{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Cast civitai trained LoRa in torch.bfloat16 to Tensor Art Compatible torch.float16 dtype\n",
        "\n",
        "Created by Adcom: https://tensor.art/u/743241123023077878"
      ],
      "metadata": {
        "id": "YDCnQpDdqDe4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#initialize\n",
        "import torch\n",
        "from safetensors.torch import load_file, save_file\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "id": "CBVTifA_ZwdC",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1eda0d91-edb2-4c2a-ddb0-aa2174c25519"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "from safetensors.torch import load_file, save_file\n",
        "import torch.nn as nn\n",
        "from torch import linalg as LA\n",
        "import os\n",
        "import math\n",
        "import random\n",
        "import numpy as np\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "\n",
        "# This method rescales a _lora to a given ratio. I haven't tested it\n",
        "# yet but this is more or less how it works\n",
        "def rescale_and_save(_lora , savefile_name, new_ratio , rank):\n",
        "    count = 0\n",
        "    lora = {}\n",
        "    for key in _lora:count = count + 1\n",
        "    NUM_ITEMS = count\n",
        "    count = 0\n",
        "    decimals = 6\n",
        "    for key in _lora:\n",
        "        if not f'{key}'.find('alpha') > -1: continue\n",
        "        alpha = f'{key}'\n",
        "        up = f'{key}'.replace('alpha' , 'lora_up.weight')\n",
        "        down = f'{key}'.replace('alpha' , 'lora_down.weight')\n",
        "        #------#\n",
        "        new_alpha = torch.tensor(new_ratio*rank).to(device = device , dtype=torch.float32)\n",
        "        lora[up] = torch.round(torch.sqrt(_lora[alpha]/new_alpha)*_lora[up], decimals = decimals).to(device = device , dtype=torch.float32)\n",
        "        lora[down] = torch.round(torch.sqrt(_lora[alpha]/new_alpha)*_lora[down], decimals = decimals).to(device = device , dtype=torch.float32)\n",
        "        #-----#\n",
        "        lora[alpha] = (new_alpha/_lora[alpha])*_lora[alpha].to(device = device , dtype=torch.float32)\n",
        "        count = count + 3\n",
        "        print(f'{count} / {NUM_ITEMS}')\n",
        "    #--------#\n",
        "    print(f'done!')\n",
        "    print(f'casting params to fp16....')\n",
        "    for key in lora: lora[f'{key}'] = lora[f'{key}'].to(device = device , dtype=torch.float16)\n",
        "    #-------#\n",
        "    print(f'done!')\n",
        "    print(f'saving {savefile_name}...')\n",
        "    save_file(lora , f'{savefile_name}')\n",
        "  #-----------#\n",
        "\n",
        "tgt = load_file('/content/drive/MyDrive/Saved from Chrome/window-voyeur- F.safetensors')\n",
        "for key in tgt:\n",
        "  if f'{key}'.find('alpha')>-1: print(tgt[f'{key}'])\n",
        "  print(f\" {key} : {tgt[f'{key}'].shape}\")\n",
        "\n",
        "name = 'window'\n",
        "savefile_name = f'{name}.safetensors'\n",
        "new_ratio = 0.5\n",
        "rank = 4\n",
        "\n",
        "rescale_and_save(tgt , savefile_name, new_ratio , rank)\n",
        "  #(alpha/scale) = (32/16)\n",
        "\n"
      ],
      "metadata": {
        "collapsed": true,
        "id": "PiXcXp7krKMt",
        "outputId": "6e15f27a-ae62-4f58-c4a0-fba448bb7028",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_0_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_0_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_0_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_0_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_0_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_0_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_0_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_10_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_10_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_10_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_10_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_10_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_10_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_10_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_11_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_11_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_11_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_11_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_11_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_11_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_11_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_1_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_1_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_1_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_1_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_1_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_1_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_1_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_2_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_2_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_2_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_2_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_2_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_2_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_2_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_3_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_3_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_3_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_3_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_3_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_3_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_3_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_4_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_4_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_4_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_4_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_4_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_4_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_4_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_5_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_5_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_5_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_5_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_5_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_5_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_5_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_6_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_6_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_6_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_6_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_6_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_6_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_6_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_7_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_7_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_7_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_7_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_7_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_7_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_7_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_8_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_8_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_8_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_8_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_8_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_8_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_8_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_9_mlp_fc1.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_9_mlp_fc1.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_9_mlp_fc1.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_9_mlp_fc2.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_9_mlp_fc2.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_te1_text_model_encoder_layers_9_mlp_fc2.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_k_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_k_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_k_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_out_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_out_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_out_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_q_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_q_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_q_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_v_proj.alpha : torch.Size([])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_v_proj.lora_down.weight : torch.Size([4, 768])\n",
            " lora_te1_text_model_encoder_layers_9_self_attn_v_proj.lora_up.weight : torch.Size([768, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_0_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_0_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_0_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_0_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_0_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_0_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_0_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_0_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_0_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_10_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_10_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_10_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_10_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_10_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_10_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_10_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_10_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_10_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_11_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_11_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_11_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_11_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_11_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_11_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_11_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_11_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_11_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_12_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_12_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_12_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_12_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_12_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_12_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_12_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_12_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_12_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_13_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_13_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_13_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_13_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_13_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_13_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_13_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_13_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_13_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_14_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_14_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_14_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_14_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_14_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_14_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_14_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_14_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_14_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_15_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_15_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_15_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_15_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_15_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_15_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_15_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_15_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_15_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_16_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_16_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_16_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_16_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_16_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_16_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_16_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_16_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_16_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_17_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_17_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_17_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_17_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_17_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_17_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_17_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_17_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_17_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_18_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_18_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_18_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_18_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_18_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_18_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_18_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_18_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_18_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_19_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_19_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_19_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_19_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_19_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_19_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_19_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_19_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_19_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_1_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_1_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_1_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_1_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_1_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_1_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_1_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_1_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_1_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_20_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_20_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_20_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_20_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_20_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_20_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_20_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_20_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_20_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_21_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_21_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_21_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_21_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_21_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_21_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_21_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_21_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_21_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_22_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_22_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_22_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_22_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_22_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_22_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_22_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_22_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_22_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_23_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_23_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_23_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_23_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_23_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_23_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_23_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_23_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_23_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_24_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_24_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_24_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_24_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_24_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_24_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_24_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_24_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_24_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_25_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_25_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_25_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_25_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_25_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_25_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_25_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_25_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_25_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_26_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_26_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_26_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_26_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_26_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_26_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_26_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_26_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_26_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_27_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_27_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_27_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_27_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_27_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_27_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_27_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_27_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_27_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_28_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_28_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_28_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_28_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_28_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_28_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_28_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_28_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_28_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_29_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_29_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_29_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_29_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_29_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