feat: switch hosted OCR CPU actor defaults to Nemotron OCR v2

[kheiss-uwzoo]

Summary

Switch hosted/remote CPU OCR actor DEFAULT_INVOKE_URL values from nemotron-ocr-v1 to nemotron-ocr-v2, and align README/CLI examples, library-mode CI, and tests.

Draft — for eng review. Split from #2259 per Randy's guidance: docs PR stays docs-only; this change waits for eng sign-off on hosted endpoint readiness.

On main, hosted defaults remain v1 intentionally (#2021: v2 available locally and in Helm, remote flip deferred).

Changes

• Operators: ocr/cpu_ocr.py, chart/cpu_actor.py, table/cpu_actor.py, video/ocr_actor.py
• README/CLI example invoke URLs
• Library-mode CI: OCR_INVOKE_URL → v2
• Tests asserting default remote URLs

PR scope check

• Base: main
• Files changed: 12 — 4 operators, 5 tests, README/CLI, 1 CI workflow
• Red flags: .github/workflows/integration-test-library-mode.yml (OCR invoke URL only)

Test plan

• pytest nemo_retriever/tests/test_operator_flags_and_cpu_actors.py nemo_retriever/tests/test_video_frame_ocr_actor.py
• Confirm library-mode integration workflow uses nemotron-ocr-v2
• Verify hosted v2 endpoint is GA before merging

[kheiss-uwzoo]

Addressed the nemoretriever-ocr-v2 callout in dde7529a: library-mode CI and README video tests now use nemotron-ocr-v2, matching operator DEFAULT_INVOKE_URL values and build.nvidia.com.

[kheiss-uwzoo]

Addressed both review comments in \dde7529a\ and rebased onto latest \main\ (\�5404c05):

• Library-mode CI (.github/workflows/integration-test-library-mode.yml): \OCR_INVOKE_URL\ now uses
  emotron-ocr-v2\ (not
  emoretriever-ocr-v2).
• README video pipeline tests (\ est_readme_video_pipeline_example.py): both \ocr_invoke_url\ fixtures updated to
  emotron-ocr-v2.

Review threads resolved. PR diff remains scoped to the 12 OCR-default files.

[greptile-apps]

Greptile Summary

This PR switches hosted OCR defaults and examples to Nemotron OCR v2. The main changes are:

• Hosted CPU OCR actor defaults now point at nemotron-ocr-v2.
• Video frame CPU OCR defaults now point at nemotron-ocr-v2.
• README, CLI examples, workflow env, and tests now use the v2 OCR URL.

Confidence Score: 5/5

This looks safe to merge.

• No blocking issues found in the changed code.

Important Files Changed

Filename	Overview
nemo_retriever/src/nemo_retriever/operators/extract/ocr/cpu_ocr.py	Updates the hosted CPU OCR default endpoint to nemotron-ocr-v2.
nemo_retriever/src/nemo_retriever/operators/extract/video/ocr_actor.py	Updates the video frame CPU OCR default endpoint to nemotron-ocr-v2.
.github/workflows/integration-test-library-mode.yml	Updates the library-mode integration OCR invoke URL to the hosted v2 endpoint.
nemo_retriever/src/nemo_retriever/operators/extract/table/cpu_actor.py	Clarifies that table OCR is only used when an OCR invoke URL is provided.

_{Reviews (2): Last reviewed commit: "Merge remote-tracking branch 'upstream/m..." | Re-trigger Greptile}

[kheiss-uwzoo]

Rebased onto latest \main\ (\�ec74047) — merge conflicts are resolved.

Conflict resolutions:

• .github/workflows/integration-test-library-mode.yml: kept
  emotron-ocr-v2; dropped \GRAPHIC_ELEMENTS_INVOKE_URL\ (chart operator removed on main)
• \chart/cpu_actor.py: accepted deletion (entire chart/graphic-elements operator removed on main)
• \ est_operator_flags_and_cpu_actors.py: removed \TestGraphicElementsCPUActor\ (tested deleted actor)

Hosted OCR v2 defaults remain on the surviving CPU actors (\ocr/cpu_ocr.py, \�ideo/ocr_actor.py), README/CLI examples, and CI. Unit tests green.

@charlesbluca — ready for your re-review when you have a moment.