From d3cce99c850c94bc799bac99b87ac3253a64b47b Mon Sep 17 00:00:00 2001 From: s444415 Date: Fri, 24 Mar 2023 11:47:10 +0000 Subject: [PATCH] streaming training, from zero configuration --- .gitignore | 4 +- config-eval.yaml | 16 +- config-train.yaml | 20 +- eval.py | 6 +- notepads/donut-from-zero-train.ipynb | 494 +++++++++++++++++++++++++++ notepads/workbook.ipynb | 0 train_stream.py | 11 + 7 files changed, 531 insertions(+), 20 deletions(-) create mode 100644 notepads/donut-from-zero-train.ipynb create mode 100644 notepads/workbook.ipynb diff --git a/.gitignore b/.gitignore index e7809c4..bf11051 100644 --- a/.gitignore +++ b/.gitignore @@ -5,4 +5,6 @@ wandb __pycache__/ checkpoint .vscode -donut_env \ No newline at end of file +donut_env +env_donut +*.out \ No newline at end of file diff --git a/config-eval.yaml b/config-eval.yaml index 77586b4..c5bb1a8 100644 --- a/config-eval.yaml +++ b/config-eval.yaml @@ -1,11 +1,11 @@ -pretrained_processor_path: "Zombely/plwiki-proto-fine-tuned-v2" -pretrained_model_path: "Zombely/plwiki-proto-fine-tuned-v2" -validation_dataset_path: "Zombely/diachronia-ocr" +pretrained_processor_path: "Zombely/pl-donut-v1.2" +pretrained_model_path: "Zombely/pl-donut-v1.2" +validation_dataset_path: "Zombely/diachronia-ocr-train" validation_dataset_split: "train" -has_metadata: False +has_metadata: True print_output: True -output_file_dir: "../../gonito-outs" -test_name: "fine-tuned-test" -image_size: [1920, 2560] +output_file_dir: "" +test_name: "fine-tuned" +image_size: [1280, 960] use_enc_dec_config: False -max_dec_length: 768 \ No newline at end of file +max_dec_length: 768 diff --git a/config-train.yaml b/config-train.yaml index d3857e3..7cad7a1 100644 --- a/config-train.yaml +++ b/config-train.yaml @@ -1,22 +1,22 @@ -dataset_path: "Zombely/wikisource-yellow" -pretrained_model_path: "Zombely/plwiki-proto-fine-tuned-v3.2" -start_model_path: "Zombely/plwiki-proto-fine-tuned-v3.2" -output_model_path: "Zombely/pl-donut" -wandb_test_name: "wikisource-small" +dataset_path: "Zombely/wikisource-red" +pretrained_model_path: "Zombely/pl-donut-v1.1" +start_model_path: "Zombely/pl-donut-v1.1" +output_model_path: "Zombely/pl-donut-v1.2" +wandb_test_name: "pl-donut-v1.2" checkpoint_path: "./checkpoint" max_length: 768 -image_size: [1280, 960] +image_size: [1260, 960] train_config: max_epochs: 1 val_check_interval: 1.0 check_val_every_n_epoch: 1 gradient_clip_val: 1.0 num_training_samples_per_epoch: 800 - lr: 3.0e-5 + lr: 1.0e-4 train_batch_sizes: [8] val_batch_sizes: [1] - seed: 2022 + seed: 2023 num_nodes: 1 - warmup_steps: 300 + warmup_steps: 10 result_path: "./result" - verbose: True \ No newline at end of file + verbose: True diff --git a/eval.py b/eval.py index 8fd5268..72aa761 100644 --- a/eval.py +++ b/eval.py @@ -69,7 +69,11 @@ def main(config): accs.append(score) if config.print_output: - print(seq) + if 'ground_truth' in sample: + ground_truth = json.loads(sample["ground_truth"]) + ground_truth = str(ground_truth["gt_parse"]) + print("Original: ", ground_truth + "\n") + print("Prediction: ", str(seq) + "\n") output_list.append(seq) if config.output_file_dir: df = pd.DataFrame(map(lambda x: x.get('text_sequence', ''), output_list)) diff --git a/notepads/donut-from-zero-train.ipynb b/notepads/donut-from-zero-train.ipynb new file mode 100644 index 0000000..06cce37 --- /dev/null +++ b/notepads/donut-from-zero-train.ipynb @@ -0,0 +1,494 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "from transformers import VisionEncoderDecoderConfig, DonutProcessor, VisionEncoderDecoderModel, DonutImageProcessor, XLMRobertaTokenizerFast, BertConfig, ViTConfig\n", + "from datasets import load_dataset, interleave_datasets\n", + "import json\n", + "import random\n", + "from typing import Any, List, Tuple\n", + "import torch\n", + "from torch.utils.data import Dataset, DataLoader\n", + "import re\n", + "from nltk import edit_distance\n", + "import numpy as np\n", + "from pytorch_lightning.loggers import WandbLogger\n", + "from pytorch_lightning.callbacks import Callback\n", + "import pytorch_lightning as pl\n", + "import os\n", + "from huggingface_hub import login\n" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [], + "source": [ + "import torch\n", + "import pytorch_lightning as pl\n", + "from nltk import edit_distance\n", + "import re\n", + "import numpy as np\n", + "\n", + "\n", + "class DonutModelPLModuleStream(pl.LightningModule):\n", + " def __init__(self, config, processor, model, max_length, train_dataloader, val_dataloader):\n", + " super().