_29_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_29_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_29_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_29_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_2_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_2_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_2_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_2_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_2_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_2_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_2_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_2_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_2_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_30_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_30_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_30_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_30_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_30_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_30_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_30_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_30_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_30_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_31_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_31_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_31_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_31_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_31_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_31_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_31_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_31_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_31_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_32_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_32_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_32_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_32_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_32_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_32_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_32_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_32_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_32_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_33_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_33_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_33_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_33_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_33_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_33_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_33_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_33_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_33_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_34_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_34_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_34_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_34_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_34_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_34_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_34_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_34_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_34_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_35_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_35_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_35_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_35_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_35_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_35_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_35_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_35_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_35_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_36_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_36_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_36_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_36_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_36_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_36_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_36_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_36_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_36_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_37_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_37_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_37_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_37_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_37_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_37_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_37_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_37_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_37_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_3_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_3_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_3_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_3_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_3_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_3_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_3_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_3_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_3_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_4_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_4_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_4_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_4_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_4_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_4_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_4_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_4_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_4_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_5_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_5_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_5_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_5_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_5_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_5_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_5_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_5_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_5_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_6_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_6_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_6_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_6_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_6_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_6_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_6_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_6_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_6_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_7_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_7_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_7_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_7_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_7_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_7_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_7_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_7_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_7_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_8_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_8_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_8_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_8_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_8_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_8_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_8_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_8_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_8_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_9_linear1.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_9_linear1.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_9_linear1.lora_up.weight : torch.Size([21504, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_9_linear2.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_9_linear2.lora_down.weight : torch.Size([4, 15360])\n",
            " lora_unet_single_blocks_9_linear2.lora_up.weight : torch.Size([3072, 4])\n",
            "tensor(1., dtype=torch.bfloat16)\n",
            " lora_unet_single_blocks_9_modulation_lin.alpha : torch.Size([])\n",
            " lora_unet_single_blocks_9_modulation_lin.lora_down.weight : torch.Size([4, 3072])\n",
            " lora_unet_single_blocks_9_modulation_lin.lora_up.weight : torch.Size([9216, 4])\n",
            "3 / 558\n",
            "6 / 558\n",
            "9 / 558\n",
            "12 / 558\n",
            "15 / 558\n",
            "18 / 558\n",
            "21 / 558\n",
            "24 / 558\n",
            "27 / 558\n",
            "30 / 558\n",
            "33 / 558\n",
            "36 / 558\n",
            "39 / 558\n",
            "42 / 558\n",
            "45 / 558\n",
            "48 / 558\n",
            "51 / 558\n",
            "54 / 558\n",
            "57 / 558\n",
            "60 / 558\n",
            "63 / 558\n",
            "66 / 558\n",
            "69 / 558\n",
            "72 / 558\n",
            "75 / 558\n",
            "78 / 558\n",
            "81 / 558\n",
            "84 / 558\n",
            "87 / 558\n",
            "90 / 558\n",
            "93 / 558\n",
            "96 / 558\n",
            "99 / 558\n",
            "102 / 558\n",
            "105 / 558\n",
            "108 / 558\n",
            "111 / 558\n",
            "114 / 558\n",
            "117 / 558\n",
            "120 / 558\n",
            "123 / 558\n",
            "126 / 558\n",
            "129 / 558\n",
            "132 / 558\n",
            "135 / 558\n",
            "138 / 558\n",
            "141 / 558\n",
            "144 / 558\n",
            "147 / 558\n",
            "150 / 558\n",
            "153 / 558\n",
            "156 / 558\n",
            "159 / 558\n",
            "162 / 558\n",
            "165 / 558\n",
            "168 / 558\n",
            "171 / 558\n",
            "174 / 558\n",
            "177 / 558\n",
            "180 / 558\n",
            "183 / 558\n",
            "186 / 558\n",
            "189 / 558\n",
            "192 / 558\n",
            "195 / 558\n",
            "198 / 558\n",
            "201 / 558\n",
            "204 / 558\n",
            "207 / 558\n",
            "210 / 558\n",
            "213 / 558\n",
            "216 / 558\n",
            "219 / 558\n",
            "222 / 558\n",
            "225 / 558\n",
            "228 / 558\n",
            "231 / 558\n",
            "234 / 558\n",
            "237 / 558\n",
            "240 / 558\n",
            "243 / 558\n",
            "246 / 558\n",
            "249 / 558\n",
            "252 / 558\n",
            "255 / 558\n",
            "258 / 558\n",
            "261 / 558\n",
            "264 / 558\n",
            "267 / 558\n",
            "270 / 558\n",
            "273 / 558\n",
            "276 / 558\n",
            "279 / 558\n",
            "282 / 558\n",
            "285 / 558\n",
            "288 / 558\n",
            "291 / 558\n",
            "294 / 558\n",
            "297 / 558\n",
            "300 / 558\n",
            "303 / 558\n",
            "306 / 558\n",
            "309 / 558\n",
            "312 / 558\n",
            "315 / 558\n",
            "318 / 558\n",
            "321 / 558\n",
            "324 / 558\n",
            "327 / 558\n",
            "330 / 558\n",
            "333 / 558\n",
            "336 / 558\n",
            "339 / 558\n",
            "342 / 558\n",
            "345 / 558\n",
            "348 / 558\n",
            "351 / 558\n",
            "354 / 558\n",
            "357 / 558\n",
            "360 / 558\n",
            "363 / 558\n",
            "366 / 558\n",
            "369 / 558\n",
            "372 / 558\n",
            "375 / 558\n",
            "378 / 558\n",
            "381 / 558\n",
            "384 / 558\n",
            "387 / 558\n",
            "390 / 558\n",
            "393 / 558\n",
            "396 / 558\n",
            "399 / 558\n",
            "402 / 558\n",
            "405 / 558\n",
            "408 / 558\n",
            "411 / 558\n",
            "414 / 558\n",
            "417 / 558\n",
            "420 / 558\n",
            "423 / 558\n",
            "426 / 558\n",
            "429 / 558\n",
            "432 / 558\n",
            "435 / 558\n",
            "438 / 558\n",
            "441 / 558\n",
            "444 / 558\n",
            "447 / 558\n",
            "450 / 558\n",
            "453 / 558\n",
            "456 / 558\n",
            "459 / 558\n",
            "462 / 558\n",
            "465 / 558\n",
            "468 / 558\n",
            "471 / 558\n",
            "474 / 558\n",
            "477 / 558\n",
            "480 / 558\n",
            "483 / 558\n",
            "486 / 558\n",
            "489 / 558\n",
            "492 / 558\n",
            "495 / 558\n",
            "498 / 558\n",
            "501 / 558\n",
            "504 / 558\n",
            "507 / 558\n",
            "510 / 558\n",
            "513 / 558\n",
            "516 / 558\n",
            "519 / 558\n",
            "522 / 558\n",
            "525 / 558\n",
            "528 / 558\n",
            "531 / 558\n",
            "534 / 558\n",
            "537 / 558\n",
            "540 / 558\n",
            "543 / 558\n",
            "546 / 558\n",
            "549 / 558\n",
            "552 / 558\n",
            "555 / 558\n",
            "558 / 558\n",
            "done!\n",
            "casting params to fp16....\n",
            "done!\n",
            "saving window.safetensors...\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tgt = load_file('/content/doggy.safetensors')\n",
        "for key in tgt:\n",
        "  if f'{key}'.find('alpha')>-1: print(tgt[f'{key}'])\n",
        "  print(tgt[f'{key}'].shape)"
      ],
      "metadata": {
        "id": "qk0Lbf27vBjl",
        "outputId": "109eaa9f-6941-4cb7-d084-ca54eac1ac6c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 768])\n",
            "torch.Size([768, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([21504, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "tensor(16., dtype=torch.float16)\n",
            "torch.Size([])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "import torch\n",
        "from safetensors.torch import load_file, save_file\n",
        "import torch.nn as nn\n",
        "from torch import linalg as LA\n",
        "import os\n",
        "import math\n",
        "import random\n",
        "import numpy as np\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "\n",
        "# For pcnt = 30 , 'filter_and_save'  will keep all top 30 % values\n",
        "#, and the lowest (negative) 30% values for each layer delta_W in this lora\n",
        "# Then save the new filtered lora as a .safetensor file\n",
        "def filter_and_save(_lora , savefile_name, new_rank , new_alpha, resolution):\n",
        "    lora = {}\n",
        "    count = 0\n",
        "    for key in _lora:count = count + 1\n",
        "    NUM_ITEMS = count\n",
        "    count = 0\n",
        "    thresh = resolution*0.000001 # 1e-6\n",
        "    #-------#\n",
        "    for key in _lora:\n",
        "        if f'{key}'.find('alpha') > -1:\n",
        "            lora[f'{key}'] = torch.tensor(new_alpha).to(device = device , dtype = torch.float32)\n",
        "            count = count + 1\n",
        "            print(f'{count} / {NUM_ITEMS}')\n",
        "            continue\n",
        "        #------#\n",
        "        if not f'{key}'.find('lora_down') > -1: continue\n",
        "        up = f'{key}'.replace('lora_down' , 'lora_up')\n",
        "        down = f'{key}'\n",
        "        #-------#\n",
        "        delta_W = torch.matmul(_lora[up],_lora[down]).to(device = device , dtype=torch.float32)\n",
        "        #---#\n",
        "        N = delta_W.numel()\n",
        "        y = delta_W.flatten().to(device = device , dtype=torch.float32)\n",
        "        values,indices = torch.sort(y, descending = False) # smallest -> largest elements\n",
        "        y = torch.zeros(y.shape).to(device = device , dtype=torch.float32)\n",
        "        y[indices[values>thresh]] = 1\n",
        "        y[indices[values<-thresh]] = 1\n",
        "        y = y.unflatten(0,delta_W.shape).to(device = device , dtype=torch.float32)\n",
        "        delta_W = torch.mul(delta_W,y).to(device = device , dtype=torch.float32)\n",
        "        #------#\n",
        "        tmp={}\n",
        "        tmp['u'], tmp['s'], tmp['Vh'] = torch.svd(delta_W)\n",
        "        tmp['u'] = tmp['u'][:,: new_rank]\n",
        "        tmp['s'] = tmp['s'][: new_rank]\n",
        "        #-------#\n",
        "        tmp['u'] = torch.round(torch.matmul(tmp['u'], torch.diag(tmp['s'])),decimals=6)\n",
        "        tmp['Vh'] = torch.round(tmp['Vh'].t()[: new_rank,:],decimals=6)\n",
        "        #-------#\n",
        "        for key in tmp:tmp[f'{key}'] = tmp[f'{key}'].contiguous()\n",
        "        lora[up] = tmp['u'].to(device = device , dtype=torch.float32)\n",
        "        lora[down] = tmp['Vh'].to(device = device , dtype=torch.float32)\n",
        "        #-------#\n",
        "        count  = count  + 2\n",
        "        print(f'{count} / {NUM_ITEMS}')\n",
        "    #-------#\n",
        "    print(f'done!')\n",
        "    print(f'casting params to fp16....')\n",
        "    for key in _lora: lora[f'{key}'] = lora[f'{key}'].to(device = device , dtype=torch.float16)\n",
        "    #-------#\n",
        "    print(f'done!')\n",
        "    print(f'saving {savefile_name}...')\n",
        "    save_file(lora , f'{savefile_name}')\n",
        "#--------#\n",
        "\n",
        "def count_zeros(_lora, resolution):\n",
        "    count = 0\n",
        "    for key in _lora:count = count + 1\n",
        "    NUM_ITEMS = count\n",
        "    count = 0\n",
        "    #-----#\n",
        "    thresh = resolution*0.000001 # 1e-6\n",
        "\n",
        "    print(f'at resolution = {resolution}e-6 :')\n",
        "    for key in _lora:\n",
        "        if f'{key}'.find('alpha') > -1:\n",
        "            count = count + 1\n",
        "            continue\n",
        "        #------#\n",
        "        if not f'{key}'.find('lora_down') > -1: continue\n",
        "        up = f'{key}'.replace('lora_down' , 'lora_up')\n",
        "        down = f'{key}'\n",
        "        #-------#\n",
        "        delta_W = torch.matmul(_lora[up],_lora[down]).to(device = device , dtype=torch.float32)\n",
        "        N = delta_W.numel()\n",
        "        y = delta_W.flatten().to(device = device , dtype=torch.float32)\n",
        "        values,indices = torch.sort(y, descending = False) # smallest -> largest elements\n",
        "        y = torch.ones(y.shape).to(device = device , dtype=torch.float32)\n",
        "        y[indices[values>thresh]] = 0\n",
        "        neg_pcnt  = round((100*torch.sum(y) / N).item(),2)\n",
        "        y[indices[values<-thresh]] = 0\n",
        "        count  = count  + 2\n",
        "        pcnt  = round((100*torch.sum(y) / N).item(),2)\n",
        "        neg_pcnt = round(neg_pcnt - pcnt,2) # remove zero % from neg_pcnt\n",
        "        pos_pcnt = round(100- pcnt - neg_pcnt,2)\n",
        "        print(f'at {count} / {NUM_ITEMS} : {pcnt} % zeros ,{pos_pcnt} % pos. , {neg_pcnt} % neg ')\n",
        "    #------#\n",
        "#-----#\n",
        "\n",
        "def merge_and_save(_lora1 , _lora2 , _lora3, savefile_name, new_rank , new_alpha, resolution):\n",
        "    lora = {}\n",
        "    count = 0\n",
        "    for key in _lora1:count = count + 1\n",
        "    NUM_ITEMS = count\n",
        "    count = 0\n",
        "    thresh = resolution*0.000001 # 1e-6\n",
        "\n",
        "    #-------#\n",
