[Model] Add Ernie4.5 and Ernie4.5MoE Model Support #20220

CSWYF3634076 · 2025-06-29T07:20:01Z

Purpose

Support Baidu Ernie4.5 model for vllm
In this PR, I have provided the implementation of the Ernie4.5 and Ernie4.5MoE model structure through two files: ernie45.py
and ernie45_moe. py.

github-actions · 2025-06-29T07:20:10Z

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

gemini-code-assist

Summary of Changes

Hello @CSWYF3634076, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly expands vLLM's model compatibility by integrating the Ernie 4.5 and Ernie 4.5 MoE architectures. The changes involve adding the necessary model definitions and logic to enable efficient inference for these new large language models, including their specialized Mixture of Experts components.

Highlights

New Model Support: I've added comprehensive support for the Ernie 4.5 and Ernie 4.5 MoE models. This includes their specific architectural components like MLP, Attention, and DecoderLayer, tailored for vLLM's efficient inference.
Mixture of Experts (MoE) Integration: For the Ernie 4.5 MoE model, I've implemented the Mixture of Experts (MoE) layer, incorporating support for both expert routing and shared experts, which is crucial for its unique architecture.
Documentation and Test Coverage: The pull request updates the documentation to list the newly supported Ernie models and adds corresponding entries to the test registry. This ensures that the models are discoverable and can be loaded correctly for testing and usage.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces support for Ernie4.5 and Ernie4.5MoE models, including their implementations, documentation updates, and test registry entries. The code is well-structured and aligns with existing patterns in the vLLM codebase. I've provided a minor suggestion to enhance code clarity and consistency. Overall, this is a solid contribution.

vllm/model_executor/models/ernie45_moe.py

Isotr0py

Thanks for adding these models! Just added some initial comments about dense model, PTAL!

vllm/model_executor/models/ernie45.py

docs/models/supported_models.md

mergify · 2025-06-30T03:44:03Z

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @CSWYF3634076.

https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork

xjpang · 2025-06-30T09:28:31Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.

request:
curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'

response:
{"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

prnake · 2025-06-30T09:49:15Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.

request: curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'

response: {"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

Same issue, and I'm using the branch CSWYF3634076:dev, so it's very likely a implementation problem.

vllm/model_executor/models/ernie45_moe.py

CSWYF3634076 · 2025-06-30T10:52:04Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.

request: curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'

response: {"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

I just executed the entire process once and didn't reproduce it. Could you please post your vllm serve command

prnake · 2025-06-30T12:17:11Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.
request: curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'
response: {"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

I just executed the entire process once and didn't reproduce it. Could you please post your vllm serve command

python3 -m vllm.entrypoints.openai.api_server \
  --model /models/ERNIE-4.5-21B-A3B-PT     \
  --served-model-name ERNIE-4.5-21B-A3B-PT    \
  --tensor-parallel-size 8     \
  --dtype auto \
  --trust-remote-code

xjpang · 2025-06-30T12:30:24Z

/models/ERNIE-4.5-21B-A3B-PT

['python3', '-m', 'vllm.entrypoints.openai.api_server', '--model=/models/ERNIE-4.5-21B-A3B-PT', '--host=127.0.0.1', '--port=11522', '--trust-remote-code', '--tensor-parallel-size=2', '--gpu-memory-utilization=0.95', '--max-logprobs=20', '--swap-space=10', '--max-num-seqs=32', '--no-enable-prefix-caching', '--enable-chunked-prefill', '--max-num-batched-tokens=2048']

CSWYF3634076 · 2025-06-30T13:25:03Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.
request: curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'
response: {"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

I just executed the entire process once and didn't reproduce it. Could you please post your vllm serve command
python3 -m vllm.entrypoints.openai.api_server \
  --model /models/ERNIE-4.5-21B-A3B-PT     \
  --served-model-name ERNIE-4.5-21B-A3B-PT    \
  --tensor-parallel-size 8     \
  --dtype auto \
  --trust-remote-code

I have reproduced the issue, which occurred after I rebased the main branch. Please use branch CSWYF3634076:ernie to experience it first. I am currently investigating the cause of the issue and apologize for any inconvenience caused

