Update README.md

README.md

@@ -16,10 +16,6 @@ In addition to its ability to act as a virtual medical consultant, we are still
 ## Model Details
 
 ### Model Description
-**BioChat** is a language model fine-tuned using the [ChatDoctor-5k](https://huggingface.co/datasets/LinhDuong/chatdoctor-5k) dataset, specifically designed for medical conversations. It acts as a virtual doctor, allowing users to ask questions about symptoms, discuss potential health concerns, or seek general medical information. With its foundation in biomedical language understanding, BioChat delivers responses tailored to the medical domain.
-
-The model also supports natural language generation, producing coherent and context-aware replies to user queries. However, these generation capabilities are intended for research purposes and are not yet suitable for real-world clinical applications. BioChat represents a step forward in making healthcare-related NLP tools more accessible while emphasizing the importance of reliability, truthfulness, and explainability in its design and use.
-
 
 |      Model Name     |             Base Model             |      Model Type     | Sequence Length |                          Download                          |
 |:-------------------:|:----------------------------------:|:-------------------:|:---------------:|:-----------------------------------------------------:|
@@ -30,13 +26,24 @@ The model also supports natural language generation, producing coherent and cont
 
 - **Finetuned from model [optional]:** [BioMistral-7B](https://huggingface.co/BioMistral/BioMistral-7B).
 
-### Model Sources [optional]
+### Using BioChat
+
+You can use BioMistral with [Hugging Face's Transformers library](https://github.com/huggingface/transformers) as follow.
+
+Loading the model and tokenizer :
 
-<!-- Provide the basic links for the model. -->
+```python
+from transformers import AutoModel, AutoTokenizer
 
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+tokenizer = AutoTokenizer.from_pretrained("BioMistral/BioMistral-7B")
+model = AutoModelForCausalLM.from_pretrained(
+    "BioMistral/BioMistral-7B",
+    load_in_8bit=True,
+    device_map="auto",
+    output_hidden_states=True  # Ensure hidden states are available
+)
+model = PeftModel.from_pretrained(model, "Indah1/BioChat10")
+```
 
 ## Bias, Risks, and Limitations