Update prompts.

demo/sample_generator.ipynb

@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -20,30 +20,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/home/yale/work/meta-prompt/.venv/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
-      "  from .autonotebook import tqdm as notebook_tqdm\n"
-     ]
-    },
-    {
-     "ename": "TypeError",
-     "evalue": "Dataframe.__init__() got an unexpected keyword argument 'format'",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
-      "Cell \u001b[0;32mIn[2], line 139\u001b[0m\n\u001b[1;32m    131\u001b[0m     \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[1;32m    132\u001b[0m         \u001b[38;5;28;01mraise\u001b[39;00m gr\u001b[38;5;241m.\u001b[39mError(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mAn error occurred: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mstr\u001b[39m(e)\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m    134\u001b[0m demo \u001b[38;5;241m=\u001b[39m gr\u001b[38;5;241m.\u001b[39mInterface(\n\u001b[1;32m    135\u001b[0m     fn\u001b[38;5;241m=\u001b[39mprocess_json,\n\u001b[1;32m    136\u001b[0m     inputs\u001b[38;5;241m=\u001b[39m[gr\u001b[38;5;241m.\u001b[39mTextbox(label\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mInput JSON\u001b[39m\u001b[38;5;124m\"\u001b[39m),\n\u001b[1;32m    137\u001b[0m             gr\u001b[38;5;241m.\u001b[39mSlider(label\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mGenerating Batch Size\u001b[39m\u001b[38;5;124m\"\u001b[39m, value\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m3\u001b[39m, minimum\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m, maximum\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m10\u001b[39m, step\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m)],\n\u001b[1;32m    138\u001b[0m     outputs\u001b[38;5;241m=\u001b[39m[gr\u001b[38;5;241m.\u001b[39mTextbox(label\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mDescription\u001b[39m\u001b[38;5;124m\"\u001b[39m),\n\u001b[0;32m--> 139\u001b[0m              \u001b[43mgr\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mDataFrame\u001b[49m\u001b[43m(\u001b[49m\u001b[43mlabel\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mExamples\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\n\u001b[1;32m    140\u001b[0m \u001b[43m                           \u001b[49m\u001b[43mheaders\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m[\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mInput\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mOutput\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    141\u001b[0m \u001b[43m                           \u001b[49m\u001b[38;5;28;43mformat\u001b[39;49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mmarkdown\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m],\n\u001b[1;32m    142\u001b[0m     title\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mTask Description Generator\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    143\u001b[0m     description\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mEnter a JSON object with \u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124minput\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m and \u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124moutput\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m fields to generate a task description and additional examples.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    144\u001b[0m )\n\u001b[1;32m    146\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;18m__name__\u001b[39m \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m__main__\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[1;32m    147\u001b[0m     demo\u001b[38;5;241m.\u001b[39mlaunch()\n",
-      "File \u001b[0;32m~/work/meta-prompt/.venv/lib/python3.10/site-packages/gradio/component_meta.py:163\u001b[0m, in \u001b[0;36mupdateable.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    161\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m    162\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m--> 163\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
-      "\u001b[0;31mTypeError\u001b[0m: Dataframe.__init__() got an unexpected keyword argument 'format'"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "import json\n",
     "import yaml\n",
@@ -55,14 +34,16 @@
     "from langchain.output_parsers import YamlOutputParser\n",
     "\n",
     "# Define prompt strings as constants\n",
-    "DESCRIPTION_PROMPT = \"\"\"Given the following JSON example for a task type:\n",
+    "DESCRIPTION_PROMPT = [\n",
+    "    (\"system\", \"\"\"Given the following JSON example for a task type:\n",
     "{raw_example}\n",
     "\n",
     "Provide a concise description of the task type, including the format and style of the output.\n",
     "\n",
     "Format your response as follows:\n",