        "    for key in _lora1:\n",
        "        if f'{key}'.find('alpha') > -1:\n",
        "            lora[f'{key}'] = torch.tensor(new_alpha).to(device = device , dtype = torch.float32)\n",
        "            count = count + 1\n",
        "            print(f'{count} / {NUM_ITEMS}')\n",
        "            continue\n",
        "        #------#\n",
        "        #if count<462:\n",
        "        #    count = count + 2\n",
        "        #    continue\n",
        "        if not f'{key}'.find('lora_down') > -1: continue\n",
        "        up = f'{key}'.replace('lora_down' , 'lora_up')\n",
        "        down = f'{key}'\n",
        "        #-------#\n",
        "\n",
        "        # Setup\n",
        "        delta_W = torch.matmul(_lora1[up]*0,_lora1[down]*0).to(device = device, dtype=torch.float32)\n",
        "        tgt_shape = delta_W.shape\n",
        "        N = delta_W.numel()\n",
        "        delta_W = torch.zeros(N).to(device = device , dtype=torch.float32)\n",
        "        #-----#\n",
        "\n",
        "        #Positives\n",
        "        Y = torch.zeros(3,N).to(device = device , dtype=torch.float32)\n",
        "        Y[0] = torch.matmul(_lora1[up],_lora1[down]).flatten().to(device = device , dtype=torch.float32)\n",
        "        Y[1] = torch.matmul(_lora2[up],_lora2[down]).flatten().to(device = device , dtype=torch.float32)\n",
        "        Y[2] = torch.matmul(_lora3[up],_lora3[down]).flatten().to(device = device , dtype=torch.float32)\n",
        "        Y[torch.abs(Y)<thresh] = 0 #Trim\n",
        "        Y = torch.round(Y,decimals=6).t()\n",
        "        num = torch.sum(Y>0,dim=1) + 0.001\n",
        "        elect = torch.sum(Y<0,dim=1) + 0.001\n",
        "        elect = (num>=elect)\n",
        "        Y[Y<0] = 0\n",
        "        Y = torch.sum(Y, dim=1).to(device = device , dtype=torch.float32)\n",
        "        delta_W[elect] = torch.round((Y[elect]/num[elect]),decimals=6).to(device = device , dtype=torch.float32)\n",
        "        #-----#\n",
        "\n",
        "        #Negatives\n",
        "        Y = torch.zeros(3,N).to(device = device , dtype=torch.float32)\n",
        "        Y[0] = torch.matmul(_lora1[up],_lora1[down]).flatten().to(device = device , dtype=torch.float32)\n",
        "        Y[1] = torch.matmul(_lora2[up],_lora2[down]).flatten().to(device = device , dtype=torch.float32)\n",
        "        Y[2] = torch.matmul(_lora3[up],_lora3[down]).flatten().to(device = device , dtype=torch.float32)\n",
        "        Y[torch.abs(Y)<thresh] = 0 #Trim\n",
        "        Y = torch.round(Y,decimals=6).t()\n",
        "        num = torch.sum(Y<0,dim=1) + 0.001  # <-- to prevent divide by 0\n",
        "        elect = torch.sum(Y>0,dim=1) + 0.001\n",
        "        elect = (elect<num)\n",
        "        Y[Y>0] = 0\n",
        "        Y = torch.sum(Y, dim=1).to(device = device , dtype=torch.float32)\n",
        "        delta_W[elect] = torch.round(Y[elect]/num[elect],decimals=6).to(device = device , dtype=torch.float32)\n",
        "        #----#\n",
        "\n",
        "        # Free up memory prior to SVD\n",
        "        delta_W = delta_W.unflatten(0,tgt_shape).to(device = device , dtype=torch.float32)\n",
        "        delta_W = delta_W.clone().detach()\n",
        "        Y = {}\n",
        "        num = {}\n",
        "        num = {}\n",
        "        elect = {}\n",
        "        #-----#\n",
        "\n",
        "        # Run SVD (Single Value Decomposition)\n",
        "        #to get the new lora_up and lora_down for delta_W\n",
        "        tmp={}\n",
        "        tmp['u'], tmp['s'], tmp['Vh'] = torch.svd(delta_W)\n",
        "        tmp['u'] = tmp['u'][:,: new_rank]\n",
        "        tmp['s'] = tmp['s'][: new_rank]\n",
        "        tmp['u'] = torch.matmul(tmp['u'], torch.diag(tmp['s']))\n",
        "        tmp['Vh'] = tmp['Vh'].t()[: new_rank,:]\n",
        "        for key in tmp:tmp[f'{key}'] = tmp[f'{key}'].contiguous()\n",
        "        lora[up] = torch.round(tmp['u'],decimals=6).to(device = device , dtype=torch.float32)\n",
        "        lora[down] = torch.round(tmp['Vh'],decimals=6).to(device = device , dtype=torch.float32)\n",
        "        #-------#\n",
        "\n",
        "        count  = count  + 2\n",
        "        print(f'{count} / {NUM_ITEMS}')\n",
        "        #----#\n",
        "    #--------#\n",
        "    print(f'done!')\n",
        "    print(f'casting params to fp16....')\n",
        "    for key in lora: lora[f'{key}'] = lora[f'{key}'].to(device = device , dtype=torch.float16)\n",
        "    #-------#\n",
        "    print(f'done!')\n",
        "    print(f'saving {savefile_name}...')\n",
        "    save_file(lora , f'{savefile_name}')\n",
        "#------#\n",
        "\n",
        "new_rank = 32\n",
        "new_alpha = math.floor(new_rank/2)\n",
        "resolution = 200\n",
        "name = 'yeero_euro_puff'\n",
        "yeero = load_file('/content/drive/MyDrive/Saved from Chrome/yeero_100_r32_16alpha.safetensors')\n",
        "euro = load_file('/content/drive/MyDrive/Saved from Chrome/euro_100_r32_16alpha.safetensors')\n",
        "puff = load_file('/content/drive/MyDrive/Saved from Chrome/puff_200_r32_16alpha.safetensors')\n",
        "savefile_name = f'{name}_{resolution}_r{new_rank}_a{new_alpha}.safetensors'\n",
        "\n",
        "#tgt = load_file(f'/kaggle/input/flux-loras/{name}_{resolution}_r32_16alpha.safetensors')\n",
        "for key in yeero:\n",
        "    yeero[f'{key}'] = yeero[f'{key}'].to(device = device , dtype = torch.float32)\n",
        "    euro[f'{key}'] = euro[f'{key}'].to(device = device , dtype = torch.float32)\n",
        "    puff[f'{key}'] = puff[f'{key}'].to(device = device , dtype = torch.float32)\n",
        "#-----#\n",
        "print(f'for {name}.safetensors at scale = (rank/alpha) = 0.5')\n",
        "merge_and_save(yeero , euro , puff, savefile_name, new_rank , new_alpha, resolution)\n",
        "\n",
        "\n",
        "#Yeero + Scale + Puff\n",
        "#filter_and_save(tgt , f'{name}_{resolution}_r{new_rank}_{new_alpha}alpha.safetensors' , new_rank , new_alpha, resolution)\n"
      ],
      "metadata": {
        "id": "SKYzFxehkfG8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "70f308e8-6632-42ca-e3ce-607e56813778",
        "collapsed": true
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "for yeero_euro_puff.safetensors at scale = (rank/alpha) = 0.5\n",
            "1 / 912\n",
            "3 / 912\n",
            "4 / 912\n",
            "6 / 912\n",
            "7 / 912\n",
            "9 / 912\n",
            "10 / 912\n",
            "12 / 912\n",
            "13 / 912\n",
            "15 / 912\n",
            "16 / 912\n",
            "18 / 912\n",
            "19 / 912\n",
            "21 / 912\n",
            "22 / 912\n",
            "24 / 912\n",
            "25 / 912\n",
            "27 / 912\n",
            "28 / 912\n",
            "30 / 912\n",
            "31 / 912\n",
            "33 / 912\n",
            "34 / 912\n",
            "36 / 912\n",
            "37 / 912\n",
            "39 / 912\n",
            "40 / 912\n",
            "42 / 912\n",
            "43 / 912\n",
            "45 / 912\n",
            "46 / 912\n",
            "48 / 912\n",
            "49 / 912\n",
            "51 / 912\n",
            "52 / 912\n",
            "54 / 912\n",
            "55 / 912\n",
            "57 / 912\n",
            "58 / 912\n",
            "60 / 912\n",
            "61 / 912\n",
            "63 / 912\n",
            "64 / 912\n",
            "66 / 912\n",
            "67 / 912\n",
            "69 / 912\n",
            "70 / 912\n",
            "72 / 912\n",
            "73 / 912\n",
            "75 / 912\n",
            "76 / 912\n",
            "78 / 912\n",
            "79 / 912\n",
            "81 / 912\n",
            "82 / 912\n",
            "84 / 912\n",
            "85 / 912\n",
            "87 / 912\n",
            "88 / 912\n",
            "90 / 912\n",
            "91 / 912\n",
            "93 / 912\n",
            "94 / 912\n",
            "96 / 912\n",
            "97 / 912\n",
            "99 / 912\n",
            "100 / 912\n",
            "102 / 912\n",
            "103 / 912\n",
            "105 / 912\n",
            "106 / 912\n",
            "108 / 912\n",
            "109 / 912\n",
            "111 / 912\n",
            "112 / 912\n",
            "114 / 912\n",
            "115 / 912\n",
            "117 / 912\n",
            "118 / 912\n",
            "120 / 912\n",
            "121 / 912\n",
            "123 / 912\n",
            "124 / 912\n",
            "126 / 912\n",
            "127 / 912\n",
            "129 / 912\n",
            "130 / 912\n",
            "132 / 912\n",
            "133 / 912\n",
            "135 / 912\n",
            "136 / 912\n",
            "138 / 912\n",
            "139 / 912\n",
            "141 / 912\n",
            "142 / 912\n",
            "144 / 912\n",
            "145 / 912\n",
            "147 / 912\n",
            "148 / 912\n",
            "150 / 912\n",
            "151 / 912\n",
            "153 / 912\n",
            "154 / 912\n",
            "156 / 912\n",
            "157 / 912\n",
            "159 / 912\n",
            "160 / 912\n",
            "162 / 912\n",
            "163 / 912\n",
            "165 / 912\n",
            "166 / 912\n",
            "168 / 912\n",
            "169 / 912\n",
            "171 / 912\n",
            "172 / 912\n",
            "174 / 912\n",
            "175 / 912\n",
            "177 / 912\n",
            "178 / 912\n",
            "180 / 912\n",
            "181 / 912\n",
            "183 / 912\n",
            "184 / 912\n",
            "186 / 912\n",
            "187 / 912\n",
            "189 / 912\n",
            "190 / 912\n",
            "192 / 912\n",
            "193 / 912\n",
            "195 / 912\n",
            "196 / 912\n",
            "198 / 912\n",
            "199 / 912\n",
            "201 / 912\n",
            "202 / 912\n",
            "204 / 912\n",
            "205 / 912\n",
            "207 / 912\n",
            "208 / 912\n",
            "210 / 912\n",
            "211 / 912\n",
            "213 / 912\n",
            "214 / 912\n",
            "216 / 912\n",
            "217 / 912\n",
            "219 / 912\n",
            "220 / 912\n",
            "222 / 912\n",
            "223 / 912\n",
            "225 / 912\n",
            "226 / 912\n",
            "228 / 912\n",
            "229 / 912\n",
            "231 / 912\n",
            "232 / 912\n",
            "234 / 912\n",
            "235 / 912\n",
            "237 / 912\n",
            "238 / 912\n",
            "240 / 912\n",
            "241 / 912\n",
            "243 / 912\n",
            "244 / 912\n",
            "246 / 912\n",
            "247 / 912\n",
            "249 / 912\n",
            "250 / 912\n",
            "252 / 912\n",
            "253 / 912\n",
            "255 / 912\n",
            "256 / 912\n",
            "258 / 912\n",
            "259 / 912\n",
            "261 / 912\n",
            "262 / 912\n",
            "264 / 912\n",
            "265 / 912\n",
            "267 / 912\n",
            "268 / 912\n",
            "270 / 912\n",
            "271 / 912\n",
            "273 / 912\n",
            "274 / 912\n",
            "276 / 912\n",
            "277 / 912\n",
            "279 / 912\n",
            "280 / 912\n",
            "282 / 912\n",
            "283 / 912\n",
            "285 / 912\n",
            "286 / 912\n",
            "288 / 912\n",
            "289 / 912\n",
            "291 / 912\n",
            "292 / 912\n",
            "294 / 912\n",
            "295 / 912\n",
            "297 / 912\n",
            "298 / 912\n",
            "300 / 912\n",
            "301 / 912\n",
            "303 / 912\n",
            "304 / 912\n",
            "306 / 912\n",
            "307 / 912\n",
            "309 / 912\n",
            "310 / 912\n",
            "312 / 912\n",
            "313 / 912\n",
            "315 / 912\n",
            "316 / 912\n",
            "318 / 912\n",
            "319 / 912\n",
            "321 / 912\n",
            "322 / 912\n",
            "324 / 912\n",
            "325 / 912\n",
            "327 / 912\n",
            "328 / 912\n",
            "330 / 912\n",
            "331 / 912\n",
            "333 / 912\n",
            "334 / 912\n",
            "336 / 912\n",
            "337 / 912\n",
            "339 / 912\n",
            "340 / 912\n",
            "342 / 912\n",
            "343 / 912\n",
            "345 / 912\n",
            "346 / 912\n",
            "348 / 912\n",
            "349 / 912\n",
            "351 / 912\n",
            "352 / 912\n",
            "354 / 912\n",
            "355 / 912\n",
            "357 / 912\n",
            "358 / 912\n",
            "360 / 912\n",
            "361 / 912\n",
            "363 / 912\n",
            "364 / 912\n",
            "366 / 912\n",
            "367 / 912\n",
            "369 / 912\n",
            "370 / 912\n",
            "372 / 912\n",
            "373 / 912\n",
            "375 / 912\n",
            "376 / 912\n",
            "378 / 912\n",
            "379 / 912\n",
            "381 / 912\n",
            "382 / 912\n",
            "384 / 912\n",
            "385 / 912\n",
            "387 / 912\n",
            "388 / 912\n",
            "390 / 912\n",
            "391 / 912\n",
            "393 / 912\n",
            "394 / 912\n",
            "396 / 912\n",
            "397 / 912\n",
            "399 / 912\n",
            "400 / 912\n",
            "402 / 912\n",
            "403 / 912\n",
            "405 / 912\n",
            "406 / 912\n",
            "408 / 912\n",
            "409 / 912\n",
            "411 / 912\n",
            "412 / 912\n",
            "414 / 912\n",
            "415 / 912\n",
            "417 / 912\n",
            "418 / 912\n",
            "420 / 912\n",
            "421 / 912\n",
            "423 / 912\n",
            "424 / 912\n",
            "426 / 912\n",
            "427 / 912\n",
            "429 / 912\n",
            "430 / 912\n",
            "432 / 912\n",
            "433 / 912\n",
            "435 / 912\n",
            "436 / 912\n",
            "438 / 912\n",
            "439 / 912\n",
            "441 / 912\n",
            "442 / 912\n",
            "444 / 912\n",
            "445 / 912\n",
            "447 / 912\n",
            "448 / 912\n",
            "450 / 912\n",
            "451 / 912\n",
            "453 / 912\n",
            "454 / 912\n",
            "456 / 912\n",
            "457 / 912\n",
            "459 / 912\n",
            "460 / 912\n",
            "462 / 912\n",
            "463 / 912\n",
            "465 / 912\n",
            "466 / 912\n",
            "468 / 912\n",
            "469 / 912\n",
            "471 / 912\n",
            "472 / 912\n",
            "474 / 912\n",
            "475 / 912\n",
            "477 / 912\n",
            "478 / 912\n",
            "480 / 912\n",
            "481 / 912\n",
            "483 / 912\n",
            "484 / 912\n",
            "486 / 912\n",
            "487 / 912\n",
            "489 / 912\n",
            "490 / 912\n",
            "492 / 912\n",
            "493 / 912\n",
            "495 / 912\n",
            "496 / 912\n",
            "498 / 912\n",
            "499 / 912\n",
            "501 / 912\n",
            "502 / 912\n",
            "504 / 912\n",
            "505 / 912\n",
            "507 / 912\n",
            "508 / 912\n",
            "510 / 912\n",
            "511 / 912\n",
            "513 / 912\n",
            "514 / 912\n",
            "516 / 912\n",
            "517 / 912\n",
            "519 / 912\n",
            "520 / 912\n",
            "522 / 912\n",
            "523 / 912\n",
            "525 / 912\n",
            "526 / 912\n",
            "528 / 912\n",
            "529 / 912\n",
            "531 / 912\n",
            "532 / 912\n",
            "534 / 912\n",
            "535 / 912\n",
            "537 / 912\n",
            "538 / 912\n",
            "540 / 912\n",
            "541 / 912\n",
            "543 / 912\n",
            "544 / 912\n",
            "546 / 912\n",
            "547 / 912\n",
            "549 / 912\n",
            "550 / 912\n",
            "552 / 912\n",
            "553 / 912\n",
            "555 / 912\n",
            "556 / 912\n",
            "558 / 912\n",
            "559 / 912\n",
            "561 / 912\n",
            "562 / 912\n",
            "564 / 912\n",
            "565 / 912\n",
            "567 / 912\n",
            "568 / 912\n",
            "570 / 912\n",
            "571 / 912\n",
            "573 / 912\n",
            "574 / 912\n",
            "576 / 912\n",
            "577 / 912\n",
            "579 / 912\n",
            "580 / 912\n",
            "582 / 912\n",
            "583 / 912\n",
            "585 / 912\n",
            "586 / 912\n",
            "588 / 912\n",
            "589 / 912\n",
            "591 / 912\n",
            "592 / 912\n",
            "594 / 912\n",
            "595 / 912\n",
            "597 / 912\n",
            "598 / 912\n",
            "600 / 912\n",
            "601 / 912\n",
            "603 / 912\n",
            "604 / 912\n",
            "606 / 912\n",
            "607 / 912\n",
            "609 / 912\n",
            "610 / 912\n",
            "612 / 912\n",
            "613 / 912\n",
            "615 / 912\n",
            "616 / 912\n",
            "618 / 912\n",
            "619 / 912\n",
            "621 / 912\n",
            "622 / 912\n",
            "624 / 912\n",
            "625 / 912\n",
            "627 / 912\n",
            "628 / 912\n",
            "630 / 912\n",
            "631 / 912\n",
            "633 / 912\n",
            "634 / 912\n",
            "636 / 912\n",
            "637 / 912\n",
            "639 / 912\n",
            "640 / 912\n",
            "642 / 912\n",
            "643 / 912\n",
            "645 / 912\n",
            "646 / 912\n",
            "648 / 912\n",
            "649 / 912\n",
            "651 / 912\n",
            "652 / 912\n",
            "654 / 912\n",
            "655 / 912\n",
            "657 / 912\n",
            "658 / 912\n",
            "660 / 912\n",
            "661 / 912\n",
            "663 / 912\n",
            "664 / 912\n",
            "666 / 912\n",
            "667 / 912\n",
            "669 / 912\n",
            "670 / 912\n",
            "672 / 912\n",
            "673 / 912\n",
            "675 / 912\n",
            "676 / 912\n",
            "678 / 912\n",
            "679 / 912\n",
            "681 / 912\n",
            "682 / 912\n",
            "684 / 912\n",
            "685 / 912\n",
            "687 / 912\n",
            "688 / 912\n",
            "690 / 912\n",
            "691 / 912\n",
            "693 / 912\n",
            "694 / 912\n",
            "696 / 912\n",
            "697 / 912\n",
            "699 / 912\n",
            "700 / 912\n",
            "702 / 912\n",
            "703 / 912\n",
            "705 / 912\n",
            "706 / 912\n",
            "708 / 912\n",
            "709 / 912\n",
            "711 / 912\n",
            "712 / 912\n",
            "714 / 912\n",
            "715 / 912\n",
            "717 / 912\n",
            "718 / 912\n",
            "720 / 912\n",
            "721 / 912\n",
            "723 / 912\n",
            "724 / 912\n",
            "726 / 912\n",
            "727 / 912\n",
            "729 / 912\n",
            "730 / 912\n",
            "732 / 912\n",
            "733 / 912\n",
            "735 / 912\n",
            "736 / 912\n",
            "738 / 912\n",
            "739 / 912\n",
            "741 / 912\n",
            "742 / 912\n",
            "744 / 912\n",
            "745 / 912\n",
            "747 / 912\n",
            "748 / 912\n",
            "750 / 912\n",
            "751 / 912\n",
            "753 / 912\n",
            "754 / 912\n",
            "756 / 912\n",
