---
license: apache-2.0
language:
- en
library_name: elm
---
# SliceX AI™ ELM (Efficient Language Models)
**ELM** (which stands for **E**fficient **L**anguage **M**odels) is the first version in the series of cutting-edge language models from [SliceX AI](https://slicex.ai) that is designed to achieve the best in class performance in terms of _quality_, _throughput_ & _memory_.
ELM is designed to be a modular and customizable family of neural networks that are highly efficient and performant. Today we are sharing the first version in this series: **ELM-v0.1** models (named _Rambutan_).
_Model:_ ELM introduces a new type of _(de)-composable LLM model architecture_ along with the algorithmic optimizations required to learn (training) and run (inference) these models. At a high level, we train a single ELM model in a self-supervised manner (during pre-training phase) but once trained the ELM model can be sliced in many ways to fit different user/task needs. The optimizations can be applied to the model either during the pre-training and/or fine-tuning stage.
_Fast Inference with Customization:_ Once trained, the ELM model architecture permits flexible inference strategies at runtime depending on the deployment needs. For instance, the ELM model can be _decomposed_ into smaller slices, i.e., smaller (or larger) models can be extracted from the original model to create multiple inference endpoints. Alternatively, the original (single) ELM model can be loaded _as is_ for inference and different slices within the model can be queried directly to power faster inference. This provides an additional level of flexibility for users to make compute/memory tradeoffs depending on their application and runtime needs.
- **Blog:** [Medium](https://medium.com/sujith-ravi/introducing-elm-efficient-customizable-privacy-preserving-llms-cea56e4f727d)
- **Github:** https://github.com/slicex-ai/elm
- **Demo** (try it out): https://huggingface.co/spaces/slicexai/elm-demo-v1
- **HuggingFace** (access ELM Model cards, code & app from HF): https://huggingface.co/slicexai
## ELM-v0.1 Model Release
This repository contains code to run our ELM models. The current ELM model `elm-v0.1` (named _Rambutan_) was pre-trained (an intermediate checkpoint was used) and then instruction fine-tuned for downstream tasks.
ELM models (in the `models` folder) in this repository come in three sizes (`elm-1.0`, `elm-0.75` and `elm-0.25`). **All these different slices are extracted from the same ELM finetuned checkpoint for inference** and supports the following use-case.
- news_classification
- toxicity_detection
- news_content_generation
- news_summarization
**NOTE: ELM-v0.1 release is an early version finetuned from an intermediate pretrained checkpoint & without any KV caching, decoding optimizations, or quantization applied.**
## Setup ELM
### Download ELM repo
```bash
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/slicexai/elm-v0.1
```
### Installation
```bash
cd elm-v0.1
pip install -r requirements.txt
```
## Download ELM task-specific model checkpoints
### Install git-lfs
```bash
sudo apt-get install git-lfs
git lfs install
```
For Macbook, replace `sudo apt-get install git-lfs` with `brew install git-lfs`
(Optional) Installing git-lfs without sudo,
```bash
wget https://github.com/git-lfs/git-lfs/releases/download/v3.2.0/git-lfs-linux-amd64-v3.2.0.tar.gz
tar -xzf git-lfs-linux-amd64-v3.2.0.tar.gz
PATH=$PATH://git-lfs-3.2.0/
git lfs install
```
### Download ELM checkpoints
To download all checkpoints
```bash
git lfs pull
```
```note
NOTE: Please allow a few minutes for the full download of all model checkpoints.
```
To download elm-1.0 model checkpoints individually
```bash
git lfs pull -I models/elm-1.0_news_classification/ckpt.pt
git lfs pull -I models/elm-1.0_toxicity_detection/ckpt.pt
git lfs pull -I models/elm-1.0_news_content_generation/ckpt.pt
git lfs pull -I models/elm-1.0_news_summarization/ckpt.pt
```
To download elm-0.75 model checkpoints individually
```bash
git lfs pull -I models/elm-0.75_news_classification/ckpt.pt
git lfs pull -I models/elm-0.75_toxicity_detection/ckpt.pt
git lfs pull -I models/elm-0.75_news_content_generation/ckpt.pt
git lfs pull -I models/elm-0.75_news_summarization/ckpt.pt
```
To download elm-0.25 model checkpoints individually
```bash
git lfs pull -I models/elm-0.25_news_classification/ckpt.pt
git lfs pull -I models/elm-0.25_toxicity_detection/ckpt.pt
git lfs pull -I models/elm-0.25_news_content_generation/ckpt.pt
```
## How to use: Run ELM on a sample task (e.g., news classification)
```bash
python run.py
E.g. python run.py models/elm-0.75_news_classification
```
Prompts for the specific tasks can be found in the corresponding checkpoint directory. See an example below from `models/elm-0.75_news_classification/example_prompts.json`.
```json
{
"inputs": ["GM May Close Plant in Europe DETROIT (Reuters) - General Motors Corp. <A HREF=\"http://www.investor.reuters.com/FullQuote.aspx?ticker=GM.N target=/stocks/quickinfo/fullquote\">GM.N</A> will likely cut some jobs in Europe and may close a plant there as part of a restructuring plan under development to try to return the region to profitability, the U.S. automaker said on Wednesday."],
"template": "[INST]Below is a news article. Please classify it under one of the following classes (World, Business, Sports, Sci/Tech). Please format your response as a JSON payload.\n\n### Article: {input}\n\n### JSON Response:[/INST]"
}
```
Running the above command returns the following response
```json
{
"prompt": "[INST]Below is a news article. Please classify it under one of the following classes (World, Business, Sports, Sci/Tech). Please format your response as a JSON payload.\n\n### Article: GM May Close Plant in Europe DETROIT (Reuters) - General Motors Corp. <A HREF=\"http://www.investor.reuters.com/FullQuote.aspx?ticker=GM.N target=/stocks/quickinfo/fullquote\">GM.N</A> will likely cut some jobs in Europe and may close a plant there as part of a restructuring plan under development to try to return the region to profitability, the U.S. automaker said on Wednesday.\n\n### JSON Response:[/INST]",
"response": "{'text_label': 'Business'}"
}
```