Initial commit for BERT

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.csv filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,160 @@
+# 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
+MANIFEST
+
+# 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
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .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
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__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
+.spyproject
+
+# 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/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/

data_1/Fake.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bebf8bcfe95678bf2c732bf413a2ce5f621af0102c82bf08083b2e5d3c693d0c
+size 62789876

data_1/True.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba0844414a65dc6ae7402b8eee5306da24b6b56488d6767135af466c7dcb2775
+size 53582940

data_2/WELFake_Dataset.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:665331424230fc452e9482c3547a6a199a2c29745ade8d236950d1d105223773
+size 245086152

data_3/news_articles.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53855240e9036a7d6c204e72bd0fa9d37a10f8e1bd2b2fdf34b962569ef271c6
+size 10969548

inference_main.py

@@ -0,0 +1,116 @@
+import torch
+from transformers import (
+    DistilBertTokenizer,
+    DistilBertForSequenceClassification,
+    Trainer,
+)
+from datasets import Dataset, load_from_disk
+import pandas as pd
+from sklearn.metrics import accuracy_score, f1_score, roc_auc_score
+from functools import partial
+import os
+
+version = 3
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+print("Device:", device)
+
+
+if __name__ == "__main__":
+    # Load the tokenizer
+    tokenizer = DistilBertTokenizer.from_pretrained("distilbert-base-uncased")
+
+    cleaned_path = (
+        f"./output/version_{version}/tokenized_data_{version}/inference_{version}"
+    )
+
+    # Load the model
+    model = DistilBertForSequenceClassification.from_pretrained(
+        f"./output/version_{version}/best_model_{version}"
+    )
+    model.to(device)
+
+    # Load and prepare the new dataset
+    if os.path.exists(cleaned_path):
+        print("Loading dataset...")
+        dataset = load_from_disk(cleaned_path)
+    else:
+        print("No dataset found. Loading and preparing dataset now...")
+        # # Load the datasets
+        # true_news = pd.read_csv("data_1/True.csv")
+        # fake_news = pd.read_csv("data_1/Fake.csv")
+
+        # # Add labels
+        # true_news["label"] = 1
+        # fake_news["label"] = 0
+
+        # # Combine the datasets
+        # df = pd.concat([true_news, fake_news], ignore_index=True)
+
+        df = pd.read_csv("./data_3/news_articles.csv")
+        df.drop(
+            columns=[
+                "author",
+                "published",
+                "site_url",
+                "main_img_url",
+                "type",
+                "text_without_stopwords",
+                "title_without_stopwords",
+                "hasImage",
+            ],
+            inplace=True,
+        )
+        # Map Real to 1 and Fake to 0
+        df["label"] = df["label"].map({"Real": 1, "Fake": 0})
+        df = df[df["label"].isin([1, 0])]
+
+        # Drop rows where the language is not 'english'
+        df = df[df["language"] == "english"]
+        df.drop(columns=["language"], inplace=True)
+
+        # Convert "no title" to empty string
+        df["title"] = df["title"].apply(lambda x: "" if x == "no title" else x)
+
+        df.dropna(inplace=True)
+
+        # Remove patterns like "COUNTRY or STATE NAME (Reuters) -"
+        df["text"] = df["text"].str.replace(
+            r"(\b[A-Z]{2,}(?:\s[A-Z]{2,})*\s\(Reuters\)\s-|\(Reuters\))", "", regex=True
+        )
+
+        # Remove patterns like "Featured image via author name / image place"
+        df["text"] = df["text"].str.replace(
+            r"Featured image via .+?\.($|\s)", "", regex=True
+        )
+
+        df["text"] = df["title"] + " " + df["text"]
+        df = df[["text", "label"]]
+        df["label"] = df["label"].astype(int)
+
+        dataset = Dataset.from_pandas(df)
+
+        def tokenize(tokenizer, examples):
+            return tokenizer(
+                examples["text"], padding=True, truncation=True, max_length=512
+            )
+
+        # Use partial to create a new function that has tokenizer as its first argument
+        tokenize_with_tokenizer = partial(tokenize, tokenizer)
+
+        # Use tokenize_with_tokenizer in the map function
+        dataset = dataset.map(tokenize_with_tokenizer, batched=True, num_proc=8)
+        dataset.save_to_disk(cleaned_path)
+
+    dataset.set_format("torch", columns=["input_ids", "attention_mask", "label"])
+    print("Finished tokenizing.")
+
+    trainer = Trainer(model=model)
+    predictions = trainer.predict(dataset)
+    accuracy = accuracy_score(predictions.label_ids, predictions.predictions.argmax(-1))
+    f1 = f1_score(predictions.label_ids, predictions.predictions.argmax(-1))
+    auc_roc = roc_auc_score(predictions.label_ids, predictions.predictions.argmax(-1))
+
+    print(f"Accuracy: {accuracy}")
+    print(f"F1 Score: {f1}")
+    print(f"AUC-ROC: {auc_roc}")

logs/version_1/events.out.tfevents.1703257066.DESKTOP-GO6MFA2.30996.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1dbf89aad28f266f87bfa73a16bc3abdc943d37d52648ce19e8794a98f23529
+size 4242

logs/version_1/events.out.tfevents.1703257151.DESKTOP-GO6MFA2.6540.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06d1e1df6a205c5d5cd7bdd500e3e58fe9047b6df4b1caa5dcfefa5851baf825
+size 4857

logs/version_1/events.out.tfevents.1703295125.DESKTOP-GO6MFA2.19276.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b08b981a25e1595659c28689c044b14ae3082d01b210244b05294dfedab8abc6
+size 110669

logs/version_2/events.out.tfevents.1703301313.DESKTOP-GO6MFA2.21900.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb80af9b9f1369ce1da98450c4b69a226895d778c82505500527c8687ab9da3f
+size 216859

logs/version_3/events.out.tfevents.1703420244.DESKTOP-GO6MFA2.19124.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcb41e4b83c630f9095c66046a5235182cdd0d897681efd81b7e27c46b81ea6e
+size 51381

output/version_1/checkpoint-2683/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "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,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "vocab_size": 30522
+}

output/version_1/checkpoint-2683/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14622ad5fea653baca561acc5e78d4bb3768ba9c86eed9209b04c707443e0bc5
+size 267832560

output/version_1/checkpoint-2683/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee5cfd3f5a843f9adb86791abd2aa8168107190ae28e2d5954d58643d3802e65
+size 535727290

output/version_1/checkpoint-2683/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0d8428d28280b58ca6a556a9e67c2242441cdf14b3e162d276775ac31e158e1
+size 14244

output/version_1/checkpoint-2683/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7179810020251db8397847c284285a5c29a8d3d69ce0b35c2766110fa44e2a14
+size 1064

