SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2504.04060 · 2025

VocalNet: Speech LLM with Multi-Token Prediction for Faster and High-Quality Generation

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2504.04060.json · badge: /api/ran/2504.04060.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
sjtu-omniagent/vocalnet canonical (classifier label)
classifier basis not recorded on the link
2 ran · 0 no recorded run
FunctionStatusWhere it lives
mel_loss Ran
ran; our contract draft was wrong, not the code
sjtu-omniagent/vocalnet/cosyvoice/utils/losses.py
code served (permissive licence) · get_code("2430375d11b90639")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
import torch.nn.functional as F

torch.manual_seed(0)

# Create small mel transform modules
class MelTransform(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x):
        return self.linear(x)

# Create sample audio tensors (batch_size=2, time_steps=4)
real_speech = torch.randn(2, 4)
generated_speech = torch.randn(2, 4)

# Create list of mel transforms with different output dimensions for variety
mel_transforms = [
    MelTransform(4, 3),
    MelTransform(4, 5),
]

ARGS = {
    "real_speech": real_speech,
    "generated_speech": generated_speech,
    "mel_transforms": mel_transforms,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
tpr_loss Ran
ran; our contract draft was wrong, not the code
sjtu-omniagent/vocalnet/cosyvoice/utils/losses.py
code served (permissive licence) · get_code("af423678123eb287")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn.functional as F

torch.manual_seed(0)

# Create small tensors for disc_real_outputs and disc_generated_outputs
# These are lists of tensors (one per batch or discriminator output)
disc_real_outputs = [
    torch.randn(2, 4),  # small batch of 2, dimension 4
    torch.randn(2, 4),
]

disc_generated_outputs = [
    torch.randn(2, 4),
    torch.randn(2, 4),
]

# tau is a scalar threshold parameter
tau = 0.5

ARGS = {
    "disc_real_outputs": disc_real_outputs,
    "disc_generated_outputs": disc_generated_outputs,
    "tau": tau,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 0 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
mel_loss 1 other paper
1 later
2608.30325 (2026-08)
Sequential Trajectories and Simultaneous Blending: Multi-Emotion Modeling for Instruction-Following TTS
tpr_loss 1 other paper
1 later
2608.30325 (2026-08)
Sequential Trajectories and Simultaneous Blending: Multi-Emotion Modeling for Instruction-Following TTS

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Speech large language models (LLMs) have emerged as a prominent research focus in speech processing. We introduce VocalNet-1B and VocalNet-8B, a series of high-performance, low-latency speech LLMs enabled by a scalable and model-agnostic training framework designed for real-time voice interaction. Central to our contribution is the first application of multi-token prediction (MTP) to speech LLMs. This approach represents a paradigm shift from standard next-token prediction (NTP), offering simultaneous improvements in generation speed and quality. Informed by analysis of MTP's effect on speech generation and experimental comparisons, we designed a straightforward and highly effective MTP implementation. Experiments demonstrate that VocalNet performs on par with mainstream Omni LLMs even with limited training data, and significantly surpasses existing open-source speech LLMs. To foster reproducibility and community advancement, all model weights, inference code, training data, and framework implementations have been made publicly available at https://github.com/SJTU-OmniAgent/VocalNet

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2504.04060")
get_code_for_paper("2504.04060")
have("2504.04060")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2504.04060.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2504.04060.svg)](https://syntology.ai/paper/2504.04060)

Connect an agent — have() is free.