Using Multi-layer Perceptron and Convolutional Neural Networks for Satellite image classification - 2023

Antonio Fonseca

Packages to be installed:

conda install -c conda-forge umap-learn
pip install phate
conda install -c conda-forge imageio
pip install wandb
[ ]:
import numpy as np
import codecs
import copy
import json
import scipy.io
from scipy.spatial.distance import cdist, pdist, squareform
from scipy.linalg import eigh
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans
import random
from sklearn import manifold
import os
# import phate
# import umap
import pandas as pd
# import scprep
from torch.nn import functional as F


import pandas as pd
from sklearn.metrics import r2_score
from sklearn.preprocessing import MinMaxScaler
# import seaborn as sns

import torch
from torch.utils.data import Dataset, DataLoader
from torch.utils.data.sampler import SubsetRandomSampler,RandomSampler
from torchvision import datasets, transforms
from torch.nn.functional import softmax
from torch import optim, nn
import torchvision
import torchvision.transforms as transforms
import torchvision.datasets as datasets
import time

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(device)
/gpfs/gibbs/project/dijk/ahf38/conda_envs/geo_comp2/lib/python3.8/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
  from .autonotebook import tqdm as notebook_tqdm
/gpfs/gibbs/project/dijk/ahf38/conda_envs/geo_comp2/lib/python3.8/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: /gpfs/gibbs/project/dijk/ahf38/conda_envs/geo_comp2/lib/python3.8/site-packages/torchvision/image.so: undefined symbol: _ZN5torch3jit17parseSchemaOrNameERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE
  warn(f"Failed to load image Python extension: {e}")
cuda

Now that we have an autoencoder working on MNIST, let’s use this model to visualize some geodata. For the next section we will use the SAT-6 (https://csc.lsu.edu/~saikat/deepsat/)

SAT-6 consists of a total of 405,000 image patches each of size 28x28 and covering 6 landcover classes - barren land, trees, grassland, roads, buildings and water bodies. 324,000 images (comprising of four-fifths of the total dataset) were chosen as the training dataset and 81,000 (one fifths) were chosen as the testing dataset. Similar to SAT-4, the training and test sets were selected from disjoint NAIP tiles. Once generated, the images in the dataset were randomized in the same way as that for SAT-4. The specifications for the various landcover classes of SAT-4 and SAT-6 were adopted from those used in the National Land Cover Data (NLCD) algorithm.

The datasets are encoded as MATLAB .mat files that can be read using the standard load command in MATLAB. Each sample image is 28x28 pixels and consists of 4 bands - red, green, blue and near infrared. The training and test labels are 1x4 and 1x6 vectors for SAT-4 and SAT-6 respectively having a single 1 indexing a particular class from 0 through 4 or 6 and 0 values at all other indices.

The MAT file for the SAT-6 dataset contains the following variables:

  • train_x 28x28x4x324000 uint8 (containing 324000 training samples of 28x28 images each with 4 channels)

  • train_y 324000x6 uint8 (containing 6x1 vectors having labels for the 324000 training samples)

  • test_x 28x28x4x81000 uint8 (containing 81000 test samples of 28x28 images each with 4 channels)

  • test_y 81000x6 uint8 (containing 6x1 vectors having labels for the 81000 test samples)

Labels: - Building = 0 - Barren_land = 1 - Tree=2 - Grassland=3 - Road = 4 - Water = 5

[ ]:
import numpy as np
import scipy.io
import matplotlib.pyplot as plt
import torch
from torch import optim, nn
import wandb
import datetime

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(device)
cuda
[ ]:
# Using the satelite images dataset
###############################################################################
#load the data
data = scipy.io.loadmat("./SAT-4_and_SAT-6_datasets/sat-6-full.mat")
train_images = data['train_x']
train_labels = data['train_y']

test_images = data['test_x']
test_labels = data['test_y']
[ ]:

[ ]:
####################################################################
#Checkout the data
print('Training data shape : ', train_images.shape, train_labels.shape)
print('Testing data shape : ', test_images.shape, test_labels.shape)
Training data shape :  (28, 28, 4, 324000) (6, 324000)
Testing data shape :  (28, 28, 4, 81000) (6, 81000)
[ ]:
#Change the dimension to fit into the model
x_train = train_images.transpose(3,0,1,2)
t_train = train_labels.transpose()

# x_test = test_images.transpose(3,0,1,2)
# t_test = test_labels.transpose()
print('Training data shape : ', x_train.shape, t_train.shape)
# print('Testing data shape : ', x_test.shape, t_test.shape)
Training data shape :  (324000, 28, 28, 4) (324000, 6)
[ ]:
#Check what is in each channel
fig,ax = plt.subplots(4,4, figsize=(10,10))
ax = ax.ravel()
list_idx = np.linspace(0,100,num=16,dtype=np.int64)
for count, idx in enumerate(list_idx):
#     print(idx)
    print('count, t_train[count,:]: {}, {}'.format(count, t_train[count,:]))
#     print(x_train[idx,:,:,0:3])
    ax[count].imshow(x_train[count,:,:,0:3])
    ax[count].set_title(str(np.argmax(t_train[count,:])))
count, t_train[count,:]: 0, [0 0 1 0 0 0]
count, t_train[count,:]: 1, [0 1 0 0 0 0]
count, t_train[count,:]: 2, [0 0 0 0 0 1]
count, t_train[count,:]: 3, [0 0 0 0 0 1]
count, t_train[count,:]: 4, [0 0 0 0 0 1]
count, t_train[count,:]: 5, [1 0 0 0 0 0]
count, t_train[count,:]: 6, [1 0 0 0 0 0]
count, t_train[count,:]: 7, [0 0 0 0 0 1]
count, t_train[count,:]: 8, [0 1 0 0 0 0]
count, t_train[count,:]: 9, [0 0 1 0 0 0]
count, t_train[count,:]: 10, [0 0 0 0 0 1]
count, t_train[count,:]: 11, [0 1 0 0 0 0]
count, t_train[count,:]: 12, [0 1 0 0 0 0]
count, t_train[count,:]: 13, [0 0 0 0 1 0]
count, t_train[count,:]: 14, [0 0 0 0 0 1]
count, t_train[count,:]: 15, [0 0 1 0 0 0]
../_images/CASESTUDY_CNN_satelite_2023_10_1.png
[ ]:
# split in training and testing
from torch.utils.data import Dataset, DataLoader
from torch.utils.data.sampler import SubsetRandomSampler
import torchvision.transforms as transforms
from scipy.ndimage import zoom


class MyDataset(Dataset):
    def __init__(self, data, target):
        print('data.dtype: {}'.format(data.dtype))
        print('target.dtype: {}'.format(target.dtype))
        self.data = torch.from_numpy(data).float()
        self.target = torch.from_numpy(target).float()


    def __getitem__(self, index):
        x = self.data[index]
        y = self.target[index]
        return x, y

    def __len__(self):
        return len(self.data)

print('x_train.shape: {}'.format(x_train.shape))
n_samples = 50000
dataset = MyDataset(x_train[:n_samples,:,:,:], np.argmax(t_train[:n_samples],axis=1))
del x_train, t_train
dataset_size  = len(dataset)
print('dataset_size: {}'.format(dataset_size))
test_split=0.2

batch_size=1024

# -- split dataset
indices       = list(range(dataset_size))
split         = int(np.floor(test_split*dataset_size))
print('split: {}'.format(split))
# np.random.shuffle(indices) # Randomizing the indices is not a good idea if you want to model the sequence
train_indices, val_indices = indices[split:], indices[:split]

# -- create dataloaders
# #Original
train_sampler = SubsetRandomSampler(train_indices)
valid_sampler = SubsetRandomSampler(val_indices)

dataloaders   = {
    'train': torch.utils.data.DataLoader(dataset, batch_size=batch_size, sampler=train_sampler),
    'test': torch.utils.data.DataLoader(dataset, batch_size=batch_size, sampler=valid_sampler),
    'all': torch.utils.data.DataLoader(dataset,  batch_size=5000, shuffle=False),
    }
x_train.shape: (324000, 28, 28, 4)
data.dtype: uint8
target.dtype: int64
dataset_size: 50000
split: 10000
[ ]:
class FFnet(nn.Module):
    '''
    Linear activation in the middle (instead of an activation function)
    '''
    def __init__(self):
        super(FFnet, self).__init__()
        self.enc_lin1 = nn.Linear(3136, 1000) # 28 x 28 x 4
        self.enc_lin2 = nn.Linear(1000, 500)
        self.enc_lin3 = nn.Linear(500, 250)
        self.enc_lin4 = nn.Linear(250, 6)

        self.relu = nn.ReLU()
        self.tanh = nn.Tanh()

    def encode(self, x):
        x = self.enc_lin1(x)
        x = self.relu(x)
        x = self.enc_lin2(x)
        x = self.relu(x)
        x = self.enc_lin3(x)
        x = self.relu(x)
        output = self.enc_lin4(x)
        return output

    def forward(self, x):
        z = self.encode(x)
        return z
[ ]:

[ ]:
## Second routine for training and evaluation (using the )
# Training and Evaluation routines
import time
def train(model,loss_fn, optimizer, train_loader, test_loader, num_epochs=None, verbose=False):
    """
    This is a standard training loop, which leaves some parts to be filled in.
    INPUT:
    :param model: an untrained pytorch model
    :param loss_fn: e.g. Cross Entropy loss or Mean Squared Error.
    :param optimizer: the model optimizer, initialized with a learning rate.
    :param training_set: The training data, in a dataloader for easy iteration.
    :param test_loader: The testing data, in a dataloader for easy iteration.
    """
    print('optimizer: {}'.format(optimizer))
    if num_epochs is None:
        num_epochs = 100
    print('n. of epochs: {}'.format(num_epochs))
    for epoch in range(num_epochs+1):
        start = time.time()
        # loop through each data point in the training set
        for data, targets in train_loader:
            # run the model on the data
            model_input = data.view(data.size(0),-1).to(device)# TODO: Turn the 28 by 28 image tensors into a 784 dimensional tensor.
            if verbose: print('model_input.shape: {}'.format(model_input.shape))

            # Clear gradients w.r.t. parameters
            optimizer.zero_grad()

            out = model(model_input) # The second output is the latent representation
            if verbose:
                print('targets.shape: {}'.format(targets.shape))
                print('targets: {}'.format(targets))
                print('out.shape: {}'.format(out.shape))
                print('out: {}'.format(out))

            # Calculate the loss
            targets = targets.type(torch.LongTensor).to(device) # add an extra dimension to keep CrossEntropy happy.
            if verbose: print('targets.shape: {}'.format(targets.shape))
            loss = loss_fn(out,targets)
            if verbose: print('loss: {}'.format(loss))

            # Find the gradients of our loss via backpropogation
            loss.backward()

            # Adjust accordingly with the optimizer
            optimizer.step()

        # Give status reports every 100 epochs
        loss_train, acc_train = evaluate(model,train_loader,verbose)
        loss_test, acc_test = evaluate(model,test_loader,verbose)
        if epoch % 10==0:
            print(f" EPOCH {epoch}. Progress: {epoch/num_epochs*100}%. ")
            print(" Train loss: {:.4f}. Train Acc: {:.4f}, Test loss: {:.4f}. Test Acc: {:.4f}. Time/epoch: {:.4f}".format(loss_train, acc_train, loss_test, acc_test, (time.time() - start))) #TODO: implement the evaluate function to provide performance statistics during training.

        wandb.log({
            "Loss/train": loss_train,
            "Loss/val": loss_test,
            "Accuracy/train": acc_train,
            "Accuracy/val": acc_test,
            "epoch": epoch
        }, step=epoch)

def evaluate(model, evaluation_set, verbose=False):
    """
    Evaluates the given model on the given dataset.
    Returns the percentage of correct classifications out of total classifications.
    """
    with torch.no_grad(): # this disables backpropogation, which makes the model run much more quickly.
        correct = 0
        total = 0
        loss_all=0

        for data, targets in evaluation_set:
            targets= targets.to(device)
            # run the model on the data
            model_input = data.view(data.size(0),-1).to(device)# TODO: Turn the 28 by 28 image tensors into a 784 dimensional tensor.
            if verbose:
                print('model_input.shape: {}'.format(model_input.shape))
                print('targets.shape: {}'.format(targets.shape))
            out = model(model_input)
            targets = targets.type(torch.LongTensor).to(device)
            loss = loss_fn(out,targets)

            if verbose: print('out[:5]: {}'.format(out[:5]))
            loss_all+=loss.item()

            # the class with the highest energy is what we choose as prediction
            _, predicted = torch.max(out.data, 1)
            total += targets.size(0)
            correct += (predicted == targets).sum().item()
        acc = 100 * correct / total
        loss = loss_all/total
    return loss, acc

[ ]:
lr_range = [0.01,0.005,0.001, 0.0001]

for lr in lr_range:
    if 'model' in globals():
        print('Deleting previous model')
        del model, loss_fn, optimizer
    model = FFnet().to(device)
    DATETIME = datetime.datetime.now().strftime('%Y-%m-%d-%H_%M_%S')
    wandb.init(project="GeoComp_Matera2023", entity="ahof1704", name="CNN_sat_FFnet_{}".format(DATETIME))
    wandb.watch(model, log="all", log_freq=1)
    optimizer = torch.optim.Adam(model.parameters(), lr = lr)

    loss_fn = nn.CrossEntropyLoss().to(device)
    train(model,loss_fn, optimizer, dataloaders['train'], dataloaders['test'],verbose=False)
Deleting previous model
Finishing last run (ID:tk5smslm) before initializing another...
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▄▆▆▇▇████▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▄▆▆▇▇████▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▂▂▂▁▁▁▁▁▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇
Loss/val█▂▂▁▁▁▁▁▁▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.00154
Loss/val0.00153
epoch58

View run CNN_sat_FFnet_2023-05-18-14_10_05 at: https://wandb.ai/ahof1704/GeoComp_Matera2023/runs/tk5smslm
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_141005-tk5smslm/logs
Successfully finished last run (ID:tk5smslm). Initializing new run:
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_141211-bdbijsdw
optimizer: Adam (
Parameter Group 0
    amsgrad: False
    betas: (0.9, 0.999)
    capturable: False
    eps: 1e-08
    foreach: None
    lr: 0.01
    maximize: False
    weight_decay: 0
)
n. of epochs: 100
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0008. Train Acc: 66.3475, Test loss: 0.0008. Test Acc: 66.8600. Time/epoch: 4.3759
 EPOCH 10. Progress: 10.0%.
 Train loss: 0.0004. Train Acc: 81.9250, Test loss: 0.0004. Test Acc: 81.8900. Time/epoch: 3.9911
 EPOCH 20. Progress: 20.0%.
 Train loss: 0.0002. Train Acc: 89.1425, Test loss: 0.0003. Test Acc: 88.6500. Time/epoch: 3.9714
 EPOCH 30. Progress: 30.0%.
 Train loss: 0.0002. Train Acc: 93.1425, Test loss: 0.0002. Test Acc: 92.5700. Time/epoch: 3.7580
 EPOCH 40. Progress: 40.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 3.8162
 EPOCH 50. Progress: 50.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 3.9956
 EPOCH 60. Progress: 60.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 3.7411
 EPOCH 70. Progress: 70.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 4.1185
 EPOCH 80. Progress: 80.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 3.8451
 EPOCH 90. Progress: 90.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 4.0117
 EPOCH 100. Progress: 100.0%.
 Train loss: 0.0015. Train Acc: 36.8250, Test loss: 0.0015. Test Acc: 37.5400. Time/epoch: 3.9103
Deleting previous model
Finishing last run (ID:bdbijsdw) before initializing another...
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▅▇▇█▇▇▅████▇██▄▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▅▇▇█▇▇▅████▇██▄▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.00154
Loss/val0.00153
epoch100

View run CNN_sat_FFnet_2023-05-18-14_12_11 at: https://wandb.ai/ahof1704/GeoComp_Matera2023/runs/bdbijsdw
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_141211-bdbijsdw/logs
Successfully finished last run (ID:bdbijsdw). Initializing new run:
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_141901-0dxs47tg
optimizer: Adam (
Parameter Group 0
    amsgrad: False
    betas: (0.9, 0.999)
    capturable: False
    eps: 1e-08
    foreach: None
    lr: 0.005
    maximize: False
    weight_decay: 0
)
n. of epochs: 100
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0005. Train Acc: 75.6150, Test loss: 0.0005. Test Acc: 75.5200. Time/epoch: 4.0944
 EPOCH 10. Progress: 10.0%.
 Train loss: 0.0003. Train Acc: 84.7850, Test loss: 0.0003. Test Acc: 84.7500. Time/epoch: 3.9586
 EPOCH 20. Progress: 20.0%.
 Train loss: 0.0002. Train Acc: 92.3075, Test loss: 0.0002. Test Acc: 91.8300. Time/epoch: 3.8414
 EPOCH 30. Progress: 30.0%.
 Train loss: 0.0002. Train Acc: 93.0650, Test loss: 0.0002. Test Acc: 92.7300. Time/epoch: 4.0199
 EPOCH 40. Progress: 40.0%.
 Train loss: 0.0002. Train Acc: 89.9625, Test loss: 0.0002. Test Acc: 89.8600. Time/epoch: 4.3224
 EPOCH 50. Progress: 50.0%.
 Train loss: 0.0002. Train Acc: 90.8100, Test loss: 0.0002. Test Acc: 90.1700. Time/epoch: 4.3020
 EPOCH 60. Progress: 60.0%.
 Train loss: 0.0005. Train Acc: 77.6900, Test loss: 0.0005. Test Acc: 77.7400. Time/epoch: 3.9768
 EPOCH 70. Progress: 70.0%.
 Train loss: 0.0003. Train Acc: 87.5625, Test loss: 0.0003. Test Acc: 87.5900. Time/epoch: 3.8873
 EPOCH 80. Progress: 80.0%.
 Train loss: 0.0003. Train Acc: 87.9400, Test loss: 0.0003. Test Acc: 87.7400. Time/epoch: 4.0226
 EPOCH 90. Progress: 90.0%.
 Train loss: 0.0002. Train Acc: 93.2150, Test loss: 0.0002. Test Acc: 92.6300. Time/epoch: 4.4531
 EPOCH 100. Progress: 100.0%.
 Train loss: 0.0002. Train Acc: 92.7275, Test loss: 0.0002. Test Acc: 92.2400. Time/epoch: 3.9699
Deleting previous model
Finishing last run (ID:0dxs47tg) before initializing another...
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▄▇▄▇▇▅▇▇▇▇█▇▇▆▇███▇▃▄▂▆▆▅▆▇▅▇▇▆▇▇▇▇▇██
Accuracy/val▁▁▄▇▅▇▇▅▇▇▇▇█▇▇▆▇███▇▃▄▂▆▆▅▆▇▅▇▇▆▇▇▇▇▇▇▇
Loss/train▇█▄▂▄▂▂▄▂▂▂▂▁▂▃▃▂▁▁▁▁█▄▆▃▃▄▃▂▃▂▂▂▂▂▁▂▂▁▁
Loss/val▇█▄▂▄▂▂▄▂▂▂▂▁▂▃▃▂▁▁▁▁█▄▆▃▃▄▃▂▃▂▂▂▂▂▂▂▂▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇███

Run summary:


Accuracy/train92.7275
Accuracy/val92.24
Loss/train0.00016
Loss/val0.00017
epoch100

View run CNN_sat_FFnet_2023-05-18-14_19_01 at: https://wandb.ai/ahof1704/GeoComp_Matera2023/runs/0dxs47tg
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_141901-0dxs47tg/logs
Successfully finished last run (ID:0dxs47tg). Initializing new run:
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_142559-stjo6gi9
optimizer: Adam (
Parameter Group 0
    amsgrad: False
    betas: (0.9, 0.999)
    capturable: False
    eps: 1e-08
    foreach: None
    lr: 0.001
    maximize: False
    weight_decay: 0
)
n. of epochs: 100
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0005. Train Acc: 78.5275, Test loss: 0.0005. Test Acc: 78.4300. Time/epoch: 4.0875
 EPOCH 10. Progress: 10.0%.
 Train loss: 0.0003. Train Acc: 88.7000, Test loss: 0.0003. Test Acc: 88.5600. Time/epoch: 3.8260
 EPOCH 20. Progress: 20.0%.
 Train loss: 0.0002. Train Acc: 92.8550, Test loss: 0.0002. Test Acc: 92.3100. Time/epoch: 3.9107
 EPOCH 30. Progress: 30.0%.
 Train loss: 0.0003. Train Acc: 85.9700, Test loss: 0.0003. Test Acc: 85.9500. Time/epoch: 3.7582
 EPOCH 40. Progress: 40.0%.
 Train loss: 0.0002. Train Acc: 92.8800, Test loss: 0.0002. Test Acc: 92.4900. Time/epoch: 3.9415
 EPOCH 50. Progress: 50.0%.
 Train loss: 0.0001. Train Acc: 94.0175, Test loss: 0.0002. Test Acc: 93.3500. Time/epoch: 3.8053
 EPOCH 60. Progress: 60.0%.
 Train loss: 0.0002. Train Acc: 92.9325, Test loss: 0.0002. Test Acc: 92.3500. Time/epoch: 4.0517
 EPOCH 70. Progress: 70.0%.
 Train loss: 0.0002. Train Acc: 92.4325, Test loss: 0.0002. Test Acc: 91.9600. Time/epoch: 4.0073
 EPOCH 80. Progress: 80.0%.
 Train loss: 0.0002. Train Acc: 92.7200, Test loss: 0.0002. Test Acc: 92.0100. Time/epoch: 3.7292
 EPOCH 90. Progress: 90.0%.
 Train loss: 0.0001. Train Acc: 96.1825, Test loss: 0.0001. Test Acc: 95.4600. Time/epoch: 4.0354
 EPOCH 100. Progress: 100.0%.
 Train loss: 0.0001. Train Acc: 93.8075, Test loss: 0.0002. Test Acc: 92.5900. Time/epoch: 4.2393
Deleting previous model
Finishing last run (ID:stjo6gi9) before initializing another...
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▅▃▇▇▇▇▆▇▄▆▇▆▇█▅▇▇▇▇▄▇▅▇▇▆▆▇▇▇▇▇▇▇██▇
Accuracy/val▁▂▂▂▅▃▇▇▇▇▆▇▄▆▇▆▇█▅▇▇█▇▄█▆▇▇▆▆▇▇▇▇▇▇███▇
Loss/train▇█▅█▄▅▃▃▂▃▂▂▄▃▂▃▂▁▃▂▂▂▂▄▁▃▂▂▃▃▂▂▁▂▁▂▁▁▁▂
Loss/val▇█▅█▄▅▃▂▂▂▂▂▄▂▂▂▂▁▃▁▂▂▂▄▁▃▂▂▃▃▂▂▁▂▁▂▁▁▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇███

Run summary:


Accuracy/train93.8075
Accuracy/val92.59
Loss/train0.00015
Loss/val0.00018
epoch100

View run CNN_sat_FFnet_2023-05-18-14_25_59 at: https://wandb.ai/ahof1704/GeoComp_Matera2023/runs/stjo6gi9
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_142559-stjo6gi9/logs
Successfully finished last run (ID:stjo6gi9). Initializing new run:
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_143250-rhx17fj2
optimizer: Adam (
Parameter Group 0
    amsgrad: False
    betas: (0.9, 0.999)
    capturable: False
    eps: 1e-08
    foreach: None
    lr: 0.0001
    maximize: False
    weight_decay: 0
)
n. of epochs: 100
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0004. Train Acc: 81.5800, Test loss: 0.0004. Test Acc: 81.4000. Time/epoch: 4.6343
 EPOCH 10. Progress: 10.0%.
 Train loss: 0.0002. Train Acc: 91.4275, Test loss: 0.0002. Test Acc: 91.2600. Time/epoch: 3.9659
 EPOCH 20. Progress: 20.0%.
 Train loss: 0.0003. Train Acc: 86.6925, Test loss: 0.0003. Test Acc: 86.5700. Time/epoch: 4.0900
 EPOCH 30. Progress: 30.0%.
 Train loss: 0.0002. Train Acc: 93.3275, Test loss: 0.0002. Test Acc: 93.0000. Time/epoch: 3.7836
 EPOCH 40. Progress: 40.0%.
 Train loss: 0.0003. Train Acc: 86.7875, Test loss: 0.0003. Test Acc: 86.6300. Time/epoch: 3.9666
 EPOCH 50. Progress: 50.0%.
 Train loss: 0.0003. Train Acc: 86.1525, Test loss: 0.0003. Test Acc: 86.0300. Time/epoch: 4.4136
 EPOCH 60. Progress: 60.0%.
 Train loss: 0.0003. Train Acc: 89.4050, Test loss: 0.0003. Test Acc: 88.9200. Time/epoch: 4.2230
 EPOCH 70. Progress: 70.0%.
 Train loss: 0.0001. Train Acc: 95.5700, Test loss: 0.0001. Test Acc: 94.6200. Time/epoch: 4.1061
 EPOCH 80. Progress: 80.0%.
 Train loss: 0.0001. Train Acc: 96.5025, Test loss: 0.0001. Test Acc: 95.6900. Time/epoch: 4.3656
 EPOCH 90. Progress: 90.0%.
 Train loss: 0.0002. Train Acc: 92.6275, Test loss: 0.0002. Test Acc: 91.9200. Time/epoch: 4.0278
 EPOCH 100. Progress: 100.0%.
 Train loss: 0.0001. Train Acc: 97.1175, Test loss: 0.0001. Test Acc: 96.3000. Time/epoch: 3.7401

