|
--- |
|
license: other |
|
license_name: apple-sample-code-license |
|
license_link: LICENSE |
|
--- |
|
|
|
# OpenELM |
|
|
|
*Sachin Mehta, Mohammad Hossein Sekhavat, Qingqing Cao, Maxwell Horton, Yanzi Jin, Chenfan Sun, Iman Mirzadeh, Mahyar Najibi, Dmitry Belenko, Peter Zatloukal, Mohammad Rastegari* |
|
|
|
We introduce **OpenELM**, a family of **Open**-source **E**fficient **L**anguage **M**odels. We release both pretrained and instruction tuned models with 270M, 450M, 1.1B and 3B parameters. |
|
|
|
Our pre-training dataset contains RefinedWeb, deduplicated PILE, a subset of RedPajama, and a subset of Dolma v1.6, totaling approximately 1.8 trillion tokens. |
|
|
|
|
|
|
|
## Usage |
|
|
|
Below we provide an example of loading the model via [HuggingFace Hub](https://huggingface.co/docs/hub/) as: |
|
|
|
```python |
|
import torch |
|
from transformers import AutoTokenizer, AutoModelForCausalLM |
|
# obtain access to "meta-llama/Llama-2-7b-hf", then see https://huggingface.co/docs/hub/security-tokens to get a token |
|
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b-hf", token="hf_xxxx") |
|
|
|
model_path = "apple/OpenELM-450M" |
|
|
|
model = AutoModelForCausalLM.from_pretrained(model_path, trust_remote_code=True) |
|
model = model.cuda().eval() |
|
prompt = "Once upon a time there was" |
|
tokenized_prompt = tokenizer(prompt) |
|
prompt_tensor = torch.tensor(tokenized_prompt["input_ids"], device="cuda").unsqueeze(0) |
|
output_ids = model.generate(prompt_tensor, max_new_tokens=256, repetition_penalty=1.2, pad_token_id=0) |
|
output_ids = output_ids[0].tolist() |
|
output_text = tokenizer.decode(output_ids, skip_special_tokens=True) |
|
print(f'{model_path=}, {prompt=}\n') |
|
print(output_text) |
|
|
|
# below is the output: |
|
""" |
|
model_path='apple/OpenELM-450M', prompt='Once upon a time there was' |
|
|
|
Once upon a time there was a little girl who lived in the woods. She had a big heart and she loved to play with her friends. One day, she decided to go for a walk in the woods. As she walked, she saw a beautiful tree. It was so tall that it looked like a mountain. The tree was covered with leaves and flowers. |
|
The little girl thought that this tree was very pretty. She wanted to climb up to the tree and see what was inside. So, she went up to the tree and climbed up to the top. She was very excited when she saw that the tree was full of beautiful flowers. She also |
|
""" |
|
``` |
|
|
|
|
|
## Main Results |
|
|
|
### Zero-Shot |
|
|
|
| **Model Size** | **ARC-c** | **ARC-e** | **BoolQ** | **HellaSwag** | **PIQA** | **SciQ** | **WinoGrande** | **Average** | |
|
|-----------------------------------------------------------------------------|-----------|-----------|-----------|---------------|-----------|-----------|----------------|-------------| |
|
| [OpenELM-270M](https://huggingface.co/apple/OpenELM-270M) | 26.45 | 45.08 | **53.98** | 46.71 | 69.75 | **84.70** | **53.91** | 54.37 | |
|
| [OpenELM-270M-Instruct](https://huggingface.co/apple/OpenELM-270M-Instruct) | **30.55** | **46.68** | 48.56 | **52.07** | **70.78** | 84.40 | 52.72 | **55.11** | |
|
| [OpenELM-450M](https://huggingface.co/apple/OpenELM-450M) | 27.56 | 48.06 | 55.78 | 53.97 | 72.31 | 87.20 | 58.01 | 57.56 | |
|
| [OpenELM-450M-Instruct](https://huggingface.co/apple/OpenELM-450M-Instruct) | **30.38** | **50.00** | **60.37** | **59.34** | **72.63** | **88.00** | **58.96** | **59.95** | |
|
