注意
点击这里下载完整的示例代码
使用Ray Tune进行超参数调优
创建日期：2020年8月31日 | 最后更新：2024年10月31日 | 最后验证：2024年11月5日
超参数调优可以使一个普通模型和一个高精度模型之间产生显著差异。通常，像选择不同的学习率或改变网络层大小这样简单的事情，都会对模型性能产生巨大影响。
幸运的是，有一些工具可以帮助找到最佳参数组合。 Ray Tune 是一个行业标准的分布式超参数调优工具。Ray Tune 包含了最新的超参数搜索算法，与各种分析库集成，并通过 Ray 的分布式机器学习引擎 原生支持分布式训练。
在本教程中，我们将向您展示如何将Ray Tune集成到您的PyTorch训练工作流程中。我们将扩展PyTorch文档中的这个教程，用于训练一个CIFAR10图像分类器。
正如你将看到的，我们只需要进行一些轻微的修改。特别是，我们需要
将数据加载和训练封装在函数中，
使一些网络参数可配置，
添加检查点（可选），
并定义模型调优的搜索空间
要运行本教程，请确保已安装以下软件包：
ray[tune]: 分布式超参数调优库torchvision: 用于数据转换器
设置 / 导入
让我们从导入开始：
from functools import partial
import os
import tempfile
from pathlib import Path
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torch.utils.data import random_split
import torchvision
import torchvision.transforms as transforms
from ray import tune
from ray import train
from ray.train import Checkpoint, get_checkpoint
from ray.tune.schedulers import ASHAScheduler
import ray.cloudpickle as pickle
大多数导入是为了构建PyTorch模型。只有最后的导入是用于Ray Tune的。
数据加载器
我们将数据加载器封装在它们自己的函数中，并传递一个全局数据目录。 这样我们可以在不同的试验之间共享一个数据目录。
def load_data(data_dir="./data"):
transform = transforms.Compose(
[transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]
)
trainset = torchvision.datasets.CIFAR10(
root=data_dir, train=True, download=True, transform=transform
)
testset = torchvision.datasets.CIFAR10(
root=data_dir, train=False, download=True, transform=transform
)
return trainset, testset
可配置的神经网络
我们只能调整那些可配置的参数。 在这个例子中，我们可以指定 全连接层的层大小：
class Net(nn.Module):
def __init__(self, l1=120, l2=84):
super(Net, self).__init__()
self.conv1 = nn.Conv2d(3, 6, 5)
self.pool = nn.MaxPool2d(2, 2)
self.conv2 = nn.Conv2d(6, 16, 5)
self.fc1 = nn.Linear(16 * 5 * 5, l1)
self.fc2 = nn.Linear(l1, l2)
self.fc3 = nn.Linear(l2, 10)
def forward(self, x):
x = self.pool(F.relu(self.conv1(x)))
x = self.pool(F.relu(self.conv2(x)))
x = torch.flatten(x, 1) # flatten all dimensions except batch
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = self.fc3(x)
return x
训练函数
现在变得有趣了，因为我们引入了一些对示例的更改来自PyTorch文档。
我们将训练脚本封装在一个函数 train_cifar(config, data_dir=None) 中。
config 参数将接收我们想要训练的超参数。
data_dir 指定我们加载和存储数据的目录，以便多个运行可以共享相同的数据源。
如果提供了检查点，我们还会在运行开始时加载模型和优化器状态。在本教程的后面部分，您将找到有关如何保存检查点及其用途的信息。
net = Net(config["l1"], config["l2"])
checkpoint = get_checkpoint()
if checkpoint:
with checkpoint.as_directory() as checkpoint_dir:
data_path = Path(checkpoint_dir) / "data.pkl"
with open(data_path, "rb") as fp:
checkpoint_state = pickle.load(fp)
start_epoch = checkpoint_state["epoch"]
net.load_state_dict(checkpoint_state["net_state_dict"])
optimizer.load_state_dict(checkpoint_state["optimizer_state_dict"])
else:
start_epoch = 0
优化器的学习率也是可配置的：
optimizer = optim.SGD(net.parameters(), lr=config["lr"], momentum=0.9)
我们还将训练数据分为训练和验证子集。因此，我们在80%的数据上进行训练，并在剩余的20%上计算验证损失。我们遍历训练和测试集的批量大小也是可配置的。
使用DataParallel添加（多）GPU支持
图像分类在很大程度上受益于GPU。幸运的是，我们可以继续在Ray Tune中使用PyTorch的抽象。因此，我们可以将我们的模型包装在nn.DataParallel中，以支持在多个GPU上进行数据并行训练：
device = "cpu"
if torch.cuda.is_available():
device = "cuda:0"
if torch.cuda.device_count() > 1:
net = nn.DataParallel(net)
net.to(device)
通过使用device变量，我们确保在没有GPU可用时训练也能正常工作。PyTorch要求我们显式地将数据发送到GPU内存，如下所示：
for i, data in enumerate(trainloader, 0):
inputs, labels = data
inputs, labels = inputs.to(device), labels.to(device)
代码现在支持在CPU、单个GPU和多个GPU上进行训练。值得注意的是，Ray还支持部分GPU，因此只要模型仍然适合GPU内存，我们就可以在试验之间共享GPU。我们稍后会再讨论这个问题。
与Ray Tune通信
最有趣的部分是与Ray Tune的通信：
checkpoint_data = {
"epoch": epoch,
"net_state_dict": net.state_dict(),
"optimizer_state_dict": optimizer.state_dict(),
}
with tempfile.TemporaryDirectory() as checkpoint_dir:
data_path = Path(checkpoint_dir) / "data.pkl"
with open(data_path, "wb") as fp:
pickle.dump(checkpoint_data, fp)
checkpoint = Checkpoint.from_directory(checkpoint_dir)
train.report(
{"loss": val_loss / val_steps, "accuracy": correct / total},
checkpoint=checkpoint,
)
在这里，我们首先保存一个检查点，然后将一些指标报告回Ray Tune。具体来说，我们将验证损失和准确率发送回Ray Tune。Ray Tune随后可以使用这些指标来决定哪种超参数配置带来了最佳结果。这些指标也可以用于早期停止表现不佳的试验，以避免在这些试验上浪费资源。
