projekt-glebokie/GPT_2.ipynb at b4227f7732afc8a72e0a90846a09cac10857e686

Andrzej Preibisz 77c31c5657 GPT-2 Fixed test dataset size

2023-02-12 23:36:28 +01:00

73 KiB

Raw Blame History

! pip install datasets transformers torch scikit-learn evaluate

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting datasets
  Downloading datasets-2.9.0-py3-none-any.whl (462 kB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m462.8/462.8 KB[0m [31m8.1 MB/s[0m eta [36m0:00:00[0m
[?25hCollecting transformers
  Downloading transformers-4.26.1-py3-none-any.whl (6.3 MB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m6.3/6.3 MB[0m [31m55.6 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: torch in /usr/local/lib/python3.8/dist-packages (1.13.1+cu116)
Requirement already satisfied: scikit-learn in /usr/local/lib/python3.8/dist-packages (1.0.2)
Collecting evaluate
  Downloading evaluate-0.4.0-py3-none-any.whl (81 kB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m81.4/81.4 KB[0m [31m7.5 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.8/dist-packages (from datasets) (1.21.6)
Collecting xxhash
  Downloading xxhash-3.2.0-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (213 kB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m213.0/213.0 KB[0m [31m2.7 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: fsspec[http]>=2021.11.1 in /usr/local/lib/python3.8/dist-packages (from datasets) (2023.1.0)
Requirement already satisfied: aiohttp in /usr/local/lib/python3.8/dist-packages (from datasets) (3.8.3)
Collecting responses<0.19
  Downloading responses-0.18.0-py3-none-any.whl (38 kB)
Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.8/dist-packages (from datasets) (6.0)
Requirement already satisfied: pyarrow>=6.0.0 in /usr/local/lib/python3.8/dist-packages (from datasets) (9.0.0)
Collecting huggingface-hub<1.0.0,>=0.2.0
  Downloading huggingface_hub-0.12.0-py3-none-any.whl (190 kB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m190.3/190.3 KB[0m [31m16.6 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: dill<0.3.7 in /usr/local/lib/python3.8/dist-packages (from datasets) (0.3.6)
Requirement already satisfied: tqdm>=4.62.1 in /usr/local/lib/python3.8/dist-packages (from datasets) (4.64.1)
Requirement already satisfied: requests>=2.19.0 in /usr/local/lib/python3.8/dist-packages (from datasets) (2.25.1)
Requirement already satisfied: pandas in /usr/local/lib/python3.8/dist-packages (from datasets) (1.3.5)
Requirement already satisfied: packaging in /usr/local/lib/python3.8/dist-packages (from datasets) (23.0)
Collecting multiprocess
  Downloading multiprocess-0.70.14-py38-none-any.whl (132 kB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m132.0/132.0 KB[0m [31m8.6 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.8/dist-packages (from transformers) (2022.6.2)
Requirement already satisfied: filelock in /usr/local/lib/python3.8/dist-packages (from transformers) (3.9.0)
Collecting tokenizers!=0.11.3,<0.14,>=0.11.1
  Downloading tokenizers-0.13.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (7.6 MB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m7.6/7.6 MB[0m [31m51.8 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: typing-extensions in /usr/local/lib/python3.8/dist-packages (from torch) (4.4.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (3.1.0)
Requirement already satisfied: scipy>=1.1.0 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (1.7.3)
Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.8/dist-packages (from scikit-learn) (1.2.0)
Requirement already satisfied: yarl<2.0,>=1.0 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (1.8.2)
Requirement already satisfied: attrs>=17.3.0 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (22.2.0)
Requirement already satisfied: frozenlist>=1.1.1 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (1.3.3)
Requirement already satisfied: async-timeout<5.0,>=4.0.0a3 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (4.0.2)
Requirement already satisfied: multidict<7.0,>=4.5 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (6.0.4)
Requirement already satisfied: charset-normalizer<3.0,>=2.0 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (2.1.1)
Requirement already satisfied: aiosignal>=1.1.2 in /usr/local/lib/python3.8/dist-packages (from aiohttp->datasets) (1.3.1)
Requirement already satisfied: chardet<5,>=3.0.2 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->datasets) (4.0.0)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->datasets) (1.24.3)
Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->datasets) (2.10)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->datasets) (2022.12.7)
Collecting urllib3<1.27,>=1.21.1
  Downloading urllib3-1.26.14-py2.py3-none-any.whl (140 kB)
[2K     [90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━[0m [32m140.6/140.6 KB[0m [31m9.0 MB/s[0m eta [36m0:00:00[0m
[?25hRequirement already satisfied: pytz>=2017.3 in /usr/local/lib/python3.8/dist-packages (from pandas->datasets) (2022.7.1)
Requirement already satisfied: python-dateutil>=2.7.3 in /usr/local/lib/python3.8/dist-packages (from pandas->datasets) (2.8.2)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.8/dist-packages (from python-dateutil>=2.7.3->pandas->datasets) (1.15.0)
Installing collected packages: tokenizers, xxhash, urllib3, multiprocess, responses, huggingface-hub, transformers, datasets, evaluate
  Attempting uninstall: urllib3
    Found existing installation: urllib3 1.24.3
    Uninstalling urllib3-1.24.3:
      Successfully uninstalled urllib3-1.24.3
Successfully installed datasets-2.9.0 evaluate-0.4.0 huggingface-hub-0.12.0 multiprocess-0.70.14 responses-0.18.0 tokenizers-0.13.2 transformers-4.26.1 urllib3-1.26.14 xxhash-3.2.0