output/version_1/checkpoint-2683/trainer_state.json

@@ -0,0 +1,1639 @@
+{
+  "best_metric": 0.9918227564998602,
+  "best_model_checkpoint": "./output/version_1\\checkpoint-2683",
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2683,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.692,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.693,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.6909,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6717,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 0.6707,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-06,
+      "loss": 0.6373,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.605,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5036,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9e-06,
+      "loss": 0.4553,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.3397,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.3443,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2e-05,
+      "loss": 0.227,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.1942,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.1926,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2876,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2707,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.098,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8e-05,
+      "loss": 0.1699,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9e-05,
+      "loss": 0.2029,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.1e-05,
+      "loss": 0.1786,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.1401,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.0326,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4e-05,
+      "loss": 0.2433,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1275,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.1329,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.125,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.1456,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1305,
+      "step": 290
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3e-05,
+      "loss": 0.0463,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.1e-05,
+      "loss": 0.1595,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.17,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.3e-05,
+      "loss": 0.2112,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1009,
+      "step": 340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.5e-05,
+      "loss": 0.0959,
+      "step": 350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1272,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.7e-05,
+      "loss": 0.0558,
+      "step": 370
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.8e-05,
+      "loss": 0.0698,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.1341,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4e-05,
+      "loss": 0.1771,
+      "step": 400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.1e-05,
+      "loss": 0.0634,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1192,
+      "step": 420
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1196,
+      "step": 430
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1244,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.5e-05,
+      "loss": 0.0939,
+      "step": 450
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.0321,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7e-05,
+      "loss": 0.0461,
+      "step": 470
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.8e-05,
+      "loss": 0.1511,
+      "step": 480
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9e-05,
+      "loss": 0.1864,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.0409,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.996128532713899e-05,
+      "loss": 0.0847,
+      "step": 510
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9922570654277976e-05,
+      "loss": 0.1341,
+      "step": 520
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.988385598141696e-05,
+      "loss": 0.0931,
+      "step": 530
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.984514130855594e-05,
+      "loss": 0.1214,
+      "step": 540
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.980642663569493e-05,
+      "loss": 0.0728,
+      "step": 550
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9767711962833916e-05,
+      "loss": 0.1141,
+      "step": 560
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.97289972899729e-05,
+      "loss": 0.1548,
+      "step": 570
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.969028261711189e-05,
+      "loss": 0.0486,
+      "step": 580
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.965156794425087e-05,
+      "loss": 0.1351,
+      "step": 590
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9612853271389856e-05,
+      "loss": 0.0929,
+      "step": 600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.957413859852884e-05,
+      "loss": 0.0646,
+      "step": 610
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.953542392566783e-05,
+      "loss": 0.004,
+      "step": 620
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.949670925280682e-05,
+      "loss": 0.0459,
+      "step": 630
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9457994579945803e-05,
+      "loss": 0.1922,
+      "step": 640
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9419279907084783e-05,
+      "loss": 0.0862,
+      "step": 650
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.938056523422378e-05,
+      "loss": 0.0087,
+      "step": 660
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.934185056136276e-05,
+      "loss": 0.1043,
+      "step": 670
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9303135888501744e-05,
+      "loss": 0.0537,
+      "step": 680
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.926442121564073e-05,
+      "loss": 0.0762,
+      "step": 690
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.922570654277972e-05,
+      "loss": 0.0516,
+      "step": 700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9186991869918704e-05,
+      "loss": 0.169,
+      "step": 710
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9148277197057684e-05,
+      "loss": 0.1627,
+      "step": 720
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.910956252419667e-05,
+      "loss": 0.0761,
+      "step": 730
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.907084785133566e-05,
+      "loss": 0.1429,
+      "step": 740
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.9032133178474644e-05,
+      "loss": 0.0719,
+      "step": 750
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.899341850561363e-05,
+      "loss": 0.0664,
+      "step": 760
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.895470383275261e-05,
+      "loss": 0.0187,
+      "step": 770
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.89159891598916e-05,
+      "loss": 0.0019,
+      "step": 780
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.887727448703059e-05,
+      "loss": 0.0499,
+      "step": 790
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.883855981416957e-05,
+      "loss": 0.0736,
+      "step": 800
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.879984514130856e-05,
+      "loss": 0.0217,
+      "step": 810
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.876113046844754e-05,
+      "loss": 0.0641,
+      "step": 820
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.872241579558653e-05,
+      "loss": 0.1733,
+      "step": 830
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.868370112272552e-05,
+      "loss": 0.076,
+      "step": 840
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.86449864498645e-05,
+      "loss": 0.008,
+      "step": 850
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8606271777003485e-05,
+      "loss": 0.0813,
+      "step": 860
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.856755710414247e-05,
+      "loss": 0.0159,
+      "step": 870
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.852884243128146e-05,
+      "loss": 0.0756,
+      "step": 880
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8490127758420445e-05,
+      "loss": 0.0847,
+      "step": 890
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8451413085559425e-05,
+      "loss": 0.1892,
+      "step": 900
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.0188,
+      "step": 910
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8373983739837406e-05,
+      "loss": 0.0139,
+      "step": 920
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8335269066976386e-05,
+      "loss": 0.0648,
+      "step": 930
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.829655439411537e-05,
+      "loss": 0.0329,
+      "step": 940
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.825783972125435e-05,
+      "loss": 0.0677,
+      "step": 950
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8219125048393346e-05,
+      "loss": 0.0296,
+      "step": 960
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.818041037553233e-05,
+      "loss": 0.1641,
+      "step": 970
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.814169570267131e-05,
+      "loss": 0.0519,
+      "step": 980
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.81029810298103e-05,
+      "loss": 0.0495,
+      "step": 990
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8064266356949286e-05,
+      "loss": 0.0451,
+      "step": 1000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.802555168408827e-05,
+      "loss": 0.0554,
+      "step": 1010
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.798683701122726e-05,
+      "loss": 0.0014,
+      "step": 1020
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.794812233836624e-05,
+      "loss": 0.0637,
+      "step": 1030
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.7909407665505226e-05,
+      "loss": 0.0273,
+      "step": 1040
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.787069299264422e-05,
+      "loss": 0.0013,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.78319783197832e-05,
+      "loss": 0.0189,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.779326364692219e-05,
+      "loss": 0.027,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.775454897406117e-05,
+      "loss": 0.0833,
+      "step": 1080
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.771583430120016e-05,
+      "loss": 0.0613,
+      "step": 1090
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.767711962833915e-05,
+      "loss": 0.0025,
+      "step": 1100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.763840495547813e-05,
+      "loss": 0.1202,
+      "step": 1110
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.7599690282617114e-05,
+      "loss": 0.0706,
+      "step": 1120
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.75609756097561e-05,
+      "loss": 0.0716,
+      "step": 1130
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.752226093689509e-05,
+      "loss": 0.0095,
+      "step": 1140
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7483546264034074e-05,
+      "loss": 0.0434,
+      "step": 1150
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7444831591173054e-05,
+      "loss": 0.11,
+      "step": 1160
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.740611691831204e-05,
+      "loss": 0.0091,
+      "step": 1170
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.736740224545103e-05,
+      "loss": 0.0755,
+      "step": 1180
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7328687572590014e-05,
+      "loss": 0.1525,
+      "step": 1190
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7289972899729e-05,
+      "loss": 0.1114,
+      "step": 1200
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.725125822686798e-05,
+      "loss": 0.1406,
+      "step": 1210
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7212543554006975e-05,
+      "loss": 0.0352,
+      "step": 1220
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7173828881145955e-05,
+      "loss": 0.0719,
+      "step": 1230
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.713511420828494e-05,
+      "loss": 0.0155,
+      "step": 1240
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.709639953542393e-05,
+      "loss": 0.1089,
+      "step": 1250
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7057684862562915e-05,
+      "loss": 0.0731,
+      "step": 1260
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.70189701897019e-05,
+      "loss": 0.0866,
+      "step": 1270
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.698025551684089e-05,
+      "loss": 0.0021,
+      "step": 1280
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.694154084397987e-05,
+      "loss": 0.1304,
+      "step": 1290
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6902826171118855e-05,
+      "loss": 0.0476,
+      "step": 1300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.686411149825784e-05,
+      "loss": 0.0354,
+      "step": 1310
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.0501,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6786682152535815e-05,
+      "loss": 0.1255,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6747967479674795e-05,
+      "loss": 0.0657,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.670925280681379e-05,
+      "loss": 0.0414,
+      "step": 1350
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.667053813395277e-05,
+      "loss": 0.1237,
+      "step": 1360
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6631823461091756e-05,
+      "loss": 0.0022,
+      "step": 1370
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.659310878823074e-05,
+      "loss": 0.0478,
+      "step": 1380
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.655439411536973e-05,
+      "loss": 0.0711,
+      "step": 1390
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.6515679442508716e-05,
+      "loss": 0.0217,
+      "step": 1400
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6476964769647696e-05,
+      "loss": 0.0856,
+      "step": 1410
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.643825009678668e-05,
+      "loss": 0.119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.639953542392567e-05,
+      "loss": 0.03,
+      "step": 1430
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6360820751064656e-05,
+      "loss": 0.0426,
+      "step": 1440
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.632210607820364e-05,
