books/LLMs-from-scratch
1
0
Fork 0
mirror of https://github.com/rasbt/LLMs-from-scratch.git synced 2026-04-10 12:33:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

fixes for code (#206)

Browse Source 
* updated .gitignore

* removed unused GELU import

* fixed model_configs, fixed all tensors on same device

* removed unused tiktoken

* update

* update hparam search

* remove redundant tokenizer argument

---------

Co-authored-by: rasbt <mail@sebastianraschka.com>
This commit is contained in:
Daniel Kleine
2024-06-12 03:59:48 +02:00
committed by GitHub
parent 1a65020d81
commit dcbdc1d2e5
12 changed files with 33 additions and 46 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
ch06/01_main-chapter-code/ch06.ipynb
View File

@@ -1861,7 +1861,7 @@
"source": [
"# Overall the same as `train_model_simple` in chapter 5\n",
"def train_classifier_simple(model, train_loader, val_loader, optimizer, device, num_epochs,\n",
" eval_freq, eval_iter, tokenizer):\n",
" eval_freq, eval_iter):\n",
" # Initialize lists to track losses and examples seen\n",
" train_losses, val_losses, train_accs, val_accs = [], [], [], []\n",
" examples_seen, global_step = 0, -1\n",
@@ -1982,7 +1982,6 @@
"train_losses, val_losses, train_accs, val_accs, examples_seen = train_classifier_simple(\n",
" model, train_loader, val_loader, optimizer, device,\n",
" num_epochs=num_epochs, eval_freq=50, eval_iter=5,\n",
" tokenizer=tokenizer\n",
")\n",
"\n",
"end_time = time.time()\n",
@@ -2371,7 +2370,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.6"
"version": "3.11.4"
}
},
"nbformat": 4,

4
ch06/02_bonus_additional-experiments/additional-experiments.py
View File

@@ -235,7 +235,7 @@ def evaluate_model(model, train_loader, val_loader, device,
def train_classifier_simple(model, train_loader, val_loader, optimizer, device, num_epochs,
eval_freq, eval_iter, tokenizer, max_steps=None, trainable_token_pos=-1,
eval_freq, eval_iter, max_steps=None, trainable_token_pos=-1,
accumulation_steps=1, ignore_index=-100):
# Initialize lists to track losses and tokens seen
train_losses, val_losses, train_accs, val_accs = [], [], [], []
@@ -565,7 +565,7 @@ if __name__ == "__main__":
train_losses, val_losses, train_accs, val_accs, examples_seen = train_classifier_simple(
model, train_loader, val_loader, optimizer, device,
num_epochs=args.num_epochs, eval_freq=50, eval_iter=5,
tokenizer=tokenizer, max_steps=None, trainable_token_pos=args.trainable_token_pos,
max_steps=None, trainable_token_pos=args.trainable_token_pos,
accumulation_steps=args.accumulation_steps
)

4
ch06/03_bonus_imdb-classification/train-bert-hf.py
View File

@@ -110,7 +110,7 @@ def evaluate_model(model, train_loader, val_loader, device, eval_iter):
def train_classifier_simple(model, train_loader, val_loader, optimizer, device, num_epochs,
eval_freq, eval_iter, tokenizer, max_steps=None):
eval_freq, eval_iter, max_steps=None):
# Initialize lists to track losses and tokens seen
train_losses, val_losses, train_accs, val_accs = [], [], [], []
examples_seen, global_step = 0, -1
@@ -279,7 +279,7 @@ if __name__ == "__main__":
train_losses, val_losses, train_accs, val_accs, examples_seen = train_classifier_simple(
model, train_loader, val_loader, optimizer, device,
num_epochs=num_epochs, eval_freq=50, eval_iter=20,
tokenizer=tokenizer, max_steps=None
max_steps=None
)
end_time = time.time()

4
ch06/03_bonus_imdb-classification/train-gpt.py
View File

@@ -139,7 +139,7 @@ def evaluate_model(model, train_loader, val_loader, device, eval_iter, trainable
def train_classifier_simple(model, train_loader, val_loader, optimizer, device, num_epochs,
eval_freq, eval_iter, tokenizer, max_steps=None, trainable_token=-1):
eval_freq, eval_iter, max_steps=None, trainable_token=-1):
# Initialize lists to track losses and tokens seen
train_losses, val_losses, train_accs, val_accs = [], [], [], []
examples_seen, global_step = 0, -1
@@ -344,7 +344,7 @@ if __name__ == "__main__":
train_losses, val_losses, train_accs, val_accs, examples_seen = train_classifier_simple(
model, train_loader, val_loader, optimizer, device,
num_epochs=num_epochs, eval_freq=50, eval_iter=20,
tokenizer=tokenizer, max_steps=None, trainable_token=args.trainable_token
max_steps=None, trainable_token=args.trainable_token
)
end_time = time.time()