Upload ft-test.ipynb

ft-test.ipynb

@@ -0,0 +1,931 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "be94e6d6-4096-4d1a-aa58-5afd89f33bff",
+   "metadata": {},
+   "source": [
+    "# Fine-tuning Sandbox\n",
+    "\n",
+    "Code authored by: Shawhin Talebi <br>\n",
+    "Blog link: https://medium.com/towards-data-science/fine-tuning-large-language-models-llms-23473d763b91"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "4ef8ea85-d04d-4217-99a3-21c446bf2ffa",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "WARNING:tensorflow:From C:\\Users\\Administrator\\AppData\\Roaming\\Python\\Python39\\site-packages\\keras\\src\\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "from datasets import load_dataset, DatasetDict, Dataset\n",
+    "\n",
+    "from transformers import (\n",
+    "    AutoTokenizer,\n",
+    "    AutoConfig, \n",
+    "    AutoModelForSequenceClassification,\n",
+    "    DataCollatorWithPadding,\n",
+    "    TrainingArguments,\n",
+    "    Trainer)\n",
+    "# PEFT的全称是Parameter-Efficient Fine-Tuning，是transform开发的一个参数高效微调的库\n",
+    "from peft import PeftModel, PeftConfig, get_peft_model, LoraConfig\n",
+    "import evaluate\n",
+    "import torch\n",
+    "import numpy as np"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "aa6a4484-07d8-49dd-81ef-672105f53ebe",
+   "metadata": {},
+   "source": [
+    "### dataset"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "fa9722d3-0609-4aea-9585-9aa2cfc1fc9a",
+   "metadata": {
+    "jupyter": {
+     "source_hidden": true
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# # how dataset was generated\n",
+    "\n",
+    "# # load imdb data\n",
+    "# imdb_dataset = load_dataset(\"imdb\")\n",
+    "\n",
+    "# # define subsample size\n",
+    "# N = 1000 \n",
+    "# # generate indexes for random subsample\n",
+    "# rand_idx = np.random.randint(24999, size=N) \n",
+    "\n",
+    "# # extract train and test data\n",
+    "# x_train = imdb_dataset['train'][rand_idx]['text']\n",
+    "# y_train = imdb_dataset['train'][rand_idx]['label']\n",
+    "\n",
+    "# x_test = imdb_dataset['test'][rand_idx]['text']\n",
+    "# y_test = imdb_dataset['test'][rand_idx]['label']\n",
+    "\n",
+    "# # create new dataset\n",
+    "# dataset = DatasetDict({'train':Dataset.from_dict({'label':y_train,'text':x_train}),\n",
+    "#                              'validation':Dataset.from_dict({'label':y_test,'text':x_test})})"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "de226234-c521-4577-802c-0e7079ef4364",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "DatasetDict({\n",
+       "    train: Dataset({\n",
+       "        features: ['label', 'text'],\n",
+       "        num_rows: 1000\n",
+       "    })\n",
+       "    validation: Dataset({\n",
+       "        features: ['label', 'text'],\n",
+       "        num_rows: 1000\n",
+       "    })\n",
+       "    test: Dataset({\n",
+       "        features: ['label', 'text'],\n",
+       "        num_rows: 1000\n",
+       "    })\n",
+       "})"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# 加载数据集 训练 验证 测试\n",
+    "dataset = load_dataset('shawhin/imdb-truncated')\n",
+    "dataset"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "d5625faa-5fea-4334-bd38-b77de983d8a8",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.5"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# 得出训练集标签的平均值\n",
+    "np.array(dataset['train']['label']).sum()/len(dataset['train']['label'])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "3644c68d-9adf-48a4-90a2-8fd89555a302",
+   "metadata": {},
+   "source": [
+    "### model"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "a60dd1fe-8144-4678-b018-20891e49237a",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Some weights of DistilBertForSequenceClassification were not initialized from the model checkpoint at distilbert-base-uncased and are newly initialized: ['classifier.bias', 'classifier.weight', 'pre_classifier.bias', 'pre_classifier.weight']\n",
+      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
+     ]
+    }
+   ],
+   "source": [
+    "model_checkpoint = 'distilbert-base-uncased'\n",
+    "\n",
+    "# 类别的映射关系\n",
+    "id2label = {0: \"Negative\", 1: \"Positive\"}\n",
+    "label2id = {\"Negative\":0, \"Positive\":1}\n",
+    "\n",
+    "# 加载预训练的权重 num_labels指明是二分类任务 model_checkpoint 预训练模型的名称\n",
+    "model = AutoModelForSequenceClassification.from_pretrained(\n",
+    "    model_checkpoint, num_labels=2, id2label=id2label, label2id=label2id)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "853002f8-d39c-4bc4-8d07-e44a47de3b47",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# display architecture\n",
+    "model = model.cuda()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "4bc98609-873d-455c-bac4-155632cda484",
+   "metadata": {},
+   "source": [
+    "### 预处理数据"
+   ]
+  },
+  {
+   "cell_type": "raw",
+   "id": "93e728f3-9e12-400d-950e-f7f2e29fe19e",
+   "metadata": {},
+   "source": [
+    "add_prefix_space参数告诉 tokenizer 在处理单词和标点符号之间添加一个前缀空格 前缀空格（表示为 Ġ）\n",
+    "# 原始句子\n",
+    "sentence = \"Hello, world!\"\n",
+    "['ĠHello', ',', 'Ġworld', '!']"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "7fe08707-657f-4e66-aa72-84899c54bf8d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# 创建分词器\n",
+    "tokenizer = AutoTokenizer.from_pretrained(model_checkpoint, add_prefix_space=True)\n",
+    "\n",
+    "# 判断是否有填充标记 通过 resize_token_embeddings 方法调整模型的 token embeddings，以包含新添加的 pad token。\n",
+    "if tokenizer.pad_token is None:\n",
