initial commit

.gitignore

@@ -0,0 +1,2 @@
+build
+*.egg-info

ThirdPartyNotices.txt

@@ -0,0 +1,32 @@
+THIRD-PARTY SOFTWARE NOTICES AND INFORMATION
+Do Not Translate or Localize
+
+This project incorporates components from the projects listed below.
+
+1.	langchain (https://github.com/langchain-ai/langchain)
+
+langchain NOTICES AND INFORMATION BEGIN HERE
+=========================================
+The MIT License
+
+Copyright (c) Harrison Chase
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.
+=========================================
+END OF langchain NOTICES AND INFORMATION

notebooks/demo.ipynb

@@ -0,0 +1,147 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "! pip install ../\n",
+    "! pip install matplotlib"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import learn_to_pick"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "class fake_llm_caller:\n",
+    "    def predict(self, message):\n",
+    "        return \"5\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "pick = learn_to_pick.PickBest.create(llm=fake_llm_caller(), metrics_step=5, metrics_window_size=5)\n",
+    "random_pick = learn_to_pick.PickBest.create(\n",
+    "    llm=fake_llm_caller(),\n",
+    "    metrics_step=5,\n",
+    "    metrics_window_size=5, # rolling window average\n",
+    "    policy=learn_to_pick.PickBestRandomPolicy # set the random policy instead of default\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# four meals defined, some vegetarian some not\n",
+    "\n",
+    "meals = [\n",
+    "    \"Beef Enchiladas with Feta cheese. Mexican-Greek fusion\",\n",
+    "    \"Chicken Flatbreads with red sauce. Italian-Mexican fusion\",\n",
+    "    \"Veggie sweet potato quesadillas with vegan cheese\",\n",
+    "    \"One-Pan Tortelonni bake with peppers and onions\",\n",
+    "]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# TODO function hook to call LLM between predict and learn?\n",
+    "# TODO how to pass extra args to scorer? How to get LLMs response back to user and in scorer?\n",
+    "# TODO slates from the get go?\n",
+    "for _ in range(100):\n",
+    "    r = pick.run(meal = learn_to_pick.ToSelectFrom(meals),\n",
+    "        user = learn_to_pick.BasedOn(\"Anna\"),\n",
+    "        preference = learn_to_pick.BasedOn([\"Meat eater\", \"loves beef\"]),\n",
+    "        text_to_personalize = \"This is the weeks specialty dish, our master chefs \\\n",
+    "            believe you will love it!\",)\n",
+    "\n",
+    "    r = pick.run(meal = learn_to_pick.ToSelectFrom(meals),\n",
+    "        user = learn_to_pick.BasedOn(\"Tom\"),\n",
+    "        preference = learn_to_pick.BasedOn([\"Vegetarian\", \"regular dairy is ok\"]),\n",
+    "        text_to_personalize = \"This is the weeks specialty dish, our master chefs \\\n",
+    "            believe you will love it!\",)\n",
+    "    \n",
+    "    r = random_pick.run(meal = learn_to_pick.ToSelectFrom(meals),\n",
+    "        user = learn_to_pick.BasedOn(\"Anna\"),\n",
+    "        preference = learn_to_pick.BasedOn([\"Meat eater\", \"loves beef\"]),\n",
+    "        text_to_personalize = \"This is the weeks specialty dish, our master chefs \\\n",
+    "            believe you will love it!\",)\n",
+    "\n",
+    "    r = random_pick.run(meal = learn_to_pick.ToSelectFrom(meals),\n",
+    "        user = learn_to_pick.BasedOn(\"Tom\"),\n",
+    "        preference = learn_to_pick.BasedOn([\"Vegetarian\", \"regular dairy is ok\"]),\n",
+    "        text_to_personalize = \"This is the weeks specialty dish, our master chefs \\\n",
+    "            believe you will love it!\",)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from matplotlib import pyplot as plt\n",
+    "pick.metrics.to_pandas()['score'].plot(label=\"vw learning policy\")\n",
+    "random_pick.metrics.to_pandas()['score'].plot(label=\"random learning policy\")\n",
+    "plt.legend()\n",
+    "\n",
+    "print(f\"VW, calculated over a rolling window, is: {pick.metrics.to_pandas()['score'].iloc[-1]}\")\n",
+    "print(f\"Random, calculated over a rolling window, is: {random_pick.metrics.to_pandas()['score'].iloc[-1]}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(r[\"picked_metadata\"].selected.score)\n",
+    "print(r[\"picked\"])\n",
+    "print(r[\"picked_metadata\"].to_select_from)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

setup.py

@@ -0,0 +1,21 @@
+from setuptools import setup, find_packages
+
+setup(
+    name="learn_to_pick",
+    version="0.1",
+    install_requires=[
+        'numpy',
+        'pandas',
+        'vowpal-wabbit-next',
+        'sentence-transformers',
+        'torch',
+        'pyskiplist',
+        'parameterfree',
+    ],
+    author="VowpalWabbit",
+    description="",
+    packages=find_packages(where="src"),
+    package_dir={"": "src"},
+    url="https://github.com/VowpalWabbit/learn_to_pick",
+    python_requires='>=3.8',
+)

src/learn_to_pick/__init__.py

@@ -0,0 +1,54 @@
+import logging
+
+from learn_to_pick.base import (
+    AutoSelectionScorer,
+    BasedOn,
+    Embed,
+    Embedder,
+    Policy,
+    SelectionScorer,
+    ToSelectFrom,
+    VwPolicy,
+    embed,
+    stringify_embedding,
+)
+from learn_to_pick.pick_best import (
+    PickBest,
+    PickBestEvent,
+    PickBestFeatureEmbedder,
+    PickBestRandomPolicy,
+    PickBestSelected,
+)
+
+
+def configure_logger() -> None:
+    logger = logging.getLogger(__name__)
+    logger.setLevel(logging.INFO)
+    ch = logging.StreamHandler()
+    formatter = logging.Formatter(
+        "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
+    )
+    ch.setFormatter(formatter)
+    ch.setLevel(logging.INFO)
+    logger.addHandler(ch)
+
+
+configure_logger()
+
+__all__ = [
+    "PickBest",
+    "PickBestEvent",
+    "PickBestSelected",
+    "PickBestFeatureEmbedder",
+    "PickBestRandomPolicy",
+    "Embed",
+    "BasedOn",
+    "ToSelectFrom",
+    "SelectionScorer",
+    "AutoSelectionScorer",
+    "Embedder",
+    "Policy",
+    "VwPolicy",
+    "embed",
+    "stringify_embedding",
+]