__init__()\n", + " self.config = config\n", + " self.processor = processor\n", + " self.model = model\n", + " self.max_length = max_length\n", + " self._train_dataloader = train_dataloader\n", + " self._val_dataloader = val_dataloader\n", + "\n", + " def training_step(self, batch, batch_idx):\n", + " # pixel_values, labels, _ = batch\n", + " pixel_values = batch['pixel_values']\n", + " labels = batch['labels']\n", + " outputs = self.model(pixel_values, labels=labels)\n", + " loss = outputs.loss\n", + " self.log_dict({\"train_loss\": loss}, sync_dist=True)\n", + " return loss\n", + "\n", + " def validation_step(self, batch, batch_idx, dataset_idx=0):\n", + " # pixel_values, labels, answers = batch\n", + "\n", + " pixel_values = batch['pixel_values']\n", + " labels = batch['labels']\n", + " answers = batch['target_sequence'][0]\n", + " batch_size = pixel_values.shape[0]\n", + " # we feed the prompt to the model\n", + " decoder_input_ids = torch.full((batch_size, 1), self.model.config.decoder_start_token_id, device=self.device)\n", + " \n", + " outputs = self.model.generate(pixel_values,\n", + " decoder_input_ids=decoder_input_ids,\n", + " max_length=self.max_length,\n", + " early_stopping=True,\n", + " pad_token_id=self.processor.tokenizer.pad_token_id,\n", + " eos_token_id=self.processor.tokenizer.eos_token_id,\n", + " use_cache=True,\n", + " num_beams=1,\n", + " bad_words_ids=[[self.processor.tokenizer.unk_token_id]],\n", + " return_dict_in_generate=True,)\n", + " \n", + " predictions = []\n", + " for seq in self.processor.tokenizer.batch_decode(outputs.sequences):\n", + " seq = seq.replace(self.processor.tokenizer.eos_token, \"\").replace(self.processor.tokenizer.pad_token, \"\")\n", + " seq = re.sub(r\"<.*?>\", \"\", seq, count=1).strip() # remove first task start token\n", + " predictions.append(seq)\n", + "\n", + " scores = list()\n", + " for pred, answer in zip(predictions, answers):\n", + " pred = re.sub(r\"(?:(?<=>) | (?=\", \"\", answer, count=1)\n", + " answer = answer.replace(self.processor.tokenizer.eos_token, \"\")\n", + " scores.append(edit_distance(pred, answer) / max(len(pred), len(answer)))\n", + "\n", + " if self.config.get(\"verbose\", False) and len(scores) == 1:\n", + " print(f\"Prediction: {pred}\")\n", + " print(f\" Answer: {answer}\")\n", + " print(f\" Normed ED: {scores[0]}\")\n", + "\n", + " return scores\n", + "\n", + " def validation_epoch_end(self, validation_step_outputs):\n", + " # I set this to 1 manually\n", + " # (previously set to len(self.config.dataset_name_or_paths))\n", + " num_of_loaders = 1\n", + " if num_of_loaders == 1:\n", + " validation_step_outputs = [validation_step_outputs]\n", + " assert len(validation_step_outputs) == num_of_loaders\n", + " cnt = [0] * num_of_loaders\n", + " total_metric = [0] * num_of_loaders\n", + " val_metric = [0] * num_of_loaders\n", + " for i, results in enumerate(validation_step_outputs):\n", + " for scores in results:\n", + " cnt[i] += len(scores)\n", + " total_metric[i] += np.sum(scores)\n", + " val_metric[i] = total_metric[i] / cnt[i]\n", + " val_metric_name = f\"val_metric_{i}th_dataset\"\n", + " self.log_dict({val_metric_name: val_metric[i]}, sync_dist=True)\n", + " self.log_dict({\"val_metric\": np.sum(total_metric) / np.sum(cnt)}, sync_dist=True)\n", + "\n", + " def configure_optimizers(self):\n", + " # TODO add scheduler\n", + " optimizer = torch.optim.Adam(self.parameters(), lr=self.config.get(\"lr\"))\n", + " \n", + " return optimizer\n", + "\n", + " def train_dataloader(self):\n", + " return self._train_dataloader\n", + "\n", + " def val_dataloader(self):\n", + " return self._val_dataloader" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [], + "source": [ + "# dataset = load_dataset" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [], + "source": [ + "image_processor = DonutImageProcessor(do_resize=True, do_align_long_axis=False, size=[960, 1260])\n", + "tokenizer = XLMRobertaTokenizerFast.from_pretrained('xlm-roberta-base')" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [], + "source": [ + "config_encoder = ViTConfig(image_size=[1260, 960])\n", + "config_decoder = BertConfig()\n", + "config = VisionEncoderDecoderConfig.from_encoder_decoder_configs(config_encoder, config_decoder)" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [], + "source": [ + "processor = DonutProcessor(image_processor=image_processor, tokenizer=tokenizer)\n", + "model = VisionEncoderDecoderModel(config=config)" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [], + "source": [ + "added_tokens = []\n", + "\n", + "### PROCESS FUNC START ###\n", + "\n", + "def add_tokens(list_of_tokens: List[str]):\n", + " \"\"\"\n", + " Add special tokens to tokenizer and resize the token embeddings of the decoder\n", + " \"\"\"\n", + " newly_added_num = processor.tokenizer.add_tokens(list_of_tokens)\n", + " if newly_added_num > 0:\n", + " model.decoder.resize_token_embeddings(len(processor.tokenizer))\n", + " added_tokens.extend(list_of_tokens)\n", + "\n", + "def json2token(obj: Any, update_special_tokens_for_json_key: bool = True, sort_json_key: bool = True):\n", + " \"\"\"\n", + " Convert an ordered JSON object into a token sequence\n", + " \"\"\"\n", + " if type(obj) == dict:\n", + " if len(obj) == 1 and \"text_sequence\" in obj:\n", + " return obj[\"text_sequence\"]\n", + " else:\n", + " output = \"\"\n", + " if sort_json_key:\n", + " keys = sorted(obj.keys(), reverse=True)\n", + " else:\n", + " keys = obj.keys()\n", + " for k in keys:\n", + " if update_special_tokens_for_json_key:\n", + " add_tokens([fr\"\", fr\"\"])\n", + " output += (\n", + " fr\"\"\n", + " + json2token(obj[k], update_special_tokens_for_json_key, sort_json_key)\n", + " + fr\"\"\n", + " )\n", + " return output\n", + " elif type(obj) == list:\n", + " return r\"\".join(\n", + " [json2token(item, update_special_tokens_for_json_key, sort_json_key) for item in obj]\n", + " )\n", + " else:\n", + " obj = str(obj)\n", + " if f\"<{obj}/>\" in added_tokens:\n", + " obj = f\"<{obj}/>\" # for categorical special tokens\n", + " return obj\n", + "\n", + "def process(row, split):\n", + " task_start_token, prompt_end_token = \"\", \"\"\n", + " ground_truth = json.loads(row[\"ground_truth\"])\n", + " if \"gt_parses\" in ground_truth: # when multiple ground truths are available, e.g., docvqa\n", + " assert isinstance(ground_truth[\"gt_parses\"], list)\n", + " gt_jsons = ground_truth[\"gt_parses\"]\n", + " else:\n", + " assert \"gt_parse\" in ground_truth and isinstance(ground_truth[\"gt_parse\"], dict)\n", + " gt_jsons = [ground_truth[\"gt_parse\"]]\n", + "\n", + " gt_token_sequences = (\n", + " [\n", + " json2token(\n", + " gt_json,\n", + " update_special_tokens_for_json_key=split == \"train\",\n", + " sort_json_key=False,\n", + " )\n", + " + processor.tokenizer.eos_token\n", + " for gt_json in gt_jsons # load json from list of json\n", + " ]\n", + " )\n", + "\n", + " add_tokens([task_start_token, prompt_end_token])\n", + " prompt_end_token_id = processor.tokenizer.convert_tokens_to_ids(prompt_end_token)\n", + "\n", + " # change if not 3 channels\n", + " if row['image'].mode != \"RGB\":\n", + " row['image'] = row['image'].convert(\"RGB\")\n", + " # inputs\n", + " pixel_values = processor(row[\"image\"], random_padding=split == \"train\", return_tensors=\"pt\").pixel_values\n", + " pixel_values = pixel_values.squeeze()\n", + "\n", + " # targets\n", + " input_ids = processor.tokenizer(\n", + " gt_token_sequences,\n", + " add_special_tokens=False,\n", + " max_length=config.max_length,\n", + " padding=\"max_length\",\n", + " truncation=True,\n", + " return_tensors=\"pt\",\n", + " )[\"input_ids\"].squeeze(0)\n", + "\n", + " labels = input_ids.clone()\n", + " labels[labels == processor.tokenizer.pad_token_id] = -100 # model doesn't need to predict pad token\n", + " return {\"pixel_values\": pixel_values, \"labels\": labels, 'target_sequence': gt_token_sequences }\n", + "\n", + "def proces_train(row):\n", + " return process(row, 'train')\n", + "\n", + "def proces_val(row):\n", + " return process(row, 'validation')\n" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Using custom data configuration Zombely--wikisource-red-98affb32ced5f2c5\n" + ] + } + ], + "source": [ + "dataset = load_dataset('Zombely/wikisource-red', streaming=True)\n", + "val_dataset = dataset.pop('validation') \n", + "train_dataset = interleave_datasets(list(dataset.values()))\n", + "# train_length = sum(split.num_examples for split in dataset[list(dataset.keys())[0]].info.splits.values() if split.name != 'validation')\n", + "# val_length = list(val_dataset.info.splits.values())[-1].num_examples\n", + "\n", + "\n", + "train_dataset = train_dataset.map(proces_train, remove_columns = ['image', 'ground_truth'])\n", + "val_dataset = val_dataset.map(proces_val, remove_columns = ['image', 'ground_truth'])\n", + "\n", + "train_dataset = train_dataset.with_format('torch')\n", + "val_dataset = val_dataset.with_format('torch')\n", + "\n", + "# train_dataset = CustomWrapperIterator(train_dataset, total_len=train_length)\n", + "# val_dataset = CustomWrapperIterator(val_dataset, total_len=val_length)\n", + "\n", + "model.config.pad_token_id = processor.tokenizer.pad_token_id\n", + "model.config.decoder_start_token_id = processor.tokenizer.convert_tokens_to_ids([''])[0]\n", + "\n", + "train_dataloader = DataLoader(train_dataset, batch_size=1, num_workers=0)\n", + "val_dataloader = DataLoader(val_dataset, batch_size=1, num_workers=0)\n" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": {}, + "outputs": [], + "source": [ + "train_config = {\n", + " \"max_epochs\": 1,\n", + " \"val_check_interval\": 1.0,\n", + " \"check_val_every_n_epoch\": 1,\n", + " \"gradient_clip_val\": 1.0,\n", + " \"num_training_samples_per_epoch\": 800,\n", + " \"lr\": 1.0e-4,\n", + " \"train_batch_sizes\": [8],\n", + " \"val_batch_sizes\": [1],\n", + " \"seed\": 2023,\n", + " \"num_nodes\": 1,\n", + " \"warmup_steps\": 10,\n", + " \"result_path\": \"./result\",\n", + " \"verbose\": True\n", + "}\n" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, + "outputs": [], + "source": [ + "model_module = DonutModelPLModuleStream(train_config, processor, model, max_length=config.max_length, train_dataloader=train_dataloader, val_dataloader=val_dataloader)" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Using 16bit native Automatic Mixed Precision (AMP)\n", + "GPU available: True (cuda), used: True\n", + "TPU available: False, using: 0 TPU cores\n", + "IPU available: False, using: 0 IPUs\n", + "HPU available: False, using: 0 HPUs\n", + "`Trainer(val_check_interval=1.0)` was configured so validation will run at the end of the training epoch..