Skip to content

[None][fix] Eagle: Attention DP #7939

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
mikeiovine merged 1 commit into from
Oct 6, 2025
Merged

[None][fix] Eagle: Attention DP #7939

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
29 changes: 20 additions & 9 deletions tensorrt_llm/_torch/models/modeling_speculative.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,14 +48,16 @@ def __init__(
)

tp_size = model_config.mapping.tp_size
if model_config.mapping.enable_attention_dp:
tp_size = 1
# Override the QKV projection. The number of input features
# is twice as big for EAGLE3 draft models.
self.qkv_proj = Linear(
2 * self.hidden_size,
tp_size * self.q_size + 2 * tp_size * self.kv_size,
bias=config.attention_bias,
dtype=config.torch_dtype,
mapping=model_config.mapping,
mapping=self.qkv_proj.mapping,
Comment thread
IzzyPutterman marked this conversation as resolved.
tensor_parallel_mode=TensorParallelMode.COLUMN,
weights_loading_config=WeightsLoadingConfig(
weight_mode=WeightMode.FUSED_QKV_LINEAR),
Expand Down Expand Up @@ -89,6 +91,8 @@ def __init__(
bias=getattr(config, "mlp_bias", False),
dtype=config.torch_dtype,
config=model_config,
overridden_tp_size=1
if model_config.mapping.enable_attention_dp else None,
)
self.input_layernorm = RMSNorm(hidden_size=config.hidden_size,
eps=config.rms_norm_eps,
Expand Down Expand Up @@ -182,14 +186,21 @@ def __init__(
requires_grad=False)

if self.hidden_size_in != config.hidden_size:
self.embed_tokens = Embedding(
config.vocab_size,
config.hidden_size,
dtype=config.torch_dtype,
mapping=model_config.mapping,
tensor_parallel_mode=TensorParallelMode.COLUMN,
gather_output=True,
)
if model_config.mapping.enable_attention_dp:
self.embed_tokens = Embedding(
config.vocab_size,
config.hidden_size,
dtype=config.torch_dtype,
)
else:
self.embed_tokens = Embedding(
config.vocab_size,
config.hidden_size,
dtype=config.torch_dtype,
mapping=model_config.mapping,
tensor_parallel_mode=TensorParallelMode.COLUMN,
gather_output=True,
)
else:
# Shared with target model.
self.embed_tokens = None
Expand Down
51 changes: 27 additions & 24 deletions tests/unittest/_torch/speculative/test_eagle3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,38 +24,40 @@ def enforce_single_worker(monkeypatch):


@pytest.mark.parametrize(
"use_cuda_graph,attn_backend,disable_overlap_scheduler,enable_block_reuse,use_one_model,enable_chunked_prefill,use_chain_drafter,multi_batch",
"use_cuda_graph,attn_backend,disable_overlap_scheduler,enable_block_reuse,use_one_model,enable_chunked_prefill,use_chain_drafter,multi_batch,attention_dp",
[
[True, "TRTLLM", True, False, False, False, True, False],
[True, "TRTLLM", True, False, False, False, False, False],
[False, "TRTLLM", True, False, False, False, True, False],
[False, "TRTLLM", True, False, False, False, False, False],
[True, "FLASHINFER", True, False, False, False, True, False],
[False, "FLASHINFER", True, False, False, False, True, False],
[False, "TRTLLM", False, True, True, False, True, False],
[True, "TRTLLM", False, True, True, False, True, False],
[True, "TRTLLM", True, False, True, True, True, False],
[True, "TRTLLM", True, False, True, False, True, False],
[True, "TRTLLM", True, False, False, False, True, False, False],
[True, "TRTLLM", True, False, False, False, False, False, False],
[False, "TRTLLM", True, False, False, False, True, False, False],
[False, "TRTLLM", True, False, False, False, False, False, False],
[True, "FLASHINFER", True, False, False, False, True, False, False],
[False, "FLASHINFER", True, False, False, False, True, False, False],
[False, "TRTLLM", False, True, True, False, True, False, False],
[True, "TRTLLM", False, True, True, False, True, False, False],
[True, "TRTLLM", True, False, True, True, True, False, False],
[True, "TRTLLM", True, False, True, False, True, False, False],
# TODO: nvbugs/5461761
# [True, "TRTLLM", True, False, False, True, True, False],
[True, "TRTLLM", False, False, False, False, True, False],
[False, "TRTLLM", False, False, False, False, True, False],
[True, "TRTLLM", False, False, False, False, False, True],
[False, "TRTLLM", False, False, False, False, False, True],
[True, "TRTLLM", False, False, False, False, True, True],
[False, "TRTLLM", False, False, False, False, True, True],
[True, "TRTLLM", False, False, False, False, False, False],
[False, "TRTLLM", False, False, False, False, False, False],
[True, "TRTLLM", False, False, False, True, True, False],
[True, "TRTLLM", False, False, False, True, False, False],
[True, "FLASHINFER", False, False, False, False, True, False],
[False, "FLASHINFER", False, False, False, False, True, False],
[True, "TRTLLM", False, False, False, False, True, False, False],
[False, "TRTLLM", False, False, False, False, True, False, False],
[True, "TRTLLM", False, False, False, False, False, True, False],
[True, "TRTLLM", False, False, False, False, False, True, True],
[False, "TRTLLM", False, False, False, False, False, True, False],
[True, "TRTLLM", False, False, False, False, True, True, False],
[False, "TRTLLM", False, False, False, False, True, True, False],
[True, "TRTLLM", False, False, False, False, False, False, False],
[False, "TRTLLM", False, False, False, False, False, False, False],
[True, "TRTLLM", False, False, False, True, True, False, False],
[True, "TRTLLM", False, False, False, True, False, False, False],
[True, "FLASHINFER", False, False, False, False, True, False, False],
[False, "FLASHINFER", False, False, False, False, True, False, False],
])
@pytest.mark.high_cuda_memory
def test_llama_eagle3(use_cuda_graph: bool, attn_backend: str,
disable_overlap_scheduler: bool, enable_block_reuse: bool,
use_one_model: bool, enable_chunked_prefill: bool,
use_chain_drafter: bool, multi_batch: bool, request):
use_chain_drafter: bool, multi_batch: bool,
attention_dp: bool, request):
# Use enforce_single_worker fixture only when use_chain_drafter is False.
# Otherwise, we can't modify the returned value of _get_allow_chain_drafter in multiprocessing.
if not use_chain_drafter:
Expand Down Expand Up @@ -98,6 +100,7 @@ def test_llama_eagle3(use_cuda_graph: bool, attn_backend: str,
cuda_graph_config=cuda_graph_config,
max_batch_size=max_batch_size,
kv_cache_config=kv_cache_config,
enable_attention_dp=attention_dp,
# This max_seq_len is larger than the one specified
# in the llama 3 8B eagle's config. We want to make sure
# that the draft model won't go above its max in warmup
Expand Down