检查点的保存是可选的，但是如果我们想要使用像基于人口的训练这样的高级调度器，则是必要的。此外，通过保存检查点，我们以后可以加载训练好的模型并在测试集上验证它们。最后，保存检查点对于容错非常有用，它允许我们中断训练并在以后继续训练。
完整训练函数
完整的代码示例如下所示：
def train_cifar(config, data_dir=None):
net = Net(config["l1"], config["l2"])
device = "cpu"
if torch.cuda.is_available():
device = "cuda:0"
if torch.cuda.device_count() > 1:
net = nn.DataParallel(net)
net.to(device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=config["lr"], momentum=0.9)
checkpoint = get_checkpoint()
if checkpoint:
with checkpoint.as_directory() as checkpoint_dir:
data_path = Path(checkpoint_dir) / "data.pkl"
with open(data_path, "rb") as fp:
checkpoint_state = pickle.load(fp)
start_epoch = checkpoint_state["epoch"]
net.load_state_dict(checkpoint_state["net_state_dict"])
optimizer.load_state_dict(checkpoint_state["optimizer_state_dict"])
else:
start_epoch = 0
trainset, testset = load_data(data_dir)
test_abs = int(len(trainset) * 0.8)
train_subset, val_subset = random_split(
trainset, [test_abs, len(trainset) - test_abs]
)
trainloader = torch.utils.data.DataLoader(
train_subset, batch_size=int(config["batch_size"]), shuffle=True, num_workers=8
)
valloader = torch.utils.data.DataLoader(
val_subset, batch_size=int(config["batch_size"]), shuffle=True, num_workers=8
)
for epoch in range(start_epoch, 10): # loop over the dataset multiple times
running_loss = 0.0
epoch_steps = 0
for i, data in enumerate(trainloader, 0):
# get the inputs; data is a list of [inputs, labels]
inputs, labels = data
inputs, labels = inputs.to(device), labels.to(device)
# zero the parameter gradients
optimizer.zero_grad()
# forward + backward + optimize
outputs = net(inputs)
loss = criterion(outputs, labels)
loss.backward()
optimizer.step()
# print statistics
running_loss += loss.item()
epoch_steps += 1
if i % 2000 == 1999: # print every 2000 mini-batches
print(
"[%d, %5d] loss: %.3f"
% (epoch + 1, i + 1, running_loss / epoch_steps)
)
running_loss = 0.0
# Validation loss
val_loss = 0.0
val_steps = 0
total = 0
correct = 0
for i, data in enumerate(valloader, 0):
with torch.no_grad():
inputs, labels = data
inputs, labels = inputs.to(device), labels.to(device)
outputs = net(inputs)
_, predicted = torch.max(outputs.data, 1)
total += labels.size(0)
correct += (predicted == labels).sum().item()
loss = criterion(outputs, labels)
val_loss += loss.cpu().numpy()
val_steps += 1
checkpoint_data = {
"epoch": epoch,
"net_state_dict": net.state_dict(),
"optimizer_state_dict": optimizer.state_dict(),
}
with tempfile.TemporaryDirectory() as checkpoint_dir:
data_path = Path(checkpoint_dir) / "data.pkl"
with open(data_path, "wb") as fp:
pickle.dump(checkpoint_data, fp)
checkpoint = Checkpoint.from_directory(checkpoint_dir)
train.report(
{"loss": val_loss / val_steps, "accuracy": correct / total},
checkpoint=checkpoint,
)
print("Finished Training")
正如你所见，大部分代码都是直接从原始示例中改编而来的。
测试集准确率
通常，机器学习模型的性能是在一个保留的测试集上进行测试的，该测试集包含未用于训练模型的数据。我们也将此封装在一个函数中：
def test_accuracy(net, device="cpu"):
trainset, testset = load_data()
testloader = torch.utils.data.DataLoader(
testset, batch_size=4, shuffle=False, num_workers=2
)
correct = 0
total = 0
with torch.no_grad():
for data in testloader:
images, labels = data
images, labels = images.to(device), labels.to(device)
outputs = net(images)
_, predicted = torch.max(outputs.data, 1)
total += labels.size(0)
correct += (predicted == labels).sum().item()
return correct / total
该函数还期望一个device参数，这样我们就可以在GPU上进行测试集验证。
配置搜索空间
最后，我们需要定义Ray Tune的搜索空间。以下是一个示例：
config = {
"l1": tune.choice([2 ** i for i in range(9)]),
"l2": tune.choice([2 ** i for i in range(9)]),
"lr": tune.loguniform(1e-4, 1e-1),
"batch_size": tune.choice([2, 4, 8, 16])
}
tune.choice() 接受一个值列表，这些值是从中均匀采样的。
在这个例子中，l1 和 l2 参数
应该是4到256之间的2的幂，即4、8、16、32、64、128或256。
lr（学习率）应该在0.0001到0.1之间均匀采样。最后，
批量大小是2、4、8和16之间的选择。
在每次试验中，Ray Tune 现在将从这些搜索空间中随机抽取一组参数。然后它将并行训练多个模型，并从中找出表现最好的一个。我们还使用了 ASHAScheduler，它会提前终止表现不佳的试验。
我们使用functools.partial来包装train_cifar函数，以设置常量data_dir参数。我们还可以告诉Ray Tune每个试验应该有哪些资源可用：
