Update README.md

README.md

@@ -14,4 +14,92 @@ pipeline_tag: text-generation
 # Llama 3 8B Instruct (Financial RAG)
 
 This model is a fine-tuned version of the original [Llama 3 8B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) model
-on 4000 examples from the [virattt/financial-qa-10K](https://huggingface.co/datasets/virattt/financial-qa-10K) dataset.
+on 4000 examples from the [virattt/financial-qa-10K](https://huggingface.co/datasets/virattt/financial-qa-10K) dataset.
+
+The model is fine-tuned using a LoRA adapter for RAG use cases. It is optimized to answer a question based on a context:
+
+```txt
+Answer the question:
+{question}
+
+Using the information:
+{context}
+```
+
+## Example Usage
+
+Load the model:
+
+```py
+MODEL_NAME = "curiousily/Llama-3-8B-Instruct-Finance-RAG"
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, use_fast=True)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_NAME,
+    device_map="auto"
+)
+
+pipe = pipeline(
+    task="text-generation",
+    model=model,
+    tokenizer=tokenizer,
+    max_new_tokens=128,
+    return_full_text=False,
+)
+```
+
+Format the prompt (uses the original Instruct prompt format):
+
+````py
+prompt = """
+<|begin_of_text|><|start_header_id|>system<|end_header_id|>
+
+Use only the information to answer the question<|eot_id|><|start_header_id|>user<|end_header_id|>
+
+How much did the company's net earnings amount to in fiscal 2022?
+
+Information:
+
+```
+Net earnings were $17.1 billion in fiscal 2022.
+```<|eot_id|><|start_header_id|>assistant<|end_header_id|>
+"""
+````
+
+And make a prediction:
+
+```py
+print(outputs[0]["generated_text"])
+```
+
+```
+$17.1 billion
+```
+
+Here's a helper function to build your prompts:
+
+```py
+def create_test_prompt(data_row):
+    prompt = dedent(f"""
+    {data_row["question"]}
+
+    Information:
+
+    ```
+    {data_row["context"]}
+    ```
+    """)
+    messages = [
+        {"role": "system", "content": "Use only the information to answer the question"},
+        {"role": "user", "content": prompt},
+    ]
+    return tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+```
+
+Where `data_row` must be a dict:
+
+```
+data_row = {
+  "question": "...",
+  "context": "..."
+}
+```