Readme Updated

README.md

@@ -1,21 +1,39 @@
 ---
-base_model: unsloth/meta-llama-3.1-8b-instruct-bnb-4bit
-tags:
-- text-generation-inference
-- transformers
-- unsloth
-- llama
-license: apache-2.0
+license: llama3.1
 language:
 - en
+- sql
+- py
+tags:
+- gis
+- geospatial
+- fine-tuned
+- unsloth
+pipeline_tag: text-generation
 ---
 
-# Uploaded finetuned  model
+# 🗺️ Llama-3.1-8B-GIS-Expert
+
+This model is a fine-tuned version of **Llama-3.1-8B-Instruct**, specialized for **Geospatial Analysis (GIS)**.
+
+## 🧠 Capabilities
+It has been trained on a multi-task dataset to handle three specific roles:
+1.  **PostGIS Expert:** Converts natural language questions into valid `PostgreSQL` / `PostGIS` SQL queries.
+2.  **Python GIS Developer:** Writes Python scripts using `geopandas`, `shapely`, and `rasterio`.
+3.  **Geospatial Analyst:** Explains spatial relationships and topology logic.
+
+## 💻 How to Use
 
-- **Developed by:** kumarprince070107
-- **License:** apache-2.0
-- **Finetuned from model :** unsloth/meta-llama-3.1-8b-instruct-bnb-4bit
+### System Prompts (Crucial)
+To get the best results, you must use the correct System Prompt for the task:
+* **For SQL:** "You are a PostGIS expert. Convert the question into a SQL query."
+* **For Python:** "You are a Python GIS developer. Write a script to solve the geospatial problem."
+* **For Reasoning:** "You are a Geospatial Analyst. Explain the spatial relationship."
 
-This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+### Example Code
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
+model_id = "kumarprince070107/Llama-3.1-8B-GIS-v2"
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")