+      "loss": 0.1365,
+      "step": 1450
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.628339140534262e-05,
+      "loss": 0.1428,
+      "step": 1460
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.624467673248161e-05,
+      "loss": 0.0264,
+      "step": 1470
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.62059620596206e-05,
+      "loss": 0.0289,
+      "step": 1480
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.616724738675958e-05,
+      "loss": 0.0949,
+      "step": 1490
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.612853271389857e-05,
+      "loss": 0.0195,
+      "step": 1500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.608981804103756e-05,
+      "loss": 0.1042,
+      "step": 1510
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.6051103368176543e-05,
+      "loss": 0.0666,
+      "step": 1520
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.601238869531553e-05,
+      "loss": 0.0529,
+      "step": 1530
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.597367402245451e-05,
+      "loss": 0.047,
+      "step": 1540
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.59349593495935e-05,
+      "loss": 0.0338,
+      "step": 1550
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5896244676732484e-05,
+      "loss": 0.0107,
+      "step": 1560
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.585753000387147e-05,
+      "loss": 0.0304,
+      "step": 1570
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.581881533101046e-05,
+      "loss": 0.1013,
+      "step": 1580
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.578010065814944e-05,
+      "loss": 0.0352,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.5741385985288424e-05,
+      "loss": 0.0075,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.570267131242742e-05,
+      "loss": 0.0509,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.56639566395664e-05,
+      "loss": 0.0297,
+      "step": 1620
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5625241966705384e-05,
+      "loss": 0.0653,
+      "step": 1630
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5586527293844364e-05,
+      "loss": 0.0193,
+      "step": 1640
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.554781262098336e-05,
+      "loss": 0.2175,
+      "step": 1650
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5509097948122345e-05,
+      "loss": 0.0316,
+      "step": 1660
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5470383275261325e-05,
+      "loss": 0.0665,
+      "step": 1670
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.543166860240031e-05,
+      "loss": 0.0717,
+      "step": 1680
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.53929539295393e-05,
+      "loss": 0.0631,
+      "step": 1690
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5354239256678285e-05,
+      "loss": 0.0521,
+      "step": 1700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.531552458381727e-05,
+      "loss": 0.057,
+      "step": 1710
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.527680991095625e-05,
+      "loss": 0.0823,
+      "step": 1720
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.0572,
+      "step": 1730
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.5199380565234225e-05,
+      "loss": 0.1083,
+      "step": 1740
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.516066589237321e-05,
+      "loss": 0.0838,
+      "step": 1750
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.51219512195122e-05,
+      "loss": 0.0967,
+      "step": 1760
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.508323654665118e-05,
+      "loss": 0.0446,
+      "step": 1770
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5044521873790165e-05,
+      "loss": 0.0796,
+      "step": 1780
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.500580720092916e-05,
+      "loss": 0.0722,
+      "step": 1790
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.496709252806814e-05,
+      "loss": 0.1105,
+      "step": 1800
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4928377855207126e-05,
+      "loss": 0.0389,
+      "step": 1810
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4889663182346106e-05,
+      "loss": 0.0448,
+      "step": 1820
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.48509485094851e-05,
+      "loss": 0.0394,
+      "step": 1830
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4812233836624086e-05,
+      "loss": 0.0233,
+      "step": 1840
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4773519163763066e-05,
+      "loss": 0.0172,
+      "step": 1850
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.473480449090205e-05,
+      "loss": 0.077,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.469608981804104e-05,
+      "loss": 0.0259,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4657375145180026e-05,
+      "loss": 0.0846,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.461866047231901e-05,
+      "loss": 0.1179,
+      "step": 1890
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.457994579945799e-05,
+      "loss": 0.0322,
+      "step": 1900
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.454123112659698e-05,
+      "loss": 0.0288,
+      "step": 1910
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.450251645373597e-05,
+      "loss": 0.0802,
+      "step": 1920
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.446380178087495e-05,
+      "loss": 0.0765,
+      "step": 1930
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.442508710801394e-05,
+      "loss": 0.0112,
+      "step": 1940
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.438637243515292e-05,
+      "loss": 0.0645,
+      "step": 1950
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.4347657762291913e-05,
+      "loss": 0.0346,
+      "step": 1960
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.43089430894309e-05,
+      "loss": 0.0432,
+      "step": 1970
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.427022841656988e-05,
+      "loss": 0.06,
+      "step": 1980
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.423151374370887e-05,
+      "loss": 0.0404,
+      "step": 1990
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.4192799070847854e-05,
+      "loss": 0.0866,
+      "step": 2000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.415408439798684e-05,
+      "loss": 0.032,
+      "step": 2010
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.411536972512583e-05,
+      "loss": 0.0524,
+      "step": 2020
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.407665505226481e-05,
+      "loss": 0.0845,
+      "step": 2030
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.4037940379403794e-05,
+      "loss": 0.0361,
+      "step": 2040
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.399922570654278e-05,
+      "loss": 0.0688,
+      "step": 2050
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.396051103368177e-05,
+      "loss": 0.0151,
+      "step": 2060
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.3921796360820754e-05,
+      "loss": 0.0645,
+      "step": 2070
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.3883081687959734e-05,
+      "loss": 0.0024,
+      "step": 2080
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.384436701509873e-05,
+      "loss": 0.0026,
+      "step": 2090
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.380565234223771e-05,
+      "loss": 0.0604,
+      "step": 2100
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3766937669376695e-05,
+      "loss": 0.0012,
+      "step": 2110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.372822299651568e-05,
+      "loss": 0.0417,
+      "step": 2120
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.368950832365467e-05,
+      "loss": 0.1647,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.0347,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.361207897793264e-05,
+      "loss": 0.036,
+      "step": 2150
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.357336430507162e-05,
+      "loss": 0.0155,
+      "step": 2160
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.353464963221061e-05,
+      "loss": 0.0682,
+      "step": 2170
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.3495934959349595e-05,
+      "loss": 0.0258,
+      "step": 2180
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.345722028648858e-05,
+      "loss": 0.0407,
+      "step": 2190
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.341850561362757e-05,
+      "loss": 0.0182,
+      "step": 2200
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.337979094076655e-05,
+      "loss": 0.0362,
+      "step": 2210
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.334107626790554e-05,
+      "loss": 0.0388,
+      "step": 2220
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.330236159504452e-05,
+      "loss": 0.0652,
+      "step": 2230
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.326364692218351e-05,
+      "loss": 0.0167,
+      "step": 2240
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.3224932249322496e-05,
+      "loss": 0.0021,
+      "step": 2250
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.318621757646148e-05,
+      "loss": 0.0624,
+      "step": 2260
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.314750290360047e-05,
+      "loss": 0.0713,
+      "step": 2270
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.310878823073945e-05,
+      "loss": 0.0622,
+      "step": 2280
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.3070073557878436e-05,
+      "loss": 0.0238,
+      "step": 2290
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.303135888501742e-05,
+      "loss": 0.0425,
+      "step": 2300
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.299264421215641e-05,
+      "loss": 0.0177,
+      "step": 2310
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.2953929539295396e-05,
+      "loss": 0.097,
+      "step": 2320
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.291521486643438e-05,
+      "loss": 0.0612,
+      "step": 2330
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.287650019357336e-05,
+      "loss": 0.0444,
+      "step": 2340
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.2837785520712356e-05,
+      "loss": 0.082,
+      "step": 2350
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.2799070847851336e-05,
+      "loss": 0.0308,
+      "step": 2360
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.276035617499032e-05,
+      "loss": 0.0809,
+      "step": 2370
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.272164150212931e-05,
+      "loss": 0.0042,
+      "step": 2380
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.26829268292683e-05,
+      "loss": 0.0282,
+      "step": 2390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.2644212156407283e-05,
+      "loss": 0.043,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.2605497483546263e-05,
+      "loss": 0.1414,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.256678281068525e-05,
+      "loss": 0.1418,
+      "step": 2420
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.252806813782424e-05,
+      "loss": 0.0145,
+      "step": 2430
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.2489353464963224e-05,
+      "loss": 0.0514,
+      "step": 2440
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.245063879210221e-05,
+      "loss": 0.0134,
+      "step": 2450
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.241192411924119e-05,
+      "loss": 0.0253,
+      "step": 2460
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.237320944638018e-05,
+      "loss": 0.089,
+      "step": 2470
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.233449477351917e-05,
+      "loss": 0.063,
+      "step": 2480
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.229578010065815e-05,
+      "loss": 0.0488,
+      "step": 2490
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.225706542779714e-05,
+      "loss": 0.0024,
+      "step": 2500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.221835075493612e-05,
+      "loss": 0.0398,
+      "step": 2510
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.217963608207511e-05,
+      "loss": 0.115,
+      "step": 2520
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.21409214092141e-05,
+      "loss": 0.0305,
+      "step": 2530
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.210220673635308e-05,
+      "loss": 0.0667,
+      "step": 2540
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.0683,
+      "step": 2550
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.202477739063105e-05,
+      "loss": 0.0736,
+      "step": 2560
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.198606271777004e-05,
+      "loss": 0.013,
+      "step": 2570
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.1947348044909025e-05,
+      "loss": 0.0184,
+      "step": 2580
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.1908633372048005e-05,
+      "loss": 0.0778,
+      "step": 2590
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.186991869918699e-05,
+      "loss": 0.0018,
+      "step": 2600
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.1831204026325985e-05,
+      "loss": 0.0419,
+      "step": 2610
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.1792489353464965e-05,
+      "loss": 0.0129,
+      "step": 2620
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.175377468060395e-05,
+      "loss": 0.0254,
+      "step": 2630
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.171506000774293e-05,
+      "loss": 0.0337,
+      "step": 2640
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.1676345334881925e-05,
+      "loss": 0.0766,
+      "step": 2650
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.163763066202091e-05,
+      "loss": 0.0338,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.159891598915989e-05,
+      "loss": 0.0081,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.156020131629888e-05,
+      "loss": 0.1303,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9918227564998602,
+      "eval_f1": 0.9915358460536786,
+      "eval_loss": 0.030720438808202744,
+      "eval_runtime": 162.5131,
+      "eval_samples_per_second": 88.042,
+      "eval_steps_per_second": 1.378,
+      "step": 2683
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 13415,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 5685633217714176.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

