--- license: cc-by-nc-4.0 tags: - generated_from_trainer base_model: lightblue/suzume-llama-3-8B-multilingual model-index: - name: workspace/llm_training/axolotl/llama3-multilingual-orpo/output_mitsu_full_borda results: [] --- # Suzume ORPO

Suzume with Mitsu - a Japanese tree sparrow with honey on it

[[Paper]](https://arxiv.org/abs/2405.18952) [[Dataset]](https://huggingface.co/datasets/lightblue/mitsu) This is Suzume ORPO, an ORPO trained fine-tune of the [lightblue/suzume-llama-3-8B-multilingual](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual) model using our [lightblue/mitsu](https://huggingface.co/datasets/lightblue/mitsu) dataset. We have trained several versions of this model using ORPO and so recommend that you use the best performing model from our tests, [lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half). Note that this model has a non-commerical license as we used the Command R and Command R+ models to generate our training data for this model ([lightblue/mitsu](https://huggingface.co/datasets/lightblue/mitsu)). We are currently working on a developing a commerically usable model, so stay tuned for that! # Model list We have ORPO trained the following models using different proportions of the [lightblue/mitsu](https://huggingface.co/datasets/lightblue/mitsu) dataset: * Trained on the top/bottom responses of all prompts in the dataset: [lightblue/suzume-llama-3-8B-multilingual-orpo-borda-full](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-full) * Trained on the top/bottom responses of the prompts of the 75\% most consistently ranked responses in the dataset: [lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75) * Trained on the top/bottom responses of the prompts of the 50\% most consistently ranked responses in the dataset: [lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half) * Trained on the top/bottom responses of the prompts of the 25\% most consistently ranked responses in the dataset: [lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25) # Model results We compare the MT-Bench scores across 6 languages for our 4 ORPO trained models, as well as some baselines: * [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) - The foundation model that our models are ultimately built upon * [Nexusflow/Starling-LM-7B-beta](https://huggingface.co/Nexusflow/Starling-LM-7B-beta) - The highest performing open model on the Chatbot arena that is of a similar size to ours * gpt-3.5-turbo - A fairly high quality (although not state-of-the-art) proprietary LLM * [lightblue/suzume-llama-3-8B-multilingual](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual) - The base model which we train our ORPO finetunes from | **MT-Bench language** | **meta-llama/Meta-Llama-3-8B-Instruct** | **Nexusflow/Starling-LM-7B-beta** | **gpt-3.5-turbo** | **lightblue/suzume-llama-3-8B-multilingual** | **lightblue/suzume-llama-3-8B-multilingual-orpo-borda-full** | **lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75** | **lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half** | **lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25** | |-----------------------|-----------------------------------------|-----------------------------------|-------------------|----------------------------------------------|--------------------------------------------------------------|---------------------------------------------------------------|--------------------------------------------------------------|---------------------------------------------------------------| | **Chinese πŸ‡¨πŸ‡³** | NaN | 6.97 | 7.55 | 7.11 | 7.65 | **7.77** | 7.74 | 7.44 | | **English πŸ‡ΊπŸ‡Έ** | 7.98 | 7.92 | **8.26** | 7.73 | 7.98 | 7.94 | 7.98 | 8.22 | | **French πŸ‡«πŸ‡·** | NaN | 7.29 | 7.74 | 7.66 | **7.84** | 7.46 | 7.78 | 7.81 | | **German πŸ‡©πŸ‡ͺ** | NaN | 6.99 | 7.68 | 7.26 | 7.28 | 7.64 | 7.7 | **7.71** | | **Japanese πŸ‡―πŸ‡΅** | NaN | 6.22 | **7.84** | 6.56 | 7.2 | 7.12 | 7.34 | 7.04 | | **Russian πŸ‡·πŸ‡Ί** | NaN | 8.28 | 7.94 | 8.19 | 8.3 | 8.74 | **8.94** | 8.81 | We can see noticable improvement on most languages compared to the base model. We also find that our ORPO models achieve the highest score out of all the models we evaluated for a number of languages. # Training data We trained this model using the [lightblue/mitsu_full_borda](https://huggingface.co/datasets/lightblue/mitsu_full_borda) dataset. # Training configuration [Built with Axolotl](https://github.com/OpenAccess-AI-Collective/axolotl)
See axolotl config axolotl version: `0.4.0` ```yaml base_model: lightblue/suzume-llama-3-8B-multilingual model_type: LlamaForCausalLM tokenizer_type: AutoTokenizer # PreTrainedTokenizerFast load_in_8bit: false load_in_4bit: false strict: false rl: orpo orpo_alpha: 0.1 remove_unused_columns: false chat_template: chatml datasets: - path: lightblue/mitsu_full_borda type: orpo.chat_template conversation: llama-3 dataset_prepared_path: /workspace/llm_training/axolotl/llama3-multilingual-orpo/prepared_mitsu_full_borda val_set_size: 0.02 output_dir: /workspace/llm_training/axolotl/llama3-multilingual-orpo/output_mitsu_full_borda sequence_len: 8192 sample_packing: false pad_to_sequence_len: true use_wandb: true wandb_project: axolotl wandb_entity: peterd wandb_name: mitsu_full_borda gradient_accumulation_steps: 8 micro_batch_size: 1 num_epochs: 1 optimizer: paged_adamw_8bit lr_scheduler: cosine learning_rate: 8e-6 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_steps: 10 evals_per_epoch: 20 eval_table_size: saves_per_epoch: 1 debug: deepspeed: /workspace/axolotl/deepspeed_configs/zero3_bf16.json weight_decay: 0.0 special_tokens: pad_token: <|end_of_text|> ```

