unsloth/studio/backend/core
Daniel Han b60b0740c2 Tighten comments across PR 5711 (no behaviour change)
Comments-only pass. Drops verbose docstrings to single-line form,
removes repetitive "null = unset" / "Local only" tails (already
encoded by the type signature and capability map), keeps every
authoritative source URL but cuts surrounding prose, and removes
fully-redundant per-field comments where the field name already
says what the comment says.

Touches:
  - types/runtime.ts (InferenceParams)
  - types/api.ts (OpenAIChatCompletionsRequest wire shape)
  - provider-capabilities.ts (ProviderCapabilities interface + bucket
    inline blocks + per-model resolvers + reasoning helpers)
  - api/chat-adapter.ts (external + local forwarding stanzas)
  - backend models/inference.py (Field descriptions)
  - backend llama_cpp.py (3rd payload-builder inline comments)
  - backend routes/inference.py (_build_passthrough_payload)
  - backend external_provider.py (4.7 sampling-removed header,
    _is_openai_family_cloud docstring)

Net 508 lines deleted across 8 files; 65/65 sampling_params_routing
tests still pass; frontend tsc clean.
2026-05-27 11:14:45 +00:00
..
data_recipe Recipe Studio local model selector (#5769) 2026-05-26 02:37:24 -07:00
export fix: repair mlx studio base export save_method (#5727) 2026-05-25 04:04:07 -07:00
inference Tighten comments across PR 5711 (no behaviour change) 2026-05-27 11:14:45 +00:00
training studio: install flash-linear-attention and tilelang for Qwen3.5 family (#5434) 2026-05-18 03:49:06 -07:00
__init__.py [Studio] Show non exported models in chat UI (#4892) 2026-04-14 15:03:58 +04:00
tool_healing.py studio: extract tool-call XML parser into a reusable helper module (#5583) 2026-05-19 05:06:17 -07:00