\n" + ] + } + ], + "source": [ + "\n", + "trainer = pl.Trainer(\n", + " accelerator=\"gpu\" if torch.cuda.is_available() else 'cpu', # change to gpu\n", + " devices=1,\n", + " max_epochs=train_config['max_epochs'],\n", + " val_check_interval=train_config['val_check_interval'],\n", + " check_val_every_n_epoch=train_config['check_val_every_n_epoch'],\n", + " gradient_clip_val=train_config['gradient_clip_val'],\n", + " precision=16, # we'll use mixed precision\n", + " num_sanity_val_steps=0,\n", + ")\n" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Missing logger folder: /home/wmi/project/donut/notepads/lightning_logs\n", + "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n", + "\n", + " | Name | Type | Params\n", + "----------------------------------------------------\n", + "0 | model | VisionEncoderDecoderModel | 227 M \n", + "----------------------------------------------------\n", + "227 M Trainable params\n", + "0 Non-trainable params\n", + "227 M Total params\n", + "455.428 Total estimated model params size (MB)\n", + "/home/wmi/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:224: PossibleUserWarning: The dataloader, train_dataloader, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 14 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.\n", + " rank_zero_warn(\n", + "/home/wmi/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:224: PossibleUserWarning: The dataloader, val_dataloader 0, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 14 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.\n", + " rank_zero_warn(\n" + ] + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "15d82804a8fe4aa6b2a02a16ce144496", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Training: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(1024, 1579)\n", + "(1024, 1473)\n" + ] + }, + { + "ename": "OutOfMemoryError", + "evalue": "CUDA out of memory. Tried to allocate 368.00 MiB (GPU 0; 23.70 GiB total capacity; 22.07 GiB already allocated; 260.56 MiB free; 22.09 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mOutOfMemoryError\u001b[0m Traceback (most recent call last)", + "\u001b[0;32m/tmp/ipykernel_32385/828374167.py\u001b[0m in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel_module\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)\u001b[0m\n\u001b[1;32m 580\u001b[0m \u001b[0;32mraise\u001b[0m \u001b[0mTypeError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"`Trainer.fit()` requires a `LightningModule`, got: {model.__class__.__qualname__}\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 581\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_lightning_module\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 582\u001b[0;31m call._call_and_handle_interrupt(\n\u001b[0m\u001b[1;32m 583\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_fit_impl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_dataloaders\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_dataloaders\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdatamodule\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mckpt_path\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 584\u001b[0m )\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/call.py\u001b[0m in \u001b[0;36m_call_and_handle_interrupt\u001b[0;34m(trainer, trainer_fn, *args, **kwargs)\u001b[0m\n\u001b[1;32m 36\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlauncher\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlaunch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrainer_fn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrainer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 37\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 38\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mtrainer_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 39\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 40\u001b[0m \u001b[0;32mexcept\u001b[0m \u001b[0m_TunerExitException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_fit_impl\u001b[0;34m(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)\u001b[0m\n\u001b[1;32m 622\u001b[0m \u001b[0mmodel_connected\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 623\u001b[0m )\n\u001b[0;32m--> 624\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mckpt_path\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mckpt_path\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 625\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 626\u001b[0m \u001b[0;32massert\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstopped\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_run\u001b[0;34m(self, model, ckpt_path)\u001b[0m\n\u001b[1;32m 1059\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_checkpoint_connector\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mresume_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1060\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1061\u001b[0;31m \u001b[0mresults\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_stage\u001b[0m\u001b[0;34m(\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 1062\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1063\u001b[0m \u001b[0mlog\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdetail\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"{self.