chore(container): bump tensorrt-llm pin from 1.3.0rc11 to 1.3.0rc13 by tanmayv25 · Pull Request #9126 · ai-dynamo/dynamo

tanmayv25 · 2026-05-04T23:59:03Z

This is the original PR: #8870

Summary by CodeRabbit

Chores
- Updated TensorRT-LLM dependency from version 1.3.0rc11 to 1.3.0rc13.
Documentation
- Updated support matrix to reflect the new TensorRT-LLM version compatibility.

The current pip_wheel pin in container/context.yaml is tensorrt-llm==1.3.0rc11 (tagged 2026-04-07). The C++ admission-overflow fix for issue NVIDIA/TensorRT-LLM#13318 (PR NVIDIA/TensorRT-LLM#12976, "Fix compute token accounting for KV cache reuse with context chunking") merged to TRT-LLM main on 2026-04-18. Tag-by-tag status, verified by checking the cherry-picked microBatchScheduler.cpp on each tag tree: - v1.3.0rc11 (2026-04-07, 4e69c14f7): reuse_adjusted_compute count = 0 -- does NOT include the admission fix. - v1.3.0rc12 (2026-04-16, 61cef212a): reuse_adjusted_compute count = 0 -- does NOT include the admission fix (rc12 was cut 2 days before PR #12976 merged). - v1.3.0rc13 (2026-04-26, b9ce4b69d): reuse_adjusted_compute count = 10 (1 static helper + 9 call sites) -- this is the first published rc tag with the admission fix. Bumping rc11 -> rc13 directly (skipping rc12) gives users of dynamo's default container path (has_trtllm_context: "0") the upstream fix for issue NVIDIA/TensorRT-LLM#13318 without forcing them onto a custom-built wheel. Files updated: - container/context.yaml: pip_wheel and github_trtllm_commit - container/deps/requirements.common.txt: comment that documents the transformers version constraint per backend (rc11 -> rc13; same ==4.57.3 constraint applies, no transformers pin change needed) - docs/reference/support-matrix.md: "main (ToT)" row TRT-LLM column (the historical "v1.1.0-dev.3" row stays at rc11 because that release shipped against rc11 -- that is a fact about the past) docs/reference/release-artifacts.md is intentionally NOT changed: every rc11 reference in that file is in the v1.1.0-dev.3 release-history section (lines 169 / 587 / 594), which describes what that past release shipped. Those are historical facts. Verified locally: - ARM64 (GB200) container build with this change pulls tensorrt-llm==1.3.0rc13 from pypi.nvidia.com without conflict. - Static cross-check: libtensorrt_llm.so in the resulting venv has the reuse_adjusted_compute static helper compiled in (rc13 source tree has 10 occurrences; the helper is optimizer-inlined so strings-grep returns 0, but provenance is established via byte-identity of the shipped library against the rc13 wheel build). Signed-off-by: Yifan Jiang <19356972+yifjiang@users.noreply.github.com>

github-actions · 2026-05-05T00:00:59Z

🌿 Fern Docs Preview: https://nvidia-preview-a818c37a-40d3-4ec3-81ff-65640a243b6d.docs.buildwithfern.com/dynamo/dev

coderabbitai · 2026-05-05T00:01:23Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: c9f8d5d3-2f01-4437-8ab1-4ed1a1d20479

📥 Commits

Reviewing files that changed from the base of the PR and between 26645cc and 8aa083f.

📒 Files selected for processing (3)

container/context.yaml
container/deps/requirements.common.txt
docs/reference/support-matrix.md

Walkthrough

TensorRT-LLM dependencies are bumped from version 1.3.0rc11 to 1.3.0rc13 across build configuration, dependency comments, and user-facing documentation.

Changes

TensorRT-LLM Version Bump (1.3.0rc11 → 1.3.0rc13)

Layer / File(s)	Summary
Build Configuration `container/context.yaml`	TensorRT-LLM version updated in `pip_wheel` and `github_trtllm_commit` fields from 1.3.0rc11 to 1.3.0rc13.
Dependency Documentation `container/deps/requirements.common.txt`	Version pin comment for TensorRT-LLM updated from 1.3.0rc11 to 1.3.0rc13.
Support Documentation `docs/reference/support-matrix.md`	Backend Dependencies table updated to reflect TensorRT-LLM 1.3.0rc13 for main (ToT) release.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description lacks substantive detail and does not follow the provided template structure with required sections (Overview, Details, Where to start, Related Issues).	Provide a complete description following the template: add Overview section, detailed explanation of changes, reviewer guidance, and any related issue links.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and specifically describes the main change: bumping tensorrt-llm from 1.3.0rc11 to 1.3.0rc13 across container configuration files.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

Tip

💬 Introducing Slack Agent: The best way for teams to turn conversations into code.

Slack Agent is built on CodeRabbit's deep understanding of your code, so your team can collaborate across the entire SDLC without losing context.

Generate code and open pull requests
Plan features and break down work
Investigate incidents and troubleshoot customer tickets together
Automate recurring tasks and respond to alerts with triggers
Summarize progress and report instantly

Built for teams:

Shared memory across your entire org—no repeating context
Per-thread sandboxes to safely plan and execute work
Governance built-in—scoped access, auditability, and budget controls

One agent for your entire SDLC. Right inside Slack.

