![Magpie](https://cdn-uploads.huggingface.co/production/uploads/653df1323479e9ebbe3eb6cc/FWWILXrAGNwWr52aghV0S.png) ## 🔥 Chat with Magpie [Here](https://huggingface.co/spaces/flydust/Chat-with-Magpie)! # 🐦 Llama-3-8B-Magpie-Align-v0.3 Project Web: [https://magpie-align.github.io/](https://magpie-align.github.io/) Online Model Demo: [https://huggingface.co/spaces/flydust/Chat-with-Magpie](https://huggingface.co/spaces/flydust/Chat-with-Magpie) Arxiv Technical Report: [https://arxiv.org/abs/2406.08464](https://arxiv.org/abs/2406.08464) Codes: [https://github.com/magpie-align/magpie](https://github.com/magpie-align/magpie) ## 🧐 About This Model This model is an aligned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B). We apply the following pipeline: We first perform SFT using: * [Magpie-Align/Magpie-Pro-MT-300K-v0.1](https://huggingface.co/datasets/Magpie-Align/Magpie-Pro-MT-300K-v0.1) * [Magpie-Align/Magpie-Reasoning-150K](https://huggingface.co/datasets/Magpie-Align/Magpie-Reasoning-150K) * [Magpie-Align/Magpie-Qwen2-Pro-200K-Chinese](https://huggingface.co/datasets/Magpie-Align/Magpie-Qwen2-Pro-200K-Chinese) * **SFT Model Checkpoint:** [Magpie-Align/Llama-3-8B-Magpie-Align-SFT-v0.3](https://huggingface.co/Magpie-Align/Llama-3-8B-Magpie-Align-SFT-v0.3) We then perform DPO on the [princeton-nlp/llama3-ultrafeedback-armorm](https://huggingface.co/datasets/princeton-nlp/llama3-ultrafeedback-armorm) dataset. The overall performance is much better than the official Llama-3-8B-Instruct Model! Plus, it can answer Chinese queries frequently, thanks to our new [Chinese instruction dataset](https://huggingface.co/datasets/Magpie-Align/Magpie-Qwen2-Pro-200K-Chinese)! - **Alpaca Eval 2 (vs GPT-4-Turbo-1106): 48.58 (LC), 50.36 (WR)** - **Alpaca Eval 2 (vs Llama-3-8B-Instruct): 73.65 (LC), 75.81 (WR)** - **Arena Hard: 42.2** - **WildBench WB-Score: 41.1** - **Zero-Eval GSM: 50.0** ## 🔥 Model Performance We compare our Llama-3-8B-Magpie-Align with official and other **open-aligned LLMs** that have been fine-tuned from base models and have publicly released their training datasets. The results are as follows: ``` +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | Aligned Model ID | MT-Bench | | | Alpaca Eval 2 | | Alpaca Eval 2 | | Arena Hard | | | | | | (GPT-4-Turbo-1106) | | (Llama-3-8B-Instruct) | | | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | | R1 | R2 | AVG | LC WR | WR | LC WR | WR | Score | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | meta-llama/Meta-Llama-3-8B-Instruct | 8.31 | 7.65 | 7.98 | 22.92 | 22.57 | 50 | 50 | 20.6 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | princeton-nlp/Llama-3-Base-8B-SFT-DPO | 8.12 | 7.23 | 7.67 | 17.71 | 15.34 | 43.73 | 38.80 | 14.8 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | NousResearch/Hermes-2-Pro-Llama-3-8B | 8.05 | 7.35 | 7.70 | 15.60 | 12.86 | 36.37 | 30.52 | 11.5 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | allenai/llama-3-tulu-2-dpo-8b | 7.71 | 7.15 | 7.43 | 14.89 | 14.8 | 35.43 | 35.42 | 11.7 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | cognitivecomputations/dolphin-2.9-llama3-8b | 7.97 | 6.98 | 7.47 | 12.50 | 8.79 | 32.67 | 22.80 | 8.2 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | openchat/openchat-3.6-8b-20240522 | 7.83 | 7.23 | 7.53 | 17.70 | 12.53 | 41.30 | 30.79 | 6.7 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | Magpie-Align/Llama-3-8B-Magpie-Align-v0.1 | 8.01 | 7.63 | 7.82 | 38.52 | 38.47 | 69.37 | 70.05 | 32.4 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | Magpie-Align/Llama-3-8B-Magpie-Align-v0.2 | 7.81 | 7.64 | 7.73 | 49.86 | 51.98 | 75.17 | 78.20 | 37.5 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ | Magpie-Align/Llama-3-8B-Magpie-Align-v0.3 | 7.82 | 7.51 | 7.67 | 48.58 | 50.36 | 73.65 | 75.81 | 42.2 | +---------------------------------------------+----------+------+------+--------------------+-------+-----------------------+-------+------------+ ``` ## 👀 Other Information **License**: Please follow [Meta Llama 3 Community License](https://llama.meta.com/llama3/license). **Conversation Template**: Please use Llama 3 **official chat template** for the best performance. **How to use it?** Please check the official [Llama 3 repository](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct#how-to-use) for detailed instructions. Simply replace the original `model_id` with `Magpie-Align/Llama-3-8B-Magpie-Align-SFT-v1.0`. The detailed training pipeline is as follows. ## Stage 1: Supervised Fine-tuning We use [Axolotl](https://github.com/OpenAccess-AI-Collective/axolotl) for SFT. ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 1 - eval_batch_size: 1 - seed: 42 - distributed_type: multi-GPU - num_devices: 4 - gradient_accumulation_steps: 32 - total_train_batch_size: 128 - total_eval_batch_size: 4 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_steps: 98 - num_epochs: 2 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:------:|:----:|:---------------:| | 0.8616 | 0.0019 | 1 | 0.8870 | | 0.5554 | 0.2013 | 106 | 0.5568 | | 0.5067 | 0.4027 | 212 | 0.5065 | | 0.4728 | 0.6040 | 318 | 0.4865 | | 0.4681 | 0.8054 | 424 | 0.4740 | | 0.4563 | 1.0067 | 530 | 0.4662 | | 0.4115 | 1.1944 | 636 | 0.4642 | | 0.3993 | 1.3957 | 742 | 0.4620 | | 0.4048 | 1.5971 | 848 | 0.4613 | | 0.4167 | 1.7984 | 954 | 0.4611 | ### Framework versions - Transformers 4.42.3 - Pytorch 2.3.1+cu121 - Datasets 2.19.1 - Tokenizers 0.19.1 *Internal name for identification: Llama-3-8B-Magpie-Mix-RC*. Please change the model name in the below Axolotl config. [