__class__.__name__}: trainer tearing down\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_run_stage\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 1138\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredicting\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1139\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_predict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1140\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_train\u001b[0m\u001b[0;34m(\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 1141\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1142\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m_pre_training_routine\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_run_train\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 1161\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1162\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_detect_anomaly\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_detect_anomaly\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1163\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\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 1164\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1165\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m_run_evaluate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0m_EVALUATE_OUTPUT\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/loop.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 197\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 198\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 199\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 200\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 201\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/fit_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 265\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_data_fetcher\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msetup\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdataloader\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_to_device\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mbatch_to_device\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 266\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"run_training_epoch\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 267\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_outputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mepoch_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_data_fetcher\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 268\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 269\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/loop.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 197\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 198\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 199\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 200\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 201\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/epoch/training_epoch_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self, data_fetcher)\u001b[0m\n\u001b[1;32m 212\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 213\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"run_training_batch\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 214\u001b[0;31m \u001b[0mbatch_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbatch_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\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 215\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 216\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbatch_progress\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mincrement_processed\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/loop.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 197\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 198\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 199\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 200\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 201\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/batch/training_batch_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self, kwargs)\u001b[0m\n\u001b[1;32m 86\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_frequencies\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"batch_idx\"\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[0m\n\u001b[1;32m 87\u001b[0m )\n\u001b[0;32m---> 88\u001b[0;31m \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moptimizers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\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 89\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 90\u001b[0m \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmanual_loop\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/loop.