BioSilk_Llama3.1-8b-Base-CPT-SFT-DPO

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

BioSilk_Llama3.1-8b-Base-CPT-SFT-DPO

This model is a fine-tuned version of lamm-mit/BioSilk_Llama3.1-8b-Base-CPT-SFT on the lamm-mit/orpo-dpo-mix-40k, the lamm-mit/bioinspired_gpt-4o-mini-silk_dataset_V10-orpo_dataset_V10_messageformat_filtered and the lamm-mit/bioinspired_gpt-4o-mini-dataset_V1_B-cleaned-orpo_dataset_V10_messageformat_filtered datasets.

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: 5e-07
• train_batch_size: 2
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• total_eval_batch_size: 8
• 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

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1