output/version_1/checkpoint-2683/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74a88796cb182fbcbbd9333e4cc7f379372db09868e9a6a8fc694de5e7ebb51a
+size 4664

output/version_1/checkpoint-5366/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "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,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "vocab_size": 30522
+}

output/version_1/checkpoint-5366/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a1fb3f15e62570ec9ec4559ed016a2ea0dfd42a0b1ca7a7094b09b9d30988be
+size 267832560

output/version_1/checkpoint-5366/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e46886b98bf1f487faa7731e12645287ca075990053bdb24234166bb6e8094c1
+size 535727290

output/version_1/checkpoint-5366/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0ad96408ee695de286c6d7f3048d3b97ad07b18461f569b020d5e7c6f14180b
+size 14244

output/version_1/checkpoint-5366/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcfb6a9d3d7006a51bd7e9ca7e2e3a6bd61344cd981cefd77495fee0196c4252
+size 1064

output/version_1/checkpoint-5366/trainer_state.json

@@ -0,0 +1,3257 @@
+{
+  "best_metric": 0.9931506849315068,
+  "best_model_checkpoint": "./output/version_1\\checkpoint-5366",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 5366,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.692,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.693,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.6909,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6717,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 0.6707,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-06,
+      "loss": 0.6373,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.605,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5036,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9e-06,
+      "loss": 0.4553,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.3397,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.3443,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2e-05,
+      "loss": 0.227,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.1942,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.1926,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2876,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.2707,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.098,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8e-05,
+      "loss": 0.1699,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9e-05,
+      "loss": 0.2029,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.1e-05,
+      "loss": 0.1786,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.1401,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.0326,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.4e-05,
+      "loss": 0.2433,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1275,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.1329,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.125,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.1456,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1305,
+      "step": 290
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3e-05,
+      "loss": 0.0463,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.1e-05,
+      "loss": 0.1595,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.17,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.3e-05,
+      "loss": 0.2112,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1009,
+      "step": 340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.5e-05,
+      "loss": 0.0959,
+      "step": 350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1272,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.7e-05,
+      "loss": 0.0558,
+      "step": 370
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 3.8e-05,
+      "loss": 0.0698,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.1341,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4e-05,
+      "loss": 0.1771,
+      "step": 400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.1e-05,
+      "loss": 0.0634,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1192,
+      "step": 420
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1196,
+      "step": 430
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1244,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.5e-05,
+      "loss": 0.0939,
+      "step": 450
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.0321,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7e-05,
+      "loss": 0.0461,
+      "step": 470
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.8e-05,
+      "loss": 0.1511,
+      "step": 480
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9e-05,
+      "loss": 0.1864,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.0409,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.996128532713899e-05,
+      "loss": 0.0847,
+      "step": 510
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9922570654277976e-05,
+      "loss": 0.1341,
+      "step": 520
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.988385598141696e-05,
+      "loss": 0.0931,
+      "step": 530
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.984514130855594e-05,
+      "loss": 0.1214,
+      "step": 540
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.980642663569493e-05,
+      "loss": 0.0728,
+      "step": 550
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9767711962833916e-05,
+      "loss": 0.1141,
+      "step": 560
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.97289972899729e-05,
+      "loss": 0.1548,
+      "step": 570
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.969028261711189e-05,
+      "loss": 0.0486,
+      "step": 580
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.965156794425087e-05,
+      "loss": 0.1351,
+      "step": 590
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9612853271389856e-05,
+      "loss": 0.0929,
+      "step": 600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.957413859852884e-05,
+      "loss": 0.0646,
+      "step": 610
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.953542392566783e-05,
+      "loss": 0.004,
+      "step": 620
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.949670925280682e-05,
+      "loss": 0.0459,
+      "step": 630
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9457994579945803e-05,
+      "loss": 0.1922,
+      "step": 640
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9419279907084783e-05,
+      "loss": 0.0862,
+      "step": 650
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.938056523422378e-05,
+      "loss": 0.0087,
+      "step": 660
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.934185056136276e-05,
+      "loss": 0.1043,
+      "step": 670
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9303135888501744e-05,
+      "loss": 0.0537,
+      "step": 680
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.926442121564073e-05,
+      "loss": 0.0762,
+      "step": 690
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.922570654277972e-05,
+      "loss": 0.0516,
+      "step": 700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9186991869918704e-05,
+      "loss": 0.169,
+      "step": 710
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9148277197057684e-05,
+      "loss": 0.1627,
+      "step": 720
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.910956252419667e-05,
+      "loss": 0.0761,
+      "step": 730
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.907084785133566e-05,
+      "loss": 0.1429,
+      "step": 740
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.9032133178474644e-05,
+      "loss": 0.0719,
+      "step": 750
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.899341850561363e-05,
+      "loss": 0.0664,
+      "step": 760
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.895470383275261e-05,
+      "loss": 0.0187,
+      "step": 770
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.89159891598916e-05,
+      "loss": 0.0019,
+      "step": 780
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.887727448703059e-05,
+      "loss": 0.0499,
+      "step": 790
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.883855981416957e-05,
+      "loss": 0.0736,
+      "step": 800
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.879984514130856e-05,
+      "loss": 0.0217,
+      "step": 810
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.876113046844754e-05,
+      "loss": 0.0641,
+      "step": 820
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.872241579558653e-05,
+      "loss": 0.1733,
+      "step": 830
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.868370112272552e-05,
+      "loss": 0.076,
+      "step": 840
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.86449864498645e-05,
+      "loss": 0.008,
+      "step": 850
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8606271777003485e-05,
+      "loss": 0.0813,
+      "step": 860
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.856755710414247e-05,
+      "loss": 0.0159,
+      "step": 870
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.852884243128146e-05,
+      "loss": 0.0756,
+      "step": 880
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8490127758420445e-05,
+      "loss": 0.0847,
+      "step": 890
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8451413085559425e-05,
+      "loss": 0.1892,
+      "step": 900
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.0188,
+      "step": 910
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8373983739837406e-05,
+      "loss": 0.0139,
+      "step": 920
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8335269066976386e-05,
+      "loss": 0.0648,
+      "step": 930
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.829655439411537e-05,
+      "loss": 0.0329,
+      "step": 940
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.825783972125435e-05,
+      "loss": 0.0677,
+      "step": 950
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8219125048393346e-05,
+      "loss": 0.0296,
+      "step": 960
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.818041037553233e-05,
+      "loss": 0.1641,
+      "step": 970
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.814169570267131e-05,
+      "loss": 0.0519,
+      "step": 980
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.81029810298103e-05,
+      "loss": 0.0495,
+      "step": 990
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8064266356949286e-05,
+      "loss": 0.0451,
+      "step": 1000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.802555168408827e-05,
+      "loss": 0.0554,
+      "step": 1010
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.798683701122726e-05,
+      "loss": 0.0014,
+      "step": 1020
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.794812233836624e-05,
+      "loss": 0.0637,
+      "step": 1030
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.7909407665505226e-05,
+      "loss": 0.0273,
+      "step": 1040
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.787069299264422e-05,
+      "loss": 0.0013,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.78319783197832e-05,
+      "loss": 0.0189,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.779326364692219e-05,
+      "loss": 0.027,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.775454897406117e-05,
+      "loss": 0.0833,
+      "step": 1080
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.771583430120016e-05,
+      "loss": 0.0613,
+      "step": 1090
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.767711962833915e-05,
+      "loss": 0.0025,
+      "step": 1100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.763840495547813e-05,
+      "loss": 0.1202,
+      "step": 1110
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.7599690282617114e-05,
+      "loss": 0.0706,
+      "step": 1120
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.75609756097561e-05,
+      "loss": 0.0716,
+      "step": 1130
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.752226093689509e-05,
+      "loss": 0.0095,
+      "step": 1140
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7483546264034074e-05,
+      "loss": 0.0434,
+      "step": 1150
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7444831591173054e-05,
+      "loss": 0.11,
+      "step": 1160
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.740611691831204e-05,
+      "loss": 0.0091,
+      "step": 1170
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.736740224545103e-05,
+      "loss": 0.0755,
+      "step": 1180
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7328687572590014e-05,
+      "loss": 0.1525,
+      "step": 1190
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7289972899729e-05,
+      "loss": 0.1114,
+      "step": 1200
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.725125822686798e-05,
+      "loss": 0.1406,
+      "step": 1210
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7212543554006975e-05,
+      "loss": 0.0352,
+      "step": 1220
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7173828881145955e-05,
+      "loss": 0.0719,
+      "step": 1230
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.713511420828494e-05,
+      "loss": 0.0155,
+      "step": 1240
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.709639953542393e-05,
+      "loss": 0.1089,
+      "step": 1250
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7057684862562915e-05,
+      "loss": 0.0731,
+      "step": 1260
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.70189701897019e-05,
+      "loss": 0.0866,
+      "step": 1270
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.698025551684089e-05,
+      "loss": 0.0021,
+      "step": 1280
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.694154084397987e-05,
+      "loss": 0.1304,
+      "step": 1290
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6902826171118855e-05,
+      "loss": 0.0476,
+      "step": 1300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.686411149825784e-05,
+      "loss": 0.0354,
+      "step": 1310
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.0501,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6786682152535815e-05,
+      "loss": 0.1255,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6747967479674795e-05,
+      "loss": 0.0657,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.670925280681379e-05,
+      "loss": 0.0414,
+      "step": 1350
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.667053813395277e-05,
+      "loss": 0.1237,
+      "step": 1360
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6631823461091756e-05,
+      "loss": 0.0022,
+      "step": 1370
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.659310878823074e-05,
+      "loss": 0.0478,
+      "step": 1380
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.655439411536973e-05,
+      "loss": 0.0711,
+      "step": 1390
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.6515679442508716e-05,
+      "loss": 0.0217,
+      "step": 1400
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6476964769647696e-05,
+      "loss": 0.0856,
+      "step": 1410
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.643825009678668e-05,
+      "loss": 0.119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.639953542392567e-05,
+      "loss": 0.03,
+      "step": 1430
