README.md

---
library_name: transformers
tags:
- medical-qa
- healthcare
- llama
- fine-tuned
- llama-cpp
- gguf-my-repo
license: llama3.2
datasets:
- ruslanmv/ai-medical-chatbot
base_model: Ellbendls/llama-3.2-3b-chat-doctor
---

# Triangle104/llama-3.2-3b-chat-doctor-Q5_K_M-GGUF
This model was converted to GGUF format from [`Ellbendls/llama-3.2-3b-chat-doctor`](https://huggingface.co/Ellbendls/llama-3.2-3b-chat-doctor) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
Refer to the [original model card](https://huggingface.co/Ellbendls/llama-3.2-3b-chat-doctor) for more details on the model.

---
Model details:
-
Llama-3.2-3B-Chat-Doctor is a specialized medical question-answering model based on the Llama 3.2 3B architecture. This model has been fine-tuned specifically for providing accurate and helpful responses to medical-related queries.

    Developed by: Ellbendl Satria
    Model type: Language Model (Conversational AI)
    Language: English
    Base Model: Meta Llama-3.2-3B-Instruct
    Model Size: 3 Billion Parameters
    Specialization: Medical Question Answering
    License: llama3.2

Model Capabilities

    Provides informative responses to medical questions
    Assists in understanding medical terminology and health-related concepts
    Offers preliminary medical information (not a substitute for professional medical advice)

Direct Use

This model can be used for:

    Providing general medical information
    Explaining medical conditions and symptoms
    Offering basic health-related guidance
    Supporting medical education and patient communication

Limitations and Important Disclaimers

⚠️ CRITICAL WARNINGS:

    NOT A MEDICAL PROFESSIONAL: This model is NOT a substitute for professional medical advice, diagnosis, or treatment.
    Always consult a qualified healthcare provider for medical concerns.
    The model's responses should be treated as informational only and not as medical recommendations.

Out-of-Scope Use

The model SHOULD NOT be used for:

    Providing emergency medical advice
    Diagnosing specific medical conditions
    Replacing professional medical consultation
    Making critical healthcare decisions

Bias, Risks, and Limitations
Potential Biases

    May reflect biases present in the training data
    Responses might not account for individual patient variations
    Limited by the comprehensiveness of the training dataset

Technical Limitations

    Accuracy is limited to the knowledge in the training data
    May not capture the most recent medical research or developments
    Cannot perform physical examinations or medical tests

Recommendations

    Always verify medical information with professional healthcare providers
    Use the model as a supplementary information source
    Be aware of potential inaccuracies or incomplete information

Training Details
Training Data

    Source Dataset: ruslanmv/ai-medical-chatbot
    Base Model: Meta Llama-3.2-3B-Instruct

Training Procedure

[Provide details about the fine-tuning process, if available]

    Fine-tuning approach
    Computational resources used
    Training duration
    Specific techniques applied during fine-tuning

How to Use the Model
Hugging Face Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Ellbendls/llama-3.2-3b-chat-doctor"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Example usage
input_text = "I had a surgery which ended up with some failures. What can I do to fix it?"

# Prepare inputs with explicit padding and attention mask
inputs = tokenizer(input_text, return_tensors="pt", padding=True, truncation=True)

# Generate response with more explicit parameters
outputs = model.generate(
    input_ids=inputs['input_ids'], 
    attention_mask=inputs['attention_mask'],
    max_new_tokens=150,  # Specify max new tokens to generate
    do_sample=True,      # Enable sampling for more diverse responses
    temperature=0.7,     # Control randomness of output
    top_p=0.9,           # Nucleus sampling to maintain quality
    num_return_sequences=1  # Number of generated sequences
)

# Decode the generated response
response = tokenizer.decode(outputs[0], skip_special_tokens=True)

print(response)

Ethical Considerations

This model is developed with the intent to provide helpful, accurate, and responsible medical information. Users are encouraged to:

    Use the model responsibly
    Understand its limitations
    Seek professional medical advice for serious health concerns

---
## Use with llama.cpp
Install llama.cpp through brew (works on Mac and Linux)

```bash
brew install llama.cpp

```
Invoke the llama.cpp server or the CLI.

### CLI:
```bash
llama-cli --hf-repo Triangle104/llama-3.2-3b-chat-doctor-Q5_K_M-GGUF --hf-file llama-3.2-3b-chat-doctor-q5_k_m.gguf -p "The meaning to life and the universe is"
```

### Server:
```bash
llama-server --hf-repo Triangle104/llama-3.2-3b-chat-doctor-Q5_K_M-GGUF --hf-file llama-3.2-3b-chat-doctor-q5_k_m.gguf -c 2048
```

Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.

Step 1: Clone llama.cpp from GitHub.
```
git clone https://github.com/ggerganov/llama.cpp
```

Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
```
cd llama.cpp && LLAMA_CURL=1 make
```

Step 3: Run inference through the main binary.
```
./llama-cli --hf-repo Triangle104/llama-3.2-3b-chat-doctor-Q5_K_M-GGUF --hf-file llama-3.2-3b-chat-doctor-q5_k_m.gguf -p "The meaning to life and the universe is"
```
or 
```
./llama-server --hf-repo Triangle104/llama-3.2-3b-chat-doctor-Q5_K_M-GGUF --hf-file llama-3.2-3b-chat-doctor-q5_k_m.gguf -c 2048
```