Adapted from this PR: vllm-project/vllm#20220 --------- Co-authored-by: Corle-hyz <heyongzhe18@mails.ucas.edu.cn>

jeejeelee

Some NIT comments, Please provide feedback on https://github.com/vllm-project/vllm/pull/20220/files#r2173637449

docs/models/supported_models.md

tests/models/registry.py

vllm/model_executor/models/ernie45.py

vllm/model_executor/models/ernie45_moe.py

vllm/model_executor/models/registry.py

xjpang · 2025-07-01T03:32:56Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.
request: curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'
response: {"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

I just executed the entire process once and didn't reproduce it. Could you please post your vllm serve command
python3 -m vllm.entrypoints.openai.api_server \
  --model /models/ERNIE-4.5-21B-A3B-PT     \
  --served-model-name ERNIE-4.5-21B-A3B-PT    \
  --tensor-parallel-size 8     \
  --dtype auto \
  --trust-remote-code
I have reproduced the issue, which occurred after I rebased the main branch. Please use branch CSWYF3634076:ernie to experience it first. I am currently investigating the cause of the issue and apologize for any inconvenience caused

I used CSWYF3634076:ernie branch, and asked "你能做什么？"。
The outputs seems still have something wrong.

CSWYF3634076 · 2025-07-01T04:37:47Z

I merged this mr to v0.9.1, and served the ERNIE-4.5-21B-A3B-PT model. The output is garbled text, just like the text below.
request: curl -X POST "http://127.0.0.1:11522/v1/chat/completions" -H "Content-Type: application/json" -d '{ "model":"baidu/ERNIE-4.5-21B-A3B-PT", "messages": [ { "role": "user", "content": "你好" } ], "top_p": 1.0, "top_k": 40, "temperature": 1.0, "max_tokens": 128, "repetition_penalty": 1.0 }'
response: {"id":"chatcmpl-8b62dd62ffa0400b82254ac0030c7fe5","object":"chat.completion","created":1751275598,"model":"baidu/ERNIE-4.5-21B-A3B-PT","choices":[{"index":0,"message":{"role":"assistant","reasoning_content":null,"content":"彩 PC分 ⇒南昌 –“**о ** arrows企 shows – – –拖画拖 \"拖挤 \r丢 – –琶**________ arrows彩**丢④****** ****** **** \r**** ** ** **丢**** ** **②** ** ** ************ **潮②** ** **摇 ** **** **** ** **媒 ** ****②** \r** ** ****** ** ********① ****/ ******，、③→、→→ \r →“ heading-->\r **→→*/\r。② />\r。","tool_calls":[]},"logprobs":null,"finish_reason":"length","stop_reason":null}],"usage":{"prompt_tokens":9,"total_tokens":137,"completion_tokens":128,"prompt_tokens_details":null},"prompt_logprobs":null,"kv_transfer_params":null}

I just executed the entire process once and didn't reproduce it. Could you please post your vllm serve command
python3 -m vllm.entrypoints.openai.api_server \
  --model /models/ERNIE-4.5-21B-A3B-PT     \
  --served-model-name ERNIE-4.5-21B-A3B-PT    \
  --tensor-parallel-size 8     \
  --dtype auto \
  --trust-remote-code
I have reproduced the issue, which occurred after I rebased the main branch. Please use branch CSWYF3634076:ernie to experience it first. I am currently investigating the cause of the issue and apologize for any inconvenience caused
I used CSWYF3634076:ernie branch, and asked "你能做什么？"。 The outputs seems still have something wrong.

May I ask if the recommended hyperparameters are being used. I have created an issue specifically for collecting questions, and you can continue to ask questions here. CSWYF3634076#1

Adapted from this PR: vllm-project/vllm#20220 --------- Co-authored-by: Corle-hyz <heyongzhe18@mails.ucas.edu.cn>

Isotr0py

LGTM now! Thanks for adding this model!

CSWYF3634076 · 2025-07-02T13:58:24Z

#20376

SongDI911 · 2025-07-02T14:09:36Z

When I was testing Model Ernie4_5_VLMoeForConditionalGeneration, I found that it seemed not to support the visual model

ValueError: Ernie4_5_VLMoeForConditionalGeneration has no vLLM implementation and the Transformers implementation is not compatible with vLLM. Try setting VLLM_USE_V1=0.