            "757 / 912\n",
            "759 / 912\n",
            "760 / 912\n",
            "762 / 912\n",
            "763 / 912\n",
            "765 / 912\n",
            "766 / 912\n",
            "768 / 912\n",
            "769 / 912\n",
            "771 / 912\n",
            "772 / 912\n",
            "774 / 912\n",
            "775 / 912\n",
            "777 / 912\n",
            "778 / 912\n",
            "780 / 912\n",
            "781 / 912\n",
            "783 / 912\n",
            "784 / 912\n",
            "786 / 912\n",
            "787 / 912\n",
            "789 / 912\n",
            "790 / 912\n",
            "792 / 912\n",
            "793 / 912\n",
            "795 / 912\n",
            "796 / 912\n",
            "798 / 912\n",
            "799 / 912\n",
            "801 / 912\n",
            "802 / 912\n",
            "804 / 912\n",
            "805 / 912\n",
            "807 / 912\n",
            "808 / 912\n",
            "810 / 912\n",
            "811 / 912\n",
            "813 / 912\n",
            "814 / 912\n",
            "816 / 912\n",
            "817 / 912\n",
            "819 / 912\n",
            "820 / 912\n",
            "822 / 912\n",
            "823 / 912\n",
            "825 / 912\n",
            "826 / 912\n",
            "828 / 912\n",
            "829 / 912\n",
            "831 / 912\n",
            "832 / 912\n",
            "834 / 912\n",
            "835 / 912\n",
            "837 / 912\n",
            "838 / 912\n",
            "840 / 912\n",
            "841 / 912\n",
            "843 / 912\n",
            "844 / 912\n",
            "846 / 912\n",
            "847 / 912\n",
            "849 / 912\n",
            "850 / 912\n",
            "852 / 912\n",
            "853 / 912\n",
            "855 / 912\n",
            "856 / 912\n",
            "858 / 912\n",
            "859 / 912\n",
            "861 / 912\n",
            "862 / 912\n",
            "864 / 912\n",
            "865 / 912\n",
            "867 / 912\n",
            "868 / 912\n",
            "870 / 912\n",
            "871 / 912\n",
            "873 / 912\n",
            "874 / 912\n",
            "876 / 912\n",
            "877 / 912\n",
            "879 / 912\n",
            "880 / 912\n",
            "882 / 912\n",
            "883 / 912\n",
            "885 / 912\n",
            "886 / 912\n",
            "888 / 912\n",
            "889 / 912\n",
            "891 / 912\n",
            "892 / 912\n",
            "894 / 912\n",
            "895 / 912\n",
            "897 / 912\n",
            "898 / 912\n",
            "900 / 912\n",
            "901 / 912\n",
            "903 / 912\n",
            "904 / 912\n",
            "906 / 912\n",
            "907 / 912\n",
            "909 / 912\n",
            "910 / 912\n",
            "912 / 912\n",
            "done!\n",
            "casting params to fp16....\n",
            "done!\n",
            "saving yeero_euro_puff_200_r32_a16.safetensors...\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from safetensors.torch import load_file, save_file\n",
        "_puff = load_file('/content/drive/MyDrive/Saved from Chrome/pfbkFLUX.safetensors')\n",
        "puff = {}\n",
        "\n",
        "#alpha = 64\n",
        "#rank = 64\n",
        "\n",
        "# = > so scale = 1\n",
        "#desired scale = 0.5\n",
        "# so multiply matrices by 2 and set alpha to 32\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "for key in _puff:\n",
        "  if f'{key}'.find('alpha')>-1:\n",
        "    puff[f'{key}'] = torch.tensor(32).to(device=device , dtype = torch.float16)\n",
        "    #print(puff[f'{key}'])\n",
        "    continue\n",
        "  puff[f'{key}'] = 2*_puff[f'{key}'].to(device=device , dtype = torch.float16)\n",
        "\n",
        "  #print(puff[f'{key}'].shape)\n",
        "\n",
        "save_file(puff, 'puff.safetensors')"
      ],
      "metadata": {
        "id": "U8fCk78GimS8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from safetensors.torch import load_file, save_file\n",
        "_tongue = load_file('/content/drive/MyDrive/Saved from Chrome/tongue-flux-v2.1.safetensors')\n",
        "tongue = {}\n",
        "# Scale = 32/16 = 2\n",
        "# Desired scale = 0.5  => multiply all matrices by 4 and set alpha to 8\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "for key in _tongue:\n",
        "  if f'{key}'.find('alpha')>-1:\n",
        "    tongue[f'{key}'] = torch.tensor(8).to(device=device , dtype = torch.float16)\n",
        "    continue\n",
        "  #-------#\n",
        "  tongue[f'{key}'] = 4*_tongue[f'{key}'].to(device=device , dtype = torch.float16)\n",
        "#-------#\n",
        "save_file(tongue, 'tongue.safetensors')"
      ],
      "metadata": {
        "id": "lFNa6vgrgdSA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "\n",
        "_oily = load_file('/content/drive/MyDrive/Saved from Chrome/OiledSkin_FluxDev.safetensors')\n",
        "\n",
        "star = load_file('/content/drive/MyDrive/Saved from Chrome/star_100_r32_16alpha.safetensors')\n",
        "#A = vs , B = u\n",
        "#lora_down = A , lora_up = B\n",
        "\n",
        "oily = {}\n",
        "for key in _oily:\n",
        "  if not f'{key}'.find('_A.')>-1:continue\n",
        "  A = f'{key}'\n",
        "  B = f'{key}'.replace('_A.','_B.')\n",
        "  down = f'{key}'.replace('_A.','_down.')\n",
        "  up = f'{key}'.replace('_A.','_up.')\n",
        "  #-----#\n",
        "  oily[f'{up}'] = _oily[f'{B}'].to(device = device , dtype=torch.float16)\n",
        "  oily[f'{down}'] = _oily[f'{A}'].to(device = device , dtype=torch.float16)\n",
        "  #------#\n",
        "  if not f'{key}'.find('to_k.')>-1:continue\n",
        "  k = f'{key}'\n",
        "  q = k.replace('to_k.','to_q.')\n",
        "  v = k.replace('to_k.','to_v.')\n",
        "\n",
        "print(\"---------OILY---------\")\n",
        "for key in oily:\n",
        "  print(key)\n",
        "  #if f'{key}'.find('alpha')>-1:print(key)\n",
        "\n",
        "print(\"---------STAR---------\")\n",
        "for key in star:\n",
        "  break\n",
        "  print(key)"
      ],
      "metadata": {
        "id": "1oxeJYHRqxQC",
        "collapsed": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "12e3a407-f9d1-403e-949b-31330be59577"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "---------OILY---------\n",
            "transformer.single_transformer_blocks.0.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.0.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.0.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.0.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.0.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.0.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.0.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.0.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.0.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.0.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.0.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.0.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.1.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.1.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.1.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.1.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.1.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.1.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.1.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.1.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.1.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.1.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.1.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.1.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.10.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.10.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.10.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.10.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.10.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.10.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.10.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.10.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.10.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.10.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.10.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.10.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.11.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.11.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.11.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.11.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.11.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.11.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.11.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.11.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.11.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.11.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.11.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.11.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.12.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.12.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.12.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.12.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.12.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.12.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.12.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.12.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.12.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.12.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.12.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.12.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.13.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.13.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.13.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.13.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.13.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.13.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.13.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.13.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.13.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.13.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.13.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.13.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.14.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.14.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.14.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.14.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.14.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.14.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.14.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.14.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.14.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.14.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.14.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.14.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.15.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.15.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.15.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.15.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.15.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.15.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.15.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.15.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.15.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.15.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.15.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.15.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.16.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.16.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.16.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.16.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.16.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.16.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.16.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.16.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.16.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.16.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.16.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.16.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.17.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.17.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.17.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.17.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.17.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.17.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.17.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.17.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.17.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.17.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.17.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.17.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.18.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.18.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.18.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.18.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.18.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.18.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.18.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.18.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.18.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.18.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.18.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.18.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.19.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.19.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.19.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.19.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.19.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.19.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.19.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.19.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.19.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.19.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.19.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.19.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.2.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.2.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.2.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.2.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.2.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.2.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.2.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.2.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.2.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.2.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.2.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.2.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.20.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.20.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.20.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.20.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.20.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.20.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.20.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.20.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.20.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.20.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.20.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.20.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.21.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.21.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.21.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.21.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.21.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.21.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.21.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.21.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.21.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.21.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.21.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.21.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.22.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.22.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.22.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.22.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.22.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.22.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.22.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.22.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.22.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.22.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.22.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.22.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.23.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.23.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.23.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.23.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.23.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.23.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.23.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.23.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.23.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.23.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.23.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.23.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.24.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.24.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.24.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.24.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.24.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.24.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.24.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.24.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.24.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.24.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.24.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.24.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.25.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.25.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.25.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.25.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.25.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.25.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.25.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.25.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.25.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.25.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.25.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.25.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.26.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.26.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.26.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.26.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.26.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.26.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.26.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.26.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.26.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.26.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.26.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.26.