Skip to content

[https://nvbugs/5441438][fix] Set correct draft length for the cuda graph dummy request #6701

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
merged 2 commits into from
Aug 12, 2025
Merged

[https://nvbugs/5441438][fix] Set correct draft length for the cuda graph dummy request #6701

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9c5d663
https://nvbugs/5441438: Set correct draft length for the cuda graph d…
ziyixiong-nv Aug 7, 2025
b21206e
Replace the usage of self.max_draft_len with runtime_draft_len
ziyixiong-nv Aug 8, 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
3 changes: 2 additions & 1 deletion tensorrt_llm/_torch/pyexecutor/model_engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -875,11 +875,12 @@ def _get_padded_batch(
if available_blocks < 1:
return 0

max_draft_len = self.max_draft_len if self.enable_spec_decode else 0
cuda_graph_dummy_request_ids = [MAX_UINT64 - 1]
self.cuda_graph_dummy_request = kv_cache_manager.add_dummy_requests(
cuda_graph_dummy_request_ids,
is_gen=True,
max_num_draft_tokens=self.max_draft_len,
max_num_draft_tokens=max_draft_len,
use_mrope=self.use_mrope,
max_beam_width=self.max_beam_width)[0]
self.cuda_graph_dummy_request.is_cuda_graph_dummy = True
Expand Down
2 changes: 2 additions & 0 deletions tests/integration/defs/accuracy/accuracy_core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -155,6 +155,8 @@ def evaluate(self,
spec_dec_algo = None
elif isinstance(llm.args.speculative_config, DecodingBaseConfig):
spec_dec_algo = llm.args.speculative_config.decoding_type
if spec_dec_algo == 'AUTO':
spec_dec_algo = 'NGram'
else:
raise ValueError(
f"Not recognized speculative_config: {llm.args.speculative_config}."
Expand Down
25 changes: 21 additions & 4 deletions tests/integration/defs/accuracy/test_llm_api_pytorch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,10 +21,10 @@
from tensorrt_llm._torch.modules.fused_moe.fused_moe_triton import \
IS_TRITON_KERNELS_AVAILABLE
from tensorrt_llm._torch.pyexecutor.config import MoeLoadBalancerConfig
from tensorrt_llm.llmapi import (CudaGraphConfig, EagleDecodingConfig,
KvCacheConfig, MoeConfig, MTPDecodingConfig,
NGramDecodingConfig, SamplingParams,
TorchCompileConfig)
from tensorrt_llm.llmapi import (AutoDecodingConfig, CudaGraphConfig,
EagleDecodingConfig, KvCacheConfig, MoeConfig,
MTPDecodingConfig, NGramDecodingConfig,
SamplingParams, TorchCompileConfig)
from tensorrt_llm.quantization import QuantAlgo

from ..conftest import (llm_models_root, parametrize_with_ids, skip_no_hopper,
Expand Down Expand Up @@ -356,6 +356,23 @@ def test_guided_decoding_with_ngram(self, backend: str, mocker):
task = JsonModeEval(self.MODEL_NAME)
task.evaluate(llm)

@skip_pre_hopper
def test_auto_spec_decode(self):
pytorch_config = {
"cuda_graph_config":
CudaGraphConfig(batch_sizes=[1, 32, 64], enable_padding=True)
}
kv_cache_config = KvCacheConfig(enable_block_reuse=False,
free_gpu_memory_fraction=0.5)
spec_config = AutoDecodingConfig()
with LLM(model=self.MODEL_PATH,
**pytorch_config,
kv_cache_config=kv_cache_config,
speculative_config=spec_config,
max_batch_size=64) as llm:
task = GSM8K(self.MODEL_NAME)
task.evaluate(llm)


class TestLlama3_2_1B(LlmapiAccuracyTestHarness):
MODEL_NAME = "meta-llama/Llama-3.2-1B"
Expand Down