Using CNNs for a image dataset

[ ]:
class CNNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(4, 6, 5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1 = nn.Linear(16 * 4 * 4, 120)
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 6)

    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


model = CNNet()
[ ]:
# Training and Evaluation routines
import time
def train(model,loss_fn, optimizer, train_loader, test_loader, num_epochs=None, verbose=False):
    """
    This is a standard training loop, which leaves some parts to be filled in.
    INPUT:
    :param model: an untrained pytorch model
    :param loss_fn: e.g. Cross Entropy loss of Mean Squared Error.
    :param optimizer: the model optimizer, initialized with a learning rate.
    :param training_set: The training data, in a dataloader for easy iteration.
    :param test_loader: The testing data, in a dataloader for easy iteration.
    """

    print('optimizer: {}'.format(optimizer))
    if num_epochs is None:
        num_epochs = 100
    print('n. of epochs: {}'.format(num_epochs))
    for epoch in range(num_epochs+1):
        start = time.time()
        # loop through each data point in the training set
        for data, targets in train_loader:

            # run the model on the data
            model_input = data.permute(0, 3, 2, 1).to(device)
            if verbose: print('model_input.shape: {}'.format(model_input.shape))

            # Clear gradients w.r.t. parameters
            optimizer.zero_grad()

            out = model(model_input) # The second output is the latent representation
            if verbose:
                print('targets.shape: {}'.format(targets.shape))
                print('targets: {}'.format(targets))
                print('out.shape: {}'.format(out.shape))
                print('out: {}'.format(out))

            # Calculate the loss
            targets = targets.type(torch.LongTensor).to(device) # add an extra dimension to keep CrossEntropy happy.
            if verbose: print('targets.shape: {}'.format(targets.shape))
            loss = loss_fn(out,targets)
            if verbose: print('loss: {}'.format(loss))

            # Find the gradients of our loss via backpropogation
            loss.backward()

            # Adjust accordingly with the optimizer
            optimizer.step()

        # Give status reports every 100 epochs
        if epoch % 10==0:
            print(f" EPOCH {epoch}. Progress: {epoch/num_epochs*100}%. ")
            loss_train, acc_train = evaluate(model,train_loader,verbose)
            loss_test, acc_test = evaluate(model,test_loader,verbose)
            print(" Train loss: {:.4f}. Train Acc: {:.4f}, Test loss: {:.4f}. Test Acc: {:.4f}. Time/epoch: {:.4f}".format(loss_train, acc_train, loss_test, acc_test, (time.time() - start))) #TODO: implement the evaluate function to provide performance statistics during training.


def evaluate(model, evaluation_set, verbose=False):
    """
    Evaluates the given model on the given dataset.
    Returns the percentage of correct classifications out of total classifications.
    """
    with torch.no_grad(): # this disables backpropogation, which makes the model run much more quickly.
        correct = 0
        total = 0
        loss_all=0

        for data, targets in evaluation_set:

            # run the model on the data
            model_input = data.permute(0, 3, 2, 1).to(device)
            if verbose:
                print('model_input.shape: {}'.format(model_input.shape))
                print('targets.shape: {}'.format(targets.shape))
            out = model(model_input)
            targets = targets.type(torch.LongTensor)
            loss = loss_fn(out,targets)

            if verbose: print('out[:5]: {}'.format(out[:5]))
            loss_all+=loss.item()

            # the class with the highest energy is what we choose as prediction
            _, predicted = torch.max(out.data, 1)
            total += targets.size(0)
            correct += (predicted == targets).sum().item()
        acc = 100 * correct / total
        loss = loss_all/total
    return loss, acc

[ ]:
lr_range = [0.01,0.005,0.001]

for lr in lr_range:
    if 'model' in globals():
        print('Deleting previous model')
        del model, loss_fn, optimizer
    model = CNNet().to(device)
    optimizer = torch.optim.Adam(model.parameters(), lr = lr)

    loss_fn = nn.CrossEntropyLoss().to(device)
    train(model,loss_fn, optimizer, dataloaders['train'], dataloaders['test'],verbose=False)
Deleting previous model
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[14], line 6
      4 if 'model' in globals():
      5     print('Deleting previous model')
----> 6     del model, loss_fn, optimizer
      7 model = CNNet().to(device)
      8 optimizer = torch.optim.Adam(model.parameters(), lr = lr)

NameError: name 'loss_fn' is not defined

Define Your Neural Network

Before we can run the sweep, let’s define a function that creates and trains our neural network.

In the function below, we define a simple CNN in Pytorch, and add the following lines of code to log models metrics, visualize performance and output and track our experiments easily:

  • wandb.init() – Initialize a new W&B run. Each run is single execution of the training script.

  • wandb.config – Save all your hyperparameters in a config object. This lets you use our app to sort and compare your runs by hyperparameter values.

  • callbacks=[WandbCallback()] – Fetch all layer dimensions, model parameters and log them automatically to your W&B dashboard.

  • wandb.log() – Logs custom objects – these can be images, videos, audio files, HTML, plots, point clouds etc. Here we use wandb.log to log images of Simpson characters overlaid with actual and predicted labels.

Sweeping with WandB

[ ]:
# Configure the sweep – specify the parameters to search through, the search strategy, the optimization metric et all.
sweep_config = {
    'method': 'random', #grid, random
    'metric': {
      'name': 'Accuracy/val',
      'goal': 'maximize'
    },
    'parameters': {
        'epochs': {
            'values': [10, 20, 50]
        },
        'batch_size': {
            'values': [32,64,128]
        },
        'weight_decay': {
            'values': [0.0005, 0.005, 0.05]
        },
        'learning_rate': {
            'values': [1e-2, 1e-3, 1e-4, 3e-4, 3e-5, 1e-5]
        },
        'optimizer': {
            'values': ['adam', 'sgd', 'rmsprop']
        }
    }
}

[ ]:
# Initialize a new sweep
# Arguments:
#     – sweep_config: the sweep config dictionary defined above
#     – entity: Set the username for the sweep
#     – project: Set the project name for the sweep
sweep_id = wandb.sweep(sweep_config, entity="ahof1704", project="CNN_Sat_sweep")
Create sweep with ID: 4bmop0um
Sweep URL: https://wandb.ai/ahof1704/CNN_Sat_sweep/sweeps/4bmop0um
[ ]:
class CNNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(4, 6, 5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1 = nn.Linear(16 * 4 * 4, 120)
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 6)

    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


model = CNNet()
[ ]:
# Training and Evaluation routines for Sweeping
import time

loss_fn = nn.CrossEntropyLoss().to(device)

def train(config=None):
    """
    This is a standard training loop, which leaves some parts to be filled in.
    INPUT:
    :param model: an untrained pytorch model
    :param loss_fn: e.g. Cross Entropy loss of Mean Squared Error.
    :param optimizer: the model optimizer, initialized with a learning rate.
    :param training_set: The training data, in a dataloader for easy iteration.
    :param test_loader: The testing data, in a dataloader for easy iteration.
    """

    with wandb.init(config=config):
        verbose=False
        model = CNNet().to(device)
        model.train()

        # Config is a variable that holds and saves hyperparameters and inputs
        config = wandb.config

        print('optimizer: {}'.format(config.optimizer))
        # Define the optimizer
        if config.optimizer=='sgd':
            optimizer =  torch.optim.SGD(model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay, momentum=0.9, nesterov=True)
        elif config.optimizer=='rmsprop':
            optimizer =  torch.optim.RMSprop(model.parameters(), lr=config.learning_rate, weight_decay=config.weight_decay)
        elif config.optimizer=='adam':
            optimizer =  torch.optim.Adam(model.parameters(), lr=config.learning_rate, betas=(0.9, 0.999))

        # -- create dataloaders
        train_sampler = SubsetRandomSampler(train_indices)
        valid_sampler = SubsetRandomSampler(val_indices)

        dataloaders   = {
            'train': torch.utils.data.DataLoader(dataset, batch_size=config.batch_size, sampler=train_sampler),
            'test': torch.utils.data.DataLoader(dataset, batch_size=config.batch_size, sampler=valid_sampler),
            'all': torch.utils.data.DataLoader(dataset,  batch_size=5000, shuffle=False),
            }
        train_loader = dataloaders['train']
        test_loader = dataloaders['test']

        for epoch in range(config.epochs+1):
            start = time.time()
            # loop through each data point in the training set
            for data, targets in train_loader:

                # run the model on the data
                model_input = data.permute(0, 3, 2, 1).to(device)
                if verbose: print('model_input.shape: {}'.format(model_input.shape))

                # Clear gradients w.r.t. parameters
                optimizer.zero_grad()

                out = model(model_input) # The second output is the latent representation
                if verbose:
                    print('targets.shape: {}'.format(targets.shape))
                    print('targets: {}'.format(targets))
                    print('out.shape: {}'.format(out.shape))
                    print('out: {}'.format(out))

                # Calculate the loss
                targets = targets.type(torch.LongTensor).to(device) # add an extra dimension to keep CrossEntropy happy.
                if verbose: print('targets.shape: {}'.format(targets.shape))
                loss = loss_fn(out,targets)
                if verbose: print('loss: {}'.format(loss))

                # Find the gradients of our loss via backpropogation
                loss.backward()

                # Adjust accordingly with the optimizer
                optimizer.step()

            loss_train, acc_train = evaluate(model,train_loader,verbose)
            loss_test, acc_test = evaluate(model,test_loader,verbose)

            # Give status reports every 100 epochs
            if epoch % 10==0:
                print(f" EPOCH {epoch}. Progress: {epoch/config.epochs*100}%. ")
                print(" Train loss: {:.4f}. Train Acc: {:.4f}, Test loss: {:.4f}. Test Acc: {:.4f}. Time/epoch: {:.4f}".format(loss_train, acc_train, loss_test, acc_test, (time.time() - start))) #TODO: implement the evaluate function to provide performance statistics during training.

            wandb.log({
                "Loss/train": loss_train,
                "Loss/val": loss_test,
                "Accuracy/train": acc_train,
                "Accuracy/val": acc_test,
                "epoch": epoch
            }, step=epoch)

def evaluate(model, evaluation_set, verbose=False):
    """
    Evaluates the given model on the given dataset.
    Returns the percentage of correct classifications out of total classifications.
    """
    model.eval()
    with torch.no_grad(): # this disables backpropogation, which makes the model run much more quickly.
        correct = 0
        total = 0
        loss_all=0

        for data, targets in evaluation_set:

            # run the model on the data
            model_input = data.permute(0, 3, 2, 1).to(device)
            if verbose:
                print('model_input.shape: {}'.format(model_input.shape))
                print('targets.shape: {}'.format(targets.shape))
            out = model(model_input)
            targets = targets.type(torch.LongTensor).to(device)
            loss = loss_fn(out,targets)

            if verbose: print('out[:5]: {}'.format(out[:5]))
            loss_all+=loss.item()

            # the class with the highest energy is what we choose as prediction
            _, predicted = torch.max(out.data, 1)
            total += targets.size(0)
            correct += (predicted == targets).sum().item()
        acc = 100 * correct / total
        loss = loss_all/total
    return loss, acc

[ ]:
wandb.agent(sweep_id, train)
wandb: Agent Starting Run: zzam8458 with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230517_235613-zzam8458
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0088. Train Acc: 83.1675, Test loss: 0.0088. Test Acc: 83.3300. Time/epoch: 3.3380
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0028. Train Acc: 92.3525, Test loss: 0.0029. Test Acc: 92.4100. Time/epoch: 3.2960
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▆▆▇▇▇▇██
Accuracy/val▁▄▅▆▆▇▇▇███
Loss/train█▃▂▂▂▂▁▁▁▁▁
Loss/val█▃▂▂▂▂▁▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train92.3525
Accuracy/val92.41
Loss/train0.00284
Loss/val0.00293
epoch10

View run dutiful-sweep-1 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/zzam8458
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230517_235613-zzam8458/logs
wandb: Agent Starting Run: oft3pjvq with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230517_235703-oft3pjvq
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0099. Train Acc: 87.1850, Test loss: 0.0103. Test Acc: 86.9400. Time/epoch: 4.9065
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0049. Train Acc: 94.0975, Test loss: 0.0054. Test Acc: 93.6300. Time/epoch: 4.8879
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0035. Train Acc: 95.4450, Test loss: 0.0042. Test Acc: 94.4000. Time/epoch: 4.8633
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0038. Train Acc: 95.4750, Test loss: 0.0047. Test Acc: 94.5100. Time/epoch: 4.6823
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0052. Train Acc: 93.3300, Test loss: 0.0058. Test Acc: 92.8400. Time/epoch: 4.6711
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0033. Train Acc: 96.1875, Test loss: 0.0042. Test Acc: 95.2100. Time/epoch: 4.8509
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▆▆▇▆▇▆▇▇▆▇▇▇▆▇▇▆█▇█▇▅▇▇▆█▇▆███▇██▇███
Accuracy/val▁▅▆▆▇▇▇▇▆█▇▇▇█▇▅▇▇▆█▇██▄▇█▆█▇▆█▇█▇██▇███
Loss/train█▅▄▄▃▂▃▂▃▂▂▃▂▂▂▃▂▂▂▁▂▂▂▅▂▁▃▁▂▃▁▁▁▂▁▁▂▁▁▁
Loss/val█▄▃▃▂▂▂▂▃▂▂▃▂▁▂▄▂▂▂▁▂▂▁▅▂▁▃▁▃▃▁▂▁▂▁▁▂▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.1875
Accuracy/val95.21
Loss/train0.00333
Loss/val0.00418
epoch50

View run icy-sweep-2 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/oft3pjvq
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230517_235703-oft3pjvq/logs
wandb: Agent Starting Run: 5l40buem with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_000122-5l40buem
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0067. Train Acc: 91.3225, Test loss: 0.0069. Test Acc: 91.1800. Time/epoch: 4.6403
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0040. Train Acc: 94.8575, Test loss: 0.0042. Test Acc: 94.8800. Time/epoch: 4.6858
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0036. Train Acc: 95.1500, Test loss: 0.0038. Test Acc: 95.2500. Time/epoch: 4.6908
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0034. Train Acc: 95.7000, Test loss: 0.0035. Test Acc: 95.7100. Time/epoch: 4.5836
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0035. Train Acc: 95.5450, Test loss: 0.0036. Test Acc: 95.4700. Time/epoch: 4.5669
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0033. Train Acc: 95.9025, Test loss: 0.0035. Test Acc: 95.9200. Time/epoch: 4.6939
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▅▅▄▇▆▇▆▆▆▇▇▆█▄▇▇▇▇█▇██████▇▇█▇▇█▇▇████
Accuracy/val▁▃▅▅▅▇▆▇▆▇▆▇▇▆█▅▇▇▇▆▇▇▇█████▇▇█▇▇█▇█████
Loss/train█▆▄▄▅▃▃▃▂▂▃▂▂▃▂▄▂▂▂▂▂▂▂▁▁▁▁▁▁▂▁▁▂▁▁▁▁▁▁▁
Loss/val█▆▄▄▅▃▃▃▃▂▂▂▂▃▂▄▂▂▂▂▂▂▂▁▁▁▁▁▁▂▁▁▂▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.9025
Accuracy/val95.92
Loss/train0.00333
Loss/val0.00346
epoch50

View run gentle-sweep-3 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/5l40buem
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_000122-5l40buem/logs
wandb: Agent Starting Run: ycw8i34p with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_000536-ycw8i34p
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.6145
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0479. Test Acc: 37.5400. Time/epoch: 4.7618
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▆▄▃█▆▂▂▂▂▃▁
Loss/val▆▅▃█▃▃▁▂▂▃▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04804
Loss/val0.04794
epoch10

View run eager-sweep-4 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ycw8i34p
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_000536-ycw8i34p/logs
wandb: Agent Starting Run: jc5xf3r3 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_000642-jc5xf3r3
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.5855
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7293
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▄█▁▃▄▁▂▆█▂▃
Loss/val▃█▁▃▃▂▄▆█▁▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.0481
Loss/val0.04798
epoch10

View run vague-sweep-5 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/jc5xf3r3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_000642-jc5xf3r3/logs
wandb: Agent Starting Run: i7nmk94y with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_000748-i7nmk94y
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0065. Train Acc: 73.8450, Test loss: 0.0065. Test Acc: 73.8100. Time/epoch: 2.1962
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 90.5125, Test loss: 0.0020. Test Acc: 90.4400. Time/epoch: 2.3259
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▆▆▇▇▇███
Accuracy/val▁▄▅▆▆▇▇▇███
Loss/train█▅▃▃▂▂▂▁▁▁▁
Loss/val█▅▃▃▂▂▂▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train90.5125
Accuracy/val90.44
Loss/train0.00192
Loss/val0.00196
epoch10

View run glowing-sweep-6 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/i7nmk94y
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_000748-i7nmk94y/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: 8e2wk0j2 with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_000835-8e2wk0j2
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0068. Train Acc: 85.0775, Test loss: 0.0068. Test Acc: 85.7200. Time/epoch: 3.0562
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0025. Train Acc: 93.6550, Test loss: 0.0027. Test Acc: 93.5200. Time/epoch: 3.1775
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0022. Train Acc: 94.7350, Test loss: 0.0023. Test Acc: 94.4900. Time/epoch: 3.1338
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▆▆▇▇▇▇▇▇▇█████████
Accuracy/val▁▄▅▆▆▇▇▇▇▇▇██████████
Loss/train█▄▃▃▂▂▂▂▂▁▂▁▁▁▁▁▁▁▁▁▁
Loss/val█▄▃▃▂▂▂▂▂▁▂▁▁▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train94.735
Accuracy/val94.49
Loss/train0.00217
Loss/val0.0023
epoch20

View run deep-sweep-7 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/8e2wk0j2
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_000835-8e2wk0j2/logs
wandb: Agent Starting Run: 9olidkws with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_000955-9olidkws
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0068. Train Acc: 91.0925, Test loss: 0.0069. Test Acc: 90.8100. Time/epoch: 4.7435
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0047. Train Acc: 93.6450, Test loss: 0.0048. Test Acc: 93.8200. Time/epoch: 4.8731
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0039. Train Acc: 94.9900, Test loss: 0.0040. Test Acc: 95.1200. Time/epoch: 4.8602
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0052. Train Acc: 92.5125, Test loss: 0.0053. Test Acc: 92.6600. Time/epoch: 4.6839
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0036. Train Acc: 95.6250, Test loss: 0.0037. Test Acc: 95.6400. Time/epoch: 4.8334
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0036. Train Acc: 95.3300, Test loss: 0.0037. Test Acc: 95.5700. Time/epoch: 4.8268
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▃▅▄▄▆▅▅▆▅▆▇▇▆▇▆▇▇▇▆▆▇█▃▇▇▇▇▆█▇█▆████▇▇
Accuracy/val▁▃▄▅▅▅▆▅▅▆▄▇▇▇▆▇▇▇▇▇▆▆▇█▃▇▇▇▇▆█▇█▇████▇▇
Loss/train█▆▅▄▄▄▃▄▄▃▄▃▂▂▃▂▂▂▂▂▃▂▁▁▅▂▂▂▂▃▁▂▁▃▁▁▁▁▂▂
Loss/val█▆▅▄▄▄▃▄▄▃▄▂▂▂▃▂▂▂▂▂▃▂▁▁▅▂▂▂▂▃▁▂▁▂▁▁▁▁▂▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.33
Accuracy/val95.57
Loss/train0.00357
Loss/val0.00366
epoch50

View run morning-sweep-8 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9olidkws
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_000955-9olidkws/logs
wandb: Agent Starting Run: xwgaxq12 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_001419-xwgaxq12
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0018. Train Acc: 90.3100, Test loss: 0.0019. Test Acc: 90.2400. Time/epoch: 2.3505
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0012. Train Acc: 93.2350, Test loss: 0.0013. Test Acc: 93.2100. Time/epoch: 2.3292
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0009. Train Acc: 95.3875, Test loss: 0.0011. Test Acc: 94.9800. Time/epoch: 2.1884
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0006. Train Acc: 96.9850, Test loss: 0.0007. Test Acc: 96.5600. Time/epoch: 2.3183
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0006. Train Acc: 97.2925, Test loss: 0.0007. Test Acc: 96.8100. Time/epoch: 2.3052
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 97.6450, Test loss: 0.0006. Test Acc: 97.1600. Time/epoch: 2.1686
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▅▆▆▆▄▆▆▇▅▆▆▇▆▇▇▇▅▇▆▇▇▇▇▇███▇███▇▇███
Accuracy/val▁▃▄▅▅▆▆▆▄▆▆▇▅▆▆▇▆▇▇▇▅▇▆▇▇▇▇▇▇██▇█▇█▇▇█▇█
Loss/train█▆▅▄▄▃▃▃▅▃▃▂▄▃▃▂▃▂▂▂▃▂▃▂▂▂▂▂▁▁▁▂▁▁▁▁▂▁▁▁
Loss/val█▆▅▄▄▃▃▃▅▃▃▂▄▃▃▂▃▂▂▂▃▂▂▂▁▂▂▂▁▁▁▂▁▁▁▂▃▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.645
Accuracy/val97.16
Loss/train0.00049
Loss/val0.00063
epoch50

View run royal-sweep-9 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xwgaxq12
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_001419-xwgaxq12/logs
wandb: Agent Starting Run: 3qr8irec with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_001627-3qr8irec
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0119. Train Acc: 88.0000, Test loss: 0.0122. Test Acc: 87.8200. Time/epoch: 4.7784
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0055. Train Acc: 92.7175, Test loss: 0.0056. Test Acc: 92.7200. Time/epoch: 4.6236
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0050. Train Acc: 93.7725, Test loss: 0.0052. Test Acc: 93.5600. Time/epoch: 4.5725
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0047. Train Acc: 93.7725, Test loss: 0.0049. Test Acc: 93.9700. Time/epoch: 4.7329
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0040. Train Acc: 95.1700, Test loss: 0.0042. Test Acc: 95.1100. Time/epoch: 4.7364
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0041. Train Acc: 95.0375, Test loss: 0.0043. Test Acc: 94.8900. Time/epoch: 4.7085
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▄▅▅▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇██▇████▇█▇█▇████▆█
Accuracy/val▁▄▄▅▅▆▆▆▆▇▆▇▇▇▇▇▆▇▇▇▇▇█▇▇████▇█▇█▇████▆█
Loss/train█▄▄▃▃▂▂▂▂▂▂▂▂▂▂▁▂▂▂▂▁▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▂▁
Loss/val█▄▄▃▃▃▂▂▂▂▂▂▂▂▂▁▂▂▂▂▁▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▂▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.0375
Accuracy/val94.89
Loss/train0.0041
Loss/val0.00427
epoch50

View run lucky-sweep-10 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/3qr8irec
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_001627-3qr8irec/logs
wandb: Agent Starting Run: t4osxxhb with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002039-t4osxxhb
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0016. Train Acc: 91.7150, Test loss: 0.0016. Test Acc: 91.6600. Time/epoch: 2.3046
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0007. Train Acc: 96.3875, Test loss: 0.0008. Test Acc: 95.9900. Time/epoch: 2.4241
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0004. Train Acc: 97.8550, Test loss: 0.0006. Test Acc: 97.0100. Time/epoch: 2.4088
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▅▆▆▆▇▆▆▆▇▆▇▇███▇█
Accuracy/val▁▄▄▆▅▇▆▇▇▇▇▆▇▆█▇███▇█
Loss/train█▅▅▄▄▃▃▃▂▃▃▃▂▃▂▂▁▁▁▂▁
Loss/val█▅▄▃▃▂▃▂▂▂▂▃▁▃▁▂▁▁▁▃▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train97.855
Accuracy/val97.01
Loss/train0.00044
Loss/val0.00063
epoch20

