[docs] add STACK.md + CO-AUTHORS.md trackers for the will/ltx2_sr_port stack
* STACK.md tracks the 10-PR split of will/ltx2_sr_port: per-PR slice ranges, tip SHAs (live, not hardcoded), independence map, landing strategy, fast-track candidates, and re-slice commands to run after any rebase or trailer injection. * CO-AUTHORS.md documents the 4 GitHub users credited as co-authors via Co-authored-by trailers on every commit in the stack: - @Davids048 (Junda Su) - @RandNMR73 (Matthew Noto) - @XOR-op - @jzhang38 (Zhang Peiyuan) All 4 verified active in FastVideo-internal git history. Trailer emails use GitHub's <id>+<username>@users.noreply.github.com form for reliable account linkage. Both files are temporary trackers (delete after the full stack merges or as documented at the bottom of each file). Co-authored-by: Junda (David) Su <90978028+Davids048@users.noreply.github.com> Co-authored-by: Matthew Noto <99706358+RandNMR73@users.noreply.github.com> Co-authored-by: XOR-op <17672363+XOR-op@users.noreply.github.com> Co-authored-by: Zhang Peiyuan <42993249+jzhang38@users.noreply.github.com>
This commit is contained in:
co-authored by
Junda Su
Matthew Noto
XOR-op
Zhang Peiyuan
parent
67fdb4f438
commit
4c5144163c
@@ -0,0 +1,88 @@
|
||||
# Co-Authors — `will/ltx2_sr_port` Stack
|
||||
|
||||
**Status:** PERMANENT — keep around as the source of truth for who collaborated on this work, even after the stack merges.
|
||||
**Last updated:** 2026-05-04
|
||||
|
||||
This file documents the human co-authors credited on every commit in the
|
||||
`will/ltx2_sr_port` stack and its 10 split PRs. The 4 collaborators below
|
||||
worked on the FastVideo-internal precursor of this code (LTX-2 streaming
|
||||
server, NVFP4 wire-up, GPU pool, prompt enhancer, etc.) and are credited as
|
||||
co-authors on the public-side upstream commits via Git's standard
|
||||
[`Co-authored-by`](https://docs.github.com/en/pull-requests/committing-changes-to-your-project/creating-and-editing-commits/creating-a-commit-with-multiple-authors)
|
||||
trailer convention.
|
||||
|
||||
The trailers are added to every commit on `will/ltx2_sr_port` (see
|
||||
[`STACK.md`](STACK.md)), which means GitHub will:
|
||||
|
||||
- Show the 4 co-authors on every commit detail page
|
||||
- Show them on the merge commit / squash commit summary
|
||||
- Display their avatars in the PR's "Contributors" sidebar
|
||||
- Surface them in [`/contributors`](https://github.com/hao-ai-lab/FastVideo/contributors) once the stack lands
|
||||
|
||||
## Co-author roster
|
||||
|
||||
| GitHub user | Real name | GitHub ID | Trailer email |
|
||||
|---|---|---|---|
|
||||
| [`@Davids048`](https://github.com/Davids048) | Junda (David) Su | 90978028 | `90978028+Davids048@users.noreply.github.com` |
|
||||
| [`@RandNMR73`](https://github.com/RandNMR73) | Matthew Noto | 99706358 | `99706358+RandNMR73@users.noreply.github.com` |
|
||||
| [`@XOR-op`](https://github.com/XOR-op) | (unset) | 17672363 | `17672363+XOR-op@users.noreply.github.com` |
|
||||
| [`@jzhang38`](https://github.com/jzhang38) | Zhang Peiyuan | 42993249 | `42993249+jzhang38@users.noreply.github.com` |
|
||||
|
||||
## Why no-reply emails
|
||||
|
||||
GitHub's `<id>+<username>@users.noreply.github.com` form is the most reliable
|
||||
way to link a `Co-authored-by` trailer to a GitHub account. It:
|
||||
|
||||
- Always works regardless of whether the user has a public verified email
|
||||
- Survives the user changing their primary email
|
||||
- Doesn't expose anyone's personal email to git history
|
||||
- Is the format GitHub itself produces when you click "Add co-author" in the
|
||||
web UI
|
||||
|
||||
(All 4 collaborators have this email already used in `FastVideo-internal`
|
||||
git history, verified via `git log --all` on that repo.)