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.27.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.27.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.27.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.27.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.27.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.27.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.27.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.27.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.27.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.27.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.27.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.27.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.28.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.28.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.28.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.28.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.28.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.28.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.28.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.28.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.28.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.28.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.28.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.28.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.29.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.29.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.29.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.29.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.29.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.29.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.29.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.29.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.29.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.29.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.29.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.29.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.3.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.3.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.3.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.3.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.3.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.3.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.3.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.3.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.3.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.3.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.3.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.3.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.30.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.30.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.30.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.30.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.30.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.30.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.30.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.30.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.30.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.30.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.30.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.30.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.31.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.31.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.31.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.31.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.31.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.31.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.31.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.31.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.31.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.31.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.31.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.31.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.32.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.32.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.32.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.32.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.32.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.32.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.32.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.32.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.32.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.32.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.32.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.32.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.33.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.33.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.33.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.33.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.33.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.33.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.33.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.33.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.33.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.33.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.33.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.33.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.34.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.34.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.34.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.34.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.34.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.34.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.34.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.34.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.34.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.34.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.34.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.34.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.35.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.35.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.35.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.35.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.35.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.35.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.35.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.35.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.35.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.35.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.35.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.35.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.36.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.36.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.36.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.36.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.36.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.36.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.36.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.36.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.36.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.36.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.36.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.36.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.37.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.37.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.37.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.37.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.37.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.37.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.37.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.37.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.37.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.37.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.37.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.37.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.4.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.4.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.4.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.4.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.4.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.4.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.4.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.4.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.4.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.4.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.4.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.4.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.5.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.5.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.5.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.5.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.5.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.5.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.5.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.5.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.5.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.5.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.5.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.5.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.6.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.6.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.6.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.6.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.6.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.6.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.6.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.6.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.6.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.6.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.6.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.6.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.7.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.7.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.7.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.7.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.7.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.7.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.7.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.7.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.7.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.7.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.7.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.7.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.8.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.8.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.8.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.8.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.8.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.8.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.8.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.8.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.8.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.8.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.8.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.8.proj_out.lora_down.weight\n",
            "transformer.single_transformer_blocks.9.attn.to_k.lora_up.weight\n",
            "transformer.single_transformer_blocks.9.attn.to_k.lora_down.weight\n",
            "transformer.single_transformer_blocks.9.attn.to_q.lora_up.weight\n",
            "transformer.single_transformer_blocks.9.attn.to_q.lora_down.weight\n",
            "transformer.single_transformer_blocks.9.attn.to_v.lora_up.weight\n",
            "transformer.single_transformer_blocks.9.attn.to_v.lora_down.weight\n",
            "transformer.single_transformer_blocks.9.norm.linear.lora_up.weight\n",
            "transformer.single_transformer_blocks.9.norm.linear.lora_down.weight\n",
            "transformer.single_transformer_blocks.9.proj_mlp.lora_up.weight\n",
            "transformer.single_transformer_blocks.9.proj_mlp.lora_down.weight\n",
            "transformer.single_transformer_blocks.9.proj_out.lora_up.weight\n",
            "transformer.single_transformer_blocks.9.proj_out.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.0.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.0.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.0.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.0.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.0.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.0.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.0.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.0.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.0.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.0.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.0.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.0.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.0.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.0.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.1.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.1.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.1.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.1.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.1.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.1.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.1.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.1.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.1.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.1.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.1.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.1.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.1.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.1.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.10.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.10.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.10.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.10.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.10.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.10.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.10.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.10.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.10.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.10.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.10.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.10.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.10.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.10.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.11.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.11.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.11.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.11.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.11.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.11.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.11.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.11.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.11.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.11.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.11.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.11.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.11.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.11.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.12.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.12.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.12.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.12.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.12.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.12.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.12.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.12.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.12.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.12.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.12.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.12.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.12.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.12.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.13.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.13.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.13.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.13.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.13.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.13.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.13.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.13.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.13.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.13.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.13.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.13.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.13.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.13.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.14.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.14.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.14.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.14.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.14.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.14.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.14.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.14.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.14.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.14.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.14.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.14.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.14.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.14.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.15.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.15.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.15.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.15.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.15.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.15.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.15.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.15.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.15.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.15.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.15.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.15.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.15.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.15.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.16.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.16.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.16.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.16.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.16.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.16.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.16.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.16.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.16.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.16.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.16.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.16.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.16.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.16.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.17.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.17.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.17.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.17.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.17.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.17.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.17.