基于DiscoPOP对公共和合成数据集混合进行微调的7B参数GPT-like模型
21 拉取 2个月前更新
2个月前更新
2个月前
524c42220d3b · 6.3GB
Readme
DiscoPOP-zephyr-7b-gemma
此模型是在argilla/dpo-mix-7k数据集上对HuggingFaceH4/zephyr-7b-gemma-sft-v0.1进行微调的版本。
此模型来自论文“Discovering Preference Optimization Algorithms with and for Large Language Models”
在这里阅读关于它的博客文章!
在这里查看生成它的代码库: https://github.com/SakanaAI/DiscoPOP
模型描述
这个模型在训练上与HuggingFaceH4/zephyr-7b-gemma-v0.1相同,但使用的是DiscoPOP替代直接偏好优化(DPO)。
DiscoPOP是我们发现偏好优化的算法,定义如下
def log_ratio_modulated_loss(
self,
policy_chosen_logps: torch.FloatTensor,
policy_rejected_logps: torch.FloatTensor,
reference_chosen_logps: torch.FloatTensor,
reference_rejected_logps: torch.FloatTensor,
) -> torch.FloatTensor:
pi_logratios = policy_chosen_logps - policy_rejected_logps
ref_logratios = reference_chosen_logps - reference_rejected_logps
logits = pi_logratios - ref_logratios
# Modulate the mixing coefficient based on the log ratio magnitudes
log_ratio_modulation = torch.sigmoid(logits)
logistic_component = -F.logsigmoid(self.beta * logits)
exp_component = torch.exp(-self.beta * logits)
# Blend between logistic and exponential component based on log ratio modulation
losses = logistic_component * (1 - log_ratio_modulation) + exp_component * log_ratio_modulation
return losses
训练超参数
以下是在训练期间使用的超参数
- learning_rate: 5e-07
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2
框架版本
- Transformers 4.40.1
- Pytorch 2.1.2+cu121
- 数据集 2.19.0
- 分词器 0.19.1
Zephyr 7B Gemma 的模型卡片
Zephyr 是一系列经过训练,可以作为有帮助的助手的语言模型。Zephyr 7B Gemma 是该系列的第三个模型,是通过对google/gemma-7b
进行微调得到的版本,该模型在公开可用和合成数据集的混合上进行训练,并使用了直接偏好优化(DPO)。您可以通过在Alignment Handbook中提供的配方来重现该模型的训练。
模型描述
- 模型类型: 在公开可用和合成数据集混合上微调的 7B 参数类似的 GPT 模型。
- 语言(NLP): 主要为英语
- 许可证: Gemma 使用条款
- 微调自模型: google/gemma-7b
模型来源
- 仓库: https://github.com/huggingface/alignment-handbook
- 演示: https://hugging-face.cn/spaces/HuggingFaceH4/zephyr-7b-gemma-chat
性能
模型 | MT Bench⬇️ | IFEval |
---|---|---|
zephyr-7b-gemma-v0.1 | 7.81 | 28.76 |
zephyr-7b-beta | 7.34 | 43.81 |
google/gemma-7b-it | 6.38 | 38.01 |
模型 | AGIEval | GPT4All | TruthfulQA | BigBench | Average ⬇️ |
---|---|---|---|---|---|
zephyr-7b-beta | 37.52 | 71.77 | 55.26 | 39.77 | 51.08 |
zephyr-7b-gemma-v0.1 | 34.22 | 66.37 | 52.19 | 37.10 | 47.47 |
mlabonne/Gemmalpaca-7B | 21.6 | 40.87 | 44.85 | 30.49 | 34.45 |
google/gemma-7b-it | 21.33 | 40.84 | 41.70 | 30.25 | 33.53 |
AGIEval、GPT4All、TruthfulQA、BigBench 的详细情况
### AGIEval | Task |Version| Metric |Value| |Stderr| |------------------------------|------:|--------|----:|---|-----:| |agieval_aqua_rat | 0|acc |21.65|± | 2.59| | | |acc_norm|25.20|± | 2.73| |agieval_logiqa_en | 0|acc |34.72|± | 1.87| | | |acc_norm|35.94|± | 1.88| |agieval_lsat_ar | 0|acc |19.57|± | 2.62| | | |acc_norm|21.74|± | 2.73| |agival_lsat_lr | 0|acc |30.59|± | 2.04| | | |acc_norm|32.55|± | 2.08| |agival_lsat_rc | 0|acc |49.07|± | 3.05| | | |acc_norm|42.75|± | 3.02| |agival_sat_en | 0|acc |54.85|± | 3.48| | | |acc_norm|53.40|± | 3.48| |agival_sat_en_without_passage| 0|acc |37.38|± | 3.38| | | |acc_norm|33.98|± | 3.31| |agival_sat_math | 0|acc |30.91|± | 