!wget 'https://git.wmi.amu.edu.pl/s444465/projekt-glebokie/raw/branch/master/run_glue.py' -O 'run_glue.py'
!wget 'https://git.wmi.amu.edu.pl/s444465/projekt-glebokie/raw/branch/master/roberta.py' -O 'roberta.py'
!wget 'https://git.wmi.amu.edu.pl/s444465/projekt-glebokie/raw/branch/master/gpt2.py' -O 'gpt2.py'

--2023-02-12 21:57:57--  https://git.wmi.amu.edu.pl/s444465/projekt-glebokie/raw/branch/master/run_glue.py
Resolving git.wmi.amu.edu.pl (git.wmi.amu.edu.pl)... 150.254.78.40
Connecting to git.wmi.amu.edu.pl (git.wmi.amu.edu.pl)|150.254.78.40|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 30601 (30K) [text/plain]
Saving to: ‘run_glue.py’

run_glue.py         100%[===================>]  29.88K  --.-KB/s    in 0.1s    

2023-02-12 21:57:58 (248 KB/s) - ‘run_glue.py’ saved [30601/30601]

--2023-02-12 21:57:58--  https://git.wmi.amu.edu.pl/s444465/projekt-glebokie/raw/branch/master/roberta.py
Resolving git.wmi.amu.edu.pl (git.wmi.amu.edu.pl)... 150.254.78.40
Connecting to git.wmi.amu.edu.pl (git.wmi.amu.edu.pl)|150.254.78.40|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 12783 (12K) [text/plain]
Saving to: ‘roberta.py’

roberta.py          100%[===================>]  12.48K  --.-KB/s    in 0s      

2023-02-12 21:57:58 (265 MB/s) - ‘roberta.py’ saved [12783/12783]

--2023-02-12 21:57:58--  https://git.wmi.amu.edu.pl/s444465/projekt-glebokie/raw/branch/master/gpt2.py
Resolving git.wmi.amu.edu.pl (git.wmi.amu.edu.pl)... 150.254.78.40
Connecting to git.wmi.amu.edu.pl (git.wmi.amu.edu.pl)|150.254.78.40|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 7976 (7.8K) [text/plain]
Saving to: ‘gpt2.py’

gpt2.py             100%[===================>]   7.79K  --.-KB/s    in 0s      

2023-02-12 21:57:59 (1.37 GB/s) - ‘gpt2.py’ saved [7976/7976]

import json
from pathlib import Path
from typing import Dict, List
from datasets import load_dataset

loaded_data = load_dataset('emotion')
!mkdir -v -p data

train_path = Path('data/train.json')
valid_path = Path('data/valid.json')
test_path = Path('data/test.json')
data_train, data_valid, data_test = [], [], []

for source_data, dataset, max_size in [
  (loaded_data['train'], data_train, None),
  (loaded_data['validation'], data_valid, None),
  (loaded_data['test'], data_test, None),
]:
  for i, data in enumerate(source_data):
    if max_size is not None and i >= max_size:
      break
    data_line = {
      'label': int(data['label']),
      'text': data['text'],
    }
    dataset.append(data_line)

print(f'Train: {len(data_train):6d}')
print(f'Valid: {len(data_valid):6d}')

data_class_1, data_class_2 = [], []

"""for data in data_valid:
  label = data['label']
  if label == 0:
    data_class_1.append(data)
  elif label == 1:
    data_class_2.append(data)

print(f'Label 1: {len(data_class_1):6d}')
print(f'Label 2: {len(data_class_2):6d}')

size_half_class_1 = int(len(data_class_1) / 2)
size_half_class_2 = int(len(data_class_2) / 2)

data_valid = data_class_1[:size_half_class_1] + data_class_2[:size_half_class_2]
data_test = data_class_1[size_half_class_1:] + data_class_2[size_half_class_2:]
"""

print(f'Valid: {len(data_valid):6d}')
print(f'Test : {len(data_test):6d}')

MAP_LABEL_TRANSLATION = {
    0: 'sadness',
    1: 'joy',
    2: 'love',
    3: 'anger',
    4: 'fear',
    5: 'surprise',
}

def save_as_translations(original_save_path: Path, data_to_save: List[Dict]) -> None:
    file_name = 's2s-' + original_save_path.name
    file_path = original_save_path.parent / file_name

    print(f'Saving into: {file_path}')
    with open(file_path, 'wt') as f_write:
        for data_line in data_to_save:
            label = data_line['label']
            new_label = MAP_LABEL_TRANSLATION[label]
            data_line['label'] = new_label
            data_line_str = json.dumps(data_line)
            f_write.write(f'{data_line_str}\n')

for file_path, data_to_save in [(train_path, data_train), (valid_path, data_valid), (test_path, data_test)]:
  print(f'Saving into: {file_path}')
  with open(file_path, 'wt') as f_write:
    for data_line in data_to_save:
      data_line_str = json.dumps(data_line)
      f_write.write(f'{data_line_str}\n')
  
  save_as_translations(file_path, data_to_save)

WARNING:datasets.builder:No config specified, defaulting to: emotion/split
WARNING:datasets.builder:Found cached dataset emotion (/root/.cache/huggingface/datasets/emotion/split/1.0.0/cca5efe2dfeb58c1d098e0f9eeb200e9927d889b5a03c67097275dfb5fe463bd)

  0%|          | 0/3 [00:00<?, ?it/s]

mkdir: created directory 'data'
Train:  16000
Valid:   2000
Valid:   2000
Test :   2000
Saving into: data/train.json
Saving into: data/s2s-train.json
Saving into: data/valid.json
Saving into: data/s2s-valid.json
Saving into: data/test.json
Saving into: data/s2s-test.json


!head -n 4500 data/train.json > data/train-5k.json
!tail -n 2500 data/train.json >> data/train-5k.json
!wc -l data/train-5k.json