# workspace/llm_training/axolotl/llama3-multilingual-orpo/output_mitsu_full_borda This model is a fine-tuned version of [lightblue/suzume-llama-3-8B-multilingual](https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.1005 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 8e-06 - train_batch_size: 1 - eval_batch_size: 1 - seed: 42 - distributed_type: multi-GPU - num_devices: 4 - gradient_accumulation_steps: 8 - total_train_batch_size: 32 - 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: 10 - num_epochs: 1 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:----:|:---------------:| | 7.6836 | 0.01 | 1 | 7.6268 | | 3.9429 | 0.05 | 4 | 2.3993 | | 0.1193 | 0.1 | 8 | 0.1331 | | 0.1039 | 0.15 | 12 | 0.1209 | | 0.1082 | 0.2 | 16 | 0.1189 | | 0.1209 | 0.25 | 20 | 0.1180 | | 0.1106 | 0.3 | 24 | 0.1157 | | 0.103 | 0.35 | 28 | 0.1156 | | 0.1141 | 0.41 | 32 | 0.1123 | | 0.1156 | 0.46 | 36 | 0.1104 | | 0.0925 | 0.51 | 40 | 0.1089 | | 0.1113 | 0.56 | 44 | 0.1052 | | 0.1146 | 0.61 | 48 | 0.1073 | | 0.1029 | 0.66 | 52 | 0.1086 | | 0.1198 | 0.71 | 56 | 0.1072 | | 0.1205 | 0.76 | 60 | 0.1062 | | 0.1209 | 0.81 | 64 | 0.1041 | | 0.1047 | 0.86 | 68 | 0.1020 | | 0.0798 | 0.91 | 72 | 0.1008 | | 0.1007 | 0.96 | 76 | 0.1005 | ### Framework versions - Transformers 4.38.2 - Pytorch 2.2.1+cu121 - Datasets 2.18.0 - Tokenizers 0.15.0 # How to cite ```tex @article{devine2024sure, title={Are You Sure? Rank Them Again: Repeated Ranking For Better Preference Datasets}, author={Devine, Peter}, journal={arXiv preprint arXiv:2405.18952}, year={2024} } ``` # Developer Peter Devine - ([ptrdvn](https://huggingface.co/ptrdvn))