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6360820751064656e-05,
+      "loss": 0.0426,
+      "step": 1440
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.632210607820364e-05,
+      "loss": 0.1365,
+      "step": 1450
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.628339140534262e-05,
+      "loss": 0.1428,
+      "step": 1460
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.624467673248161e-05,
+      "loss": 0.0264,
+      "step": 1470
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.62059620596206e-05,
+      "loss": 0.0289,
+      "step": 1480
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.616724738675958e-05,
+      "loss": 0.0949,
+      "step": 1490
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.612853271389857e-05,
+      "loss": 0.0195,
+      "step": 1500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.608981804103756e-05,
+      "loss": 0.1042,
+      "step": 1510
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.6051103368176543e-05,
+      "loss": 0.0666,
+      "step": 1520
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.601238869531553e-05,
+      "loss": 0.0529,
+      "step": 1530
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.597367402245451e-05,
+      "loss": 0.047,
+      "step": 1540
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.59349593495935e-05,
+      "loss": 0.0338,
+      "step": 1550
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5896244676732484e-05,
+      "loss": 0.0107,
+      "step": 1560
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.585753000387147e-05,
+      "loss": 0.0304,
+      "step": 1570
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.581881533101046e-05,
+      "loss": 0.1013,
+      "step": 1580
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.578010065814944e-05,
+      "loss": 0.0352,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.5741385985288424e-05,
+      "loss": 0.0075,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.570267131242742e-05,
+      "loss": 0.0509,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.56639566395664e-05,
+      "loss": 0.0297,
+      "step": 1620
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5625241966705384e-05,
+      "loss": 0.0653,
+      "step": 1630
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5586527293844364e-05,
+      "loss": 0.0193,
+      "step": 1640
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.554781262098336e-05,
+      "loss": 0.2175,
+      "step": 1650
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5509097948122345e-05,
+      "loss": 0.0316,
+      "step": 1660
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5470383275261325e-05,
+      "loss": 0.0665,
+      "step": 1670
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.543166860240031e-05,
+      "loss": 0.0717,
+      "step": 1680
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.53929539295393e-05,
+      "loss": 0.0631,
+      "step": 1690
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5354239256678285e-05,
+      "loss": 0.0521,
+      "step": 1700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.531552458381727e-05,
+      "loss": 0.057,
+      "step": 1710
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.527680991095625e-05,
+      "loss": 0.0823,
+      "step": 1720
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.0572,
+      "step": 1730
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.5199380565234225e-05,
+      "loss": 0.1083,
+      "step": 1740
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.516066589237321e-05,
+      "loss": 0.0838,
+      "step": 1750
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.51219512195122e-05,
+      "loss": 0.0967,
+      "step": 1760
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.508323654665118e-05,
+      "loss": 0.0446,
+      "step": 1770
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5044521873790165e-05,
+      "loss": 0.0796,
+      "step": 1780
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.500580720092916e-05,
+      "loss": 0.0722,
+      "step": 1790
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.496709252806814e-05,
+      "loss": 0.1105,
+      "step": 1800
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4928377855207126e-05,
+      "loss": 0.0389,
+      "step": 1810
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4889663182346106e-05,
+      "loss": 0.0448,
+      "step": 1820
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.48509485094851e-05,
+      "loss": 0.0394,
+      "step": 1830
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4812233836624086e-05,
+      "loss": 0.0233,
+      "step": 1840
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4773519163763066e-05,
+      "loss": 0.0172,
+      "step": 1850
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.473480449090205e-05,
+      "loss": 0.077,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.469608981804104e-05,
+      "loss": 0.0259,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4657375145180026e-05,
+      "loss": 0.0846,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.461866047231901e-05,
+      "loss": 0.1179,
+      "step": 1890
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.457994579945799e-05,
+      "loss": 0.0322,
+      "step": 1900
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.454123112659698e-05,
+      "loss": 0.0288,
+      "step": 1910
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.450251645373597e-05,
+      "loss": 0.0802,
+      "step": 1920
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.446380178087495e-05,
+      "loss": 0.0765,
+      "step": 1930
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.442508710801394e-05,
+      "loss": 0.0112,
+      "step": 1940
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.438637243515292e-05,
+      "loss": 0.0645,
+      "step": 1950
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.4347657762291913e-05,
+      "loss": 0.0346,
+      "step": 1960
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.43089430894309e-05,
+      "loss": 0.0432,
+      "step": 1970
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.427022841656988e-05,
+      "loss": 0.06,
+      "step": 1980
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.423151374370887e-05,
+      "loss": 0.0404,
+      "step": 1990
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.4192799070847854e-05,
+      "loss": 0.0866,
+      "step": 2000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.415408439798684e-05,
+      "loss": 0.032,
+      "step": 2010
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.411536972512583e-05,
+      "loss": 0.0524,
+      "step": 2020
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.407665505226481e-05,
+      "loss": 0.0845,
+      "step": 2030
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.4037940379403794e-05,
+      "loss": 0.0361,
+      "step": 2040
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.399922570654278e-05,
+      "loss": 0.0688,
+      "step": 2050
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.396051103368177e-05,
+      "loss": 0.0151,
+      "step": 2060
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.3921796360820754e-05,
+      "loss": 0.0645,
+      "step": 2070
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.3883081687959734e-05,
+      "loss": 0.0024,
+      "step": 2080
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.384436701509873e-05,
+      "loss": 0.0026,
+      "step": 2090
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.380565234223771e-05,
+      "loss": 0.0604,
+      "step": 2100
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3766937669376695e-05,
+      "loss": 0.0012,
+      "step": 2110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.372822299651568e-05,
+      "loss": 0.0417,
+      "step": 2120
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.368950832365467e-05,
+      "loss": 0.1647,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.0347,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.361207897793264e-05,
+      "loss": 0.036,
+      "step": 2150
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.357336430507162e-05,
+      "loss": 0.0155,
+      "step": 2160
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.353464963221061e-05,
+      "loss": 0.0682,
+      "step": 2170
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.3495934959349595e-05,
+      "loss": 0.0258,
+      "step": 2180
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.345722028648858e-05,
+      "loss": 0.0407,
+      "step": 2190
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.341850561362757e-05,
+      "loss": 0.0182,
+      "step": 2200
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.337979094076655e-05,
+      "loss": 0.0362,
+      "step": 2210
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.334107626790554e-05,
+      "loss": 0.0388,
+      "step": 2220
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.330236159504452e-05,
+      "loss": 0.0652,
+      "step": 2230
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.326364692218351e-05,
+      "loss": 0.0167,
+      "step": 2240
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.3224932249322496e-05,
+      "loss": 0.0021,
+      "step": 2250
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.318621757646148e-05,
+      "loss": 0.0624,
+      "step": 2260
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.314750290360047e-05,
+      "loss": 0.0713,
+      "step": 2270
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.310878823073945e-05,
+      "loss": 0.0622,
+      "step": 2280
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.3070073557878436e-05,
+      "loss": 0.0238,
+      "step": 2290
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.303135888501742e-05,
+      "loss": 0.0425,
+      "step": 2300
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.299264421215641e-05,
+      "loss": 0.0177,
+      "step": 2310
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.2953929539295396e-05,
+      "loss": 0.097,
+      "step": 2320
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.291521486643438e-05,
+      "loss": 0.0612,
+      "step": 2330
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.287650019357336e-05,
+      "loss": 0.0444,
+      "step": 2340
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.2837785520712356e-05,
+      "loss": 0.082,
+      "step": 2350
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.2799070847851336e-05,
+      "loss": 0.0308,
+      "step": 2360
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.276035617499032e-05,
+      "loss": 0.0809,
+      "step": 2370
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.272164150212931e-05,
+      "loss": 0.0042,
+      "step": 2380
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.26829268292683e-05,
+      "loss": 0.0282,
+      "step": 2390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.2644212156407283e-05,
+      "loss": 0.043,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.2605497483546263e-05,
+      "loss": 0.1414,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.256678281068525e-05,
+      "loss": 0.1418,
+      "step": 2420
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.252806813782424e-05,
+      "loss": 0.0145,
+      "step": 2430
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.2489353464963224e-05,
+      "loss": 0.0514,
+      "step": 2440
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.245063879210221e-05,
+      "loss": 0.0134,
+      "step": 2450
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.241192411924119e-05,
+      "loss": 0.0253,
+      "step": 2460
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.237320944638018e-05,
+      "loss": 0.089,
+      "step": 2470
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.233449477351917e-05,
+      "loss": 0.063,
+      "step": 2480
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.229578010065815e-05,
+      "loss": 0.0488,
+      "step": 2490
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.225706542779714e-05,
+      "loss": 0.0024,
+      "step": 2500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.221835075493612e-05,
+      "loss": 0.0398,
+      "step": 2510
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.217963608207511e-05,
+      "loss": 0.115,
+      "step": 2520
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.21409214092141e-05,
+      "loss": 0.0305,
+      "step": 2530
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.210220673635308e-05,
+      "loss": 0.0667,
+      "step": 2540
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.0683,
+      "step": 2550
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.202477739063105e-05,
+      "loss": 0.0736,
+      "step": 2560
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.198606271777004e-05,
+      "loss": 0.013,
+      "step": 2570
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.1947348044909025e-05,
+      "loss": 0.0184,
+      "step": 2580
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.1908633372048005e-05,
+      "loss": 0.0778,
+      "step": 2590
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.186991869918699e-05,
+      "loss": 0.0018,
+      "step": 2600
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.1831204026325985e-05,
+      "loss": 0.0419,
+      "step": 2610
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.1792489353464965e-05,
+      "loss": 0.0129,
+      "step": 2620
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.175377468060395e-05,
+      "loss": 0.0254,
+      "step": 2630
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.171506000774293e-05,
+      "loss": 0.0337,
+      "step": 2640
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.1676345334881925e-05,
+      "loss": 0.0766,
+      "step": 2650
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.163763066202091e-05,
+      "loss": 0.0338,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.159891598915989e-05,
+      "loss": 0.0081,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.156020131629888e-05,
+      "loss": 0.1303,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9918227564998602,
+      "eval_f1": 0.9915358460536786,
+      "eval_loss": 0.030720438808202744,
+      "eval_runtime": 162.5131,
+      "eval_samples_per_second": 88.042,
+      "eval_steps_per_second": 1.378,
+      "step": 2683
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1521486643437866e-05,
+      "loss": 0.0511,
+      "step": 2690
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.148277197057685e-05,
+      "loss": 0.0236,
+      "step": 2700
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.144405729771584e-05,
+      "loss": 0.0021,
+      "step": 2710
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.140534262485482e-05,
+      "loss": 0.0025,
+      "step": 2720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1366627951993806e-05,
+      "loss": 0.0691,
+      "step": 2730
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132791327913279e-05,
+      "loss": 0.0077,
+      "step": 2740
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.128919860627178e-05,
+      "loss": 0.0004,
+      "step": 2750