](https://github.com/axolotl-ai-cloud/axolotl)

See axolotl config

axolotl version: `0.4.1` ```yaml base_model: meta-llama/Meta-Llama-3-8B model_type: LlamaForCausalLM tokenizer_type: AutoTokenizer load_in_8bit: false load_in_4bit: false strict: false datasets: - path: Magpie-Align/Magpie-Reasoning-150K type: sharegpt conversation: llama3 - path: Magpie-Align/Magpie-Qwen2-Pro-200K-Chinese type: sharegpt conversation: llama3 - path: Magpie-Align/Magpie-Pro-MT-300K-v0.1 type: sharegpt conversation: llama3 dataset_prepared_path: last_run_prepared val_set_size: 0.001 output_dir: axolotl_out/Llama-3-8B-Magpie-Mix-RC sequence_len: 8192 sample_packing: true eval_sample_packing: false pad_to_sequence_len: true wandb_project: SynDa wandb_entity: wandb_watch: wandb_name: Llama-3-8B-Magpie-Mix-RC wandb_log_model: hub_model_id: Magpie-Align/Llama-3-8B-Magpie-Mix-RC gradient_accumulation_steps: 32 micro_batch_size: 1 num_epochs: 2 optimizer: paged_adamw_8bit lr_scheduler: cosine learning_rate: 2e-5 train_on_inputs: false group_by_length: false bf16: auto fp16: tf32: false gradient_checkpointing: true gradient_checkpointing_kwargs: use_reentrant: false early_stopping_patience: resume_from_checkpoint: logging_steps: 1 xformers_attention: flash_attention: true warmup_ratio: 0.1 evals_per_epoch: 5 eval_table_size: saves_per_epoch: 1 debug: deepspeed: weight_decay: 0.0 fsdp: fsdp_config: special_tokens: pad_token: <|end_of_text|> ```

## Stage 2: Direct Preference Optimization ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 7e-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: 1 ### Training results | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | |:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:| | 0.5813 | 0.2137 | 100 | 0.5238 | -2.6816 | -3.4539 | 0.7298 | 0.7723 | -612.4234 | -541.2933 | -1.1244 | -1.1082 | | 0.5021 | 0.4275 | 200 | 0.4483 | -3.4053 | -4.4858 | 0.8024 | 1.0805 | -715.6146 | -613.6641 | -1.1035 | -1.0844 | | 0.3802 | 0.6412 | 300 | 0.4069 | -3.7974 | -5.1705 | 0.8427 | 1.3731 | -784.0882 | -652.8716 | -1.1310 | -1.1105 | | 0.3827 | 0.8549 | 400 | 0.3872 | -4.3693 | -5.9670 | 0.8710 | 1.5976 | -863.7308 | -710.0647 | -1.1495 | -1.1283 | ### Framework versions - Transformers 4.42.3 - Pytorch 2.3.1+cu121 - Datasets 2.20.0 - Tokenizers 0.19.1

