Release: QnA Module V1.0

.gitignore

@@ -0,0 +1,138 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyderworkspace
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+datasets/
+
+*.bin

Dockerfile

@@ -0,0 +1,36 @@
+ARG PYTORCH="2.2.0"
+ARG CUDA="12.1"
+ARG CUDNN="8"
+
+FROM pytorch/pytorch:${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel
+
+ENV TORCH_CUDA_ARCH_LIST="6.0 6.1 7.0+PTX"
+ENV TORCH_NVCC_FLAGS="-Xfatbin -compress-all"
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Update package list and install software-properties-common
+RUN apt update && apt install -y software-properties-common
+
+# Add deadsnakes PPA for Python 3.9
+RUN add-apt-repository ppa:deadsnakes/ppa
+RUN apt update
+
+RUN apt install -y git vim libgl1-mesa-glx libglib2.0-0 ninja-build libsm6 libxrender-dev libxext6 libgl1-mesa-glx python3.9 python3.9-dev python3.9-distutils wget net-tools zip unzip
+RUN apt-get clean && rm -rf /var/lib/apt/lists/*
+
+# Set Python 3.9 as the default python version
+RUN update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.9 1
+RUN update-alternatives --install /usr/bin/python python /usr/bin/python3.9 1
+
+# Install pip for Python 3.9
+RUN wget https://bootstrap.pypa.io/get-pip.py
+RUN python3.9 get-pip.py
+
+# Install Python Library
+RUN pip install --upgrade pip
+RUN pip install --upgrade setuptools
+RUN pip install Pillow Flask Flask-Cors tensorflow>=2.0.0 transformers
+
+# Set the default command to run when the container starts
+WORKDIR /app
+

README.md

@@ -0,0 +1,5 @@
+![thumbnail](./resource/thumb.png)
+
+> 릴리즈 버전 [Release v1.0](https://github.com/DGU-ITRC/PLASS_QNA/releases/tag/v1.6)
+
+TBU

model.py

@@ -0,0 +1,54 @@
+import argparse
+import os
+import torch
+import random
+import numpy as np
+from transformers import DistilBertTokenizerFast
+from transformers import DistilBertForQuestionAnswering
+
+def init_args(context, question):
+    if context is None or question is None:
+        context = "Stephen Silvagni (born 31 May 1967) is a former Australian rules footballer for the Carlton Football Club." 
+        question = "What was the name of Stephen Silvagni's team?"
+    args = {
+        'context': context,
+        'question': question,
+        'seed': 42,
+        'save_dir': 'save/baseline-01',
+    }
+    return args
+
+def set_seed(seed):
+    random.seed(seed)
+    np.random.seed(seed)
+    torch.manual_seed(seed)
+    torch.cuda.manual_seed_all(seed)
+
+def predict(context, question):
+    args = init_args(context, question)
+    set_seed(args['seed'])
+    checkpoint_path = os.path.join(args['save_dir'], 'checkpoint')
+    model = DistilBertForQuestionAnswering.from_pretrained(checkpoint_path)
+    tokenizer = DistilBertTokenizerFast.from_pretrained('distilbert-base-uncased')
+    args['device'] = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu') 
+    os.environ["TOKENIZERS_PARALLELISM"] = "false"
+    context = args['context']
+    question = args['question']
+    inputs = tokenizer(question, context, return_tensors="pt")
+    with torch.no_grad():
+        outputs = model(**inputs, )
+    start_index = outputs.start_logits.argmax()
+    end_index = outputs.end_logits.argmax()
+    start_idx = int(start_index.numpy())
+    end_idx = int(end_index.numpy())   
+    predict_tokens = inputs.input_ids[0,start_index:end_index+1]
+    predict_answer = tokenizer.decode(predict_tokens)
+    result = {'context': context, 'question': question, 'start_idx': start_idx, 'end_idx': end_idx, 'answer': predict_answer}
+    print(result)
+    return result
+
+
+if __name__ == '__main__':
+    context = "Stephen Silvagni (born 31 May 1967) is a former Australian rules footballer for the Carlton Football Club." 
+    question = "What was the name of Stephen Silvagni's team?"
+    predict(context, question)

save/baseline-01/baseline-01/log_train.txt

@@ -0,0 +1,26 @@
+[06.08.23 20:20:39] Args: {
+    "batch_size": 16,
+    "do_eval": false,
+    "do_test": false,
+    "do_train": true,
+    "eval": false,
+    "eval_datasets": "race,relation_extraction,duorc",
+    "eval_dir": "datasets/oodomain_test",
+    "eval_every": 5000,
+    "lr": 3e-05,
+    "num_epochs": 3,
+    "num_visuals": 10,
+    "recompute_features": false,
+    "run_name": "baseline",
+    "save_dir": "save/baseline-01/baseline-01",
+    "seed": 42,
+    "sub_file": "",
+    "train": false,
+    "train_datasets": "squad,nat_questions,newsqa",
+    "train_dir": "datasets/indomain_train",
+    "val_dir": "datasets/indomain_val",
+    "visualize_predictions": false
+}
+[06.08.23 20:20:39] Preparing Training Data...
+[06.08.23 20:21:46] Preparing Validation Data...
+[06.08.23 20:22:00] Epoch: 0

