Quantization made by Richard Erkhov.

tulu-2-dpo-13b - GGUF

Model creator: https://huggingface.co/allenai/
Original model: https://huggingface.co/allenai/tulu-2-dpo-13b/

Name	Quant method	Size
tulu-2-dpo-13b.Q2_K.gguf	Q2_K	4.52GB
tulu-2-dpo-13b.IQ3_XS.gguf	IQ3_XS	4.99GB
tulu-2-dpo-13b.IQ3_S.gguf	IQ3_S	5.27GB
tulu-2-dpo-13b.Q3_K_S.gguf	Q3_K_S	5.27GB
tulu-2-dpo-13b.IQ3_M.gguf	IQ3_M	5.57GB
tulu-2-dpo-13b.Q3_K.gguf	Q3_K	5.9GB
tulu-2-dpo-13b.Q3_K_M.gguf	Q3_K_M	5.9GB
tulu-2-dpo-13b.Q3_K_L.gguf	Q3_K_L	6.45GB
tulu-2-dpo-13b.IQ4_XS.gguf	IQ4_XS	6.54GB
tulu-2-dpo-13b.Q4_0.gguf	Q4_0	6.86GB
tulu-2-dpo-13b.IQ4_NL.gguf	IQ4_NL	6.9GB
tulu-2-dpo-13b.Q4_K_S.gguf	Q4_K_S	6.91GB
tulu-2-dpo-13b.Q4_K.gguf	Q4_K	7.33GB
tulu-2-dpo-13b.Q4_K_M.gguf	Q4_K_M	7.33GB
tulu-2-dpo-13b.Q4_1.gguf	Q4_1	7.61GB
tulu-2-dpo-13b.Q5_0.gguf	Q5_0	8.36GB
tulu-2-dpo-13b.Q5_K_S.gguf	Q5_K_S	8.36GB
tulu-2-dpo-13b.Q5_K.gguf	Q5_K	8.6GB
tulu-2-dpo-13b.Q5_K_M.gguf	Q5_K_M	8.6GB
tulu-2-dpo-13b.Q5_1.gguf	Q5_1	9.1GB
tulu-2-dpo-13b.Q6_K.gguf	Q6_K	9.95GB
tulu-2-dpo-13b.Q8_0.gguf	Q8_0	12.88GB

Original model description:

model-index: - name: tulu-2-dpo-13b results: [] datasets: - HuggingFaceH4/ultrafeedback_binarized - allenai/tulu-v2-sft-mixture language: - en base_model: meta-llama/Llama-2-13b-hf license: other license_name: ai2-impact-license-low-risk license_link: https://allenai.org/impact-license

Model Card for Tulu V2 DPO 13B

Tulu is a series of language models that are trained to act as helpful assistants. Tulu V2 DPO 13B is a fine-tuned version of Llama 2 that was trained on on a mix of publicly available, synthetic and human datasets using Direct Preference Optimization (DPO). This model is a strong alternative to Llama 2 13b Chat.

For more details, read the paper: Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2 .

Model description

Model type: A model belonging to a suite of instruction and RLHF tuned chat models on a mix of publicly available, synthetic and human-created datasets.
Language(s) (NLP): Primarily English
License: AI2 ImpACT Low-risk license.
Finetuned from model: meta-llama/Llama-2-13b-hf

Model Sources

Repository: https://github.com/allenai/https://github.com/allenai/open-instruct
DPO Recipe: The DPO recipe is from the Zephyr Beta model
Model Family: Other models and the dataset are found in the Tulu V2 collection.

Performance

Model	Size	Alignment	MT-Bench (score)	AlpacaEval (win rate %)
Tulu-v2-7b 🐪	7B	SFT	6.30	73.9
Tulu-v2-dpo-7b 🐪	7B	DPO	6.29	85.1
Tulu-v2-13b 🐪	13B	SFT	6.70	78.9
Tulu-v2-dpo-13b 🐪	13B	DPO	7.00	89.5
Tulu-v2-70b 🐪	70B	SFT	7.49	86.6
Tulu-v2-dpo-70b 🐪	70B	DPO	7.89	95.1

Input Format

The model is trained to use the following format (note the newlines):

<|user|>
Your message here!
<|assistant|>

For best results, format all inputs in this manner. Make sure to include a newline after <|assistant|>, this can affect generation quality quite a bit.

Intended uses & limitations

The model was initially fine-tuned on a filtered and preprocessed of the Tulu V2 mix dataset, which contains a diverse range of human created instructions and synthetic dialogues generated primarily by other LLMs. We then further aligned the model with a Jax DPO trainer built on EasyLM on the openbmb/UltraFeedback dataset, which contains 64k prompts and model completions that are ranked by GPT-4.

Bias, Risks, and Limitations

The Tulu models have not been aligned to generate safe completions within the RLHF phase or deployed with in-the-loop filtering of responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so). It is also unknown what the size and composition of the corpus was used to train the base Llama 2 models, however it is likely to have included a mix of Web data and technical sources like books and code. See the Falcon 180B model card for an example of this.

Training hyperparameters

The following hyperparameters were used during DPO training:

learning_rate: 5e-07
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 3.0

Citation

If you find Tulu 2 is useful in your work, please cite it with:

@misc{ivison2023camels,
      title={Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2}, 
      author={Hamish Ivison and Yizhong Wang and Valentina Pyatkin and Nathan Lambert and Matthew Peters and Pradeep Dasigi and Joel Jang and David Wadden and Noah A. Smith and Iz Beltagy and Hannaneh Hajishirzi},
      year={2023},
      eprint={2311.10702},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Model card adapted from Zephyr Beta