Merge pull request #667 from DhivyaBharathy-web/add-Phi_4_14B_GRPO

Add Phi_4_14B_GRPO notebook

Author: MervinPraison

examples/cookbooks/Phi_4_14B_GRPO.ipynb

@@ -0,0 +1,123 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "id": "cc1b7810",
+      "metadata": {
+        "id": "cc1b7810"
+      },
+      "source": [
+        "# Phi-4 (14B) Conversational with GRPO"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "9c882a2d",
+      "metadata": {
+        "id": "9c882a2d"
+      },
+      "source": [
+        "**Description:**\n",
+        "\n",
+        "This notebook demonstrates inference using the Phi-4 14B parameter model with GRPO optimization strategy."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/DhivyaBharathy-web/PraisonAI/blob/main/examples/cookbooks/Phi_4_14B_GRPO.ipynb)\n"
+      ],
+      "metadata": {
+        "id": "Z7azojscP0Ax"
+      },
+      "id": "Z7azojscP0Ax"
+    },
+    {
+      "cell_type": "markdown",
+      "id": "e8616784",
+      "metadata": {
+        "id": "e8616784"
+      },
+      "source": [
+        "**Dependencies**\n",
+        "\n",
+        "```python\n",
+        "!pip install transformers accelerate\n",
+        "!pip install torch\n",
+        "```"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "09866448",
+      "metadata": {
+        "id": "09866448"
+      },
+      "source": [
+        "**Tools Used**\n",
+        "\n",
+        "- HuggingFace Transformers\n",
+        "- GRPO Optimization\n",
+        "- PyTorch"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "403b0c37",
+      "metadata": {
+        "id": "403b0c37"
+      },
+      "source": [
+        "**YAML Prompt**\n",
+        "\n",
+        "```yaml\n",
+        "system: Act as a professional consultant.\n",
+        "user: How can AI help in healthcare?\n",
+        "```"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "e1785bcd",
+      "metadata": {
+        "id": "e1785bcd"
+      },
+      "outputs": [],
+      "source": [
+        "from transformers import AutoTokenizer, AutoModelForCausalLM\n",
+        "\n",
+        "tokenizer = AutoTokenizer.from_pretrained(\"microsoft/phi-4-14b\")\n",
+        "model = AutoModelForCausalLM.from_pretrained(\"microsoft/phi-4-14b\")\n",
+        "\n",
+        "prompt = \"How can AI help in healthcare?\"\n",
+        "inputs = tokenizer(prompt, return_tensors=\"pt\")\n",
+        "outputs = model.generate(**inputs, max_new_tokens=60)\n",
+        "print(tokenizer.decode(outputs[0], skip_special_tokens=True))"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "074e4e6e",
+      "metadata": {
+        "id": "074e4e6e"
+      },
+      "source": [
+        "**Output**\n",
+        "\n",
+        "The model generates thoughtful insights on AI applications in healthcare.\n",
+        "\n",
+        "How can AI help in healthcare?\n",
+        "\n",
+        "AI can assist in healthcare by improving diagnostics, predicting patient outcomes, personalizing treatments, and enhancing administrative workflows. It enables faster data analysis and decision-making, leading to better patient care.\n"
+      ]
+    }
+  ],
+  "metadata": {
+    "colab": {
+      "provenance": []
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 5
+}