|
||||
|
||||
## Trailer block (copy-paste ready)
|
||||
|
||||
The trailers added to every commit on `will/ltx2_sr_port`:
|
||||
|
||||
```
|
||||
Co-authored-by: Junda (David) Su <90978028+Davids048@users.noreply.github.com>
|
||||
Co-authored-by: Matthew Noto <99706358+RandNMR73@users.noreply.github.com>
|
||||
Co-authored-by: XOR-op <17672363+XOR-op@users.noreply.github.com>
|
||||
Co-authored-by: Zhang Peiyuan <42993249+jzhang38@users.noreply.github.com>
|
||||
```
|
||||
|
||||
## How the trailers were applied
|
||||
|
||||
```bash
|
||||
git rebase --exec '
|
||||
git commit --amend --no-edit \
|
||||
--trailer "Co-authored-by: Junda (David) Su <90978028+Davids048@users.noreply.github.com>" \
|
||||
--trailer "Co-authored-by: Matthew Noto <99706358+RandNMR73@users.noreply.github.com>" \
|
||||
--trailer "Co-authored-by: XOR-op <17672363+XOR-op@users.noreply.github.com>" \
|
||||
--trailer "Co-authored-by: Zhang Peiyuan <42993249+jzhang38@users.noreply.github.com>"
|
||||
' origin/main will/ltx2_sr_port
|
||||
```
|
||||
|
||||
Git's `--trailer` flag is idempotent (it dedupes by the full `key: value`
|
||||
string), so re-running the rebase is safe and won't add duplicates.
|
||||
|
||||
## How to add a new co-author later
|
||||
|
||||
1. Add the user to the roster table above.
|
||||
2. Append their `Co-authored-by` line to the trailer block.
|
||||
3. Re-run the rebase command above on `will/ltx2_sr_port` — git's
|
||||
trailer dedupe handles the existing 4; the new one gets appended.
|
||||
4. Re-slice all 10 split branches per [`STACK.md`](STACK.md).
|
||||
5. Force-push `will/api_7.6`, `will/api_7.7`, and `will/ltx2_sr_port`.
|
||||
|
||||
## What we do NOT add
|
||||
|
||||
Per [`AGENTS.md`](AGENTS.md):
|
||||
|
||||
> Never add any coding agent or models such as Claude (or Claude Code), GPT,
|
||||
> Codex or others as a co-author in commits or PRs.
|
||||
|
||||
So no `Co-authored-by: Claude <noreply@anthropic.com>` or similar. Only
|
||||
human collaborators.
|
||||
@@ -0,0 +1,145 @@
|
||||
# PR Stack — `will/ltx2_sr_port` Split Tracker
|
||||
|
||||
**Status:** TEMPORARY — delete after the full stack lands on `main`.
|
||||
**Last updated:** 2026-05-04
|
||||
**Canonical source:** [`will/ltx2_sr_port`](https://github.com/hao-ai-lab/FastVideo/tree/will/ltx2_sr_port) — the unified branch holding all 41 commits in dependency order.
|
||||
|
||||
This file tracks the 10-PR split of `will/ltx2_sr_port`. Each PR is a slice
|
||||
of the canonical branch, stacked on its predecessor.
|
||||
|
||||
For per-PR design rationale see [`.agents/memory/dreamverse-integration/pr-roadmap.md`](.agents/memory/dreamverse-integration/pr-roadmap.md).
|
||||
For decisions log see [`.agents/memory/dreamverse-integration/decisions-log.md`](.agents/memory/dreamverse-integration/decisions-log.md).
|
||||
For co-author attribution see [`CO-AUTHORS.md`](CO-AUTHORS.md).
|
||||
|
||||
---
|
||||
|
||||
## Stack diagram
|
||||
|
||||
```
|
||||
origin/main
|
||||
└─ will/api_7.6 [open PR #1257] GPU pool + worker subprocess (7 commits)
|
||||
└─ will/api_7.7 [open PR #1258] Prompt enhancer + LLMProvider (3 commits)
|
||||
└─ will/api_7.8 Streaming auxiliaries (2 commits)
|
||||
└─ will/api_7.9 Router upstream (3 commits)
|
||||
└─ will/api_7.10 generate_async + Dynamo (3 commits)
|
||||
└─ will/api_8 Server contract docs (3 commits)
|
||||
└─ will/ltx2_sr_runtime LTX-2 SR + i2v + harness (9 commits)
|
||||
└─ will/ltx2_nvfp4 NVFP4 + per-component compile (6 commits)
|
||||
└─ will/ltx2_post_fixes Post-handoff parity fixes (2 commits)
|
||||
└─ will/agents_cleanup .agents/ cleanup + STACK.md (3 commits)
|
||||
```
|
||||
|
||||
Total: **41 commits across 10 PRs** (mean 4.1 per PR — reviewable size).
