Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
114 commits
Select commit Hold shift + click to select a range
a6db760
feat: run native H3 video sweeps from YAML recipes
edwingao28 Nov 3, 2026
b7edc5f
feat: verify GPU ownership and seal H3 execution evidence
edwingao28 Nov 3, 2026
a4c6cd8
feat: define reproducible H3 workload sets
edwingao28 Nov 3, 2026
d9353c8
feat: measure H3 requests and validate generated media
edwingao28 Nov 3, 2026
84243e2
chore: merge origin/main into videogenx native ship
edwingao28 Nov 3, 2026
0c27900
refactor: delete unused OwnedProcess.owns helper
edwingao28 Nov 3, 2026
f3b94f9
chore: merge GPU evidence cleanup into native H3 sweep
edwingao28 Nov 3, 2026
78072a7
refactor: delete unused H3 bundle JSON digest helper
edwingao28 Nov 3, 2026
21f5637
chore: merge GPU evidence into native H3 sweep
edwingao28 Nov 3, 2026
42b1b16
chore: merge origin/main into H3 workload contract
edwingao28 Nov 3, 2026
d73bf24
chore: merge measurement core into GPU evidence
edwingao28 Nov 3, 2026
72cfc5a
chore: merge H3 workload contract into measurement core
edwingao28 Nov 3, 2026
884b5f2
test: restore shared launch and workflow test cases
edwingao28 Nov 3, 2026
39b16e8
refactor: fold H3 spec digest and case helpers
edwingao28 Nov 3, 2026
0b49d0c
test: table-drive H3 spec rejection cases
edwingao28 Nov 3, 2026
7da1e60
chore: compact H3 spec YAML and drop stale references
edwingao28 Nov 3, 2026
58d90e2
fix: use warmup case in GPU report slots
edwingao28 Nov 3, 2026
8acb65a
refactor: drop unreachable GPU job guards
edwingao28 Nov 3, 2026
f48424f
test: drop dead parametrize branches in measurement tests
edwingao28 Nov 3, 2026
86b4d6a
refactor: share no-follow open and canonical JSON helpers
edwingao28 Nov 3, 2026
1af606c
test: inline single-case video parametrize and share reseal fixtures
edwingao28 Nov 3, 2026
2c798e7
refactor: share deadline budget in GPU probes
edwingao28 Nov 3, 2026
e82ff60
refactor: reuse runner slots and hashing in compare
edwingao28 Nov 3, 2026
b57162d
test: drop dead parametrize branches in GPU tests
edwingao28 Nov 3, 2026
8b1d9c5
refactor: dedupe evaluator encoding, status, and quantile rules
edwingao28 Nov 3, 2026
a19e13f
refactor: drop duplicated video digest, srun, schema and env helpers
edwingao28 Nov 3, 2026
d85b37d
refactor: share video srt recipe defaults like agentic
edwingao28 Nov 3, 2026
66324c7
docs: link video pins and timeouts to their sources
edwingao28 Nov 3, 2026
21d0239
chore: merge measurement core into GPU evidence
edwingao28 Nov 3, 2026
2a2fef7
chore: merge GPU evidence into native H3 sweep
edwingao28 Nov 3, 2026
d3a41d4
chore: merge H3 workload contract into measurement core
edwingao28 Nov 3, 2026
03a26e9
chore: merge measurement core into native H3 sweep
edwingao28 Nov 3, 2026
4c0b861
refactor: move offline compare and report tools out of measurement core
edwingao28 Nov 3, 2026
e18ad7b
refactor: move offline GPU job tools out of native H3 sweep
edwingao28 Nov 3, 2026
0fccd68
refactor: move manual srt-yaml commands out of native H3 sweep
edwingao28 Nov 4, 2026
d20fd9d
style: format native H3 evaluator modules with ruff
edwingao28 Nov 4, 2026
91ee12b
refactor: expose canonical_json_bytes as the one H3 hash encoding
edwingao28 Nov 4, 2026
a965d6e
fix: check the video readiness port and drop the empty prepare descri…
edwingao28 Nov 4, 2026
7db9735
refactor: drop citations and per-engine notes from H3 specs
edwingao28 Nov 4, 2026
d4e6f34
test: drop the hand-built fairness hash test
edwingao28 Nov 4, 2026
17e4c63
refactor: validate H3 v0.1 policy once in the spec loader by schema v…
edwingao28 Nov 4, 2026
feb6af6
ci: lint and format all of experimental/video-generation and drop mea…
edwingao28 Nov 4, 2026
5660b77