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.17.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.17.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.17.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.17.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.17.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.17.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.17.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.18.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.18.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.18.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.18.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.18.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.18.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.18.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.18.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.18.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.18.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.18.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.18.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.18.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.18.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.2.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.2.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.2.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.2.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.2.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.2.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.2.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.2.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.2.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.2.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.2.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.2.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.2.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.2.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.3.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.3.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.3.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.3.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.3.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.3.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.3.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.3.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.3.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.3.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.3.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.3.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.3.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.3.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.4.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.4.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.4.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.4.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.4.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.4.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.4.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.4.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.4.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.4.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.4.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.4.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.4.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.4.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.5.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.5.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.5.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.5.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.5.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.5.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.5.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.5.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.5.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.5.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.5.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.5.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.5.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.5.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.6.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.6.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.6.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.6.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.6.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.6.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.6.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.6.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.6.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.6.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.6.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.6.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.6.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.6.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.7.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.7.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.7.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.7.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.7.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.7.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.7.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.7.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.7.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.7.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.7.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.7.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.7.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.7.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.8.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.8.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.8.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.8.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.8.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.8.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.8.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.8.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.8.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.8.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.8.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.8.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.8.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.8.norm1_context.linear.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.add_k_proj.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.add_k_proj.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.add_q_proj.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.add_q_proj.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.add_v_proj.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.add_v_proj.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.to_add_out.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.to_add_out.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.to_k.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.to_k.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.to_out.0.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.to_out.0.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.to_q.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.to_q.lora_down.weight\n",
            "transformer.transformer_blocks.9.attn.to_v.lora_up.weight\n",
            "transformer.transformer_blocks.9.attn.to_v.lora_down.weight\n",
            "transformer.transformer_blocks.9.ff.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.9.ff.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.9.ff.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.9.ff.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.9.ff_context.net.0.proj.lora_up.weight\n",
            "transformer.transformer_blocks.9.ff_context.net.0.proj.lora_down.weight\n",
            "transformer.transformer_blocks.9.ff_context.net.2.lora_up.weight\n",
            "transformer.transformer_blocks.9.ff_context.net.2.lora_down.weight\n",
            "transformer.transformer_blocks.9.norm1.linear.lora_up.weight\n",
            "transformer.transformer_blocks.9.norm1.linear.lora_down.weight\n",
            "transformer.transformer_blocks.9.norm1_context.linear.lora_up.weight\n",
            "transformer.transformer_blocks.9.norm1_context.linear.lora_down.weight\n",
            "---------STAR---------\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "down = 'lora_unet_double_blocks_0_img_attn_qkv.lora_down.weight'\n",
        "up = 'lora_unet_double_blocks_0_img_attn_qkv.lora_up.weight'\n",
        "tgt = star\n",
        "print(\"STAR\")\n",
        "print(tgt[f'{up}'].shape)\n",
        "#print(torch.matmul(tgt[f'{up}'],tgt[f'{down}']).shape)\n",
        "\n",
        "down = 'transformer.transformer_blocks.0.attn.to_k.lora_down.weight'\n",
        "up = 'transformer.transformer_blocks.0.attn.to_k.lora_up.weight'\n",
        "tgt = oily\n",
        "print(\"VS. OILY\")\n",
        "print(tgt[f'{up}'].shape)\n",
        "#print(torch.matmul(tgt[f'{up}'],tgt[f'{down}']).shape)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GoDfgENYaWD7",
        "outputId": "9336ae1a-6244-4e76-f291-82cda4482831"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "STAR\n",
            "torch.Size([9216, 32])\n",
            "VS. OILY\n",
            "torch.Size([3072, 32])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for key in oily:\n",
        "  print(oily[f'{key}'].shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xQhVLouEfmGE",
        "outputId": "662176b3-480d-48eb-f5db-97ec71b5e970"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([9216, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 15360])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([12288, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([3072, 32])\n",
            "torch.Size([32, 12288])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n",
            "torch.Size([18432, 32])\n",
            "torch.Size([32, 3072])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "import torch\n",
        "from safetensors.torch import load_file, save_file\n",
        "import torch.nn as nn\n",
        "from torch import linalg as LA\n",
        "import os\n",
        "import math\n",
        "import random\n",
        "import numpy as np\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "def _filter(tgt , pcnt, largest):\n",
        "    num_topk = math.floor(tgt.numel()*(pcnt/100))\n",
        "    y = tgt.flatten().to(device = device , dtype=torch.float32)\n",
        "    values,indices = torch.topk( y ,  num_topk , largest=largest)\n",
        "    _values,_indices = torch.topk( -y , num_topk , largest=largest)\n",
        "    y = y*0\n",
        "    y[indices] = 1\n",
        "    y[_indices] = 1\n",
        "    y = y.unflatten(0,tgt.shape).to(device = device , dtype=torch.float32)\n",
        "    return torch.mul(tgt,y)\n",
        "\n",
        "#----#\n",
        "\n",
        "# For pcnt = 30 , 'filter_and_save'  will keep all top 30 % values\n",
        "#, and the lowest (negative) 30% values for each layer delta_W in this lora\n",
        "# Then save the new filtered lora as a .safetensor file\n",
        "def filter_and_save(_lora , savefile_name, new_rank , new_alpha, thresh):\n",
        "    lora = {}\n",
        "    count = 0\n",
        "    for key in _lora:count = count + 1\n",
        "    NUM_ITEMS = count\n",
        "    count = 0\n",
        "    thresh = resolution*0.000001 # 1e-6\n",
        "    #-------#\n",
        "    for key in _lora:\n",
        "        if f'{key}'.find('alpha') > -1:\n",
        "            lora[f'{key}'] = torch.tensor(new_alpha).to(device = device , dtype = torch.float32)\n",
        "            count = count + 1\n",
        "            print(f'{count} / {NUM_ITEMS}')\n",
        "            continue\n",
        "        #------#\n",
        "        if not f'{key}'.find('lora_down') > -1: continue\n",
        "        up = f'{key}'.replace('lora_down' , 'lora_up')\n",
        "        down = f'{key}'\n",
        "        #-------#\n",
        "        delta_W = torch.matmul(_lora[up],_lora[down]).to(device = device , dtype=torch.float32)\n",
        "        #---#\n",
        "        N = delta_W.numel()\n",
        "        y = delta_W.flatten().to(device = device , dtype=torch.float32)\n",
        "        values,indices = torch.sort(y, descending = False) # smallest -> largest elements\n",
        "        y = torch.zeros(y.shape).to(device = device , dtype=torch.float32)\n",
        "        y[indices[values>thresh]] = 1\n",
        "        y[indices[values<-thresh]] = 1\n",
        "        y = y.unflatten(0,delta_W.shape).to(device = device , dtype=torch.float32)\n",
        "        delta_W = torch.mul(delta_W,y).to(device = device , dtype=torch.float32)\n",
        "        #------#\n",
        "        tmp={}\n",
        "        tmp['u'], tmp['s'], tmp['Vh'] = torch.svd(delta_W)\n",
        "        tmp['u'] = tmp['u'][:,: new_rank]\n",
        "        tmp['s'] = tmp['s'][: new_rank]\n",
        "        #-------#\n",
        "        tmp['u'] = torch.round(torch.matmul(tmp['u'], torch.diag(tmp['s'])),decimals=6)\n",
        "        tmp['Vh'] = torch.round(tmp['Vh'].t()[: new_rank,:],decimals=6)\n",
        "        #-------#\n",
        "        for key in tmp:tmp[f'{key}'] = tmp[f'{key}'].contiguous()\n",
        "        lora[up] = tmp['u'].to(device = device , dtype=torch.float32)\n",
        "        lora[down] = tmp['Vh'].to(device = device , dtype=torch.float32)\n",
        "        #-------#\n",
        "        count  = count  + 2\n",
        "        print(f'{count} / {NUM_ITEMS}')\n",
        "    #-------#\n",
        "    print(f'done!')\n",
        "    print(f'casting params to fp16....')\n",
        "    for key in _lora: lora[f'{key}'] = lora[f'{key}'].to(device = device , dtype=torch.float16)\n",
        "    #-------#\n",
        "    print(f'done!')\n",
        "    print(f'saving {savefile_name}...')\n",
        "    save_file(lora , f'{savefile_name}')\n",
        "#--------#\n",
        "\n",
        "\n",
        "new_rank = 32\n",
        "new_alpha = new_rank/2\n",
        "resolution = 100\n",
        "star = load_file('/kaggle/input/flux-loras/yeero.safetensors')\n",
        "for key in star:\n",
        "    star[f'{key}'] = star[f'{key}'].to(device = device , dtype = torch.float32)\n",
        "\n",
        "filter_and_save(star , f'yeero_{resolution}_r{new_rank}_{new_alpha}alpha.safetensors' , new_rank , new_alpha, resolution)\n",
        "\n",
        "#pcnt = 30\n",
        "#new_rank = 6\n",
        "#filter_and_save(yeero , f'yeero_topk{pcnt}_r{new_rank}.safetensors' , pcnt , new_rank)\n",
        "#filter_and_save(euro , f'euro_topk{pcnt}_r{new_rank}.safetensors' , pcnt , new_rank)\n",
        "#filter_and_save(star , f'star_topk{pcnt}_r{new_rank}.safetensors' , pcnt , new_rank)\n"
      ],
      "metadata": {
        "id": "f46xbSVkUlDl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cgi = load_file('/content/drive/MyDrive/Saved from Chrome/cgi_style.safetensors')"
      ],
      "metadata": {
        "id": "JuGDCX5272Bh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#cgi = load_file('/content/drive/MyDrive/Saved from Chrome/cgi_style.safetensors')\n",
        "doll = load_file('/content/drive/MyDrive/Saved from Chrome/dolls.safetensors')\n",
        "euro = load_file('/content/drive/MyDrive/Saved from Chrome/euro.safetensors')\n",
        "scale = load_file('/content/drive/MyDrive/Saved from Chrome/scale.safetensors')\n",
        "cgi = load_file('/content/drive/MyDrive/Saved from Chrome/cgi.safetensors')\n",
        "guns = load_file('/content/drive/MyDrive/Saved from Chrome/guns.safetensors')\n",
        "iris = load_file('/content/drive/MyDrive/Saved from Chrome/iris.safetensors')"
      ],
      "metadata": {
        "id": "FftDdBRG7su6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for key in doll:\n",
        "  doll[f'{key}'] = doll[f'{key}'].to(dtype=torch.float16)\n",
        "  euro[f'{key}'] = euro[f'{key}'].to(dtype=torch.float16)\n",
        "  scale[f'{key}'] = scale[f'{key}'].to(dtype=torch.float16)"
      ],
      "metadata": {
        "id": "RII9SEqh8KH2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "#define metric for similarity\n",
        "tgt_dim = torch.Size([64, 3072])\n",
        "cos0 = nn.CosineSimilarity(dim=1)\n",
        "cos = nn.CosineSimilarity(dim=1)\n",
        "\n",
        "\n",
        "def sim(tgt , ref ,key):\n",
        "  return torch.sum(torch.abs(cos(tgt, ref[f'{key}'])))  + torch.sum(torch.abs(cos0(tgt, ref[f'{key}'])))\n",
        "#-----#\n",
        "\n",
        "from torch import linalg as LA\n",
        "\n",
        "LA.matrix_norm\n",
        "def rand_search(A , B , key , iters):\n",
        "  tgt_norm = (LA.matrix_norm(A[f'{key}']) + LA.matrix_norm(B[f'{key}']))/2\n",
        "  tgt_avg = (A[f'{key}'] + B[f'{key}'])/2\n",
        "\n",
        "  max_sim = (sim(tgt_avg , A , key) + sim(tgt_avg , B , key))\n",
        "  cand = tgt_avg\n",
        "\n",
        "  for iter in range(iters):\n",
        "    rand = torch.ones(tgt_dim)*(-0.5)  + torch.rand(tgt_dim)\n",
        "    rand = rand  * (tgt_norm/LA.matrix_norm(rand))\n",
        "    #rand = (rand + tgt_avg)/2\n",
        "    #rand = rand  * (tgt_norm/LA.matrix_norm(rand))\n",
        "\n",
        "    tmp = sim(rand,A, key) + sim(rand , B, key)\n",
        "    if (tmp > max_sim):\n",
        "      max_sim = tmp\n",
        "      cand = rand\n",
        "      print('found!')\n",
        "      break\n",
        "  #------#\n",
        "  print('returning')\n",
        "  return cand  , max_sim\n",
        "#-----#"
      ],
      "metadata": {
        "id": "hJL6QEclHdHn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cand , max_sim = rand_search(cgi , iris , 'lora_unet_double_blocks_0_img_attn_proj.lora_down.weight' , 1000)\n",
        "print(sim(cand , iris , key))\n",
        "print(sim(cand , cgi , key))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ckyBSQi5Ll4F",
        "outputId": "341f7192-083d-4423-f61f-4f49d5756e79"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "returning\n",
            "tensor(91.1875, dtype=torch.float16)\n",
            "tensor(90.2500, dtype=torch.float16)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(torch.rand(1).to(dtype=torch.float16)*3).item()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XLwslN61hiIJ",
        "outputId": "9e3cbba6-3727-4772-f453-fecf8a408790"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.2138671875"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "torch.rand(1).to(dtype=torch.float16)*10"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AKwh0lZ1f8dJ",
        "outputId": "59186526-bd73-4efe-925a-3e7a9c738e53"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([6.8555], dtype=torch.float16)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "#define metric for similarity\n",
        "tgt_dim = torch.Size([64, 3072])\n",
        "cos0 = nn.CosineSimilarity(dim=0)\n",
        "\n",
        "\n",
        "\n",
        "cos = nn.CosineSimilarity(dim=1)\n",
        "\n",
        "\n",
        "def sim(tgt , ref ,key):\n",
        "  return torch.sum(torch.abs(cos(tgt, ref[f'{key}'])))  + torch.sum(torch.abs(cos0(tgt, ref[f'{key}'])))\n",
        "#-----#"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SNCvvkb2h3Zb",
