[PR #15159] [CLOSED] integration: improve vision test robustness and add thinking tests #20326

Closed
opened 2026-04-16 07:33:18 -05:00 by GiteaMirror · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/15159
Author: @dhiltgen
Created: 3/31/2026
Status: Closed

Base: mainHead: int_vision


📝 Commits (1)

  • 69a27d5 integration: improve vision test robustness and add thinking tests

📊 Changes

12 files changed (+1213 additions, -21 deletions)

View changed files

📝 integration/context_test.go (+2 -0)
📝 integration/llm_image_test.go (+3 -4)
integration/thinking_test.go (+155 -0)
📝 integration/vision_test.go (+53 -17)
📝 integration/vision_test_data_test.go (+159 -0)
video/video.go (+73 -0)
video/video_darwin.go (+118 -0)
video/video_darwin.m (+154 -0)
video/video_darwin_test.go (+48 -0)
video/video_linux.go (+162 -0)
video/video_test.go (+131 -0)
video/video_windows.go (+155 -0)

📄 Description

Add skipIfNoVisionOverride() to skip vision tests when OLLAMA_TEST_MODEL is set to a non-vision model. Add Think:false to context exhaustion test to prevent thinking models from using all context before the test can measure it. Add third test image (ollama homepage) and replace OCR test with ImageDescription test using it. Relax match strings for broader model compatibility. Add TestThinkingEnabled and TestThinkingSuppressed to verify thinking output and channel tag handling.


🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/ollama/ollama/pull/15159 **Author:** [@dhiltgen](https://github.com/dhiltgen) **Created:** 3/31/2026 **Status:** ❌ Closed **Base:** `main` ← **Head:** `int_vision` --- ### 📝 Commits (1) - [`69a27d5`](https://github.com/ollama/ollama/commit/69a27d59ef17a24f22f7077af9a516cc7e5eb9af) integration: improve vision test robustness and add thinking tests ### 📊 Changes **12 files changed** (+1213 additions, -21 deletions) <details> <summary>View changed files</summary> 📝 `integration/context_test.go` (+2 -0) 📝 `integration/llm_image_test.go` (+3 -4) ➕ `integration/thinking_test.go` (+155 -0) 📝 `integration/vision_test.go` (+53 -17) 📝 `integration/vision_test_data_test.go` (+159 -0) ➕ `video/video.go` (+73 -0) ➕ `video/video_darwin.go` (+118 -0) ➕ `video/video_darwin.m` (+154 -0) ➕ `video/video_darwin_test.go` (+48 -0) ➕ `video/video_linux.go` (+162 -0) ➕ `video/video_test.go` (+131 -0) ➕ `video/video_windows.go` (+155 -0) </details> ### 📄 Description Add skipIfNoVisionOverride() to skip vision tests when OLLAMA_TEST_MODEL is set to a non-vision model. Add Think:false to context exhaustion test to prevent thinking models from using all context before the test can measure it. Add third test image (ollama homepage) and replace OCR test with ImageDescription test using it. Relax match strings for broader model compatibility. Add TestThinkingEnabled and TestThinkingSuppressed to verify thinking output and channel tag handling. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
GiteaMirror added the pull-request label 2026-04-16 07:33:18 -05:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#20326