Conversation
…efaults) Move preserve_all_thinking and preserve_thinking_between_tool_calls from ``render`` / ``render_ids`` call-site kwargs to constructor kwargs on every renderer. Each renderer stores them as ``_preserve_all_thinking`` / ``_preserve_thinking_between_tool_calls`` instance attributes and consults them internally via ``should_preserve_past_thinking``. This kills the method-monkey-patching ``_bind_preserve_defaults`` helper introduced in #4: with construction-time configuration there is nothing left to bind after the fact. The Renderer Protocol's ``render`` / ``render_ids`` signatures shrink back to their pre-#3 shape. Why: in real usage downstream of this package (verifiers ``RendererClient``, prime-rl ``RendererConfig``) the flags are config-time, not per-render. Call-site kwargs were over-engineered for that shape — and the wrapper that made them work was a method-rebind trick that hid the real method signature from type-checkers and IDE tooling. "Renderer = configuration" is cleaner: build a different renderer (or pool) when you want a different configuration. Changes: - ``renderers/base.py``: drop ``_bind_preserve_defaults``; ``create_renderer`` / ``create_renderer_pool`` forward the flags directly to the constructor. Protocol's ``render`` / ``render_ids`` no longer expose the flags. - Every concrete renderer (``glm5``, ``glm45``, ``qwen3``, ``qwen35``, ``qwen36``, ``qwen3_vl``, ``minimax_m2``, ``nemotron3``, ``kimi_k2``, ``kimi_k25``, ``gpt_oss``, ``deepseek_v3``, ``default``): accept the flags in ``__init__``, store as attrs, drop them from render signatures. - ``DefaultRenderer`` raises NotImplementedError at construction now (fail fast) rather than at first render call. - README §1: new "Preserving past reasoning (constructor-only overrides)" subsection, with a worked compaction example motivating why one would flip ``preserve_all_thinking=True`` (a user-turn summariser would otherwise lose all of the trajectory's reasoning). - README §3.d: cross-link to the constructor-only override. - Tests: rewrite to build per-flag renderers (no call-site kwargs). - Bump 0.1.8 → 0.1.9. 893 passed, 48 skipped, 1 xfailed locally — no parity regressions. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Pre-commit auto-applied line wrapping on the long ``preserve_thinking_between_tool_calls`` attribute assignments. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
3 tasks
hallerite
added a commit
that referenced
this pull request
May 7, 2026
PyPI's latest published renderers is 0.1.5 — the in-tree bumps in #3 → #5 → this PR (0.1.6 → 0.1.7 → 0.1.8 → 0.1.9 → 0.1.10) accumulated against PyPI without ever publishing. Roll forward to 0.1.6 once for the actual release; tag ``renderers-v0.1.6`` post-merge to fire the publish workflow (.github/workflows/publish.yml asserts the tag's ``-vX.Y.Z`` suffix matches pyproject.toml). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
hallerite
added a commit
that referenced
this pull request
May 7, 2026
…preserve_thinking) (#6) * feat: drop legacy chat-template-pass-through kwargs ``GLM5Renderer.__init__`` no longer accepts ``clear_thinking``, and ``Qwen36Renderer.__init__`` no longer accepts ``preserve_thinking``. Both were chat-template-kwarg pass-throughs whose default values are now baked straight into the renderer: - GLM-5 default: drop ``<think>`` on past-cycle assistants, keep on in-flight cycle and trailing assistant. (What ``clear_thinking=True`` did before.) The render gate is now ``msg_idx > last_user_index or preserve_thinking`` — clean and consistent with the rest of the family. - Qwen3.6 default: identical to Qwen3.5 (``msg_idx > last_query_index`` via the inherited ``_should_render_thinking``). The class shrinks to just the ``_render_arg_value`` override, which is the only real template delta vs Qwen3.5. Callers who want the toggled-on behaviour pass ``preserve_all_thinking=True`` to ``create_renderer`` — the renderer-agnostic spelling of the same intent landed in #3/#5. Two regression tests pin the dropped kwargs (constructor must raise ``TypeError`` when either is passed). The README's §1 "Preserving past reasoning" subsection (added in #5) already motivates the override flags; no new prose needed here. §3.d is unchanged (still cross-links to the override flag for compaction). Bump 0.1.9 → 0.1.10. 895 tests pass locally. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> * chore: consolidate version to 0.1.6 (PyPI-aligned) PyPI's latest published renderers is 0.1.5 — the in-tree bumps in #3 → #5 → this PR (0.1.6 → 0.1.7 → 0.1.8 → 0.1.9 → 0.1.10) accumulated against PyPI without ever publishing. Roll forward to 0.1.6 once for the actual release; tag ``renderers-v0.1.6`` post-merge to fire the publish workflow (.github/workflows/publish.yml asserts the tag's ``-vX.Y.Z`` suffix matches pyproject.toml). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Moves `preserve_all_thinking` / `preserve_thinking_between_tool_calls` from `render` / `render_ids` call-site kwargs to constructor-only kwargs on every renderer. Each renderer stores them as `_preserve_all_thinking` / `_preserve_thinking_between_tool_calls` instance attributes and consults them internally via `should_preserve_past_thinking`.
Kills the method-monkey-patching `_bind_preserve_defaults` helper from #4: with construction-time configuration there is nothing left to bind after the fact. `Renderer.render` / `render_ids` signatures shrink back to their pre-#3 shape.
Why
In real downstream usage (`verifiers.RendererClient`, prime-rl's `RendererConfig`), the flags are config-time, not per-render — bound once at pool creation, never changed per render. Call-site kwargs were over-engineered for that shape, and the wrapper that made them work hid the real method signature from type-checkers and IDE tooling.
"Renderer = configuration" is cleaner: build a different renderer (or pool) when you want a different configuration.
What changed
Test plan
🤖 Generated with Claude Code