7000 data/train-5k.json

from pathlib import Path

for file_name in ["train", "valid", "test", "s2s-train", "s2s-valid", "s2s-test"]:
  print(f"=== {file_name} ===")
  all_text = Path(f"data/{file_name}.json").read_text().split('\n')
  text = all_text[:2500] + all_text[-2500:]
  Path(f"data/{file_name}-5k.json").write_text("\n".join(text))

=== train ===
=== valid ===
=== test ===
=== s2s-train ===
=== s2s-valid ===
=== s2s-test ===

import os

os.environ['TOKENIZERS_PARALLELISM'] = 'true'

!python run_glue.py \
--cache_dir .cache_training \
--model_name_or_path gpt2 \
--custom_model gpt2_hidden \
--freeze_weights \
--train_file data/s2s-train.json  \
--validation_file data/s2s-valid.json \
--test_file data/s2s-test.json \
--per_device_train_batch_size 24 \
--per_device_eval_batch_size 24 \
--do_train \
--do_eval \
--do_predict \
--max_seq_length 128 \
--learning_rate 2e-5 \
--num_train_epochs 5 \
--output_dir out/imdb-5k/gpt2

2023-02-12 22:00:15.880386: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2023-02-12 22:00:16.771169: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/lib64-nvidia
2023-02-12 22:00:16.771276: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/lib64-nvidia
2023-02-12 22:00:16.771294: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
WARNING:__main__:Process rank: -1, device: cuda:0, n_gpu: 1distributed training: False, 16-bits training: False
INFO:__main__:Training/evaluation parameters TrainingArguments(
_n_gpu=1,
adafactor=False,
adam_beta1=0.9,
adam_beta2=0.999,
adam_epsilon=1e-08,
auto_find_batch_size=False,
bf16=False,
bf16_full_eval=False,
data_seed=None,
dataloader_drop_last=False,
dataloader_num_workers=0,
dataloader_pin_memory=True,
ddp_bucket_cap_mb=None,
ddp_find_unused_parameters=None,
ddp_timeout=1800,
debug=[],
deepspeed=None,
disable_tqdm=False,
do_eval=True,
do_predict=True,
do_train=True,
eval_accumulation_steps=None,
eval_delay=0,
eval_steps=None,
evaluation_strategy=no,
fp16=False,
fp16_backend=auto,
fp16_full_eval=False,
fp16_opt_level=O1,
fsdp=[],
fsdp_min_num_params=0,
fsdp_transformer_layer_cls_to_wrap=None,
full_determinism=False,
gradient_accumulation_steps=1,
gradient_checkpointing=False,
greater_is_better=None,
group_by_length=False,
half_precision_backend=auto,
hub_model_id=None,
hub_private_repo=False,
hub_strategy=every_save,
hub_token=<HUB_TOKEN>,
ignore_data_skip=False,
include_inputs_for_metrics=False,
jit_mode_eval=False,
label_names=None,
label_smoothing_factor=0.0,
learning_rate=2e-05,
length_column_name=length,
load_best_model_at_end=False,
local_rank=-1,
log_level=passive,
log_level_replica=passive,
log_on_each_node=True,
logging_dir=out/imdb-5k/gpt2/runs/Feb12_22-00-19_506c7abe63fb,
logging_first_step=False,
logging_nan_inf_filter=True,
logging_steps=500,
logging_strategy=steps,
lr_scheduler_type=linear,
max_grad_norm=1.0,
max_steps=-1,
metric_for_best_model=None,
mp_parameters=,
no_cuda=False,
num_train_epochs=5.0,
optim=adamw_hf,
optim_args=None,
output_dir=out/imdb-5k/gpt2,
overwrite_output_dir=False,
past_index=-1,
per_device_eval_batch_size=24,
per_device_train_batch_size=24,
prediction_loss_only=False,
push_to_hub=False,
push_to_hub_model_id=None,
push_to_hub_organization=None,
push_to_hub_token=<PUSH_TO_HUB_TOKEN>,
ray_scope=last,
remove_unused_columns=True,
report_to=['tensorboard'],
resume_from_checkpoint=None,
run_name=out/imdb-5k/gpt2,
save_on_each_node=False,
save_steps=500,
save_strategy=steps,
save_total_limit=None,
seed=42,
sharded_ddp=[],
skip_memory_metrics=True,
tf32=None,
torch_compile=False,
torch_compile_backend=None,
torch_compile_mode=None,
torchdynamo=None,
tpu_metrics_debug=False,
tpu_num_cores=None,
use_ipex=False,
use_legacy_prediction_loop=False,
use_mps_device=False,
warmup_ratio=0.0,
warmup_steps=0,
weight_decay=0.0,
xpu_backend=None,
)
INFO:__main__:load a local file for train: data/s2s-train.json
INFO:__main__:load a local file for validation: data/s2s-valid.json
INFO:__main__:load a local file for test: data/s2s-test.json
WARNING:datasets.builder:Using custom data configuration default-623c8a7b15a2e58a
INFO:datasets.info:Loading Dataset Infos from /usr/local/lib/python3.8/dist-packages/datasets/packaged_modules/json
INFO:datasets.builder:Generating dataset json (/content/.cache_training/json/default-623c8a7b15a2e58a/0.0.0/0f7e3662623656454fcd2b650f34e886a7db4b9104504885bd462096cc7a9f51)
Downloading and preparing dataset json/default to /content/.cache_training/json/default-623c8a7b15a2e58a/0.0.0/0f7e3662623656454fcd2b650f34e886a7db4b9104504885bd462096cc7a9f51...
Downloading data files: 100% 3/3 [00:00<00:00, 14873.42it/s]
INFO:datasets.download.download_manager:Downloading took 0.0 min
INFO:datasets.download.download_manager:Checksum Computation took 0.0 min
Extracting data files: 100% 3/3 [00:00<00:00, 1763.55it/s]
INFO:datasets.utils.info_utils:Unable to verify checksums.
INFO:datasets.builder:Generating train split
INFO:datasets.builder:Generating validation split
INFO:datasets.builder:Generating test split
INFO:datasets.utils.info_utils:Unable to verify splits sizes.
Dataset json downloaded and prepared to /content/.cache_training/json/default-623c8a7b15a2e58a/0.0.0/0f7e3662623656454fcd2b650f34e886a7db4b9104504885bd462096cc7a9f51. Subsequent calls will reuse this data.
100% 3/3 [00:00<00:00, 1028.86it/s]
Downloading (…)lve/main/config.json: 100% 665/665 [00:00<00:00, 126kB/s]
[INFO|configuration_utils.py:660] 2023-02-12 22:00:20,342 >> loading configuration file config.json from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/config.json
[INFO|configuration_utils.py:712] 2023-02-12 22:00:20,343 >> Model config GPT2Config {
  "_name_or_path": "gpt2",
  "activation_function": "gelu_new",
  "architectures": [
    "GPT2LMHeadModel"
  ],
  "attn_pdrop": 0.1,
  "bos_token_id": 50256,
  "embd_pdrop": 0.1,
  "eos_token_id": 50256,
  "id2label": {
    "0": "LABEL_0",
    "1": "LABEL_1",
    "2": "LABEL_2",
    "3": "LABEL_3",
    "4": "LABEL_4",
    "5": "LABEL_5"
  },
  "initializer_range": 0.02,
  "label2id": {
    "LABEL_0": 0,
    "LABEL_1": 1,
    "LABEL_2": 2,
    "LABEL_3": 3,
    "LABEL_4": 4,
    "LABEL_5": 5
  },
  "layer_norm_epsilon": 1e-05,
  "model_type": "gpt2",
  "n_ctx": 1024,
  "n_embd": 768,
  "n_head": 12,
  "n_inner": null,
  "n_layer": 12,
  "n_positions": 1024,
  "reorder_and_upcast_attn": false,
  "resid_pdrop": 0.1,
  "scale_attn_by_inverse_layer_idx": false,
  "scale_attn_weights": true,
  "summary_activation": null,
  "summary_first_dropout": 0.1,
  "summary_proj_to_labels": true,
  "summary_type": "cls_index",
  "summary_use_proj": true,
  "task_specific_params": {
    "text-generation": {
      "do_sample": true,
      "max_length": 50
    }
  },
  "transformers_version": "4.26.1",
  "use_cache": true,
  "vocab_size": 50257
}

