Skip to content

test: add nanov3 prefill decode test #2141

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
yuki-97 merged 2 commits into from
Mar 24, 2026
+346 −37
Merged

test: add nanov3 prefill decode test #2141

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
30f2253
test: add vLLM prefill-decode logprob consistency CI test for Nano v3
ZhiyuLi-Nvidia Mar 23, 2026
d255ec9
lint
ZhiyuLi-Nvidia Mar 23, 2026
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
2 changes: 1 addition & 1 deletion docs/adding-new-models.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -148,7 +148,7 @@ Test that vLLM yields near-identical token log-probabilities when comparing deco

```sh
# Run that is expected to pass
uv run --extra vllm tools/model_diagnostics/2.long_generation_decode_vs_prefill.py Qwen/Qwen2.5-1.5B
uv run --extra vllm tools/model_diagnostics/2.long_generation_decode_vs_prefill.py --model Qwen/Qwen2.5-1.5B
# ...
# [Qwen/Qwen2.5-1.5B] ALL GOOD!
```
Expand Down
1 change: 1 addition & 0 deletions tests/functional/L1_Functional_Tests_GPU.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,6 +73,7 @@ run_test uv run --no-sync bash ./tests/functional/sft_megatron_lora.sh
run_test uv run --no-sync bash ./tests/functional/sft_resume_diamond.sh
run_test uv run --no-sync bash ./tests/functional/test_automodel_extra_installed_correctly.sh
run_test fast uv run --no-sync bash ./tests/functional/test_converters.sh
run_test uv run --no-sync bash ./tests/functional/test_decode_vs_prefill.sh
run_test uv run --no-sync bash ./tests/functional/test_mcore_extra_installed_correctly.sh
run_test uv run --no-sync bash ./tests/functional/vlm_grpo.sh

Expand Down
6 changes: 6 additions & 0 deletions tests/functional/test_decode_vs_prefill.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
uv run --extra vllm python tools/model_diagnostics/2.long_generation_decode_vs_prefill.py \
--model nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16 \
--prompts arc \
--max-tokens 8192 \
--num-batches 4 \
--tensor-parallel-size 2 \
Loading
Loading