131 lines
4.0 KiB
Markdown
131 lines
4.0 KiB
Markdown
---
|
|
datasets:
|
|
- brucewayne0459/Skin_diseases_and_care
|
|
language:
|
|
- en
|
|
license: mit
|
|
tags:
|
|
- medical
|
|
- dermatology
|
|
- skin_disease
|
|
- skin_care
|
|
- unsloth
|
|
- trl
|
|
- sft
|
|
---
|
|
|
|
## Model Details
|
|
|
|
### Model Description
|
|
|
|
<!-- Provide a longer summary of what this model is. -->
|
|
|
|
|
|
|
|
- **Developed by:** Bruce_Wayne(The Batman)
|
|
- **Model type:** Text Generation
|
|
- **Finetuned from model [optional]:** OpenBioLLM(llama-3)(aaditya/Llama3-OpenBioLLM-8B)
|
|
|
|
## You can find the gguf versions here --> https://huggingface.co/brucewayne0459/OpenBioLLm-Derm-gguf
|
|
### please let me know how the model works -->https://forms.gle/N14zZTkLpUr6Hf4BA
|
|
### Thank you!
|
|
## Uses
|
|
|
|
### Direct Use
|
|
|
|
This model is fine-tuned on skin diseases and dermatology data and is used for a dermatology chatbot to provide clear, accurate, and helpful information about various skin diseases, skin care routines, treatments, and related dermatological advice.
|
|
|
|
|
|
## Bias, Risks, and Limitations
|
|
|
|
This model is trained on dermatology data, which might contain inherent biases. It is important to note that the model's responses should not be considered a substitute for professional medical advice. There may be limitations in understanding rare skin conditions or those not well-represented in the training data.
|
|
The model still need to be fine-tuned further to get accurate answers.
|
|
|
|
### Recommendations
|
|
|
|
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
|
|
|
## How to Get Started with the Model
|
|
|
|
Use the code below to get started with the model.
|
|
|
|
```python
|
|
from transformers import AutoTokenizer, AutoModelForCausalLM
|
|
|
|
model_name = "brucewayne0459/OpenBioLLm-Derm"
|
|
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
|
model = AutoModelForCausalLM.from_pretrained(model_name)
|
|
```
|
|
|
|
## Training Details
|
|
|
|
### Training Data
|
|
|
|
The model is fine-tuned on a dataset containing information about various skin diseases and dermatology care.
|
|
brucewayne0459/Skin_diseases_and_care
|
|
|
|
### Training Procedure
|
|
|
|
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
|
|
|
#### Preprocessing [optional]
|
|
|
|
"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
|
|
|
|
### Instruction:
|
|
You are a highly knowledgeable and empathetic dermatologist. Provide clear, accurate, and helpful information about various skin diseases, skin care routines, treatments, and related dermatological advice.
|
|
|
|
### Input:
|
|
{}
|
|
|
|
### Response:
|
|
{}
|
|
"""
|
|
EOS_TOKEN = tokenizer.eos_token # Must add EOS_TOKEN
|
|
|
|
def formatting_prompts_func(examples):
|
|
inputs = examples["Topic"]
|
|
outputs = examples["Information"]
|
|
texts = []
|
|
|
|
Prompt passed while fine tuning the model
|
|
#### Training Hyperparameters
|
|
|
|
Training regime: The model was trained using the following hyperparameters:
|
|
Per device train batch size: 2
|
|
Gradient accumulation steps: 4
|
|
Warmup steps: 5
|
|
Max steps: 120
|
|
Learning rate: 2e-4
|
|
Optimizer: AdamW (8-bit)
|
|
Weight decay: 0.01
|
|
LR scheduler type: Linear
|
|
|
|
|
|
|
|
## Environmental Impact
|
|
|
|
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
|
|
|
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
|
|
|
- **Hardware Type:** Tesls T4 gpu
|
|
- **Hours used:** 1hr
|
|
- **Cloud Provider:** Google Colab
|
|
|
|
|
|
## Technical Specifications [optional]
|
|
|
|
### Model Architecture and Objective
|
|
|
|
This model is based on the LLaMA (Large Language Model Meta AI) architecture and fine-tuned to provide dermatological advice.
|
|
|
|
|
|
#### Hardware
|
|
|
|
The training was performed on Tesla T4 gpu with 4-bit quantization and gradient checkpointing to optimize memory usage.
|
|
|
|
|
|
### Feel free to provide any missing details or correct the assumptions made, and I'll update the model card accordingly.
|
|
|