gpus_per_trial = 2
# ...
result = tune.run(
partial(train_cifar, data_dir=data_dir),
resources_per_trial={"cpu": 8, "gpu": gpus_per_trial},
config=config,
num_samples=num_samples,
scheduler=scheduler,
checkpoint_at_end=True)
您可以指定CPU的数量，这些CPU随后可用于增加PyTorch DataLoader实例的num_workers。在每个试验中，所选的GPU数量对PyTorch可见。试验无法访问未为其请求的GPU - 因此您不必担心两个试验使用相同的资源集。
在这里我们也可以指定部分GPU，所以像gpus_per_trial=0.5这样的设置是完全有效的。试验将共享GPU资源。你只需要确保模型仍然适合GPU内存。
在训练模型之后，我们将找到表现最好的一个，并从检查点文件加载训练好的网络。然后我们获得测试集的准确率，并通过打印报告所有内容。
完整的主函数如下所示：
def main(num_samples=10, max_num_epochs=10, gpus_per_trial=2):
data_dir = os.path.abspath("./data")
load_data(data_dir)
config = {
"l1": tune.choice([2**i for i in range(9)]),
"l2": tune.choice([2**i for i in range(9)]),
"lr": tune.loguniform(1e-4, 1e-1),
"batch_size": tune.choice([2, 4, 8, 16]),
}
scheduler = ASHAScheduler(
metric="loss",
mode="min",
max_t=max_num_epochs,
grace_period=1,
reduction_factor=2,
)
result = tune.run(
partial(train_cifar, data_dir=data_dir),
resources_per_trial={"cpu": 2, "gpu": gpus_per_trial},
config=config,
num_samples=num_samples,
scheduler=scheduler,
)
best_trial = result.get_best_trial("loss", "min", "last")
print(f"Best trial config: {best_trial.config}")
print(f"Best trial final validation loss: {best_trial.last_result['loss']}")
print(f"Best trial final validation accuracy: {best_trial.last_result['accuracy']}")
best_trained_model = Net(best_trial.config["l1"], best_trial.config["l2"])
device = "cpu"
if torch.cuda.is_available():
device = "cuda:0"
if gpus_per_trial > 1:
best_trained_model = nn.DataParallel(best_trained_model)
best_trained_model.to(device)
best_checkpoint = result.get_best_checkpoint(trial=best_trial, metric="accuracy", mode="max")
with best_checkpoint.as_directory() as checkpoint_dir:
data_path = Path(checkpoint_dir) / "data.pkl"
with open(data_path, "rb") as fp:
best_checkpoint_data = pickle.load(fp)
best_trained_model.load_state_dict(best_checkpoint_data["net_state_dict"])
test_acc = test_accuracy(best_trained_model, device)
print("Best trial test set accuracy: {}".format(test_acc))
if __name__ == "__main__":
# You can change the number of GPUs per trial here:
main(num_samples=10, max_num_epochs=10, gpus_per_trial=0)
Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to /var/lib/workspace/beginner_source/data/cifar-10-python.tar.gz
0% 0.00/170M [00:00<?, ?B/s]
0% 459k/170M [00:00<00:37, 4.53MB/s]
5% 7.96M/170M [00:00<00:03, 45.7MB/s]
11% 18.8M/170M [00:00<00:02, 74.4MB/s]
17% 29.5M/170M [00:00<00:01, 87.2MB/s]
24% 40.6M/170M [00:00<00:01, 95.7MB/s]
30% 51.7M/170M [00:00<00:01, 101MB/s]
37% 62.8M/170M [00:00<00:01, 104MB/s]
43% 73.9M/170M [00:00<00:00, 106MB/s]
50% 84.9M/170M [00:00<00:00, 107MB/s]
56% 96.0M/170M [00:01<00:00, 108MB/s]
63% 107M/170M [00:01<00:00, 109MB/s]
69% 118M/170M [00:01<00:00, 110MB/s]
76% 130M/170M [00:01<00:00, 111MB/s]
83% 141M/170M [00:01<00:00, 111MB/s]
89% 152M/170M [00:01<00:00, 111MB/s]
96% 163M/170M [00:01<00:00, 110MB/s]
100% 170M/170M [00:01<00:00, 102MB/s]
Extracting /var/lib/workspace/beginner_source/data/cifar-10-python.tar.gz to /var/lib/workspace/beginner_source/data
Files already downloaded and verified
2025-01-02 21:58:16,732 WARNING services.py:1889 -- WARNING: The object store is using /tmp instead of /dev/shm because /dev/shm has only 2147479552 bytes available. This will harm performance! You may be able to free up space by deleting files in /dev/shm. If you are inside a Docker container, you can increase /dev/shm size by passing '--shm-size=10.24gb' to 'docker run' (or add it to the run_options list in a Ray cluster config). Make sure to set this to more than 30% of available RAM.
2025-01-02 21:58:16,992 INFO worker.py:1642 -- Started a local Ray instance.
2025-01-02 21:58:18,354 INFO tune.py:228 -- Initializing Ray automatically. For cluster usage or custom Ray initialization, call `ray.init(...)` before `tune.run(...)`.