|
||||
|
||||
## Branch table
|
||||
|
||||
| # | PR branch | Open PR | Slice | Tip SHA | Commits | Concern |
|
||||
|---|---|---|---|---|---|---|
|
||||
| 1 | `will/api_7.6` | **#1257** | 1–7 | _SEE NOTES_ | 7 | GPU pool + worker subprocess + test coverage + 4 review-fixup commits |
|
||||
| 2 | `will/api_7.7` | **#1258** | 8–10 | _SEE NOTES_ | 3 | Prompt enhancer + LLMProvider abstraction (cerebras, groq) |
|
||||
| 3 | `will/api_7.8` | (not yet) | 11–12 | _SEE NOTES_ | 2 | Streaming auxiliaries (safety, rewrite, logger, mock) |
|
||||
| 4 | `will/api_7.9` | (not yet) | 13–15 | _SEE NOTES_ | 3 | Router upstream + `fastvideo router-serve` CLI + test coverage |
|
||||
| 5 | `will/api_7.10` | (not yet) | 16–18 | _SEE NOTES_ | 3 | `VideoEvent` hierarchy + `VideoGenerator.generate_async` + Dynamo health helper |
|
||||
| 6 | `will/api_8` | (not yet) | 19–21 | _SEE NOTES_ | 3 | Server contract docs (OpenAI HTTP + Dynamo) + Dreamverse/Dynamo shape contract tests |
|
||||
| 7 | `will/ltx2_sr_runtime` | (not yet) | 22–30 | _SEE NOTES_ | 9 | LTX-2 SR runtime port + i2v conditioning + alignment harness against FastVideo-internal |
|
||||
| 8 | `will/ltx2_nvfp4` | (not yet) | 31–36 | _SEE NOTES_ | 6 | NVFP4 wire-up + per-component compile + typed `transformer_quant` flow |
|
||||
| 9 | `will/ltx2_post_fixes` | (not yet) | 37–38 | _SEE NOTES_ | 2 | Post-handoff parity fixes (Gemma `to()` round-trip, list-of-generators) |
|
||||
| 10 | `will/agents_cleanup` | (not yet) | 39–41 | _SEE NOTES_ | 3 | `.agents/` Phase 1 cleanup + dreamverse-integration memory + STACK/CO-AUTHORS docs |
|
||||
|
||||
**Tip SHAs**: see `git log --oneline --reverse origin/main..will/ltx2_sr_port` for the
|
||||
authoritative sequence. SHAs change whenever `will/ltx2_sr_port` is rebased
|
||||
or amended (e.g. when injecting new `Co-authored-by` trailers). Rather than
|
||||
hardcode them here (which goes stale), use the live re-slicing commands
|
||||
below.
|
||||
|
||||
## Re-slice commands (run when `will/ltx2_sr_port` SHAs change)
|
||||
|
||||
After any rebase or trailer injection on `will/ltx2_sr_port`, recompute the
|
||||
boundary SHAs and reset all split branches:
|
||||
|
||||
```bash
|
||||
# Show the 41 commits with their slice numbers
|
||||
git log --oneline --reverse origin/main..will/ltx2_sr_port | nl -ba
|
||||
|
||||
# Capture boundary tips by index
|
||||
SHA_7_6=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '7p' | awk '{print $1}')
|
||||
SHA_7_7=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '10p' | awk '{print $1}')
|
||||
SHA_7_8=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '12p' | awk '{print $1}')
|
||||
SHA_7_9=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '15p' | awk '{print $1}')
|
||||
SHA_7_10=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '18p' | awk '{print $1}')
|
||||
SHA_8=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '21p' | awk '{print $1}')
|
||||
SHA_LTX2_SR=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '30p' | awk '{print $1}')
|
||||
SHA_LTX2_NVFP4=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '36p' | awk '{print $1}')
|
||||
SHA_LTX2_POST=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '38p' | awk '{print $1}')
|
||||
SHA_AGENTS=$(git log --oneline --reverse origin/main..will/ltx2_sr_port | sed -n '41p' | awk '{print $1}')
