Skip to content

feat: Enable QLoRA and FSDP compatibility #974

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
IgorKasianenko merged 2 commits into from
Jul 30, 2025
Merged

feat: Enable QLoRA and FSDP compatibility #974

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
34ae0ed
feat: Enable QLoRA and FSDP compatibility
jscaldwell55 Jul 17, 2025
7e17d89
Update finetuning.py
jscaldwell55 Jul 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 15 additions & 5 deletions src/llama_cookbook/finetuning.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -116,7 +116,7 @@ def main(**kwargs):
if not train_config.enable_fsdp or rank == 0:
wandb_run = setup_wandb(train_config, fsdp_config, **kwargs)

# setting quantization configs
# setting quantization configs
bnb_config = None
if train_config.quantization:
if type(train_config.quantization) == type(True):
Expand All @@ -135,6 +135,17 @@ def main(**kwargs):
update_config(quant_config, **kwargs)
bnb_config = quant_config.create_bnb_config(train_config.quantization)


if train_config.enable_fsdp:
if train_config.quantization == "4bit":
bnb_config.bnb_4bit_quant_storage = bnb_config.bnb_4bit_compute_dtype
from logging import getLogger
logger = getLogger()
logger.warning(
"FSDP and 4-bit QLoRA enabled. Setting `bnb_4bit_quant_storage` "
f"to {bnb_config.bnb_4bit_compute_dtype} for compatibility."
)

# Load the pre-trained model and setup its configuration
use_cache = False if train_config.enable_fsdp else None
config = AutoConfig.from_pretrained(train_config.model_name)
Expand Down Expand Up @@ -264,10 +275,9 @@ def main(**kwargs):
elif torch.cuda.is_available():
device_id = torch.cuda.current_device()

if train_config.freeze_LLM_only:
use_orig_params = True
else:
use_orig_params = False

use_orig_params = train_config.freeze_LLM_only or train_config.use_peft

model = FSDP(
model,
auto_wrap_policy=(
Expand Down
Loading