CSWYF3634076 · 2025-07-02T14:20:58Z

When I was testing Model Ernie4_5_VLMoeForConditionalGeneration, I found that it seemed not to support the visual model

ValueError: Ernie4_5_VLMoeForConditionalGeneration has no vLLM implementation and the Transformers implementation is not compatible with vLLM. Try setting VLLM_USE_V1=0.

The vllm running the Ernie45 multimodal model is still under development. Support will be provided in the near future, You can use FastDeploy to deploy the paddle version for priority experience

SongDI911 · 2025-07-02T15:19:47Z

When I was testing Model Ernie4_5_VLMoeForConditionalGeneration, I found that it seemed not to support the visual model
ValueError: Ernie4_5_VLMoeForConditionalGeneration has no vLLM implementation and the Transformers implementation is not compatible with vLLM. Try setting VLLM_USE_V1=0.

The vllm running the Ernie45 multimodal model is still under development. Support will be provided in the near future, You can use FastDeploy to deploy the paddle version for priority experience

FastDeploy has other bugs such as memory sharing。。。。。

CSWYF3634076 · 2025-07-03T06:20:27Z

@xjpang @prnake Using the latest vllm-project/vllm main branch code will no longer output in garbled texts

CSWYF3634076 · 2025-07-03T06:24:37Z

When I was testing Model Ernie4_5_VLMoeForConditionalGeneration, I found that it seemed not to support the visual model
ValueError: Ernie4_5_VLMoeForConditionalGeneration has no vLLM implementation and the Transformers implementation is not compatible with vLLM. Try setting VLLM_USE_V1=0.

The vllm running the Ernie45 multimodal model is still under development. Support will be provided in the near future, You can use FastDeploy to deploy the paddle version for priority experience

FastDeploy has other bugs such as memory sharing。。。。。

@SongDI911 Can you open an issue in FastDeploy, describe the problem, or join our ernie user WeChat group

SongDI911 · 2025-07-03T07:29:34Z

When I was testing Model Ernie4_5_VLMoeForConditionalGeneration, I found that it seemed not to support the visual model
ValueError: Ernie4_5_VLMoeForConditionalGeneration has no vLLM implementation and the Transformers implementation is not compatible with vLLM. Try setting VLLM_USE_V1=0.

The vllm running the Ernie45 multimodal model is still under development. Support will be provided in the near future, You can use FastDeploy to deploy the paddle version for priority experience

FastDeploy has other bugs such as memory sharing。。。。。

@SongDI911 Can you open an issue in FastDeploy, describe the problem, or join our ernie user WeChat group

PaddlePaddle/FastDeploy#2683 here
I've applied for a wechat group, name is gxs

Signed-off-by: wangyafeng <wangyafeng@baidu.com>

CSWYF3634076 requested review from hmellor, DarkLight1337 and ywang96 as code owners June 29, 2025 07:20

gemini-code-assist bot reviewed Jun 29, 2025

View reviewed changes

mergify bot added the documentation Improvements or additions to documentation label Jun 29, 2025

gemini-code-assist bot reviewed Jun 29, 2025

View reviewed changes

vllm/model_executor/models/ernie45_moe.py Outdated Show resolved Hide resolved

Isotr0py reviewed Jun 29, 2025

View reviewed changes

vllm/model_executor/models/ernie45.py Outdated Show resolved Hide resolved

docs/models/supported_models.md Outdated Show resolved Hide resolved

Isotr0py added the new-model Requests to new models label Jun 29, 2025

jzhang533 mentioned this pull request Jun 29, 2025

model : add support for ERNIE 4.5 0.3B model ggml-org/llama.cpp#14408

Merged

mergify bot added the needs-rebase label Jun 30, 2025

prnake reviewed Jun 30, 2025

View reviewed changes

vllm/model_executor/models/ernie45_moe.py Show resolved Hide resolved

Hchnr mentioned this pull request Jun 30, 2025

[Model] Add Ernie-4.5 FlagOpen/FlagScale#637

Merged

EliEron mentioned this pull request Jun 30, 2025

Feature Request: Add Ernie4.5MoE support ggml-org/llama.cpp#14465

Closed

4 tasks

ceci3 pushed a commit to FlagOpen/FlagScale that referenced this pull request Jul 1, 2025