save/baseline-01/baseline-02/log_train.txt

@@ -0,0 +1,26 @@
+[06.08.23 20:33:01] Args: {
+    "batch_size": 16,
+    "do_eval": false,
+    "do_test": false,
+    "do_train": true,
+    "eval": false,
+    "eval_datasets": "race,relation_extraction,duorc",
+    "eval_dir": "datasets/oodomain_test",
+    "eval_every": 5000,
+    "lr": 3e-05,
+    "num_epochs": 3,
+    "num_visuals": 10,
+    "recompute_features": false,
+    "run_name": "baseline",
+    "save_dir": "save/baseline-01/baseline-02",
+    "seed": 42,
+    "sub_file": "",
+    "train": false,
+    "train_datasets": "squad,nat_questions,newsqa",
+    "train_dir": "datasets/indomain_train",
+    "val_dir": "datasets/indomain_val",
+    "visualize_predictions": false
+}
+[06.08.23 20:33:01] Preparing Training Data...
+[06.08.23 20:34:05] Preparing Validation Data...
+[06.08.23 20:34:18] Epoch: 0

save/baseline-01/baseline-03/log_train.txt

@@ -0,0 +1,26 @@
+[06.08.23 20:35:04] Args: {
+    "batch_size": 16,
+    "do_eval": false,
+    "do_test": false,
+    "do_train": true,
+    "eval": false,
+    "eval_datasets": "race,relation_extraction,duorc",
+    "eval_dir": "datasets/oodomain_test",
+    "eval_every": 5000,
+    "lr": 3e-05,
+    "num_epochs": 3,
+    "num_visuals": 10,
+    "recompute_features": false,
+    "run_name": "baseline",
+    "save_dir": "save/baseline-01/baseline-03",
+    "seed": 42,
+    "sub_file": "",
+    "train": false,
+    "train_datasets": "squad,nat_questions,newsqa",
+    "train_dir": "datasets/indomain_train",
+    "val_dir": "datasets/indomain_val",
+    "visualize_predictions": false
+}
+[06.08.23 20:35:04] Preparing Training Data...
+[06.08.23 20:36:06] Preparing Validation Data...
+[06.08.23 20:36:19] Epoch: 0

save/baseline-01/baseline-04/log_train.txt

@@ -0,0 +1,27 @@
+[06.08.23 20:40:51] Args: {
+    "batch_size": 16,
+    "do_eval": false,
+    "do_test": false,
+    "do_train": true,
+    "eval": false,
+    "eval_datasets": "race,relation_extraction,duorc",
+    "eval_dir": "datasets/oodomain_test",
+    "eval_every": 5000,
+    "lr": 3e-05,
+    "num_epochs": 3,
+    "num_visuals": 10,
+    "recompute_features": false,
+    "run_name": "baseline",
+    "save_dir": "save/baseline-01/baseline-04",
+    "seed": 42,
+    "sub_file": "",
+    "train": false,
+    "train_datasets": "squad,nat_questions,newsqa",
+    "train_dir": "datasets/indomain_train",
+    "val_dir": "datasets/indomain_val",
+    "visualize_predictions": false
+}
+[06.08.23 20:40:51] Preparing Training Data...
+[06.08.23 20:42:57] Preparing Validation Data...
+[06.08.23 20:43:16] Epoch: 0
+[06.08.23 20:43:17] Evaluating at step 0...

save/baseline-01/checkpoint/config.json

@@ -0,0 +1,23 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForQuestionAnswering"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "pad_token_id": 0,
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "transformers_version": "4.2.2",
+  "vocab_size": 30522
+}

save/baseline-01/log_test.txt

@@ -0,0 +1,8 @@
+[06.09.23 01:22:42] Eval F1: -1.00, EM: -1.00
+[06.09.23 01:22:42] Writing submission file to save/baseline-01/test_mtl_submission.csv...
+[06.12.23 16:54:13] Eval F1: -1.00, EM: -1.00
+[06.12.23 16:54:13] Writing submission file to save/baseline-01/test_mtl_submission.csv...
+[06.13.23 17:38:30] Eval F1: -1.00, EM: -1.00
+[06.13.23 17:38:30] Writing submission file to save/baseline-01/test_mtl_submission.csv...
+[06.15.23 20:54:55] Eval F1: -1.00, EM: -1.00
+[06.15.23 20:54:55] Writing submission file to save/baseline-01/test_mtl_submission.csv...