学習したpthファイルから事前学習モデルsafetensorsを作るやつ

safetensors.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "\n",
    "jvs_root = \"Data/jvs/models\"\n",
    "step = 50000\n",
    "\n",
    "g_model = torch.load(f\"{jvs_root}/G_{step}.pth\", map_location=\"cpu\")\n",
    "d_model = torch.load(f\"{jvs_root}/D_{step}.pth\", map_location=\"cpu\")\n",
    "dur_model = torch.load(f\"{jvs_root}/DUR_{step}.pth\", map_location=\"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "emb_g.weight\n"
     ]
    }
   ],
   "source": [
    "g_dict = {}\n",
    "for key in g_model[\"model\"].keys():\n",
    "    if key.startswith(\"emb_g\"):\n",
    "        print(key)\n",
    "    else:\n",
    "        g_dict[key] = g_model[\"model\"][key]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_dict = {}\n",
    "for key in d_model[\"model\"].keys():\n",
    "    d_dict[key] = d_model[\"model\"][key]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "dur_dict = {}\n",
    "for key in dur_model[\"model\"].keys():\n",
    "    dur_dict[key] = dur_model[\"model\"][key]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from safetensors.torch import save_file\n",
    "\n",
    "\n",
    "save_file(g_dict, f\"G_0.safetensors\")\n",
    "save_file (d_dict, f\"D_0.safetensors\")\n",
    "save_file (dur_dict, f\"DUR_0.safetensors\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}