[INFO|tokenization_auto.py:458] 2023-02-12 22:00:20,434 >> Could not locate the tokenizer configuration file, will try to use the model config instead.
[INFO|configuration_utils.py:660] 2023-02-12 22:00:20,525 >> loading configuration file config.json from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/config.json
[INFO|configuration_utils.py:712] 2023-02-12 22:00:20,525 >> Model config GPT2Config {
  "_name_or_path": "gpt2",
  "activation_function": "gelu_new",
  "architectures": [
    "GPT2LMHeadModel"
  ],
  "attn_pdrop": 0.1,
  "bos_token_id": 50256,
  "embd_pdrop": 0.1,
  "eos_token_id": 50256,
  "initializer_range": 0.02,
  "layer_norm_epsilon": 1e-05,
  "model_type": "gpt2",
  "n_ctx": 1024,
  "n_embd": 768,
  "n_head": 12,
  "n_inner": null,
  "n_layer": 12,
  "n_positions": 1024,
  "reorder_and_upcast_attn": false,
  "resid_pdrop": 0.1,
  "scale_attn_by_inverse_layer_idx": false,
  "scale_attn_weights": true,
  "summary_activation": null,
  "summary_first_dropout": 0.1,
  "summary_proj_to_labels": true,
  "summary_type": "cls_index",
  "summary_use_proj": true,
  "task_specific_params": {
    "text-generation": {
      "do_sample": true,
      "max_length": 50
    }
  },
  "transformers_version": "4.26.1",
  "use_cache": true,
  "vocab_size": 50257
}

Downloading (…)olve/main/vocab.json: 100% 1.04M/1.04M [00:00<00:00, 9.19MB/s]
Downloading (…)olve/main/merges.txt: 100% 456k/456k [00:00<00:00, 4.93MB/s]
Downloading (…)/main/tokenizer.json: 100% 1.36M/1.36M [00:00<00:00, 11.9MB/s]
[INFO|tokenization_utils_base.py:1802] 2023-02-12 22:00:21,743 >> loading file vocab.json from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/vocab.json
[INFO|tokenization_utils_base.py:1802] 2023-02-12 22:00:21,743 >> loading file merges.txt from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/merges.txt
[INFO|tokenization_utils_base.py:1802] 2023-02-12 22:00:21,743 >> loading file tokenizer.json from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/tokenizer.json
[INFO|tokenization_utils_base.py:1802] 2023-02-12 22:00:21,744 >> loading file added_tokens.json from cache at None
[INFO|tokenization_utils_base.py:1802] 2023-02-12 22:00:21,744 >> loading file special_tokens_map.json from cache at None
[INFO|tokenization_utils_base.py:1802] 2023-02-12 22:00:21,744 >> loading file tokenizer_config.json from cache at None
[INFO|configuration_utils.py:660] 2023-02-12 22:00:21,744 >> loading configuration file config.json from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/config.json
[INFO|configuration_utils.py:712] 2023-02-12 22:00:21,745 >> Model config GPT2Config {
  "_name_or_path": "gpt2",
  "activation_function": "gelu_new",
  "architectures": [
    "GPT2LMHeadModel"
  ],
  "attn_pdrop": 0.1,
  "bos_token_id": 50256,
  "embd_pdrop": 0.1,
  "eos_token_id": 50256,
  "initializer_range": 0.02,
  "layer_norm_epsilon": 1e-05,
  "model_type": "gpt2",
  "n_ctx": 1024,
  "n_embd": 768,
  "n_head": 12,
  "n_inner": null,
  "n_layer": 12,
  "n_positions": 1024,
  "reorder_and_upcast_attn": false,
  "resid_pdrop": 0.1,
  "scale_attn_by_inverse_layer_idx": false,
  "scale_attn_weights": true,
  "summary_activation": null,
  "summary_first_dropout": 0.1,
  "summary_proj_to_labels": true,
  "summary_type": "cls_index",
  "summary_use_proj": true,
  "task_specific_params": {
    "text-generation": {
      "do_sample": true,
      "max_length": 50
    }
  },
  "transformers_version": "4.26.1",
  "use_cache": true,
  "vocab_size": 50257
}

