chop-model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b0374368-eb08-4ceb-8c8e-c8f409aa7213",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "from huggingface_hub import PyTorchModelHubMixin\n",
    "from transformers import AutoConfig, AutoModel\n",
    "from transformers.models.mixtral.modeling_mixtral import MixtralDecoderLayer\n",
    "from transformers.models.gpt_neo.modeling_gpt_neo import GPTNeoBlock, GPTNeoModel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "94c38ffa-afbf-4056-a0f0-9015a3e18fc9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "resolved_archive_file='/Users/joel/.cache/huggingface/hub/models--EleutherAI--gpt-neo-125M/snapshots/21def0189f5705e2521767faed922f1f15e7d7db/model.safetensors'\n",
      "f=<builtins.safe_open object at 0x1677db7b0>\n"
     ]
    }
   ],
   "source": [
    "hf_model = AutoModel.from_pretrained(\"EleutherAI/gpt-neo-125M\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "defef29c-671a-43af-861a-c64eda75ec8c",
   "metadata": {},
   "outputs": [],
   "source": [
    "class MyModel(GPTNeoModel):\n",
    "    def __init__(self, config):\n",
    "        super().__init__(config)\n",
    "        self.h = nn.ModuleList([GPTNeoBlock(config, 0)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "f313e455-503d-4c36-b736-d3154b136f50",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6dd45c59fcbd4f0ba1fc003f558b4e6e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "model.safetensors:   0%|          | 0.00/189M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "CommitInfo(commit_url='https://huggingface.co/joelb/my-awesome-model/commit/f1090d0a00028679dedc0a4efea9d11f434d5556', commit_message='Upload model', commit_description='', oid='f1090d0a00028679dedc0a4efea9d11f434d5556', pr_url=None, pr_revision=None, pr_num=None)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "config = AutoConfig.from_pretrained(\"EleutherAI/gpt-neo-125M\")\n",
    "config.num_layers = 1\n",
    "config.attention_layers = config.attention_layers[:1]\n",
    "config.attention_types = [[['global'], 1]]\n",
    "model = MyModel(config)\n",
    "model.push_to_hub(\"joelb/my-awesome-model\", config=config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "dc5a44cd-a4ac-4c70-a4a9-a04b57e5b189",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d3c7daec4a0743e89ed2cce9c132f409",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "config.json:   0%|          | 0.00/930 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "33fd4573a95e450a94f5586d5aa3f27f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "model.safetensors:   0%|          | 0.00/189M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "resolved_archive_file='/Users/joel/.cache/huggingface/hub/models--joelb--my-awesome-model/snapshots/f1090d0a00028679dedc0a4efea9d11f434d5556/model.safetensors'\n",
      "f=<builtins.safe_open object at 0x169b09bf0>\n"
     ]
    }
   ],
   "source": [
    "model2 = AutoModel.from_pretrained(\"joelb/my-awesome-model\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}