ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

slash

Browse source
This commit is contained in:
Daniel Han-Chen 2024-02-18 19:15:41 +11:00
commit 7573ee2c22
2 changed files with 7 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

6
unsloth/models/llama.py
View file

@ -985,9 +985,9 @@ class FastLlamaModel:
debug_info = """debug_info = \\
f"==((====))== Unsloth - Free Apache OSS license | Num GPUs = {args.world_size}\\n"\\
f" \\\ /| Num examples = {num_examples:,} | Num Epochs = {num_train_epochs:,}\\n"\\
f"O^O/ \_/ \\ Batch size per device = {self._train_batch_size:,} | Gradient Accumulation steps = {args.gradient_accumulation_steps}\\n"\\
f"\ / Total batch size = {total_train_batch_size:,} | Total steps = {max_steps:,}\\n"\\
f" \\\\ /| Num examples = {num_examples:,} | Num Epochs = {num_train_epochs:,}\\n"\\
f"O^O/ \\_/ \\ Batch size per device = {self._train_batch_size:,} | Gradient Accumulation steps = {args.gradient_accumulation_steps}\\n"\\
f"\\ / Total batch size = {total_train_batch_size:,} | Total steps = {max_steps:,}\\n"\\
f' "-____-" Number of trainable parameters = {get_model_param_count(model, trainable_only=True):,}'
logger.warning_once(debug_info)"""

8
unsloth/models/mistral.py
View file

@ -387,12 +387,12 @@ class FastMistralModel(FastLlamaModel):
original_debug = inner_training_loop[start:end]
spaces = re.search('\n([\s\t]{1,})', original_debug).group(0)[1:]
front_spaces = re.match('([\s\t]{1,})', inner_training_loop).group(0)
debug_info = """debug_info = \\
f"==((====))== Unsloth - Free Apache OSS license | Num GPUs = {args.world_size}\\n"\\
f" \\\ /| Num examples = {num_examples:,} | Num Epochs = {num_train_epochs:,}\\n"\\
f"O^O/ \_/ \\ Batch size per device = {self._train_batch_size:,} | Gradient Accumulation steps = {args.gradient_accumulation_steps}\\n"\\
f"\ / Total batch size = {total_train_batch_size:,} | Total steps = {max_steps:,}\\n"\\
f" \\\\ /| Num examples = {num_examples:,} | Num Epochs = {num_train_epochs:,}\\n"\\
f"O^O/ \\_/ \\ Batch size per device = {self._train_batch_size:,} | Gradient Accumulation steps = {args.gradient_accumulation_steps}\\n"\\
f"\\ / Total batch size = {total_train_batch_size:,} | Total steps = {max_steps:,}\\n"\\
f' "-____-" Number of trainable parameters = {get_model_param_count(model, trainable_only=True):,}'
logger.warning_once(debug_info)"""