2025-01-02 21:58:18,356 INFO tune.py:654 -- [output] This will use the new output engine with verbosity 2. To disable the new output and use the legacy output engine, set the environment variable RAY_AIR_NEW_OUTPUT=0. For more information, please see https://github.com/ray-project/ray/issues/36949
+--------------------------------------------------------------------+
| Configuration for experiment train_cifar_2025-01-02_21-58-18 |
+--------------------------------------------------------------------+
| Search algorithm BasicVariantGenerator |
| Scheduler AsyncHyperBandScheduler |
| Number of trials 10 |
+--------------------------------------------------------------------+
View detailed results here: /var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18
To visualize your results with TensorBoard, run: `tensorboard --logdir /var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18`
Trial status: 10 PENDING
Current time: 2025-01-02 21:58:18. Total running time: 0s
Logical resource usage: 0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+-------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size |
+-------------------------------------------------------------------------------+
| train_cifar_acb57_00000 PENDING 16 1 0.00213327 2 |
| train_cifar_acb57_00001 PENDING 1 2 0.013416 4 |
| train_cifar_acb57_00002 PENDING 256 64 0.0113784 2 |
| train_cifar_acb57_00003 PENDING 64 256 0.0274071 8 |
| train_cifar_acb57_00004 PENDING 16 2 0.056666 4 |
| train_cifar_acb57_00005 PENDING 8 64 0.000353097 4 |
| train_cifar_acb57_00006 PENDING 16 4 0.000147684 8 |
| train_cifar_acb57_00007 PENDING 256 256 0.00477469 8 |
| train_cifar_acb57_00008 PENDING 128 256 0.0306227 8 |
| train_cifar_acb57_00009 PENDING 2 16 0.0286986 2 |
+-------------------------------------------------------------------------------+
Trial train_cifar_acb57_00007 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00007 config |
+--------------------------------------------------+
| batch_size 8 |
| l1 256 |
| l2 256 |
| lr 0.00477 |
+--------------------------------------------------+
Trial train_cifar_acb57_00001 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00001 config |
+--------------------------------------------------+
| batch_size 4 |
| l1 1 |
| l2 2 |
| lr 0.01342 |
+--------------------------------------------------+
Trial train_cifar_acb57_00003 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00003 config |
+--------------------------------------------------+
| batch_size 8 |
| l1 64 |
| l2 256 |
| lr 0.02741 |
+--------------------------------------------------+
Trial train_cifar_acb57_00000 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00000 config |
+--------------------------------------------------+
| batch_size 2 |
| l1 16 |
| l2 1 |
| lr 0.00213 |
+--------------------------------------------------+
Trial train_cifar_acb57_00002 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00002 config |
+--------------------------------------------------+
| batch_size 2 |
| l1 256 |
| l2 64 |
| lr 0.01138 |
+--------------------------------------------------+
Trial train_cifar_acb57_00004 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00004 config |
+--------------------------------------------------+
| batch_size 4 |
| l1 16 |
| l2 2 |
| lr 0.05667 |
+--------------------------------------------------+
(func pid=4869) Files already downloaded and verified
Trial train_cifar_acb57_00006 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00006 config |
+--------------------------------------------------+
| batch_size 8 |
| l1 16 |
| l2 4 |
| lr 0.00015 |
+--------------------------------------------------+
Trial train_cifar_acb57_00005 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00005 config |
+--------------------------------------------------+
| batch_size 4 |
| l1 8 |
| l2 64 |
| lr 0.00035 |
+--------------------------------------------------+
(func pid=4868) [1, 2000] loss: 2.321
(func pid=4886) Files already downloaded and verified [repeated 15x across cluster] (Ray deduplicates logs by default. Set RAY_DEDUP_LOGS=0 to disable log deduplication, or see https://docs.ray.io/en/master/ray-observability/ray-logging.html#log-deduplication for more options.)
Trial status: 8 RUNNING | 2 PENDING
Current time: 2025-01-02 21:58:48. Total running time: 30s
Logical resource usage: 16.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+-------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size |
+-------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 |
| train_cifar_acb57_00001 RUNNING 1 2 0.013416 4 |
| train_cifar_acb57_00002 RUNNING 256 64 0.0113784 2 |
| train_cifar_acb57_00003 RUNNING 64 256 0.0274071 8 |
| train_cifar_acb57_00004 RUNNING 16 2 0.056666 4 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 |
| train_cifar_acb57_00006 RUNNING 16 4 0.000147684 8 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 |
| train_cifar_acb57_00008 PENDING 128 256 0.0306227 8 |
| train_cifar_acb57_00009 PENDING 2 16 0.0286986 2 |
+-------------------------------------------------------------------------------+
(func pid=4868) [1, 4000] loss: 1.153 [repeated 8x across cluster]
(func pid=4871) [1, 4000] loss: 1.047 [repeated 7x across cluster]
(func pid=4868) [1, 6000] loss: 0.768
(func pid=4869) [1, 6000] loss: 0.770
Trial train_cifar_acb57_00007 finished iteration 1 at 2025-01-02 21:59:18. Total running time: 1min 0s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 54.19759 |
| time_total_s 54.19759 |
| training_iteration 1 |
| accuracy 0.4812 |
| loss 1.46991 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000000
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000000)
Trial status: 8 RUNNING | 2 PENDING
Current time: 2025-01-02 21:59:18. Total running time: 1min 0s
Logical resource usage: 16.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+----------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+----------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 |
| train_cifar_acb57_00001 RUNNING 1 2 0.013416 4 |
| train_cifar_acb57_00002 RUNNING 256 64 0.0113784 2 |
| train_cifar_acb57_00003 RUNNING 64 256 0.0274071 8 |
| train_cifar_acb57_00004 RUNNING 16 2 0.056666 4 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 |
| train_cifar_acb57_00006 RUNNING 16 4 0.000147684 8 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 1 54.1976 1.46991 0.4812 |
