feat: initial very ugly submission code (still prettier than the default tho)

.gitignore

@@ -0,0 +1,3 @@
+.env
+downloaded_files
+.idea

notebooks/01-vi-questions.ipynb

@@ -0,0 +1,246 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "id": "initial_id",
+   "metadata": {
+    "collapsed": true
+   },
+   "source": [
+    "import json\n",
+    "from pathlib import Path\n",
+    "\n",
+    "import pandas as pd\n",
+    "import requests\n",
+    "from dotenv import dotenv_values\n",
+    "\n",
+    "test_api_base = \"https://agents-course-unit4-scoring.hf.space\"\n",
+    "\n",
+    "def get_random_question():\n",
+    "    url = f\"{test_api_base}/random-question\"\n",
+    "\n",
+    "\n",
+    "    try:\n",
+    "        # Fetch the random question\n",
+    "        response = requests.get(url, timeout=10)\n",
+    "        response.raise_for_status()\n",
+    "        question_data = response.json()\n",
+    "\n",
+    "        # Check if there's an associated file to download\n",
+    "        if question_data.get(\"file_name\") and question_data.get(\"task_id\"):\n",
+    "            task_id = question_data[\"task_id\"]\n",
+    "            file_url = f\"{test_api_base}/files/{task_id}\"\n",
+    "\n",
+    "            # Create a directory for downloaded files if it doesn't exist\n",
+    "            download_dir = Path(\"downloaded_files\")\n",
+    "            download_dir.mkdir(exist_ok=True)\n",
+    "\n",
+    "            # Download the file\n",
+    "            file_response = requests.get(file_url, timeout=30)\n",
+    "            file_response.raise_for_status()\n",
+    "\n",
+    "            # Get filename from content-disposition header or use task_id\n",
+    "            content_disposition = file_response.headers.get('content-disposition', '')\n",
+    "            if 'filename=' in content_disposition:\n",
+    "                filename = content_disposition.split('filename=')[1].strip('\"')\n",
+    "            else:\n",
+    "                # Default filename with extension based on content-type\n",
+    "                content_type = file_response.headers.get('content-type', '')\n",
+    "                ext = '.bin'  # default\n",
+    "                if 'excel' in content_type or 'spreadsheet' in content_type:\n",
+    "                    ext = '.xlsx'\n",
+    "                elif 'csv' in content_type:\n",
+    "                    ext = '.csv'\n",
+    "                elif 'json' in content_type:\n",
+    "                    ext = '.json'\n",
+    "                elif 'text' in content_type:\n",
+    "                    ext = '.txt'\n",
+    "                filename = f\"{task_id}{ext}\"\n",
+    "\n",
+    "            file_path = download_dir / filename\n",
+    "\n",
+    "            # Save the file\n",
+    "            with open(file_path, 'wb') as f:\n",
+    "                f.write(file_response.content)\n",
+    "\n",
+    "            # Add the file path to the question data\n",
+    "            question_data['downloaded_file_path'] = str(file_path)\n",
+    "            print(f\"Downloaded file to: {file_path}\")\n",
+    "\n",
+    "        return question_data\n",
+    "\n",
+    "    except requests.exceptions.RequestException as e:\n",
+    "        print(f\"Error fetching question: {e}\")\n",
+    "        return None\n",
+    "    except json.JSONDecodeError as e:\n",
+    "        print(f\"Error parsing JSON response: {e}\")\n",
+    "        return None\n",
+    "    except Exception as e:\n",
+    "        print(f\"Unexpected error: {e}\")\n",
+    "        return None"
+   ],
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": [
+    "# if __name__ == \"__main__\":\n",
+    "question = get_random_question()\n",
+    "if question:\n",
+    "    print(f\"Task ID: {question.get('task_id')}\")\n",
+    "    print(f\"Question: {question.get('question')}\")\n",
+    "    print(f\"Level: {question.get('Level')}\")\n",
+    "    if 'downloaded_file_path' in question:\n",
+    "        print(f\"Downloaded file: {question['downloaded_file_path']}\")"
+   ],
+   "id": "55d7941445304e9b",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": [
+    "from smolagents import AzureOpenAIServerModel, CodeAgent, DuckDuckGoSearchTool, VisitWebpageTool\n",
+    "\n",
+    "config = dotenv_values()\n",
+    "\n",
+    "model = AzureOpenAIServerModel(\n",
+    "    model_id=config[\"AZURE_OPENAI_CHAT_MODEL\"],\n",
+    "    api_key=config[\"AZURE_OPENAI_API_KEY\"],\n",
+    "    api_version=config[\"AZURE_OPENAI_API_VERSION\"],\n",
+    "    azure_endpoint=config[\"AZURE_OPENAI_API_BASE\"],\n",
+    ")\n",
+    "\n",
+    "agent = CodeAgent(tools=[], model=model, max_steps=10, verbosity_level=0)"
+   ],
+   "id": "99393f634f21563f",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "agent.run(question.get(\"question\"))",
+   "id": "e761157828f8ebb1",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "",
+   "id": "47901ff79b3bed13",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "## Question Processing",
+   "id": "1290570b730fda4"
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": [
+    "response = requests.get(f\"{test_api_base}/questions\", timeout=15)\n",
+    "response.raise_for_status()\n",
+    "\n",
+    "questions_data = response.json()"
+   ],
+   "id": "9f6fe414bc8fb090",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "questions_data",
+   "id": "ca50d2e29bfbc34a",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": [
+    "\n",
+    "def run_agents(questions_data: list[{}]):\n",
+    "    answers = []\n",
+    "    results_log = []\n",
+    "    for question_data in questions_data:\n",
+    "        task_id = question_data.get(\"task_id\")\n",
+    "        question_text = question_data.get(\"question\")\n",
+    "        level = question_data.get(\"Level\")\n",
+    "        print(f\"Task ID: {task_id}, Question: {question_text}, Level: {level}\")\n",
+    "\n",
+    "        answer = agent.run(task=question_text)\n",
+    "        print(answer)\n",
+    "\n",
+    "        answers.append({\"task_id\": task_id, \"submitted_answer\": answer})\n",
+    "        results_log.append({\"Task ID\": task_id, \"Question\": question_text, \"Answer\": answer})\n",
+    "\n",
+    "    submission_data = {\n",
+    "        \"username\": \"vladi\",\n",
+    "        \"agent_code\": \"https://huggingface.co/spaces/vladi/AgentsGAIAFun\",\n",
+    "        \"answers\": answers\n",
+    "    }\n",
+    "\n",
+    "    return submission_data, results_log\n",
+    "\n",
+    "def submit_answers(submission_data: dict):\n",
+    "    print(f\"Submitting {len(submission_data['answers'])} answers\")\n",
+    "\n",
+    "    response = requests.post(f\"{test_api_base}/submit\", json=submission_data, timeout=60)\n",
+    "    response.raise_for_status()\n",
+    "    result_data = response.json()\n",
+    "\n",
+    "    return result_data\n",
+    "\n",
+    "\n",
+    "submission_data, results_log = run_agents(questions_data[:4])\n",
+    "results_df = pd.DataFrame(results_log)\n",
+    "\n",
+    "result = submit_answers(submission_data)\n",
+    "\n",
+    "print(results_df)\n",
+    "print(result)\n"
+   ],
+   "id": "74bce95503481798",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "results_df",
+   "id": "57e1c5515e9bf8a1",
+   "outputs": [],
+   "execution_count": null
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}