chore: merge H3 workload contract into measurement core
edwingao28 Nov 4, 2026
bd16ff8
revert: keep run comparison and PSNR in the measurement core
edwingao28 Nov 4, 2026
0d93090
refactor: move SGLang server timing out of the measurement core
edwingao28 Nov 4, 2026
abaf289
chore: drop the unreferenced smoke plan and the nonexistent mvp extra…
edwingao28 Nov 4, 2026
2112dae
style: ruff format the measurement core
edwingao28 Nov 4, 2026
d03b073
refactor: share the native video file contract from infx.bench.video
edwingao28 Nov 4, 2026
e9371be
refactor: hash run bundles with the spec's canonical JSON
edwingao28 Nov 4, 2026
75c2eb7
refactor: name each rejected field of the sealed v0.1 server launch
edwingao28 Nov 4, 2026
8eb8447
refactor: trust loader-validated plans and drop the redundant run_pla…
edwingao28 Nov 4, 2026
debb90f
refactor: derive engine encoding requests and media checks from the p…
edwingao28 Nov 4, 2026
2fc9b41
refactor: keep native video checks at the binder, executor and ingest…
edwingao28 Nov 4, 2026
8ee78c5
chore: drop unconsumed prose and constant fields from run data
edwingao28 Nov 4, 2026
3c03a0c
fix: keep a bounded redacted HTTP error excerpt and refuse every paid…
edwingao28 Nov 4, 2026
670580e
refactor: default the single-choice reference dtype and attention bac…
edwingao28 Nov 4, 2026
ba59a25
refactor: record native video job status as JSON
edwingao28 Nov 4, 2026
ad9c4d1
fix: count malformed power samples in the video bundle instead of ski…
edwingao28 Nov 4, 2026
ffaeb74
fix: drop unread constant bundle fields and validate each power sampl…
edwingao28 Nov 4, 2026
93c2163
refactor: resolve the video harness at one path beside the project
edwingao28 Nov 4, 2026
93d7db6
chore: merge measurement core into native H3 sweep
edwingao28 Nov 4, 2026
6ab00f7
fix: hash the measurement window run with the shared canonical JSON
edwingao28 Nov 4, 2026
fe49aaa
refactor: seal the image digest for every native video plan and check…
edwingao28 Nov 4, 2026
bc12c6c
fix: keep the video workload specs in nested matrix snapshots
edwingao28 Nov 4, 2026
4de6e7a
fix: report the native video bundle outcome as its CI phase
edwingao28 Nov 4, 2026
4348071
fix: count video sweep jobs in run stats after they finish
edwingao28 Nov 4, 2026
d8f9ee9
fix: drop the static protocol source pin so run configuration names o…
edwingao28 Nov 4, 2026
02fba81
refactor: rename generic H3 spec and workflow names to video
edwingao28 Nov 4, 2026
4ec9d98
chore: merge video workload contract into measurement core
edwingao28 Nov 4, 2026
62a58ff
refactor: rename generic H3 measurement-core naming to video
edwingao28 Nov 4, 2026
1ca9352
chore: merge measurement core into native video sweep
edwingao28 Nov 4, 2026
f03d6e1
refactor: rename generic H3 native-sweep naming to video
edwingao28 Nov 4, 2026
e13bacc
chore: merge origin/main into H3 workload contract
edwingao28 Nov 4, 2026
a87f5c3
chore: merge H3 workload contract into measurement core
edwingao28 Nov 4, 2026
6828a9b
chore: merge measurement core into native H3 sweep
edwingao28 Nov 4, 2026
bfc5812
fix: record approved head and video matrix in trusted sweeps
edwingao28 Nov 4, 2026
7401ad4
fix: validate trusted video sweeps with the dispatching workflow's va…
edwingao28 Nov 4, 2026
8be56f9
feat: run H3 on SGLang v0.5.23 for B200 and H100 (#4776)
edwingao28 Nov 4, 2026
7558694
fix: record engine parallelism in native video spec.json (#4777)
edwingao28 Nov 4, 2026
4b3a222
chore: fold the v0.5.23 changelog entry into the H3 sweep entry
edwingao28 Nov 4, 2026
5fdb865
chore: merge main into H3 workload contract
edwingao28 Nov 4, 2026
b8afd6b
chore: merge H3 workload contract into measurement core
edwingao28 Nov 4, 2026
4fa88e5
chore: merge measurement core into native H3 sweep
edwingao28 Nov 4, 2026