        "outputId": "725fabd1-3fe2-4ac2-f24c-5f9309d45e4a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "7.715576171875"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from safetensors.torch import load_file , save_file\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "from torch import linalg as LA\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "#define metric for similarity\n",
        "cos0 = nn.CosineSimilarity(dim=0).to(device)\n",
        "final_score = 0\n",
        "highest_score = 0\n",
        "w_cgi = 1\n",
        "w_doll = 2\n",
        "w_euro = 2\n",
        "w_guns = 1\n",
        "w_iris = 2\n",
        "w_scale = 1\n",
        "\n",
        "w_noise = 0.00001 * (w_cgi + w_doll + w_euro  + w_guns + w_iris + w_scale)\n",
        "fixed_noise = {}\n",
        "\n",
        "#for key in doll:\n",
        "#  fixed_noise[f'{key}'] = torch.zeros(doll[f'{key}'].shape).to(device = device , dtype=torch.float16)\n",
        "#------#\n",
        "#w_offset = 0* (w1+w2+w3)\n",
        "#_w_offset = 0\n",
        "\n",
        "W =  (w_cgi + w_doll + w_euro  + w_guns + w_iris + w_scale + w_noise)*torch.ones(1).to(device = device,dtype=torch.float16)\n",
        "\n",
        "SCALE = 0.0001\n",
        "one = torch.ones(1).to(dtype=torch.float16).to(device)\n",
        "\n",
        "for attempt in range(1000):\n",
        "  print(f'attempt no : {attempt+1} ')\n",
        "  merge = load_file('/content/drive/MyDrive/Saved from Chrome/dolls.safetensors')\n",
        "  for key in doll:\n",
        "    tgt_dim = doll[f'{key}'].shape\n",
        "    if tgt_dim == torch.Size([]): continue\n",
        "    r_cgi = torch.rand(1).to(device = device,dtype=torch.float16).item()*w_cgi\n",
        "    r_doll = torch.rand(1).to(device = device,dtype=torch.float16).item()*w_doll\n",
        "    r_euro = torch.rand(1).to(device = device,dtype=torch.float16).item()*w_euro\n",
        "    r_guns = torch.rand(1).to(device = device,dtype=torch.float16).item()*w_guns\n",
        "    r_iris = torch.rand(1).to(device = device,dtype=torch.float16).item()*w_iris\n",
        "    r_scale = torch.rand(1).to(device = device,dtype=torch.float16).item()*w_scale\n",
        "    #------#\n",
        "    noise = torch.rand(tgt_dim).to(device = device,dtype=torch.float16)\n",
        "    noise_norm = LA.matrix_norm(noise).to(device = device,dtype=torch.float16).item()\n",
        "    noise = (w_noise/noise_norm)*noise.to(device = device,dtype=torch.float16)\n",
        "    #-----#\n",
        "    merge[f'{key}'] = r_cgi * cgi[f'{key}'] #overwrite\n",
        "    merge[f'{key}'] = merge[f'{key}'] + r_doll * doll[f'{key}']\n",
        "    merge[f'{key}'] = merge[f'{key}'] + r_euro * euro[f'{key}']\n",
        "    merge[f'{key}'] = merge[f'{key}'] + r_guns * guns[f'{key}']\n",
        "    merge[f'{key}'] = merge[f'{key}'] + r_iris * iris[f'{key}']\n",
        "    merge[f'{key}'] = merge[f'{key}'] + r_scale * scale[f'{key}']\n",
        "    merge[f'{key}'] = ((merge[f'{key}'] + noise)/W).to(device = device,dtype=torch.float16)\n",
        "  #-------#\n",
        "  score = torch.zeros(1).to(device = device, dtype=torch.float32)\n",
        "  #----#\n",
        "  NUM_ITERS = 10\n",
        "  for iter in range(NUM_ITERS):\n",
        "    for key in doll:\n",
        "      tgt_dim = doll[f'{key}'].shape\n",
        "      if tgt_dim == torch.Size([]): continue\n",
        "      vec = torch.rand(tgt_dim[0]).to(device = device,dtype=torch.float16)\n",
        "      cgi_out = torch.matmul(vec , cgi[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      doll_out = torch.matmul(vec , doll[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      euro_out = torch.matmul(vec , euro[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      guns_out = torch.matmul(vec , guns[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      iris_out = torch.matmul(vec , iris[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      scale_out = torch.matmul(vec , scale[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      merge_out = torch.matmul(vec , merge[f'{key}']).to(device = device,dtype=torch.float16)\n",
        "      #-------#\n",
        "      sim_value_cgi = torch.abs(cos0(cgi_out , merge_out)).to(device = device,dtype=torch.float32)*SCALE\n",
        "      sim_value_doll = torch.abs(cos0(doll_out , merge_out)).to(device = device,dtype=torch.float32)*SCALE\n",
        "      sim_value_euro = torch.abs(cos0(euro_out , merge_out)).to(device = device,dtype=torch.float32)*SCALE\n",
        "      sim_value_guns = torch.abs(cos0(guns_out , merge_out)).to(device = device,dtype=torch.float32)*SCALE\n",
        "      sim_value_iris = torch.abs(cos0(iris_out , merge_out)).to(device = device,dtype=torch.float32)*SCALE\n",
        "      sim_value_scale = torch.abs(cos0(scale_out , merge_out)).to(device = device,dtype=torch.float32)*SCALE\n",
        "      score = score  + SCALE*(sim_value_cgi + 2*sim_value_doll +  2*sim_value_euro + sim_value_guns + 2*sim_value_iris + sim_value_scale)/9  #<--- This score can be anything at all\n",
        "      #----#\n",
        "  #-----#\n",
        "\n",
        "  final_score = (1000/(NUM_ITERS * SCALE))*score.to(device = 'cpu' , dtype=torch.float32).item()\n",
        "  if (final_score>highest_score) :\n",
        "    highest_score = final_score\n",
        "    print('new highscore!')\n",
        "    print(f'score : {final_score} pts')\n",
        "    #------#\n",
        "    save_file(merge , 'all_merge_R4.safetensors')\n",
        "  #------#\n",
        "\n",
        "print(f'------------')\n",
        "print(f'Final score : {highest_score} pts')\n",
        "\n",
        "\n",
        "#all R1 23.190992578747682\n",
        "\n",
        "#all R2 23.333244826062582\n",
        "\n",
        "#all R3 23.34471355425194\n",
        "\n",
        "#all R4 23.402637452818453"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "9L_g5Zp9Du2E",
        "outputId": "a3aa2bde-061e-43f5-ca35-96bdc470be80"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "attempt no : 1 \n",
            "new highscore!\n",
            "score : 23.264414267032407 pts\n",
            "attempt no : 2 \n",
            "attempt no : 3 \n",
            "attempt no : 4 \n",
            "new highscore!\n",
            "score : 23.29399467271287 pts\n",
            "attempt no : 5 \n",
            "attempt no : 6 \n",
            "attempt no : 7 \n",
            "attempt no : 8 \n",
            "attempt no : 9 \n",
            "attempt no : 10 \n",
            "attempt no : 11 \n",
            "new highscore!\n",
            "score : 23.362628780887462 pts\n",
            "attempt no : 12 \n",
            "attempt no : 13 \n",
            "attempt no : 14 \n",
            "attempt no : 15 \n",
            "attempt no : 16 \n",
            "attempt no : 17 \n",
            "attempt no : 18 \n",
            "attempt no : 19 \n",
            "attempt no : 20 \n",
            "attempt no : 21 \n",
            "attempt no : 22 \n",
            "attempt no : 23 \n",
            "new highscore!\n",
            "score : 23.37011210329365 pts\n",
            "attempt no : 24 \n",
            "attempt no : 25 \n",
            "attempt no : 26 \n",
            "attempt no : 27 \n",
            "attempt no : 28 \n",
            "attempt no : 29 \n",
            "attempt no : 30 \n",
            "attempt no : 31 \n",
            "attempt no : 32 \n",
            "attempt no : 33 \n",
            "attempt no : 34 \n",
            "new highscore!\n",
            "score : 23.402637452818453 pts\n",
            "attempt no : 35 \n",
            "attempt no : 36 \n",
            "attempt no : 37 \n",
            "attempt no : 38 \n",
            "attempt no : 39 \n",
            "attempt no : 40 \n",
            "attempt no : 41 \n",
            "attempt no : 42 \n",
            "attempt no : 43 \n",
            "attempt no : 44 \n",
            "attempt no : 45 \n",
            "attempt no : 46 \n",
            "attempt no : 47 \n",
            "attempt no : 48 \n",
            "attempt no : 49 \n",
            "attempt no : 50 \n",
            "attempt no : 51 \n",
            "attempt no : 52 \n",
            "attempt no : 53 \n",
            "attempt no : 54 \n",
            "attempt no : 55 \n",
            "attempt no : 56 \n",
            "attempt no : 57 \n",
            "attempt no : 58 \n",
            "attempt no : 59 \n",
            "attempt no : 60 \n",
            "attempt no : 61 \n",
            "attempt no : 62 \n",
            "attempt no : 63 \n",
            "attempt no : 64 \n",
            "attempt no : 65 \n",
            "attempt no : 66 \n",
            "attempt no : 67 \n",
            "attempt no : 68 \n",
            "attempt no : 69 \n",
            "attempt no : 70 \n",
            "attempt no : 71 \n",
            "attempt no : 72 \n",
            "attempt no : 73 \n",
            "attempt no : 74 \n",
            "attempt no : 75 \n",
            "attempt no : 76 \n",
            "attempt no : 77 \n",
            "attempt no : 78 \n",
            "attempt no : 79 \n",
            "attempt no : 80 \n",
            "attempt no : 81 \n",
            "attempt no : 82 \n",
            "attempt no : 83 \n",
            "attempt no : 84 \n",
            "attempt no : 85 \n",
            "attempt no : 86 \n"
          ]
        },
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-33-037249458db5>\u001b[0m in \u001b[0;36m<cell line: 31>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     61\u001b[0m       \u001b[0mtgt_dim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdoll\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34mf'{key}'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m       \u001b[0;32mif\u001b[0m \u001b[0mtgt_dim\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mSize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mcontinue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m       \u001b[0mvec\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrand\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtgt_dim\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdevice\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat16\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m       \u001b[0mcgi_out\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvec\u001b[0m \u001b[0;34m,\u001b[0m \u001b[0mcgi\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34mf'{key}'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdevice\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat16\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m       \u001b[0mdoll_out\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvec\u001b[0m \u001b[0;34m,\u001b[0m \u001b[0mdoll\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34mf'{key}'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdevice\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat16\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "    for key in doll:\n",
        "      if final_score<38.5: break\n",
        "      _w_offset = w_offset\n",
        "      W =  (w1+w2+w3 + w_noise + _w_offset)*torch.ones(1).to(device = device,dtype=torch.float16)\n",
        "      tgt_dim = doll[f'{key}'].shape\n",
        "      if tgt_dim == torch.Size([]): continue\n",
        "      fixed_noise[f'{key}'] = fixed_noise[f'{key}'] + merge[f'{key}']\n",
        "      fixed_noise[f'{key}'] = (fixed_noise[f'{key}'] * (w_offset*torch.ones(1).to(device = device,dtype=torch.float16)/LA.matrix_norm(fixed_noise[f'{key}']))).to(device = device,dtype=torch.float16)"
      ],
      "metadata": {
        "id": "jWFHMJN6TqDq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "  vec = torch.rand(tgt_dim[0]).to(dtype=torch.float16)\n",
        "  same = torch.abs(cos0(vec ,vec))"
      ],
      "metadata": {
        "id": "k7Pq-kDbuNnQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "same"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ANBPfP7tuOoa",
        "outputId": "24300487-f874-4f1b-beb7-0f441ec7df4a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor(1., dtype=torch.float16)"
            ]
          },
          "metadata": {},
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "torch.ones(1).to(dtype=torch.float16)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zN92j8JJuQ6G",
        "outputId": "b810f4e6-a8f3-426a-ae52-ffbd44fb3f00"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([1.], dtype=torch.float16)"
            ]
          },
          "metadata": {},
          "execution_count": 66
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "py-JMJzhsAI4",
        "outputId": "207cd809-031c-48e3-af0a-98bc114d910e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "score : 45.8125 pts\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "%cd /content/\n",
        "save_file(merge , 'doll_euro_scale_R_merge.safetensors')"
      ],
      "metadata": {
        "id": "7qogsYsAr2QU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "9wzLwurSpwpL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test = torch.rand(tgt_dim)\n",
        "vec = torch.rand(tgt_dim[0])"
      ],
      "metadata": {
        "id": "DHdy4DptowYG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tgt_dim[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WeNJ0bquphtx",
        "outputId": "442bfb2e-c1ab-4549-a4ea-ca80d3cc9a7d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9216"
            ]
          },
          "metadata": {},
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(torch.matmul(vec,test)).shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xqZp3Xo8pQuW",
        "outputId": "68e5c25e-3391-45e7-9c73-45e0174ddbc1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([64])"
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tgt_dim = torch.Size([64, 3072])\n",
        "cosa = nn.CosineSimilarity(dim=0)\n",
        "cos_dim1 = nn.CosineSimilarity(dim=1)\n",
        "\n",
        "for key in cgi:\n",
        "  if not cgi[f'{key}'].shape == torch.Size([64, 3072]): continue\n",
        "  print(f'{key} : ')\n",
        "  print(torch.sum(torch.abs(cos_dim1(cgi[f'{key}'] , iris[f'{key}']))))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VFNw0Nck8V6Q",
        "outputId": "e48bab98-18f7-43bb-d1cf-89f3e00f7ccf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "lora_unet_double_blocks_0_img_attn_proj.lora_down.weight : \n",
            "tensor(1.6982, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.8145, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_img_mlp_0.lora_down.weight : \n",
            "tensor(1.6309, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_img_mod_lin.lora_down.weight : \n",
            "tensor(2.6211, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_txt_attn_proj.lora_down.weight : \n",
            "tensor(2.3203, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.3027, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.5898, dtype=torch.float16)\n",
            "lora_unet_double_blocks_0_txt_mod_lin.lora_down.weight : \n",
            "tensor(2.7402, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_img_attn_proj.lora_down.weight : \n",
            "tensor(2.0410, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.3350, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_img_mlp_0.lora_down.weight : \n",
            "tensor(2.0020, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_img_mod_lin.lora_down.weight : \n",
            "tensor(2.6562, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.1816, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.1348, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_txt_mlp_0.lora_down.weight : \n",
            "tensor(3.0156, dtype=torch.float16)\n",
            "lora_unet_double_blocks_10_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.4746, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_img_attn_proj.lora_down.weight : \n",
            "tensor(1.8359, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.5312, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_img_mlp_0.lora_down.weight : \n",
            "tensor(2.1465, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_img_mod_lin.lora_down.weight : \n",
            "tensor(3.9277, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.7246, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.8594, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_txt_mlp_0.lora_down.weight : \n",
            "tensor(3.6465, dtype=torch.float16)\n",
            "lora_unet_double_blocks_11_txt_mod_lin.lora_down.weight : \n",
            "tensor(2.6152, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_img_attn_proj.lora_down.weight : \n",
            "tensor(1.7295, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4795, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_img_mlp_0.lora_down.weight : \n",
            "tensor(3.4043, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_img_mod_lin.lora_down.weight : \n",
            "tensor(2.0137, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.4375, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.8994, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.1152, dtype=torch.float16)\n",
            "lora_unet_double_blocks_12_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.2744, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_img_attn_proj.lora_down.weight : \n",
            "tensor(3.0742, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4980, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_img_mlp_0.lora_down.weight : \n",
            "tensor(1.9609, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_img_mod_lin.lora_down.weight : \n",
            "tensor(2.6133, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.6904, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.1680, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.8574, dtype=torch.float16)\n",
            "lora_unet_double_blocks_13_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.9053, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_img_attn_proj.lora_down.weight : \n",
            "tensor(1.8135, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4033, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_img_mlp_0.lora_down.weight : \n",
            "tensor(1.5547, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_img_mod_lin.lora_down.weight : \n",
            "tensor(2.8906, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.1328, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.3701, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_txt_mlp_0.lora_down.weight : \n",
            "tensor(3.3145, dtype=torch.float16)\n",
            "lora_unet_double_blocks_14_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.2031, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_img_attn_proj.lora_down.weight : \n",
            "tensor(1.5137, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.3809, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_img_mlp_0.lora_down.weight : \n",
            "tensor(1.4834, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_img_mod_lin.lora_down.weight : \n",
            "tensor(1.6465, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.7256, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.8672, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.1953, dtype=torch.float16)\n",
            "lora_unet_double_blocks_15_txt_mod_lin.lora_down.weight : \n",
            "tensor(0.9858, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_img_attn_proj.lora_down.weight : \n",
            "tensor(1.5703, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4648, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_img_mlp_0.lora_down.weight : \n",
            "tensor(1.5537, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_img_mod_lin.lora_down.weight : \n",
            "tensor(2.6133, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_txt_attn_proj.lora_down.weight : \n",