|
||||
|
||||
# Reset all 10 branches to the new boundary tips
|
||||
git branch -f will/api_7.6 "$SHA_7_6"
|
||||
git branch -f will/api_7.7 "$SHA_7_7"
|
||||
git branch -f will/api_7.8 "$SHA_7_8"
|
||||
git branch -f will/api_7.9 "$SHA_7_9"
|
||||
git branch -f will/api_7.10 "$SHA_7_10"
|
||||
git branch -f will/api_8 "$SHA_8"
|
||||
git branch -f will/ltx2_sr_runtime "$SHA_LTX2_SR"
|
||||
git branch -f will/ltx2_nvfp4 "$SHA_LTX2_NVFP4"
|
||||
git branch -f will/ltx2_post_fixes "$SHA_LTX2_POST"
|
||||
git branch -f will/agents_cleanup "$SHA_AGENTS"
|
||||
```
|
||||
|
||||
After resetting, force-push the branches whose PRs are open:
|
||||
|
||||
```bash
|
||||
git push --force-with-lease origin will/api_7.6
|
||||
git push --force-with-lease origin will/api_7.7
|
||||
git push --force-with-lease origin will/ltx2_sr_port
|
||||
```
|
||||
|
||||
## Independence map
|
||||
|
||||
| PR | Hard-depends on | Could rebase to land independently? |
|
||||
|---|---|---|
|
||||
| 7.6 | `origin/main` only | ✅ already independent |
|
||||
| 7.7 | 7.6 (uses `GpuPool` from 7.6) | ❌ |
|
||||
| 7.8 | 7.7 (auxiliaries plug into prompt enhancer) | ❌ |
|
||||
| 7.9 | `origin/main` (router is orthogonal to streaming) | ✅ — could rebase onto 7.6 directly if 7.7/7.8 stall |
|
||||
| 7.10 | 7.6 (`GpuPool`) + on-main `ContinuationState` (PR 7) | ✅ — only needs 7.6, can fast-track over the 7.7/7.8 stack |
|
||||
| 8 | 7.10 (consumes `generate_async` + `VideoEvent`) | ❌ |
|
||||
| `ltx2_sr_runtime` | `origin/main` + (independent of 7.x streaming work) | ✅ — purely LTX-2-specific |
|
||||
| `ltx2_nvfp4` | `ltx2_sr_runtime` + per-component compile (which is on `ltx2_sr_runtime`) | ❌ |
|
||||
| `ltx2_post_fixes` | `ltx2_nvfp4` | ❌ |
|
||||
| `agents_cleanup` | NOTHING — pure infra | ✅ — fully independent of every other PR |
|
||||
|
||||
## Landing strategy
|
||||
|
||||
The default landing path is bottom-up through the stack: 7.6 → 7.7 → 7.8 →
|
||||
7.9 → 7.10 → 8 → ltx2_sr_runtime → ltx2_nvfp4 → ltx2_post_fixes →
|
||||
agents_cleanup.
|
||||
|
||||
After a PR merges, the canonical branch and remaining splits should be
|
||||
rebased onto the new `origin/main` (which now contains the just-merged PR),
|
||||
then re-sliced via the commands above.
|
||||
|
||||
If review on a mid-stack PR stalls, **fast-track candidates** are:
|
||||
|
||||
- `will/agents_cleanup` (no dependency)
|
||||
- `will/ltx2_sr_runtime` (no dependency on 7.x streaming work)
|
||||
- `will/api_7.9` (router is orthogonal — rebase to land directly on 7.6)
|
||||
- `will/api_7.10` (only needs 7.6 — rebase to land directly on 7.6)
|
||||
|
||||
## Verification per PR
|
||||
|
||||
For each PR's branch, verify the slice content with:
|
||||
|
||||
```bash
|
||||
# Show commits + verify count
|
||||
git log --oneline <PARENT>..<BRANCH> | wc -l # expect the count from the table
|
||||
|
||||
# Confirm tree state at the tip matches the canonical branch's slice tip
|
||||
git diff <slice-tip-SHA-on-canonical> <BRANCH> # expect empty
|
||||
```
|
||||
|
||||
## When to delete this file
|
||||
|
||||
When all 10 PRs have merged to `main`. Until then, keep this file fresh:
|
||||
re-run the re-slice commands and update any commit-count cells whose
|
||||
sequence shifts.
|
||||
Reference in New Issue
Block a user