+    "    tokenizer.add_special_tokens({'pad_token': '[PAD]'})\n",
+    "    model.resize_token_embeddings(len(tokenizer))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "20f4adb9-ce8f-4f54-9b94-300c9daae1b8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# 创建分词器函数\n",
+    "def tokenize_function(examples):\n",
+    "    # 提取文本\n",
+    "    text = examples[\"text\"]\n",
+    "\n",
+    "    # 设置 tokenizer 的截断位置为左侧。这意味着如果文本超过指定的 max_length，则在左侧截断。这是为了确保重要的文本内容被保留下来。\n",
+    "    tokenizer.truncation_side = \"left\"\n",
+    "    tokenized_inputs = tokenizer(\n",
+    "        text,\n",
+    "        # 返回numpy 类型\n",
+    "        return_tensors=\"np\",\n",
+    "        # 是否进行文本截断\n",
+    "        truncation=True,\n",
+    "        max_length=512\n",
+    "    )\n",
+    "\n",
+    "    return tokenized_inputs"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "b7600bcd-7e93-4fb4-bd8d-ffc76bed1ac2",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "c029f605df0e4e3c9484aa97af255052",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Map:   0%|          | 0/1000 [00:00<?, ? examples/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "DatasetDict({\n",
+       "    train: Dataset({\n",
+       "        features: ['label', 'text', 'input_ids', 'attention_mask'],\n",
+       "        num_rows: 1000\n",
+       "    })\n",
+       "    validation: Dataset({\n",
+       "        features: ['label', 'text', 'input_ids', 'attention_mask'],\n",
+       "        num_rows: 1000\n",
+       "    })\n",
+       "    test: Dataset({\n",
+       "        features: ['label', 'text', 'input_ids', 'attention_mask'],\n",
+       "        num_rows: 1000\n",
+       "    })\n",
+       "})"
+      ]
+     },
+     "execution_count": 9,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# tokenize training and validation datasets\n",
+    "tokenized_dataset = dataset.map(tokenize_function, batched=True)\n",
+    "tokenized_dataset"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "3f8e85f9-1804-4f49-a783-4da59580ea1e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# 创建数据收集器\n",
+    "data_collator = DataCollatorWithPadding(tokenizer=tokenizer)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "3cd9a120-580d-470c-a981-7c7e22604865",
+   "metadata": {},
+   "source": [
+    "### evaluation"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "2a894819-2e9c-4a53-9790-32130c182bca",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Using the latest cached version of the module from C:\\Users\\Administrator\\.cache\\huggingface\\modules\\evaluate_modules\\metrics\\evaluate-metric--accuracy\\f887c0aab52c2d38e1f8a215681126379eca617f96c447638f751434e8e65b14 (last modified on Fri Mar 15 09:54:33 2024) since it couldn't be found locally at evaluate-metric--accuracy, or remotely on the Hugging Face Hub.\n"
+     ]
+    }
+   ],
+   "source": [
+    "# import accuracy evaluation metric\n",
+    "accuracy = evaluate.load(\"accuracy\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "c07b9be2-a3f6-4b38-b9e8-6a2bc8aa945a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# define an evaluation function to pass into trainer later\n",
+    "def compute_metrics(p):\n",
+    "    predictions, labels = p\n",
+    "    predictions = np.argmax(predictions, axis=1)\n",
+    "    # 计算预测结果和真实标签 返回准确率\n",
+    "    return {\"accuracy\": accuracy.compute(predictions=predictions, references=labels)}"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "47500035-a555-46e0-83dc-440586d96b7e",
+   "metadata": {},
+   "source": [
+    "### Apply untrained model to text"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "8f3761c1-a297-45c8-882e-d74856259810",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Untrained model predictions:\n",
+      "----------------------------\n",
+      "I'm sorry. - Negative\n",
+      "You areedespicable person - Negative\n",
+      "Better than the first one. - Negative\n",
+      "This is not worth watching even once. - Negative\n",
+      "This one is a pass. - Negative\n"
+     ]
+    }
+   ],
+   "source": [
+    "# define list of examples\n",
+    "text_list = [\"I'm sorry.\", \"You areedespicable person\", \"Better than the first one.\", \"This is not worth watching even once.\", \"This one is a pass.\"]\n",
+    "\n",
+    "print(\"Untrained model predictions:\")\n",
+    "print(\"----------------------------\")\n",
+    "for text in text_list:\n",
+    "    # 将文本转化为可以理解的编码 并返回pytorch张量\n",
+    "    inputs = tokenizer.encode(text, return_tensors=\"pt\")\n",
+    "    # 计算对数\n",
+    "    logits = model(inputs.cuda()).logits\n",
+    "    # convert logits to label\n",
+    "    predictions = torch.argmax(logits)\n",
+    "\n",
+    "    print(text + \" - \" + id2label[predictions.tolist()])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "ff356f78-c9fd-4f2b-8f5b-097cf29c1c08",
+   "metadata": {},
+   "source": [
+    "### Train model"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "id": "e4dde538-cd7f-4ab5-a96d-c30f3003822e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "peft_config = LoraConfig(task_type=\"SEQ_CLS\", # 序列分类任务\n",
+    "                        r = 4, # 递归深度\n",
+    "                        lora_alpha = 32, # alpha 值表示 LORA 模块的影响更大。\n",
+    "                        lora_dropout = 0.01,\n",