3.12| | | |acc_norm|28.18|± | 3.04| 平均值:34.22% ### GPT4All | Task |Version| Metric |Value| |Stderr| |-------------|------:|------|----:|---|-----:| |arc_challenge| 0|acc |49.15|± | 1.46| | | |acc_norm|52.47|± | 1.46| |arc_easy | 0|acc |77.44|± | 0.86| | | |acc_norm|74.75|± | 0.89| |boolq | 1|acc |79.69|± | 0.70| | | |acc_norm|78.00|± | 0.41| |hellaswag | 0|acc |60.59|± | 0.49| | | |acc_norm|78.00|± | 0.41| |openbookqa | 0|acc |29.20|± | 2.04| | | |acc_norm|37.80|± | 2.17| |piqa | 0|acc |76.82|± | 0.98| | | |acc_norm|77.80|± | 0.97| |winogrande | 0|acc |64.09|± | 1.35| | | |acc_norm|64.09|± | 1.35| 平均值:66.37% ### TruthfulQA | Task |Version|Metric|Value| |Stderr| |-------------|------:|------|----:|---|-----:| |truthfulqa_mc| 1|mc1 |35.74|± | 1.68| | | |mc2 |52.19|± | 1.59| 平均值:52.19% ### Bigbench | Task |Version| Metric |Value| |Stderr| |------------------------------------------------|------:|---------------------|----:|---|-----:| |bigbench_causal_judgement | 0|multiple_choice_grade|53.68|± | 3.63| |bigbench_date_understanding | 0|multiple_choice_grade|59.89|± | 2.55| |bigbench_disambiguation_qa | 0|multiple_choice_grade|30.23|± | 2.86| |bigbench_geometric_shapes | 0|multiple_choice_grade|11.42|± | 1.68| | | |exact_str_match | 0.00|± | 0.00| |bigbench_logic_deduction_five_objects | 0|multiple_choice_grade|28.40|± | 2.02| |bigbench_logic_deduction_seven_objects | 0|multiple_choice_grade|19.14|± | 1.49| |bigbench_logic_deduction_three_objects | 0|multiple_choice_grade|44.67|± | 2.88| |bigbench_movie_recommendation | 0|multiple_choice_grade|26.80|± | 1.98| |bigbench_navigate | 0|multiple_choice_grade|50.00|± | 1.58| |bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|52.75|± | 1.12| |bigbench_ruins_names | 0|multiple_choice_grade|33.04|± | 2.22| |bigbench_salient_translation_error_detection | 0|multiple_choice_grade|33.37|± | 1.49| |bigbench_snarks | 0|multiple_choice_grade|48.62|± | 3.73| |bigbench_sports_understanding | 0|multiple_choice_grade|58.11|± | 1.57| |bigbench_temporal_sequences | 0|multiple_choice_grade|37.20|± | 1.53| |bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|20.08|± | 1.13| |bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|15.77|± | 0.87| |bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|44.67|± | 2.88| 平均值:37.1%预期用途及限制
该模型最初在包含ChatGPT生成的多种合成对话的DEITA 10K数据集上进行了微调。
然后我们进一步使用🤗 TRL的 DPOTrainer
在argilla/dpo-mix-7k数据集上对模型进行了对齐,该数据集包含7k个提示和模型完成,这些完成由GPT-4进行排序。因此,该模型可用于聊天,您可以查看我们的演示以测试其功能。
以下是使用🤗 Transformers中的pipeline()
函数运行该模型的方法:
# pip install transformers>=4.38.2
# pip install accelerate
import torch
from transformers import pipeline
pipe = pipeline(
"text-generation",
model="HuggingFaceH4/zephyr-7b-gemma-v0.1",
device_map="auto",
torch_dtype=torch.bfloat16,
)
messages = [
{
"role": "system",
"content": "", # Model not yet trained for follow this
},
{"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
]
outputs = pipe(
messages,
max_new_tokens=128,
do_sample=True,
temperature=0.7,
top_k=50,
top_p=0.95,
stop_sequence="<|im_end|>",
)
print(outputs[0]["generated_text"][-1]["content"])