INFO:__main__:Using hidden states in model: True
INFO:__main__:Using implementation from class: GPT2ForSequenceClassificationCustom
Downloading (…)"pytorch_model.bin";: 100% 548M/548M [00:05<00:00, 103MB/s]
[INFO|modeling_utils.py:2275] 2023-02-12 22:00:27,304 >> loading weights file pytorch_model.bin from cache at .cache_training/models--gpt2/snapshots/e7da7f221d5bf496a48136c0cd264e630fe9fcc8/pytorch_model.bin
[INFO|modeling_utils.py:2857] 2023-02-12 22:00:30,150 >> All model checkpoint weights were used when initializing GPT2ForSequenceClassificationCustom.

[WARNING|modeling_utils.py:2859] 2023-02-12 22:00:30,150 >> Some weights of GPT2ForSequenceClassificationCustom were not initialized from the model checkpoint at gpt2 and are newly initialized: ['score.dense_1_hidden.weight', 'score.dense_2.weight', 'score.out_proj.weight', 'score.dense_2.bias', 'score.dense_1_hidden.bias', 'score.dense_1_input.bias', 'score.dense_1_input.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
INFO:__main__:Freezing encoder weights
INFO:__main__:Freezing layer 1
INFO:__main__:Freezing layer 2
INFO:__main__:Freezing layer 3
INFO:__main__:Freezing layer 4
INFO:__main__:Freezing layer 5
INFO:__main__:Freezing layer 6
INFO:__main__:Freezing layer 7
INFO:__main__:Freezing layer 8
INFO:__main__:Freezing layer 9
INFO:__main__:Freezing layer 10
INFO:__main__:Freezing layer 11
INFO:__main__:Freezing layer 12
INFO:__main__:Freezing layer 13
INFO:__main__:Freezing layer 14
INFO:__main__:Freezing layer 15
INFO:__main__:Freezing layer 16
INFO:__main__:Freezing layer 17
INFO:__main__:Freezing layer 18
INFO:__main__:Freezing layer 19
INFO:__main__:Freezing layer 20
INFO:__main__:Freezing layer 21
INFO:__main__:Freezing layer 22
INFO:__main__:Freezing layer 23
INFO:__main__:Freezing layer 24
INFO:__main__:Freezing layer 25
INFO:__main__:Freezing layer 26
INFO:__main__:Freezing layer 27
INFO:__main__:Freezing layer 28
INFO:__main__:Freezing layer 29
INFO:__main__:Freezing layer 30
INFO:__main__:Freezing layer 31
INFO:__main__:Freezing layer 32
INFO:__main__:Freezing layer 33
INFO:__main__:Freezing layer 34
INFO:__main__:Freezing layer 35
INFO:__main__:Freezing layer 36
INFO:__main__:Freezing layer 37
INFO:__main__:Freezing layer 38
INFO:__main__:Freezing layer 39
INFO:__main__:Freezing layer 40
INFO:__main__:Ignoring layer 41
INFO:__main__:Ignoring layer 42
INFO:__main__:Ignoring layer 43
INFO:__main__:Ignoring layer 44
INFO:__main__:Ignoring layer 45
INFO:__main__:Ignoring layer 46
INFO:__main__:Ignoring layer 47
INFO:__main__:Ignoring layer 48
INFO:__main__:Ignoring layer 49
INFO:__main__:Ignoring layer 50
INFO:__main__:Ignoring layer 51
INFO:__main__:Ignoring layer 52
INFO:__main__:Ignoring layer 53
INFO:__main__:Ignoring layer 54
INFO:__main__:Ignoring layer 55
INFO:__main__:Ignoring layer 56
INFO:__main__:Ignoring layer 57
INFO:__main__:Ignoring layer 58
INFO:__main__:Ignoring layer 59
INFO:__main__:Ignoring layer 60
INFO:__main__:Ignoring layer 61
INFO:__main__:Ignoring layer 62
INFO:__main__:Ignoring layer 63
INFO:__main__:Ignoring layer 64
INFO:__main__:Ignoring layer 65
INFO:__main__:Ignoring layer 66
INFO:__main__:Ignoring layer 67
INFO:__main__:Ignoring layer 68
INFO:__main__:Ignoring layer 69
INFO:__main__:Ignoring layer 70
INFO:__main__:Ignoring layer 71
INFO:__main__:Ignoring layer 72
INFO:__main__:Ignoring layer 73
INFO:__main__:Ignoring layer 74
INFO:__main__:Ignoring layer 75
INFO:__main__:Ignoring layer 76
INFO:__main__:Ignoring layer 77
INFO:__main__:Ignoring layer 78
INFO:__main__:Ignoring layer 79
INFO:__main__:Ignoring layer 80
INFO:__main__:Ignoring layer 81
INFO:__main__:Ignoring layer 82
INFO:__main__:Ignoring layer 83
INFO:__main__:Ignoring layer 84
INFO:__main__:Ignoring layer 85
INFO:__main__:Ignoring layer 86
INFO:__main__:Ignoring layer 87
INFO:__main__:Ignoring layer 88
INFO:__main__:Ignoring layer 89
INFO:__main__:Ignoring layer 90
INFO:__main__:Ignoring layer 91
INFO:__main__:Ignoring layer 92
INFO:__main__:Ignoring layer 93
INFO:__main__:Ignoring layer 94
INFO:__main__:Ignoring layer 95
INFO:__main__:Ignoring layer 96
INFO:__main__:Ignoring layer 97
INFO:__main__:Ignoring layer 98
INFO:__main__:Ignoring layer 99
INFO:__main__:Ignoring layer 100
INFO:__main__:Ignoring layer 101
INFO:__main__:Ignoring layer 102
INFO:__main__:Ignoring layer 103
INFO:__main__:Ignoring layer 104
INFO:__main__:Ignoring layer 105
INFO:__main__:Ignoring layer 106
INFO:__main__:Ignoring layer 107
INFO:__main__:Ignoring layer 108
INFO:__main__:Ignoring layer 109
INFO:__main__:Ignoring layer 110
INFO:__main__:Ignoring layer 111
INFO:__main__:Ignoring layer 112
INFO:__main__:Ignoring layer 113
INFO:__main__:Ignoring layer 114
INFO:__main__:Ignoring layer 115
INFO:__main__:Ignoring layer 116
INFO:__main__:Ignoring layer 117
INFO:__main__:Ignoring layer 118
INFO:__main__:Ignoring layer 119