| train_cifar_acb57_00008 PENDING 128 256 0.0306227 8 |
| train_cifar_acb57_00009 PENDING 2 16 0.0286986 2 |
+----------------------------------------------------------------------------------------------------------------------------------+
Trial train_cifar_acb57_00006 finished iteration 1 at 2025-01-02 21:59:18. Total running time: 1min 0s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00006 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 53.9304 |
| time_total_s 53.9304 |
| training_iteration 1 |
| accuracy 0.1185 |
| loss 2.30605 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00006 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00006_6_batch_size=8,l1=16,l2=4,lr=0.0001_2025-01-02_21-58-18/checkpoint_000000
Trial train_cifar_acb57_00006 completed after 1 iterations at 2025-01-02 21:59:18. Total running time: 1min 0s
Trial train_cifar_acb57_00008 started with configuration:
+--------------------------------------------------+
| Trial train_cifar_acb57_00008 config |
+--------------------------------------------------+
| batch_size 8 |
| l1 128 |
| l2 256 |
| lr 0.03062 |
+--------------------------------------------------+
(func pid=4886) Files already downloaded and verified
Trial train_cifar_acb57_00003 finished iteration 1 at 2025-01-02 21:59:20. Total running time: 1min 1s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00003 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 55.97928 |
| time_total_s 55.97928 |
| training_iteration 1 |
| accuracy 0.2109 |
| loss 2.082 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00003 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00003_3_batch_size=8,l1=64,l2=256,lr=0.0274_2025-01-02_21-58-18/checkpoint_000000
Trial train_cifar_acb57_00003 completed after 1 iterations at 2025-01-02 21:59:20. Total running time: 1min 1s
Trial train_cifar_acb57_00009 started with configuration:
+-------------------------------------------------+
| Trial train_cifar_acb57_00009 config |
+-------------------------------------------------+
| batch_size 2 |
| l1 2 |
| l2 16 |
| lr 0.0287 |
+-------------------------------------------------+
(func pid=4886) Files already downloaded and verified
(func pid=4870) [1, 6000] loss: 0.734 [repeated 3x across cluster]
(func pid=4871) Files already downloaded and verified [repeated 2x across cluster]
(func pid=4868) [1, 8000] loss: 0.576
(func pid=4869) [1, 8000] loss: 0.577
(func pid=4887) [2, 2000] loss: 1.389 [repeated 4x across cluster]
(func pid=4868) [1, 10000] loss: 0.441 [repeated 3x across cluster]
(func pid=4872) [1, 10000] loss: 0.467 [repeated 3x across cluster]
Trial status: 8 RUNNING | 2 TERMINATED
Current time: 2025-01-02 21:59:48. Total running time: 1min 30s
Logical resource usage: 16.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 |
| train_cifar_acb57_00001 RUNNING 1 2 0.013416 4 |
| train_cifar_acb57_00002 RUNNING 256 64 0.0113784 2 |
| train_cifar_acb57_00004 RUNNING 16 2 0.056666 4 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 1 54.1976 1.46991 0.4812 |
| train_cifar_acb57_00008 RUNNING 128 256 0.0306227 8 |
| train_cifar_acb57_00009 RUNNING 2 16 0.0286986 2 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
+------------------------------------------------------------------------------------------------------------------------------------+
(func pid=4871) [1, 4000] loss: 1.169
(func pid=4870) [1, 10000] loss: 0.463
Trial train_cifar_acb57_00005 finished iteration 1 at 2025-01-02 22:00:00. Total running time: 1min 41s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00005 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 95.20012 |
| time_total_s 95.20012 |
| training_iteration 1 |
| accuracy 0.3406 |
| loss 1.74255 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00005 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000000
(func pid=4885) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000000) [repeated 3x across cluster]
Trial train_cifar_acb57_00001 finished iteration 1 at 2025-01-02 22:00:00. Total running time: 1min 42s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00001 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 96.32778 |
| time_total_s 96.32778 |
| training_iteration 1 |
| accuracy 0.0989 |
| loss 2.30402 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00001 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00001_1_batch_size=4,l1=1,l2=2,lr=0.0134_2025-01-02_21-58-18/checkpoint_000000
Trial train_cifar_acb57_00001 completed after 1 iterations at 2025-01-02 22:00:00. Total running time: 1min 42s
Trial train_cifar_acb57_00004 finished iteration 1 at 2025-01-02 22:00:01. Total running time: 1min 42s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00004 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 96.47434 |
| time_total_s 96.47434 |
| training_iteration 1 |
| accuracy 0.101 |
| loss 2.33135 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00004 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00004_4_batch_size=4,l1=16,l2=2,lr=0.0567_2025-01-02_21-58-18/checkpoint_000000
Trial train_cifar_acb57_00004 completed after 1 iterations at 2025-01-02 22:00:01. Total running time: 1min 42s
(func pid=4871) [1, 6000] loss: 0.777 [repeated 4x across cluster]
Trial train_cifar_acb57_00007 finished iteration 2 at 2025-01-02 22:00:08. Total running time: 1min 50s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000001 |
| time_this_iter_s 50.0624 |
| time_total_s 104.25999 |
| training_iteration 2 |
| accuracy 0.5473 |
| loss 1.28808 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 2 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000001
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000001) [repeated 3x across cluster]
Trial train_cifar_acb57_00008 finished iteration 1 at 2025-01-02 22:00:10. Total running time: 1min 52s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00008 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 51.9493 |
| time_total_s 51.9493 |
| training_iteration 1 |
| accuracy 0.2172 |
| loss 2.05322 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00008 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00008_8_batch_size=8,l1=128,l2=256,lr=0.0306_2025-01-02_21-58-18/checkpoint_000000
(func pid=4885) [2, 2000] loss: 1.727 [repeated 3x across cluster]
(func pid=4871) [1, 8000] loss: 0.584
Trial status: 6 RUNNING | 4 TERMINATED
Current time: 2025-01-02 22:00:18. Total running time: 2min 0s