See alignment handbook config

```yaml # Customized Configs model_name_or_path: Magpie-Align/Llama-3-8B-Magpie-Align-SFT-v0.3 hub_model_id: Magpie-Align/Llama-3-8B-Magpie-Align-v0.3-RC output_dir: alignment_handbook_out/Llama-3-8B-Magpie-Align-v0.3-RC run_name: Llama-3-8B-Magpie-Align-v0.3-RC dataset_mixer: princeton-nlp/llama3-ultrafeedback-armorm: 1.0 dataset_splits: - train - test preprocessing_num_workers: 24 # DPOTrainer arguments bf16: true beta: 0.01 learning_rate: 0.7e-6 gradient_accumulation_steps: 8 per_device_train_batch_size: 2 per_device_eval_batch_size: 4 num_train_epochs: 1 max_length: 2048 max_prompt_length: 1800 warmup_ratio: 0.1 logging_steps: 1 lr_scheduler_type: cosine optim: adamw_torch torch_dtype: null use_flash_attention_2: true do_eval: true evaluation_strategy: steps eval_steps: 100 gradient_checkpointing: true gradient_checkpointing_kwargs: use_reentrant: False log_level: info push_to_hub: true save_strategy: "steps" save_steps: 100 save_total_limit: 1 seed: 42 report_to: - wandb ```

## Downstream Performance (Lighteval) | Datasets | Llama-3-8B-Magpie-Align-v0.3 | | :--- | :---: | | MMLU (5) | 65.69 | | ARC (25) | 63.23 | | HellaSwag (25) | 82.15 | | TruthfulQA (0) | 60.97 | | Winogrande (5) | 73.64 | ## Paper Abstract

Click Here

High-quality instruction data is critical for aligning large language models (LLMs). Although some models, such as Llama-3-Instruct, have open weights, their alignment data remain private, which hinders the democratization of AI. High human labor costs and a limited, predefined scope for prompting prevent existing open-source data creation methods from scaling effectively, potentially limiting the diversity and quality of public alignment datasets. Is it possible to synthesize high-quality instruction data at scale by extracting it directly from an aligned LLM? We present a self-synthesis method for generating large-scale alignment data named Magpie. Our key observation is that aligned LLMs like Llama-3-Instruct can generate a user query when we input only the left-side templates up to the position reserved for user messages, thanks to their auto-regressive nature. We use this method to prompt Llama-3-Instruct and generate 4 million instructions along with their corresponding responses. We perform a comprehensive analysis of the extracted data and select 300K high-quality instances. To compare Magpie data with other public instruction datasets, we fine-tune Llama-3-8B-Base with each dataset and evaluate the performance of the fine-tuned models. Our results indicate that in some tasks, models fine-tuned with Magpie perform comparably to the official Llama-3-8B-Instruct, despite the latter being enhanced with 10 million data points through supervised fine-tuning (SFT) and subsequent feedback learning. We also show that using Magpie solely for SFT can surpass the performance of previous public datasets utilized for both SFT and preference optimization, such as direct preference optimization with UltraFeedback. This advantage is evident on alignment benchmarks such as AlpacaEval, ArenaHard, and WildBench.

## 📚 Citation If you find the model, data, or code useful, please cite our paper: ``` @article{xu2024magpie, title={Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing}, author={Zhangchen Xu and Fengqing Jiang and Luyao Niu and Yuntian Deng and Radha Poovendran and Yejin Choi and Bill Yuchen Lin}, year={2024}, eprint={2406.08464}, archivePrefix={arXiv}, primaryClass={cs.CL} } ``` Please also cite the creators of preference datasets: SimPO paper: ``` @article{meng2024simpo, title={{SimPO}: Simple preference optimization with a reference-free reward}, author={Meng, Yu and Xia, Mengzhou and Chen, Danqi}, journal={arXiv preprint arXiv:2405.14734}, year={2024} } ``` UltraFeedback paper: ``` @article{cui2023ultrafeedback, title={{UltraFeedback}: Boosting language models with high-quality feedback}, author={Cui, Ganqu and Yuan, Lifan and Ding, Ning and Yao, Guanming and Zhu, Wei and Ni, Yuan and Xie, Guotong and Liu, Zhiyuan and Sun, Maosong}, journal={arXiv preprint arXiv:2310.01377}, year={2023} } ``` ArmoRM paper: ``` @article{wang2024interpretable, title={Interpretable Preferences via Multi-Objective Reward Modeling and Mixture-of-Experts}, author={Wang, Haoxiang and Xiong, Wei and Xie, Tengyang and Zhao, Han and Zhang, Tong}, journal={arXiv preprint arXiv:2406.12845}, year={2024} } ``` **Questions?** Please contact [Zhangchen](https://zhangchenxu.com/) by email.