INFO:__main__:Ignoring layer 120
INFO:__main__:Ignoring layer 121
INFO:__main__:Ignoring layer 122
INFO:__main__:Ignoring layer 123
INFO:__main__:Ignoring layer 124
INFO:__main__:Ignoring layer 125
INFO:__main__:Ignoring layer 126
INFO:__main__:Ignoring layer 127
INFO:__main__:Ignoring layer 128
INFO:__main__:Ignoring layer 129
INFO:__main__:Ignoring layer 130
INFO:__main__:Ignoring layer 131
INFO:__main__:Ignoring layer 132
INFO:__main__:Ignoring layer 133
INFO:__main__:Ignoring layer 134
INFO:__main__:Ignoring layer 135
INFO:__main__:Ignoring layer 136
INFO:__main__:Ignoring layer 137
INFO:__main__:Ignoring layer 138
INFO:__main__:Ignoring layer 139
INFO:__main__:Ignoring layer 140
INFO:__main__:Ignoring layer 141
INFO:__main__:Ignoring layer 142
INFO:__main__:Ignoring layer 143
INFO:__main__:Ignoring layer 144
INFO:__main__:Ignoring layer 145
INFO:__main__:Ignoring layer 146
INFO:__main__:Ignoring layer 147
INFO:__main__:Ignoring layer 148
INFO:__main__:Ignoring layer 149
INFO:__main__:Ignoring layer 150
INFO:__main__:Ignoring layer 151
INFO:__main__:Ignoring layer 152
INFO:__main__:Ignoring layer 153
INFO:__main__:Ignoring layer 154
INFO:__main__:Ignoring layer 155
[ERROR|tokenization_utils_base.py:1042] 2023-02-12 22:00:30,162 >> Using pad_token, but it is not set yet.
INFO:__main__:Set PAD token to EOS: <|endoftext|>
Running tokenizer on dataset:   0% 0/16 [00:00<?, ?ba/s]INFO:datasets.arrow_dataset:Caching processed dataset at /content/.cache_training/json/default-623c8a7b15a2e58a/0.0.0/0f7e3662623656454fcd2b650f34e886a7db4b9104504885bd462096cc7a9f51/cache-c8be49ce7f75c3ae.arrow
Running tokenizer on dataset: 100% 16/16 [00:01<00:00, 11.14ba/s]
Running tokenizer on dataset:   0% 0/2 [00:00<?, ?ba/s]INFO:datasets.arrow_dataset:Caching processed dataset at /content/.cache_training/json/default-623c8a7b15a2e58a/0.0.0/0f7e3662623656454fcd2b650f34e886a7db4b9104504885bd462096cc7a9f51/cache-01058c80c2124fd4.arrow
Running tokenizer on dataset: 100% 2/2 [00:00<00:00,  6.04ba/s]
Running tokenizer on dataset:   0% 0/2 [00:00<?, ?ba/s]INFO:datasets.arrow_dataset:Caching processed dataset at /content/.cache_training/json/default-623c8a7b15a2e58a/0.0.0/0f7e3662623656454fcd2b650f34e886a7db4b9104504885bd462096cc7a9f51/cache-53937ce94683c2e2.arrow
Running tokenizer on dataset: 100% 2/2 [00:00<00:00, 11.40ba/s]
INFO:__main__:Sample 10476 of the training set: {'label': 4, 'text': 'i do find new friends i m going to try extra hard to make them stay and if i decide that i don t want to feel hurt again and just ride out the last year of school on my own i m going to have to try extra hard not to care what people think of me being a loner', 'input_ids': [72, 466, 1064, 649, 2460, 1312, 285, 1016, 284, 1949, 3131, 1327, 284, 787, 606, 2652, 290, 611, 1312, 5409, 326, 1312, 836, 256, 765, 284, 1254, 5938, 757, 290, 655, 6594, 503, 262, 938, 614, 286, 1524, 319, 616, 898, 1312, 285, 1016, 284, 423, 284, 1949, 3131, 1327, 407, 284, 1337, 644, 661, 892, 286, 502, 852, 257, 300, 14491, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256], 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]}.
INFO:__main__:Sample 1824 of the training set: {'label': 2, 'text': 'i asked them to join me in creating a world where all year old girls could grow up feeling hopeful and powerful', 'input_ids': [72, 1965, 606, 284, 4654, 502, 287, 4441, 257, 995, 810, 477, 614, 1468, 4813, 714, 1663, 510, 4203, 17836, 290, 3665, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256], 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]}.
INFO:__main__:Sample 409 of the training set: {'label': 3, 'text': 'i feel when you are a caring person you attract other caring people into your life', 'input_ids': [72, 1254, 618, 345, 389, 257, 18088, 1048, 345, 4729, 584, 18088, 661, 656, 534, 1204, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256], 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]}.
Downloading builder script: 100% 4.20k/4.20k [00:00<00:00, 4.82MB/s]
[INFO|trainer.py:710] 2023-02-12 22:00:38,984 >> The following columns in the training set don't have a corresponding argument in `GPT2ForSequenceClassificationCustom.forward` and have been ignored: text. If text are not expected by `GPT2ForSequenceClassificationCustom.forward`,  you can safely ignore this message.
/usr/local/lib/python3.8/dist-packages/transformers/optimization.py:306: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning
  warnings.warn(
[INFO|trainer.py:1650] 2023-02-12 22:00:38,993 >> ***** Running training *****
[INFO|trainer.py:1651] 2023-02-12 22:00:38,993 >>   Num examples = 16000
[INFO|trainer.py:1652] 2023-02-12 22:00:38,993 >>   Num Epochs = 5
[INFO|trainer.py:1653] 2023-02-12 22:00:38,993 >>   Instantaneous batch size per device = 24
[INFO|trainer.py:1654] 2023-02-12 22:00:38,993 >>   Total train batch size (w. parallel, distributed & accumulation) = 24
