[GH-ISSUE #15632] [0.21.0-rc0] qwen3.6:35b-a3b-nvfp4 fails to load: layer 0 missing linear attention projections #56484

Closed
opened 2026-04-29 10:53:31 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @danieltomasz on GitHub (Apr 16, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/15632

What is the issue?

Title: qwen3.6:35b-a3b-nvfp4 fails to load: layer 0 missing linear attention projections

What is the issue?

Running the newly released qwen3.6:35b-a3b-nvfp4 fails immediately with an MLX runner error about missing linear attention projections on layer 0. The model pulls successfully but cannot be loaded.

Command:

ollama run qwen3.6:35b-a3b-nvfp4

Error:

Error: 500 Internal Server Error: mlx runner failed: Error: layer 0: missing linear attention projections (exit: exit status 1)

Environment:

  • Ollama version: 0.21.0-rc0
  • OS: macOS 26.4.1
  • Hardware: Apple Silicon

Happy to test other variants (mxfp8, mlx-bf16) or provide additional logs if helpful.

Relevant log output


OS

No response

GPU

No response

CPU

No response

Ollama version

No response

Originally created by @danieltomasz on GitHub (Apr 16, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/15632 ### What is the issue? **Title:** qwen3.6:35b-a3b-nvfp4 fails to load: layer 0 missing linear attention projections **What is the issue?** Running the newly released `qwen3.6:35b-a3b-nvfp4` fails immediately with an MLX runner error about missing linear attention projections on layer 0. The model pulls successfully but cannot be loaded. **Command:** ``` ollama run qwen3.6:35b-a3b-nvfp4 ``` **Error:** ``` Error: 500 Internal Server Error: mlx runner failed: Error: layer 0: missing linear attention projections (exit: exit status 1) ``` **Environment:** - Ollama version: 0.21.0-rc0 - OS: macOS 26.4.1 - Hardware: Apple Silicon Happy to test other variants (`mxfp8`, `mlx-bf16`) or provide additional logs if helpful. ### Relevant log output ```shell ``` ### OS _No response_ ### GPU _No response_ ### CPU _No response_ ### Ollama version _No response_
GiteaMirror added the bug label 2026-04-29 10:53:31 -05:00
Author
Owner

@PureBlissAK commented on GitHub (Apr 18, 2026):

🤖 Automated Triage & Analysis Report

Issue: #15632
Analyzed: 2026-04-18T18:19:20.984810

Analysis

  • Type: unknown
  • Severity: medium
  • Components: unknown

Implementation Plan

  • Effort: medium
  • Steps:

This issue has been triaged and marked for implementation.

<!-- gh-comment-id:4274304730 --> @PureBlissAK commented on GitHub (Apr 18, 2026): <!-- ollama-issue-orchestrator:v1 issue:15632 --> ## 🤖 Automated Triage & Analysis Report **Issue**: #15632 **Analyzed**: 2026-04-18T18:19:20.984810 ### Analysis - **Type**: unknown - **Severity**: medium - **Components**: unknown ### Implementation Plan - **Effort**: medium - **Steps**: *This issue has been triaged and marked for implementation.*
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#56484