dataset similar to the C4 or Falcon dataset used during training, use the build_dataset.py
script. This script pre-tokenizes the data, chunks it into blocks of a specified sequence length, and uploads it to the Huggingface hub.
Example command:
python3 Andromeda/build_dataset.py --seed 42 --seq_len 8192 --hf_account "HUGGINGFACE APIKEY" --tokenizer "EleutherAI/gpt-neox-20b" --dataset_name "EleutherAI/the_pile_deduplicated"
๐ Why Andromeda?
Andromeda offers several advantages:
- Andromeda offers reliable processing of 100,000+ sequence lengths extremely fast under 300ms
- Andromeda's dataset strategy was crafted with atomic precision and attention to detail for creativity and quantitative reasoning.
- Andromeda is extremely intelligent with the ability to think like a poet or make API Calls to your favorite apps.
For detailed information about the model architecture and methods, refer to the Model Architecture documentation.
๐ฏ Andromeda Principles
- Efficiency: Optimize with techniques like attention flashing, rotary position encodings, and deep normalization.
- Flexibility: Adapt to various tasks and domains for wide applications.
- Scalability: Designed to scale with resources and data sizes.
- Community-Driven: Thrives on contributions from the open-source community.
๐ Get Involved
We're just at the beginning of our journey. As we continue to develop and refine Andromeda, we invite you to join us. Whether you're a developer, researcher, or simply an enthusiast, your insights and contributions can help shape the future of Andromeda.
๐ค Contributing to Andromeda
We are thrilled to invite you to be a part of the Andromeda project. This is not just an open-source project but a community initiative, and we value your expertise and creativity. To show our appreciation, we have instituted a unique rewards system that directly compensates contributors from the revenue generated by the Andromeda API.
๐ Why Contribute
Contributing to Andromeda not only enhances your skills and profile but also comes with financial rewards. When you contribute code, documentation, or any form of improvement to the Andromeda project, you are adding value. As such, we believe it's only fair that you share in the rewards.
๐ฐ Rewards Program
Here's how the Andromeda Rewards Program works:
Submit a Pull Request: This can be a code enhancement, bug fix, documentation update, new feature, or any improvement to the project.
Review and Approval: Our team will review your contribution. If it gets approved and merged, you become eligible for the rewards program.
Revenue Share: Once your pull request is merged, you will receive a percentage of the revenue generated by the Andromeda API. The percentage will be determined based on the significance and impact of your contribution.
This means you're not just contributing to an open-source project; you're becoming a part of the Andromeda ecosystem. Your efforts can yield ongoing benefits as the Andromeda API grows and evolves.
๐ Becoming a Paid API
As part of our growth strategy, we will be deploying Andromeda as a Paid API. The revenue generated from this API will not only sustain and further the project but also fund the rewards program. If you contribute anything to make Andromeda, you will receive recurring revenue from paid API requests!
๐ How to Start Contributing
If you're ready to become a part of Andromeda and contribute to the future of multimodal embeddings, here's what you need to do:
Fork the repository.
Make your improvements or additions in your forked repository.
Submit a pull request detailing the changes you've made.
Our team will review your submission. If it's approved, it will be merged into the main repository, and you will become part of the Andromeda Rewards Program.
Thank you for considering contributing to Andromeda. Your expertise and commitment to this project are what make it thrive. Let's build the future of multimodal embeddings together.
๐บ๏ธ Roadmap
Training phase: Train Andromeda on a large-scale dataset to achieve SOTA performance in various natural language processing tasks.
World-class inference infrastructure: Establish a robust and efficient infrastructure that leverages techniques such as:
- Model quantization: Reduce memory and computational requirements without significant loss in performance.
- Distillation: Train smaller, faster models that retain the knowledge of the larger model.
- Optimized serving frameworks: Deploy Andromeda using efficient serving frameworks, such as NVIDIA Triton or TensorFlow Serving, for rapid inference.
Continuous improvement: Continuously fine-tune Andromeda on diverse data sources and adapt it to new tasks and domains.
Community-driven development: Encourage open-source contributions, including pre-processing improvements, advanced training techniques, and novel use cases.
๐ Benchmarks
Speed
- Andromeda utilizes one of the most reliable Attentions ever, flash attention 2.0 Triton. It consumes 50x less memory than GPT-3 and 10x less than LLAMA.
- We can speed this up even more with dynamic sparse flash attention 2.0.
๐ฎ Join the Journey
We're just getting started, and we invite you to join the journey. Let's revolutionize the NLP landscape together! ๐๐
- Join Agora and work with 2,000+ AI Engineers to implement all new features.
- Provide compute and help train Andromeda.
- Share the message on how we're liberating this superintelligent AI and seizing the power from the corrupt, providing it back to you.