Latest Posts

  • DBRX: A New State-of-the-Art Open LLM by Databricks


    DBRX: A New State-of-the-Art Open LLM by Databricks

    DBRX utilizes a transformer-based decoder-only architecture with a fine-grained Mixture-of-Experts (MoE) design. This means it uses a large number of smaller expert models to process different parts of the input, rather than relying on a single massive model.

  • Fine-tune an Instruct model over raw text data


    Fine-tune an Instruct model over raw text data

    This experiment seeks to discover a lighter approach that navigates between the constraints of a 128K context window and the complexities of a model fine-tuned on billions of tokens, perhaps more in the realm of tens of millions of tokens. For a smaller-scale test, I’ll fine-tune Mistral’s 7B Instruct v0.2 model on The Guardian’s manage-frontend…

  • Crew AI Tutorial


    Crew AI Tutorial

    In the realm of artificial intelligence, the adoption of multi-agent systems (MAS) via crew ai represents a paradigm shift towards more dynamic and complex problem-solving capabilities. This blog dives into the essence of Multi Agent Systems, highlighting the necessity for such systems in today’s technological landscape and exploring the CrewAI framework as a possible solution.

  • Fine Tuning LLMs on a Single Consumer Graphic Card


    Fine Tuning LLMs on a Single Consumer Graphic Card

    Unlock the secrets of fine-tuning large language models (LLMs) on a single consumer graphics card. Discover how to leverage your GPU for enhanced AI model performance without breaking the bank. Ideal for AI enthusiasts and developers seeking practical insights.

  • LangChain Cheatsheet — All Secrets on a Single Page


    LangChain Cheatsheet — All Secrets on a Single Page

    Unlock the full potential of LangChain with our ultimate cheatsheet! Discover all the secrets and tips you need, condensed into a single, easy-to-navigate page. Save time and boost your efficiency today!

  • TaCo: Enhancing Cross-Lingual Transfer for Low-Resource Languages in LLM


    TaCo: Enhancing Cross-Lingual Transfer for Low-Resource Languages in LLM

    TaCo, short for Translation-Assisted Cross-Linguality, utilizes a clever combination of translation and chain-of-thought processes to efficiently train LLMs on new languages. This blog post delves into the details of TaCo, exploring its approach, datasets, training process, and evaluation results.

  • RAFT: Finetuning LLM for Domain-Specific Knowledge


    RAFT: Finetuning LLM for Domain-Specific Knowledge

    This blog post dives into RAFT (Retrieval-Augmented Fine-Tuning), a novel training recipe that enhances the ability of LLMs to perform well in domain-specific RAG tasks. We’ll explore the challenges of adapting LLMs to specialized domains, delve into the details of RAFT, and analyze its effectiveness through evaluation results and qualitative examples.

  • Open Source Techniques Unlock Hidden Potential in LLM


    Open Source Techniques Unlock Hidden Potential in LLM

    Matt Shumer, CEO of Hyperwrite AI, introduced the “Claude Opus to Haiku” technique, which allows users to achieve Claude 3 Opus-level quality at a fraction of the cost and latency. This open-source method involves providing Claude 3 Opus with a task description and a single input-output example.

  • MoonDream: A Tiny Vision Language Model for Edge Devices


    MoonDream: A Tiny Vision Language Model for Edge Devices

    MoonDream is a small, efficient vision language model (VLM) that can operate with less than 6GB of memory. This makes it ideal for use on edge devices in a variety of applications, such as: