# AetherCortex/Llama-X

Open Academic Research on Improving LLaMA to SOTA LLM

Repository: https://github.com/AetherCortex/Llama-X
Canonical: https://ross.abutalabs.com/products/llama-x
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2023-08-30T00:25:38+00:00

## Health v2 (maintenance only)
Score: 30/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 89
- inputs: {"age_days": 1252, "days_push": 1100, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1602, forks 104 (observed 2026-08-28T04:05:09.758854+00:00)

## What it is
Llama-X is an open academic research project aiming to progressively improve Meta's LLaMA into a state-of-the-art open LLM through community collaboration. It publishes code, data, model weights, and experiment details across ten research areas including instruction tuning, RLHF, long-context transformers, and multimodal modeling.

## Use cases
- improve llama into a sota open-source llm
- train instruction-following models from llama
- run rlhf experiments on open models
- research long-context transformer training
- study open academic llm research methods
- find a curated list of llm research papers

## When to choose
- you want to follow or contribute to systematic open LLM research built on LLaMA
- you need reference code, data, and experiment details for instruction tuning and RLHF
- you want a structured research roadmap covering ten LLM improvement areas

## When to avoid
- you need a production-ready, stable model or library - the project is early-stage and experimental
- you just want to run inference with a finished LLM rather than participate in research
- you require guaranteed maintenance - the last release was in 2023

## Facets
- artifact type: learning-resource
- maturity: experimental
- function: llm-training, machine-learning, deep-learning
- domain: large-language-models, deep-learning, artificial-intelligence
- platform: python
- tags: llama, open-research, instruction-tuning, rlhf, fine-tuning, academic-project, research, gpu, linux

## Member repositories
- AetherCortex/Llama-X (main) score 30

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:09.758854+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:52:33.582991+00:00, confidence not recorded.
  - readme: https://github.com/AetherCortex/Llama-X (fetched 2026-08-28T04:05:09.758854+00:00, sha fe4084bd6822)
- Data as of 2026-08-30T08:39:29.467469+00:00.