View run scarlet-sweep-11 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/t4osxxhb
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002039-t4osxxhb/logs
wandb: Agent Starting Run: 3d41qa4k with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002141-3d41qa4k
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 5.1893
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 5.1383
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0479. Test Acc: 37.5400. Time/epoch: 5.1727
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▆▅▄▄▄▆▅▅▂▆▂▂▁▇▄█▅▄▅▇▂
Loss/val▅█▂▁▃▇▆▇▄▆▅▅▂█▅▆▅▆▅▄▂
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04805
Loss/val0.04793
epoch20

View run proud-sweep-12 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/3d41qa4k
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002141-3d41qa4k/logs
wandb: Agent Starting Run: xvc9j71t with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002342-xvc9j71t
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0060. Train Acc: 91.9650, Test loss: 0.0062. Test Acc: 91.7200. Time/epoch: 4.8781
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0121. Train Acc: 84.9550, Test loss: 0.0121. Test Acc: 85.3500. Time/epoch: 4.8153
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0035. Train Acc: 95.7625, Test loss: 0.0036. Test Acc: 95.7500. Time/epoch: 4.8847
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▆▆▆▆▇▇▇▇▇█▁▇███▇█▇███
Accuracy/val▅▅▆▆▇▇▇▇▇█▁▇▇████▇███
Loss/train▃▃▂▂▂▂▂▁▂▂█▂▁▁▁▁▁▂▁▁▁
Loss/val▃▃▂▂▂▂▂▁▂▂█▂▁▁▁▁▁▂▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train95.7625
Accuracy/val95.75
Loss/train0.00353
Loss/val0.0036
epoch20

View run atomic-sweep-13 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xvc9j71t
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002342-xvc9j71t/logs
wandb: Agent Starting Run: w3ttjtb4 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002539-w3ttjtb4
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0018. Train Acc: 90.5200, Test loss: 0.0019. Test Acc: 90.5500. Time/epoch: 2.3806
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0010. Train Acc: 95.0050, Test loss: 0.0011. Test Acc: 94.8300. Time/epoch: 2.3062
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0009. Train Acc: 95.5125, Test loss: 0.0009. Test Acc: 95.4000. Time/epoch: 2.1862
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0008. Train Acc: 96.2875, Test loss: 0.0008. Test Acc: 96.1700. Time/epoch: 2.3307
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0009. Train Acc: 95.3725, Test loss: 0.0010. Test Acc: 95.1500. Time/epoch: 2.3377
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 95.3175, Test loss: 0.0009. Test Acc: 95.4600. Time/epoch: 2.3253
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▁▅▆▃▆▅▆▇▆▅▆▇▇▆▇▇▇▅▇▇█▇█▆▆▆▇█▇█▇█▇██▇▅▇
Accuracy/val▁▅▂▅▆▃▆▆▆▆▆▅▆▇▇▆▇▇▇▅▇▇█▇█▆▇▆▇███▇█▇██▇▅▇
Loss/train█▅▇▄▃▅▃▄▃▂▃▄▃▂▂▂▂▂▂▃▂▂▁▂▁▃▂▃▂▁▁▁▂▁▂▁▁▂▃▂
Loss/val█▅▇▄▃▅▃▃▃▃▂▄▃▂▂▂▂▂▂▃▂▂▁▂▁▃▂▃▁▁▁▁▂▁▂▁▁▂▃▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.3175
Accuracy/val95.46
Loss/train0.0009
Loss/val0.00093
epoch50

View run lilac-sweep-14 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/w3ttjtb4
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002539-w3ttjtb4/logs
wandb: Agent Starting Run: 6kgluzmq with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002749-6kgluzmq
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0046. Train Acc: 93.9450, Test loss: 0.0048. Test Acc: 93.8200. Time/epoch: 4.6201
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0035. Train Acc: 95.2925, Test loss: 0.0039. Test Acc: 95.0600. Time/epoch: 4.7475
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▄▁▄▆▇▄▇▇██▆
Accuracy/val▄▁▄▆▇▃▇▇██▆
Loss/train▅█▅▃▂▅▂▃▁▁▃
Loss/val▅█▆▃▂▅▂▃▁▁▃
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train95.2925
Accuracy/val95.06
Loss/train0.0035
Loss/val0.00385
epoch10

View run pleasant-sweep-15 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/6kgluzmq
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002749-6kgluzmq/logs
wandb: Agent Starting Run: vn8tnx3k with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002855-vn8tnx3k
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0042. Train Acc: 76.6925, Test loss: 0.0043. Test Acc: 76.9400. Time/epoch: 2.2380
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0169. Train Acc: 58.1875, Test loss: 0.0170. Test Acc: 58.5100. Time/epoch: 2.3528
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▆▇▇█▅▆██▅█▁
Accuracy/val▆▇██▅▆██▅█▁
Loss/train▂▂▁▂▂▂▁▁▂▁█
Loss/val▂▂▁▂▂▂▁▁▂▁█
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train58.1875
Accuracy/val58.51
Loss/train0.0169
Loss/val0.01697
epoch10

View run kind-sweep-16 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/vn8tnx3k
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002855-vn8tnx3k/logs
wandb: Agent Starting Run: 7m5xr9i9 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_002936-7m5xr9i9
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.0138
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 2.9350
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0977
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.1260
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.1197
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.0731
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▃▂▂▁▅▂▁▃▅▁█▅▁▄▃▃▇▂▃▅▂▂▂▁▁▃▁▂▃▁▂▁▁▄▃▂▁▁▃▂
Loss/val▃▃▂▂▆▄▂▆▅▂█▅▂▅▄▄▄▂▃▅▅▆▃▄▂▆▁▃▃▄▃▂▄▄▂▃▃▄▃▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02402
Loss/val0.02403
epoch50

View run astral-sweep-17 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/7m5xr9i9
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_002936-7m5xr9i9/logs
wandb: Agent Starting Run: do2ct7u1 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_003223-do2ct7u1
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0021. Train Acc: 90.2725, Test loss: 0.0022. Test Acc: 90.5000. Time/epoch: 2.4773
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0009. Train Acc: 95.6300, Test loss: 0.0009. Test Acc: 95.6800. Time/epoch: 2.4208
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0007. Train Acc: 96.4575, Test loss: 0.0008. Test Acc: 96.4900. Time/epoch: 2.4178
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0006. Train Acc: 96.9950, Test loss: 0.0007. Test Acc: 96.5600. Time/epoch: 2.2943
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0006. Train Acc: 97.2800, Test loss: 0.0006. Test Acc: 96.9200. Time/epoch: 2.4210
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 97.6825, Test loss: 0.0006. Test Acc: 97.2500. Time/epoch: 2.4330
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▄▅▅▅▆▆▆▆▆▇▇▆▇▇▇▇▇▇▆▇▇▇▇▇▇▇▇█▇████████
Accuracy/val▁▃▄▄▅▅▅▆▆▆▆▇▇▇▆▇▇▇▇▇▇▆▇▇▇▇██▇▇██████████
Loss/train█▅▅▄▄▃▃▃▃▃▃▂▂▂▃▂▂▂▂▂▂▃▂▂▁▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▅▄▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.6825
Accuracy/val97.25
Loss/train0.00051
Loss/val0.00062
epoch50

View run robust-sweep-18 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/do2ct7u1
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_003223-do2ct7u1/logs
wandb: Agent Starting Run: uchsrn45 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_003440-uchsrn45
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0176. Train Acc: 55.3525, Test loss: 0.0178. Test Acc: 55.5500. Time/epoch: 3.2013
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1492
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1419
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0273
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0241. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1642
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1493
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▁███████████████████████████████████████
Loss/val▁███████████████████████████████████████
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02405
Loss/val0.02409
epoch50

View run clear-sweep-19 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/uchsrn45
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_003440-uchsrn45/logs
wandb: Agent Starting Run: oc0l1pdk with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_003732-oc0l1pdk
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0018. Train Acc: 88.6375, Test loss: 0.0018. Test Acc: 88.6900. Time/epoch: 2.2337
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0008. Train Acc: 96.1925, Test loss: 0.0008. Test Acc: 96.0400. Time/epoch: 2.3739
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0005. Train Acc: 97.7425, Test loss: 0.0006. Test Acc: 97.5600. Time/epoch: 2.3606
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0004. Train Acc: 98.0750, Test loss: 0.0006. Test Acc: 97.5400. Time/epoch: 2.3336
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0003. Train Acc: 98.4550, Test loss: 0.0005. Test Acc: 97.6800. Time/epoch: 2.2281
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0003. Train Acc: 98.6000, Test loss: 0.0006. Test Acc: 97.7000. Time/epoch: 2.3394
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▂▅▆▆▆▆▆▆▇▆▅▆▆▇▇▇▇▇▇▇▅▇███▇█▇████████▇█
Accuracy/val▁▂▂▅▆▆▆▆▇▆▇▆▅▆▇▇███▇▇▇▅▇███▇█▇███▇▇███▇█
Loss/train███▅▄▄▃▃▃▃▃▄▄▃▃▂▂▂▂▂▂▂▅▂▂▁▁▂▂▂▁▁▁▁▁▁▁▁▂▁
Loss/val███▅▃▃▃▃▃▃▂▃▄▃▂▂▂▁▁▂▂▂▅▂▁▁▁▂▂▂▁▁▁▁▁▁▁▁▂▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.6
Accuracy/val97.7
Loss/train0.0003
Loss/val0.00055
epoch50

View run unique-sweep-20 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/oc0l1pdk
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_003732-oc0l1pdk/logs
wandb: Agent Starting Run: xdwakrae with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_003945-xdwakrae
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0104. Train Acc: 70.3125, Test loss: 0.0103. Test Acc: 70.9600. Time/epoch: 3.0324
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0019. Train Acc: 95.6000, Test loss: 0.0021. Test Acc: 95.4500. Time/epoch: 3.1860
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0016. Train Acc: 95.8350, Test loss: 0.0021. Test Acc: 95.1400. Time/epoch: 3.1535
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▅▇▇▇▆▆▇██▇███████▇█
Accuracy/val▁▃▅▇▇▇▆▆▇██▇███████▇█
Loss/train█▆▄▂▂▂█▃▂▂▁▂▁▁▁▁▁▁▁▂▁
Loss/val▇▅▄▂▂▂█▃▂▁▁▁▁▁▁▁▁▁▁▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train95.835
Accuracy/val95.14
Loss/train0.00164
Loss/val0.00205
epoch20

View run deep-sweep-21 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xdwakrae
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_003945-xdwakrae/logs
wandb: Agent Starting Run: cr02yzq9 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004106-cr02yzq9
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0045. Train Acc: 94.1700, Test loss: 0.0047. Test Acc: 94.2100. Time/epoch: 5.1689
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0025. Train Acc: 97.0225, Test loss: 0.0027. Test Acc: 96.5200. Time/epoch: 5.1706
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▃▅▇▆▅▅▇▇█
Accuracy/val▁▄▃▅▇▇▅▅▇▇█
Loss/train█▅▆▃▂▃▃▄▂▂▁
Loss/val█▆▆▄▂▃▃▅▂▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train97.0225
Accuracy/val96.52
Loss/train0.00246
Loss/val0.00275
epoch10

View run wandering-sweep-22 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/cr02yzq9
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004106-cr02yzq9/logs
wandb: Agent Starting Run: cmkt7j2h with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004217-cmkt7j2h
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0034. Train Acc: 90.8500, Test loss: 0.0035. Test Acc: 91.1100. Time/epoch: 3.2042
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0014. Train Acc: 96.5125, Test loss: 0.0016. Test Acc: 96.3200. Time/epoch: 3.1972
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0009. Train Acc: 97.7800, Test loss: 0.0011. Test Acc: 97.3000. Time/epoch: 3.1353
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0009. Train Acc: 97.5925, Test loss: 0.0012. Test Acc: 97.0900. Time/epoch: 3.0095
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0005. Train Acc: 98.8900, Test loss: 0.0009. Test Acc: 98.0400. Time/epoch: 2.9984
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 98.8875, Test loss: 0.0010. Test Acc: 97.9200. Time/epoch: 3.1653
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▄▅▄▃▅▆▆▆▆▆▆▇▄▇▇▇▇▇▇▇▇▇▇▆▇▇██▇████████
Accuracy/val▁▄▄▄▅▄▃▆▆▆▇▇▆▆▇▄▇▇▇▇▇▇██▇▇▆▇▆██▇█████▇▇█
Loss/train█▆▅▅▄▅▆▄▃▃▃▃▃▄▂▅▂▂▂▂▂▂▂▂▂▂▃▂▂▁▁▂▁▁▁▁▁▁▁▁
Loss/val█▅▅▅▄▄▆▃▃▃▂▂▃▃▂▅▂▂▂▂▂▂▁▁▂▂▃▂▂▁▁▂▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.8875
Accuracy/val97.92
Loss/train0.00047
Loss/val0.001
epoch50

View run devout-sweep-23 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/cmkt7j2h
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004217-cmkt7j2h/logs
wandb: Agent Starting Run: mq1zs0j2 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004510-mq1zs0j2
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0015. Train Acc: 91.7725, Test loss: 0.0015. Test Acc: 92.0400. Time/epoch: 2.2943
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0008. Train Acc: 96.0750, Test loss: 0.0009. Test Acc: 95.9700. Time/epoch: 2.3073
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▃▁▅▄▆▆▇▆██▇
Accuracy/val▄▁▆▄▆▆▇▇██▇
Loss/train▆█▄▆▄▃▂▃▁▁▂
Loss/val▆█▃▆▄▃▂▃▁▁▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train96.075
Accuracy/val95.97
Loss/train0.00077
Loss/val0.00085
epoch10

View run olive-sweep-24 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/mq1zs0j2
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004510-mq1zs0j2/logs
wandb: Agent Starting Run: t25nlhl0 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004550-t25nlhl0
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0017. Train Acc: 90.7050, Test loss: 0.0018. Test Acc: 90.5800. Time/epoch: 2.3931
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0029. Train Acc: 85.2775, Test loss: 0.0030. Test Acc: 85.6400. Time/epoch: 2.2056
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0011. Train Acc: 93.8225, Test loss: 0.0011. Test Acc: 94.1100. Time/epoch: 2.3408
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0015. Train Acc: 91.7100, Test loss: 0.0016. Test Acc: 91.5500. Time/epoch: 2.3471
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0010. Train Acc: 94.9125, Test loss: 0.0010. Test Acc: 95.0100. Time/epoch: 2.3499
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0011. Train Acc: 94.4100, Test loss: 0.0011. Test Acc: 94.7400. Time/epoch: 2.2061
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▆▆▅▅▇▃▇▇▃█▅▇▅▆▆▁▇█▇▆▇█▅▆▆▆█▇█▇██████████
Accuracy/val▅▆▅▅▇▃▇▇▃█▅▇▅▆▆▁▇█▇▆▇█▅▆▆▆█▇█▇██████████
Loss/train▃▃▃▃▂▆▂▂▆▁▄▂▄▂▃█▁▁▂▂▂▁▃▃▃▂▁▂▁▂▁▁▁▁▁▁▁▁▁▁
Loss/val▃▃▃▃▃▆▂▂▆▁▄▂▄▃▃█▁▁▂▂▂▁▃▃▃▂▁▂▁▂▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train94.41
Accuracy/val94.74
Loss/train0.00108
Loss/val0.0011
epoch50

View run restful-sweep-25 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/t25nlhl0
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004550-t25nlhl0/logs
wandb: Agent Starting Run: xiqj6i11 with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004759-xiqj6i11
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0075. Train Acc: 77.7025, Test loss: 0.0074. Test Acc: 78.1000. Time/epoch: 3.1311
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0031. Train Acc: 92.9450, Test loss: 0.0034. Test Acc: 92.3900. Time/epoch: 3.0820
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▅▅▄▇▆██
Accuracy/val▁▂▄▅▅▅▄▆▆██
Loss/train█▅▅▆▄▄▄▃▄▂▁
Loss/val█▅▅▆▄▄▄▃▄▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train92.945
Accuracy/val92.39
Loss/train0.00312
Loss/val0.00339
epoch10

View run visionary-sweep-26 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xiqj6i11
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004759-xiqj6i11/logs
wandb: Agent Starting Run: famk9f31 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004848-famk9f31
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0062. Train Acc: 90.5375, Test loss: 0.0063. Test Acc: 90.7300. Time/epoch: 4.9297
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0032. Train Acc: 96.3000, Test loss: 0.0034. Test Acc: 96.2200. Time/epoch: 4.8901
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▃▂▁▇▆█▅██▇█
Accuracy/val▃▂▁▇▆█▅██▇█
Loss/train▆█▇▂▃▁▄▁▁▂▂
Loss/val▆█▇▂▃▁▄▁▁▂▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train96.3
Accuracy/val96.22
Loss/train0.0032
Loss/val0.00339
epoch10

View run dark-sweep-27 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/famk9f31
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004848-famk9f31/logs
wandb: Agent Starting Run: ytk8g35l with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_004953-ytk8g35l
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0082. Train Acc: 89.1525, Test loss: 0.0083. Test Acc: 89.5100. Time/epoch: 5.2215
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0031. Train Acc: 96.1425, Test loss: 0.0042. Test Acc: 94.8700. Time/epoch: 5.1365
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▆▅▆▅▇████
Accuracy/val▁▃▆▅▆▅▇█▇█▇
Loss/train█▆▄▄▃▄▂▁▁▁▁
Loss/val█▆▃▃▃▄▂▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train96.1425
Accuracy/val94.87
Loss/train0.00307
Loss/val0.00423
epoch10

View run deft-sweep-28 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ytk8g35l
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_004953-ytk8g35l/logs
wandb: Agent Starting Run: aec4ug5t with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 3e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_005103-aec4ug5t
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0020. Train Acc: 90.2800, Test loss: 0.0020. Test Acc: 90.5100. Time/epoch: 2.3997
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0012. Train Acc: 94.1325, Test loss: 0.0012. Test Acc: 94.1300. Time/epoch: 2.3393
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0012. Train Acc: 94.0325, Test loss: 0.0012. Test Acc: 94.1000. Time/epoch: 2.3384
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▃▁▆▆▇▆▆▆▆▇▇███▇▇▆█▆
Accuracy/val▁▂▃▁▆▆▇▆▆▇▆▇▇███▇▇▇█▆
Loss/train█▆▅▇▃▂▂▃▃▂▃▂▂▂▁▁▂▂▂▁▂
Loss/val█▆▅▇▃▂▂▃▃▂▃▂▂▂▁▁▂▂▂▁▂
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train94.0325
Accuracy/val94.1
Loss/train0.00116
Loss/val0.00121
epoch20

View run brisk-sweep-29 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/aec4ug5t
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_005103-aec4ug5t/logs
wandb: Agent Starting Run: pjh7m7v6 with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_005204-pjh7m7v6
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0039. Train Acc: 89.7025, Test loss: 0.0041. Test Acc: 89.3400. Time/epoch: 3.3796
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0016. Train Acc: 96.1050, Test loss: 0.0018. Test Acc: 96.0800. Time/epoch: 3.3007
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0013. Train Acc: 96.8500, Test loss: 0.0015. Test Acc: 96.6600. Time/epoch: 3.3318
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▆▆▆▇▇▇▇▇▇▇▇█▇▇▇▇▇█
Accuracy/val▁▄▄▆▆▆▇▇▇▇▇▇▇██▇▇▇▇██
Loss/train█▅▅▃▃▃▃▂▂▂▂▂▂▂▂▃▂▂▂▂▁
Loss/val█▅▄▃▃▃▂▂▂▂▂▂▂▂▁▂▂▂▂▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train96.85
Accuracy/val96.66
Loss/train0.00132
Loss/val0.00147
epoch20

View run bumbling-sweep-30 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/pjh7m7v6
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_005204-pjh7m7v6/logs
wandb: Agent Starting Run: k19vq8g3 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_005326-k19vq8g3
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0031. Train Acc: 87.0525, Test loss: 0.0031. Test Acc: 86.9800. Time/epoch: 2.3651
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0012. Train Acc: 93.9275, Test loss: 0.0013. Test Acc: 94.1100. Time/epoch: 2.3194
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0011. Train Acc: 94.6950, Test loss: 0.0011. Test Acc: 94.6200. Time/epoch: 2.3236
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0011. Train Acc: 93.9350, Test loss: 0.0012. Test Acc: 94.1100. Time/epoch: 2.1896
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0009. Train Acc: 95.4750, Test loss: 0.0010. Test Acc: 95.4700. Time/epoch: 2.3240
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 95.7675, Test loss: 0.0010. Test Acc: 95.7500. Time/epoch: 2.3105
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇███▇█▇███████████████
Accuracy/val▁▃▄▅▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇█▇███████████████
Loss/train█▅▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.7675
Accuracy/val95.75
Loss/train0.00091
Loss/val0.00097
epoch50

View run winter-sweep-31 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/k19vq8g3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_005326-k19vq8g3/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: 41ow5vvs with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.001
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_005543-41ow5vvs
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0013. Train Acc: 92.6025, Test loss: 0.0014. Test Acc: 92.5400. Time/epoch: 2.4629
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0008. Train Acc: 96.0200, Test loss: 0.0010. Test Acc: 95.3400. Time/epoch: 2.4240
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0005. Train Acc: 97.5950, Test loss: 0.0009. Test Acc: 96.1700. Time/epoch: 2.3089
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0004. Train Acc: 97.9350, Test loss: 0.0010. Test Acc: 96.0000. Time/epoch: 2.2992
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0004. Train Acc: 98.0250, Test loss: 0.0011. Test Acc: 95.8700. Time/epoch: 2.4359
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0007. Train Acc: 96.9100, Test loss: 0.0016. Test Acc: 94.9100. Time/epoch: 2.4494
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▂▃▄▅▆▄▅▆▆▆▆▃▆▆▇▆▅▇▇▇▇▇▇▇▇▇▆▇█▇▇██▆▇█▇▆
Accuracy/val▁▄▂▄▅▇▇▅▆▇▇█▇▃▇▇▇▆▅▇█▇▇▇▇▇▇█▆▇█▆▆██▅▇▇▆▅
Loss/train█▇▇▇▆▄▄▅▄▃▃▃▃▆▃▃▂▃▄▂▂▂▂▂▂▂▂▂▄▂▁▃▂▁▁▄▂▁▂▄
Loss/val▆▄▅▄▄▂▁▃▂▁▁▁▁▅▁▂▁▃▄▂▂▂▂▂▂▃▄▂▅▄▂▅▅▄▃▆▄▅▆█
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.91
Accuracy/val94.91
Loss/train0.0007
Loss/val0.00165
epoch50

View run mild-sweep-32 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/41ow5vvs
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_005543-41ow5vvs/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: rma15prl with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_005805-rma15prl
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0021. Train Acc: 88.6875, Test loss: 0.0021. Test Acc: 88.5300. Time/epoch: 2.2395
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0010. Train Acc: 94.8100, Test loss: 0.0011. Test Acc: 94.8900. Time/epoch: 2.3652
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0006. Train Acc: 97.1975, Test loss: 0.0007. Test Acc: 96.8800. Time/epoch: 2.3607
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0005. Train Acc: 97.9025, Test loss: 0.0006. Test Acc: 97.4800. Time/epoch: 2.3467
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0004. Train Acc: 97.8625, Test loss: 0.0006. Test Acc: 97.2900. Time/epoch: 2.2306
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0015. Train Acc: 92.9100, Test loss: 0.0017. Test Acc: 92.7400. Time/epoch: 2.2342
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▅▄▄▆▄▆▅▇▄▇▆▇▇▇▇▇▆▇▇▇█▇█▇▇█▇███▇▇██▆▇▇▄
Accuracy/val▁▃▅▄▅▇▄▆▆▇▄▇▆▇▇▇▇▇▆▇▇▇█▇█▇██▇▇██▇▇██▆▇▇▄
Loss/train█▆▄▅▅▃▅▃▄▃▅▂▃▂▂▂▂▂▃▂▂▂▁▂▁▂▂▁▂▂▁▁▂▂▁▁▄▂▁▅
Loss/val█▆▄▄▅▃▅▃▃▂▅▂▃▂▂▂▂▂▃▂▂▂▁▂▁▂▂▁▂▂▁▁▂▂▁▁▄▂▂▆
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train92.91
Accuracy/val92.74
Loss/train0.00147
Loss/val0.00167
epoch50