+    "                        target_modules = ['q_lin'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "id": "f1391303-1e16-4d5c-b2b4-799997eff9f8",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "LoraConfig(peft_type=<PeftType.LORA: 'LORA'>, auto_mapping=None, base_model_name_or_path=None, revision=None, task_type='SEQ_CLS', inference_mode=False, r=4, target_modules={'q_lin'}, lora_alpha=32, lora_dropout=0.01, fan_in_fan_out=False, bias='none', use_rslora=False, modules_to_save=None, init_lora_weights=True, layers_to_transform=None, layers_pattern=None, rank_pattern={}, alpha_pattern={}, megatron_config=None, megatron_core='megatron.core', loftq_config={}, use_dora=False)"
+      ]
+     },
+     "execution_count": 15,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "peft_config"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "id": "3e0d9408-9fc4-4bd3-8d35-4d8217fe01e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "trainable params: 628,994 || all params: 67,584,004 || trainable%: 0.9306847223789819\n"
+     ]
+    }
+   ],
+   "source": [
+    "# 对模型进行配置\n",
+    "model = get_peft_model(model, peft_config)\n",
+    "model.print_trainable_parameters()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "id": "5db78059-e5ae-4807-89db-b58ef6abedd1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# hyperparameters\n",
+    "lr = 1e-3\n",
+    "batch_size = 4\n",
+    "num_epochs = 10"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "9244ed55-65a4-4c66-8388-55efd87bceb8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# define training arguments\n",
+    "training_args = TrainingArguments(\n",
+    "    output_dir= model_checkpoint + \"-lora-text-classification\",\n",
+    "    learning_rate=lr,\n",
+    "    per_device_train_batch_size=batch_size,\n",
+    "    per_device_eval_batch_size=batch_size,\n",
+    "    num_train_epochs=num_epochs,\n",
+    "    weight_decay=0.01, #  权重衰减，一种正则化技术，用于控制模型参数的大小。\n",
+    "    evaluation_strategy=\"epoch\",\n",
+    "    save_strategy=\"epoch\",\n",
+    "    load_best_model_at_end=True, # 是否在训练结束加载最佳模型\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6e21aa23-a366-4606-b13b-ad22e4639272",
+   "metadata": {},
+   "source": [
+    "### "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "id": "fc8bc705-5dd7-4305-a797-399b2b0fa2c7",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "D:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\accelerate\\accelerator.py:432: FutureWarning: Passing the following arguments to `Accelerator` is deprecated and will be removed in version 1.0 of Accelerate: dict_keys(['dispatch_batches', 'split_batches', 'even_batches', 'use_seedable_sampler']). Please pass an `accelerate.DataLoaderConfiguration` instead: \n",
+      "dataloader_config = DataLoaderConfiguration(dispatch_batches=None, split_batches=False, even_batches=True, use_seedable_sampler=True)\n",
+      "  warnings.warn(\n",
+      "\u001b[34m\u001b[1mwandb\u001b[0m: Currently logged in as: \u001b[33m1321416285\u001b[0m (\u001b[33mxuuuu\u001b[0m). Use \u001b[1m`wandb login --relogin`\u001b[0m to force relogin\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "wandb version 0.16.4 is available!  To upgrade, please run:\n",
+       " $ pip install wandb --upgrade"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Tracking run with wandb version 0.15.12"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Run data is saved locally in <code>D:\\software\\Anaconda\\jupyterfile\\AIfinetuning\\wandb\\run-20240315_211852-07azjtzv</code>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Syncing run <strong><a href='https://wandb.ai/xuuuu/huggingface/runs/07azjtzv' target=\"_blank\">fast-firefly-2</a></strong> to <a href='https://wandb.ai/xuuuu/huggingface' target=\"_blank\">Weights & Biases</a> (<a href='https://wandb.me/run' target=\"_blank\">docs</a>)<br/>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       " View project at <a href='https://wandb.ai/xuuuu/huggingface' target=\"_blank\">https://wandb.ai/xuuuu/huggingface</a>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       " View run at <a href='https://wandb.ai/xuuuu/huggingface/runs/07azjtzv' target=\"_blank\">https://wandb.ai/xuuuu/huggingface/runs/07azjtzv</a>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "\n",
+       "    <div>\n",
+       "      \n",
+       "      <progress value='2500' max='2500' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
+       "      [2500/2500 02:44, Epoch 10/10]\n",
+       "    </div>\n",
+       "    <table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       " <tr style=\"text-align: left;\">\n",
+       "      <th>Epoch</th>\n",
+       "      <th>Training Loss</th>\n",
+       "      <th>Validation Loss</th>\n",
+       "      <th>Accuracy</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <td>1</td>\n",
+       "      <td>No log</td>\n",
+       "      <td>0.438809</td>\n",
+       "      <td>{'accuracy': 0.855}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>2</td>\n",
+       "      <td>0.427600</td>\n",
+       "      <td>0.648398</td>\n",
+       "      <td>{'accuracy': 0.859}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>3</td>\n",
+       "      <td>0.427600</td>\n",
+       "      <td>0.637398</td>\n",
+       "      <td>{'accuracy': 0.877}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>4</td>\n",
+       "      <td>0.218100</td>\n",
+       "      <td>0.689158</td>\n",
+       "      <td>{'accuracy': 0.889}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>5</td>\n",
+       "      <td>0.218100</td>\n",
+       "      <td>0.774748</td>\n",
+       "      <td>{'accuracy': 0.897}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>6</td>\n",
+       "      <td>0.073100</td>\n",