[INFO|trainer.py:1655] 2023-02-12 22:00:38,993 >>   Gradient Accumulation steps = 1
[INFO|trainer.py:1656] 2023-02-12 22:00:38,993 >>   Total optimization steps = 3335
[INFO|trainer.py:1657] 2023-02-12 22:00:38,994 >>   Number of trainable parameters = 68517888
{'loss': 1.0593, 'learning_rate': 1.7001499250374815e-05, 'epoch': 0.75}
 15% 500/3335 [04:16<24:04,  1.96it/s][INFO|trainer.py:2709] 2023-02-12 22:04:55,709 >> Saving model checkpoint to out/imdb-5k/gpt2/checkpoint-500
[INFO|configuration_utils.py:453] 2023-02-12 22:04:55,710 >> Configuration saved in out/imdb-5k/gpt2/checkpoint-500/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:04:57,444 >> Model weights saved in out/imdb-5k/gpt2/checkpoint-500/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:04:57,444 >> tokenizer config file saved in out/imdb-5k/gpt2/checkpoint-500/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:04:57,444 >> Special tokens file saved in out/imdb-5k/gpt2/checkpoint-500/special_tokens_map.json
{'loss': 0.3829, 'learning_rate': 1.4002998500749626e-05, 'epoch': 1.5}
 30% 1000/3335 [08:36<19:51,  1.96it/s][INFO|trainer.py:2709] 2023-02-12 22:09:15,813 >> Saving model checkpoint to out/imdb-5k/gpt2/checkpoint-1000
[INFO|configuration_utils.py:453] 2023-02-12 22:09:15,814 >> Configuration saved in out/imdb-5k/gpt2/checkpoint-1000/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:09:17,628 >> Model weights saved in out/imdb-5k/gpt2/checkpoint-1000/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:09:17,629 >> tokenizer config file saved in out/imdb-5k/gpt2/checkpoint-1000/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:09:17,630 >> Special tokens file saved in out/imdb-5k/gpt2/checkpoint-1000/special_tokens_map.json
{'loss': 0.256, 'learning_rate': 1.100449775112444e-05, 'epoch': 2.25}
 45% 1500/3335 [12:56<15:43,  1.95it/s][INFO|trainer.py:2709] 2023-02-12 22:13:36,008 >> Saving model checkpoint to out/imdb-5k/gpt2/checkpoint-1500
[INFO|configuration_utils.py:453] 2023-02-12 22:13:36,009 >> Configuration saved in out/imdb-5k/gpt2/checkpoint-1500/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:13:37,703 >> Model weights saved in out/imdb-5k/gpt2/checkpoint-1500/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:13:37,704 >> tokenizer config file saved in out/imdb-5k/gpt2/checkpoint-1500/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:13:37,704 >> Special tokens file saved in out/imdb-5k/gpt2/checkpoint-1500/special_tokens_map.json
{'loss': 0.2101, 'learning_rate': 8.005997001499251e-06, 'epoch': 3.0}
 60% 2000/3335 [17:17<11:23,  1.95it/s][INFO|trainer.py:2709] 2023-02-12 22:17:56,308 >> Saving model checkpoint to out/imdb-5k/gpt2/checkpoint-2000
[INFO|configuration_utils.py:453] 2023-02-12 22:17:56,309 >> Configuration saved in out/imdb-5k/gpt2/checkpoint-2000/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:17:58,005 >> Model weights saved in out/imdb-5k/gpt2/checkpoint-2000/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:17:58,006 >> tokenizer config file saved in out/imdb-5k/gpt2/checkpoint-2000/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:17:58,006 >> Special tokens file saved in out/imdb-5k/gpt2/checkpoint-2000/special_tokens_map.json
{'loss': 0.17, 'learning_rate': 5.0074962518740634e-06, 'epoch': 3.75}
 75% 2500/3335 [21:37<07:05,  1.96it/s][INFO|trainer.py:2709] 2023-02-12 22:22:16,686 >> Saving model checkpoint to out/imdb-5k/gpt2/checkpoint-2500
[INFO|configuration_utils.py:453] 2023-02-12 22:22:16,687 >> Configuration saved in out/imdb-5k/gpt2/checkpoint-2500/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:22:18,356 >> Model weights saved in out/imdb-5k/gpt2/checkpoint-2500/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:22:18,357 >> tokenizer config file saved in out/imdb-5k/gpt2/checkpoint-2500/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:22:18,357 >> Special tokens file saved in out/imdb-5k/gpt2/checkpoint-2500/special_tokens_map.json
{'loss': 0.1569, 'learning_rate': 2.008995502248876e-06, 'epoch': 4.5}
 90% 3000/3335 [25:57<02:51,  1.95it/s][INFO|trainer.py:2709] 2023-02-12 22:26:36,938 >> Saving model checkpoint to out/imdb-5k/gpt2/checkpoint-3000
[INFO|configuration_utils.py:453] 2023-02-12 22:26:36,939 >> Configuration saved in out/imdb-5k/gpt2/checkpoint-3000/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:26:38,608 >> Model weights saved in out/imdb-5k/gpt2/checkpoint-3000/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:26:38,608 >> tokenizer config file saved in out/imdb-5k/gpt2/checkpoint-3000/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:26:38,608 >> Special tokens file saved in out/imdb-5k/gpt2/checkpoint-3000/special_tokens_map.json
100% 3335/3335 [28:53<00:00,  2.15it/s][INFO|trainer.py:1901] 2023-02-12 22:29:32,259 >> 