Logical resource usage: 12.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 |
| train_cifar_acb57_00002 RUNNING 256 64 0.0113784 2 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 1 95.2001 1.74255 0.3406 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 2 104.26 1.28808 0.5473 |
| train_cifar_acb57_00008 RUNNING 128 256 0.0306227 8 1 51.9493 2.05322 0.2172 |
| train_cifar_acb57_00009 RUNNING 2 16 0.0286986 2 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
+------------------------------------------------------------------------------------------------------------------------------------+
(func pid=4868) [1, 16000] loss: 0.247
(func pid=4871) [1, 10000] loss: 0.467 [repeated 5x across cluster]
(func pid=4885) [2, 6000] loss: 0.536 [repeated 2x across cluster]
(func pid=4868) [1, 20000] loss: 0.197 [repeated 5x across cluster]
Trial status: 6 RUNNING | 4 TERMINATED
Current time: 2025-01-02 22:00:48. Total running time: 2min 30s
Logical resource usage: 12.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 |
| train_cifar_acb57_00002 RUNNING 256 64 0.0113784 2 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 1 95.2001 1.74255 0.3406 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 2 104.26 1.28808 0.5473 |
| train_cifar_acb57_00008 RUNNING 128 256 0.0306227 8 1 51.9493 2.05322 0.2172 |
| train_cifar_acb57_00009 RUNNING 2 16 0.0286986 2 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
+------------------------------------------------------------------------------------------------------------------------------------+
Trial train_cifar_acb57_00007 finished iteration 3 at 2025-01-02 22:00:48. Total running time: 2min 30s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000002 |
| time_this_iter_s 40.35045 |
| time_total_s 144.61044 |
| training_iteration 3 |
| accuracy 0.5602 |
| loss 1.258 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 3 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000002
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000002) [repeated 2x across cluster]
(func pid=4870) [1, 18000] loss: 0.257 [repeated 2x across cluster]
Trial train_cifar_acb57_00008 finished iteration 2 at 2025-01-02 22:00:53. Total running time: 2min 34s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00008 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000001 |
| time_this_iter_s 42.5872 |
| time_total_s 94.5365 |
| training_iteration 2 |
| accuracy 0.2199 |
| loss 2.0719 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00008 saved a checkpoint for iteration 2 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00008_8_batch_size=8,l1=128,l2=256,lr=0.0306_2025-01-02_21-58-18/checkpoint_000001
Trial train_cifar_acb57_00008 completed after 2 iterations at 2025-01-02 22:00:53. Total running time: 2min 34s
(func pid=4885) [2, 10000] loss: 0.305 [repeated 2x across cluster]
Trial train_cifar_acb57_00000 finished iteration 1 at 2025-01-02 22:01:01. Total running time: 2min 42s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00000 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 156.91532 |
| time_total_s 156.91532 |
| training_iteration 1 |
| accuracy 0.2024 |
| loss 1.95374 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00000 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00000_0_batch_size=2,l1=16,l2=1,lr=0.0021_2025-01-02_21-58-18/checkpoint_000000
(func pid=4868) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00000_0_batch_size=2,l1=16,l2=1,lr=0.0021_2025-01-02_21-58-18/checkpoint_000000) [repeated 2x across cluster]
(func pid=4870) [1, 20000] loss: 0.232 [repeated 3x across cluster]
Trial train_cifar_acb57_00005 finished iteration 2 at 2025-01-02 22:01:08. Total running time: 2min 49s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00005 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000001 |
| time_this_iter_s 68.21913 |
| time_total_s 163.41925 |
| training_iteration 2 |
| accuracy 0.449 |
| loss 1.50449 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00005 saved a checkpoint for iteration 2 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000001
(func pid=4885) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000001)
(func pid=4868) [2, 2000] loss: 1.954
(func pid=4871) [1, 18000] loss: 0.260
Trial status: 5 RUNNING | 5 TERMINATED
Current time: 2025-01-02 22:01:18. Total running time: 3min 0s
Logical resource usage: 10.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 1 156.915 1.95374 0.2024 |
| train_cifar_acb57_00002 RUNNING 256 64 0.0113784 2 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 2 163.419 1.50449 0.449 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 3 144.61 1.258 0.5602 |
| train_cifar_acb57_00009 RUNNING 2 16 0.0286986 2 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
+------------------------------------------------------------------------------------------------------------------------------------+
(func pid=4885) [3, 2000] loss: 1.484 [repeated 2x across cluster]
Trial train_cifar_acb57_00002 finished iteration 1 at 2025-01-02 22:01:21. Total running time: 3min 3s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00002 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 177.56105 |
| time_total_s 177.56105 |
| training_iteration 1 |
| accuracy 0.0985 |
| loss 2.3243 |
+------------------------------------------------------------+
(func pid=4870) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00002_2_batch_size=2,l1=256,l2=64,lr=0.0114_2025-01-02_21-58-18/checkpoint_000000)
Trial train_cifar_acb57_00002 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00002_2_batch_size=2,l1=256,l2=64,lr=0.0114_2025-01-02_21-58-18/checkpoint_000000
Trial train_cifar_acb57_00002 completed after 1 iterations at 2025-01-02 22:01:21. Total running time: 3min 3s
(func pid=4871) [1, 20000] loss: 0.233 [repeated 2x across cluster]
Trial train_cifar_acb57_00007 finished iteration 4 at 2025-01-02 22:01:26. Total running time: 3min 8s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000003 |
| time_this_iter_s 37.85795 |
| time_total_s 182.46839 |
| training_iteration 4 |
| accuracy 0.5818 |
| loss 1.22092 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 4 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000003
(func pid=4885) [3, 4000] loss: 0.735
(func pid=4868) [2, 6000] loss: 0.645
(func pid=4887) [5, 2000] loss: 1.068
