RichardErkhov/UCLA-AGI_-_zephyr-7b-sft-full-SPIN-iter2-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

zephyr-7b-sft-full-SPIN-iter2 - GGUF

• Model creator: https://huggingface.co/UCLA-AGI/
• Original model: https://huggingface.co/UCLA-AGI/zephyr-7b-sft-full-SPIN-iter2/

Name	Quant method	Size
zephyr-7b-sft-full-SPIN-iter2.Q2_K.gguf	Q2_K	2.53GB
zephyr-7b-sft-full-SPIN-iter2.IQ3_XS.gguf	IQ3_XS	2.81GB
zephyr-7b-sft-full-SPIN-iter2.IQ3_S.gguf	IQ3_S	2.96GB
zephyr-7b-sft-full-SPIN-iter2.Q3_K_S.gguf	Q3_K_S	2.95GB
zephyr-7b-sft-full-SPIN-iter2.IQ3_M.gguf	IQ3_M	3.06GB
zephyr-7b-sft-full-SPIN-iter2.Q3_K.gguf	Q3_K	3.28GB
zephyr-7b-sft-full-SPIN-iter2.Q3_K_M.gguf	Q3_K_M	3.28GB
zephyr-7b-sft-full-SPIN-iter2.Q3_K_L.gguf	Q3_K_L	3.56GB
zephyr-7b-sft-full-SPIN-iter2.IQ4_XS.gguf	IQ4_XS	3.67GB
zephyr-7b-sft-full-SPIN-iter2.Q4_0.gguf	Q4_0	3.83GB
zephyr-7b-sft-full-SPIN-iter2.IQ4_NL.gguf	IQ4_NL	3.87GB
zephyr-7b-sft-full-SPIN-iter2.Q4_K_S.gguf	Q4_K_S	3.86GB
zephyr-7b-sft-full-SPIN-iter2.Q4_K.gguf	Q4_K	4.07GB
zephyr-7b-sft-full-SPIN-iter2.Q4_K_M.gguf	Q4_K_M	4.07GB
zephyr-7b-sft-full-SPIN-iter2.Q4_1.gguf	Q4_1	4.24GB
zephyr-7b-sft-full-SPIN-iter2.Q5_0.gguf	Q5_0	4.65GB
zephyr-7b-sft-full-SPIN-iter2.Q5_K_S.gguf	Q5_K_S	4.65GB
zephyr-7b-sft-full-SPIN-iter2.Q5_K.gguf	Q5_K	4.78GB
zephyr-7b-sft-full-SPIN-iter2.Q5_K_M.gguf	Q5_K_M	4.78GB
zephyr-7b-sft-full-SPIN-iter2.Q5_1.gguf	Q5_1	5.07GB
zephyr-7b-sft-full-SPIN-iter2.Q6_K.gguf	Q6_K	5.53GB

Original model description:

license: mit datasets: - UCLA-AGI/SPIN_iter2 language: - en pipeline_tag: text-generation

Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models (https://arxiv.org/abs/2401.01335)

zephyr-7b-sft-full-spin-iter2

This model is a self-play fine-tuned model at iteration 2 from alignment-handbook/zephyr-7b-sft-full using synthetic data based on on the HuggingFaceH4/ultrachat_200k dataset.

Model Details

Model Description

• Model type: A 7B parameter GPT-like model fine-tuned on synthetic datasets.
• Language(s) (NLP): Primarily English
• License: MIT
• Finetuned from model: alignment-handbook/zephyr-7b-sft-full (based on mistralai/Mistral-7B-v0.1)

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-07
• train_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 64
• optimizer: RMSProp
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 2.0

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	63.54
ARC (25-shot)	66.47
HellaSwag (10-shot)	85.82
MMLU (5-shot)	61.48
TruthfulQA (0-shot)	57.75
Winogrande (5-shot)	76.95
GSM8K (5-shot)	32.75

Citation

@misc{chen2024selfplay,
      title={Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models}, 
      author={Zixiang Chen and Yihe Deng and Huizhuo Yuan and Kaixuan Ji and Quanquan Gu},
      year={2024},
      eprint={2401.01335},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}