| [OpenELM-1_1B](https://huggingface.co/apple/OpenELM-1_1B) | 32.34 | **55.43** | 63.58 | 64.81 | **75.57** | **90.60** | 61.72 | 63.44 | |
|
| [OpenELM-1_1B-Instruct](https://huggingface.co/apple/OpenELM-1_1B-Instruct) | **37.97** | 52.23 | **70.00** | **71.20** | 75.03 | 89.30 | **62.75** | **65.50** | |
|
| [OpenELM-3B](https://huggingface.co/apple/OpenELM-3B) | 35.58 | 59.89 | 67.40 | 72.44 | 78.24 | **92.70** | 65.51 | 67.39 | |
|
| [OpenELM-3B-Instruct](https://huggingface.co/apple/OpenELM-3B-Instruct) | **39.42** | **61.74** | **68.17** | **76.36** | **79.00** | 92.50 | **66.85** | **69.15** | |
|
|
|
### LLM360 |
|
|
|
| **Model Size** | **ARC-c** | **HellaSwag** | **MMLU** | **TruthfulQA** | **WinoGrande** | **Average** | |
|
|-----------------------------------------------------------------------------|-----------|---------------|-----------|----------------|----------------|-------------| |
|
| [OpenELM-270M](https://huggingface.co/apple/OpenELM-270M) | 27.65 | 47.15 | 25.72 | **39.24** | **53.83** | 38.72 | |
|
| [OpenELM-270M-Instruct](https://huggingface.co/apple/OpenELM-270M-Instruct) | **32.51** | **51.58** | **26.70** | 38.72 | 53.20 | **40.54** | |
|
| [OpenELM-450M](https://huggingface.co/apple/OpenELM-450M) | 30.20 | 53.86 | **26.01** | 40.18 | 57.22 | 41.50 | |
|
| [OpenELM-450M-Instruct](https://huggingface.co/apple/OpenELM-450M-Instruct) | **33.53** | **59.31** | 25.41 | **40.48** | **58.33** | **43.41** | |
|
| [OpenELM-1_1B](https://huggingface.co/apple/OpenELM-1_1B) | 36.69 | 65.71 | **27.05** | 36.98 | 63.22 | 45.93 | |
|
| [OpenELM-1_1B-Instruct](https://huggingface.co/apple/OpenELM-1_1B-Instruct) | **41.55** | **71.83** | 25.65 | **45.95** | **64.72** | **49.94** | |
|
| [OpenELM-3B](https://huggingface.co/apple/OpenELM-3B) | 42.24 | 73.28 | **26.76** | 34.98 | 67.25 | 48.90 | |
|
| [OpenELM-3B-Instruct](https://huggingface.co/apple/OpenELM-3B-Instruct) | **47.70** | **76.87** | 24.80 | **38.76** | **67.96** | **51.22** | |
|
|
|
|
|
### OpenLLM Leaderboard |
|
|
|
| **Model Size** | **ARC-c** | **CrowS-Pairs** | **HellaSwag** | **MMLU** | **PIQA** | **RACE** | **TruthfulQA** | **WinoGrande** | **Average** | |
|
|-----------------------------------------------------------------------------|-----------|-----------------|---------------|-----------|-----------|-----------|----------------|----------------|-------------| |
|
| [OpenELM-270M](https://huggingface.co/apple/OpenELM-270M) | 27.65 | **66.79** | 47.15 | 25.72 | 69.75 | 30.91 | **39.24** | **53.83** | 45.13 | |
|
| [OpenELM-270M-Instruct](https://huggingface.co/apple/OpenELM-270M-Instruct) | **32.51** | 66.01 | **51.58** | **26.70** | **70.78** | 33.78 | 38.72 | 53.20 | **46.66** | |
|
| [OpenELM-450M](https://huggingface.co/apple/OpenELM-450M) | 30.20 | **68.63** | 53.86 | **26.01** | 72.31 | 33.11 | 40.18 | 57.22 | 47.69 | |
|
| [OpenELM-450M-Instruct](https://huggingface.co/apple/OpenELM-450M-Instruct) | **33.53** | 67.44 | **59.31** | 25.41 | **72.63** | **36.84** | **40.48** | **58.33** | **49.25** | |
|
| [OpenELM-1_1B](https://huggingface.co/apple/OpenELM-1_1B) | 36.69 | **71.74** | 65.71 | **27.05** | **75.57** | 36.46 | 36.98 | 63.22 | 51.68 | |
|
| [OpenELM-1_1B-Instruct](https://huggingface.co/apple/OpenELM-1_1B-Instruct) | **41.55** | 71.02 | **71.83** | 25.65 | 75.03 | **39.43** | **45.95** | **64.72** | **54.40** | |
|
| [OpenELM-3B](https://huggingface.co/apple/OpenELM-3B) | 42.24 | **73.29** | 73.28 | **26.76** | 78.24 | **38.76** | 34.98 | 67.25 | 54.35 | |