View run eager-sweep-33 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/rma15prl
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_005805-rma15prl/logs
wandb: Agent Starting Run: 9g0xstpx with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.001
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_010017-9g0xstpx
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0024. Train Acc: 93.3225, Test loss: 0.0025. Test Acc: 93.3200. Time/epoch: 3.3641
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0016. Train Acc: 95.8700, Test loss: 0.0019. Test Acc: 95.5100. Time/epoch: 3.1685
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0013. Train Acc: 96.5725, Test loss: 0.0018. Test Acc: 96.1900. Time/epoch: 3.1581
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0014. Train Acc: 96.4300, Test loss: 0.0021. Test Acc: 95.7500. Time/epoch: 3.3234
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0011. Train Acc: 96.9625, Test loss: 0.0018. Test Acc: 96.0000. Time/epoch: 3.3414
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 98.0200, Test loss: 0.0022. Test Acc: 96.9100. Time/epoch: 3.3105
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▂▂▄▂▃▄▄▅▅▅▆▄▅▅▆▅▆▆▆▇▆▇▇▇▆▇█▇▇▇▇▇▇▁▇▅▆█▆█
Accuracy/val▃▃▅▃▄▄▅▅▆▆▆▅▅▅▆▆▇▇▇▇▆█▇▇▆▇██▇█▇▇▇▁█▆▆█▆█
Loss/train██▆█▇▆▆▅▄▅▄▆▅▄▄▅▃▃▃▂▃▂▃▂▃▂▁▂▂▂▂▂▂█▂▅▃▁▆▁
Loss/val▆▆▄▆▅▅▅▃▃▃▂▅▄▃▃▃▂▂▃▂▃▁▁▂▄▂▁▂▂▂▂▁▂█▂▅▅▂▇▅
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.02
Accuracy/val96.91
Loss/train0.00087
Loss/val0.00223
epoch50

View run cosmic-sweep-34 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9g0xstpx
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_010017-9g0xstpx/logs
wandb: Agent Starting Run: hljjsi11 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_010315-hljjsi11
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0017. Train Acc: 90.4575, Test loss: 0.0017. Test Acc: 90.5300. Time/epoch: 2.2875
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0008. Train Acc: 95.6600, Test loss: 0.0009. Test Acc: 95.4700. Time/epoch: 2.4250
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0006. Train Acc: 96.7150, Test loss: 0.0008. Test Acc: 95.9700. Time/epoch: 2.4188
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0003. Train Acc: 98.7750, Test loss: 0.0005. Test Acc: 97.8700. Time/epoch: 2.2719
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0003. Train Acc: 98.6125, Test loss: 0.0006. Test Acc: 97.6300. Time/epoch: 2.2986
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0002. Train Acc: 98.8025, Test loss: 0.0007. Test Acc: 97.2700. Time/epoch: 2.4233
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▄▅▆▆▆▆▅▇▆▇▇▆▇▇▆▇▇▇▇█▇▇█▆▇█▇██▇█▇██▇███
Accuracy/val▁▄▅▅▆▆▇▇▆▇▇▇▇▆▇▇▆█▇▇▇█▇▇█▆▇█▇██▇█▇██▇▇█▇
Loss/train█▆▅▅▄▄▃▃▄▂▃▂▂▃▃▃▃▂▂▂▂▂▂▂▁▃▂▁▂▁▁▂▁▂▁▁▂▁▁▁
Loss/val█▅▄▄▃▃▂▂▃▂▂▁▁▂▂▂▂▁▁▂▂▁▂▃▁▄▁▁▂▁▁▂▁▂▁▁▂▂▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.8025
Accuracy/val97.27
Loss/train0.00024
Loss/val0.00066
epoch50

View run whole-sweep-35 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/hljjsi11
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_010315-hljjsi11/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: nt3b11pr with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_010539-nt3b11pr
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0111. Train Acc: 86.6875, Test loss: 0.0114. Test Acc: 86.5300. Time/epoch: 4.6194
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0052. Train Acc: 93.3725, Test loss: 0.0054. Test Acc: 93.3400. Time/epoch: 4.7821
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▇▇▄▇▇█
Accuracy/val▁▃▄▅▆▇▇▄█▇█
Loss/train█▅▄▃▂▂▂▄▁▂▁
Loss/val█▅▄▃▂▂▂▄▁▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.3725
Accuracy/val93.34
Loss/train0.00515
Loss/val0.00545
epoch10

View run apricot-sweep-36 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/nt3b11pr
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_010539-nt3b11pr/logs
wandb: Agent Starting Run: zvbsrho9 with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_010644-zvbsrho9
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0242. Train Acc: 36.8250, Test loss: 0.0243. Test Acc: 37.5400. Time/epoch: 3.0688
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0244. Train Acc: 36.8250, Test loss: 0.0244. Test Acc: 37.5400. Time/epoch: 3.0125
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▂█▂▁▆▄▇▁▇▆▇
Loss/val▂█▂▁▅▄▇▁▆▆▇
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02439
Loss/val0.02443
epoch10

View run soft-sweep-37 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/zvbsrho9
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_010644-zvbsrho9/logs
wandb: Agent Starting Run: js3qtjb0 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_010733-js3qtjb0
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0067. Train Acc: 85.3000, Test loss: 0.0068. Test Acc: 85.8900. Time/epoch: 2.9926
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0017. Train Acc: 95.5425, Test loss: 0.0018. Test Acc: 95.5200. Time/epoch: 2.9669
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0018. Train Acc: 95.4600, Test loss: 0.0020. Test Acc: 95.2000. Time/epoch: 3.1107
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0016. Train Acc: 96.0800, Test loss: 0.0017. Test Acc: 95.7800. Time/epoch: 3.0781
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0013. Train Acc: 96.7250, Test loss: 0.0015. Test Acc: 96.4200. Time/epoch: 3.1107
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0012. Train Acc: 97.1975, Test loss: 0.0014. Test Acc: 96.9500. Time/epoch: 3.0805
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▅▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇█▇▇█▇██████████▇████
Accuracy/val▁▄▅▅▆▆▇▇▇▇▇▇▇▇▇▇▇█▇▇█▇▇█▇▇█████████▇████
Loss/train█▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▂▁▁▂▁▁▁▁▁▁▁▁▁▁▂▁▁▁▁
Loss/val█▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▂▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.1975
Accuracy/val96.95
Loss/train0.00116
Loss/val0.00137
epoch50

View run silvery-sweep-38 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/js3qtjb0
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_010733-js3qtjb0/logs
wandb: Agent Starting Run: 2svh3oa2 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_011021-2svh3oa2
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0060. Train Acc: 92.1275, Test loss: 0.0061. Test Acc: 91.9300. Time/epoch: 4.8761
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0027. Train Acc: 96.5675, Test loss: 0.0030. Test Acc: 96.0500. Time/epoch: 4.8464
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▃▁▅▇▇█▇▇▇██
Accuracy/val▃▁▅▇▇█▇█▇█▇
Loss/train▇█▄▂▂▁▂▁▂▁▁
Loss/val▇█▄▂▂▁▂▁▂▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train96.5675
Accuracy/val96.05
Loss/train0.00269
Loss/val0.00296
epoch10

View run different-sweep-39 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/2svh3oa2
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_011021-2svh3oa2/logs
wandb: Agent Starting Run: pcr7q62f with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_011127-pcr7q62f
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0039. Train Acc: 88.8100, Test loss: 0.0039. Test Acc: 89.1000. Time/epoch: 3.0240
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0017. Train Acc: 95.9075, Test loss: 0.0018. Test Acc: 95.7000. Time/epoch: 3.1668
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0014. Train Acc: 96.3800, Test loss: 0.0016. Test Acc: 96.1700. Time/epoch: 3.0218
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0011. Train Acc: 97.4375, Test loss: 0.0013. Test Acc: 97.1200. Time/epoch: 2.9951
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0010. Train Acc: 97.6875, Test loss: 0.0012. Test Acc: 97.4000. Time/epoch: 3.1581
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 97.8525, Test loss: 0.0011. Test Acc: 97.5400. Time/epoch: 3.1793
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▂▅▆▆▆▆▆▇▆▇▇▇▇▇▇█▇▇█▆▇█▇▆█▇██████████
Accuracy/val▁▄▅▅▁▅▆▆▆▆▆▇▆▇▇▇▇▇▇█▇▇█▆███▆█▇██████████
Loss/train█▅▄▄▇▄▄▃▃▃▃▂▃▂▂▂▂▂▂▁▂▂▁▃▂▁▂▃▁▃▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▄▇▄▄▃▃▃▃▂▃▂▂▂▂▂▂▁▂▂▁▃▂▁▁▃▁▂▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.8525
Accuracy/val97.54
Loss/train0.00087
Loss/val0.0011
epoch50

View run golden-sweep-40 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/pcr7q62f
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_011127-pcr7q62f/logs
wandb: Agent Starting Run: tq2yb5rq with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_011419-tq2yb5rq
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0269. Train Acc: 62.7025, Test loss: 0.0267. Test Acc: 63.3500. Time/epoch: 4.7851
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0141. Train Acc: 79.1350, Test loss: 0.0141. Test Acc: 79.7000. Time/epoch: 4.7103
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▂▁▂▄▇█▇████
Accuracy/val▂▁▂▄▇█▇█▇██
Loss/train██▇▇▃▁▂▁▂▁▁
Loss/val██▇▇▃▁▂▁▂▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train79.135
Accuracy/val79.7
Loss/train0.01414
Loss/val0.01411
epoch10

View run vibrant-sweep-41 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/tq2yb5rq
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_011419-tq2yb5rq/logs
wandb: Agent Starting Run: xm0g4eyc with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_011526-xm0g4eyc
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0983
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1080
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 2.9523
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▂▂▃▃▄▂▃▃▂█▇▇▂▂▁▁▃▅▃▁▃
Loss/val▅▇▃▄▅▅▃▅▅▇▅█▅▅▁▄▅▆▅▃█
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02403
Loss/val0.02408
epoch20

View run robust-sweep-42 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xm0g4eyc
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_011526-xm0g4eyc/logs
wandb: Agent Starting Run: oibs8tj5 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_011642-oibs8tj5
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.1351
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0241. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1293
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 2.9743
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 2.9681
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0801
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1191
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁█▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁█▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train██████████████████████████▁█████████████
Loss/val██████████████████████████▁█████████████
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02402
Loss/val0.02406
epoch50

View run olive-sweep-43 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/oibs8tj5
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_011642-oibs8tj5/logs
wandb: Agent Starting Run: td3idvxx with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_011929-td3idvxx
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0060. Train Acc: 80.2150, Test loss: 0.0060. Test Acc: 80.0600. Time/epoch: 2.3139
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0013. Train Acc: 93.2350, Test loss: 0.0014. Test Acc: 93.0400. Time/epoch: 2.4386
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0011. Train Acc: 93.9750, Test loss: 0.0012. Test Acc: 94.0900. Time/epoch: 2.4276
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▇▇▇▇▇▇███████████
Accuracy/val▁▅▆▆▇▇▇▇▇█▇██████████
Loss/train█▄▃▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▄▃▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train93.975
Accuracy/val94.09
Loss/train0.00114
Loss/val0.0012
epoch20

View run dauntless-sweep-44 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/td3idvxx
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_011929-td3idvxx/logs
wandb: Agent Starting Run: cet2jps7 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_012032-cet2jps7
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0066. Train Acc: 91.0700, Test loss: 0.0067. Test Acc: 91.1100. Time/epoch: 5.0446
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0029. Train Acc: 96.4875, Test loss: 0.0034. Test Acc: 96.0800. Time/epoch: 5.1598
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0022. Train Acc: 97.2775, Test loss: 0.0034. Test Acc: 96.4800. Time/epoch: 5.1723
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0009. Train Acc: 99.0950, Test loss: 0.0022. Test Acc: 97.9000. Time/epoch: 5.1286
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0011. Train Acc: 98.7550, Test loss: 0.0031. Test Acc: 97.5600. Time/epoch: 5.1539
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0007. Train Acc: 99.1825, Test loss: 0.0029. Test Acc: 97.8000. Time/epoch: 5.0007
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▃▄▄▄▅▆▆▆▆▇▆▄▇▅▆▇▅▇▇▆▇▇█▇█▇▇▇█▇▇█▇▆▇███
Accuracy/val▁▄▄▄▅▄▆▇▆▆▇▇▇▄▇▅▆▇▅▇▇▆▇▇█▇█▇▇▇█▇▇█▇▆▇███
Loss/train█▅▅▅▅▅▄▃▄▃▃▃▃▅▃▄▃▂▄▂▂▃▃▃▁▂▁▂▂▂▁▂▃▁▃▃▂▁▁▁
Loss/val█▄▅▅▄▅▃▁▃▂▂▂▂▅▂▅▃▂▅▁▂▃▃▃▁▂▁▂▂▂▁▂▃▁▄▅▂▁▂▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train99.1825
Accuracy/val97.8
Loss/train0.00071
Loss/val0.00285
epoch50

View run charmed-sweep-45 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/cet2jps7
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_012032-cet2jps7/logs
wandb: Agent Starting Run: wgxc0r89 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_012506-wgxc0r89
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0057. Train Acc: 88.1100, Test loss: 0.0058. Test Acc: 87.8200. Time/epoch: 3.1096
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0025. Train Acc: 93.7300, Test loss: 0.0026. Test Acc: 93.8200. Time/epoch: 2.9596
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0022. Train Acc: 94.6750, Test loss: 0.0023. Test Acc: 94.5100. Time/epoch: 2.9497
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0020. Train Acc: 95.2350, Test loss: 0.0021. Test Acc: 95.1200. Time/epoch: 3.0862
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0019. Train Acc: 95.4075, Test loss: 0.0020. Test Acc: 95.2700. Time/epoch: 3.0906
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0018. Train Acc: 95.4550, Test loss: 0.0019. Test Acc: 95.4600. Time/epoch: 2.9645
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▆▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇████▇██████▇▇█████
Accuracy/val▁▃▄▅▆▆▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇██▇▇██████▇▇█████
Loss/train█▅▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.455
Accuracy/val95.46
Loss/train0.0018
Loss/val0.00191
epoch50

View run stilted-sweep-46 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/wgxc0r89
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_012506-wgxc0r89/logs
wandb: Agent Starting Run: wpego9ua with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_012754-wpego9ua
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0150. Train Acc: 58.0100, Test loss: 0.0149. Test Acc: 58.5900. Time/epoch: 3.1601
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0320
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train█▁▁▁▁▁▁▁▁▁▁
Accuracy/val█▁▁▁▁▁▁▁▁▁▁
Loss/train▁██████████
Loss/val▁██████████
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02403
Loss/val0.02405
epoch10

View run visionary-sweep-47 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/wpego9ua
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_012754-wpego9ua/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: ntymp54i with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_012850-ntymp54i
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0062. Train Acc: 75.4750, Test loss: 0.0063. Test Acc: 75.7900. Time/epoch: 2.4605
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0018. Train Acc: 89.3325, Test loss: 0.0019. Test Acc: 89.1500. Time/epoch: 2.4217
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0014. Train Acc: 92.4650, Test loss: 0.0015. Test Acc: 92.3500. Time/epoch: 2.2873
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0011. Train Acc: 94.1775, Test loss: 0.0012. Test Acc: 94.1200. Time/epoch: 2.2804
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0010. Train Acc: 95.0900, Test loss: 0.0010. Test Acc: 95.1300. Time/epoch: 2.4266
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 95.5025, Test loss: 0.0010. Test Acc: 95.5600. Time/epoch: 2.4387
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▅▅▅▅▆▆▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇█████████████████
Accuracy/val▁▃▅▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇███████████████
Loss/train█▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.5025
Accuracy/val95.56
Loss/train0.0009
Loss/val0.00096
epoch50

View run bright-sweep-48 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ntymp54i
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_012850-ntymp54i/logs
wandb: Agent Starting Run: 8e8t06bq with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_013103-8e8t06bq
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0130. Train Acc: 72.5800, Test loss: 0.0128. Test Acc: 72.8100. Time/epoch: 3.1799
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0030. Train Acc: 92.1975, Test loss: 0.0032. Test Acc: 91.9100. Time/epoch: 3.1378
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0027. Train Acc: 93.4900, Test loss: 0.0028. Test Acc: 93.1500. Time/epoch: 3.1438
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0021. Train Acc: 95.3625, Test loss: 0.0023. Test Acc: 95.0000. Time/epoch: 3.0164
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0020. Train Acc: 95.2875, Test loss: 0.0022. Test Acc: 94.8700. Time/epoch: 3.1641
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0020. Train Acc: 95.5525, Test loss: 0.0022. Test Acc: 95.2500. Time/epoch: 3.1630
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▆▇▂▄▅▆▇▇▇▇▇▇█▇▇▇██▇████████████▇██████
Accuracy/val▁▃▇▇▂▄▅▆▇▇▇▇▇▇█▇▇▇██▇████████████▇██████
Loss/train█▄▃▂▅▃▃▂▂▂▁▁▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▄▃▂▅▄▃▂▂▂▁▁▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▂▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.5525
Accuracy/val95.25
Loss/train0.00197
Loss/val0.00217
epoch50

View run confused-sweep-49 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/8e8t06bq
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_013103-8e8t06bq/logs
wandb: Agent Starting Run: rvcbs4dg with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_013355-rvcbs4dg
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0038. Train Acc: 82.9650, Test loss: 0.0038. Test Acc: 82.7400. Time/epoch: 2.1988
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0013. Train Acc: 93.0050, Test loss: 0.0013. Test Acc: 93.0800. Time/epoch: 2.3553
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 95.3475, Test loss: 0.0010. Test Acc: 95.1500. Time/epoch: 2.3193
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0010. Train Acc: 94.9400, Test loss: 0.0010. Test Acc: 94.9600. Time/epoch: 2.3394
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0009. Train Acc: 95.6050, Test loss: 0.0009. Test Acc: 95.5400. Time/epoch: 2.1788
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0010. Train Acc: 95.0950, Test loss: 0.0010. Test Acc: 95.0300. Time/epoch: 2.1963
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▇▇▇▇▇▇▆▇██▅███▇██▇▇███████▇▇█▇▇██▇█████
Accuracy/val▁▇▇▇█▇▇▆▇██▅▇█████▇▇█████▇█▇▇█▇▇████████
Loss/train█▂▂▂▁▁▁▂▂▁▁▃▁▁▁▁▁▁▂▂▁▁▁▁▁▁▁▁▂▁▁▂▁▁▁▁▁▁▁▁
Loss/val█▂▂▂▁▁▁▃▂▁▁▃▁▁▁▁▁▁▂▂▁▁▁▁▁▁▁▁▂▁▁▂▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.095
Accuracy/val95.03
Loss/train0.00097
Loss/val0.001
epoch50

View run peach-sweep-50 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/rvcbs4dg
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_013355-rvcbs4dg/logs
wandb: Agent Starting Run: ykrhvl2b with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.001
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_013602-ykrhvl2b
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0014. Train Acc: 92.3400, Test loss: 0.0014. Test Acc: 92.4800. Time/epoch: 2.4528
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0008. Train Acc: 95.9200, Test loss: 0.0010. Test Acc: 95.4400. Time/epoch: 2.4371
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 97.6450, Test loss: 0.0008. Test Acc: 96.5700. Time/epoch: 2.2827
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▂▃▂▄▅▁▆▆▆▅▆▆▆▅▆▇▇▅▇██
Accuracy/val▂▃▂▄▆▁▇▆▆▅▆▆▆▆▆▇▇▅▇██
Loss/train▇▇▇▅▅█▃▃▃▅▄▃▄▄▄▂▂▄▃▁▁
Loss/val▇▆▇▄▄█▂▂▂▄▃▄▃▄▄▂▂▄▃▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train97.645
Accuracy/val96.57
Loss/train0.00049
Loss/val0.00079
epoch20

View run different-sweep-51 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ykrhvl2b
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_013602-ykrhvl2b/logs
wandb: Agent Starting Run: ru1sh0ei with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_013704-ru1sh0ei
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0076. Train Acc: 77.3225, Test loss: 0.0075. Test Acc: 78.0900. Time/epoch: 3.1420
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0026. Train Acc: 93.9200, Test loss: 0.0029. Test Acc: 93.4000. Time/epoch: 3.1229
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0020. Train Acc: 95.1150, Test loss: 0.0024. Test Acc: 94.2000. Time/epoch: 2.9649
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0017. Train Acc: 96.0825, Test loss: 0.0023. Test Acc: 95.0600. Time/epoch: 2.9383
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0017. Train Acc: 96.3525, Test loss: 0.0022. Test Acc: 95.1500. Time/epoch: 3.1102
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0016. Train Acc: 95.8775, Test loss: 0.0022. Test Acc: 94.9000. Time/epoch: 3.0799
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▆▇▇▇▇▇▇▇▇▇▇█▇▇█▇████▇████▇██████▇████
Accuracy/val▁▅▅▆▇▇▇▇▇▇▇▇▇██▇▇█▇████▇████▇██████▇████
Loss/train█▄▄▃▃▂▂▂▃▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▂▁▂▁▂▁▁▁▁▁▁▂▁▁▁▁
Loss/val█▄▄▃▂▂▂▂▂▂▂▂▂▂▁▂▂▁▂▁▁▂▁▂▂▁▁▁▂▁▁▁▁▁▁▃▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.8775
Accuracy/val94.9
Loss/train0.00158
Loss/val0.00222
epoch50

View run polished-sweep-52 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ru1sh0ei
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_013704-ru1sh0ei/logs
wandb: Agent Starting Run: pxhxw92y with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_013951-pxhxw92y
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0042. Train Acc: 87.8425, Test loss: 0.0043. Test Acc: 87.7200. Time/epoch: 3.0504
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0023. Train Acc: 94.3600, Test loss: 0.0024. Test Acc: 94.1600. Time/epoch: 3.1346
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▆▆▆▅▆▆█▇█
Accuracy/val▁▄▆▅▆▆▆▆█▇█
Loss/train█▅▄▄▃▄▃▃▂▂▁
Loss/val█▅▄▄▃▄▃▃▁▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train94.36
Accuracy/val94.16
Loss/train0.00226
Loss/val0.00243
epoch10

View run pretty-sweep-53 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/pxhxw92y
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_013951-pxhxw92y/logs
wandb: Agent Starting Run: um2tdmbb with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_014042-um2tdmbb
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0174. Train Acc: 60.3375, Test loss: 0.0172. Test Acc: 60.8500. Time/epoch: 3.0297
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0046. Train Acc: 87.9275, Test loss: 0.0048. Test Acc: 87.9800. Time/epoch: 3.1036
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0024. Train Acc: 93.5025, Test loss: 0.0025. Test Acc: 93.5100. Time/epoch: 3.1106
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▂▂▂▂▂▂▂▅▇▇█████████
Accuracy/val▁▂▂▂▂▂▂▂▂▅▇▇█████████
Loss/train█▇▇▇▇▆▆▆▆▃▂▁▁▁▁▁▁▁▁▁▁
Loss/val█▇▇▇▇▆▆▆▆▃▂▁▁▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train93.5025
Accuracy/val93.51
Loss/train0.00244
Loss/val0.00249
epoch20

View run robust-sweep-54 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/um2tdmbb
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_014042-um2tdmbb/logs
wandb: Agent Starting Run: hcn68694 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_014159-hcn68694
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0060. Train Acc: 91.3650, Test loss: 0.0064. Test Acc: 91.2400. Time/epoch: 4.5928
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0039. Train Acc: 95.1425, Test loss: 0.0040. Test Acc: 95.1500. Time/epoch: 4.7674
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0026. Train Acc: 96.7850, Test loss: 0.0028. Test Acc: 96.5700. Time/epoch: 4.7427
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0034. Train Acc: 95.6125, Test loss: 0.0036. Test Acc: 95.4800. Time/epoch: 4.5831
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0026. Train Acc: 96.7900, Test loss: 0.0031. Test Acc: 96.3800. Time/epoch: 4.5847
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 97.6125, Test loss: 0.0024. Test Acc: 97.1300. Time/epoch: 4.5635
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▃▅▅▆▅▅▅▆▇▇▆▆▆▇▇▇▇▇▇▇▆▇▆█▇█▇▆▇██▇▇█▇█▇█
Accuracy/val▁▁▃▅▆▆▅▅▆▆▇▇▇▇▇▇▇▇▇▇█▇▆▇▆███▇▇▇███▇█▇█▇█
Loss/train▇█▆▄▄▃▄▄▄▃▃▃▃▃▃▂▂▂▂▂▂▂▃▂▃▂▂▁▂▃▂▁▁▁▂▁▂▁▂▁
Loss/val██▆▄▄▃▄▄▄▃▂▂▃▃▂▂▂▂▂▂▂▁▃▂▃▁▁▁▂▃▂▁▁▁▂▁▂▁▂▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.6125
Accuracy/val97.13
Loss/train0.00193
Loss/val0.00244
epoch50