4e1c7c4
chore: drop the H200 vLLM-Omni cell from the H3 sweep
edwingao28 Nov 4, 2026
4ff7777
chore: run H3 H200 on SGLang v0.5.23
edwingao28 Nov 4, 2026
035a2af
chore: drop the H200 SGLang cell from the H3 sweep
edwingao28 Nov 4, 2026
1b9af0a
refactor: move the video harness to top-level video-generation
edwingao28 Nov 4, 2026
df4d1f2
refactor: name the 20-prompt H3 spec by its schema version
edwingao28 Nov 4, 2026
8629a92
chore: merge video workload contract into measurement core
edwingao28 Nov 4, 2026
bdcea42
docs: drop the release label from the runner's policy note
edwingao28 Nov 4, 2026
09454fa
fix: let vgbench run record the server's numerical identity
edwingao28 Nov 4, 2026
a519a26
fix: check reference case IDs before selecting all cases
edwingao28 Nov 4, 2026
952e047
chore: merge measurement core into native H3 sweep
edwingao28 Nov 4, 2026
ce4efe2
docs: drop the release label from the H3 docs and changelog entry
edwingao28 Nov 4, 2026
7f218d4
fix: accept the newest video bundle per deployment across rerun attempts
edwingao28 Nov 4, 2026
9118bbd
fix: retry the video results upload like other result uploads
edwingao28 Nov 4, 2026
be3315a
fix: stop serving-mode video runs promptly on interrupt
edwingao28 Nov 4, 2026
4ca79ff
chore: merge measurement core into native H3 sweep
edwingao28 Nov 4, 2026
e7d6d55
chore: merge main into H3 workload contract
edwingao28 Nov 4, 2026
6ce7cd7
chore: merge video workload contract into measurement core
edwingao28 Nov 4, 2026
1b2078a
chore: merge measurement core into native H3 sweep
edwingao28 Nov 4, 2026
fca4366
fix: keep a single video bundle in its named directory
edwingao28 Nov 5, 2026
689c9b8
chore: merge main into the video workload stack
edwingao28 Nov 5, 2026
b6fde8e
chore: merge main into the video measurement stack
edwingao28 Nov 5, 2026
7fce18f
chore: merge main into the H3 cross-hardware stack
edwingao28 Nov 5, 2026
f667f38
refactor: share single-node srt steps between points and video
edwingao28 Nov 5, 2026
2069d51
chore: merge main into the H3 cross-hardware branch
edwingao28 Nov 5, 2026
e7fe0ed
fix: run video sweeps only when a selection names them
edwingao28 Nov 5, 2026
4c6a8c7
refactor: drop the unused services-only recipe node count
edwingao28 Nov 5, 2026
d75760f
ci: raise video job limits to fit 1-GPU points
edwingao28 Nov 5, 2026
b4ee120
chore: merge main into the H3 cross-hardware branch
edwingao28 Nov 5, 2026
80c3c84
test: prune 46 slop cases from the video sweep tests
edwingao28 Nov 5, 2026
35c1748
docs: fix video harness README paths to the repo root
edwingao28 Nov 5, 2026
f47845b
Merge remote-tracking branch 'origin/main' into pr-2916-reuse-39771
edwingao28 Nov 5, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions .github/workflows/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ usage: python -m infx.matrix.generate full-sweep
[--no-evals | --evals-only] [--all-evals]
[--smoke] [--trim-conc]
[--runner-node-filter RUNNER_NODE_FILTER]
[--scenario-type {fixed-seq-len,agentic-coding} [{fixed-seq-len,agentic-coding} ...]]
[--scenario-type {fixed-seq-len,agentic-coding,video-generation} [{fixed-seq-len,agentic-coding,video-generation} ...]]
[--model-prefix MODEL_PREFIX [MODEL_PREFIX ...]]
[--precision PRECISION [PRECISION ...]]
[--framework FRAMEWORK [FRAMEWORK ...]]
Expand Down Expand Up @@ -109,7 +109,7 @@ usage: python -m infx.matrix.generate test-config
[--no-evals | --evals-only] [--all-evals]
[--smoke] [--trim-conc]
[--runner-node-filter RUNNER_NODE_FILTER]
[--scenario-type {fixed-seq-len,agentic-coding} [{fixed-seq-len,agentic-coding} ...]]
[--scenario-type {fixed-seq-len,agentic-coding,video-generation} [{fixed-seq-len,agentic-coding,video-generation} ...]]
--config-keys CONFIG_KEYS [CONFIG_KEYS ...]