Trial train_cifar_acb57_00009 finished iteration 1 at 2025-01-02 22:01:39. Total running time: 3min 21s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00009 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000000 |
| time_this_iter_s 139.43723 |
| time_total_s 139.43723 |
| training_iteration 1 |
| accuracy 0.1015 |
| loss 2.3257 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00009 saved a checkpoint for iteration 1 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00009_9_batch_size=2,l1=2,l2=16,lr=0.0287_2025-01-02_21-58-18/checkpoint_000000
Trial train_cifar_acb57_00009 completed after 1 iterations at 2025-01-02 22:01:39. Total running time: 3min 21s
(func pid=4871) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00009_9_batch_size=2,l1=2,l2=16,lr=0.0287_2025-01-02_21-58-18/checkpoint_000000) [repeated 2x across cluster]
(func pid=4885) [3, 6000] loss: 0.487
Trial status: 3 RUNNING | 7 TERMINATED
Current time: 2025-01-02 22:01:48. Total running time: 3min 30s
Logical resource usage: 6.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 1 156.915 1.95374 0.2024 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 2 163.419 1.50449 0.449 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 4 182.468 1.22092 0.5818 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
(func pid=4885) [3, 8000] loss: 0.357 [repeated 2x across cluster]
(func pid=4885) [3, 10000] loss: 0.284 [repeated 3x across cluster]
Trial train_cifar_acb57_00007 finished iteration 5 at 2025-01-02 22:01:59. Total running time: 3min 41s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000004 |
| time_this_iter_s 32.97702 |
| time_total_s 215.44542 |
| training_iteration 5 |
| accuracy 0.5554 |
| loss 1.30401 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 5 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000004
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000004)
Trial train_cifar_acb57_00005 finished iteration 3 at 2025-01-02 22:02:06. Total running time: 3min 48s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00005 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000002 |
| time_this_iter_s 58.1904 |
| time_total_s 221.60965 |
| training_iteration 3 |
| accuracy 0.4754 |
| loss 1.45713 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00005 saved a checkpoint for iteration 3 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000002
(func pid=4885) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000002)
(func pid=4868) [2, 14000] loss: 0.276 [repeated 2x across cluster]
(func pid=4885) [4, 2000] loss: 1.381 [repeated 2x across cluster]
Trial status: 3 RUNNING | 7 TERMINATED
Current time: 2025-01-02 22:02:18. Total running time: 4min 0s
Logical resource usage: 6.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 1 156.915 1.95374 0.2024 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 3 221.61 1.45713 0.4754 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 5 215.445 1.30401 0.5554 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
(func pid=4887) [6, 4000] loss: 0.534 [repeated 2x across cluster]
Trial train_cifar_acb57_00007 finished iteration 6 at 2025-01-02 22:02:32. Total running time: 4min 13s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000005 |
| time_this_iter_s 32.56537 |
| time_total_s 248.01078 |
| training_iteration 6 |
| accuracy 0.5841 |
| loss 1.24011 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 6 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000005
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000005)
(func pid=4885) [4, 6000] loss: 0.450 [repeated 3x across cluster]
(func pid=4887) [7, 2000] loss: 0.965 [repeated 2x across cluster]
Trial status: 3 RUNNING | 7 TERMINATED
Current time: 2025-01-02 22:02:48. Total running time: 4min 30s
Logical resource usage: 6.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 RUNNING 16 1 0.00213327 2 1 156.915 1.95374 0.2024 |
| train_cifar_acb57_00005 RUNNING 8 64 0.000353097 4 3 221.61 1.45713 0.4754 |
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 6 248.011 1.24011 0.5841 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
Trial train_cifar_acb57_00000 finished iteration 2 at 2025-01-02 22:02:49. Total running time: 4min 30s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00000 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000001 |
| time_this_iter_s 107.74446 |
| time_total_s 264.65979 |
| training_iteration 2 |
| accuracy 0.2192 |
| loss 1.91832 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00000 saved a checkpoint for iteration 2 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00000_0_batch_size=2,l1=16,l2=1,lr=0.0021_2025-01-02_21-58-18/checkpoint_000001
Trial train_cifar_acb57_00000 completed after 2 iterations at 2025-01-02 22:02:49. Total running time: 4min 30s
(func pid=4868) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00000_0_batch_size=2,l1=16,l2=1,lr=0.0021_2025-01-02_21-58-18/checkpoint_000001)
(func pid=4885) [4, 10000] loss: 0.266 [repeated 2x across cluster]
Trial train_cifar_acb57_00005 finished iteration 4 at 2025-01-02 22:02:59. Total running time: 4min 41s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00005 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000003 |
| time_this_iter_s 53.03366 |
| time_total_s 274.64331 |
| training_iteration 4 |
| accuracy 0.5127 |
| loss 1.37224 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00005 saved a checkpoint for iteration 4 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000003
Trial train_cifar_acb57_00005 completed after 4 iterations at 2025-01-02 22:02:59. Total running time: 4min 41s
(func pid=4885) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00005_5_batch_size=4,l1=8,l2=64,lr=0.0004_2025-01-02_21-58-18/checkpoint_000003)
Trial train_cifar_acb57_00007 finished iteration 7 at 2025-01-02 22:03:02. Total running time: 4min 44s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000006 |
| time_this_iter_s 30.3697 |
| time_total_s 278.38048 |
| training_iteration 7 |
| accuracy 0.5745 |
| loss 1.33752 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 7 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000006
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000006)
(func pid=4887) [8, 2000] loss: 0.951 [repeated 2x across cluster]
Trial status: 9 TERMINATED | 1 RUNNING
Current time: 2025-01-02 22:03:18. Total running time: 5min 0s