View run lunar-sweep-55 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/hcn68694
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_014159-hcn68694/logs
wandb: Agent Starting Run: 1j8th984 with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.0001
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_014612-1j8th984
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0014. Train Acc: 92.4975, Test loss: 0.0014. Test Acc: 92.5700. Time/epoch: 2.4395
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0006. Train Acc: 97.0575, Test loss: 0.0007. Test Acc: 96.7800. Time/epoch: 2.2884
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 97.5575, Test loss: 0.0006. Test Acc: 97.2100. Time/epoch: 2.4340
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▄▅▅▆▆▅▇▆▇▇█▇█▇▇█▇▇█
Accuracy/val▁▄▅▅▄▆▆▅▇▆▇▇████▇█▇▇█
Loss/train█▆▅▄▅▃▃▄▂▃▂▂▁▁▁▂▂▁▂▂▁
Loss/val█▆▅▄▅▃▃▄▂▂▂▂▁▁▁▂▂▁▂▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train97.5575
Accuracy/val97.21
Loss/train0.00052
Loss/val0.0006
epoch20

View run driven-sweep-56 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/1j8th984
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_014612-1j8th984/logs
wandb: Agent Starting Run: qhzg268e with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_014718-qhzg268e
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0026. Train Acc: 93.4925, Test loss: 0.0026. Test Acc: 93.5700. Time/epoch: 3.1672
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0016. Train Acc: 96.0975, Test loss: 0.0018. Test Acc: 95.4600. Time/epoch: 3.1687
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 97.6525, Test loss: 0.0014. Test Acc: 97.0900. Time/epoch: 3.3189
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0007. Train Acc: 98.3975, Test loss: 0.0013. Test Acc: 97.3400. Time/epoch: 3.3105
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0006. Train Acc: 98.4825, Test loss: 0.0015. Test Acc: 97.0300. Time/epoch: 3.3044
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0004. Train Acc: 99.1600, Test loss: 0.0013. Test Acc: 97.5900. Time/epoch: 3.3094
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▃▄▄▅▅▅▄▆▆▅▆▆▆▆▆▄▆▆▇▇▇▆▇█▇▇▇██▇▇████▇██
Accuracy/val▁▂▄▄▄▆▆▆▄▇▇▅▇▇▇▇▇▅▅▆▇██▇▇█▇█▆██▆▇█▇██▇██
Loss/train█▇▅▅▅▄▄▄▅▃▃▄▃▃▃▃▃▄▃▃▂▂▂▂▂▂▂▂▂▁▁▂▂▁▁▁▁▁▁▁
Loss/val█▆▄▄▅▃▃▂▄▂▂▄▂▂▂▂▂▄▄▂▁▁▁▂▂▁▂▂▃▂▂▄▄▂▂▂▂▃▃▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train99.16
Accuracy/val97.59
Loss/train0.00038
Loss/val0.00129
epoch50

View run young-sweep-57 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/qhzg268e
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_014718-qhzg268e/logs
wandb: Agent Starting Run: a1co11pt with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015021-a1co11pt
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0025. Train Acc: 87.2775, Test loss: 0.0025. Test Acc: 87.2400. Time/epoch: 2.4782
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0011. Train Acc: 94.2025, Test loss: 0.0012. Test Acc: 93.7700. Time/epoch: 2.4222
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 94.8650, Test loss: 0.0011. Test Acc: 94.3300. Time/epoch: 2.4377
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0008. Train Acc: 96.0625, Test loss: 0.0009. Test Acc: 95.4800. Time/epoch: 2.3026
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0007. Train Acc: 96.4650, Test loss: 0.0008. Test Acc: 95.9900. Time/epoch: 2.4351
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0006. Train Acc: 97.0775, Test loss: 0.0008. Test Acc: 96.6400. Time/epoch: 2.4276
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▃▄▅▅▅▅▆▆▆▇▆▇▇▇▆▇▇▇▇▇▇▇▇▇██████▇███████
Accuracy/val▁▃▄▄▅▅▅▅▆▆▆▇▆▇▇▇▆▇▇▇▇▇▇▇▇▇▇████▇▇███████
Loss/train█▆▅▅▄▃▃▃▃▃▃▂▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▂▁▁▁▁▁▁▁
Loss/val█▆▅▅▃▃▃▃▃▃▃▂▃▂▂▂▃▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.0775
Accuracy/val96.64
Loss/train0.00062
Loss/val0.00075
epoch50

View run dutiful-sweep-58 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/a1co11pt
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015021-a1co11pt/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: 8lbqnae6 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015243-8lbqnae6
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0481. Test Acc: 37.5400. Time/epoch: 4.9159
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.9081
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▇█▄▂▁▅▄▄█▂▂
Loss/val▇█▃▁▁▂▃▄▅▃▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04806
Loss/val0.04796
epoch10

View run toasty-sweep-59 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/8lbqnae6
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015243-8lbqnae6/logs
wandb: Agent Starting Run: 8lxesosr with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015350-8lxesosr
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.1360. Train Acc: 45.6450, Test loss: 0.1391. Test Acc: 45.3500. Time/epoch: 4.7017
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0040. Train Acc: 95.3500, Test loss: 0.0046. Test Acc: 94.9800. Time/epoch: 4.6806
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0043. Train Acc: 95.4000, Test loss: 0.0054. Test Acc: 94.7700. Time/epoch: 4.8745
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▇▆██▇▇█████████▇████
Accuracy/val▁▇▆██▇██████████▇████
Loss/train█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train95.4
Accuracy/val94.77
Loss/train0.00427
Loss/val0.00538
epoch20

View run bright-sweep-60 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/8lxesosr
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015350-8lxesosr/logs
wandb: Agent Starting Run: 24buqjcp with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015551-24buqjcp
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0036. Train Acc: 89.9200, Test loss: 0.0037. Test Acc: 89.8600. Time/epoch: 3.1996
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0018. Train Acc: 95.2600, Test loss: 0.0019. Test Acc: 95.3200. Time/epoch: 3.3460
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0016. Train Acc: 96.1525, Test loss: 0.0017. Test Acc: 95.7300. Time/epoch: 3.3045
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▅▆▆▇▇▆▇▇██▇▇█████
Accuracy/val▁▄▄▅▆▆▇▇▇▇▇███▇▇█████
Loss/train█▆▅▄▃▃▃▂▂▂▂▂▁▁▂▂▁▁▁▁▁
Loss/val█▆▅▄▃▃▂▂▂▂▂▂▁▁▂▂▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train96.1525
Accuracy/val95.73
Loss/train0.00157
Loss/val0.00169
epoch20

View run smart-sweep-61 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/24buqjcp
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015551-24buqjcp/logs
wandb: Agent Starting Run: t7ub61uh with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 3e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015712-t7ub61uh
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0074. Train Acc: 89.7000, Test loss: 0.0076. Test Acc: 89.8600. Time/epoch: 4.9325
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0042. Train Acc: 94.5575, Test loss: 0.0043. Test Acc: 94.6700. Time/epoch: 4.8575
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0039. Train Acc: 94.9975, Test loss: 0.0040. Test Acc: 95.0800. Time/epoch: 4.8236
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▄▆▆▆▆▇▆▆▇▇▆▆█▇█▇██▇
Accuracy/val▁▄▄▆▆▆▆▇▆▆▇▇▆▆█▇█▇█▇█
Loss/train█▅▅▃▃▃▃▂▃▃▂▂▃▃▁▂▁▁▁▂▁
Loss/val█▅▅▃▃▃▃▂▃▃▂▂▃▃▁▂▁▁▁▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train94.9975
Accuracy/val95.08
Loss/train0.00388
Loss/val0.00399
epoch20

View run dandy-sweep-62 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/t7ub61uh
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015712-t7ub61uh/logs
wandb: Agent Starting Run: dlyi88g3 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015909-dlyi88g3
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0025. Train Acc: 87.0400, Test loss: 0.0025. Test Acc: 86.9100. Time/epoch: 2.3753
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0011. Train Acc: 93.8025, Test loss: 0.0011. Test Acc: 93.9700. Time/epoch: 2.3173
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▁▄▇▅▆█▇█▇
Accuracy/val▁▃▁▅▇▆▆█▇█▇
Loss/train█▅▆▄▂▃▃▁▂▁▁
Loss/val█▅▆▄▂▃▃▁▂▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.8025
Accuracy/val93.97
Loss/train0.00112
Loss/val0.00113
epoch10

View run genial-sweep-63 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/dlyi88g3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015909-dlyi88g3/logs
wandb: Agent Starting Run: ak4nng79 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_015949-ak4nng79
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0119. Train Acc: 82.9050, Test loss: 0.0118. Test Acc: 83.2500. Time/epoch: 5.0588
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0042. Train Acc: 94.6050, Test loss: 0.0044. Test Acc: 94.5800. Time/epoch: 5.1459
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0034. Train Acc: 95.9725, Test loss: 0.0035. Test Acc: 95.8900. Time/epoch: 5.1579
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0031. Train Acc: 96.1425, Test loss: 0.0034. Test Acc: 95.9500. Time/epoch: 5.0730
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0027. Train Acc: 96.7825, Test loss: 0.0030. Test Acc: 96.4700. Time/epoch: 4.9959
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0025. Train Acc: 97.0150, Test loss: 0.0028. Test Acc: 96.8200. Time/epoch: 5.0264
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇█████▇█▇█████████████
Accuracy/val▁▄▅▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇█████▇███████████████
Loss/train█▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.015
Accuracy/val96.82
Loss/train0.00253
Loss/val0.0028
epoch50

View run super-sweep-64 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ak4nng79
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_015949-ak4nng79/logs
wandb: Agent Starting Run: 9605oqnw with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_020424-9605oqnw
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0122. Train Acc: 36.8250, Test loss: 0.0123. Test Acc: 37.5400. Time/epoch: 2.2035
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0122. Train Acc: 36.8250, Test loss: 0.0122. Test Acc: 37.5400. Time/epoch: 2.1998
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▄▂▁▁▁▅▃▃▃▃
Loss/val█▄▁▁▃▂▇▂▂▃▃
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01216
Loss/val0.01224
epoch10

View run glorious-sweep-65 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9605oqnw
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_020424-9605oqnw/logs
wandb: Agent Starting Run: 32wrwz33 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_020501-32wrwz33
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0241. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.2372
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.1972
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1618
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.3073
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.3188
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.3351
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▂▂▅▃▁▂▂▁▂▃▂▂▂▂▁▂▂▃▁▂▃▂▁▂▂▂▃▁▁▁▂▁▃▂▂▂▁▁▂
Loss/val█▂▂▅▄▃▂▃▂▃▃▃▅▃▃▂▃▃▂▃▃▆▃▂▂▃▃▃▂▂▁▃▃▃▃▅▁▃▁▃
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02403
Loss/val0.02406
epoch50

View run solar-sweep-66 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/32wrwz33
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_020501-32wrwz33/logs
wandb: Agent Starting Run: ould8q90 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_020804-ould8q90
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0051. Train Acc: 77.3275, Test loss: 0.0052. Test Acc: 77.4300. Time/epoch: 2.4906
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0012. Train Acc: 94.1025, Test loss: 0.0012. Test Acc: 94.1800. Time/epoch: 2.4152
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 94.8325, Test loss: 0.0010. Test Acc: 95.0300. Time/epoch: 2.2949
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0009. Train Acc: 95.5700, Test loss: 0.0009. Test Acc: 95.6500. Time/epoch: 2.2816
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0008. Train Acc: 95.9950, Test loss: 0.0008. Test Acc: 95.9600. Time/epoch: 2.4423
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0007. Train Acc: 96.2450, Test loss: 0.0008. Test Acc: 96.1800. Time/epoch: 2.4232
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▆▇▇▇▇▇▇▇▇▇▇▇▇▇███████████████████████
Accuracy/val▁▄▄▆▇▇▇▇▇▇▇▇▇▇▇▇████████████████████████
Loss/train█▄▄▃▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▄▄▃▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.245
Accuracy/val96.18
Loss/train0.00073
Loss/val0.00079
epoch50

View run fresh-sweep-67 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ould8q90
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_020804-ould8q90/logs
wandb: Agent Starting Run: 7fx470wl with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_021021-7fx470wl
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7811
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7463
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7215
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▄▄▃▁▁▁▅▃▁▄▄▁▂▆▆▆█▁▅▄▃
Loss/val▆▃▃▂▂▂▆▄▁▆▃▂▄▅██▇▁▄▆▂
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04808
Loss/val0.04795
epoch20

View run fine-sweep-68 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/7fx470wl
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_021021-7fx470wl/logs
wandb: Agent Starting Run: 5wjouk0d with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_021213-5wjouk0d
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0013. Train Acc: 92.9325, Test loss: 0.0013. Test Acc: 92.8700. Time/epoch: 2.4512
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0006. Train Acc: 96.8700, Test loss: 0.0007. Test Acc: 96.4300. Time/epoch: 2.4318
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0007. Train Acc: 96.5075, Test loss: 0.0008. Test Acc: 96.0000. Time/epoch: 2.4348
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▄▁▆▅▆▆▇▇▇█▇█▇▇█▇▆▇▇
Accuracy/val▁▂▄▁▅▅▇▇▇▇▇██▇▇▇█▇▅▇▆
Loss/train█▇▅█▄▄▂▂▂▂▂▁▂▁▁▁▁▁▃▁▂
Loss/val█▇▆█▄▄▂▂▂▂▂▁▂▁▂▂▁▂▃▂▃
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train96.5075
Accuracy/val96.0
Loss/train0.00069
Loss/val0.00085
epoch20

View run robust-sweep-69 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/5wjouk0d
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_021213-5wjouk0d/logs
wandb: Agent Starting Run: ualodyxv with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_021315-ualodyxv
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4980
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4229
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4428
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3018
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.2865
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.2946
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▄▂▄▃▅▂▅▃▂▃▁▃▇▂▁▂▄▂▂▂▂▂▂▂▁▂▂▂▂▂▂▁▁▃▂▁▂▁▁
Loss/val▅▄▃▃▃▆▄▄▄█▄▃▅▄▄▂▃▅▃▃▆▂▂▅▄▄▄▃▃▄▃▃▄▃▃▂▆▂▂▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01203
Loss/val0.01208
epoch50

View run worthy-sweep-70 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ualodyxv
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_021315-ualodyxv/logs
wandb: Agent Starting Run: aica9whg with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_021530-aica9whg
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0047. Train Acc: 78.7500, Test loss: 0.0047. Test Acc: 78.8200. Time/epoch: 2.3694
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0017. Train Acc: 89.8825, Test loss: 0.0018. Test Acc: 89.6000. Time/epoch: 2.2194
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0014. Train Acc: 92.1825, Test loss: 0.0015. Test Acc: 91.9300. Time/epoch: 2.3605
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0013. Train Acc: 93.3900, Test loss: 0.0013. Test Acc: 93.2300. Time/epoch: 2.3704
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0012. Train Acc: 93.8375, Test loss: 0.0012. Test Acc: 93.8600. Time/epoch: 2.2167
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0011. Train Acc: 94.0700, Test loss: 0.0012. Test Acc: 94.1000. Time/epoch: 2.2159
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇████████████████
Accuracy/val▁▄▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇███████████████
Loss/train█▅▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train94.07
Accuracy/val94.1
Loss/train0.00113
Loss/val0.00118
epoch50

View run stoic-sweep-71 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/aica9whg
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_021530-aica9whg/logs
wandb: Agent Starting Run: ye109ifx with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_021743-ye109ifx
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0037. Train Acc: 89.5775, Test loss: 0.0039. Test Acc: 89.3200. Time/epoch: 3.1174
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0017. Train Acc: 95.9275, Test loss: 0.0018. Test Acc: 95.7400. Time/epoch: 3.0752
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0014. Train Acc: 96.5225, Test loss: 0.0016. Test Acc: 96.3300. Time/epoch: 2.9782
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0013. Train Acc: 96.7325, Test loss: 0.0015. Test Acc: 96.3300. Time/epoch: 3.1177
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0012. Train Acc: 96.8750, Test loss: 0.0014. Test Acc: 96.6700. Time/epoch: 3.0941
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 97.8125, Test loss: 0.0011. Test Acc: 97.4300. Time/epoch: 3.0942
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▆▆▆▆▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇█▇██▇█████
Accuracy/val▁▄▅▆▆▆▆▆▇▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇██▇██▇█████
Loss/train█▆▅▄▄▄▃▃▃▃▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▁▂▁▁▁▁▁
Loss/val█▆▅▄▄▄▃▃▃▃▃▃▂▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▁▂▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.8125
Accuracy/val97.43
Loss/train0.00092
Loss/val0.00108
epoch50

View run kind-sweep-72 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ye109ifx
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_021743-ye109ifx/logs
wandb: Agent Starting Run: k6x3u2n4 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_022031-k6x3u2n4
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0143. Train Acc: 80.5825, Test loss: 0.0144. Test Acc: 80.4000. Time/epoch: 4.7606
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0045. Train Acc: 94.2575, Test loss: 0.0047. Test Acc: 94.1900. Time/epoch: 4.7643
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0041. Train Acc: 94.9200, Test loss: 0.0043. Test Acc: 94.8200. Time/epoch: 4.7542
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0038. Train Acc: 95.3125, Test loss: 0.0039. Test Acc: 95.0100. Time/epoch: 4.6119
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0038. Train Acc: 95.4925, Test loss: 0.0039. Test Acc: 95.3200. Time/epoch: 4.6420
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0035. Train Acc: 95.5825, Test loss: 0.0037. Test Acc: 95.3700. Time/epoch: 4.6310
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▇▇▇▇▇▇▇█▇█▇▇█████▇█▇████████████████
Accuracy/val▁▅▆▆▇▇▇▇▇▇▇███▇▇█████▇█▇██████▇█████████
Loss/train█▅▄▃▂▂▂▂▂▂▂▁▁▁▁▂▁▁▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.5825
Accuracy/val95.37
Loss/train0.00352
Loss/val0.00369
epoch50

View run bumbling-sweep-73 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/k6x3u2n4
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_022031-k6x3u2n4/logs
wandb: Agent Starting Run: mty627wj with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_022444-mty627wj
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0068. Train Acc: 91.9800, Test loss: 0.0069. Test Acc: 91.9300. Time/epoch: 5.1930
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0018. Train Acc: 97.8925, Test loss: 0.0023. Test Acc: 97.3300. Time/epoch: 5.0212
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0024. Train Acc: 97.1275, Test loss: 0.0032. Test Acc: 96.6400. Time/epoch: 5.1637
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0014. Train Acc: 98.3925, Test loss: 0.0028. Test Acc: 97.4600. Time/epoch: 5.1462
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0009. Train Acc: 98.9500, Test loss: 0.0025. Test Acc: 97.5800. Time/epoch: 5.1569
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0012. Train Acc: 98.6050, Test loss: 0.0037. Test Acc: 97.0200. Time/epoch: 5.0122
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▂▄▅▅▄▇▆▆▆▆▇▆▄▆▆▇▆▆▇▇█▇█▇▇██▇██▇██▆▇██
Accuracy/val▁▅▅▂▄▅▅▄▇▆▆▇▆▇▇▄▆▆▇▆▆█▇█▇█▇▇▇█▇█▇▇█▇▆▇▇▇
Loss/train█▄▄▆▅▄▄▄▂▃▃▃▃▂▂▅▃▂▂▃▃▂▂▁▂▁▂▂▁▁▃▁▁▂▁▁▃▁▁▁
Loss/val█▄▃▆▅▃▃▄▁▂▃▂▂▁▂▅▃▂▂▃▃▁▂▁▂▁▂▃▂▂▅▂▃▃▂▃▅▄▃▃
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.605
Accuracy/val97.02
Loss/train0.00118
Loss/val0.00372
epoch50

View run wobbly-sweep-74 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/mty627wj
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_022444-mty627wj/logs
wandb: Agent Starting Run: yjbe5i0g with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_022918-yjbe5i0g
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0078. Train Acc: 84.5950, Test loss: 0.0078. Test Acc: 84.4300. Time/epoch: 3.3518
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0025. Train Acc: 93.1525, Test loss: 0.0026. Test Acc: 93.2400. Time/epoch: 3.1908
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▇▇▇▇███
Accuracy/val▁▅▆▆▇▇▇▇███
Loss/train█▄▃▂▂▂▁▁▁▁▁
Loss/val█▄▃▂▂▂▁▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.1525
Accuracy/val93.24
Loss/train0.00253
Loss/val0.00261
epoch10

View run curious-sweep-75 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/yjbe5i0g
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_022918-yjbe5i0g/logs
wandb: Agent Starting Run: 6048t5bg with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023008-6048t5bg
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0020. Train Acc: 90.9550, Test loss: 0.0020. Test Acc: 91.2400. Time/epoch: 2.4864
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0010. Train Acc: 95.1750, Test loss: 0.0010. Test Acc: 95.1600. Time/epoch: 2.4245
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▆▆▇▇▇▇▇█
Accuracy/val▁▄▅▆▆▇▇▇▇▇█
Loss/train█▅▄▃▂▂▂▂▁▂▁
Loss/val█▄▃▃▂▂▂▂▁▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train95.175
Accuracy/val95.16
Loss/train0.00097
Loss/val0.00101
epoch10

View run clean-sweep-76 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/6048t5bg
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023008-6048t5bg/logs
wandb: Agent Starting Run: mgan5bka with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023051-mgan5bka
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0055. Train Acc: 85.7550, Test loss: 0.0055. Test Acc: 86.4100. Time/epoch: 3.1400
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0028. Train Acc: 93.4300, Test loss: 0.0029. Test Acc: 93.2700. Time/epoch: 3.1073
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0024. Train Acc: 94.1525, Test loss: 0.0025. Test Acc: 94.1400. Time/epoch: 3.0903
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0022. Train Acc: 94.7525, Test loss: 0.0023. Test Acc: 94.7300. Time/epoch: 2.9781
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0020. Train Acc: 95.3200, Test loss: 0.0021. Test Acc: 95.2500. Time/epoch: 2.9601
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 95.0275, Test loss: 0.0020. Test Acc: 95.1000. Time/epoch: 2.9362
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇█▇████████▇████████
Accuracy/val▁▃▅▅▅▆▆▆▆▆▆▇▇▆▇▇▇▇▇▇▇█▇█▇█▇█████████████
Loss/train█▅▄▄▄▃▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▅▄▄▃▃▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.0275
Accuracy/val95.1
Loss/train0.00194
Loss/val0.00204
epoch50

View run dashing-sweep-77 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/mgan5bka
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023051-mgan5bka/logs
wandb: Agent Starting Run: 8yc0qcsu with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023338-8yc0qcsu
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0021. Train Acc: 88.7475, Test loss: 0.0022. Test Acc: 88.9100. Time/epoch: 2.3702
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0012. Train Acc: 93.6400, Test loss: 0.0012. Test Acc: 94.0300. Time/epoch: 2.2140
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▆▇▄▃█▇▁█
Accuracy/val▁▃▄▆▇▄▃▇▆▁█
Loss/train█▅▄▃▂▃▅▂▃▇▁
Loss/val█▅▄▃▂▃▅▂▃▇▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.64
Accuracy/val94.03
Loss/train0.00121
Loss/val0.00124
epoch10

View run olive-sweep-78 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/8yc0qcsu
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023338-8yc0qcsu/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: w3rprb9t with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023425-w3rprb9t
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.2045
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.1797
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▁█▆▃▆▄▂▄▃▄▃
Loss/val▁█▆▅▆▃▆▇▂▂▅
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02402
Loss/val0.02407
epoch10

View run noble-sweep-79 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/w3rprb9t
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023425-w3rprb9t/logs
wandb: Agent Starting Run: efloip58 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023511-efloip58
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0044. Train Acc: 90.6625, Test loss: 0.0044. Test Acc: 90.9400. Time/epoch: 3.1424
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0024. Train Acc: 94.0600, Test loss: 0.0024. Test Acc: 94.2700. Time/epoch: 3.1304
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0018. Train Acc: 95.4925, Test loss: 0.0019. Test Acc: 95.5100. Time/epoch: 2.9462
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0016. Train Acc: 96.1650, Test loss: 0.0017. Test Acc: 96.1700. Time/epoch: 3.0749
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0016. Train Acc: 95.9875, Test loss: 0.0017. Test Acc: 95.9400. Time/epoch: 3.0883
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0016. Train Acc: 95.9375, Test loss: 0.0017. Test Acc: 95.6600. Time/epoch: 3.1140
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▄▅▅▆▆▆▅▅▆▅▆▆▆▇▇▆▆▇▆▇▇▇▇▇▇█▇▇▇▇█▇█████▇
Accuracy/val▁▄▄▅▅▆▅▆▅▅▆▅▆▆▆▇▇▆▆▇▆▇▇▇█▇▇█▇▇█▇█▇█████▇
Loss/train█▅▄▄▃▃▃▃▃▃▂▃▂▂▂▂▂▂▃▂▂▂▂▂▁▂▂▁▁▂▁▁▁▂▁▁▁▁▁▁
Loss/val█▅▄▄▃▃▃▃▃▃▂▃▂▂▃▂▂▂▃▂▂▂▂▂▁▂▂▁▁▂▁▁▁▂▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.9375
Accuracy/val95.66
Loss/train0.00159
Loss/val0.00169
epoch50

