Daniel Voigt Godoy - Deep Learning with PyTorch Step-by-Step A Beginner’s Guide-leanpub
model’s graph (not quite the same as the dynamic computation graph we drew usingmake_dot(), though), and, of course, both scalars: training and validation losses.add_graphLet’s start with add_graph(): unfortunately, its documentation seems to be absent(as at the time of writing), and its default values for arguments lead you to believeyou don’t need to provide any inputs (input_to_model=None). What happens if wetry it?writer.add_graph(model)We’ll get an enormous error message that ends with:Output...TypeError: 'NoneType' object is not iterableSo, we do need to send it some inputs together with our model. Let’s fetch a minibatchof data points from our train_loader and then pass it as input toadd_graph():Adding the Model’s Graph# Fetching a tuple of feature (dummy_x) and label (dummy_y)dummy_x, dummy_y = next(iter(train_loader))# Since our model was sent to device, we need to do the same# with the data.# Even here, both model and data need to be on the same device!writer.add_graph(model, dummy_x.to(device))If you open (or refresh) your browser (or re-run the cell containing the magic%tensorboard --logdir runs inside a notebook) to look at TensorBoard, it shouldlook like this:TensorBoard | 157
Figure 2.4 - Dynamic computation graph on TensorBoardadd_scalarsWhat about sending the loss values to TensorBoard? I’m on it! We can use theadd_scalars() method to send multiple scalar values at once; it needs threearguments:• main_tag: the parent name of the tags, or the "group tag," if you will• tag_scalar_dict: the dictionary containing the key: value pairs for the scalarsyou want to keep track of (in our case, training and validation losses)• global_step: step value; that is, the index you’re associating with the valuesyou’re sending in the dictionary; the epoch comes to mind in our case, as lossesare computed for each epochHow does it translate into code? Let’s check it out:Adding Losseswriter.add_scalars(main_tag='loss',tag_scalar_dict={'training': loss,'validation': val_loss},global_step=epoch)If you run the code above after performing the model training, it will just send bothloss values computed for the last epoch (199). Your TensorBoard will look like this(don’t forget to refresh it—it may take a while if you’re running it on Google Colab):158 | Chapter 2: Rethinking the Training Loop
- Page 132 and 133: "What do we need this for?"It turns
- Page 134 and 135: 1 Instantiating a model2 What IS th
- Page 136 and 137: In the __init__() method, we create
- Page 138 and 139: LayersA Linear model can be seen as
- Page 140 and 141: There are MANY different layers tha
- Page 142 and 143: We use magic, just like that:%run -
- Page 144 and 145: • Step 1: compute model’s predi
- Page 146 and 147: RecapFirst of all, congratulations
- Page 148 and 149: Chapter 2Rethinking the Training Lo
- Page 150 and 151: Let’s take a look at the code onc
- Page 152 and 153: Higher-Order FunctionsAlthough this
- Page 154 and 155: def exponentiation_builder(exponent
- Page 156 and 157: Apart from returning the loss value
- Page 158 and 159: Our code should look like this; see
- Page 160 and 161: There is no need to load the whole
- Page 162 and 163: but if we want to get serious about
- Page 164 and 165: How does this change our code so fa
- Page 166 and 167: Run - Model Training V2%run -i mode
- Page 168 and 169: piece of code that’s going to be
- Page 170 and 171: for it. We could do the same for th
- Page 172 and 173: EvaluationHow can we evaluate the m
- Page 174 and 175: And then, we update our model confi
- Page 176 and 177: Run - Model Training V4%run -i mode
- Page 178 and 179: Loading Extension# Load the TensorB
- Page 180 and 181: browser, you’ll likely see someth
- Page 184 and 185: Figure 2.5 - Scalars on TensorBoard
- Page 186 and 187: Define - Model Training V51 %%write
- Page 188 and 189: If, by any chance, you ended up wit
- Page 190 and 191: The procedure is exactly the same,
- Page 192 and 193: soon, so please bear with me for no
- Page 194 and 195: After recovering our model’s stat
- Page 196 and 197: Run - Model Configuration V31 # %lo
- Page 198 and 199: This is the general structure you
- Page 200 and 201: Chapter 2.1Going ClassySpoilersIn t
- Page 202 and 203: # A completely empty (and useless)
- Page 204 and 205: # These attributes are defined here
- Page 206 and 207: # Creates the train_step function f
- Page 208 and 209: # Builds function that performs a s
- Page 210 and 211: setattrThe setattr function sets th
- Page 212 and 213: See? We effectively modified the un
- Page 214 and 215: the random seed as arguments.This s
- Page 216 and 217: The current state of development of
- Page 218 and 219: Lossesdef plot_losses(self):fig = p
- Page 220 and 221: Run - Data Preparation V21 # %load
- Page 222 and 223: Model TrainingWe start by instantia
- Page 224 and 225: Making PredictionsLet’s make up s
- Page 226 and 227: OutputOrderedDict([('0.weight', ten
- Page 228 and 229: Run - Data Preparation V21 # %load
- Page 230 and 231: • defining our StepByStep class
Figure 2.4 - Dynamic computation graph on TensorBoard
add_scalars
What about sending the loss values to TensorBoard? I’m on it! We can use the
add_scalars() method to send multiple scalar values at once; it needs three
arguments:
• main_tag: the parent name of the tags, or the "group tag," if you will
• tag_scalar_dict: the dictionary containing the key: value pairs for the scalars
you want to keep track of (in our case, training and validation losses)
• global_step: step value; that is, the index you’re associating with the values
you’re sending in the dictionary; the epoch comes to mind in our case, as losses
are computed for each epoch
How does it translate into code? Let’s check it out:
Adding Losses
writer.add_scalars(
main_tag='loss',
tag_scalar_dict={'training': loss,
'validation': val_loss},
global_step=epoch
)
If you run the code above after performing the model training, it will just send both
loss values computed for the last epoch (199). Your TensorBoard will look like this
(don’t forget to refresh it—it may take a while if you’re running it on Google Colab):
158 | Chapter 2: Rethinking the Training Loop