     "Task Description: [Your description here]\n",
-    "\"\"\"\n",
+    "\"\"\")\n",
+    "]\n",
     "\n",
     "BRIEFS_PROMPT = [\n",
     "    (\"system\", \"\"\"Given the task type description, generate descriptions for {generating_batch_size} new \n",
@@ -104,24 +85,31 @@
     "\"\"\")\n",
     "]\n",
     "\n",
-    "EXAMPLES_PROMPT = \"\"\"Given the following task type description, and input/output example:\n",
-    "Task Description: {description}\n",
-    "Example: {raw_example}\n",
-    "\n",
-    "Generate {generating_batch_size} new input/output examples for this task type.\n",
+    "EXAMPLES_PROMPT = [\n",
+    "    (\"system\", \"\"\"Given the task type description, and input/output example, generate {generating_batch_size}\n",
+    "new input/output examples for this task type.\n",
     "\n",
     "Format your response as a valid JSON object with a single key 'examples' \n",
     "containing a JSON array of {generating_batch_size} objects, each with 'input' and 'output' fields.\n",
-    "\"\"\"\n",
+    "\"\"\"),\n",
+    "    (\"user\", \"\"\"Task Description:\n",
+    "\n",
+    "{description}\n",
+    "\n",
+    "Example:\n",
+    "\n",
+    "{raw_example}\n",
+    "\n",
+    "\"\"\")\n",
+    "]\n",
     "\n",
     "class TaskDescriptionGenerator:\n",
-    "    def __init__(self, model_temperature=0.7):        \n",
-    "        self.description_prompt = ChatPromptTemplate.from_template(DESCRIPTION_PROMPT)\n",
+    "    def __init__(self, model):        \n",
+    "        self.description_prompt = ChatPromptTemplate.from_messages(DESCRIPTION_PROMPT)\n",
     "        self.briefs_prompt = ChatPromptTemplate.from_messages(BRIEFS_PROMPT)\n",
     "        self.examples_from_briefs_prompt = ChatPromptTemplate.from_messages(EXAMPLES_FROM_BRIEFS_PROMPT)\n",
-    "        self.examples_prompt = ChatPromptTemplate.from_template(EXAMPLES_PROMPT)\n",
+    "        self.examples_prompt = ChatPromptTemplate.from_messages(EXAMPLES_PROMPT)\n",
     "\n",
-    "        model = ChatOpenAI(model=config[\"model_name\"], temperature=model_temperature)\n",
     "        json_model = model.bind(response_format={\"type\": \"json_object\"})\n",
     "\n",
     "        output_parser = StrOutputParser()\n",
@@ -163,30 +151,43 @@
     "            return result\n",
     "\n",
     "        except Exception as e:\n",
-    "            raise RuntimeError(f\"An error occurred during processing: {str(e)}\")\n",
-    "\n",
+    "            raise RuntimeError(f\"An error occurred during processing: {str(e)}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "import gradio as gr\n",
     "\n",
-    "def process_json(input_json, generating_batch_size=3):\n",
+    "def process_json(input_json, model_name, generating_batch_size=3):\n",
     "    try:\n",
-    "        generator = TaskDescriptionGenerator()\n",
+    "        model = ChatOpenAI(model=model_name)\n",
+    "        generator = TaskDescriptionGenerator(model)\n",
     "        result = generator.process(input_json, generating_batch_size)\n",
     "        description = result[\"description\"]\n",
     "        examples = [[example[\"input\"], example[\"output\"]] for example in result[\"additional_examples\"]]\n",
-    "        print(examples)\n",
     "        return description, examples\n",
     "    except Exception as e:\n",
     "        raise gr.Error(f\"An error occurred: {str(e)}\")\n",
     "\n",
     "demo = gr.Interface(\n",
     "    fn=process_json,\n",
-    "    inputs=[gr.Textbox(label=\"Input JSON\"),\n",
-    "            gr.Slider(label=\"Generating Batch Size\", value=3, minimum=1, maximum=10, step=1)],\n",
-    "    outputs=[gr.Textbox(label=\"Description\"),\n",
-    "             gr.DataFrame(label=\"Examples\", \n",
-    "                           headers=[\"Input\", \"Output\"])],\n",
+    "    inputs=[\n",
+    "        gr.Textbox(label=\"Input JSON\"),\n",
+    "        gr.Dropdown(label=\"Model Name\", choices=[\"llama3-70b-8192\", \"llama3-8b-8192\", \"llama-3.1-70b-versatile\", \"llama-3.1-8b-instant\", \"gemma2-9b-it\"], value=\"llama3-70b-8192\"),\n",
+    "        gr.Slider(label=\"Generating Batch Size\", value=3, minimum=1, maximum=10, step=1)\n",
+    "    ],\n",
+    "    outputs=[\n",
+    "        gr.Textbox(label=\"Description\"),\n",
+    "        gr.DataFrame(label=\"Examples\", headers=[\"Input\", \"Output\"])\n",
+    "    ],\n",
     "    title=\"Task Description Generator\",\n",
-    "    description=\"Enter a JSON object with 'input' and 'output' fields to generate a task description and additional examples.\"\n",
+    "    description=\"Enter a JSON object with 'input' and 'output' fields to generate a task description and additional examples.\",\n",
+    "    allow_flagging=\"manual\",\n",
+    "    flagging_callback=gr.CSVLogger()\n",
     ")\n",
     "\n",
     "if __name__ == \"__main__\":\n",