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.1250483933410766e-05,
+      "loss": 0.0261,
+      "step": 2760
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.1211769260549746e-05,
+      "loss": 0.024,
+      "step": 2770
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.117305458768874e-05,
+      "loss": 0.0005,
+      "step": 2780
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.1134339914827726e-05,
+      "loss": 0.0025,
+      "step": 2790
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.1095625241966706e-05,
+      "loss": 0.0045,
+      "step": 2800
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.105691056910569e-05,
+      "loss": 0.082,
+      "step": 2810
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.101819589624468e-05,
+      "loss": 0.0223,
+      "step": 2820
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.097948122338367e-05,
+      "loss": 0.0535,
+      "step": 2830
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.0940766550522653e-05,
+      "loss": 0.0025,
+      "step": 2840
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.0902051877661633e-05,
+      "loss": 0.0092,
+      "step": 2850
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.086333720480062e-05,
+      "loss": 0.0121,
+      "step": 2860
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.082462253193961e-05,
+      "loss": 0.0668,
+      "step": 2870
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.0785907859078594e-05,
+      "loss": 0.0789,
+      "step": 2880
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.074719318621758e-05,
+      "loss": 0.0247,
+      "step": 2890
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.070847851335656e-05,
+      "loss": 0.0294,
+      "step": 2900
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.066976384049555e-05,
+      "loss": 0.0382,
+      "step": 2910
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.0631049167634534e-05,
+      "loss": 0.0025,
+      "step": 2920
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.059233449477352e-05,
+      "loss": 0.0018,
+      "step": 2930
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.055361982191251e-05,
+      "loss": 0.0003,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.051490514905149e-05,
+      "loss": 0.0459,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.0054,
+      "step": 2960
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.043747580332947e-05,
+      "loss": 0.0906,
+      "step": 2970
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.039876113046845e-05,
+      "loss": 0.0004,
+      "step": 2980
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.0360046457607435e-05,
+      "loss": 0.0378,
+      "step": 2990
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.032133178474642e-05,
+      "loss": 0.0383,
+      "step": 3000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.028261711188541e-05,
+      "loss": 0.0672,
+      "step": 3010
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.0243902439024395e-05,
+      "loss": 0.029,
+      "step": 3020
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.0205187766163375e-05,
+      "loss": 0.0007,
+      "step": 3030
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.016647309330236e-05,
+      "loss": 0.1203,
+      "step": 3040
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.012775842044135e-05,
+      "loss": 0.0012,
+      "step": 3050
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.0089043747580335e-05,
+      "loss": 0.0248,
+      "step": 3060
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.005032907471932e-05,
+      "loss": 0.0371,
+      "step": 3070
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.00116144018583e-05,
+      "loss": 0.0877,
+      "step": 3080
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.9972899728997295e-05,
+      "loss": 0.0044,
+      "step": 3090
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.9934185056136275e-05,
+      "loss": 0.032,
+      "step": 3100
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.989547038327526e-05,
+      "loss": 0.0465,
+      "step": 3110
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.985675571041425e-05,
+      "loss": 0.0889,
+      "step": 3120
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.9818041037553236e-05,
+      "loss": 0.0017,
+      "step": 3130
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.977932636469222e-05,
+      "loss": 0.023,
+      "step": 3140
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.97406116918312e-05,
+      "loss": 0.0203,
+      "step": 3150
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.970189701897019e-05,
+      "loss": 0.0494,
+      "step": 3160
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.9663182346109176e-05,
+      "loss": 0.0222,
+      "step": 3170
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.962446767324816e-05,
+      "loss": 0.0008,
+      "step": 3180
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.958575300038715e-05,
+      "loss": 0.0009,
+      "step": 3190
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.9547038327526136e-05,
+      "loss": 0.0005,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.9508323654665116e-05,
+      "loss": 0.0739,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.946960898180411e-05,
+      "loss": 0.0428,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.943089430894309e-05,
+      "loss": 0.0031,
+      "step": 3230
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.9392179636082076e-05,
+      "loss": 0.0003,
+      "step": 3240
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.935346496322106e-05,
+      "loss": 0.0231,
+      "step": 3250
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.931475029036005e-05,
+      "loss": 0.0773,
+      "step": 3260
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.927603561749904e-05,
+      "loss": 0.0235,
+      "step": 3270
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.923732094463802e-05,
+      "loss": 0.0553,
+      "step": 3280
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.9198606271777003e-05,
+      "loss": 0.017,
+      "step": 3290
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.915989159891599e-05,
+      "loss": 0.0003,
+      "step": 3300
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.912117692605498e-05,
+      "loss": 0.0004,
+      "step": 3310
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.9082462253193964e-05,
+      "loss": 0.0003,
+      "step": 3320
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.9043747580332944e-05,
+      "loss": 0.0002,
+      "step": 3330
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.900503290747193e-05,
+      "loss": 0.0001,
+      "step": 3340
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.8966318234610924e-05,
+      "loss": 0.0014,
+      "step": 3350
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.8927603561749904e-05,
+      "loss": 0.0782,
+      "step": 3360
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.0006,
+      "step": 3370
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.885017421602787e-05,
+      "loss": 0.0428,
+      "step": 3380
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.8811459543166864e-05,
+      "loss": 0.0493,
+      "step": 3390
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.877274487030585e-05,
+      "loss": 0.0225,
+      "step": 3400
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.873403019744483e-05,
+      "loss": 0.0006,
+      "step": 3410
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.869531552458382e-05,
+      "loss": 0.0008,
+      "step": 3420
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.8656600851722805e-05,
+      "loss": 0.0198,
+      "step": 3430
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.861788617886179e-05,
+      "loss": 0.0024,
+      "step": 3440
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.857917150600078e-05,
+      "loss": 0.0007,
+      "step": 3450
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.854045683313976e-05,
+      "loss": 0.0241,
+      "step": 3460
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.8501742160278745e-05,
+      "loss": 0.039,
+      "step": 3470
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.846302748741774e-05,
+      "loss": 0.0005,
+      "step": 3480
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.842431281455672e-05,
+      "loss": 0.031,
+      "step": 3490
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.8385598141695705e-05,
+      "loss": 0.0005,
+      "step": 3500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.8346883468834685e-05,
+      "loss": 0.0002,
+      "step": 3510
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.830816879597368e-05,
+      "loss": 0.0759,
+      "step": 3520
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.8269454123112665e-05,
+      "loss": 0.0003,
+      "step": 3530
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.8230739450251645e-05,
+      "loss": 0.054,
+      "step": 3540
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.819202477739063e-05,
+      "loss": 0.0002,
+      "step": 3550
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.815331010452962e-05,
+      "loss": 0.0368,
+      "step": 3560
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.8114595431668606e-05,
+      "loss": 0.0003,
+      "step": 3570
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.807588075880759e-05,
+      "loss": 0.0475,
+      "step": 3580
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.803716608594657e-05,
+      "loss": 0.003,
+      "step": 3590
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.799845141308556e-05,
+      "loss": 0.0009,
+      "step": 3600
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.795973674022455e-05,
+      "loss": 0.0005,
+      "step": 3610
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.792102206736353e-05,
+      "loss": 0.0727,
+      "step": 3620
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.788230739450252e-05,
+      "loss": 0.0242,
+      "step": 3630
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.78435927216415e-05,
+      "loss": 0.0233,
+      "step": 3640
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.780487804878049e-05,
+      "loss": 0.0008,
+      "step": 3650
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.776616337591948e-05,
+      "loss": 0.0332,
+      "step": 3660
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.772744870305846e-05,
+      "loss": 0.035,
+      "step": 3670
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.7688734030197446e-05,
+      "loss": 0.0191,
+      "step": 3680
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.765001935733643e-05,
+      "loss": 0.1471,
+      "step": 3690
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.761130468447542e-05,
+      "loss": 0.001,
+      "step": 3700
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.757259001161441e-05,
+      "loss": 0.0007,
+      "step": 3710
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.753387533875339e-05,
+      "loss": 0.001,
+      "step": 3720
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.7495160665892373e-05,
+      "loss": 0.027,
+      "step": 3730
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.745644599303136e-05,
+      "loss": 0.0235,
+      "step": 3740
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.741773132017035e-05,
+      "loss": 0.0068,
+      "step": 3750
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.7379016647309334e-05,
+      "loss": 0.0034,
+      "step": 3760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.7340301974448314e-05,
+      "loss": 0.0014,
+      "step": 3770
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.0351,
+      "step": 3780
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.726287262872629e-05,
+      "loss": 0.0006,
+      "step": 3790
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.7224157955865274e-05,
+      "loss": 0.1266,
+      "step": 3800
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.718544328300426e-05,
+      "loss": 0.0595,
+      "step": 3810
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.714672861014325e-05,
+      "loss": 0.002,
+      "step": 3820
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.7108013937282234e-05,
+      "loss": 0.0378,
+      "step": 3830
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.706929926442122e-05,
+      "loss": 0.001,
+      "step": 3840
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.70305845915602e-05,
+      "loss": 0.035,
+      "step": 3850
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.699186991869919e-05,
+      "loss": 0.0383,
+      "step": 3860
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.6953155245838175e-05,
+      "loss": 0.0047,
+      "step": 3870
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.691444057297716e-05,
+      "loss": 0.0004,
+      "step": 3880
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.687572590011615e-05,
+      "loss": 0.0117,
+      "step": 3890
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.683701122725513e-05,
+      "loss": 0.0917,
+      "step": 3900
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.679829655439412e-05,
+      "loss": 0.0312,
+      "step": 3910
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.67595818815331e-05,
+      "loss": 0.0463,
+      "step": 3920
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.672086720867209e-05,
+      "loss": 0.0232,
+      "step": 3930
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.6682152535811075e-05,
+      "loss": 0.0059,
+      "step": 3940
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.664343786295006e-05,
+      "loss": 0.0209,
+      "step": 3950
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.660472319008905e-05,
+      "loss": 0.0528,
+      "step": 3960
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.656600851722803e-05,
+      "loss": 0.0377,
+      "step": 3970
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.6527293844367015e-05,
+      "loss": 0.0148,
+      "step": 3980
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.6488579171506e-05,
+      "loss": 0.0324,
+      "step": 3990
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.644986449864499e-05,
+      "loss": 0.0209,
+      "step": 4000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.6411149825783976e-05,
+      "loss": 0.002,
+      "step": 4010
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.6372435152922956e-05,
+      "loss": 0.0362,
+      "step": 4020
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.633372048006194e-05,
+      "loss": 0.0067,
+      "step": 4030
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.629500580720093e-05,
+      "loss": 0.0004,
+      "step": 4040
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.6256291134339916e-05,
+      "loss": 0.018,
+      "step": 4050
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.62175764614789e-05,
+      "loss": 0.0161,
+      "step": 4060
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.617886178861789e-05,
+      "loss": 0.0005,
+      "step": 4070
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.614014711575687e-05,
+      "loss": 0.0587,
+      "step": 4080