+       "      <td>0.846054</td>\n",
+       "      <td>{'accuracy': 0.887}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>7</td>\n",
+       "      <td>0.073100</td>\n",
+       "      <td>0.946100</td>\n",
+       "      <td>{'accuracy': 0.894}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>8</td>\n",
+       "      <td>0.015500</td>\n",
+       "      <td>0.941895</td>\n",
+       "      <td>{'accuracy': 0.901}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>9</td>\n",
+       "      <td>0.015500</td>\n",
+       "      <td>0.994161</td>\n",
+       "      <td>{'accuracy': 0.898}</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <td>10</td>\n",
+       "      <td>0.006700</td>\n",
+       "      <td>0.999837</td>\n",
+       "      <td>{'accuracy': 0.897}</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table><p>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Trainer is attempting to log a value of \"{'accuracy': 0.855}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-250 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.859}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-500 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.877}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-750 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.889}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-1000 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.897}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-1250 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.887}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-1500 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.894}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-1750 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.901}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-2000 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.898}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-2250 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n",
+      "Trainer is attempting to log a value of \"{'accuracy': 0.897}\" of type <class 'dict'> for key \"eval/accuracy\" as a scalar. This invocation of Tensorboard's writer.add_scalar() is incorrect so we dropped this attribute.\n",
+      "Checkpoint destination directory distilbert-base-uncased-lora-text-classification\\checkpoint-2500 already exists and is non-empty. Saving will proceed but saved results may be invalid.\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "TrainOutput(global_step=2500, training_loss=0.14819346437454223, metrics={'train_runtime': 174.6372, 'train_samples_per_second': 57.262, 'train_steps_per_second': 14.315, 'total_flos': 1112883852759936.0, 'train_loss': 0.14819346437454223, 'epoch': 10.0})"
+      ]
+     },
+     "execution_count": 19,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# creater trainer object\n",
+    "trainer = Trainer(\n",
+    "    model=model,\n",
+    "    args=training_args,\n",
+    "    train_dataset=tokenized_dataset[\"train\"],\n",
+    "    eval_dataset=tokenized_dataset[\"validation\"],\n",
+    "    tokenizer=tokenizer,\n",
+    "    data_collator=data_collator, # this will dynamically pad examples in each batch to be equal length\n",
+    "    compute_metrics=compute_metrics, \n",
+    ")\n",
+    "\n",
+    "# train model\n",
+    "trainer.train()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6f5664d1-9bd2-4ce1-bc24-cab5adf80f49",
+   "metadata": {},
+   "source": [
+    "### Generate prediction"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "id": "e5dc029e-1c16-491d-a3f1-715f9e0adf52",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Trained model predictions:\n",
+      "--------------------------\n",
+      "I'm sorry. - Negative\n",
+      "You areedespicable person - Positive\n",
+      "Better than the first one. - Positive\n",
+      "This is not worth watching even once. - Negative\n",
+      "This one is a pass. - Negative\n"
+     ]
+    }
+   ],
+   "source": [
+    "model.to('cuda') # moving to mps for Mac (can alternatively do 'cpu')\n",
+    "\n",
+    "print(\"Trained model predictions:\")\n",
+    "print(\"--------------------------\")\n",
+    "for text in text_list:\n",
+    "    inputs = tokenizer.encode(text, return_tensors=\"pt\").to(\"cuda\") # moving to mps for Mac (can alternatively do 'cpu')\n",
+    "\n",
+    "    logits = model(inputs).logits\n",
+    "    predictions = torch.max(logits,1).indices\n",
+    "\n",
+    "    print(text + \" - \" + id2label[predictions.tolist()[0]])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c084bd9e-f7b1-4979-b753-73335ee0cede",
+   "metadata": {},
+   "source": [
+    "### Optional: push model to hub"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "id": "159eb49a-dd0d-4c9e-b9ab-27e06585fd84",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "a0e23e8a27634de78c21c18041cd010f",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "VBox(children=(HTML(value='<center> <img\\nsrc=https://huggingface.co/front/assets/huggingface_logo-noborder.sv…"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "# option 1: notebook login\n",
+    "from huggingface_hub import notebook_login\n",
+    "notebook_login() # ensure token gives write access\n",
+    "\n",
+    "# # option 2: key login\n",
+    "# from huggingface_hub import login\n",
+    "# write_key = 'hf_' # paste token here\n",
+    "# login(write_key)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "id": "09496307-e253-47e3-a46f-3f28a84c89a7",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "hf_name = 'shawhin' # your hf username or org name\n",