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 197\u001b[0m \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 198\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_start\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 199\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madvance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 200\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_advance_end\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 201\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_restarting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36madvance\u001b[0;34m(self, optimizers, kwargs)\u001b[0m\n\u001b[1;32m 198\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_build_kwargs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_hiddens\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 199\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 200\u001b[0;31m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_optimization\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizers\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptim_progress\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_position\u001b[0m\u001b[0;34m]\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 201\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloss\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 202\u001b[0m \u001b[0;31m# automatic optimization assumes a loss needs to be returned for extras to be considered as the batch\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m_run_optimization\u001b[0;34m(self, kwargs, optimizer)\u001b[0m\n\u001b[1;32m 245\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 246\u001b[0m \u001b[0;31m# the `batch_idx` is optional with inter-batch parallelism\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 247\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizer_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopt_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"batch_idx\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\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 248\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 249\u001b[0m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconsume_result\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m_optimizer_step\u001b[0;34m(self, optimizer, opt_idx, batch_idx, train_step_and_backward_closure)\u001b[0m\n\u001b[1;32m 355\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 356\u001b[0m \u001b[0;31m# model hook\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 357\u001b[0;31m self.trainer._call_lightning_module_hook(\n\u001b[0m\u001b[1;32m 358\u001b[0m \u001b[0;34m\"optimizer_step\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 359\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcurrent_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_call_lightning_module_hook\u001b[0;34m(self, hook_name, pl_module, *args, **kwargs)\u001b[0m\n\u001b[1;32m 1303\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1304\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"[LightningModule]{pl_module.__class__.__name__}.{hook_name}\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1305\u001b[0;31m \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 1306\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1307\u001b[0m \u001b[0;31m# restore current_fx when nested context\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/core/module.py\u001b[0m in \u001b[0;36moptimizer_step\u001b[0;34m(self, epoch, batch_idx, optimizer, optimizer_idx, optimizer_closure, on_tpu, using_native_amp, using_lbfgs)\u001b[0m\n\u001b[1;32m 1659\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1660\u001b[0m \"\"\"\n\u001b[0;32m-> 1661\u001b[0;31m \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0moptimizer_closure\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 1662\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1663\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0moptimizer_zero_grad\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepoch\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_idx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mOptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/core/optimizer.py\u001b[0m in \u001b[0;36mstep\u001b[0;34m(self, closure, **kwargs)\u001b[0m\n\u001b[1;32m 167\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 168\u001b[0m \u001b[0;32massert\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_strategy\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 169\u001b[0;31m \u001b[0mstep_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_strategy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_optimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 170\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 171\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_on_after_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/strategies/strategy.py\u001b[0m in \u001b[0;36moptimizer_step\u001b[0;34m(self, optimizer, opt_idx, closure, model, **kwargs)\u001b[0m\n\u001b[1;32m 232\u001b[0m \u001b[0;31m# TODO(lite): remove