View run dainty-sweep-80 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/efloip58
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023511-efloip58/logs
wandb: Agent Starting Run: 58g5zuda with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023759-58g5zuda
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 5.2325
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0479. Test Acc: 37.5400. Time/epoch: 5.1512
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▃▆▅▇▃█▄▁▃▁
Loss/val█▂▆▄█▃▄▃▁▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04805
Loss/val0.04794
epoch10

View run polar-sweep-81 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/58g5zuda
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023759-58g5zuda/logs
wandb: Agent Starting Run: 69htsr9g with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023909-69htsr9g
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0041. Train Acc: 76.7550, Test loss: 0.0041. Test Acc: 77.0200. Time/epoch: 2.2282
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0013. Train Acc: 92.8875, Test loss: 0.0014. Test Acc: 92.7800. Time/epoch: 2.3363
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▆▆▇▅▇▇███▇
Accuracy/val▁▆▆▇▅▇▇███▇
Loss/train█▃▃▂▄▂▂▁▁▁▁
Loss/val█▃▃▂▅▂▂▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train92.8875
Accuracy/val92.78
Loss/train0.00131
Loss/val0.00139
epoch10

View run stellar-sweep-82 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/69htsr9g
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023909-69htsr9g/logs
wandb: Agent Starting Run: zykk0qgn with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_023949-zykk0qgn
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0110. Train Acc: 86.2175, Test loss: 0.0112. Test Acc: 86.1000. Time/epoch: 5.0716
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0043. Train Acc: 94.6700, Test loss: 0.0045. Test Acc: 94.7200. Time/epoch: 5.0484
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0035. Train Acc: 95.6950, Test loss: 0.0037. Test Acc: 95.6200. Time/epoch: 5.1391
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▆▆▇▇▇▇▇▇▇███▇█▇█
Accuracy/val▁▃▄▅▆▆▆▇▇▇▇▇▇▇███▇█▇█
Loss/train█▅▄▄▃▃▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁
Loss/val█▅▄▄▃▃▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train95.695
Accuracy/val95.62
Loss/train0.00349
Loss/val0.00367
epoch20

View run cerulean-sweep-83 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/zykk0qgn
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_023949-zykk0qgn/logs
wandb: Agent Starting Run: lpnx987r with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_024151-lpnx987r
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0029. Train Acc: 92.3125, Test loss: 0.0029. Test Acc: 92.3300. Time/epoch: 3.0446
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0023. Train Acc: 93.5900, Test loss: 0.0024. Test Acc: 93.7000. Time/epoch: 3.0265
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▆▇▇▅▆▄█▄▄▆
Accuracy/val▁▇▇▇▅▆▅█▃▄▆
Loss/train█▅▄▅▃▃▄▁▄▄▃
Loss/val█▄▄▄▃▃▄▁▄▄▃
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.59
Accuracy/val93.7
Loss/train0.00235
Loss/val0.00236
epoch10

View run magic-sweep-84 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/lpnx987r
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_024151-lpnx987r/logs
wandb: Agent Starting Run: 1sytqfqj with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_024239-1sytqfqj
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0019. Train Acc: 91.2225, Test loss: 0.0019. Test Acc: 91.1300. Time/epoch: 2.3752
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0010. Train Acc: 95.2700, Test loss: 0.0010. Test Acc: 95.0900. Time/epoch: 2.1876
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0008. Train Acc: 95.9575, Test loss: 0.0008. Test Acc: 95.9300. Time/epoch: 2.3217
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0008. Train Acc: 95.8325, Test loss: 0.0009. Test Acc: 95.5600. Time/epoch: 2.3351
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0007. Train Acc: 96.5900, Test loss: 0.0007. Test Acc: 96.3200. Time/epoch: 2.3392
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0007. Train Acc: 96.2000, Test loss: 0.0008. Test Acc: 95.9700. Time/epoch: 2.1825
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▂▁▅▂▆▆▅▅▆▆▆▇▇▇▇▇▇▇▇▇▇▇█▇▇▇▆█▇██████▇█▇█▇
Accuracy/val▂▁▅▂▆▆▆▆▆▅▆▇▆▇▇▇▇▇▇▇▇▇██▇▇▆█▇██████▇█▇█▇
Loss/train█▇▄▆▃▃▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▁▂▂▂▂▁▂▁▁▁▁▁▁▁▁▂▁▂
Loss/val█▇▄▆▃▃▃▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▁▁▂▂▂▁▂▁▁▁▁▁▁▁▁▂▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.2
Accuracy/val95.97
Loss/train0.00074
Loss/val0.0008
epoch50

View run eager-sweep-85 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/1sytqfqj
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_024239-1sytqfqj/logs
wandb: Agent Starting Run: b6eyrf08 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_024450-b6eyrf08
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0047. Train Acc: 86.9175, Test loss: 0.0047. Test Acc: 86.9200. Time/epoch: 3.2241
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0015. Train Acc: 96.1675, Test loss: 0.0016. Test Acc: 96.0600. Time/epoch: 3.0156
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 97.7150, Test loss: 0.0012. Test Acc: 97.2600. Time/epoch: 3.0039
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0007. Train Acc: 98.4500, Test loss: 0.0009. Test Acc: 97.9400. Time/epoch: 3.1780
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0005. Train Acc: 98.8300, Test loss: 0.0009. Test Acc: 98.2600. Time/epoch: 3.1700
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0006. Train Acc: 98.6625, Test loss: 0.0009. Test Acc: 97.7700. Time/epoch: 3.1696
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▅▆▆▇▆▆▇▆▇▇▇▇▆▇▇▇▇▇▇▇██▆▇██▇███▇██████
Accuracy/val▁▅▅▅▆▆▇▆▇▇▇▇▇▇▇▇▇▇▇▇█████▆███▇██████████
Loss/train█▅▄▄▃▃▃▃▃▂▃▂▂▂▂▃▂▂▂▂▂▂▂▁▁▃▂▁▁▂▁▂▁▂▁▁▁▁▁▁
Loss/val█▄▄▄▃▃▂▃▂▂▂▂▂▂▂▃▁▁▂▂▁▁▁▁▁▃▁▁▁▂▁▁▁▂▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.6625
Accuracy/val97.77
Loss/train0.00057
Loss/val0.00095
epoch50

View run radiant-sweep-86 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/b6eyrf08
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_024450-b6eyrf08/logs
wandb: Agent Starting Run: w3vdhzz8 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_024743-w3vdhzz8
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4030
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3531
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0122. Test Acc: 37.5400. Time/epoch: 2.2054
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0122. Test Acc: 37.5400. Time/epoch: 2.2200
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0122. Test Acc: 37.5400. Time/epoch: 2.3810
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3692
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▃▆▅█▆▅▅▅▂▇▂▆▂█▄▅▄▄▅▃▅▄▂▄▃▁▂▄▂▂▆▂▁▁▂▃▂▂▄▁
Loss/val▂▆▆█▆▆▅▅▁▆▃▇▃█▄▅▄▄▅▄▅▄▂▅▅▂▂▄▁▂▆▂▂▁▃▄▃▂▃▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01205
Loss/val0.01212
epoch50

View run prime-sweep-87 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/w3vdhzz8
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_024743-w3vdhzz8/logs
wandb: Agent Starting Run: eu9lai07 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_024954-eu9lai07
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0044. Train Acc: 75.5350, Test loss: 0.0044. Test Acc: 76.2800. Time/epoch: 2.2035
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0013. Train Acc: 93.4050, Test loss: 0.0014. Test Acc: 93.2300. Time/epoch: 2.3293
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▆▇▇█▇█████
Accuracy/val▁▆▇▇█▇█████
Loss/train█▃▂▂▁▂▁▁▁▁▁
Loss/val█▃▂▂▁▂▁▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.405
Accuracy/val93.23
Loss/train0.00134
Loss/val0.00142
epoch10

View run gentle-sweep-88 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/eu9lai07
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_024954-eu9lai07/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: u7xrakak with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_025041-u7xrakak
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0033. Train Acc: 81.8575, Test loss: 0.0033. Test Acc: 81.8200. Time/epoch: 2.2008
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0013. Train Acc: 93.2875, Test loss: 0.0013. Test Acc: 93.1900. Time/epoch: 2.1925
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0013. Train Acc: 93.7175, Test loss: 0.0013. Test Acc: 93.5500. Time/epoch: 2.3314
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0009. Train Acc: 95.6000, Test loss: 0.0010. Test Acc: 95.1500. Time/epoch: 2.3217
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0008. Train Acc: 96.1400, Test loss: 0.0009. Test Acc: 95.5400. Time/epoch: 2.2083
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0007. Train Acc: 96.5650, Test loss: 0.0009. Test Acc: 95.7400. Time/epoch: 2.2010
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▆▆▆▆▅▆▆▇▆▇▇▇▇▇▇▇▆▇██▇▇██████████▇█████
Accuracy/val▁▄▆▆▇▇▆▆▇▇▇▇▇▇▇▇▇█▇███████████████▇█████
Loss/train█▆▄▃▃▃▄▃▃▂▂▂▂▂▂▂▃▂▃▂▂▁▁▂▁▁▁▁▁▁▁▁▁▁▂▁▁▁▁▁
Loss/val█▆▃▃▂▂▄▃▂▂▂▂▂▂▂▂▂▂▃▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▂▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.565
Accuracy/val95.74
Loss/train0.0007
Loss/val0.00088
epoch50

View run efficient-sweep-89 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/u7xrakak
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_025041-u7xrakak/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: p4whnt5d with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_025259-p4whnt5d
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4557
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3005
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▃▂▁▄▇▄▆▅█▃▂
Loss/val▄▄▁▄▆▅██▅▂▇
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01203
Loss/val0.01213
epoch10

View run zesty-sweep-90 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/p4whnt5d
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_025259-p4whnt5d/logs
wandb: Agent Starting Run: 9mzw55p6 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_025339-9mzw55p6
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0041. Train Acc: 89.8300, Test loss: 0.0042. Test Acc: 89.6100. Time/epoch: 3.1495
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0020. Train Acc: 94.8225, Test loss: 0.0021. Test Acc: 94.7100. Time/epoch: 3.1173
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0017. Train Acc: 95.6725, Test loss: 0.0019. Test Acc: 95.5400. Time/epoch: 2.9793
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0017. Train Acc: 95.6575, Test loss: 0.0019. Test Acc: 95.4800. Time/epoch: 3.1003
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0022. Train Acc: 94.2775, Test loss: 0.0023. Test Acc: 94.3300. Time/epoch: 3.0977
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0015. Train Acc: 96.1125, Test loss: 0.0017. Test Acc: 95.9600. Time/epoch: 3.1133
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▆▆▆▆▆▆▇▆▇▇▆▇▇▆▇▇█▇██▇▇█▇██▇██████▇██
Accuracy/val▁▄▅▅▆▆▆▆▇▆▇▆▆▇▇▇▇▆▇▇████▇▇█▇█████████▇██
Loss/train█▅▄▄▃▃▃▃▂▃▂▃▂▂▂▂▂▃▂▂▁▁▁▁▂▂▁▂▁▁▁▁▁▁▁▁▁▂▁▁
Loss/val█▅▄▄▃▃▃▃▂▃▂▃▂▂▂▂▂▃▂▂▁▁▂▁▂▂▂▂▁▁▁▁▁▁▁▁▁▂▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.1125
Accuracy/val95.96
Loss/train0.00151
Loss/val0.00168
epoch50

View run noble-sweep-91 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9mzw55p6
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_025339-9mzw55p6/logs
wandb: Agent Starting Run: migt86gf with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_025627-migt86gf
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0241. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 2.9943
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.1021
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0975
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▂▄▁▃▂▂▅▂▄▁▃▆▁▁▂▂▁▂▃▁
Loss/val█▅▃▃▄▁▄▆▅▆▃▅▇▂▄▂▃▃▂▄▅
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02402
Loss/val0.02407
epoch20

View run pleasant-sweep-92 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/migt86gf
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_025627-migt86gf/logs
wandb: Agent Starting Run: 3zun05t5 with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_025747-3zun05t5
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0106. Train Acc: 87.3925, Test loss: 0.0108. Test Acc: 86.9600. Time/epoch: 4.6284
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0048. Train Acc: 93.9650, Test loss: 0.0050. Test Acc: 93.9400. Time/epoch: 4.7316
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0058. Train Acc: 92.3425, Test loss: 0.0060. Test Acc: 92.1000. Time/epoch: 4.7718
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▅▆▆▆▇▇▇▇▇▇▇▇▇▇████▆
Accuracy/val▁▃▅▆▆▆▇▇▇▇▇▇▇▇█▇████▆
Loss/train█▅▄▃▃▃▂▂▂▂▂▁▂▁▁▂▁▁▁▁▃
Loss/val█▅▄▃▃▃▂▂▂▂▂▁▂▁▁▂▁▁▁▁▃
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train92.3425
Accuracy/val92.1
Loss/train0.00576
Loss/val0.00596
epoch20

View run vague-sweep-93 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/3zun05t5
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_025747-3zun05t5/logs
wandb: Agent Starting Run: 6742a047 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_025939-6742a047
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3244
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4324
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▂▅▁▇▂▁▃▄▂▂
Loss/val█▃▇▁▅▆▅▄█▄▆
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01203
Loss/val0.01212
epoch10

View run sleek-sweep-94 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/6742a047
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_025939-6742a047/logs
wandb: Agent Starting Run: 7xfwvxrr with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030020-7xfwvxrr
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0151. Train Acc: 63.0500, Test loss: 0.0151. Test Acc: 63.6700. Time/epoch: 3.0093
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0048. Train Acc: 88.4175, Test loss: 0.0052. Test Acc: 88.3000. Time/epoch: 3.0877
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0036. Train Acc: 92.3075, Test loss: 0.0046. Test Acc: 90.6500. Time/epoch: 3.1176
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▂▁▃▅▅▅▆▆▇▇▇▇▇▇██▇█▇██
Accuracy/val▂▁▃▅▅▅▆▆▇▇▇▇▇▇████▇██
Loss/train█▇▆▃▃▃▃▃▂▂▂▂▂▂▁▁▂▁▂▁▁
Loss/val█▇▅▃▃▃▂▃▂▂▂▂▂▂▁▁▁▁▂▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train92.3075
Accuracy/val90.65
Loss/train0.00355
Loss/val0.00455
epoch20

View run bright-sweep-95 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/7xfwvxrr
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030020-7xfwvxrr/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: agga0njv with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030146-agga0njv
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0054. Train Acc: 75.3475, Test loss: 0.0054. Test Acc: 76.1300. Time/epoch: 2.3289
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0017. Train Acc: 91.4675, Test loss: 0.0017. Test Acc: 91.8000. Time/epoch: 2.4321
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0013. Train Acc: 92.9650, Test loss: 0.0014. Test Acc: 92.8700. Time/epoch: 2.4259
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▆▇▇▇▇▇▇▇█████████
Accuracy/val▁▄▆▆▆▆▇▇▇▇█▇█████████
Loss/train█▅▄▃▃▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train92.965
Accuracy/val92.87
Loss/train0.00135
Loss/val0.00138
epoch20

View run young-sweep-96 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/agga0njv
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030146-agga0njv/logs
wandb: Agent Starting Run: tjhkbl9x with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030252-tjhkbl9x
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0072. Train Acc: 89.4800, Test loss: 0.0072. Test Acc: 89.7000. Time/epoch: 5.2330
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0038. Train Acc: 95.0225, Test loss: 0.0043. Test Acc: 94.7500. Time/epoch: 5.1294
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▆▇▅█▇█▆
Accuracy/val▁▅▆▆▆▇▅█▇▇▆
Loss/train█▄▄▃▃▂▅▁▂▁▃
Loss/val█▄▄▃▃▂▅▁▂▁▃
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train95.0225
Accuracy/val94.75
Loss/train0.00385
Loss/val0.00431
epoch10

View run curious-sweep-97 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/tjhkbl9x
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030252-tjhkbl9x/logs
wandb: Agent Starting Run: simxgk4b with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030403-simxgk4b
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.9203
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0482. Train Acc: 36.8250, Test loss: 0.0481. Test Acc: 37.5400. Time/epoch: 4.8459
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.8674
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▁▁▄▄▄▃▄▇▄▄▆▃▂▅▂█▃▃▃▄▂
Loss/val▂▁▄▅▅▂▃█▃▂▇▂▃▅▃█▂▅▅▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04808
Loss/val0.04796
epoch20

View run treasured-sweep-98 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/simxgk4b
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030403-simxgk4b/logs
wandb: Agent Starting Run: bv8vyks7 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030600-bv8vyks7
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0106. Train Acc: 87.1825, Test loss: 0.0107. Test Acc: 87.0900. Time/epoch: 4.7402
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0047. Train Acc: 94.4275, Test loss: 0.0049. Test Acc: 94.4800. Time/epoch: 4.7193
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▇▇▇▇█▇█
Accuracy/val▁▅▆▆▇▇▇▇█▇█
Loss/train█▄▃▂▂▂▂▁▁▂▁
Loss/val█▄▃▃▂▂▂▁▁▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train94.4275
Accuracy/val94.48
Loss/train0.00474
Loss/val0.00494
epoch10

View run prime-sweep-99 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/bv8vyks7
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030600-bv8vyks7/logs
wandb: Agent Starting Run: uiatywb3 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030704-uiatywb3
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0031. Train Acc: 82.7000, Test loss: 0.0031. Test Acc: 82.7900. Time/epoch: 2.2303
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0011. Train Acc: 93.6525, Test loss: 0.0011. Test Acc: 93.5700. Time/epoch: 2.3759
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0006. Train Acc: 96.8750, Test loss: 0.0007. Test Acc: 96.6500. Time/epoch: 2.3505
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0007. Train Acc: 96.8975, Test loss: 0.0007. Test Acc: 96.6100. Time/epoch: 2.3367
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0014. Train Acc: 93.4575, Test loss: 0.0015. Test Acc: 93.1800. Time/epoch: 2.2386
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0008. Train Acc: 96.0250, Test loss: 0.0008. Test Acc: 95.7100. Time/epoch: 2.2163
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▄▆▅▅▆▇▆▇▇▇▇█▇▇▇▇▇█▄▇▇█▇██▅▅█▇▆█▇▆████▇
Accuracy/val▁▆▄▆▅▅▆▇▆▇▇▇▇██▇█▇▇█▄▇▇████▅▅█▇▇█▇▆████▇
Loss/train█▃▅▃▄▅▃▂▃▂▂▂▃▂▂▂▂▂▂▂▅▂▂▁▂▁▁▅▅▂▂▃▁▂▃▁▂▁▁▂
Loss/val█▃▅▃▄▄▃▂▃▂▂▂▂▁▁▂▁▂▂▁▅▂▂▁▂▁▁▅▄▁▂▃▁▂▃▁▂▁▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.025
Accuracy/val95.71
Loss/train0.00076
Loss/val0.00084
epoch50

View run clear-sweep-100 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/uiatywb3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030704-uiatywb3/logs
wandb: Agent Starting Run: zovc4bqy with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_030915-zovc4bqy
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0035. Train Acc: 89.5725, Test loss: 0.0036. Test Acc: 89.5100. Time/epoch: 3.3588
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0017. Train Acc: 95.7350, Test loss: 0.0019. Test Acc: 95.2800. Time/epoch: 3.1717
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 97.8425, Test loss: 0.0014. Test Acc: 96.8700. Time/epoch: 3.1696
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▅▆▁▆▇▇▆▇▇▆▇▇▅▇███
Accuracy/val▁▅▅▅▆▆▁▇▇█▆▇▇▆█▇▅▇███
Loss/train█▅▄▄▄▃█▃▂▂▃▃▃▃▂▂▇▂▂▁▁
Loss/val█▄▃▃▃▂█▂▁▁▃▂▃▃▂▂█▂▂▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train97.8425
Accuracy/val96.87
Loss/train0.00089
Loss/val0.00137
epoch20

View run proud-sweep-101 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/zovc4bqy
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_030915-zovc4bqy/logs
wandb: Agent Starting Run: z5r2tlkg with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_031041-z5r2tlkg
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0026. Train Acc: 89.0050, Test loss: 0.0026. Test Acc: 88.8600. Time/epoch: 2.3906
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0011. Train Acc: 94.3475, Test loss: 0.0012. Test Acc: 94.2700. Time/epoch: 2.2251
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 95.2050, Test loss: 0.0011. Test Acc: 94.9500. Time/epoch: 2.2297
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0009. Train Acc: 95.3600, Test loss: 0.0010. Test Acc: 95.2900. Time/epoch: 2.3617
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0009. Train Acc: 95.7775, Test loss: 0.0009. Test Acc: 95.5500. Time/epoch: 2.3671
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 95.9625, Test loss: 0.0009. Test Acc: 95.7900. Time/epoch: 2.2033
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▅▆▅▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇█▇██████████▇████
Accuracy/val▁▄▅▅▅▆▅▇▆▆▇▇▇▇▇▇▇▇▇▇▇▇█▇▇██▇▇█████▇█▇███
Loss/train█▅▄▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▄▃▂▃▂▂▂▂▂▂▂▂▂▂▁▂▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.9625
Accuracy/val95.79
Loss/train0.00088
Loss/val0.00092
epoch50

View run sweet-sweep-102 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/z5r2tlkg
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_031041-z5r2tlkg/logs
wandb: Agent Starting Run: 9ogv0d1d with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_031254-9ogv0d1d
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0013. Train Acc: 93.2350, Test loss: 0.0014. Test Acc: 93.1000. Time/epoch: 2.3610
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0009. Train Acc: 95.3750, Test loss: 0.0010. Test Acc: 95.3200. Time/epoch: 2.3356
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0006. Train Acc: 97.2500, Test loss: 0.0006. Test Acc: 97.0500. Time/epoch: 2.1837
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0006. Train Acc: 96.9925, Test loss: 0.0007. Test Acc: 96.6600. Time/epoch: 2.3134
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0006. Train Acc: 96.8000, Test loss: 0.0007. Test Acc: 96.4400. Time/epoch: 2.3188
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 97.5625, Test loss: 0.0006. Test Acc: 97.2300. Time/epoch: 2.3414
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▆▁▆▇▅▇▇▇▇▇▇▇▇▇▇▇█▇▇▇██▇██████▇██████████
Accuracy/val▆▁▆▇▅▇▇▇▇▇▇▇▇▇▇██▇▇▇█████████▇██████████
Loss/train▃█▂▂▃▂▂▂▂▂▂▂▂▁▁▁▁▂▂▂▁▁▁▁▁▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁
Loss/val▃█▂▂▃▂▂▂▂▂▂▂▁▁▁▁▁▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.5625
Accuracy/val97.23
Loss/train0.0005
Loss/val0.00062
epoch50

View run wobbly-sweep-103 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9ogv0d1d
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_031254-9ogv0d1d/logs
wandb: Agent Starting Run: vhvf0jcd with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_031505-vhvf0jcd
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0047. Train Acc: 77.7925, Test loss: 0.0046. Test Acc: 78.7000. Time/epoch: 2.4174
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3413
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train█▁▁▁▁▁▁▁▁▁▁
Accuracy/val█▁▁▁▁▁▁▁▁▁▁
Loss/train▁██████████
Loss/val▁██████████
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01203
Loss/val0.01214
epoch10

View run mild-sweep-104 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/vhvf0jcd
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_031505-vhvf0jcd/logs
wandb: Agent Starting Run: 5jzm4hd6 with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_031546-5jzm4hd6
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0195. Train Acc: 78.7525, Test loss: 0.0194. Test Acc: 79.0700. Time/epoch: 4.9323
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0050. Train Acc: 94.3575, Test loss: 0.0053. Test Acc: 93.9700. Time/epoch: 4.8345
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0047. Train Acc: 94.2750, Test loss: 0.0050. Test Acc: 93.8500. Time/epoch: 4.8189
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▄▅▆▇▇▇██████▁▇█▇█████
Accuracy/val▄▅▆▇▇▇██████▁▇█▇█████
Loss/train▄▃▂▂▁▁▁▁▁▁▁▁█▁▁▁▁▁▁▁▁
Loss/val▄▃▂▂▁▁▁▁▁▁▁▁█▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train94.275
Accuracy/val93.85
Loss/train0.0047
Loss/val0.005
epoch20

