Skip to content

Porting ERNIE 4.5 and ERNIE 4.5 MoE model support #1779

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 25, 2025
Merged

Porting ERNIE 4.5 and ERNIE 4.5 MoE model support #1779

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
ba505e3
[Model] Add Ernie4.5 and Ernie4.5MoE Model Support (#20220)
CSWYF3634076 Jul 2, 2025
c0c75de
[`Ernie 4.5`] Name Change for Base 0.3B Model (#21735)
vasqu Jul 28, 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
11 changes: 6 additions & 5 deletions docs/models/supported_models.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -314,6 +314,8 @@ Specified using `--task generate`.
| `DeepseekForCausalLM` | DeepSeek | `deepseek-ai/deepseek-llm-67b-base`, `deepseek-ai/deepseek-llm-7b-chat` etc. | | ✅︎ |
| `DeepseekV2ForCausalLM` | DeepSeek-V2 | `deepseek-ai/DeepSeek-V2`, `deepseek-ai/DeepSeek-V2-Chat` etc. | | ✅︎ |
| `DeepseekV3ForCausalLM` | DeepSeek-V3 | `deepseek-ai/DeepSeek-V3-Base`, `deepseek-ai/DeepSeek-V3` etc. | | ✅︎ |
| `Ernie4_5ForCausalLM` | Ernie4.5 | `baidu/ERNIE-4.5-0.3B-PT`,etc. | | ✅︎ | ✅︎ |
| `Ernie4_5_MoeForCausalLM` | Ernie4.5MoE | `baidu/ERNIE-4.5-21B-A3B-PT`, `baidu/ERNIE-4.5-300B-A47B-PT`, etc. | | ✅︎ | ✅︎ |
| `ExaoneForCausalLM` | EXAONE-3 | `LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct`, etc. | ✅︎ | ✅︎ |
| `FalconForCausalLM` | Falcon | `tiiuae/falcon-7b`, `tiiuae/falcon-40b`, `tiiuae/falcon-rw-7b`, etc. | | ✅︎ |
| `FalconMambaForCausalLM` | FalconMamba | `tiiuae/falcon-mamba-7b`, `tiiuae/falcon-mamba-7b-instruct`, etc. | | ✅︎ |
Expand Down Expand Up @@ -371,7 +373,6 @@ Specified using `--task generate`.
| `XverseForCausalLM` | XVERSE | `xverse/XVERSE-7B-Chat`, `xverse/XVERSE-13B-Chat`, `xverse/XVERSE-65B-Chat`, etc. | ✅︎ | ✅︎ |
| `MiniMaxText01ForCausalLM` | MiniMax-Text | `MiniMaxAI/MiniMax-Text-01`, etc. | | |
| `Zamba2ForCausalLM` | Zamba2 | `Zyphra/Zamba2-7B-instruct`, `Zyphra/Zamba2-2.7B-instruct`, `Zyphra/Zamba2-1.2B-instruct`, etc. | | |

!!! note
Currently, the ROCm version of vLLM supports Mistral and Mixtral only for context lengths up to 4096.

Expand Down Expand Up @@ -556,10 +557,10 @@ Specified using `--task generate`.
| `SmolVLMForConditionalGeneration` | SmolVLM2 | T + I | `SmolVLM2-2.2B-Instruct` | ✅︎ | | ✅︎ |
| `TarsierForConditionalGeneration` | Tarsier | T + I<sup>E+</sup> | `omni-search/Tarsier-7b`,`omni-search/Tarsier-34b` | | ✅︎ | ✅︎ |

<sup>^</sup> You need to set the architecture name via `--hf-overrides` to match the one in vLLM.
&nbsp;&nbsp;&nbsp;&nbsp;• For example, to use DeepSeek-VL2 series models:
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;`--hf-overrides '{"architectures": ["DeepseekVLV2ForCausalLM"]}'`
<sup>E</sup> Pre-computed embeddings can be inputted for this modality.
<sup>^</sup> You need to set the architecture name via `--hf-overrides` to match the one in vLLM.
&nbsp;&nbsp;&nbsp;&nbsp;• For example, to use DeepSeek-VL2 series models:
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;`--hf-overrides '{"architectures": ["DeepseekVLV2ForCausalLM"]}'`
<sup>E</sup> Pre-computed embeddings can be inputted for this modality.
<sup>+</sup> Multiple items can be inputted per text prompt for this modality.

!!! warning
Expand Down
4 changes: 4 additions & 0 deletions tests/models/registry.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,6 +156,10 @@ def check_available_online(
trust_remote_code=True),
"DeepseekV3ForCausalLM": _HfExamplesInfo("deepseek-ai/DeepSeek-V3", # noqa: E501
trust_remote_code=True),
"Ernie4_5ForCausalLM": _HfExamplesInfo("baidu/ERNIE-4.5-0.3B-PT",
min_transformers_version="4.54"),
"Ernie4_5_MoeForCausalLM": _HfExamplesInfo("baidu/ERNIE-4.5-21B-A3B-PT",
trust_remote_code=True),
"ExaoneForCausalLM": _HfExamplesInfo("LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct"), # noqa: E501
"Fairseq2LlamaForCausalLM": _HfExamplesInfo("mgleize/fairseq2-dummy-Llama-3.2-1B"), # noqa: E501
"FalconForCausalLM": _HfExamplesInfo("tiiuae/falcon-7b"),
Expand Down
43 changes: 43 additions & 0 deletions vllm/model_executor/models/ernie45.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,43 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project

# Copyright 2025 The Baidu team.
# Copyright 2023 The vLLM team.
# Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommodate minor architectural differences compared
# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Inference-only Erine model compatible with HuggingFace weights."""
from vllm.config import VllmConfig
from vllm.model_executor.models.llama import LlamaForCausalLM

from .utils import PPMissingLayer


class Ernie4_5ForCausalLM(LlamaForCausalLM):

def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
super().__init__(vllm_config=vllm_config, prefix=prefix)
# Hack Llama model to fit HF format Ernie4.5 dense implementation
# Attention difference between Ernie and Llama:
# 1. rotary_dim and no Neox style.
# 2. There is no bias for o_proj in attention
for layer in self.model.layers:
if not isinstance(layer, PPMissingLayer):
layer.self_attn.rotary_emb.is_neox_style = False
layer.self_attn.o_proj.bias = None
layer.self_attn.o_proj.skip_bias_add = True
Loading