assertion once strategy's optimizer_step typing is fixed\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 233\u001b[0m \u001b[0;32massert\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpl\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mLightningModule\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 234\u001b[0;31m return self.precision_plugin.optimizer_step(\n\u001b[0m\u001b[1;32m 235\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mopt_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 236\u001b[0m )\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/plugins/precision/native_amp.py\u001b[0m in \u001b[0;36moptimizer_step\u001b[0;34m(self, optimizer, model, optimizer_idx, closure, **kwargs)\u001b[0m\n\u001b[1;32m 83\u001b[0m \u001b[0;34mf\"Native AMP and the LBFGS optimizer are not compatible (optimizer {optimizer_idx}).\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 84\u001b[0m )\n\u001b[0;32m---> 85\u001b[0;31m \u001b[0mclosure_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m(\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 86\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 87\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0m_optimizer_handles_unscaling\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moptimizer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 145\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 146\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mAny\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mAny\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mOptional\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mTensor\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 147\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 148\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_result\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 149\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36mclosure\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m 140\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 141\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_backward_fn\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mstep_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclosure_loss\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 142\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_backward_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclosure_loss\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 143\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 144\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mstep_output\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py\u001b[0m in \u001b[0;36mbackward_fn\u001b[0;34m(loss)\u001b[0m\n\u001b[1;32m 301\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 302\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mbackward_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 303\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrainer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call_strategy_hook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"backward\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mloss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopt_idx\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 304\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 305\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mbackward_fn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/trainer/trainer.py\u001b[0m in \u001b[0;36m_call_strategy_hook\u001b[0;34m(self, hook_name, *args, **kwargs)\u001b[0m\n\u001b[1;32m 1441\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1442\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"[Strategy]{self.strategy.__class__.__name__}.{hook_name}\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1443\u001b[0;31m \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 1444\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1445\u001b[0m \u001b[0;31m# restore current_fx when nested context\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/strategies/strategy.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(self, closure_loss, optimizer, optimizer_idx, *args, **kwargs)\u001b[0m\n\u001b[1;32m 205\u001b[0m \u001b[0mclosure_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprecision_plugin\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpre_backward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 206\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 207\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprecision_plugin\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 208\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 209\u001b[0m \u001b[0mclosure_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprecision_plugin\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpost_backward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclosure_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlightning_module\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/plugins/precision/precision_plugin.