View run twilight-sweep-105 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/5jzm4hd6
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_031546-5jzm4hd6/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: x5wzmb9m with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_031748-x5wzmb9m
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0136. Train Acc: 81.9050, Test loss: 0.0138. Test Acc: 81.6700. Time/epoch: 4.7866
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0047. Train Acc: 93.9850, Test loss: 0.0049. Test Acc: 94.1600. Time/epoch: 4.7301
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0041. Train Acc: 94.7500, Test loss: 0.0042. Test Acc: 95.0900. Time/epoch: 4.7042
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▃▄▅▆▇▆▇▇███████▇███
Accuracy/val▁▂▃▄▅▆▇▆▇▇███████▇███
Loss/train█▆▆▄▄▃▂▂▂▂▁▁▁▁▁▁▁▂▁▁▁
Loss/val█▆▆▅▄▃▂▂▂▂▁▂▁▁▁▁▁▂▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train94.75
Accuracy/val95.09
Loss/train0.00407
Loss/val0.00423
epoch20

View run vocal-sweep-106 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/x5wzmb9m
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_031748-x5wzmb9m/logs
wandb: Agent Starting Run: jd4zr49s with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_031940-jd4zr49s
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0078. Train Acc: 89.4400, Test loss: 0.0081. Test Acc: 89.3800. Time/epoch: 5.2359
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0036. Train Acc: 95.1000, Test loss: 0.0038. Test Acc: 95.2400. Time/epoch: 5.1533
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0028. Train Acc: 96.6600, Test loss: 0.0032. Test Acc: 96.2400. Time/epoch: 5.1585
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▃▂▆▆▆▆▆▆▆▇▆▆▇▇█▇█▇█
Accuracy/val▁▂▄▂▆▆▆▇▇▇▇▇▆▇▇██▇█▇█
Loss/train█▆▅▇▃▃▃▂▂▂▂▂▃▂▂▁▁▂▁▂▁
Loss/val█▆▅▆▃▃▃▂▂▂▂▂▃▂▂▁▁▁▁▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train96.66
Accuracy/val96.24
Loss/train0.00278
Loss/val0.00316
epoch20

View run vocal-sweep-107 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/jd4zr49s
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_031940-jd4zr49s/logs
wandb: Agent Starting Run: p9k6cxx7 with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032143-p9k6cxx7
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0018. Train Acc: 88.2000, Test loss: 0.0018. Test Acc: 88.5900. Time/epoch: 2.3159
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0007. Train Acc: 96.4925, Test loss: 0.0008. Test Acc: 96.2200. Time/epoch: 2.4209
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 97.6750, Test loss: 0.0006. Test Acc: 97.1800. Time/epoch: 2.4273
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▇▆▆▇▇▇▇▇▇█▇▇██████
Accuracy/val▁▅▇▇▆▆▇▇▇▇▇▇█████████
Loss/train█▅▄▃▄▄▃▂▃▂▂▃▁▂▁▁▁▁▁▁▁
Loss/val█▄▃▃▄▄▂▂▃▂▂▂▁▂▁▁▁▁▁▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train97.675
Accuracy/val97.18
Loss/train0.00048
Loss/val0.00061
epoch20

View run worthy-sweep-108 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/p9k6cxx7
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032143-p9k6cxx7/logs
wandb: Agent Starting Run: 9zx8scos with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032244-9zx8scos
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0028. Train Acc: 92.2725, Test loss: 0.0029. Test Acc: 92.2900. Time/epoch: 3.3926
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0018. Train Acc: 95.2300, Test loss: 0.0019. Test Acc: 95.0900. Time/epoch: 3.3430
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0011. Train Acc: 97.3800, Test loss: 0.0013. Test Acc: 96.9100. Time/epoch: 3.3288
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0009. Train Acc: 97.5450, Test loss: 0.0013. Test Acc: 97.0000. Time/epoch: 3.1773
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0008. Train Acc: 98.1075, Test loss: 0.0012. Test Acc: 97.3100. Time/epoch: 3.2019
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 98.7525, Test loss: 0.0012. Test Acc: 97.6100. Time/epoch: 3.3130
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▄▅▄▅▄▄▆▆▆▆▆▆▆▆▇▇▇▆▇▇▇▆▇▇▇▇▇▇▇▇█▆█████
Accuracy/val▁▃▄▅▆▄▅▅▄▆▇▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇█▆█████
Loss/train█▆▆▅▄▅▅▅▅▃▃▄▃▃▃▃▃▃▂▂▃▂▂▂▂▂▂▂▂▂▂▂▂▁▃▁▁▁▁▁
Loss/val█▆▅▅▃▅▄▄▄▃▂▃▂▂▃▂▂▂▂▂▂▂▂▁▂▂▁▁▁▂▂▂▂▁▃▁▁▁▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.7525
Accuracy/val97.61
Loss/train0.00054
Loss/val0.0012
epoch50

View run winter-sweep-109 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9zx8scos
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032244-9zx8scos/logs
wandb: Agent Starting Run: zyau70g3 with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032547-zyau70g3
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0241. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0341
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0240. Test Acc: 37.5400. Time/epoch: 3.0451
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0240. Train Acc: 36.8250, Test loss: 0.0241. Test Acc: 37.5400. Time/epoch: 3.0080
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train████▁████████████████
Loss/val████▁████████████████
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.02402
Loss/val0.02406
epoch20

View run daily-sweep-110 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/zyau70g3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032547-zyau70g3/logs
wandb: Agent Starting Run: 2luvylas with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032703-2luvylas
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0042. Train Acc: 89.3825, Test loss: 0.0043. Test Acc: 89.0800. Time/epoch: 3.3587
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 95.4400, Test loss: 0.0020. Test Acc: 95.3200. Time/epoch: 3.1627
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▇▇▆▆▆▇█
Accuracy/val▁▄▅▆▇▇▆▇▆▇█
Loss/train█▅▄▃▂▂▃▂▂▁▁
Loss/val█▅▄▃▂▂▃▂▂▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train95.44
Accuracy/val95.32
Loss/train0.00189
Loss/val0.00202
epoch10

View run radiant-sweep-111 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/2luvylas
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032703-2luvylas/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: tvzq27e3 with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 1e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032800-tvzq27e3
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0061. Train Acc: 86.3300, Test loss: 0.0062. Test Acc: 86.3900. Time/epoch: 3.2002
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0026. Train Acc: 93.7600, Test loss: 0.0027. Test Acc: 93.9700. Time/epoch: 3.1527
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▅▆▆▆▇██
Accuracy/val▁▃▄▅▅▆▆▆▇▇█
Loss/train█▅▄▃▃▂▂▂▂▁▁
Loss/val█▅▄▃▃▂▂▂▂▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.76
Accuracy/val93.97
Loss/train0.00261
Loss/val0.00265
epoch10

View run hopeful-sweep-112 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/tvzq27e3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032800-tvzq27e3/logs
wandb: Agent Starting Run: p67cfog1 with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032846-p67cfog1
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0026. Train Acc: 92.8100, Test loss: 0.0027. Test Acc: 92.8000. Time/epoch: 3.3845
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 94.9050, Test loss: 0.0021. Test Acc: 94.7100. Time/epoch: 3.3136
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▅▅█▇▆▇█▄
Accuracy/val▁▅▆▆▅█▇▆▇█▄
Loss/train█▄▄▄▄▁▂▃▂▁▅
Loss/val█▃▃▄▃▁▁▃▂▁▅
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train94.905
Accuracy/val94.71
Loss/train0.00194
Loss/val0.00209
epoch10

View run fluent-sweep-113 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/p67cfog1
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032846-p67cfog1/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: dvziqyxq with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_032944-dvziqyxq
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0057. Train Acc: 92.3600, Test loss: 0.0058. Test Acc: 92.3700. Time/epoch: 4.8779
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0031. Train Acc: 95.9550, Test loss: 0.0034. Test Acc: 95.5800. Time/epoch: 4.8753
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0022. Train Acc: 97.1100, Test loss: 0.0026. Test Acc: 96.7300. Time/epoch: 4.8943
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0017. Train Acc: 98.1425, Test loss: 0.0021. Test Acc: 97.7400. Time/epoch: 4.7056
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0033. Train Acc: 96.1825, Test loss: 0.0037. Test Acc: 95.9100. Time/epoch: 4.6866
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0013. Train Acc: 98.5625, Test loss: 0.0017. Test Acc: 98.2800. Time/epoch: 4.8479
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▄▆▆▁▇▇▇▆▇▇▇▇▇▇▇▇▇▇▇▇█▇███▇██▇██▇█▇██████
Accuracy/val▅▆▆▁▇▇▇▆▆▇▇▇▇▇▇▇▇▇▇▇█▇███▇██▇██▇█▇██████
Loss/train▄▃▃█▂▂▂▃▂▂▂▂▂▂▂▁▂▂▂▂▁▂▁▁▁▂▁▁▂▁▁▁▁▂▁▁▁▁▁▁
Loss/val▄▃▃█▂▂▂▂▂▂▂▂▁▂▂▁▂▂▂▂▁▂▁▁▁▁▁▁▂▁▁▁▁▂▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.5625
Accuracy/val98.28
Loss/train0.00128
Loss/val0.00174
epoch50

View run visionary-sweep-114 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/dvziqyxq
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_032944-dvziqyxq/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: cb4b5suj with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_033411-cb4b5suj
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0020. Train Acc: 89.2950, Test loss: 0.0021. Test Acc: 89.4500. Time/epoch: 2.2336
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0017. Train Acc: 90.8825, Test loss: 0.0018. Test Acc: 90.8000. Time/epoch: 2.3476
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0006. Train Acc: 96.9800, Test loss: 0.0007. Test Acc: 96.8900. Time/epoch: 2.3485
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0005. Train Acc: 97.8550, Test loss: 0.0005. Test Acc: 97.7900. Time/epoch: 2.3385
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0004. Train Acc: 98.0175, Test loss: 0.0005. Test Acc: 97.7300. Time/epoch: 2.2109
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0006. Train Acc: 97.0625, Test loss: 0.0007. Test Acc: 96.8300. Time/epoch: 2.3406
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▆▆▄▂▄▆▇▇▇▆▇▇▇█▇▇█▇▇███▇▅█▇██▇▆██▇▇▇
Accuracy/val▁▃▄▅▆▆▆▄▂▄▇▇▇▇▆▇▇██▇▇█▇▇███▇▅█▇███▆██▇▇▇
Loss/train█▆▅▄▃▃▃▅▇▅▃▂▂▂▃▂▂▂▁▂▂▂▂▂▁▁▁▂▄▁▂▁▁▂▃▁▁▂▂▂
Loss/val█▆▅▄▃▃▃▅▇▅▂▂▂▂▂▂▂▁▁▂▂▁▁▂▁▁▁▂▄▁▂▁▁▁▃▁▁▂▂▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.0625
Accuracy/val96.83
Loss/train0.00061
Loss/val0.00069
epoch50

View run trim-sweep-115 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/cb4b5suj
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_033411-cb4b5suj/logs
wandb: Agent Starting Run: 9vuy6t60 with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_033622-9vuy6t60
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0019. Train Acc: 89.9600, Test loss: 0.0020. Test Acc: 89.9600. Time/epoch: 2.2468
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0006. Train Acc: 96.7925, Test loss: 0.0007. Test Acc: 96.6300. Time/epoch: 2.2197
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0006. Train Acc: 96.9175, Test loss: 0.0007. Test Acc: 96.5100. Time/epoch: 2.3593
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0004. Train Acc: 98.1800, Test loss: 0.0005. Test Acc: 97.5100. Time/epoch: 2.3331
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0004. Train Acc: 97.9050, Test loss: 0.0006. Test Acc: 97.0400. Time/epoch: 2.3245
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0002. Train Acc: 98.9500, Test loss: 0.0005. Test Acc: 97.9400. Time/epoch: 2.2019
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▃▄▅▁▆▆▃▇▇▇▆▆▆▅▇▇▇▇▇▇▇▇▇▇██▇▇▇▆▇▇▇█▇█████
Accuracy/val▃▄▅▁▇▆▃▇▇▇▇▆▇▅▇▇▇▇▇▇▇▇█▇██▇▇█▆▇▇▇█▇█████
Loss/train▆▄▄█▃▃▇▂▂▂▃▃▃▄▂▂▂▂▂▂▂▂▁▂▁▁▂▂▂▄▂▂▂▁▂▁▁▁▁▁
Loss/val▅▄▃█▂▃▆▂▂▂▂▂▂▃▂▂▂▂▂▂▂▂▁▁▁▁▁▂▁▄▂▂▂▁▂▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.95
Accuracy/val97.94
Loss/train0.00024
Loss/val0.0005
epoch50

View run ruby-sweep-116 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/9vuy6t60
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_033622-9vuy6t60/logs
wandb: Agent Starting Run: huppqm1m with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_033833-huppqm1m
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0040. Train Acc: 88.9275, Test loss: 0.0041. Test Acc: 88.8800. Time/epoch: 3.1435
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0019. Train Acc: 95.1400, Test loss: 0.0020. Test Acc: 95.1000. Time/epoch: 2.9824
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0016. Train Acc: 95.7625, Test loss: 0.0017. Test Acc: 95.8300. Time/epoch: 2.9552
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▆▄▆▇▄▇██▇▇█▇████▇█▇█
Accuracy/val▁▆▄▆▇▄███▇▇▇▇████▇█▇█
Loss/train█▃▅▃▂▆▁▁▁▂▂▁▁▁▁▁▁▂▁▂▁
Loss/val█▃▅▃▂▆▁▁▁▂▂▁▁▁▁▁▁▂▁▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train95.7625
Accuracy/val95.83
Loss/train0.00164
Loss/val0.00169
epoch20

View run absurd-sweep-117 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/huppqm1m
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_033833-huppqm1m/logs
wandb: Agent Starting Run: jpoemwhv with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.0003
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_033950-jpoemwhv
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0025. Train Acc: 85.6625, Test loss: 0.0026. Test Acc: 85.9400. Time/epoch: 2.3635
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0008. Train Acc: 95.6950, Test loss: 0.0009. Test Acc: 95.4900. Time/epoch: 2.1930
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▆▆▇▇▇█▇▇██
Accuracy/val▁▆▅▇▇▇█▇▇██
Loss/train█▃▃▂▂▂▁▂▂▁▁
Loss/val█▃▃▂▂▂▁▂▂▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train95.695
Accuracy/val95.49
Loss/train0.00085
Loss/val0.00091
epoch10

View run eternal-sweep-118 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/jpoemwhv
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_033950-jpoemwhv/logs
wandb: Agent Starting Run: m2mggbud with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034030-m2mggbud
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0044. Train Acc: 87.3125, Test loss: 0.0044. Test Acc: 87.4500. Time/epoch: 3.1746
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 95.3650, Test loss: 0.0019. Test Acc: 95.1000. Time/epoch: 3.1354
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▆▆▇▇▇████▇
Accuracy/val▁▆▇▇▇▇▇███▇
Loss/train█▄▃▂▂▂▂▁▁▁▂
Loss/val█▃▃▂▂▂▂▁▁▁▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train95.365
Accuracy/val95.1
Loss/train0.00186
Loss/val0.00194
epoch10

View run frosty-sweep-119 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/m2mggbud
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034030-m2mggbud/logs
wandb: Agent Starting Run: 39ddzhco with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034121-39ddzhco
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0085. Train Acc: 79.1825, Test loss: 0.0087. Test Acc: 78.7600. Time/epoch: 3.1127
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0029. Train Acc: 92.4450, Test loss: 0.0031. Test Acc: 92.1300. Time/epoch: 3.0869
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0025. Train Acc: 93.7350, Test loss: 0.0027. Test Acc: 93.3700. Time/epoch: 3.0732
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▅▆▆▆▇▇▇▇▇▇▇▇▇██████
Accuracy/val▁▃▅▆▆▆▇▇▇▇▇▇▇████████
Loss/train█▅▄▃▃▂▂▂▂▂▂▁▂▁▁▁▁▁▁▁▁
Loss/val█▅▄▃▃▂▂▂▂▂▂▁▂▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train93.735
Accuracy/val93.37
Loss/train0.0025
Loss/val0.00266
epoch20

View run restful-sweep-120 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/39ddzhco
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034121-39ddzhco/logs
wandb: Agent Starting Run: s9ih4ev0 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.001
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034237-s9ih4ev0
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0038. Train Acc: 91.0400, Test loss: 0.0039. Test Acc: 90.8100. Time/epoch: 3.3814
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0024. Train Acc: 93.5725, Test loss: 0.0026. Test Acc: 93.3100. Time/epoch: 3.2949
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0009. Train Acc: 98.0000, Test loss: 0.0013. Test Acc: 97.1200. Time/epoch: 3.2832
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0010. Train Acc: 97.7750, Test loss: 0.0016. Test Acc: 96.7800. Time/epoch: 3.1826
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0006. Train Acc: 98.6025, Test loss: 0.0017. Test Acc: 96.9600. Time/epoch: 3.2852
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0013. Train Acc: 96.8175, Test loss: 0.0028. Test Acc: 95.3000. Time/epoch: 3.3088
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▂▆▂▆▅▆▃▅▆▆▆▇▆▇▇▇▇▆▆▇▇▇▇▇▇▇█▇▇██▇██▇██▆
Accuracy/val▁▆▃▇▂▇▆▇▄▆▇▆▆█▇▇█▇▇▆▆██▇█▇▇▇██▇█████▇██▆
Loss/train█▄▆▃█▃▄▃▅▄▃▃▃▂▃▂▂▂▂▃▃▂▂▃▂▂▂▂▁▂▂▁▁▁▁▁▂▁▁▃
Loss/val▇▂▅▂█▂▃▂▄▃▂▂▂▂▃▂▁▂▂▃▂▁▁▄▂▂▂▃▂▃▂▁▂▂▃▂▃▂▂▄
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.8175
Accuracy/val95.3
Loss/train0.0013
Loss/val0.00278
epoch50

View run major-sweep-121 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/s9ih4ev0
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034237-s9ih4ev0/logs
wandb: Agent Starting Run: mfega71i with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034540-mfega71i
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0121. Train Acc: 37.4350, Test loss: 0.0122. Test Acc: 38.0400. Time/epoch: 2.2479
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0122. Train Acc: 36.8250, Test loss: 0.0123. Test Acc: 37.5400. Time/epoch: 2.1978
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0122. Train Acc: 36.8250, Test loss: 0.0122. Test Acc: 37.5400. Time/epoch: 2.3595
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▆▅█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▅▅█▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▁▃▁▆▅▆▆▇▇██▆▆▆▆▇▇█▆█▇
Loss/val▁▃▁▆▆▇▅▆▅▇▇▆▅▅▆▇▆█▅█▇
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01216
Loss/val0.01224
epoch20

View run autumn-sweep-122 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/mfega71i
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034540-mfega71i/logs
wandb: Agent Starting Run: sagxbth3 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034642-sagxbth3
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.2179
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.3562
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train▄▁▄█▄▂▃▄▂▃▂
Loss/val▄▃▅█▆▃▃▆▄▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01203
Loss/val0.01209
epoch10

View run eager-sweep-123 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/sagxbth3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034642-sagxbth3/logs
wandb: Agent Starting Run: eko31h9c with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034723-eko31h9c
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 5.0831
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.9927
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 5.1511
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▂▂▄▄▄▁▅▁▁▂▃▂▂▁█▃▄▁▂▂▂
Loss/val▂▂▄▃▃▁▄▁▂▃▂▂▂▁█▃▅▂▂▃▃
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04805
Loss/val0.04797
epoch20

View run stellar-sweep-124 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/eko31h9c
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034723-eko31h9c/logs
wandb: Agent Starting Run: rc4w0al7 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_034925-rc4w0al7
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0059. Train Acc: 91.7375, Test loss: 0.0061. Test Acc: 91.7000. Time/epoch: 5.0144
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0025. Train Acc: 96.8450, Test loss: 0.0028. Test Acc: 96.5600. Time/epoch: 4.9937
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0015. Train Acc: 98.2525, Test loss: 0.0022. Test Acc: 97.6300. Time/epoch: 5.1812
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0027. Train Acc: 96.6925, Test loss: 0.0039. Test Acc: 96.0000. Time/epoch: 5.1661
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0015. Train Acc: 98.0425, Test loss: 0.0034. Test Acc: 96.6400. Time/epoch: 5.1761
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0009. Train Acc: 98.9125, Test loss: 0.0035. Test Acc: 97.4000. Time/epoch: 5.0519
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▃▅▂▄▅▅▆▆▇▅▆▆▆▆▇▇▅▅▇▆▆▇▆▇▇▇█▆▇▇███▇████
Accuracy/val▁▄▃▆▂▅▆▆▆▇▇▅▇▇▇▆█▇▆▅▇▆▆█▆████▆█▇███▇▇██▇
Loss/train█▆▆▅▇▅▄▄▃▃▃▄▃▃▃▄▂▂▄▄▂▃▄▂▄▂▂▂▂▃▁▂▁▁▁▂▁▁▁▁
Loss/val█▅▅▃▇▄▃▃▂▂▁▃▂▂▂▃▁▁▄▄▂▃▄▁▄▂▂▁▁▅▁▂▂▂▂▃▃▁▂▃
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.9125
Accuracy/val97.4
Loss/train0.00089
Loss/val0.00354
epoch50

View run floral-sweep-125 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/rc4w0al7
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_034925-rc4w0al7/logs
wandb: Agent Starting Run: 3hue7myw with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.0001
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_035401-3hue7myw
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0017. Train Acc: 90.8700, Test loss: 0.0018. Test Acc: 90.8500. Time/epoch: 2.4444
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0007. Train Acc: 96.5750, Test loss: 0.0008. Test Acc: 96.2600. Time/epoch: 2.4537
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0006. Train Acc: 97.2350, Test loss: 0.0007. Test Acc: 96.7600. Time/epoch: 2.3120
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▅▆▆▇▆▆▆▇▇▇▇█▇████▇█
Accuracy/val▁▅▅▆▆▇▆▆▇▇▇▇██▇████▇█
Loss/train█▅▄▄▃▂▃▃▃▂▂▂▂▁▂▁▁▁▁▁▁
Loss/val█▅▄▄▃▂▃▃▃▂▂▂▁▁▂▁▁▁▁▂▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train97.235
Accuracy/val96.76
Loss/train0.00056
Loss/val0.00067
epoch20

View run rare-sweep-126 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/3hue7myw
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_035401-3hue7myw/logs
wandb: Agent Starting Run: g2qv1hr9 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.01
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_035508-g2qv1hr9
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0121. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4737
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0120. Train Acc: 36.8250, Test loss: 0.0121. Test Acc: 37.5400. Time/epoch: 2.4183
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▄▅▂▁▃▂▂▅▁▁
Loss/val█▃▆▃▃▂▆▁▃▃▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.01203
Loss/val0.0121
epoch10

View run drawn-sweep-127 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/g2qv1hr9
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_035508-g2qv1hr9/logs
wandb: Agent Starting Run: 14i7qqb3 with config:
wandb:      batch_size: 32
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_035549-14i7qqb3
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0076. Train Acc: 89.9925, Test loss: 0.0079. Test Acc: 89.6700. Time/epoch: 4.7964
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0032. Train Acc: 96.0875, Test loss: 0.0036. Test Acc: 95.7100. Time/epoch: 4.7119
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▇▇▇▆█▇▇▇█
Accuracy/val▁▅▇▇█▆█▇▇▇█
Loss/train█▄▂▂▂▃▁▃▂▂▁
Loss/val█▄▂▂▂▃▁▂▂▂▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train96.0875
Accuracy/val95.71
Loss/train0.00317
Loss/val0.00357
epoch10

View run wild-sweep-128 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/14i7qqb3
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_035549-14i7qqb3/logs
wandb: Agent Starting Run: ikirp9xn with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_035655-ikirp9xn
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0043. Train Acc: 94.4825, Test loss: 0.0045. Test Acc: 94.4700. Time/epoch: 4.9078
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0020. Train Acc: 97.6225, Test loss: 0.0024. Test Acc: 97.2900. Time/epoch: 4.8971
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0015. Train Acc: 98.1250, Test loss: 0.0021. Test Acc: 97.5600. Time/epoch: 4.8745
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▂▃▂▁▄▄▄▆▆▇▇▇▇▆██▇████
Accuracy/val▃▃▂▁▅▅▃▆▇▇▇▇▇▆██▇███▇
Loss/train▇▆▇█▅▅▆▃▃▃▂▂▂▃▁▁▁▁▁▁▁
Loss/val▆▆▆█▄▄▆▃▂▂▂▂▂▃▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train98.125
Accuracy/val97.56
Loss/train0.00153
Loss/val0.00214
epoch20