+    "model_id = hf_name + \"/\" + model_checkpoint + \"-lora-text-classification\" # you can name the model whatever you want"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "id": "c56ea581-0ea3-45f3-af21-362e9093ee37",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "HfHubHTTPError",
+     "evalue": "403 Client Error: Forbidden for url: https://huggingface.co/shawhin/distilbert-base-uncased-lora-text-classification.git/info/lfs/objects/batch (Request ID: Root=1-65f44b6d-3a7059390bd0f46b3618a6e6;b93e4a6f-c6a2-4179-8d62-ec4b3235048e)\n\nAuthorization error.",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mHTTPError\u001b[0m                                 Traceback (most recent call last)",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\utils\\_errors.py:304\u001b[0m, in \u001b[0;36mhf_raise_for_status\u001b[1;34m(response, endpoint_name)\u001b[0m\n\u001b[0;32m    303\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m--> 304\u001b[0m     \u001b[43mresponse\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mraise_for_status\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    305\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m HTTPError \u001b[38;5;28;01mas\u001b[39;00m e:\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\requests\\models.py:943\u001b[0m, in \u001b[0;36mResponse.raise_for_status\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    942\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m http_error_msg:\n\u001b[1;32m--> 943\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m HTTPError(http_error_msg, response\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m)\n",
+      "\u001b[1;31mHTTPError\u001b[0m: 403 Client Error: Forbidden for url: https://huggingface.co/shawhin/distilbert-base-uncased-lora-text-classification.git/info/lfs/objects/batch",
+      "\nThe above exception was the direct cause of the following exception:\n",
+      "\u001b[1;31mHfHubHTTPError\u001b[0m                            Traceback (most recent call last)",
+      "Cell \u001b[1;32mIn[23], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m \u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mpush_to_hub\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_id\u001b[49m\u001b[43m)\u001b[49m\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\transformers\\utils\\hub.py:894\u001b[0m, in \u001b[0;36mPushToHubMixin.push_to_hub\u001b[1;34m(self, repo_id, use_temp_dir, commit_message, private, token, max_shard_size, create_pr, safe_serialization, revision, commit_description, tags, **deprecated_kwargs)\u001b[0m\n\u001b[0;32m    891\u001b[0m \u001b[38;5;66;03m# Update model card if needed:\u001b[39;00m\n\u001b[0;32m    892\u001b[0m model_card\u001b[38;5;241m.\u001b[39msave(os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39mjoin(work_dir, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mREADME.md\u001b[39m\u001b[38;5;124m\"\u001b[39m))\n\u001b[1;32m--> 894\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_upload_modified_files\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    895\u001b[0m \u001b[43m    \u001b[49m\u001b[43mwork_dir\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    896\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrepo_id\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    897\u001b[0m \u001b[43m    \u001b[49m\u001b[43mfiles_timestamps\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    898\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcommit_message\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcommit_message\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    899\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtoken\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtoken\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    900\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcreate_pr\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcreate_pr\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    901\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrevision\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrevision\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    902\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcommit_description\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcommit_description\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    903\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\transformers\\utils\\hub.py:758\u001b[0m, in \u001b[0;36mPushToHubMixin._upload_modified_files\u001b[1;34m(self, working_dir, repo_id, files_timestamps, commit_message, token, create_pr, revision, commit_description)\u001b[0m\n\u001b[0;32m    755\u001b[0m     create_branch(repo_id\u001b[38;5;241m=\u001b[39mrepo_id, branch\u001b[38;5;241m=\u001b[39mrevision, token\u001b[38;5;241m=\u001b[39mtoken, exist_ok\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m)\n\u001b[0;32m    757\u001b[0m logger\u001b[38;5;241m.\u001b[39minfo(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mUploading the following files to \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mrepo_id\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m,\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;241m.