Skip to content

[Feature] support mtp3 #33561

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

Open
mingMelody wants to merge 11 commits into
base: main
Choose a base branch
from
Open

[Feature] support mtp3 #33561

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
c7b6eac
feat: support mtp3
mingMelody Feb 2, 2026
feaafdc
fix: function signature
mingMelody Feb 2, 2026
40bf7fe
fix: mtp1 acc with pcg
mingMelody Feb 2, 2026
375a5cf
fix: remove eagle change & useless significance
mingMelody Feb 3, 2026
1b4549b
refactor: cache draft states in req_states
mingMelody Feb 5, 2026
85c581e
update unittest
mingMelody Feb 5, 2026
08b5f5c
update shift and gather 1d kernel
mingMelody Feb 6, 2026
aa8bd05
update shift and gather hidden kernel
mingMelody Feb 6, 2026
a7ed9a4
add comment for triton kernel
mingMelody Feb 9, 2026
873f252
refactor: remove duplicate code
mingMelody Feb 10, 2026
10cd5cb
tidy
mingMelody Feb 10, 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
15 changes: 14 additions & 1 deletion examples/offline_inference/spec_decode.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,6 +62,11 @@ def parse_args():
parser.add_argument("--tp", type=int, default=1)
parser.add_argument("--enforce-eager", action="store_true")
parser.add_argument("--enable-chunked-prefill", action="store_true")
parser.add_argument(
"--enable-multi-layers-mtp",
action="store_true",
help="Enable multi-layer MTP (only effective when --method=mtp).",
)
parser.add_argument("--max-model-len", type=int, default=16384)
parser.add_argument("--temp", type=float, default=0)
parser.add_argument("--top-p", type=float, default=1.0)
Expand All @@ -71,12 +76,14 @@ def parse_args():
parser.add_argument("--model-dir", type=str, default=None)
parser.add_argument("--eagle-dir", type=str, default=None)
parser.add_argument("--draft-model", type=str, default=None)
parser.add_argument("--tokenizer-dir", type=str, default=None)
parser.add_argument("--custom-mm-prompts", action="store_true")
parser.add_argument("--gpu-memory-utilization", type=float, default=0.9)
parser.add_argument("--disable-padded-drafter-batch", action="store_true")
parser.add_argument("--max-num-seqs", type=int, default=None)
parser.add_argument("--parallel-drafting", action="store_true")
parser.add_argument("--allowed-local-media-path", type=str, default="")
parser.add_argument("--trust-remote-code", action="store_true")
return parser.parse_args()


Expand All @@ -90,7 +97,11 @@ def main(args):
"please specify model_dir to give a mm based model"
)
model_dir = "meta-llama/Llama-3.1-8B-Instruct"
tokenizer = AutoTokenizer.from_pretrained(model_dir)
tokenizer_dir = args.tokenizer_dir
if tokenizer_dir is None:
tokenizer_dir = model_dir

tokenizer = AutoTokenizer.from_pretrained(tokenizer_dir)

if args.custom_mm_prompts:
prompts = llm_prompts = get_custom_mm_prompts(args.num_prompts)
Expand Down Expand Up @@ -146,6 +157,8 @@ def main(args):
"method": "mtp",
"num_speculative_tokens": args.num_spec_tokens,
}
if args.enable_multi_layers_mtp:
speculative_config["enable_multi_layers_mtp"] = True
else:
raise ValueError(f"unknown method: {args.method}")

Expand Down
Loading