View run ancient-sweep-129 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ikirp9xn
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_035655-ikirp9xn/logs
wandb: Agent Starting Run: xbpuwg2m with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0001
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_035852-xbpuwg2m
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0029. Train Acc: 92.8025, Test loss: 0.0030. Test Acc: 92.4700. Time/epoch: 3.3691
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0013. Train Acc: 96.9200, Test loss: 0.0014. Test Acc: 96.7400. Time/epoch: 3.3041
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0011. Train Acc: 97.2725, Test loss: 0.0013. Test Acc: 96.8900. Time/epoch: 3.2984
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0007. Train Acc: 98.3425, Test loss: 0.0010. Test Acc: 97.8500. Time/epoch: 3.1525
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0005. Train Acc: 98.6950, Test loss: 0.0009. Test Acc: 98.1000. Time/epoch: 3.3082
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0005. Train Acc: 98.6475, Test loss: 0.0010. Test Acc: 97.9400. Time/epoch: 3.2932
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▂▃▅▄▆▅▆▆▆▆▆▆▆▇▆▆▆▆▇▆▇▅▇▇▇█▇▇█▇██▇█▇██▇
Accuracy/val▁▄▃▃▅▅▆▅▆▆▇▇▇▇▇▇▆▇▇▆▇▇▇▆▇█▇█▇▇█▇██▇█▇██▇
Loss/train█▅▆▆▄▄▄▄▃▃▃▃▃▃▃▂▃▂▃▃▂▃▂▄▂▂▂▁▂▂▁▂▁▁▂▁▂▁▁▁
Loss/val█▅▆▆▃▄▃▃▃▂▂▂▂▂▂▂▂▂▂▃▂▂▁▃▁▁▂▁▂▂▁▂▁▁▂▁▂▁▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.6475
Accuracy/val97.94
Loss/train0.00053
Loss/val0.00104
epoch50

View run peach-sweep-130 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xbpuwg2m
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_035852-xbpuwg2m/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: bms9a5ow with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_040201-bms9a5ow
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0086. Train Acc: 83.8575, Test loss: 0.0087. Test Acc: 83.9500. Time/epoch: 2.9495
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0029. Train Acc: 92.1350, Test loss: 0.0031. Test Acc: 91.9500. Time/epoch: 3.0603
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0026. Train Acc: 93.2900, Test loss: 0.0027. Test Acc: 93.1500. Time/epoch: 3.0438
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▅▅▆▆▇▇▇▇▇▇▇████████
Accuracy/val▁▃▅▅▆▆▇▇▇▇▇▇▇████████
Loss/train█▅▃▃▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▅▃▃▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train93.29
Accuracy/val93.15
Loss/train0.00255
Loss/val0.00269
epoch20

View run effortless-sweep-131 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/bms9a5ow
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_040201-bms9a5ow/logs
wandb: Agent Starting Run: ihr6pabg with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_040317-ihr6pabg
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0016. Train Acc: 90.3525, Test loss: 0.0017. Test Acc: 90.4900. Time/epoch: 2.2637
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0007. Train Acc: 96.8375, Test loss: 0.0007. Test Acc: 96.4900. Time/epoch: 2.3961
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0007. Train Acc: 96.5400, Test loss: 0.0008. Test Acc: 95.9700. Time/epoch: 2.4040
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0004. Train Acc: 98.0225, Test loss: 0.0006. Test Acc: 97.0300. Time/epoch: 2.3816
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0004. Train Acc: 97.9450, Test loss: 0.0007. Test Acc: 96.8300. Time/epoch: 2.2781
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0003. Train Acc: 98.3350, Test loss: 0.0007. Test Acc: 97.1100. Time/epoch: 2.2493
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▃▅▅▄▅▆▆▆▅▇▆▇▆▆▆▆▇▄▇▇▇▇▇▇▇▇█▇██████▇▇██
Accuracy/val▁▄▃▅▅▅▆▆▇▇▅▇▇▇▇▇▆▇▇▃▇▇▇▇▇█▇██▇██████▇▇██
Loss/train█▆▆▅▄▅▄▄▃▃▄▃▃▂▃▃▃▃▂▆▂▂▃▂▂▂▂▂▁▂▁▂▁▁▁▁▂▂▁▁
Loss/val█▅▆▄▄▅▃▃▂▂▄▂▂▁▂▃▂▂▂▇▁▁▂▂▁▁▃▂▁▂▁▁▁▁▁▁▂▂▁▂
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.335
Accuracy/val97.11
Loss/train0.00035
Loss/val0.00074
epoch50

View run glowing-sweep-132 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ihr6pabg
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_040317-ihr6pabg/logs
wandb: Agent Starting Run: 4ecbpj83 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_040529-4ecbpj83
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0304. Train Acc: 83.0275, Test loss: 0.0309. Test Acc: 83.3500. Time/epoch: 4.7452
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0061. Train Acc: 92.8800, Test loss: 0.0062. Test Acc: 93.0900. Time/epoch: 4.5664
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0056. Train Acc: 93.4800, Test loss: 0.0057. Test Acc: 93.3600. Time/epoch: 4.5281
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0054. Train Acc: 93.1425, Test loss: 0.0054. Test Acc: 93.2300. Time/epoch: 4.7129
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0052. Train Acc: 93.7025, Test loss: 0.0053. Test Acc: 93.6900. Time/epoch: 4.6930
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0052. Train Acc: 93.7325, Test loss: 0.0052. Test Acc: 93.8400. Time/epoch: 4.7197
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▅▆▆▆▇▇▇▇▆▇▇▇▇██████▇▇▇▇█▇▇██████▇████
Accuracy/val▁▄▅▅▆▆▆▇▇▇▇▇▇▇▇▇███████▇▇▇██▇███████████
Loss/train█▇▆▆▅▅▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▇▆▆▅▅▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train93.7325
Accuracy/val93.84
Loss/train0.00516
Loss/val0.00522
epoch50

View run tough-sweep-133 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/4ecbpj83
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_040529-4ecbpj83/logs
wandb: Agent Starting Run: ddvk082j with config:
wandb:      batch_size: 128
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_040938-ddvk082j
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0033. Train Acc: 84.9600, Test loss: 0.0034. Test Acc: 85.0000. Time/epoch: 2.1726
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0012. Train Acc: 93.7450, Test loss: 0.0012. Test Acc: 93.8900. Time/epoch: 2.2998
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0010. Train Acc: 94.7875, Test loss: 0.0011. Test Acc: 95.0600. Time/epoch: 2.2911
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0010. Train Acc: 95.1850, Test loss: 0.0010. Test Acc: 95.1300. Time/epoch: 2.1656
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0009. Train Acc: 95.5050, Test loss: 0.0010. Test Acc: 95.6000. Time/epoch: 2.1626
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0008. Train Acc: 95.7200, Test loss: 0.0009. Test Acc: 95.7500. Time/epoch: 2.1521
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▃▅▆▆▆▆▇▇▇▇▇▆▇▇▇▇▇█▇█▇██▇▇▇██▇█▆███████
Accuracy/val▁▃▃▅▆▆▇▆▇▇▇▇▇▆▇▇█▇▇▇▇▇▇██▇▇▇█▇██▆███████
Loss/train█▆▅▄▃▃▂▃▂▂▂▂▂▂▂▂▁▁▁▁▂▁▁▁▁▂▁▂▁▁▁▁▂▁▁▁▁▁▁▁
Loss/val█▆▅▄▃▃▂▂▂▂▂▂▂▂▂▂▁▁▁▁▂▁▁▁▁▁▁▂▁▁▁▁▂▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.72
Accuracy/val95.75
Loss/train0.00085
Loss/val0.00092
epoch50

View run dulcet-sweep-134 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ddvk082j
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_040938-ddvk082j/logs
wandb: Agent Starting Run: ccaoqtld with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_041146-ccaoqtld
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0053. Train Acc: 63.4450, Test loss: 0.0053. Test Acc: 64.0900. Time/epoch: 2.3517
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0026. Train Acc: 86.4825, Test loss: 0.0026. Test Acc: 86.9600. Time/epoch: 2.1717
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▄▃▅▅▇▄▁█▇
Accuracy/val▁▁▄▃▅▅▇▄▁█▇
Loss/train▄▅▃▄▂▂▂▂█▁▁
Loss/val▄▅▃▄▂▂▂▂█▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train86.4825
Accuracy/val86.96
Loss/train0.00261
Loss/val0.00265
epoch10

View run good-sweep-135 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ccaoqtld
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_041146-ccaoqtld/logs
wandb: Agent Starting Run: tv6lmibh with config:
wandb:      batch_size: 32
wandb:      epochs: 20
wandb:      learning_rate: 0.01
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_041227-tv6lmibh
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0482. Train Acc: 36.8250, Test loss: 0.0481. Test Acc: 37.5400. Time/epoch: 4.8126
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.8172
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7995
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train█▃▃▂▁▁▁▃▁▂▁▂▁▃▂▃▁▃▃▂▁
Loss/val█▄▂▂▁▂▁▂▁▁▂▂▂▃▁▂▁▄▃▁▂
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04804
Loss/val0.04795
epoch20

View run radiant-sweep-136 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/tv6lmibh
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_041227-tv6lmibh/logs
wandb: Agent Starting Run: w39caajm with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 3e-05
wandb:      optimizer: adam
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_041419-w39caajm
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0036. Train Acc: 91.1700, Test loss: 0.0037. Test Acc: 91.2600. Time/epoch: 3.3021
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0017. Train Acc: 95.7500, Test loss: 0.0018. Test Acc: 95.4500. Time/epoch: 3.2997
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0014. Train Acc: 96.6900, Test loss: 0.0015. Test Acc: 96.2900. Time/epoch: 3.1159
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0013. Train Acc: 96.6850, Test loss: 0.0015. Test Acc: 96.2900. Time/epoch: 3.3020
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0010. Train Acc: 97.5825, Test loss: 0.0012. Test Acc: 97.1500. Time/epoch: 3.2864
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0010. Train Acc: 97.5925, Test loss: 0.0012. Test Acc: 97.1400. Time/epoch: 3.2681
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▃▄▅▅▅▅▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇█▇█▇▇▇███████
Accuracy/val▁▃▃▄▅▅▅▆▆▆▆▆▆▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇█▇▇▇█▇██▇██
Loss/train█▆▅▅▄▄▄▃▃▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▂▂▁▁▁▁▁▁▁
Loss/val█▆▅▅▄▄▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▁▂▁▂▂▂▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train97.5925
Accuracy/val97.14
Loss/train0.00099
Loss/val0.0012
epoch50

View run clear-sweep-137 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/w39caajm
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_041419-w39caajm/logs
wandb: Agent Starting Run: iprutlfn with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: adam
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_041716-iprutlfn
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0020. Train Acc: 90.0725, Test loss: 0.0021. Test Acc: 90.0000. Time/epoch: 2.4462
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0010. Train Acc: 94.9400, Test loss: 0.0010. Test Acc: 94.8700. Time/epoch: 2.3934
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▂▄▅▇███▇█▇
Accuracy/val▁▂▄▆▇▇██▇█▇
Loss/train█▆▅▃▂▂▂▁▂▁▂
Loss/val█▆▅▃▂▂▂▁▂▁▂
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train94.94
Accuracy/val94.87
Loss/train0.00096
Loss/val0.00104
epoch10

View run icy-sweep-138 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/iprutlfn
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_041716-iprutlfn/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: rmk7xwg2 with config:
wandb:      batch_size: 128
wandb:      epochs: 20
wandb:      learning_rate: 0.001
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_041803-rmk7xwg2
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0036. Train Acc: 77.4625, Test loss: 0.0036. Test Acc: 77.7900. Time/epoch: 2.3357
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0038. Train Acc: 78.8350, Test loss: 0.0038. Test Acc: 79.0700. Time/epoch: 2.3126
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0032. Train Acc: 81.2675, Test loss: 0.0032. Test Acc: 81.9000. Time/epoch: 2.3034
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▅▁▆▇▇██▅▅▅▅▆▆▆▆▆▆▆▆▅▆
Accuracy/val▅▁▆▇███▅▅▅▅▅▆▆▆▅▆▆▆▅▆
Loss/train▂█▄▃▂▂▁▃▃▃▂▂▂▂▂▂▂▂▂▂▂
Loss/val▂█▄▃▂▂▁▄▃▃▂▂▂▂▂▂▂▂▂▂▂
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train81.2675
Accuracy/val81.9
Loss/train0.00318
Loss/val0.00321
epoch20

View run mild-sweep-139 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/rmk7xwg2
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_041803-rmk7xwg2/logs
wandb: Agent Starting Run: 1wevhc82 with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 0.0003
wandb:      optimizer: adam
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_041904-1wevhc82
optimizer: adam
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0024. Train Acc: 93.7875, Test loss: 0.0025. Test Acc: 93.8200. Time/epoch: 3.3504
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0012. Train Acc: 96.7600, Test loss: 0.0015. Test Acc: 96.1400. Time/epoch: 3.2640
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0007. Train Acc: 98.2800, Test loss: 0.0011. Test Acc: 97.5600. Time/epoch: 3.3123
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0007. Train Acc: 98.5250, Test loss: 0.0013. Test Acc: 97.4900. Time/epoch: 3.3069
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0005. Train Acc: 98.8725, Test loss: 0.0014. Test Acc: 97.7500. Time/epoch: 3.1509
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0006. Train Acc: 98.5400, Test loss: 0.0014. Test Acc: 97.1700. Time/epoch: 3.1518
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▃▄▃▅▅▅▅▅▅▆▄▆▄▅▇▆▇▆▇▇▇▇▇▇▇▆█▅█▇▇██▇▇██▇
Accuracy/val▁▁▄▄▄▅▅▅▅▅▆▆▄▇▄▅▇▇▇▆█▇▇▇▇▇█▇█▅█▇▆▇█▇▇██▇
Loss/train██▆▅▆▄▄▄▄▄▄▃▅▃▅▅▂▃▂▃▂▂▂▂▂▂▂▃▁▅▁▂▂▁▁▂▂▁▁▂
Loss/val██▅▄▅▃▃▃▃▃▃▂▅▁▄▄▁▂▁▃▁▂▂▂▂▂▁▃▁▇▂▂▃▂▂▃▃▂▁▃
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train98.54
Accuracy/val97.17
Loss/train0.00055
Loss/val0.00143
epoch50

View run misunderstood-sweep-140 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/1wevhc82
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_041904-1wevhc82/logs
wandb: Agent Starting Run: xhucx4ro with config:
wandb:      batch_size: 64
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_042204-xhucx4ro
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0141. Train Acc: 68.0175, Test loss: 0.0140. Test Acc: 68.3800. Time/epoch: 3.1224
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0032. Train Acc: 91.7075, Test loss: 0.0033. Test Acc: 92.0200. Time/epoch: 2.9470
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0027. Train Acc: 93.5925, Test loss: 0.0029. Test Acc: 93.3000. Time/epoch: 2.9227
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0024. Train Acc: 94.1850, Test loss: 0.0025. Test Acc: 93.8900. Time/epoch: 2.9237
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0022. Train Acc: 94.2875, Test loss: 0.0023. Test Acc: 94.2000. Time/epoch: 3.0559
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0019. Train Acc: 95.2075, Test loss: 0.0020. Test Acc: 95.2100. Time/epoch: 3.0550
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▅▆▆▇▇▇▇▇▇▇▇▇▇█▇████████████████████████
Accuracy/val▁▅▆▆▇▇▇▇▇▇▇▇▇▇███▇██████████████████████
Loss/train█▄▃▂▂▂▂▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▄▃▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train95.2075
Accuracy/val95.21
Loss/train0.00193
Loss/val0.00204
epoch50

View run sweet-sweep-141 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/xhucx4ro
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_042204-xhucx4ro/logs
wandb: Agent Starting Run: iv54em4c with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 0.01
wandb:      optimizer: sgd
wandb:      weight_decay: 0.005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_042453-iv54em4c
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0480. Train Acc: 36.8250, Test loss: 0.0479. Test Acc: 37.5400. Time/epoch: 4.5864
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0479. Test Acc: 37.5400. Time/epoch: 4.5563
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.6828
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7069
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0481. Train Acc: 36.8250, Test loss: 0.0480. Test Acc: 37.5400. Time/epoch: 4.7404
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0482. Train Acc: 36.8250, Test loss: 0.0481. Test Acc: 37.5400. Time/epoch: 4.6482
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Accuracy/val▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/train▁▃▂▂▃▃▂▂▂▂▂▂▄▃▅▃▅▇▁▁█▅█▂▃▃▅▁▄▂▄▃▂▂▁▄▅▁▃▇
Loss/val▁▃▂▄▃▃▃▃▁▂▁▂▅▂▅▂▄▄▁▂█▄█▂▃▃▅▂▄▂▄▅▁▂▂▃▃▁▃▇
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train36.825
Accuracy/val37.54
Loss/train0.04817
Loss/val0.04809
epoch50

View run trim-sweep-142 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/iv54em4c
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_042453-iv54em4c/logs
wandb: Agent Starting Run: ma6pe9c9 with config:
wandb:      batch_size: 32
wandb:      epochs: 50
wandb:      learning_rate: 1e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_042901-ma6pe9c9
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0153. Train Acc: 80.8250, Test loss: 0.0152. Test Acc: 80.6200. Time/epoch: 4.7573
 EPOCH 10. Progress: 20.0%.
 Train loss: 0.0045. Train Acc: 94.4025, Test loss: 0.0046. Test Acc: 94.7300. Time/epoch: 4.7037
 EPOCH 20. Progress: 40.0%.
 Train loss: 0.0041. Train Acc: 94.7325, Test loss: 0.0043. Test Acc: 94.8600. Time/epoch: 4.6952
 EPOCH 30. Progress: 60.0%.
 Train loss: 0.0035. Train Acc: 95.7775, Test loss: 0.0037. Test Acc: 95.6700. Time/epoch: 4.6677
 EPOCH 40. Progress: 80.0%.
 Train loss: 0.0035. Train Acc: 95.6200, Test loss: 0.0036. Test Acc: 95.5900. Time/epoch: 4.5579
 EPOCH 50. Progress: 100.0%.
 Train loss: 0.0030. Train Acc: 96.2000, Test loss: 0.0033. Test Acc: 96.2100. Time/epoch: 4.5332
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▆▇▇▇▇▇▇▇▇▇▇▇█▇█▇███████████████████
Accuracy/val▁▃▄▅▆▆▇▇▇▇▇▇▇▇▇▇▇█▇█████████████████████
Loss/train█▆▅▄▃▃▂▂▂▂▂▂▂▂▂▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
Loss/val█▆▅▄▃▃▂▂▂▂▂▂▂▂▂▂▂▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
epoch▁▁▁▁▂▂▂▂▂▃▃▃▃▃▃▄▄▄▄▄▅▅▅▅▅▅▆▆▆▆▆▆▇▇▇▇▇███

Run summary:


Accuracy/train96.2
Accuracy/val96.21
Loss/train0.00305
Loss/val0.00328
epoch50

View run flowing-sweep-143 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/ma6pe9c9
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_042901-ma6pe9c9/logs
wandb: Agent Starting Run: 23nqvcc2 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.0005
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_043309-23nqvcc2
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0048. Train Acc: 73.7925, Test loss: 0.0049. Test Acc: 74.1700. Time/epoch: 2.3877
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0081. Train Acc: 57.1475, Test loss: 0.0081. Test Acc: 57.7900. Time/epoch: 2.2014
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▅▅▆▆▆▆▇███▁
Accuracy/val▅▅▆▆▆▆▇███▁
Loss/train▄▄▃▂▂▂▂▁▁▁█
Loss/val▄▄▃▂▂▃▂▁▁▁█
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train57.1475
Accuracy/val57.79
Loss/train0.00814
Loss/val0.00813
epoch10

View run resilient-sweep-144 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/23nqvcc2
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_043309-23nqvcc2/logs
wandb: Agent Starting Run: 8xl53zs7 with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 1e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_043349-8xl53zs7
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0059. Train Acc: 84.4050, Test loss: 0.0059. Test Acc: 84.3900. Time/epoch: 3.1628
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0026. Train Acc: 93.5075, Test loss: 0.0027. Test Acc: 93.4200. Time/epoch: 3.1191
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0023. Train Acc: 94.1400, Test loss: 0.0023. Test Acc: 94.0900. Time/epoch: 3.1040
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▃▄▅▆▇▇▇▇▇██▇█▇██████
Accuracy/val▁▃▄▅▆▆▇▇▇▇▇▇▇█▇██████
Loss/train█▅▄▄▃▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁
Loss/val█▅▄▄▃▂▂▂▂▂▂▂▂▁▂▁▁▁▁▁▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train94.14
Accuracy/val94.09
Loss/train0.00227
Loss/val0.00234
epoch20

View run breezy-sweep-145 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/8xl53zs7
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_043349-8xl53zs7/logs
wandb: Agent Starting Run: fzw96h8a with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: sgd
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_043506-fzw96h8a
optimizer: sgd
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0053. Train Acc: 85.2075, Test loss: 0.0055. Test Acc: 84.7800. Time/epoch: 3.1110
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0026. Train Acc: 93.2875, Test loss: 0.0027. Test Acc: 93.2000. Time/epoch: 3.0699
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▅▆▇▇▇████
Accuracy/val▁▄▆▆▇▇▇████
Loss/train█▅▃▃▂▂▂▁▁▁▁
Loss/val█▅▃▃▂▂▂▁▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.2875
Accuracy/val93.2
Loss/train0.00258
Loss/val0.00269
epoch10

View run wandering-sweep-146 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/fzw96h8a
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_043506-fzw96h8a/logs
wandb: Agent Starting Run: zyqvr0bs with config:
wandb:      batch_size: 64
wandb:      epochs: 20
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_043552-zyqvr0bs
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0035. Train Acc: 89.7875, Test loss: 0.0036. Test Acc: 89.6900. Time/epoch: 3.1379
 EPOCH 10. Progress: 50.0%.
 Train loss: 0.0024. Train Acc: 93.0025, Test loss: 0.0025. Test Acc: 92.8200. Time/epoch: 3.1572
 EPOCH 20. Progress: 100.0%.
 Train loss: 0.0018. Train Acc: 95.6375, Test loss: 0.0019. Test Acc: 95.6900. Time/epoch: 3.1587
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▁▄▃▅▄▆▅▇▅▆▅▆▅▆█▇▅▇▇▆█
Accuracy/val▁▄▃▅▄▆▅▆▅▅▅▆▅▆▇▇▅▇▇▆█
Loss/train█▅▆▄▄▃▃▂▄▃▃▂▄▃▁▁▄▁▁▃▁
Loss/val█▅▆▄▄▃▃▂▄▃▄▂▄▃▁▁▅▁▁▃▁
epoch▁▁▂▂▂▃▃▃▄▄▅▅▅▆▆▆▇▇▇██

Run summary:


Accuracy/train95.6375
Accuracy/val95.69
Loss/train0.00184
Loss/val0.00193
epoch20

View run revived-sweep-147 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/zyqvr0bs
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_043552-zyqvr0bs/logs
wandb: Agent Starting Run: dkbivil6 with config:
wandb:      batch_size: 128
wandb:      epochs: 10
wandb:      learning_rate: 0.0001
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05
Tracking run with wandb version 0.15.3
Run data is saved locally in /vast/palmer/home.mccleary/ahf38/Documents/geo_comp_offline/2022/wandb/run-20230518_043717-dkbivil6
optimizer: rmsprop
 EPOCH 0. Progress: 0.0%.
 Train loss: 0.0016. Train Acc: 92.0450, Test loss: 0.0016. Test Acc: 92.0200. Time/epoch: 2.3793
 EPOCH 10. Progress: 100.0%.
 Train loss: 0.0011. Train Acc: 93.6850, Test loss: 0.0012. Test Acc: 93.6100. Time/epoch: 2.3126
Waiting for W&B process to finish... (success).

Run history:


Accuracy/train▄▁▅▆▇▇▆▆██▇
Accuracy/val▄▁▆▆▇▇▆▆█▇▇
Loss/train▇█▄▄▃▃▄▃▁▁▁
Loss/val▇█▄▅▃▂▄▃▁▁▁
epoch▁▂▂▃▄▅▅▆▇▇█

Run summary:


Accuracy/train93.685
Accuracy/val93.61
Loss/train0.00114
Loss/val0.00117
epoch10

View run stilted-sweep-148 at: https://wandb.ai/ahof1704/CNN_Sat_sweep/runs/dkbivil6
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
Find logs at: ./wandb/run-20230518_043717-dkbivil6/logs
wandb: Sweep Agent: Waiting for job.
wandb: Job received.
wandb: Agent Starting Run: 6x9y411z with config:
wandb:      batch_size: 64
wandb:      epochs: 10
wandb:      learning_rate: 3e-05
wandb:      optimizer: rmsprop
wandb:      weight_decay: 0.05