\u001b[39mjoin(modified_files)\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m--> 758\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mcreate_commit\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    759\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrepo_id\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_id\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    760\u001b[0m \u001b[43m    \u001b[49m\u001b[43moperations\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43moperations\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    761\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcommit_message\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcommit_message\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    762\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcommit_description\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcommit_description\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    763\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtoken\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtoken\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    764\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcreate_pr\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcreate_pr\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    765\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrevision\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrevision\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    766\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\utils\\_validators.py:118\u001b[0m, in \u001b[0;36mvalidate_hf_hub_args.<locals>._inner_fn\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    115\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m check_use_auth_token:\n\u001b[0;32m    116\u001b[0m     kwargs \u001b[38;5;241m=\u001b[39m smoothly_deprecate_use_auth_token(fn_name\u001b[38;5;241m=\u001b[39mfn\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m, has_token\u001b[38;5;241m=\u001b[39mhas_token, kwargs\u001b[38;5;241m=\u001b[39mkwargs)\n\u001b[1;32m--> 118\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m fn(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\hf_api.py:1227\u001b[0m, in \u001b[0;36mfuture_compatible.<locals>._inner\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1224\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrun_as_future(fn, \u001b[38;5;28mself\u001b[39m, \u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[0;32m   1226\u001b[0m \u001b[38;5;66;03m# Otherwise, call the function normally\u001b[39;00m\n\u001b[1;32m-> 1227\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m fn(\u001b[38;5;28mself\u001b[39m, \u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\hf_api.py:3762\u001b[0m, in \u001b[0;36mHfApi.create_commit\u001b[1;34m(self, repo_id, operations, commit_message, commit_description, token, repo_type, revision, create_pr, num_threads, parent_commit, run_as_future)\u001b[0m\n\u001b[0;32m   3759\u001b[0m \u001b[38;5;66;03m# If updating twice the same file or update then delete a file in a single commit\u001b[39;00m\n\u001b[0;32m   3760\u001b[0m _warn_on_overwriting_operations(operations)\n\u001b[1;32m-> 3762\u001b[0m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mpreupload_lfs_files\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   3763\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrepo_id\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_id\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   3764\u001b[0m \u001b[43m    \u001b[49m\u001b[43madditions\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43madditions\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   3765\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtoken\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtoken\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   3766\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrepo_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_type\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   3767\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrevision\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43munquoted_revision\u001b[49m\u001b[43m,\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# first-class methods take unquoted revision\u001b[39;49;00m\n\u001b[0;32m   3768\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcreate_pr\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcreate_pr\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   3769\u001b[0m \u001b[43m    \u001b[49m\u001b[43mnum_threads\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnum_threads\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   3770\u001b[0m \u001b[43m    \u001b[49m\u001b[43mfree_memory\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mFalse\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m  \u001b[49m\u001b[38;5;66;43;03m# do not remove `CommitOperationAdd.path_or_fileobj` on LFS files for \"normal\" users\u001b[39;49;00m\n\u001b[0;32m   3771\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   3772\u001b[0m files_to_copy \u001b[38;5;241m=\u001b[39m _fetch_files_to_copy(\n\u001b[0;32m   3773\u001b[0m     copies\u001b[38;5;241m=\u001b[39mcopies,\n\u001b[0;32m   3774\u001b[0m     repo_type\u001b[38;5;241m=\u001b[39mrepo_type,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   3778\u001b[0m     endpoint\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mendpoint,\n\u001b[0;32m   3779\u001b[0m )\n\u001b[0;32m   3780\u001b[0m commit_payload \u001b[38;5;241m=\u001b[39m _prepare_commit_payload(\n\u001b[0;32m   3781\u001b[0m     operations\u001b[38;5;241m=\u001b[39moperations,\n\u001b[0;32m   3782\u001b[0m     files_to_copy\u001b[38;5;241m=\u001b[39mfiles_to_copy,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   3785\u001b[0m     parent_commit\u001b[38;5;241m=\u001b[39mparent_commit,\n\u001b[0;32m   3786\u001b[0m )\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\hf_api.py:4262\u001b[0m, in \u001b[0;36mHfApi.preupload_lfs_files\u001b[1;34m(self, repo_id, additions, token, repo_type, revision, create_pr, num_threads, free_memory, gitignore_content)\u001b[0m\n\u001b[0;32m   4256\u001b[0m     logger\u001b[38;5;241m.