|
| [OpenELM-3B-Instruct](https://huggingface.co/apple/OpenELM-3B-Instruct) | **47.70** | 72.33 | **76.87** | 24.80 | **79.00** | 38.47 | **38.76** | **67.96** | **55.73** | |
|
|
|
See the technical report for more results and comparison. |
|
|
|
## Evaluation |
|
|
|
### Setup |
|
|
|
Install the following dependencies: |
|
|
|
```bash |
|
|
|
# install public lm-eval-harness |
|
|
|
harness_repo="public-lm-eval-harness" |
|
git clone https://github.com/EleutherAI/lm-evaluation-harness ${harness_repo} |
|
cd ${harness_repo} |
|
# use main branch on 03-15-2024, SHA is dc90fec |
|
git checkout dc90fec |
|
pip install -e . |
|
cd .. |
|
|
|
# 66d6242 is the main branch on 2024-04-01 |
|
pip install datasets@git+https://github.com/huggingface/datasets.git@66d6242 |
|
pip install tokenizers>=0.15.2 transformers>=4.38.2 sentencepiece>=0.2.0 |
|
|
|
``` |
|
|
|
### Evaluate OpenELM |
|
|
|
```bash |
|
|
|
# OpenELM-270M |
|
hf_model=OpenELM-270M |
|
|
|
# this flag is needed because lm-eval-harness set add_bos_token to False by default, but OpenELM uses LLaMa tokenizer which requires add_bos_token to be True |
|
add_bos_token=True |
|
batch_size=1 |
|
|
|
mkdir lm_eval_output |
|
|
|
shot=0 |
|
task=arc_challenge,arc_easy,boolq,hellaswag,piqa,race,winogrande,sciq,truthfulqa_mc2 |
|
lm_eval --model hf \ |
|
--model_args pretrained=${hf_model},trust_remote_code=True,add_bos_token=${add_bos_token} \ |
|
--tasks ${task} \ |
|
--device cuda:0 \ |
|
--num_fewshot ${shot} \ |
|
--output_path ./lm_eval_output/${hf_model//\//_}_${task//,/_}-${shot}shot \ |
|
--batch_size ${batch_size} 2>&1 | tee ./lm_eval_output/eval-${hf_model//\//_}_${task//,/_}-${shot}shot.log |
|
|
|
shot=5 |
|
task=mmlu,winogrande |
|
lm_eval --model hf \ |
|
--model_args pretrained=${hf_model},trust_remote_code=True,add_bos_token=${add_bos_token} \ |
|
--tasks ${task} \ |
|
--device cuda:0 \ |
|
--num_fewshot ${shot} \ |
|
--output_path ./lm_eval_output/${hf_model//\//_}_${task//,/_}-${shot}shot \ |
|
--batch_size ${batch_size} 2>&1 | tee ./lm_eval_output/eval-${hf_model//\//_}_${task//,/_}-${shot}shot.log |
|
|
|
shot=25 |
|
task=arc_challenge,crows_pairs_english |
|
lm_eval --model hf \ |
|
--model_args pretrained=${hf_model},trust_remote_code=True,add_bos_token=${add_bos_token} \ |
|
--tasks ${task} \ |
|
--device cuda:0 \ |
|
--num_fewshot ${shot} \ |
|
--output_path ./lm_eval_output/${hf_model//\//_}_${task//,/_}-${shot}shot \ |
|
--batch_size ${batch_size} 2>&1 | tee ./lm_eval_output/eval-${hf_model//\//_}_${task//,/_}-${shot}shot.log |
|
|
|
shot=10 |
|
task=hellaswag |
|
lm_eval --model hf \ |
|
--model_args pretrained=${hf_model},trust_remote_code=True,add_bos_token=${add_bos_token} \ |
|
--tasks ${task} \ |
|
--device cuda:0 \ |
|
--num_fewshot ${shot} \ |
|
--output_path ./lm_eval_output/${hf_model//\//_}_${task//,/_}-${shot}shot \ |
|
--batch_size ${batch_size} 2>&1 | tee ./lm_eval_output/eval-${hf_model//\//_}_${task//,/_}-${shot}shot.log |
|
|
|
``` |
|
|
|
|
|
## Bias, Risks, and Limitations |
|
|
|
Our OpenELM models are not trained with any safety guarantees, the model outputs can be potentially inaccurate, harmful or contain biased information. produce inaccurate, biased or other objectionable responses to user prompts. Therefore, users and developers should conduct extensive safety testing and filtering suited to their specific needs. |
|
|
|
|