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(self, tensor, model, optimizer, optimizer_idx, *args, **kwargs)\u001b[0m\n\u001b[1;32m 67\u001b[0m \u001b[0;31m\\\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mKeyword\u001b[0m \u001b[0marguments\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mthe\u001b[0m \u001b[0msame\u001b[0m \u001b[0mpurpose\u001b[0m \u001b[0;32mas\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 68\u001b[0m \"\"\"\n\u001b[0;32m---> 69\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtensor\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 70\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 71\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mpost_backward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtensor\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodule\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;34m\"pl.LightningModule\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;31m# type: ignore[override]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/pytorch_lightning/core/module.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(self, loss, optimizer, optimizer_idx, *args, **kwargs)\u001b[0m\n\u001b[1;32m 1404\u001b[0m \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1405\u001b[0m \"\"\"\n\u001b[0;32m-> 1406\u001b[0;31m \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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 1407\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1408\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mtoggle_optimizer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mUnion\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mOptimizer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mLightningOptimizer\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_idx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/torch/_tensor.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(self, gradient, retain_graph, create_graph, inputs)\u001b[0m\n\u001b[1;32m 486\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 487\u001b[0m )\n\u001b[0;32m--> 488\u001b[0;31m torch.autograd.backward(\n\u001b[0m\u001b[1;32m 489\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgradient\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 490\u001b[0m )\n", + "\u001b[0;32m~/project/donut/env_donut/lib/python3.10/site-packages/torch/autograd/__init__.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)\u001b[0m\n\u001b[1;32m 195\u001b[0m \u001b[0;31m# some Python versions print out the first line of a multi-line function\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 196\u001b[0m \u001b[0;31m# calls in the traceback and some print out the last line\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 197\u001b[0;31m Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass\n\u001b[0m\u001b[1;32m 198\u001b[0m \u001b[0mtensors\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgrad_tensors_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 199\u001b[0m allow_unreachable=True, accumulate_grad=True) # Calls into the C++ engine to run the backward pass\n", + "\u001b[0;31mOutOfMemoryError\u001b[0m: CUDA out of memory. Tried to allocate 368.00 MiB (GPU 0; 23.70 GiB total capacity; 22.07 GiB already allocated; 260.56 MiB free; 22.09 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF" + ] + } + ], + "source": [ + "trainer.fit(model_module)" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "env_donut", + "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": "3.10.6" + }, + "orig_nbformat": 4 + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notepads/workbook.ipynb b/notepads/workbook.ipynb new file mode 100644 index 0000000..e69de29 diff --git a/train_stream.py b/train_stream.py index 86fb836..dd46273 100644 --- a/train_stream.py +++ b/train_stream.py @@ -167,6 +167,17 @@ def main(config, hug_token): # added_tokens=added_tokens, # sort_json_key=False, # cord dataset is preprocessed, so no need for this # ) + + # dataset_green = load_dataset("Zombely/wikisource-green", streaming=True) + # val_dataset = dataset_green.pop('validation') + # val_length = list(val_dataset.info.splits.values())[-1].num_examples + # dataset_yellow = load_dataset("Zombely/wikisource-yellow", streaming=True) + # dataset_red = load_dataset("Zombely/wikisource-red", streaming=True) + # train_dataset = interleave_datasets(list(dataset_green.values()) + list(dataset_yellow.values()) + list(dataset_red.values())) + # train_length_green = sum(split.num_examples for split in dataset_green[list(dataset_green.keys())[0]].info.splits.values() if split.name != 'validation') + # train_length_yellow = sum(split.num_examples for split in dataset_yellow[list(dataset_yellow.keys())[0]].info.splits.values()) + # train_length_red = sum(split.num_examples for split in dataset_red[list(dataset_red.keys())[0]].info.splits.values()) + # train_length = train_length_green + train_length_yellow + train_length_red dataset = load_dataset(config.dataset_path, streaming=True) val_dataset = dataset.pop('validation')