Olmo 2 0325 32B

🚀 OLMo 2 32B

OLMo 2 32B is the largest model in the OLMo 2 family, designed to advance the science of language models and released with all associated code and training details.

🚀 Quick Start

OLMo 2 32B can be used with the standard HuggingFace transformers library. First, ensure you have the appropriate version of the transformers library installed, and then follow the inference code examples to start using the model.

✨ Features

• Large-scale Model: OLMo 2 32B is the largest in the OLMo 2 family, trained on a large amount of data to provide powerful language processing capabilities.
• Open-source: All code, checkpoints, logs, and associated training details are released on GitHub, facilitating community participation and research.
• Multiple Training Stages: Through multiple stages of training, including pre-training, mid-training, and post-training, the performance and generalization ability of the model are continuously improved.

📦 Installation

OLMo 2 32B is supported in transformers v4.48 or higher:

pip install transformers>=4.48

If using vLLM, you will need to install from the main branch until v0.7.4 is released.

💻 Usage Examples

Basic Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
olmo = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0325-32B")
tokenizer = AutoTokenizer.from_pretrained("allenai/OLMo-2-0325-32B")
message = ["Language modeling is "]
inputs = tokenizer(message, return_tensors='pt', return_token_type_ids=False)
# optional verifying cuda
# inputs = {k: v.to('cuda') for k,v in inputs.items()}
# olmo = olmo.to('cuda')
response = olmo.generate(**inputs, max_new_tokens=100, do_sample=True, top_k=50, top_p=0.95)
print(tokenizer.batch_decode(response, skip_special_tokens=True)[0])
>> 'Language modeling is  a key component of any text-based application, but its effectiveness...'

Advanced Usage

For faster performance, you can quantize the model using the following method:

import torch
AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0325-32B",
    torch_dtype=torch.float16,
    load_in_8bit=True)  # Requires bitsandbytes

The quantized model is more sensitive to data types and CUDA operations. To avoid potential issues, it's recommended to pass the inputs directly to CUDA using:

inputs.input_ids.to('cuda')

To load a specific model revision with HuggingFace, simply add the argument revision:

olmo = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0325-32B", revision="step250000-tokens2098B")

Or, you can access all the revisions for the models via the following code snippet:

from huggingface_hub import list_repo_refs
out = list_repo_refs("allenai/OLMo-2-0325-32B")
branches = [b.name for b in out.branches]

Fine-tuning

Model fine-tuning can be done from the final checkpoint (the main revision of this model) or many intermediate checkpoints. Two recipes for tuning are available.

1. Fine-tune with the OLMo-core repository:

torchrun --nproc-per-node=8 ./src/scripts/official/OLMo2-0325-32B-train.py run01

You can override most configuration options from the command-line. For example, to override the learning rate you could launch the script like this:

torchrun --nproc-per-node=8 ./src/scripts/train/OLMo2-0325-32B-train.py run01 --train_module.optim.lr=6e-3

For more documentation, see the GitHub readme.

2. Further fine-tuning support is being developing in AI2's Open Instruct repository. Details are here.

📚 Documentation

Model Description

Model Sources

• Project Page: https://allenai.org/olmo
• Repositories:
  • Core repo (training, inference, fine-tuning etc.): https://github.com/allenai/OLMo-core
  • Evaluation code: https://github.com/allenai/OLMo-Eval
  • Further fine-tuning code: https://github.com/allenai/open-instruct
• Paper: https://arxiv.org/abs/2501.00656

Evaluation

Core model results for OLMo 2 32B are found below.

• Columns ARC/C through NQ represent metrics tracked during OLMo 2 development.
• Columns AGIEval through TriviaQA represent unseen evals.

Model Details

Training

Stage 1: Initial Pretraining

• Dataset: OLMo-mix-1124 (3.9T tokens)
• Coverage: 95%+ of total pretraining budget
• 32B Model: ~1.5 epoch

Stage 2: Mid-training

• Dataset: Dolmino-Mix-1124
• Two training mixes:
  • 100B tokens
  • 300B tokens
• Mix composition: 50% high-quality web data + academic/Q&A/instruction/math content

Model Merging

• 32B Model: 3 versions on 100B mix + 1 version on 300B mix, merged for final checkpoint

🔧 Technical Details

The OLMo 2 32B model is a Transformer style autoregressive language model. It is pre-trained on a large-scale dataset and goes through multiple training stages to continuously improve its performance. The model uses a series of techniques such as quantization to improve inference speed, and provides different checkpoints and fine-tuning methods to meet different application scenarios.

📄 License

The code and model are released under Apache 2.0.

Bias, Risks, and Limitations

⚠️ Important Note

Like any base or fine-tuned language model, AI can be prompted by users to generate harmful and sensitive content. Such content may also be produced unintentionally, especially in cases involving bias, so we recommend that users consider the risks when applying this technology. Additionally, many statements from OLMo or any LLM are often inaccurate, so facts should be verified.

Citation

@misc{olmo20242olmo2furious,
      title={{2 OLMo 2 Furious}},
      author={Team OLMo and Pete Walsh and Luca Soldaini and Dirk Groeneveld and Kyle Lo and Shane Arora and Akshita Bhagia and Yuling Gu and Shengyi Huang and Matt Jordan and Nathan Lambert and Dustin Schwenk and Oyvind Tafjord and Taira Anderson and David Atkinson and Faeze Brahman and Christopher Clark and Pradeep Dasigi and Nouha Dziri and Michal Guerquin and Hamish Ivison and Pang Wei Koh and Jiacheng Liu and Saumya Malik and William Merrill and Lester James V. Miranda and Jacob Morrison and Tyler Murray and Crystal Nam and Valentina Pyatkin and Aman Rangapur and Michael Schmitz and Sam Skjonsberg and David Wadden and Christopher Wilhelm and Michael Wilson and Luke Zettlemoyer and Ali Farhadi and Noah A. Smith and Hannaneh Hajishirzi},
      year={2024},
      eprint={2501.00656},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2501.00656},
}

Model Card Contact

For errors in this model card, contact olmo@allenai.org.