16 lines
954 B
Markdown
16 lines
954 B
Markdown
|
---
|
||
|
|
license: cc-by-nc-4.0
|
||
|
|
library_name: transformers
|
||
|
|
pipeline_tag: text-generation
|
||
|
|
datasets:
|
||
|
|
- Psychotherapy-LLM/PsychoCounsel-Preference
|
||
|
|
base_model:
|
||
|
|
- meta-llama/Llama-3.1-8B-Instruct
|
||
|
|
---
|
||
|
|
|
||
|
|
This model is presented in the paper [Preference Learning Unlocks LLMs' Psycho-Counseling Skills](https://hf.co/papers/2502.19731). It's a fine-tuned [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) model trained using preference learning on the [PsyCoPref](https://huggingface.co/datasets/Psychotherapy-LLM/PsychoCounsel-Preference) dataset. This dataset contains 36k high-quality preference comparison pairs aligned with the preferences of professional psychotherapists.
|
||
|
|
|
||
|
|
The model aims to improve the quality of responses in psycho-counseling sessions and achieves a win rate of 87% against GPT-4o.
|
||
|
|
|
||
|
|
|
||
|
|
This usage is the same as [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct)
|