Hugging Face Overview
PsycoLLM: Enhanced for Psychological Understanding
PsycoLLM is a 14.2 billion parameter large language model developed by MindIntLab, fine-tuned from the Qwen/Qwen1.5-14B-Chat architecture. This model is specifically designed to enhance capabilities in psychological understanding and evaluation.
Key Capabilities
- Psychological Understanding: Optimized for tasks related to mental health and psychological assessment.
- Dialogue Processing: Fine-tuned on dialogue datasets, suggesting improved conversational abilities in relevant contexts.
- Evaluation Focus: Aims to provide robust performance in evaluating psychological aspects, as indicated by its core purpose.
Training Details
The model was trained with a learning rate of 1e-05 over 3 epochs, using a total batch size of 128 across 8 GPUs. The training process involved Adam optimizer and a cosine learning rate scheduler. Further details on the specific datasets used for fine-tuning will be open-sourced in the future.
Good For
- Applications requiring psychological understanding.
- Research in mental health and AI.
- Dialogue systems with a focus on psychological context.