---
cited_by:
  - '[[thoughts/inference/qwen3.8-runtime#2-delta-vs-predecessors|qwen4next §2]]'
  - '[[thoughts/inference/qwen3.8-serving#4-pr-reading-order-vllm-main-git-show-sha|qwen4next §4]]'
date: '2026-11-11'
kind: pr
local:
  - file:///Users/aarnphm/workspace/vllm/vllm/model_executor/models/qwen3_5.py
merged: '2026-08-07'
number: 50068
source:
  - https://github.com/vllm-project/vllm/blob/f2bfad9167a29e963f29f9ea79f2811513566ea6/vllm/model_executor/models/qwen3_5.py
state: merged
tags:
  - vllm
  - vllm-pr
  - new-architectures
title: '#50068 [Model] Enable Qwen3.8 for AMD Rocm'
topic: new-architectures
url: https://github.com/vllm-project/vllm/pull/50068
created: '2026-11-11'
modified: '2026-11-11'
published: '2026-11-11'
pageLayout: default
slug: hinterland/papers/maps/prs/50068
permalink: https://aarnphm.xyz/hinterland/papers/maps/prs/50068.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
register the text-only `Qwen3_5ForCausalLM` and `Qwen3_5MoeForCausalLM` architectures · full dump: [[thoughts/inference/vllm-prs/50068|#50068]]

