Blog
-
Meta Llama 3 70B Model
Meta Llama 3 is out now and available on Replicate. This language model family comes in 8B and 70B parameter sizes, with context windows of 8K tokens. The models beat most other open source models on industry benchmarks and are licensed for commercial use. Llama 3 models include base and instruction-tuned models. You can chat…
-
Data Agents
Data Agents, empowered by LLMs are knowledge workers within Llamalndex, designed to interact with various types of data. These agents can handle both unstructured and structured data, significantly enhancing the capabilities beyond traditional query engines. Data Agents High Level Overview Functionality Data Agents can autonomously conduct searches and retrieve information across unstructured, semi-structured, and structured…
-
CodeLLama vs GPT-4
Recently Meta unveiled CodeLLama, a 70-billion-parameter open-source large language model (LLM). This release marks a significant milestone as it introduces the largest and most effective iteration of their LLM to date. CodeLLama is available in three distinct versions: the base model, the Python-specific model finely tuned for Python coding, and the Instruct model, which is…
-
MetaGPT
MetaGPT (Hong et al.) is an awesome multi-agent framework (ICLR 2024) that models agents as a software company following a structured SOP – PMs, architects, engineers, and more communicate with each other to solve the task at hand. LLamaIndex team excited to feature RAG-enhanced MetaGPT, powered by @llama_index modules – this allows agents to tap…
-
Creating Knowledge Graphs from Documents Using LlamaParse
Instead of doing naive text splitting, extract a document knowledge graph to power your advanced RAG pipeline. This tutorial by Fanghua Yu shows you a novel use case of LlamaParse – extract it into structured markdown that you can then convert into a document graph that you can store in a graph db “neo4j”. Once…
-
LlamaIndex + MistralAI Cookbooks
LlamaIndex is a data framework for LLM-based applications which benefit from context augmentation. LlamaIndex provides the essential abstractions to more easily ingest, structure, and access private or domain-specific data in order to inject these safely and reliably into LLMs for more accurate text generation. Here developers offer some cookbooks for utilizing LlamaIndex with MistralAI’s LLM’s and…
-
Llama 2 in MLPerf Inference
Introducing Llama 2 70B in MLPerf Inference v4.0. In the MLPerf Inference v4.0 cycle, the working group revisited the concept of a “larger” LLM task and established a new task force for this purpose. This task force evaluated a range of potential models for inclusion, including GPT-175B, Falcon-40B, Falcon-180B, BLOOMZ, and Llama 2 70B. After…
-
LLama-2 Explained
In this article, we will delve into LLama-2: what it is, how it is trained, its comparison to GPT variants and ChatGPT, as well as its architecture and code. We aim to offer a comprehensive understanding of this technological advancement. Introduction to LLama-2 LLama-2 is a suite of pre-trained language models, while LLama2 Chat represents…
-
LLama 3 Coming Soon
Today brought some electrifying news from Mark Zuckerberg, the CEO of Meta, who has announced the ongoing training of LLAMA 3. In a brief video, Zuckerberg shared insights into Meta’s AI research efforts, emphasizing the convergence of these efforts to achieve long-term goals in general intelligence. He articulated a vision of open-sourcing and widely distributing…