ai: Update llm pipeline

[ad-astra-video]

What does this pull request do? Explain your changes. (required)

Updates go-livepeer for LLM pipeline updates for new response format.

Adds in validation of LLM request to gateway provided in #3204

Note: checked box to allow edits by maintainers to help expedite with timezone difference. Feel free to push any updates needed to branch.

Requires ai-worker livepeer/ai-worker#387 to be merged
Specific updates (required)

• Updates needed for ai-worker changes to LLM request and response formats

How did you test each of these updates (required)

Built docker image locally for livepeer/ai-runner:llm and livepeer/go-livepeer to test streaming and non-streaming responses

Does this pull request close any open issues?

No

Checklist:

• Read the contribution guide
• make runs successfully
• All tests in ./test.sh pass
• README and other documentation updated
• Pending changelog updated

[ad-astra-video]

I will update go.mod when we get the second PR merged on ai-worker and new version tagged so can pull in final updates from ai-worker.

[leszko]

Added 2 comments. Please also check if CI passes.

[leszko]

@ad-astra-video The PR looks fine, but I see CI is failing, could you check?

https://github.com/livepeer/go-livepeer/actions/runs/12532555474/job/34951389606?pr=3336

[victorges]

LGTM but remember to update go.mod before merging

[codecov]

Codecov Report

Attention: Patch coverage is 3.57143% with 54 lines in your changes missing coverage. Please review.

Project coverage is 33.69713%. Comparing base (3acf4ea) to head (0b9cdc7).

Files with missing lines	Patch %	Lines
server/ai_process.go	0.00000%	30 Missing ⚠️
server/ai_http.go	0.00000%	9 Missing ⚠️
server/ai_mediaserver.go	0.00000%	8 Missing ⚠️
core/ai_worker.go	0.00000%	4 Missing ⚠️
server/ai_worker.go	40.00000%	2 Missing and 1 partial ⚠️

Additional details and impacted files

@@                 Coverage Diff                 @@
##              master       #3336         +/-   ##
===================================================
+ Coverage   33.68997%   33.69713%   +0.00716%     
===================================================
  Files            141         141                 
  Lines          37480       37475          -5     
===================================================
+ Hits           12627       12628          +1     
+ Misses         24131       24125          -6     
  Partials         722         722                 

Files with missing lines	Coverage Δ
core/ai.go	58.51852% <ø> (ø)
server/rpc.go	66.66667% <ø> (ø)
server/ai_worker.go	49.89059% <40.00000%> (ø)
core/ai_worker.go	31.06796% <0.00000%> (+0.12136%)	⬆️
server/ai_mediaserver.go	7.96117% <0.00000%> (+0.03080%)	⬆️
server/ai_http.go	10.03135% <0.00000%> (ø)
server/ai_process.go	0.59932% <0.00000%> (+0.00154%)	⬆️

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3acf4ea...0b9cdc7. Read the comment docs.

Files with missing lines	Coverage Δ
core/ai.go	58.51852% <ø> (ø)
server/rpc.go	66.66667% <ø> (ø)
server/ai_worker.go	49.89059% <40.00000%> (ø)
core/ai_worker.go	31.06796% <0.00000%> (+0.12136%)	⬆️
server/ai_mediaserver.go	7.96117% <0.00000%> (+0.03080%)	⬆️
server/ai_http.go	10.03135% <0.00000%> (ø)
server/ai_process.go	0.59932% <0.00000%> (+0.00154%)	⬆️

[ad-astra-video]

@rickstaa @kyriediculous Updated PR for latest updates in LLM pipeline. Please let me know if anything want changed. Built and tested locally end-to-end with no issues.

[rickstaa]

LGTM, will run one production test and merge.

Addressed changes and OOO.