MobileLLM

This repository contains the training code of MobileLLM introduced in our work: "MobileLLM: Optimizing Sub-billion Parameter Language Models for On-Device Use Cases", published in ICML 2024.

In this work, we comprehensively consider multiple design factors to obtain high-quality LLMs with fewer than a billion parameters. We integrated (1) SwiGLU activation function, (2) deep and thin architectures, (3) embedding sharing, (4) grouped-query attention to build MobileLLM. MobileLLM-125M/350M attains a remarkable 2.7%/4.3% accuracy boost over preceding 125M/350M SoTA models on zero-shot commonsense reasoning tasks. In our updated version, we further demonstrate that our design philosophy scales effectively to larger models, with SoTA results for MobileLLM-600M/1B/1.5B.

Citation

If you find our code useful for your research, please consider citing:

@article{liu2024mobilellm,
    title={MobileLLM: Optimizing Sub-billion Parameter Language Models for On-Device Use Cases},
    author={Liu, Zechun and Zhao, Changsheng and Iandola, Forrest and Lai, Chen and Tian, Yuandong and Fedorov, Igor and Xiong, Yunyang and Chang, Ernie and Shi, Yangyang and Krishnamoorthi, Raghuraman and others},
    journal={arXiv preprint arXiv:2402.14905},
    year={2024}
}

Run

Step 1. Requirements:

python 3.9, pytorch >= 2.0
pip install -r requirement.txt

Step 2. Data preprocessing

Dividing a tokenized dataset or tokenize your own dataset, and even distribute it across the total number of training nodes, where each node comprises 1x8 GPUs. Next, organize the data into the following structure:

basepath
- 1
  - xxx.jsonl
- 2
  - xxx.jsonl
- ...
- #nodes
  - xxx.jsonl

Each line of a jsonl file is a key-value pair of tokenized data {"token_ids": [1,2,3,4,...]}.

Our training code is compatible with the data pre-processing method in https://github.com/LLM360/amber-data-prep.

Step 3. Training script

The script pretrain.sh is provided to initiate training on a 1x8 node setup using torchrun. This script can be modified to adjust the --nnodes parameter and other settings to suit different multi-node configurations, such as those using slurm or torchx. The learning rate in the script is for 1x8 node with a batch size of 32. If you increase the number of nodes or the batch size, you need to increase the learning rate linearly.

Steps to run:

In pretrain.sh file, specify the --train_data_local_path to the pre-processed data in Step 2 and --input_model_filename to ./configs/{model_size}/.
Run bash pretrain.sh

Others

The model weights is still under legal review. If you have any questions, feel free to email (zechunliu at meta dot com) and (cszhao at meta dot com)

Training cost

It takes the following number of days to train MobileLLM on 1T tokens using 32 NVIDIA A100 80G GPUs.

125M	350M	600M	1B	1.5B
~3 days	~6 days	~8 days	~12 days	~18 days

Results on Zero-shot Common Sense Reasoning tasks

MobileLLM-125M

model	boolq	piqa	siqa	hellaswag	winogrande	arc_easy	arc_challenge	obqa	avg.
OPT-125M	41.3	25.2	57.5	62.0	41.9	31.1	31.2	50.8	42.6
GPT-neo-125M	40.7	24.8	61.3	62.5	41.9	29.7	31.6	50.7	42.9
Pythia-160M	40.0	25.3	59.5	62.0	41.5	29.9	31.2	50.9	42.5
MobileLLM-125M	43.9	27.1	60.2	65.3	42.4	38.9	39.5	53.1	46.3
MobileLLM-LS-125M	45.8	28.7	60.4	65.7	42.9	39.5	41.1	52.1	47.0

MobileLLM-350M

model	boolq	piqa	siqa	hellaswag	winogrande	arc_easy	arc_challenge	obqa	avg.
OPT-350M	41.9	25.7	54.0	64.8	42.6	36.2	33.3	52.4	43.9
Pythia-410M	47.1	30.3	55.3	67.2	43.1	40.1	36.2	53.4	46.6
MobileLLM-350M	53.8	33.5	62.4	68.6	44.7	49.6	40.0	57.6	51.3
MobileLLM-LS-350M	54.4	32.5	62.8	69.8	44.1	50.6	45.8	57.2	52.1

MobileLLM-600M

model	boolq	piqa	siqa	hellaswag	winogrande	arc_easy	arc_challenge	obqa	avg.
Qwen1.5-500M	54.7	32.1	46.9	68.9	46.0	48.8	37.7	55.0	48.8
BLOOM-560M	43.7	27.5	53.7	65.1	42.5	36.5	32.6	52.2	44.2
MobiLlama-800M	52.0	31.7	54.6	73.0	43.3	52.3	42.5	56.3	50.7
MobileLLM-600M	58.1	35.8	61.0	72.3	44.9	55.9	47.9	58.6	54.3

MobileLLM-1B

model	boolq	piqa	siqa	hellaswag	winogrande	arc_easy	arc_challenge	obqa	avg.
Pythia-1B	49.9	30.4	58.7	69.2	43.3	47.4	38.6	52.2	48.7
MobiLlama-1B	59.7	38.4	59.2	74.5	44.9	62.0	43.7	59.0	55.2
Falcon-1B	59.5	38.4	63.9	74.6	44.6	62.9	45.6	60.9	56.3
BLOOM-1.1B	47.6	27.3	58.6	67.0	42.4	42.2	36.6	53.8	46.9
TinyLlama-1.1B	59.2	37.1	58.1	72.9	43.9	59.1	44.7	58.8	54.2
MobileLLM-1B	63.0	39.0	66.7	74.4	45.0	61.4	46.8	62.3	57.3

MobileLLM-1.5B

model	boolq	piqa	siqa	hellaswag	winogrande	arc_easy	arc_challenge	obqa	avg.
GPT-neo-1.3B	51.3	33.0	61.8	70.9	43.7	48.6	41.2	54.5	50.6
OPT-1.3B	54.4	31.7	58.4	71.5	44.7	53.7	44.6	59.1	52.3
BLOOM-1.7B	50.9	31.2	61.7	70.0	43.2	47.2	36.2	56.1	49.6
Qwen1.5-1.8B	61.1	36.5	68.3	74.1	47.2	60.4	42.9	61.2	56.5
GPT-neo-2.7B	55.8	34.3	62.4	72.9	43.6	55.6	40.0	57.9	52.8
OPT-2.7B	56.6	34.6	61.8	74.5	45.6	60.2	48.2	59.6	55.1
Pythia-2.8B	59.4	38.9	66.1	73.8	44.5	59.6	45.0	59.4	55.8
BLOOM-3B	55.1	33.6	62.1	70.5	43.2	53.9	41.6	58.2	52.3
MobileLLM-1.5B	67.5	40.9	65.7	74.8	46.4	64.5	50.5	64.7	59.4

Acknowledgement

This code is partially based on Hugging Face transformer repo.

Contact

Zechun Liu, Meta Inc (zechunliu at meta dot com)

Changsheng Zhao, Meta Inc (cszhao at meta dot com)

Relevant Projects

SpinQuant: LLM Quantization with Learned Rotations [Paper] [Code]

LLM-QAT: Data-Free Quantization Aware Training for Large Language Models [Paper] [Code]

License

BiT is CC-BY-NC 4.0 licensed as of now.

facebookresearch/MobileLLM