[--conc CONC [CONC ...]]
[--exp-names EXP_NAMES [EXP_NAMES ...]]
Expand Down Expand Up @@ -253,6 +253,8 @@ This proof of concept produces benchmark and evaluation artifacts through the
End-to-End Tests workflow. The dispatch also uploads a `source-identity`
artifact (PR number, full approved head SHA, merge SHA, and the `evals-only`,
`agentx-fast`, `all-evals`, and fail-fast modifiers) and `changelog-metadata`.
When the matrix selects video, it also uploads `video-sweep-manifest` and runs
`validate-video-results`, because video reuse requires that manifest.
A pinned `/use <run_id>` accepts these runs; reuse and staging report the
recorded PR head and judge modifiers from the recorded dispatch rather than the
PR's current labels. Older runs without the artifact fall back to the run title
Expand Down
80 changes: 70 additions & 10 deletions .github/workflows/benchmark-tmpl.yml
Original file line number Diff line number Diff line change
Expand Up @@ -84,6 +84,11 @@ on:
required: false
type: string
default: ""
head-sha:
description: "Approved PR head a trusted dispatch tested; empty uses the triggering event's head"
required: false
type: string
default: ""
test-name:
description: "Name of the dispatched test run; empty for other triggers"
required: false
Expand All @@ -95,7 +100,7 @@ env:
PORT: '8888'
INFMAX_CONTAINER_WORKSPACE: '/workspace'
ENROOT_IMPORT_TIME_LIMIT: '120'
SALLOC_TIME_LIMIT: '480'
SALLOC_TIME_LIMIT: ${{ inputs.scenario-type == 'video-generation' && '150' || '480' }}
IS_MULTINODE: 'false'
REQUIRE_POWER: ${{ inputs.require-power && '1' || '0' }}
CPU_POWER: ${{ fromJSON(inputs.config)['cpu-power'] && '1' || '0' }}
Expand All @@ -108,9 +113,9 @@ env:
MODEL: ${{ fromJSON(inputs.config).model }}
THINKING_MODE: thinking_on
MODEL_PREFIX: ${{ fromJSON(inputs.config).model-prefix }}
ISL: ${{ inputs.scenario-type == 'agentic-coding' && '0' || fromJSON(inputs.config).isl }}
OSL: ${{ inputs.scenario-type == 'agentic-coding' && '0' || fromJSON(inputs.config).osl }}
MAX_MODEL_LEN: ${{ inputs.scenario-type == 'agentic-coding' && '0' || fromJSON(inputs.config).max-model-len }}
ISL: ${{ (inputs.scenario-type == 'agentic-coding' || inputs.scenario-type == 'video-generation') && '0' || fromJSON(inputs.config).isl }}
OSL: ${{ (inputs.scenario-type == 'agentic-coding' || inputs.scenario-type == 'video-generation') && '0' || fromJSON(inputs.config).osl }}
MAX_MODEL_LEN: ${{ (inputs.scenario-type == 'agentic-coding' || inputs.scenario-type == 'video-generation') && '0' || fromJSON(inputs.config).max-model-len }}
IMAGE: ${{ fromJSON(inputs.config).image }}
FRAMEWORK: ${{ fromJSON(inputs.config).framework }}
PRECISION: ${{ fromJSON(inputs.config).precision }}
Expand Down Expand Up @@ -139,6 +144,12 @@ env:
AIPERF_EXPERIMENTAL_FAST: ${{ inputs.agentx-fast && '1' || '0' }}
EVAL_LIMIT: ${{ inputs.eval-limit }}
AL_GATE_DATASET: ${{ inputs.eval-only && inputs.scenario-type == 'agentic-coding' && 'al_gate_dataset/speedbench_coding.jsonl' || '' }}
GITHUB_HEAD_SHA: ${{ inputs.head-sha || github.event.pull_request.head.sha || github.sha }}
VIDEO_ARTIFACT_NAME: ${{ inputs.scenario-type == 'video-generation' && format('h3-results-{0}-{1}-{2}', github.run_id, github.run_attempt, fromJSON(inputs.config).recipe-fingerprint) || '' }}
VIDEO_SPEC: ${{ inputs.scenario-type == 'video-generation' && fromJSON(inputs.config).spec || '' }}
VIDEO_SET: ${{ inputs.scenario-type == 'video-generation' && fromJSON(inputs.config).set || '' }}
VIDEO_GPUS: ${{ inputs.scenario-type == 'video-generation' && fromJSON(inputs.config).gpus || '' }}