Logical resource usage: 2.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 7 278.38 1.33752 0.5745 |
| train_cifar_acb57_00000 TERMINATED 16 1 0.00213327 2 2 264.66 1.91832 0.2192 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00005 TERMINATED 8 64 0.000353097 4 4 274.643 1.37224 0.5127 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
(func pid=4887) [8, 4000] loss: 0.502
Trial train_cifar_acb57_00007 finished iteration 8 at 2025-01-02 22:03:29. Total running time: 5min 10s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000007 |
| time_this_iter_s 26.38767 |
| time_total_s 304.76815 |
| training_iteration 8 |
| accuracy 0.5755 |
| loss 1.28171 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 8 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000007
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000007)
(func pid=4887) [9, 2000] loss: 0.934
(func pid=4887) [9, 4000] loss: 0.492
Trial status: 9 TERMINATED | 1 RUNNING
Current time: 2025-01-02 22:03:49. Total running time: 5min 30s
Logical resource usage: 2.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 8 304.768 1.28171 0.5755 |
| train_cifar_acb57_00000 TERMINATED 16 1 0.00213327 2 2 264.66 1.91832 0.2192 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00005 TERMINATED 8 64 0.000353097 4 4 274.643 1.37224 0.5127 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
Trial train_cifar_acb57_00007 finished iteration 9 at 2025-01-02 22:03:55. Total running time: 5min 37s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000008 |
| time_this_iter_s 26.51591 |
| time_total_s 331.28406 |
| training_iteration 9 |
| accuracy 0.5687 |
| loss 1.35061 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 9 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000008
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000008)
(func pid=4887) [10, 2000] loss: 0.893
(func pid=4887) [10, 4000] loss: 0.492
Trial status: 9 TERMINATED | 1 RUNNING
Current time: 2025-01-02 22:04:19. Total running time: 6min 0s
Logical resource usage: 2.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00007 RUNNING 256 256 0.00477469 8 9 331.284 1.35061 0.5687 |
| train_cifar_acb57_00000 TERMINATED 16 1 0.00213327 2 2 264.66 1.91832 0.2192 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00005 TERMINATED 8 64 0.000353097 4 4 274.643 1.37224 0.5127 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
Trial train_cifar_acb57_00007 finished iteration 10 at 2025-01-02 22:04:21. Total running time: 6min 3s
+------------------------------------------------------------+
| Trial train_cifar_acb57_00007 result |
+------------------------------------------------------------+
| checkpoint_dir_name checkpoint_000009 |
| time_this_iter_s 26.11893 |
| time_total_s 357.40299 |
| training_iteration 10 |
| accuracy 0.5642 |
| loss 1.3626 |
+------------------------------------------------------------+
Trial train_cifar_acb57_00007 saved a checkpoint for iteration 10 at: (local)/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000009
Trial train_cifar_acb57_00007 completed after 10 iterations at 2025-01-02 22:04:21. Total running time: 6min 3s
Trial status: 10 TERMINATED
Current time: 2025-01-02 22:04:21. Total running time: 6min 3s
Logical resource usage: 2.0/16 CPUs, 0/1 GPUs (0.0/1.0 accelerator_type:M60)
+------------------------------------------------------------------------------------------------------------------------------------+
| Trial name status l1 l2 lr batch_size iter total time (s) loss accuracy |
+------------------------------------------------------------------------------------------------------------------------------------+
| train_cifar_acb57_00000 TERMINATED 16 1 0.00213327 2 2 264.66 1.91832 0.2192 |
| train_cifar_acb57_00001 TERMINATED 1 2 0.013416 4 1 96.3278 2.30402 0.0989 |
| train_cifar_acb57_00002 TERMINATED 256 64 0.0113784 2 1 177.561 2.3243 0.0985 |
| train_cifar_acb57_00003 TERMINATED 64 256 0.0274071 8 1 55.9793 2.082 0.2109 |
| train_cifar_acb57_00004 TERMINATED 16 2 0.056666 4 1 96.4743 2.33135 0.101 |
| train_cifar_acb57_00005 TERMINATED 8 64 0.000353097 4 4 274.643 1.37224 0.5127 |
| train_cifar_acb57_00006 TERMINATED 16 4 0.000147684 8 1 53.9304 2.30605 0.1185 |
| train_cifar_acb57_00007 TERMINATED 256 256 0.00477469 8 10 357.403 1.3626 0.5642 |
| train_cifar_acb57_00008 TERMINATED 128 256 0.0306227 8 2 94.5365 2.0719 0.2199 |
| train_cifar_acb57_00009 TERMINATED 2 16 0.0286986 2 1 139.437 2.3257 0.1015 |
+------------------------------------------------------------------------------------------------------------------------------------+
Best trial config: {'l1': 256, 'l2': 256, 'lr': 0.00477468908087826, 'batch_size': 8}
Best trial final validation loss: 1.362598385155201
Best trial final validation accuracy: 0.5642
(func pid=4887) Checkpoint successfully created at: Checkpoint(filesystem=local, path=/var/lib/ci-user/ray_results/train_cifar_2025-01-02_21-58-18/train_cifar_acb57_00007_7_batch_size=8,l1=256,l2=256,lr=0.0048_2025-01-02_21-58-18/checkpoint_000009)
Files already downloaded and verified
Files already downloaded and verified
Best trial test set accuracy: 0.5794
如果你运行代码，一个示例输出可能看起来像这样：
Number of trials: 10/10 (10 TERMINATED)
+-----+--------------+------+------+-------------+--------+---------+------------+
| ... | batch_size | l1 | l2 | lr | iter | loss | accuracy |
|-----+--------------+------+------+-------------+--------+---------+------------|
| ... | 2 | 1 | 256 | 0.000668163 | 1 | 2.31479 | 0.0977 |
| ... | 4 | 64 | 8 | 0.0331514 | 1 | 2.31605 | 0.0983 |
| ... | 4 | 2 | 1 | 0.000150295 | 1 | 2.30755 | 0.1023 |
| ... | 16 | 32 | 32 | 0.0128248 | 10 | 1.66912 | 0.4391 |
| ... | 4 | 8 | 128 | 0.00464561 | 2 | 1.7316 | 0.3463 |
| ... | 8 | 256 | 8 | 0.00031556 | 1 | 2.19409 | 0.1736 |
| ... | 4 | 16 | 256 | 0.00574329 | 2 | 1.85679 | 0.3368 |
| ... | 8 | 2 | 2 | 0.00325652 | 1 | 2.30272 | 0.0984 |
| ... | 2 | 2 | 2 | 0.000342987 | 2 | 1.76044 | 0.292 |
| ... | 4 | 64 | 32 | 0.003734 | 8 | 1.53101 | 0.4761 |
+-----+--------------+------+------+-------------+--------+---------+------------+
Best trial config: {'l1': 64, 'l2': 32, 'lr': 0.0037339984519545164, 'batch_size': 4}
Best trial final validation loss: 1.5310075663924216
Best trial final validation accuracy: 0.4761
Best trial test set accuracy: 0.4737
大多数试验已经提前停止，以避免浪费资源。 表现最好的试验达到了约47%的验证准确率，这可以在测试集上得到确认。
就是这样！你现在可以调整你的PyTorch模型的参数了。
脚本总运行时间： (6分钟 20.572秒)
