komodo-chat

This model is a fine-tuned version of Yellow-AI-NLP/komodo-7b-base on the generator dataset.

Model description

This model understands Indonesian conversation and responds with formal Indonesian.

Intended uses & limitations

This model can be used to personalize conversations and role-play based on the persona given with the prompt PERSONA_1 is the user persona and PERSONA_2 is the assistant persona.

Training and evaluation data

This model was trained using specific prompts on the nayohan/multi_session_chat translated into Indonesian with mBART50

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

TensorBoard

Framework versions

• PEFT 0.7.2.dev0
• Transformers 4.36.2
• Pytorch 2.0.1+cu117
• Datasets 2.16.1
• Tokenizers 0.15.2