VIDEO_CONC_LIST: ${{ inputs.scenario-type == 'video-generation' && join(fromJSON(inputs.config).conc-list, ' ') || '' }}
AIPERF_FAILED_REQUEST_THRESHOLD: '0.10'
RESULT_DIR: /workspace/results
PYTHONDONTWRITEBYTECODE: '1'
Expand Down Expand Up @@ -173,14 +184,14 @@ jobs:
) ||
format('[{0}]', toJSON(inputs.runner))
) }}
timeout-minutes: ${{ inputs.runner == 'cluster:h200-dgxc' && fromJSON(inputs.config).model-prefix == 'dsv41flash' && fromJSON(inputs.config).framework == 'sglang' && inputs.scenario-type == 'agentic-coding' && !inputs.eval-only && fromJSON(inputs.config).conc >= 64 && 1470 || 500 }}
timeout-minutes: ${{ inputs.scenario-type == 'video-generation' && 165 || inputs.runner == 'cluster:h200-dgxc' && fromJSON(inputs.config).model-prefix == 'dsv41flash' && fromJSON(inputs.config).framework == 'sglang' && inputs.scenario-type == 'agentic-coding' && !inputs.eval-only && fromJSON(inputs.config).conc >= 64 && 1470 || 500 }}
name: >-
${{ inputs.klaud-run && 'klaud | ' || '' }}p${{ inputs.priority }} | ${{ fromJSON(inputs.config).model-prefix }} ${{ fromJSON(inputs.config).precision }} ${{ inputs.runner }} ${{ fromJSON(inputs.config).framework == 'sglang' && 'sgl' || fromJSON(inputs.config).framework == 'dynamo-sglang' && 'dyn-sgl' || fromJSON(inputs.config).framework == 'sglang-disagg' && 'sgl-disagg' || fromJSON(inputs.config).framework }}
TP${{ fromJSON(inputs.config).tp }}${{ format('{0}', fromJSON(inputs.config).pp) != '' && format('{0}', fromJSON(inputs.config).pp) != '1' && format('/PP{0}', fromJSON(inputs.config).pp) || '' }}${{ format('{0}', fromJSON(inputs.config).dcp-size) != '' && format('{0}', fromJSON(inputs.config).dcp-size) != '1' && format('/DCP{0}', fromJSON(inputs.config).dcp-size) || '' }}${{ format('{0}', fromJSON(inputs.config).pcp-size) != '' && format('{0}', fromJSON(inputs.config).pcp-size) != '1' && format('/PCP{0}', fromJSON(inputs.config).pcp-size) || '' }}${{ format('{0}', fromJSON(inputs.config).ep) != '' && format('{0}', fromJSON(inputs.config).ep) != '1' && format('/EP{0}', fromJSON(inputs.config).ep) || '' }}${{ inputs.dp-attn && '/DPA' || '' }}
${{ inputs.scenario-type == 'video-generation' && format('{0} GPU', fromJSON(inputs.config).gpus) || format('TP{0}', fromJSON(inputs.config).tp) }}${{ format('{0}', fromJSON(inputs.config).pp) != '' && format('{0}', fromJSON(inputs.config).pp) != '1' && format('/PP{0}', fromJSON(inputs.config).pp) || '' }}${{ format('{0}', fromJSON(inputs.config).dcp-size) != '' && format('{0}', fromJSON(inputs.config).dcp-size) != '1' && format('/DCP{0}', fromJSON(inputs.config).dcp-size) || '' }}${{ format('{0}', fromJSON(inputs.config).pcp-size) != '' && format('{0}', fromJSON(inputs.config).pcp-size) != '1' && format('/PCP{0}', fromJSON(inputs.config).pcp-size) || '' }}${{ format('{0}', fromJSON(inputs.config).ep) != '' && format('{0}', fromJSON(inputs.config).ep) != '1' && format('/EP{0}', fromJSON(inputs.config).ep) || '' }}${{ inputs.dp-attn && '/DPA' || '' }}
${{ fromJSON(inputs.config).spec-decoding != 'none' && fromJSON(inputs.config).spec-decoding || '' }}
${{ inputs.scenario-type == 'agentic-coding' && format('{0}', fromJSON(inputs.config).kv-offloading) != '' && fromJSON(inputs.config).kv-offloading != 'none' && format('{0} KV offload', fromJSON(inputs.config).kv-offloading) || '' }}
${{ inputs.scenario-type == 'agentic-coding' && fromJSON(inputs.config).kv-offload-backend.name != 'none' && fromJSON(inputs.config).kv-offload-backend.name != 'default' && format('{0}', fromJSON(inputs.config).kv-offload-backend.name) || '' }}
