Update README.md

README.md

@@ -4,4 +4,55 @@ datasets:
 - malhajar/meditron-tr
 language:
 - tr
----
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+meditron-7b-chat-turkish is a finetuned version on turkish language Alpaca of [`epfl-llm/meditron-7b`](https://huggingface.co/epfl-llm/meditron-7b) using SFT Training.
+This model can answer information about different excplicit ideas in medicine (see [`epfl-llm/meditron-7b`](https://huggingface.co/epfl-llm/meditron-7b) for more info)
+
+### Model Description
+
+- **Finetuned by:** [`Mohamad Alhajar`](https://www.linkedin.com/in/muhammet-alhajar/) 
+- **Language(s) (NLP):** Turkish,English
+- **Finetuned from model:** [`epfl-llm/meditron-7b`](https://huggingface.co/epfl-llm/meditron-7b)
+
+### Prompt Template
+```
+### Talimat:
+
+<prompt> (without the <>)
+
+### Yanıt:
+```
+
+
+## How to Get Started with the Model
+
+Use the code sample provided in the original post to interact with the model.
+```python
+from transformers import AutoTokenizer,AutoModelForCausalLM
+ 
+model_id = "malhajar/meditron-7b-chat-turkish"
+model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
+                                             device_map="auto",
+                                             torch_dtype=torch.float16,
+                                             revision="main")
+
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+
+question: "Akciğer kanseri nedir?"
+# For generating a response
+prompt = '''
+### Instruction:
+{question} 
+
+### Response:'''
+input_ids = tokenizer(prompt, return_tensors="pt").input_ids
+output = model.generate(inputs=input_ids,max_new_tokens=512,pad_token_id=tokenizer.eos_token_id,top_k=50, do_sample=True,
+        top_p=0.95)
+response = tokenizer.decode(output[0])
+
+print(response)
+```