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.610143244289586e-05,
+      "loss": 0.0614,
+      "step": 4090
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.606271777003484e-05,
+      "loss": 0.0813,
+      "step": 4100
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.602400309717383e-05,
+      "loss": 0.0004,
+      "step": 4110
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.5985288424312816e-05,
+      "loss": 0.0181,
+      "step": 4120
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.59465737514518e-05,
+      "loss": 0.0121,
+      "step": 4130
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.590785907859079e-05,
+      "loss": 0.0001,
+      "step": 4140
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.586914440572977e-05,
+      "loss": 0.0275,
+      "step": 4150
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.583042973286876e-05,
+      "loss": 0.0261,
+      "step": 4160
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.5791715060007743e-05,
+      "loss": 0.0003,
+      "step": 4170
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.575300038714673e-05,
+      "loss": 0.0003,
+      "step": 4180
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.0187,
+      "step": 4190
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.56755710414247e-05,
+      "loss": 0.0781,
+      "step": 4200
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.5636856368563684e-05,
+      "loss": 0.0011,
+      "step": 4210
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.559814169570268e-05,
+      "loss": 0.0106,
+      "step": 4220
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.555942702284166e-05,
+      "loss": 0.0006,
+      "step": 4230
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.5520712349980644e-05,
+      "loss": 0.0672,
+      "step": 4240
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.548199767711963e-05,
+      "loss": 0.0001,
+      "step": 4250
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.544328300425862e-05,
+      "loss": 0.0022,
+      "step": 4260
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.5404568331397604e-05,
+      "loss": 0.0875,
+      "step": 4270
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.5365853658536584e-05,
+      "loss": 0.0105,
+      "step": 4280
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.532713898567557e-05,
+      "loss": 0.0113,
+      "step": 4290
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.528842431281456e-05,
+      "loss": 0.0002,
+      "step": 4300
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.5249709639953545e-05,
+      "loss": 0.0011,
+      "step": 4310
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.521099496709253e-05,
+      "loss": 0.0003,
+      "step": 4320
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.517228029423151e-05,
+      "loss": 0.0001,
+      "step": 4330
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.51335656213705e-05,
+      "loss": 0.0005,
+      "step": 4340
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.509485094850949e-05,
+      "loss": 0.0001,
+      "step": 4350
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.505613627564847e-05,
+      "loss": 0.0083,
+      "step": 4360
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.501742160278746e-05,
+      "loss": 0.0001,
+      "step": 4370
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.497870692992644e-05,
+      "loss": 0.0005,
+      "step": 4380
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.493999225706543e-05,
+      "loss": 0.0498,
+      "step": 4390
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.490127758420442e-05,
+      "loss": 0.0199,
+      "step": 4400
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.48625629113434e-05,
+      "loss": 0.0021,
+      "step": 4410
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.4823848238482385e-05,
+      "loss": 0.0001,
+      "step": 4420
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.478513356562137e-05,
+      "loss": 0.0102,
+      "step": 4430
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.474641889276036e-05,
+      "loss": 0.0373,
+      "step": 4440
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.4707704219899346e-05,
+      "loss": 0.0967,
+      "step": 4450
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.4668989547038326e-05,
+      "loss": 0.0023,
+      "step": 4460
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.463027487417731e-05,
+      "loss": 0.0317,
+      "step": 4470
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4591560201316306e-05,
+      "loss": 0.096,
+      "step": 4480
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.4552845528455286e-05,
+      "loss": 0.0178,
+      "step": 4490
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.451413085559427e-05,
+      "loss": 0.0185,
+      "step": 4500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.447541618273325e-05,
+      "loss": 0.0013,
+      "step": 4510
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.4436701509872246e-05,
+      "loss": 0.0013,
+      "step": 4520
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.439798683701123e-05,
+      "loss": 0.0002,
+      "step": 4530
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.435927216415021e-05,
+      "loss": 0.0792,
+      "step": 4540
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.43205574912892e-05,
+      "loss": 0.0321,
+      "step": 4550
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.4281842818428186e-05,
+      "loss": 0.0228,
+      "step": 4560
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.424312814556717e-05,
+      "loss": 0.0112,
+      "step": 4570
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.420441347270616e-05,
+      "loss": 0.0336,
+      "step": 4580
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.416569879984514e-05,
+      "loss": 0.0205,
+      "step": 4590
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.0612,
+      "step": 4600
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.4088269454123113e-05,
+      "loss": 0.0018,
+      "step": 4610
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.40495547812621e-05,
+      "loss": 0.0287,
+      "step": 4620
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.401084010840109e-05,
+      "loss": 0.0003,
+      "step": 4630
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.397212543554007e-05,
+      "loss": 0.0308,
+      "step": 4640
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.393341076267906e-05,
+      "loss": 0.0003,
+      "step": 4650
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.389469608981804e-05,
+      "loss": 0.0002,
+      "step": 4660
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.385598141695703e-05,
+      "loss": 0.0003,
+      "step": 4670
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.3817266744096014e-05,
+      "loss": 0.0002,
+      "step": 4680
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.3778552071235e-05,
+      "loss": 0.0004,
+      "step": 4690
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.373983739837399e-05,
+      "loss": 0.0223,
+      "step": 4700
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.3701122725512974e-05,
+      "loss": 0.0001,
+      "step": 4710
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.3662408052651954e-05,
+      "loss": 0.0001,
+      "step": 4720
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.362369337979094e-05,
+      "loss": 0.0001,
+      "step": 4730
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.358497870692993e-05,
+      "loss": 0.0482,
+      "step": 4740
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.3546264034068915e-05,
+      "loss": 0.0056,
+      "step": 4750
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.35075493612079e-05,
+      "loss": 0.0001,
+      "step": 4760
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.346883468834688e-05,
+      "loss": 0.0043,
+      "step": 4770
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.3430120015485875e-05,
+      "loss": 0.0003,
+      "step": 4780
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.3391405342624855e-05,
+      "loss": 0.0729,
+      "step": 4790
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.335269066976384e-05,
+      "loss": 0.0399,
+      "step": 4800
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.331397599690283e-05,
+      "loss": 0.0369,
+      "step": 4810
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.3275261324041815e-05,
+      "loss": 0.0012,
+      "step": 4820
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.32365466511808e-05,
+      "loss": 0.0051,
+      "step": 4830
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.319783197831978e-05,
+      "loss": 0.0001,
+      "step": 4840
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.315911730545877e-05,
+      "loss": 0.04,
+      "step": 4850
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.3120402632597755e-05,
+      "loss": 0.007,
+      "step": 4860
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.308168795973674e-05,
+      "loss": 0.03,
+      "step": 4870
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.304297328687573e-05,
+      "loss": 0.016,
+      "step": 4880
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.3004258614014716e-05,
+      "loss": 0.0006,
+      "step": 4890
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.2965543941153696e-05,
+      "loss": 0.0001,
+      "step": 4900
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.292682926829269e-05,
+      "loss": 0.0004,
+      "step": 4910
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.288811459543167e-05,
+      "loss": 0.0002,
+      "step": 4920
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.2849399922570656e-05,
+      "loss": 0.0236,
+      "step": 4930
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.281068524970964e-05,
+      "loss": 0.0002,
+      "step": 4940
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.277197057684863e-05,
+      "loss": 0.0117,
+      "step": 4950
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.2733255903987616e-05,
+      "loss": 0.0335,
+      "step": 4960
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.2694541231126596e-05,
+      "loss": 0.0257,
+      "step": 4970
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.265582655826558e-05,
+      "loss": 0.027,
+      "step": 4980
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.261711188540457e-05,
+      "loss": 0.0042,
+      "step": 4990
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.2578397212543556e-05,
+      "loss": 0.068,
+      "step": 5000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.0003,
+      "step": 5010
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.250096786682152e-05,
+      "loss": 0.0068,
+      "step": 5020
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.246225319396051e-05,
+      "loss": 0.0299,
+      "step": 5030
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.2423538521099503e-05,
+      "loss": 0.0148,
+      "step": 5040
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.2384823848238483e-05,
+      "loss": 0.0016,
+      "step": 5050
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.234610917537747e-05,
+      "loss": 0.0059,
+      "step": 5060
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.230739450251645e-05,
+      "loss": 0.0375,
+      "step": 5070
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2268679829655444e-05,
+      "loss": 0.0001,
+      "step": 5080
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.222996515679443e-05,
+      "loss": 0.0009,
+      "step": 5090
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.219125048393341e-05,
+      "loss": 0.0154,
+      "step": 5100
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.21525358110724e-05,
+      "loss": 0.0693,
+      "step": 5110
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.2113821138211384e-05,
+      "loss": 0.0138,
+      "step": 5120
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.207510646535037e-05,
+      "loss": 0.0011,
+      "step": 5130
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.203639179248936e-05,
+      "loss": 0.0006,
+      "step": 5140
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.199767711962834e-05,
+      "loss": 0.0107,
+      "step": 5150
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.1958962446767324e-05,
+      "loss": 0.0166,
+      "step": 5160
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.192024777390631e-05,
+      "loss": 0.0001,
+      "step": 5170
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.18815331010453e-05,
+      "loss": 0.0364,
+      "step": 5180
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1842818428184285e-05,
+      "loss": 0.0379,
+      "step": 5190
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.1804103755323265e-05,
+      "loss": 0.0003,
+      "step": 5200
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.176538908246225e-05,
+      "loss": 0.0002,
+      "step": 5210
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1726674409601245e-05,
+      "loss": 0.0003,
+      "step": 5220
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1687959736740225e-05,
+      "loss": 0.0003,
+      "step": 5230
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.164924506387921e-05,
+      "loss": 0.0318,
+      "step": 5240
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.161053039101819e-05,
+      "loss": 0.0056,
+      "step": 5250
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.1571815718157185e-05,
+      "loss": 0.0055,
+      "step": 5260
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.153310104529617e-05,
+      "loss": 0.0124,
+      "step": 5270
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.149438637243515e-05,
+      "loss": 0.0075,
+      "step": 5280
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.145567169957414e-05,
+      "loss": 0.0425,
+      "step": 5290
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.1416957026713125e-05,
+      "loss": 0.006,
+      "step": 5300
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.137824235385211e-05,
+      "loss": 0.0003,
+      "step": 5310
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.13395276809911e-05,
+      "loss": 0.0001,
+      "step": 5320
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.130081300813008e-05,
+      "loss": 0.0002,
+      "step": 5330
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.1262098335269066e-05,
+      "loss": 0.0207,
+      "step": 5340
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.122338366240806e-05,
+      "loss": 0.0343,
+      "step": 5350
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.118466898954704e-05,
+      "loss": 0.0129,
+      "step": 5360
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9931506849315068,
+      "eval_f1": 0.9929078014184397,
+      "eval_loss": 0.03655492141842842,
+      "eval_runtime": 150.0177,
+      "eval_samples_per_second": 95.375,
+      "eval_steps_per_second": 1.493,
+      "step": 5366
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 13415,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 1.1371266435428352e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

