{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9c480696",
   "metadata": {},
   "outputs": [],
   "source": [
    "%cd /home/ubuntu/Qwen-Image-Edit-Angles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "671fc929",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import subprocess\n",
    "from pathlib import Path\n",
    "import argparse\n",
    "import warnings\n",
    "\n",
    "import yaml\n",
    "import diffusers\n",
    "\n",
    "\n",
    "from wandml.trainers.experiment_trainer import ExperimentTrainer\n",
    "from wandml import WandDataPipe\n",
    "import wandml\n",
    "from wandml import WandAuth\n",
    "from wandml import utils as wandml_utils\n",
    "from wandml.trainers.datamodels import ExperimentTrainerParameters\n",
    "from wandml.trainers.experiment_trainer import ExperimentTrainer\n",
    "\n",
    "\n",
    "from qwenimage.finetuner import QwenLoraFinetuner\n",
    "from qwenimage.sources import EditingSource, RegressionSource, StyleSourceWithRandomRef, StyleImagetoImageSource\n",
    "from qwenimage.task import RegressionTask, TextToImageWithRefTask\n",
    "from qwenimage.datamodels import QwenConfig\n",
    "from qwenimage.foundation import QwenImageFoundation, QwenImageRegressionFoundation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6a646ed7",
   "metadata": {},
   "outputs": [],
   "source": [
    "src = EditingSource(\n",
    "    data_dir=\"/data/CrispEdit\",\n",
    "    total_per=10,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35dda5f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "config = QwenConfig(\n",
    "    training_type=\"regression\",\n",
    "    regression_base_pipe_steps=8,\n",
    ")\n",
    "foundation = QwenImageRegressionFoundation(config=config)\n",
    "# finetuner = QwenLoraFinetuner(foundation, config)\n",
    "# finetuner.load(\"/data/checkpoints/reg-mse-pixel-mse_015000\", config.lora_rank)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d5ddfa5",
   "metadata": {},
   "outputs": [],
   "source": [
    "src[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "264070a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "foundation.config.regression_base_pipe_steps = 4\n",
    "\n",
    "inp = src[100]\n",
    "\n",
    "out = foundation.base_pipe(foundation.INPUT_MODEL(\n",
    "    prompt=inp[0],\n",
    "    image=[inp[2]],\n",
    "))\n",
    "out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "46a09a41",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(inp[0])\n",
    "inp[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "da4157ea",
   "metadata": {},
   "outputs": [],
   "source": [
    "out[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7446e6b3",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}