Training completed. Do not forget to share your model on huggingface.co/models =)


{'train_runtime': 1733.281, 'train_samples_per_second': 46.155, 'train_steps_per_second': 1.924, 'train_loss': 0.35007504373118614, 'epoch': 5.0}
100% 3335/3335 [28:53<00:00,  1.92it/s]
[INFO|trainer.py:2709] 2023-02-12 22:29:32,277 >> Saving model checkpoint to out/imdb-5k/gpt2
[INFO|configuration_utils.py:453] 2023-02-12 22:29:32,278 >> Configuration saved in out/imdb-5k/gpt2/config.json
[INFO|modeling_utils.py:1704] 2023-02-12 22:29:33,934 >> Model weights saved in out/imdb-5k/gpt2/pytorch_model.bin
[INFO|tokenization_utils_base.py:2160] 2023-02-12 22:29:33,934 >> tokenizer config file saved in out/imdb-5k/gpt2/tokenizer_config.json
[INFO|tokenization_utils_base.py:2167] 2023-02-12 22:29:33,934 >> Special tokens file saved in out/imdb-5k/gpt2/special_tokens_map.json
***** train metrics *****
  epoch                    =        5.0
  train_loss               =     0.3501
  train_runtime            = 0:28:53.28
  train_samples            =      16000
  train_samples_per_second =     46.155
  train_steps_per_second   =      1.924
INFO:__main__:*** Evaluate ***
[INFO|trainer.py:710] 2023-02-12 22:29:34,047 >> The following columns in the evaluation set don't have a corresponding argument in `GPT2ForSequenceClassificationCustom.forward` and have been ignored: text. If text are not expected by `GPT2ForSequenceClassificationCustom.forward`,  you can safely ignore this message.
[INFO|trainer.py:2964] 2023-02-12 22:29:34,108 >> ***** Running Evaluation *****
[INFO|trainer.py:2966] 2023-02-12 22:29:34,108 >>   Num examples = 2000
[INFO|trainer.py:2969] 2023-02-12 22:29:34,108 >>   Batch size = 24
100% 84/84 [00:17<00:00,  4.83it/s]
***** eval metrics *****
  epoch                   =        5.0
  eval_accuracy           =       0.93
  eval_loss               =     0.1531
  eval_runtime            = 0:00:17.72
  eval_samples            =       2000
  eval_samples_per_second =    112.855
  eval_steps_per_second   =       4.74
INFO:__main__:*** Predict ***
[INFO|trainer.py:710] 2023-02-12 22:29:51,834 >> The following columns in the test set don't have a corresponding argument in `GPT2ForSequenceClassificationCustom.forward` and have been ignored: text. If text are not expected by `GPT2ForSequenceClassificationCustom.forward`,  you can safely ignore this message.
[INFO|trainer.py:2964] 2023-02-12 22:29:51,836 >> ***** Running Prediction *****
[INFO|trainer.py:2966] 2023-02-12 22:29:51,836 >>   Num examples = 2000
[INFO|trainer.py:2969] 2023-02-12 22:29:51,836 >>   Batch size = 24
100% 84/84 [00:17<00:00,  4.81it/s]
INFO:__main__:***** Predict results None *****
[INFO|modelcard.py:449] 2023-02-12 22:30:09,657 >> Dropping the following result as it does not have all the necessary fields:
{'task': {'name': 'Text Classification', 'type': 'text-classification'}, 'metrics': [{'name': 'Accuracy', 'type': 'accuracy', 'value': 0.9300000071525574}]}

73 KiB Raw Blame History Unescape Escape

73 KiB

Raw Blame History