SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Clean Trainer tests and datasets dep (#8268)

Browse Source
This commit is contained in:
Sylvain Gugger
2020-11-03 15:50:55 -05:00
committed by GitHub
parent 068e6b5edd
commit 4c19f3baab
2 changed files with 18 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
.circleci/config.yml
View File

@@ -77,7 +77,6 @@ jobs:
- v0.4-torch_and_tf-{{ checksum "setup.py" }} - v0.4-torch_and_tf-{{ checksum "setup.py" }}
- v0.4-{{ checksum "setup.py" }} - v0.4-{{ checksum "setup.py" }}
- run: pip install --upgrade pip - run: pip install --upgrade pip
- run: pip install git+https://github.com/huggingface/datasets
- run: pip install .[sklearn,tf-cpu,torch,testing] - run: pip install .[sklearn,tf-cpu,torch,testing]
- save_cache: - save_cache:
key: v0.4-{{ checksum "setup.py" }} key: v0.4-{{ checksum "setup.py" }}
@@ -102,7 +101,6 @@ jobs:
- v0.4-torch-{{ checksum "setup.py" }} - v0.4-torch-{{ checksum "setup.py" }}
- v0.4-{{ checksum "setup.py" }} - v0.4-{{ checksum "setup.py" }}
- run: pip install --upgrade pip - run: pip install --upgrade pip
- run: pip install git+https://github.com/huggingface/datasets
- run: pip install .[sklearn,torch,testing] - run: pip install .[sklearn,torch,testing]
- save_cache: - save_cache:
key: v0.4-torch-{{ checksum "setup.py" }} key: v0.4-torch-{{ checksum "setup.py" }}
@@ -129,7 +127,6 @@ jobs:
- v0.4-tf-{{ checksum "setup.py" }} - v0.4-tf-{{ checksum "setup.py" }}
- v0.4-{{ checksum "setup.py" }} - v0.4-{{ checksum "setup.py" }}
- run: pip install --upgrade pip - run: pip install --upgrade pip
- run: pip install git+https://github.com/huggingface/datasets
- run: pip install .[sklearn,tf-cpu,testing] - run: pip install .[sklearn,tf-cpu,testing]
- save_cache: - save_cache:
key: v0.4-tf-{{ checksum "setup.py" }} key: v0.4-tf-{{ checksum "setup.py" }}
@@ -154,7 +151,6 @@ jobs:
- v0.4-flax-{{ checksum "setup.py" }} - v0.4-flax-{{ checksum "setup.py" }}
- v0.4-{{ checksum "setup.py" }} - v0.4-{{ checksum "setup.py" }}
- run: pip install --upgrade pip - run: pip install --upgrade pip
- run: pip install git+https://github.com/huggingface/datasets
- run: sudo pip install .[flax,sklearn,torch,testing] - run: sudo pip install .[flax,sklearn,torch,testing]
- save_cache: - save_cache:
key: v0.4-flax-{{ checksum "setup.py" }} key: v0.4-flax-{{ checksum "setup.py" }}
@@ -179,7 +175,6 @@ jobs:
- v0.4-torch-{{ checksum "setup.py" }} - v0.4-torch-{{ checksum "setup.py" }}
- v0.4-{{ checksum "setup.py" }} - v0.4-{{ checksum "setup.py" }}
- run: pip install --upgrade pip - run: pip install --upgrade pip
- run: pip install git+https://github.com/huggingface/datasets
- run: pip install .[sklearn,torch,testing] - run: pip install .[sklearn,torch,testing]
- save_cache: - save_cache:
key: v0.4-torch-{{ checksum "setup.py" }} key: v0.4-torch-{{ checksum "setup.py" }}
@@ -204,7 +199,6 @@ jobs:
- v0.4-tf-{{ checksum "setup.py" }} - v0.4-tf-{{ checksum "setup.py" }}
- v0.4-{{ checksum "setup.py" }} - v0.4-{{ checksum "setup.py" }}
- run: pip install --upgrade pip - run: pip install --upgrade pip
- run: pip install git+https://github.com/huggingface/datasets
- run: pip install .[sklearn,tf-cpu,testing] - run: pip install .[sklearn,tf-cpu,testing]
- save_cache: - save_cache:
key: v0.4-tf-{{ checksum "setup.py" }} key: v0.4-tf-{{ checksum "setup.py" }}

31
tests/test_trainer.py
View File

@@ -18,13 +18,13 @@ import os
import tempfile import tempfile
import unittest import unittest
import datasets
import numpy as np import numpy as np
from transformers import AutoTokenizer, EvaluationStrategy, PretrainedConfig, TrainingArguments, is_torch_available from transformers import AutoTokenizer, EvaluationStrategy, PretrainedConfig, TrainingArguments, is_torch_available
from transformers.file_utils import WEIGHTS_NAME from transformers.file_utils import WEIGHTS_NAME
from transformers.testing_utils import ( from transformers.testing_utils import (
get_tests_dir, get_tests_dir,
require_datasets,
require_optuna, require_optuna,
require_sentencepiece, require_sentencepiece,
require_tokenizers, require_tokenizers,
@@ -340,7 +340,10 @@ class TrainerIntegrationTest(unittest.TestCase):
self.assertTrue(np.array_equal(labels[0], trainer.eval_dataset.ys[0])) self.assertTrue(np.array_equal(labels[0], trainer.eval_dataset.ys[0]))
self.assertTrue(np.array_equal(labels[1], trainer.eval_dataset.ys[1])) self.assertTrue(np.array_equal(labels[1], trainer.eval_dataset.ys[1]))
@require_datasets
def test_trainer_with_datasets(self): def test_trainer_with_datasets(self):
import datasets
np.random.seed(42) np.random.seed(42)
x = np.random.normal(size=(64,)).astype(np.float32) x = np.random.normal(size=(64,)).astype(np.float32)
y = 2.0 * x + 3.0 + np.random.normal(scale=0.1, size=(64,)) y = 2.0 * x + 3.0 + np.random.normal(scale=0.1, size=(64,))
@@ -658,15 +661,17 @@ class TrainerHyperParameterIntegrationTest(unittest.TestCase):
def hp_name(trial): def hp_name(trial):
return MyTrialShortNamer.shortname(trial.params) return MyTrialShortNamer.shortname(trial.params)
trainer = get_regression_trainer( with tempfile.TemporaryDirectory() as tmp_dir:
learning_rate=0.1, trainer = get_regression_trainer(
logging_steps=1, output_dir=tmp_dir,
evaluation_strategy=EvaluationStrategy.EPOCH, learning_rate=0.1,
num_train_epochs=4, logging_steps=1,
disable_tqdm=True, evaluation_strategy=EvaluationStrategy.EPOCH,
load_best_model_at_end=True, num_train_epochs=4,
logging_dir="runs", disable_tqdm=True,
run_name="test", load_best_model_at_end=True,
model_init=model_init, logging_dir="runs",
) run_name="test",
trainer.hyperparameter_search(direction="minimize", hp_space=hp_space, hp_name=hp_name, n_trials=4) model_init=model_init,
)
trainer.hyperparameter_search(direction="minimize", hp_space=hp_space, hp_name=hp_name, n_trials=4)