\u001b[39minfo(\n\u001b[0;32m   4257\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSkipped upload for \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mlen\u001b[39m(new_lfs_additions)\u001b[38;5;250m \u001b[39m\u001b[38;5;241m-\u001b[39m\u001b[38;5;250m \u001b[39m\u001b[38;5;28mlen\u001b[39m(new_lfs_additions_to_upload)\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m LFS file(s) \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m   4258\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m(ignored by gitignore file).\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m   4259\u001b[0m     )\n\u001b[0;32m   4261\u001b[0m \u001b[38;5;66;03m# Upload new LFS files\u001b[39;00m\n\u001b[1;32m-> 4262\u001b[0m \u001b[43m_upload_lfs_files\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   4263\u001b[0m \u001b[43m    \u001b[49m\u001b[43madditions\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnew_lfs_additions_to_upload\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   4264\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrepo_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_type\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   4265\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrepo_id\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_id\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   4266\u001b[0m \u001b[43m    \u001b[49m\u001b[43mtoken\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtoken\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01mor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtoken\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   4267\u001b[0m \u001b[43m    \u001b[49m\u001b[43mendpoint\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mendpoint\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   4268\u001b[0m \u001b[43m    \u001b[49m\u001b[43mnum_threads\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnum_threads\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m   4269\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;66;43;03m# If `create_pr`, we don't want to check user permission on the revision as users with read permission\u001b[39;49;00m\n\u001b[0;32m   4270\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;66;43;03m# should still be able to create PRs even if they don't have write permission on the target branch of the\u001b[39;49;00m\n\u001b[0;32m   4271\u001b[0m \u001b[43m    \u001b[49m\u001b[38;5;66;43;03m# PR (i.e. `revision`).\u001b[39;49;00m\n\u001b[0;32m   4272\u001b[0m \u001b[43m    \u001b[49m\u001b[43mrevision\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrevision\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mif\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;129;43;01mnot\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mcreate_pr\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01melse\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mNone\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[0;32m   4273\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   4274\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m addition \u001b[38;5;129;01min\u001b[39;00m new_lfs_additions_to_upload:\n\u001b[0;32m   4275\u001b[0m     addition\u001b[38;5;241m.\u001b[39m_is_uploaded \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mTrue\u001b[39;00m\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\utils\\_validators.py:118\u001b[0m, in \u001b[0;36mvalidate_hf_hub_args.<locals>._inner_fn\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    115\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m check_use_auth_token:\n\u001b[0;32m    116\u001b[0m     kwargs \u001b[38;5;241m=\u001b[39m smoothly_deprecate_use_auth_token(fn_name\u001b[38;5;241m=\u001b[39mfn\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m, has_token\u001b[38;5;241m=\u001b[39mhas_token, kwargs\u001b[38;5;241m=\u001b[39mkwargs)\n\u001b[1;32m--> 118\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m fn(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\_commit_api.py:360\u001b[0m, in \u001b[0;36m_upload_lfs_files\u001b[1;34m(additions, repo_type, repo_id, token, endpoint, num_threads, revision)\u001b[0m\n\u001b[0;32m    358\u001b[0m batch_actions: List[Dict] \u001b[38;5;241m=\u001b[39m []\n\u001b[0;32m    359\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m chunk \u001b[38;5;129;01min\u001b[39;00m chunk_iterable(additions, chunk_size\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m256\u001b[39m):\n\u001b[1;32m--> 360\u001b[0m     batch_actions_chunk, batch_errors_chunk \u001b[38;5;241m=\u001b[39m \u001b[43mpost_lfs_batch_info\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    361\u001b[0m \u001b[43m        \u001b[49m\u001b[43mupload_infos\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m[\u001b[49m\u001b[43mop\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mupload_info\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mop\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mchunk\u001b[49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    362\u001b[0m \u001b[43m        \u001b[49m\u001b[43mtoken\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mtoken\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    363\u001b[0m \u001b[43m        \u001b[49m\u001b[43mrepo_id\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_id\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    364\u001b[0m \u001b[43m        \u001b[49m\u001b[43mrepo_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrepo_type\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    365\u001b[0m \u001b[43m        \u001b[49m\u001b[43mrevision\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mrevision\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    366\u001b[0m \u001b[43m        \u001b[49m\u001b[43mendpoint\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mendpoint\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    367\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    369\u001b[0m     \u001b[38;5;66;03m# If at least 1 error, we do not retrieve information for other chunks\u001b[39;00m\n\u001b[0;32m    370\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m batch_errors_chunk:\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\utils\\_validators.py:118\u001b[0m, in \u001b[0;36mvalidate_hf_hub_args.<locals>._inner_fn\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    115\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m check_use_auth_token:\n\u001b[0;32m    116\u001b[0m     kwargs \u001b[38;5;241m=\u001b[39m smoothly_deprecate_use_auth_token(fn_name\u001b[38;5;241m=\u001b[39mfn\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m, has_token\u001b[38;5;241m=\u001b[39mhas_token, kwargs\u001b[38;5;241m=\u001b[39mkwargs)\n\u001b[1;32m--> 118\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m fn(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\lfs.py:159\u001b[0m, in \u001b[0;36mpost_lfs_batch_info\u001b[1;34m(upload_infos, token, repo_type, repo_id, revision, endpoint)\u001b[0m\n\u001b[0;32m    157\u001b[0m headers \u001b[38;5;241m=\u001b[39m {\u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mLFS_HEADERS, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mbuild_hf_headers(token\u001b[38;5;241m=\u001b[39mtoken \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28;01mTrue\u001b[39;00m)}  \u001b[38;5;66;03m# Token must be provided or retrieved\u001b[39;00m\n\u001b[0;32m    158\u001b[0m resp \u001b[38;5;241m=\u001b[39m get_session()\u001b[38;5;241m.\u001b[39mpost(batch_url, headers\u001b[38;5;241m=\u001b[39mheaders, json\u001b[38;5;241m=\u001b[39mpayload)\n\u001b[1;32m--> 159\u001b[0m \u001b[43mhf_raise_for_status\u001b[49m\u001b[43m(\u001b[49m\u001b[43mresp\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    160\u001b[0m batch_info \u001b[38;5;241m=\u001b[39m resp\u001b[38;5;241m.\u001b[39mjson()\n\u001b[0;32m    162\u001b[0m objects \u001b[38;5;241m=\u001b[39m batch_info\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mobjects\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m)\n",
+      "File \u001b[1;32mD:\\software\\Anaconda\\envs\\Work1\\lib\\site-packages\\huggingface_hub\\utils\\_errors.py:362\u001b[0m, in \u001b[0;36mhf_raise_for_status\u001b[1;34m(response, endpoint_name)\u001b[0m\n\u001b[0;32m    358\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m BadRequestError(message, response\u001b[38;5;241m=\u001b[39mresponse) \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01me\u001b[39;00m\n\u001b[0;32m    360\u001b[0m \u001b[38;5;66;03m# Convert `HTTPError` into a `HfHubHTTPError` to display request information\u001b[39;00m\n\u001b[0;32m    361\u001b[0m \u001b[38;5;66;03m# as well (request id and/or server error message)\u001b[39;00m\n\u001b[1;32m--> 362\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m HfHubHTTPError(\u001b[38;5;28mstr\u001b[39m(e), response\u001b[38;5;241m=\u001b[39mresponse) \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01me\u001b[39;00m\n",
+      "\u001b[1;31mHfHubHTTPError\u001b[0m: 403 Client Error: Forbidden for url: https://huggingface.co/shawhin/distilbert-base-uncased-lora-text-classification.git/info/lfs/objects/batch (Request ID: Root=1-65f44b6d-3a7059390bd0f46b3618a6e6;b93e4a6f-c6a2-4179-8d62-ec4b3235048e)\n\nAuthorization error."
+     ]
+    }
+   ],
+   "source": [
+    "model.push_to_hub(model_id) # save model"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "f487331a-8552-4fb2-867f-985b8fe1d1ab",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "trainer.push_to_hub(model_id) # save trainer"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "00e7feaa-b70e-4b1d-a118-23c616d14639",
+   "metadata": {},
+   "source": [
+    "### Optional: load peft model"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "19cffa01-25a4-4c86-a7fa-a84353b8caae",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# how to load peft model from hub for inference\n",
+    "config = PeftConfig.from_pretrained(model_id)\n",
+    "inference_model = AutoModelForSequenceClassification.from_pretrained(\n",
+    "    config.base_model_name_or_path, num_labels=2, id2label=id2label, label2id=label2id\n",
+    ")\n",
+    "tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)\n",
+    "model = PeftModel.from_pretrained(inference_model, model_id)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "77c6ed42-8ec3-4343-9e42-405feac052ba",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Work1",
+   "language": "python",
+   "name": "work1"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.18"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}