# It is not possible for a human to eat a helicopter in one sitting, as a
# helicopter is a large and inedible machine. Helicopters are made of metal,
# plastic, and other materials that are not meant to be consumed by humans.
# Eating a helicopter would be extremely dangerous and would likely cause
# serious health problems, including choking, suffocation, and poisoning. It is
# important to only eat food that is safe and intended for human consumption.
偏见、风险和局限性
Zephyr 7B Gemma在RLHF阶段尚未与人类偏好对齐以确保安全性,也没有像ChatGPT那样部署具有在线过滤响应的功能,因此该模型可能会生成问题输出(尤其是在被提示生成时)。此外,还不知道用于训练基础模型(《google/gemma-7b》)的语料库的大小和组成,但它很可能包括Web数据以及书籍和代码等技术来源的混合。参见StarCoder2模型信息以获取示例。
培训与评估数据
此模型是在argilla/dpo-mix-7k数据集上对HuggingFaceH4/zephyr-7b-gemma-sft-v0.1进行微调的版本。
它在评估集中实现了以下结果:
- 损失:0.4695
- 奖励/选择:-3.3746
- 奖励/拒绝:-4.9715
- 奖励/准确度:0.7188
- 奖励/边际:1.5970
- 对数似然/拒绝:-459.4853
- 对数似然/选择:-429.9115
- 原始对数/拒绝:86.4684
- 原始对数/选择:92.8200
训练超参数
以下是在训练期间使用的超参数
- learning_rate: 5e-07
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2
培训结果
培训损失 | 时代 | 步数 | 验证损失 | 奖励/选择 | 奖励/拒绝 | 奖励/准确度 | 奖励/边际 | 对数似然/拒绝 | 对数似然/选择 | 原始对数/拒绝 | 原始对数/选择 |
---|---|---|---|---|---|---|---|---|---|---|---|
0.1923 | 1.9 | 100 | 0.4736 | -3.4575 | -4.9556 | 0.75 | 1.4980 | -459.1662 | -431.5707 | 86.3863 | 92.7360 |
框架版本
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- 数据集 2.14.6
- 分词器 0.15.1
引用信息
如果您在您的工作中发现此模型有用,请考虑引用Zephyr技术报告
@misc{tunstall2023zephyr,
title={Zephyr: Direct Distillation of LM Alignment},
author={Lewis Tunstall and Edward Beeching and Nathan Lambert and Nazneen Rajani and Kashif Rasul and Younes Belkada and Shengyi Huang and Leandro von Werra and Clémentine Fourrier and Nathan Habib and Nathan Sarrazin and Omar Sanseviero and Alexander M. Rush and Thomas Wolf},
year={2023},
eprint={2310.16944},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
您还希望引用该模型的创建者
@misc{zephyr_7b_gemma,
author = {Lewis Tunstall and Philipp Schmid},
title = {Zephyr 7B Gemma},
year = {2024},
publisher = {Hugging Face},
journal = {Hugging Face repository},
howpublished = {\url{https://hugging-face.cn/HuggingFaceH4/zephyr-7b-gemma-v0.1}}
}
开放LLM排行榜评估结果
指标 | 数值 |
---|---|
Avg. | 62.41 |
AI2推理挑战(25-Shot) | 58.45 |
HellaSwag(10-Shot) | 83.48 |
MMLU(5-Shot) | 60.68 |
TruthfulQA(0-shot) | 52.07 |
Winogrande(5-shot) | 74.19 |
GSM8k(5-shot) | 45.56 |