output/version_1/checkpoint-5366/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74a88796cb182fbcbbd9333e4cc7f379372db09868e9a6a8fc694de5e7ebb51a
+size 4664

output/version_1/tokenized_data_1/test/data-00000-of-00001.arrow

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff2015a0011ea37aa7a210ba102a74feab2069de421d2d5468c3e5481c4280dc
+size 85650992

output/version_1/tokenized_data_1/test/dataset_info.json

@@ -0,0 +1,34 @@
+{
+  "citation": "",
+  "description": "",
+  "features": {
+    "text": {
+      "dtype": "string",
+      "_type": "Value"
+    },
+    "label": {
+      "dtype": "int64",
+      "_type": "Value"
+    },
+    "__index_level_0__": {
+      "dtype": "int64",
+      "_type": "Value"
+    },
+    "input_ids": {
+      "feature": {
+        "dtype": "int32",
+        "_type": "Value"
+      },
+      "_type": "Sequence"
+    },
+    "attention_mask": {
+      "feature": {
+        "dtype": "int8",
+        "_type": "Value"
+      },
+      "_type": "Sequence"
+    }
+  },
+  "homepage": "",
+  "license": ""
+}

output/version_1/tokenized_data_1/test/state.json

@@ -0,0 +1,13 @@
+{
+  "_data_files": [
+    {
+      "filename": "data-00000-of-00001.arrow"
+    }
+  ],
+  "_fingerprint": "5ce9014249281adb",
+  "_format_columns": null,
+  "_format_kwargs": {},
+  "_format_type": null,
+  "_output_all_columns": false,
+  "_split": null
+}

output/version_1/tokenized_data_1/train/data-00000-of-00001.arrow

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad61882c6c08aecfc5415225f0fc9b2dfe5adc35c04f72ff13a54d325f9c5705
+size 257513392

output/version_1/tokenized_data_1/train/dataset_info.json

@@ -0,0 +1,34 @@
+{
+  "citation": "",
+  "description": "",
+  "features": {
+    "text": {
+      "dtype": "string",
+      "_type": "Value"
+    },
+    "label": {
+      "dtype": "int64",
+      "_type": "Value"
+    },
+    "__index_level_0__": {
+      "dtype": "int64",
+      "_type": "Value"
+    },
+    "input_ids": {
+      "feature": {
+        "dtype": "int32",
+        "_type": "Value"
+      },
+      "_type": "Sequence"
+    },
+    "attention_mask": {
+      "feature": {
+        "dtype": "int8",
+        "_type": "Value"
+      },
+      "_type": "Sequence"
+    }
+  },
+  "homepage": "",
+  "license": ""
+}

output/version_1/tokenized_data_1/train/state.json

@@ -0,0 +1,13 @@
+{
+  "_data_files": [
+    {
+      "filename": "data-00000-of-00001.arrow"
+    }
+  ],
+  "_fingerprint": "2df21b01ebe5c573",
+  "_format_columns": null,
+  "_format_kwargs": {},
+  "_format_type": null,
+  "_output_all_columns": false,
+  "_split": null
+}

output/version_1/tokenized_data_1/validation/data-00000-of-00001.arrow

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbb3b43bd3b94caaa1c81b7b810cdc436ba8c53ab194f5d42dbca88072250901
+size 85550904

output/version_1/tokenized_data_1/validation/dataset_info.json

@@ -0,0 +1,34 @@
+{
+  "citation": "",
+  "description": "",
+  "features": {
+    "text": {
+      "dtype": "string",
+      "_type": "Value"
+    },
+    "label": {
+      "dtype": "int64",
+      "_type": "Value"
+    },
+    "__index_level_0__": {
+      "dtype": "int64",
+      "_type": "Value"
+    },
+    "input_ids": {
+      "feature": {
+        "dtype": "int32",
+        "_type": "Value"
+      },
+      "_type": "Sequence"
+    },
+    "attention_mask": {
+      "feature": {
+        "dtype": "int8",
+        "_type": "Value"
+      },
+      "_type": "Sequence"
+    }
+  },
+  "homepage": "",
+  "license": ""
+}

output/version_1/tokenized_data_1/validation/state.json

@@ -0,0 +1,13 @@
+{
+  "_data_files": [
+    {
+      "filename": "data-00000-of-00001.arrow"
+    }
+  ],
+  "_fingerprint": "b7d247f54cb3744e",
+  "_format_columns": null,
+  "_format_kwargs": {},
+  "_format_type": null,
+  "_output_all_columns": false,
+  "_split": null
+}

output/version_2/best_model_2/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "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,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "vocab_size": 30522
+}

output/version_2/best_model_2/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b74e608df383369c4cc72882dcd3cd0fc8bb5d1efe6db62cac025600eb10a373
+size 267832560

output/version_2/best_model_2/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54b30067bc4cf97ba30a27fab6d30aeb0b853f9518803eff0f2e5fc921fbc64d
+size 4664

output/version_2/checkpoint-10732/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "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,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "vocab_size": 30522
+}

output/version_2/checkpoint-10732/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b74e608df383369c4cc72882dcd3cd0fc8bb5d1efe6db62cac025600eb10a373
+size 267832560

output/version_2/checkpoint-10732/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea677cb0ca4e2f9fb5bbae481a72ebdbda422a11d76997615afa9922e6b6b40b
+size 535727290

output/version_2/checkpoint-10732/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:347298be764b8fbe652cc68e46d74117f75628d676f53163b896b05aae266f8d
+size 14244

output/version_2/checkpoint-10732/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8079a67ad0eb59ae5c2c133a1168c8a41c37afc809bc5131a7ce594a223b0829
+size 1064

output/version_2/checkpoint-10732/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54b30067bc4cf97ba30a27fab6d30aeb0b853f9518803eff0f2e5fc921fbc64d
+size 4664

output/version_2/checkpoint-13415/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "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,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "vocab_size": 30522
+}

output/version_2/checkpoint-13415/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:044240cadf20a9ef14dc302e9ace2a86973b4a882b3038b15f47f63e87245c74
+size 267832560

output/version_2/checkpoint-13415/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67d1125d30bd38f640bf7d67d5d280846dc6ab378a05717e7eaa70784b63810a
+size 535727290

output/version_2/checkpoint-13415/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af7304ea4c6cb4369ffcfd0e5c3a5868ceacdeee06f13af06022ef5df518b967
+size 14244