Issue Station
Sign In
Tools hereToolbox

ollama - 💡(How to fix) Fix Intel GPU Arc 770 16G, gemma4:e2b gemma4:e4b, OLLAMA_VULKAN=1, Output gibberish. [9 comments, 7 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
ollama/ollama#15328•Fetched 2026-04-08 02:44:05
View on GitHub
Comments
9
Participants
7
Timeline
18
Reactions
6
Author
wqmengwqmeng
Participants
asmerasmerhidp123hidp123N3RDIUMN3RDIUMrjmalagonrjmalagonTheEarthCMSTheEarthCMSwqmengwqmengyaizawa2eyaizawa2e
Timeline (top)
commented ×9subscribed ×8labeled ×1

Code Example

ollama serve
time=2026-04-04T22:15:38.820+08:00 level=INFO source=routes.go:1744 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GGML_VK_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:0 OLLAMA_DEBUG:INFO OLLAMA_DEBUG_LOG_REQUESTS:false OLLAMA_EDITOR: OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:e:\\OllamaModels OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NO_CLOUD:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[chrome-extension://* moz-extension://* http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_REMOTES:[ollama.com] OLLAMA_SCHED_SPREAD:false OLLAMA_VULKAN:true ROCR_VISIBLE_DEVICES:]"
time=2026-04-04T22:15:38.825+08:00 level=INFO source=routes.go:1746 msg="Ollama cloud disabled: false"
time=2026-04-04T22:15:38.827+08:00 level=INFO source=images.go:499 msg="total blobs: 24"
time=2026-04-04T22:15:38.828+08:00 level=INFO source=images.go:506 msg="total unused blobs removed: 0"
time=2026-04-04T22:15:38.828+08:00 level=INFO source=routes.go:1802 msg="Listening on [::]:11434 (version 0.20.0)"
time=2026-04-04T22:15:38.829+08:00 level=INFO source=runner.go:67 msg="discovering available GPUs..."
time=2026-04-04T22:15:38.842+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10575"
time=2026-04-04T22:15:38.928+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10581"
time=2026-04-04T22:15:39.068+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10587"
time=2026-04-04T22:15:39.161+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10596"
time=2026-04-04T22:15:39.567+08:00 level=INFO source=types.go:42 msg="inference compute" id=8680a056-0800-0000-0300-000000000000 filter_id="" library=Vulkan compute=0.0 name=Vulkan0 description="Intel(R) Arc(TM) A770 Graphics" libdirs=ollama,vulkan driver=0.0 pci_id="" type=discrete total="15.9 GiB" available="14.2 GiB"
time=2026-04-04T22:15:39.567+08:00 level=INFO source=routes.go:1852 msg="vram-based default context" total_vram="15.9 GiB" default_num_ctx=4096
RAW_BUFFERClick to expand / collapse

What is the issue?

Intel GPU Arc 770 16G, gemma4:e2b gemma4:e4b, OLLAMA_VULKAN=1, Output gibberish.

When disable OLLAMA_VULKAN=1, The CPU only gemma4 output correctly.

OS

windows 11

GPU

Intel Arc 770 16Gb

CPU

Intel i5-13400F

Ollama version

0.20.0

Relevant log output


ollama serve
time=2026-04-04T22:15:38.820+08:00 level=INFO source=routes.go:1744 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GGML_VK_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:0 OLLAMA_DEBUG:INFO OLLAMA_DEBUG_LOG_REQUESTS:false OLLAMA_EDITOR: OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:e:\\OllamaModels OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NO_CLOUD:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[chrome-extension://* moz-extension://* http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_REMOTES:[ollama.com] OLLAMA_SCHED_SPREAD:false OLLAMA_VULKAN:true ROCR_VISIBLE_DEVICES:]"
time=2026-04-04T22:15:38.825+08:00 level=INFO source=routes.go:1746 msg="Ollama cloud disabled: false"
time=2026-04-04T22:15:38.827+08:00 level=INFO source=images.go:499 msg="total blobs: 24"
time=2026-04-04T22:15:38.828+08:00 level=INFO source=images.go:506 msg="total unused blobs removed: 0"
time=2026-04-04T22:15:38.828+08:00 level=INFO source=routes.go:1802 msg="Listening on [::]:11434 (version 0.20.0)"
time=2026-04-04T22:15:38.829+08:00 level=INFO source=runner.go:67 msg="discovering available GPUs..."
time=2026-04-04T22:15:38.842+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10575"
time=2026-04-04T22:15:38.928+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10581"
time=2026-04-04T22:15:39.068+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10587"
time=2026-04-04T22:15:39.161+08:00 level=INFO source=server.go:432 msg="starting runner" cmd="C:\\Users\\1\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --ollama-engine --port 10596"
time=2026-04-04T22:15:39.567+08:00 level=INFO source=types.go:42 msg="inference compute" id=8680a056-0800-0000-0300-000000000000 filter_id="" library=Vulkan compute=0.0 name=Vulkan0 description="Intel(R) Arc(TM) A770 Graphics" libdirs=ollama,vulkan driver=0.0 pci_id="" type=discrete total="15.9 GiB" available="14.2 GiB"
time=2026-04-04T22:15:39.567+08:00 level=INFO source=routes.go:1852 msg="vram-based default context" total_vram="15.9 GiB" default_num_ctx=4096

extent analysis

TL;DR

Disable the OLLAMA_VULKAN environment variable to potentially resolve the issue with gibberish output on the Intel GPU Arc 770.

Guidance

  • The issue seems to be related to the OLLAMA_VULKAN environment variable, as disabling it resolves the problem.
  • Verify that the GPU is properly configured and supported by the Ollama version 0.20.0.
  • Check the log output for any errors or warnings related to the Vulkan setup or GPU initialization.
  • Consider updating the Ollama version or seeking support from the developers if the issue persists.

Example

No code snippet is provided as the issue seems to be related to environment variables and configuration.

Notes

The provided log output does not show any explicit errors, but the fact that disabling OLLAMA_VULKAN resolves the issue suggests a potential problem with the Vulkan setup or compatibility.

Recommendation

Apply workaround: Disable the OLLAMA_VULKAN environment variable to resolve the issue, as it seems to be the most straightforward solution based on the provided information.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#LLM response#prompt template#agent execution#callback error#memory management

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING