Glm blog (#69)

Signed-off-by: zRzRzRzRzRzRzR <[email protected]>
Signed-off-by: youkaichao <[email protected]>
Co-authored-by: zRzRzRzRzRzRzR <[email protected]>

Author: youkaichao

_posts/2025-08-18-glm45-vllm.md

@@ -0,0 +1,111 @@
+---
+layout: post
+title: "GLM-4.5 Meets vLLM: Built for Intelligent Agents"
+author: "Yuxuan Zhang"
+image: /assets/logos/vllm-logo-text-light.png
+---
+
+## Introduction
+
+[General Language Model (GLM)](https://aclanthology.org/2022.acl-long.26/) is a family of foundation models created by Zhipu.ai (now renamed to [Z.ai](https://z.ai/)). The GLM team has long-term collaboration with vLLM team, dating back to the early days of vLLM and the popular [ChatGLM model series](https://github.com/zai-org/ChatGLM-6B). Recently, the GLM team released the [GLM-4.5](https://arxiv.org/abs/2508.06471) and [GLM-4.5V](https://arxiv.org/abs/2507.01006) model series, which are designed for intelligent agents. They are the top trending models in Huggingface model hub right now.
+
+GLM-4.5 has 355 billion total
+parameters with 32 billion active parameters, while GLM-4.5-Air adopts a more compact design with 106 billion total
+parameters and 12 billion active parameters. GLM-4.5 models unify reasoning, coding, and intelligent agent capabilities
+to meet the complex demands of intelligent agent applications.
+
+Both GLM-4.5 and GLM-4.5-Air are hybrid reasoning models that provide two modes: thinking mode for complex reasoning and
+tool usage, and non-thinking mode for immediate responses.
+
+As demonstrated in our comprehensive evaluation across 12 industry-standard benchmarks, GLM-4.5 achieves exceptional
+performance with a score of 63.2, in the 3rd place among all the proprietary and open-source models. Notably,
+GLM-4.5-Air delivers competitive results at 59.8 while maintaining superior efficiency.
+
+![bench_45](https://raw.githubusercontent.com/zai-org/GLM-4.5/refs/heads/main/resources/bench.png)
+
+GLM-4.5V is based on GLM-4.5-Air. It continues the technical approach of GLM-4.1V-Thinking, achieving SOTA performance
+among models of the same scale on 42 public vision-language benchmarks.
+
+![bench_45v](https://raw.githubusercontent.com/zai-org/GLM-V/refs/heads/main/resources/bench_45v.jpeg)
+
+To get more information about GLM-4.5 and GLM-4.5V, please refer to [GLM-4.5](https://github.com/zai-org/GLM-4.5)
+and [GLM-V](https://github.com/zai-org/GLM-V).
+
+This blog will guide users on how to use vLLM to accelerate inference for the GLM-4.5V and GLM-4.5 model series on
+NVIDIA Blackwell and Hopper GPUs.
+
+## Installation
+
+In the latest vLLM main branch, both the GLM-4.5V and GLM-4.5 model series are supported.
+You can install the nightly version and manually update transformers to enable model support.
+
+```shell
+pip install -U vllm --pre --extra-index-url https://wheels.vllm.ai/nightly
+pip install transformers-v4.55.0-GLM-4.5V-preview
+```
+
+## Usage
+
+GLM-4.5 and GLM-4.5V both offer FP8 and BF16 precision models.
+In vLLM, you can use the same command to run inference for either precision.
+
+For the GLM-4.5 model, you can start the service with the following command:
+
+```shell
+vllm serve zai-org/GLM-4.5-Air \
+    --tensor-parallel-size 4 \
+    --tool-call-parser glm45 \
+    --reasoning-parser glm45 \
+    --enable-auto-tool-choice
+```
+
+For the GLM-4.5V model, you can start the service with the following command:
+
+```shell
+vllm serve zai-org/GLM-4.5V \
+     --tensor-parallel-size 4   \
+     --tool-call-parser glm45   \
+     --reasoning-parser glm45   \
+     --enable-auto-tool-choice  \
+     --allowed-local-media-path / \
+     --media-io-kwargs '{"video": {"num_frames": -1}}'
+```
+
+### Important Notes
+
++ The reasoning part of the model output will be wrapped in `reasoning_content`. `content` will only contain the final
+  answer. To disable reasoning, add the following parameter:
+  `extra_body={"chat_template_kwargs": {"enable_thinking": False}}`
++ If you're using 8x H100 GPUs and encounter insufficient memory when running the GLM-4.5 model, you'll need
+  `--cpu-offload-gb 16`.
++ If you encounter `flash_infer` issues, use `VLLM_ATTENTION_BACKEND=XFORMERS` as a temporary replacement. You can also
+  specify `TORCH_CUDA_ARCH_LIST='9.0+PTX'` to use `flash_infer`, different GPUs have different TORCH_CUDA_ARCH_LIST
+  values, please check accordingly.
++ vLLM v0 is not support our model.
+
+### Grounding in GLM-4.5V
+
+GLM-4.5V equips precise grounding capabilities. Given a prompt that requests the location of a specific object, GLM-4.5V
+is able to reasoning step-by-step and identify the bounding boxes of the target object. The query prompt supports
+complex descriptions of the target object as well as specified output formats. Example prompts are:
+
+- Help me to locate `<expr>` in the image and give me its bounding boxes.
+- Please pinpoint the bounding box `[[x1,y1,x2,y2], …]` in the image as per the given description. <expr>
+
+Here, `<expr>` is the description of the target object. The output bounding box is a quadruple $$[x_1,y_1,x_2,y_2]$$
+composed of the coordinates of the top-left and bottom-right corners, where each value is normalized by the image
+width (for x) or height (for y) and scaled by 1000.
+
+In the response, the special tokens `<|begin_of_box|>` and `<|end_of_box|>` are used to mark the image bounding box in
+the answer. The bracket style may vary ([], [[]], (), <>, etc.), but the meaning is the same: to enclose the coordinates
+of the box.
+
+## Cooperation with vLLM and GLM Team
+
+Before the release of the GLM-4.5 and GLM-4.5V models, the vLLM team worked closely with the GLM team, providing
+extensive support in addressing issues related to the model launch, ensuring that the vLLM `main` branch had full
+support for the open-source GLM-4.5 series before the models were released.
+
+## Acknowledgement
+
+We would like to thank many people from the vLLM side who contributed to this effort, including: Kaichao You, Simon Mo, Zifeng Mo, Lucia Fang, Rui Qiao, Jie Le, Ce Gao, Roger Wang, Lu Fang, Wentao Ye, and Zixi Qi.