CPsyCounX

This model is a fine-tuned version of internlm/internlm2-chat-7b on the CPsyCounD dataset.

Model description

CPsyCounX is a large language model designed for Chinese Psychological Counseling. Paper: CPsyCoun.

Training and evaluation data

• Train: CPsyCounD
• Evaluate: CPsyCounE

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-06
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 28
• total_train_batch_size: 448
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 9.0
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.37.1
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1