Skip to content

Fix (brevitas_examples/llm): correct batch size for lm_eval #1430

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
Giuseppe5 merged 2 commits into from
Dec 15, 2025
Merged

Fix (brevitas_examples/llm): correct batch size for lm_eval #1430

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
aa64c7b
Fix (brevitas_examples/llm): correct batch size for lm_eval
Giuseppe5 Dec 15, 2025
7228401
Better default for lm_eval batch size
Giuseppe5 Dec 15, 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
4 changes: 3 additions & 1 deletion src/brevitas_examples/llm/main.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -654,9 +654,11 @@ def quantize_llm(args, extra_args=None):
from lm_eval.models.huggingface import HFLM
with torch.no_grad(), quant_inference_mode(model, compile=args.compile_eval):
model(**calibration_loader[0])
batch_size = 'auto' if args.few_shot_override_batch_size is None else args.few_shot_override_batch_size

wrapped_model = HFLM(
pretrained=model, add_bos_token=True) # need to wrap for LLM eval
pretrained=model, add_bos_token=True,
batch_size=batch_size) # need to wrap for LLM eval
few_shot_eval_results = evaluator.simple_evaluate(
model=wrapped_model,
model_args=None,
Expand Down
Loading