yujunjun/LLMs-from-scratch
1
0
Fork 0
mirror of https://github.com/rasbt/LLMs-from-scratch.git synced 2025-08-08 00:32:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix issue 724: unused args (#726)

Browse Source 
* Fix issue 724: unused args

* Update 02_opt_multi_gpu_ddp.py
This commit is contained in:
Matthew Hernandez 2025-07-08 04:37:39 -07:00 committed by GitHub
parent c8c6e7814a
commit 83c76891fc
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 3 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ch02/05_bpe-from-scratch/tests/tests.py
View File

@ -58,7 +58,7 @@ def gpt2_files(imported_module):
return paths
def test_tokenizer_training(imported_module, gpt2_files):
def test_tokenizer_training(imported_module):
BPETokenizerSimple = getattr(imported_module, "BPETokenizerSimple", None)
download_file_if_absent = getattr(imported_module, "download_file_if_absent", None)

5
ch05/10_llm-training-speed/02_opt_multi_gpu_ddp.py
View File

@ -312,7 +312,7 @@ def generate_and_print_sample(model, device, start_context):
def train_model_simple_with_timing(model, train_loader, val_loader, optimizer, device,
num_epochs, eval_freq, eval_iter, start_context, tokenizer):
num_epochs, eval_freq, eval_iter, start_context):
train_losses, val_losses, track_tokens = [], [], []
total_tokens, global_step, last_tokens = 0, -1, 0
@ -524,8 +524,6 @@ def main(gpt_config, settings, rank, world_size):
# Train model
##############################
tokenizer = tiktoken.get_encoding("gpt2")
train_losses, val_losses, tokens_seen = train_model_simple_with_timing(
model=model,
train_loader=train_loader,
@ -536,7 +534,6 @@ def main(gpt_config, settings, rank, world_size):
eval_freq=5,
eval_iter=1,
start_context="Every effort moves you",
tokenizer=tokenizer
)
# NEW: Clean up distributed processes

3
ch06/01_main-chapter-code/gpt_class_finetune.py
View File

@ -175,7 +175,7 @@ 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):
eval_freq, eval_iter):
# Initialize lists to track losses and tokens seen
train_losses, val_losses, train_accs, val_accs = [], [], [], []
examples_seen, global_step = 0, -1
@ -408,7 +408,6 @@ 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=5,
tokenizer=tokenizer
)
end_time = time.time()