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

Update llama.py

Browse source
This commit is contained in:
Daniel Han 2025-03-14 01:25:35 -07:00
commit 382a8aef2a
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
unsloth/models/llama.py
View file

@ -1890,9 +1890,9 @@ class FastLlamaModel:
gc.collect()
torch.cuda.empty_cache()"""
multiplier = \
"4.5 if getattr(model.config, 'quantization_config', \\{'load_in_4bit' : False\\})['load_in_4bit'] else "\
"8.0 if getattr(model.config, 'quantization_config', \\{'load_in_8bit' : False\\})['load_in_8bit'] else 1.0"
debug_info = debug_info.replace("multiplier__", "(" + multiplier + ")")
"(4.5 if getattr(model.config, 'quantization_config', \\{'load_in_4bit' : False\\})['load_in_4bit'] else "\
"(8.0 if getattr(model.config, 'quantization_config', \\{'load_in_8bit' : False\\})['load_in_8bit'] else 1.0)"
debug_info = debug_info.replace("multiplier__", multiplier)
debug_info = debug_info.replace("P__", "get_model_param_count")
debug_info = debug_info.split('\n')