👉 Get started

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

TensorRT-LLM rc13 reshaped VisualGenArgs.teacache into a polymorphic cache: Optional[CacheConfig] field, and TeaCacheConfig no longer accepts enable_teacache. Pass cache=TeaCacheConfig(...) only when enabled; otherwise leave it unset so VisualGenArgs.cache defaults to None. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

indrajit96

LGTM!
From multimodality and diffusion POV pre-merge CI should catch any failures.

DiffusionRequest fields like max_sequence_length default to None and are expected to be resolved against pipeline.default_generation_params by the executor's _merge_defaults. Dynamo calls pipeline.infer() directly, so None propagated to forward() and crashed the Wan _encode_prompt path. Replicate the merge locally so any pipeline-declared defaults (and extra_param defaults) are applied before infer(). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

…line mock The diffusion engine now reads pipeline.default_generation_params, pipeline.extra_param_specs, and req.extra_params to mirror executor default-resolution. Update the unit-test mock to expose empty maps and the request fake to default extra_params=None as on the real dataclass. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

requests 2.32.x's check_compatibility rejects chardet>=6.0.0 (and the new 6.0.0.post1 also breaks the version-tuple unpack), so transitive installs of chardet 6 cause an ImportError when conftest imports requests in the deploy-test runners. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

richardhuo-nv · 2026-05-05T06:35:17Z

https://github.com/ai-dynamo/dynamo/blob/main/components/src/dynamo/trtllm/workers/llm_worker.py#L26

I think last time when I tested this version, there is a TOKENIZER_ALIASES import error, since it is being moved to tokenizer in trtllm:
https://github.com/NVIDIA/TensorRT-LLM/blob/v1.3.0rc13/tensorrt_llm/tokenizer/tokenizer.py#L14

tanmayv25 · 2026-05-05T06:55:33Z

Thanks Richard for sharing. I do still see TOKENIZER_ALIASES here:
https://github.com/NVIDIA/TensorRT-LLM/blob/v1.3.0rc13/tensorrt_llm/llmapi/llm_args.py#L583

richardhuo-nv · 2026-05-05T06:59:34Z

Thanks Richard for sharing. I do still see TOKENIZER_ALIASES here: https://github.com/NVIDIA/TensorRT-LLM/blob/v1.3.0rc13/tensorrt_llm/llmapi/llm_args.py#L583

ah, I see. I think I tested v1.3.0rc12

tanmayv25 requested a review from a team as a code owner May 4, 2026 23:59

tanmayv25 requested a review from a team May 4, 2026 23:59

tanmayv25 requested review from a team as code owners May 4, 2026 23:59

pull-request-size Bot added the size/XS label May 4, 2026

github-actions Bot added chore documentation Improvements or additions to documentation container labels May 4, 2026

dynamo-ops approved these changes May 5, 2026

View reviewed changes

pull-request-size Bot added size/S and removed size/XS labels May 5, 2026

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 00:06 Inactive

github-actions Bot added the backend::trtllm Relates to the trtllm backend label May 5, 2026

indrajit96 approved these changes May 5, 2026

View reviewed changes

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 00:12 Inactive

pull-request-size Bot added size/M and removed size/S labels May 5, 2026

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 02:51 Inactive

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 02:52 Inactive

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 04:08 Inactive

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 04:09 Inactive

nv-yna approved these changes May 5, 2026

View reviewed changes

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 06:25 Inactive

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 06:31 Inactive

richardhuo-nv approved these changes May 5, 2026

View reviewed changes

tanmayv25 enabled auto-merge (squash) May 5, 2026 07:31

Merge branch 'main' into tanmayv-trtllm-upgrade

159cf37

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 11:58 Inactive

copy-pr-bot Bot temporarily deployed to GITLAB May 5, 2026 12:05 Inactive

tanmayv25 merged commit f05b442 into main May 5, 2026
183 of 202 checks passed

tanmayv25 deleted the tanmayv-trtllm-upgrade branch May 5, 2026 18:59

tanmayv25 mentioned this pull request May 5, 2026

chore(container): bump tensorrt-llm pin from 1.3.0rc11 to 1.3.0rc13 (… #9182

Merged

Navigation Menu

Uh oh!

chore(container): bump tensorrt-llm pin from 1.3.0rc11 to 1.3.0rc13 - #9126

chore(container): bump tensorrt-llm pin from 1.3.0rc11 to 1.3.0rc13#9126
tanmayv25 merged 6 commits into
mainfrom
tanmayv-trtllm-upgrade

tanmayv25 commented May 4, 2026 •

edited by coderabbitai Bot

Loading

Uh oh!

github-actions Bot commented May 5, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented May 5, 2026

❌ Failed checks (1 warning)

Uh oh!

indrajit96 left a comment

Uh oh!

richardhuo-nv commented May 5, 2026

Uh oh!

tanmayv25 commented May 5, 2026

Uh oh!

richardhuo-nv commented May 5, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Uh oh!

Conversation

tanmayv25 commented May 4, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Uh oh!

github-actions Bot commented May 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

coderabbitai Bot commented May 5, 2026

Walkthrough

Changes

Estimated code review effort

❌ Failed checks (1 warning)

Uh oh!

indrajit96 left a comment

Choose a reason for hiding this comment

Uh oh!

richardhuo-nv commented May 5, 2026

Uh oh!

tanmayv25 commented May 5, 2026

Uh oh!

richardhuo-nv commented May 5, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

tanmayv25 commented May 4, 2026 •

edited by coderabbitai Bot

Loading

github-actions Bot commented May 5, 2026 •

edited

Loading