            "tensor(2.2559, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.9365, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.7891, dtype=torch.float16)\n",
            "lora_unet_double_blocks_16_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.3174, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_img_attn_proj.lora_down.weight : \n",
            "tensor(2.4609, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.6240, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_img_mlp_0.lora_down.weight : \n",
            "tensor(3.1406, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_img_mod_lin.lora_down.weight : \n",
            "tensor(2.1055, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.7480, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.6436, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_txt_mlp_0.lora_down.weight : \n",
            "tensor(1.9688, dtype=torch.float16)\n",
            "lora_unet_double_blocks_17_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.8184, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_img_attn_proj.lora_down.weight : \n",
            "tensor(2.3887, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.6738, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_img_mlp_0.lora_down.weight : \n",
            "tensor(3.7500, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_img_mod_lin.lora_down.weight : \n",
            "tensor(2.7285, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_txt_attn_proj.lora_down.weight : \n",
            "tensor(2.0410, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.0586, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.0801, dtype=torch.float16)\n",
            "lora_unet_double_blocks_18_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.5684, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_img_attn_proj.lora_down.weight : \n",
            "tensor(4.9844, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.8613, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_img_mlp_0.lora_down.weight : \n",
            "tensor(2.2266, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_img_mod_lin.lora_down.weight : \n",
            "tensor(2.8164, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.7500, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.3105, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_txt_mlp_0.lora_down.weight : \n",
            "tensor(1.9639, dtype=torch.float16)\n",
            "lora_unet_double_blocks_1_txt_mod_lin.lora_down.weight : \n",
            "tensor(2.6504, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_img_attn_proj.lora_down.weight : \n",
            "tensor(4.6367, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.7988, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_img_mlp_0.lora_down.weight : \n",
            "tensor(4.6758, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_img_mod_lin.lora_down.weight : \n",
            "tensor(3.1445, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_txt_attn_proj.lora_down.weight : \n",
            "tensor(2.2285, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.4990, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.3984, dtype=torch.float16)\n",
            "lora_unet_double_blocks_2_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.4443, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_img_attn_proj.lora_down.weight : \n",
            "tensor(3.6855, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.9971, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_img_mlp_0.lora_down.weight : \n",
            "tensor(3.3301, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_img_mod_lin.lora_down.weight : \n",
            "tensor(2.3379, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_txt_attn_proj.lora_down.weight : \n",
            "tensor(2.0117, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.1621, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.7676, dtype=torch.float16)\n",
            "lora_unet_double_blocks_3_txt_mod_lin.lora_down.weight : \n",
            "tensor(3.1895, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_img_attn_proj.lora_down.weight : \n",
            "tensor(2.3848, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.7783, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_img_mlp_0.lora_down.weight : \n",
            "tensor(2.0234, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_img_mod_lin.lora_down.weight : \n",
            "tensor(1.9082, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.7588, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.9902, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_txt_mlp_0.lora_down.weight : \n",
            "tensor(1.5859, dtype=torch.float16)\n",
            "lora_unet_double_blocks_4_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.5654, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_img_attn_proj.lora_down.weight : \n",
            "tensor(2.7402, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.6221, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_img_mlp_0.lora_down.weight : \n",
            "tensor(1.6318, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_img_mod_lin.lora_down.weight : \n",
            "tensor(1.7988, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.1699, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_txt_attn_qkv.lora_down.weight : \n",
            "tensor(3.5566, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_txt_mlp_0.lora_down.weight : \n",
            "tensor(1.5791, dtype=torch.float16)\n",
            "lora_unet_double_blocks_5_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.5547, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_img_attn_proj.lora_down.weight : \n",
            "tensor(1.7988, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4531, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_img_mlp_0.lora_down.weight : \n",
            "tensor(2.4141, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_img_mod_lin.lora_down.weight : \n",
            "tensor(6.0234, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.0068, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.0098, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_txt_mlp_0.lora_down.weight : \n",
            "tensor(4.0312, dtype=torch.float16)\n",
            "lora_unet_double_blocks_6_txt_mod_lin.lora_down.weight : \n",
            "tensor(2.6309, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_img_attn_proj.lora_down.weight : \n",
            "tensor(1.4814, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4854, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_img_mlp_0.lora_down.weight : \n",
            "tensor(1.3877, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_img_mod_lin.lora_down.weight : \n",
            "tensor(2.3125, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_txt_attn_proj.lora_down.weight : \n",
            "tensor(3.4746, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.0430, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_txt_mlp_0.lora_down.weight : \n",
            "tensor(1.8018, dtype=torch.float16)\n",
            "lora_unet_double_blocks_7_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.1709, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_img_attn_proj.lora_down.weight : \n",
            "tensor(1.8857, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.8848, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_img_mlp_0.lora_down.weight : \n",
            "tensor(1.7627, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_img_mod_lin.lora_down.weight : \n",
            "tensor(4.2852, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_txt_attn_proj.lora_down.weight : \n",
            "tensor(1.3887, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_txt_attn_qkv.lora_down.weight : \n",
            "tensor(1.6289, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_txt_mlp_0.lora_down.weight : \n",
            "tensor(2.2188, dtype=torch.float16)\n",
            "lora_unet_double_blocks_8_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.5742, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_img_attn_proj.lora_down.weight : \n",
            "tensor(2.3125, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_img_attn_qkv.lora_down.weight : \n",
            "tensor(1.4854, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_img_mlp_0.lora_down.weight : \n",
            "tensor(1.9492, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_img_mod_lin.lora_down.weight : \n",
            "tensor(2.2949, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_txt_attn_proj.lora_down.weight : \n",
            "tensor(2.0781, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_txt_attn_qkv.lora_down.weight : \n",
            "tensor(2.6172, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_txt_mlp_0.lora_down.weight : \n",
            "tensor(3.1367, dtype=torch.float16)\n",
            "lora_unet_double_blocks_9_txt_mod_lin.lora_down.weight : \n",
            "tensor(1.2451, dtype=torch.float16)\n",
            "lora_unet_single_blocks_0_linear1.lora_down.weight : \n",
            "tensor(2.4375, dtype=torch.float16)\n",
            "lora_unet_single_blocks_0_modulation_lin.lora_down.weight : \n",
            "tensor(3.5684, dtype=torch.float16)\n",
            "lora_unet_single_blocks_10_linear1.lora_down.weight : \n",
            "tensor(2.6328, dtype=torch.float16)\n",
            "lora_unet_single_blocks_10_modulation_lin.lora_down.weight : \n",
            "tensor(2.9961, dtype=torch.float16)\n",
            "lora_unet_single_blocks_11_linear1.lora_down.weight : \n",
            "tensor(3.1211, dtype=torch.float16)\n",
            "lora_unet_single_blocks_11_modulation_lin.lora_down.weight : \n",
            "tensor(3.3672, dtype=torch.float16)\n",
            "lora_unet_single_blocks_12_linear1.lora_down.weight : \n",
            "tensor(3.0293, dtype=torch.float16)\n",
            "lora_unet_single_blocks_12_modulation_lin.lora_down.weight : \n",
            "tensor(3.6602, dtype=torch.float16)\n",
            "lora_unet_single_blocks_13_linear1.lora_down.weight : \n",
            "tensor(2.5918, dtype=torch.float16)\n",
            "lora_unet_single_blocks_13_modulation_lin.lora_down.weight : \n",
            "tensor(4.6367, dtype=torch.float16)\n",
            "lora_unet_single_blocks_14_linear1.lora_down.weight : \n",
            "tensor(2.0215, dtype=torch.float16)\n",
            "lora_unet_single_blocks_14_modulation_lin.lora_down.weight : \n",
            "tensor(3.5371, dtype=torch.float16)\n",
            "lora_unet_single_blocks_15_linear1.lora_down.weight : \n",
            "tensor(2.1719, dtype=torch.float16)\n",
            "lora_unet_single_blocks_15_modulation_lin.lora_down.weight : \n",
            "tensor(4.2812, dtype=torch.float16)\n",
            "lora_unet_single_blocks_16_linear1.lora_down.weight : \n",
            "tensor(2.1992, dtype=torch.float16)\n",
            "lora_unet_single_blocks_16_modulation_lin.lora_down.weight : \n",
            "tensor(4.1094, dtype=torch.float16)\n",
            "lora_unet_single_blocks_17_linear1.lora_down.weight : \n",
            "tensor(2.0703, dtype=torch.float16)\n",
            "lora_unet_single_blocks_17_modulation_lin.lora_down.weight : \n",
            "tensor(2.9277, dtype=torch.float16)\n",
            "lora_unet_single_blocks_18_linear1.lora_down.weight : \n",
            "tensor(2.0371, dtype=torch.float16)\n",
            "lora_unet_single_blocks_18_modulation_lin.lora_down.weight : \n",
            "tensor(2.6133, dtype=torch.float16)\n",
            "lora_unet_single_blocks_19_linear1.lora_down.weight : \n",
            "tensor(2.0723, dtype=torch.float16)\n",
            "lora_unet_single_blocks_19_modulation_lin.lora_down.weight : \n",
            "tensor(3.4980, dtype=torch.float16)\n",
            "lora_unet_single_blocks_1_linear1.lora_down.weight : \n",
            "tensor(1.7432, dtype=torch.float16)\n",
            "lora_unet_single_blocks_1_modulation_lin.lora_down.weight : \n",
            "tensor(2.3848, dtype=torch.float16)\n",
            "lora_unet_single_blocks_20_linear1.lora_down.weight : \n",
            "tensor(2.0137, dtype=torch.float16)\n",
            "lora_unet_single_blocks_20_modulation_lin.lora_down.weight : \n",
            "tensor(2.8203, dtype=torch.float16)\n",
            "lora_unet_single_blocks_21_linear1.lora_down.weight : \n",
            "tensor(1.8955, dtype=torch.float16)\n",
            "lora_unet_single_blocks_21_modulation_lin.lora_down.weight : \n",
            "tensor(2.7305, dtype=torch.float16)\n",
            "lora_unet_single_blocks_22_linear1.lora_down.weight : \n",
            "tensor(2.7559, dtype=torch.float16)\n",
            "lora_unet_single_blocks_22_modulation_lin.lora_down.weight : \n",
            "tensor(4.6133, dtype=torch.float16)\n",
            "lora_unet_single_blocks_23_linear1.lora_down.weight : \n",
            "tensor(2.5508, dtype=torch.float16)\n",
            "lora_unet_single_blocks_23_modulation_lin.lora_down.weight : \n",
            "tensor(4.4180, dtype=torch.float16)\n",
            "lora_unet_single_blocks_24_linear1.lora_down.weight : \n",
            "tensor(1.9219, dtype=torch.float16)\n",
            "lora_unet_single_blocks_24_modulation_lin.lora_down.weight : \n",
            "tensor(2.9453, dtype=torch.float16)\n",
            "lora_unet_single_blocks_25_linear1.lora_down.weight : \n",
            "tensor(2.7539, dtype=torch.float16)\n",
            "lora_unet_single_blocks_25_modulation_lin.lora_down.weight : \n",
            "tensor(4.5938, dtype=torch.float16)\n",
            "lora_unet_single_blocks_26_linear1.lora_down.weight : \n",
            "tensor(3.3750, dtype=torch.float16)\n",
            "lora_unet_single_blocks_26_modulation_lin.lora_down.weight : \n",
            "tensor(4.7344, dtype=torch.float16)\n",
            "lora_unet_single_blocks_27_linear1.lora_down.weight : \n",
            "tensor(2.3809, dtype=torch.float16)\n",
            "lora_unet_single_blocks_27_modulation_lin.lora_down.weight : \n",
            "tensor(4.9883, dtype=torch.float16)\n",
            "lora_unet_single_blocks_28_linear1.lora_down.weight : \n",
            "tensor(3.0859, dtype=torch.float16)\n",
            "lora_unet_single_blocks_28_modulation_lin.lora_down.weight : \n",
            "tensor(5.7539, dtype=torch.float16)\n",
            "lora_unet_single_blocks_29_linear1.lora_down.weight : \n",
            "tensor(2.3242, dtype=torch.float16)\n",
            "lora_unet_single_blocks_29_modulation_lin.lora_down.weight : \n",
            "tensor(3.9160, dtype=torch.float16)\n",
            "lora_unet_single_blocks_2_linear1.lora_down.weight : \n",
            "tensor(2.1406, dtype=torch.float16)\n",
            "lora_unet_single_blocks_2_modulation_lin.lora_down.weight : \n",
            "tensor(2.1621, dtype=torch.float16)\n",
            "lora_unet_single_blocks_30_linear1.lora_down.weight : \n",
            "tensor(2.1211, dtype=torch.float16)\n",
            "lora_unet_single_blocks_30_modulation_lin.lora_down.weight : \n",
            "tensor(4.8516, dtype=torch.float16)\n",
            "lora_unet_single_blocks_31_linear1.lora_down.weight : \n",
            "tensor(2.2773, dtype=torch.float16)\n",
            "lora_unet_single_blocks_31_modulation_lin.lora_down.weight : \n",
            "tensor(4.1367, dtype=torch.float16)\n",
            "lora_unet_single_blocks_32_linear1.lora_down.weight : \n",
            "tensor(2.5273, dtype=torch.float16)\n",
            "lora_unet_single_blocks_32_modulation_lin.lora_down.weight : \n",
            "tensor(5.0508, dtype=torch.float16)\n",
            "lora_unet_single_blocks_33_linear1.lora_down.weight : \n",
            "tensor(2.7051, dtype=torch.float16)\n",
            "lora_unet_single_blocks_33_modulation_lin.lora_down.weight : \n",
            "tensor(5.2930, dtype=torch.float16)\n",
            "lora_unet_single_blocks_34_linear1.lora_down.weight : \n",
            "tensor(2.6738, dtype=torch.float16)\n",
            "lora_unet_single_blocks_34_modulation_lin.lora_down.weight : \n",
            "tensor(4.7852, dtype=torch.float16)\n",
            "lora_unet_single_blocks_35_linear1.lora_down.weight : \n",
            "tensor(2.5117, dtype=torch.float16)\n",
            "lora_unet_single_blocks_35_modulation_lin.lora_down.weight : \n",
            "tensor(6.7734, dtype=torch.float16)\n",
            "lora_unet_single_blocks_36_linear1.lora_down.weight : \n",
            "tensor(1.8418, dtype=torch.float16)\n",
            "lora_unet_single_blocks_36_modulation_lin.lora_down.weight : \n",
            "tensor(6.5859, dtype=torch.float16)\n",
            "lora_unet_single_blocks_37_linear1.lora_down.weight : \n",
            "tensor(2.4473, dtype=torch.float16)\n",
            "lora_unet_single_blocks_37_modulation_lin.lora_down.weight : \n",
            "tensor(2.5742, dtype=torch.float16)\n",
            "lora_unet_single_blocks_3_linear1.lora_down.weight : \n",
            "tensor(2.5566, dtype=torch.float16)\n",
            "lora_unet_single_blocks_3_modulation_lin.lora_down.weight : \n",
            "tensor(4.7148, dtype=torch.float16)\n",
            "lora_unet_single_blocks_4_linear1.lora_down.weight : \n",
            "tensor(2.2832, dtype=torch.float16)\n",
            "lora_unet_single_blocks_4_modulation_lin.lora_down.weight : \n",
            "tensor(2.0566, dtype=torch.float16)\n",
            "lora_unet_single_blocks_5_linear1.lora_down.weight : \n",
            "tensor(2.2109, dtype=torch.float16)\n",
            "lora_unet_single_blocks_5_modulation_lin.lora_down.weight : \n",
            "tensor(2.7793, dtype=torch.float16)\n",
            "lora_unet_single_blocks_6_linear1.lora_down.weight : \n",
            "tensor(3.0176, dtype=torch.float16)\n",
            "lora_unet_single_blocks_6_modulation_lin.lora_down.weight : \n",
            "tensor(2.9180, dtype=torch.float16)\n",
            "lora_unet_single_blocks_7_linear1.lora_down.weight : \n",
            "tensor(2.2461, dtype=torch.float16)\n",
            "lora_unet_single_blocks_7_modulation_lin.lora_down.weight : \n",
            "tensor(2.1074, dtype=torch.float16)\n",
            "lora_unet_single_blocks_8_linear1.lora_down.weight : \n",
            "tensor(3.0391, dtype=torch.float16)\n",
            "lora_unet_single_blocks_8_modulation_lin.lora_down.weight : \n",
            "tensor(2.0039, dtype=torch.float16)\n",
            "lora_unet_single_blocks_9_linear1.lora_down.weight : \n",
            "tensor(3.8789, dtype=torch.float16)\n",
            "lora_unet_single_blocks_9_modulation_lin.lora_down.weight : \n",
            "tensor(4.0547, dtype=torch.float16)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<---- Upload your civiai trained .safetensor file to Google Colab before running the next cell\n",
        "\n"
      ],
      "metadata": {
        "id": "oDAUwfFzqzgj"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WQZ3BZn1p-pw"
      },
      "outputs": [],
      "source": [
        "civiai_lora = '' # @param {type:'string' ,placeholder:'ex. civitai_trained_e19.safetensors'}\n",
        "tensor_art_filename = '' # @param {type:'string' ,placeholder:'ex. e19.safetensors'}\n",
        "%cd /content/\n",
        "tgt = load_file(f'{civiai_lora}')\n",
        "for key in tgt:\n",
        "  tgt[f'{key}'] = tgt[f'{key}'].to(dtype=torch.float16)\n",
        "%cd /content/\n",
        "save_file(tgt , f'{tensor_art_filename}')"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Download the new .safetensor file to your device.\n",
        "\n",
        "Downloading from CoLab Notebook will seemingly do nothing for ~5min. Then the file will download , so be patient.\n",
        "\n",
        "For faster/more consistent downloads , download your .safetensor file from your Google Drive"
      ],
      "metadata": {
        "id": "blnBW-U4rAS7"
      }
    }
  ]
}