[Model] Add Ernie-4.5 (#637)

39073a5

Adapted from this PR: vllm-project/vllm#20220 --------- Co-authored-by: Corle-hyz <heyongzhe18@mails.ucas.edu.cn>

jeejeelee reviewed Jul 1, 2025

View reviewed changes

CSWYF3634076 mentioned this pull request Jul 1, 2025

[Usage]: vLLM Run Ernie45 Problem Collection CSWYF3634076/vllm#1

Open

zihugithub pushed a commit to zihugithub/FlagScale that referenced this pull request Jul 1, 2025

[Model] Add Ernie-4.5 (FlagOpen#637)

b0018fd

Adapted from this PR: vllm-project/vllm#20220 --------- Co-authored-by: Corle-hyz <heyongzhe18@mails.ucas.edu.cn>

mergify bot removed the needs-rebase label Jul 2, 2025

Isotr0py approved these changes Jul 2, 2025

View reviewed changes

Isotr0py enabled auto-merge (squash) July 2, 2025 02:57

Isotr0py removed request for simon-mo, youkaichao, aarnphm, houseroad, WoosukKwon, alexm-redhat and robertgshaw2-redhat July 2, 2025 06:43

Isotr0py removed performance Performance-related issues rocm Related to AMD ROCm frontend speculative-decoding ci/build v1 multi-modality Related to multi-modality (#4194) tool-calling llama Related to Llama models qwen Related to Qwen models labels Jul 2, 2025

Merge remote-tracking branch 'upstream/main' into dev

4dfe76a

mergify bot removed the needs-rebase label Jul 2, 2025

Isotr0py enabled auto-merge (squash) July 2, 2025 06:48

vllm-bot merged commit e303dcf into vllm-project:main Jul 2, 2025
61 of 66 checks passed

github-project-automation bot moved this to Done in Tool Calling Jul 2, 2025

huydhn pushed a commit to huydhn/vllm that referenced this pull request Jul 8, 2025

[Model] Add Ernie4.5 and Ernie4.5MoE Model Support (vllm-project#20220)

2a09eff

Signed-off-by: wangyafeng <wangyafeng@baidu.com>

Uh oh!

[Model] Add Ernie4.5 and Ernie4.5MoE Model Support #20220

[Model] Add Ernie4.5 and Ernie4.5MoE Model Support #20220

Uh oh!

Conversation

CSWYF3634076 commented Jun 29, 2025 • edited by github-actions bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Purpose

Uh oh!

github-actions bot commented Jun 29, 2025

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Isotr0py left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

mergify bot commented Jun 30, 2025

Uh oh!

xjpang commented Jun 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

prnake commented Jun 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

CSWYF3634076 commented Jun 30, 2025

Uh oh!

prnake commented Jun 30, 2025

Uh oh!

xjpang commented Jun 30, 2025

Uh oh!

CSWYF3634076 commented Jun 30, 2025

Uh oh!

jeejeelee left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

xjpang commented Jul 1, 2025

Uh oh!

CSWYF3634076 commented Jul 1, 2025

Uh oh!

Isotr0py left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

CSWYF3634076 commented Jul 2, 2025

Uh oh!

SongDI911 commented Jul 2, 2025

Uh oh!

CSWYF3634076 commented Jul 2, 2025

Uh oh!

SongDI911 commented Jul 2, 2025

Uh oh!

CSWYF3634076 commented Jul 3, 2025

Uh oh!

CSWYF3634076 commented Jul 3, 2025

Uh oh!

SongDI911 commented Jul 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

CSWYF3634076 commented Jun 29, 2025 •

edited by github-actions bot

Loading

xjpang commented Jun 30, 2025 •

edited

Loading

prnake commented Jun 30, 2025 •

edited

Loading

SongDI911 commented Jul 3, 2025 •

edited

Loading