together-ai

Introducing Together AI's New Embedding Models: From 151M to 350M Parameters

Tal Peretz

Sep 26, 2024 — 2 min read

Together AI continues to push the boundaries of AI technology with their latest offering: the Together AI/together-ai-embedding-151m-to-350m models. These new language models are designed to provide efficient and cost-effective solutions for embedding tasks, making them ideal for a wide range of applications.

One of the standout features of these models is their pricing structure. With an input price of just $0.016 per 1M tokens and an output price of $0, they offer an affordable solution for businesses of all sizes. Additionally, these models support a maximum number of tokens, further enhancing their versatility.

Together Inference Engine 2.0

The introduction of the Together Inference Engine 2.0 marks a significant leap in AI capabilities. This upgraded engine boasts a decoding throughput that is 4x faster than open-source vLLM and outperforms many commercial solutions by 1.3x to 2.5x. The engine includes the Together Turbo and Together Lite endpoints, allowing users to choose between high performance and cost-efficiency. Together Turbo matches the quality of full-precision FP16 models, while Together Lite is optimized for speed and cost-effectiveness.

Together Enterprise Platform

The Together Enterprise Platform empowers businesses to train, fine-tune, and run inference on any model, whether in the cloud or on-premise. This platform offers:

2-3x faster inference and up to 50% lower operational costs.
Continuous model optimization techniques such as auto fine-tuning and adaptive speculators.
Flexibility in deployment, ensuring data remains secure within the organization's firewall.

Model Support and Optimization

Supporting over 200 models, including leading families like Llama and custom models, the platform employs advanced optimization techniques like speculative decoding and optimized kernels (e.g., FlashAttention-3). These innovations ensure that the models deliver superior performance and efficiency.

Data Privacy and Model Ownership

Data privacy and model ownership are paramount in the Together Enterprise Platform. Organizations retain complete control over their models and proprietary data, meeting the strictest privacy and compliance policies. This focus on security makes the platform a trusted choice for enterprises.

While the specific Together AI/together-ai-embedding-151m-to-350m models are part of a broader strategy to enhance performance, cost-efficiency, and data privacy, they exemplify Together AI's commitment to providing cutting-edge AI solutions.

Unleashing Creativity with Stability.SD3-Large-V1:0 on Amazon Bedrock

The advent of Stability.SD3-Large-V1:0, a revolutionary text-to-image model, marks a significant milestone in the realm of generative AI. Integrated into Amazon Bedrock, this model is designed to transform how digital assets are created, offering unparalleled performance and quality in image generation. Model Architecture and Performance At the heart

Unlocking the Potential of Bedrock/Meta-Textgeneration-Llama-2-7B with AWS SageMaker

The Bedrock/Meta-Textgeneration-Llama-2-7B is a cutting-edge Large Language Model (LLM) developed by Meta as part of the Llama 2 series. This series ranges from 7 billion to 70 billion parameters, with the 7B model offering a balance of size and capability. Its development involved Supervised Fine-Tuning (SFT) and Reinforcement Learning

Exploring Azure AI's Phi-3-Medium-128K-Instruct: A Revolutionary LLM

Azure AI's Phi-3-Medium-128K-Instruct is a breakthrough in the field of language models, bringing forth a dense decoder-only Transformer model with 14 billion parameters. This model, part of Microsoft's Phi-3 family, stands out for its impressive performance across a wide range of benchmarks, making it a go-to

Unlocking the Potential of Azure AI's Phi-3.5-MoE-Instruct: A New Era in Artificial Intelligence

The Azure AI Phi-3.5-MoE-Instruct model, a groundbreaking addition to Microsoft's Phi-3.5 series, is revolutionizing the landscape of artificial intelligence with its innovative architecture and impressive capabilities. Model Architecture At the heart of the Phi-3.5-MoE-Instruct is the Mixture-of-Experts (MoE) model architecture, which includes 16 experts, each