Skip to content

docs: add model picker gif - #36

Closed
sublimode wants to merge 1 commit into
ltmoerdani:mainfrom
sublimode:main
Closed

docs: add model picker gif#36
sublimode wants to merge 1 commit into
ltmoerdani:mainfrom
sublimode:main

Conversation

@sublimode

Copy link
Copy Markdown
Contributor

Closes #28

@sublimode sublimode closed this Jun 13, 2026
@sublimode sublimode reopened this Jun 13, 2026
@sublimode sublimode closed this Jun 13, 2026
ltmoerdani added a commit that referenced this pull request Jul 23, 2026
…#36)

Root cause was two distinct problems:
1. Go gateway bug #37635 — ALL streaming output goes into
   reasoning_content instead of content. When MiMo thinking is OFF,
   the answer text gets swallowed by the thinking panel.
2. Model-level reasoning loop — MiMo 2.5 stuck in self-referential
   CoT, generating same fragments indefinitely.

Three-layer fix applied:
- budget_tokens cap per effort level (8K/16K/32K) in thinking.ts
- treatReasoningAsContent workaround: ONLY activates when Go gateway
  AND reasoning_effort is NOT in body (MiMo thinking OFF). When
  thinking IS on, reasoning_content is genuine CoT → thinking panel.
- Suffix-repetition detection in handleReasoning(): same 40-char
  suffix 6+ times → suppress thinking parts + emit warning.

Key regression fixes in this commit:
- Removed contentAfterReasoning guard (DeepSeek/GLM/Kimi legitimately
  use reasoning_content then content — this is normal, not degradation)
- Removed shouldSuppressTextEmit (false-positived on all reasoning
  models, not just MiMo)
- Reverted extractStreamParts delta block to main behavior

Upstream issue #37635 still open (assigned MrMushrooooom).
Workaround can be removed when gateway bug is fixed server-side.
ltmoerdani added a commit that referenced this pull request Jul 23, 2026
fix(mimo): MiMo 2.5 thinking loop fix + Go gateway workaround (#36)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Add screenshot or GIF of the model picker to README

1 participant