April 2023 Summaries

Baseten has made significant advancements in its large language model offerings with the release of Camel and StableLM, two open-source models designed for instruction completion and general-purpose applications respectively. These models are part of a growing trend of open-source large language models that are quickly gaining ground. The company is also promoting usage-based pricing for its workspaces, which eliminates monthly or annual platform fees. Additionally, Baseten is hosting AI meetups and sponsoring virtual and in-person hackathons to foster community engagement.

Comparing NVIDIA GPUs for AI workloads like fine-tuning foundation models, deploying large open-source models, and serving Large Language Models (LLMs) requires powerful GPUs with specific specs to understand when comparing cards with different architectures, core types, and memory capacity. The key specs to consider are cores, specifically CUDA cores for general-purpose computing, tensor cores optimized for machine learning calculations, and VRAM as a hard limit on model size. When selecting a GPU, price to performance is crucial, considering both the cost per minute and total cost of operation, including factors like availability, which has become increasingly scarce due to high demand. Options for scaling infrastructure vertically (increasing instance power) or horizontally (using multiple replicas of a lower-cost GPU) must also be considered. The NVIDIA T4 and A10 GPUs are two widely available options, with the T4 being less expensive but still powerful enough for many AI workloads, while the A10 offers more performance but at a higher cost per minute. Ultimately, choosing the right GPU depends on factors like model size, invocation time, and specific use cases, such as running Whisper or Stable Diffusion XL models.

Deploy StableLM with Truss Stability AI recently announced the ongoing development of the StableLM series of language models, which were released alongside a number of checkpoints for this model. These models are ideal for conversational and coding-related tasks and can be deployed using Baseten and Truss infrastructure to provide scalable and cost-efficient performance. The deployment process is made easy through the truss push command, allowing users to deploy StableLM behind a REST API for immediate use in production. With auto-scaling resources, StableLM ensures efficient and low-latency performance even in high-traffic scenarios. The models can be modified by changing the load method in model.py or by configuring GPU resources in config.yaml, and system prompts can be added for use in chatbots. Users can get started with $30 of free credits.

If You Build It, Devs will Come: How to Host an AI Meetup' discusses the benefits of hosting an in-person AI meetup, highlighting how such events can facilitate connections between developers, founders, investors, and enthusiasts, ultimately fostering a thriving community. The article provides 10 quick steps to host a successful AI meetup, including deciding on goals, selecting attendees, choosing a venue, creating an event page, promoting the event, preparing the space, welcoming guests, having fun, hustling to reach goals, and staying up late to nurture the community. By following these steps, individuals can create an engaging and effective platform for like-minded individuals to gather, network, and ideate, ultimately driving growth and success in their businesses or personal projects.