c${{ fromJSON(inputs.config).conc }}${{ inputs.eval-only && ' | eval-only' || (inputs.run-eval && ' | eval' || '') }}
c${{ inputs.scenario-type == 'video-generation' && join(fromJSON(inputs.config).conc-list, 'x') || fromJSON(inputs.config).conc }}${{ inputs.eval-only && ' | eval-only' || (inputs.run-eval && ' | eval' || '') }}
steps:
- name: Repair root-owned artifacts (pre-run)
if: ${{ inputs.runner == 'cluster:mi355x-amds' }}
Expand Down Expand Up @@ -280,7 +291,13 @@ jobs:
echo "INFERENCEX_E2E_ROOT=$PWD" >> "$GITHUB_ENV"
RESULT_FILENAME=$(python3 -m infx.results.result_filename)
export RESULT_FILENAME
export GPU_COUNT=$((TP * PP_SIZE * PCP_SIZE))
if [[ "$SCENARIO_TYPE" == video-generation ]]; then
export GPU_COUNT="$VIDEO_GPUS"
export CONC_LIST="$VIDEO_CONC_LIST"
echo "CONC_LIST=${CONC_LIST}" >> "$GITHUB_ENV"
else
export GPU_COUNT=$((TP * PP_SIZE * PCP_SIZE))
fi
echo "GPU_COUNT=${GPU_COUNT}" >> "$GITHUB_ENV"

echo "RESULT_FILENAME=${RESULT_FILENAME}" >> "$GITHUB_ENV"
Expand Down Expand Up @@ -323,8 +340,46 @@ jobs:
results/aiperf_artifacts \
--failed-request-threshold "$AIPERF_FAILED_REQUEST_THRESHOLD"

- name: Upload video results
id: upload-video-results
if: ${{ always() && inputs.scenario-type == 'video-generation' && env.RESULT_FILENAME != '' }}
continue-on-error: true
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: ${{ env.VIDEO_ARTIFACT_NAME }}
path: ${{ env.INFERENCEX_E2E_ROOT }}/h3-results
if-no-files-found: error
Comment thread
edwingao28 marked this conversation as resolved.
Resolved

- name: Wait before retrying upload video results
if: ${{ always() && steps.upload-video-results.outcome == 'failure' }}
run: sleep 10

- name: Retry upload video results (attempt 2)
id: retry-upload-video-results
if: ${{ always() && steps.upload-video-results.outcome == 'failure' }}
continue-on-error: true
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
overwrite: true
name: ${{ env.VIDEO_ARTIFACT_NAME }}
path: ${{ env.INFERENCEX_E2E_ROOT }}/h3-results
if-no-files-found: error

- name: Wait before final video results upload attempt
if: ${{ always() && steps.retry-upload-video-results.outcome == 'failure' }}
run: sleep 30

- name: Retry upload video results (attempt 3)
if: ${{ always() && steps.retry-upload-video-results.outcome == 'failure' }}
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
overwrite: true
name: ${{ env.VIDEO_ARTIFACT_NAME }}
path: ${{ env.INFERENCEX_E2E_ROOT }}/h3-results
if-no-files-found: error

- name: Process result
if: ${{ always() && env.RESULT_FILENAME != '' && !inputs.eval-only && inputs.scenario-type != 'agentic-coding' }}
if: ${{ always() && env.RESULT_FILENAME != '' && !inputs.eval-only && inputs.scenario-type != 'agentic-coding' && inputs.scenario-type != 'video-generation' }}
env:
RUNNER_TYPE: ${{ inputs.runner }}
PYTHONPATH: ${{ github.workspace }}/.result-tooling/inferencex-e2e
Expand All @@ -343,7 +398,7 @@ jobs:

- name: Upload result
id: upload-result
if: ${{ success() && env.RESULT_FILENAME != '' && !inputs.eval-only && inputs.scenario-type != 'agentic-coding' }}
if: ${{ success() && env.RESULT_FILENAME != '' && !inputs.eval-only && inputs.scenario-type != 'agentic-coding' && inputs.scenario-type != 'video-generation' }}
continue-on-error: true
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
Expand Down Expand Up @@ -468,6 +523,7 @@ jobs:
${{ env.INFERENCEX_E2E_ROOT }}/results/*.log
${{ env.INFERENCEX_E2E_ROOT }}/results/*_config.json
${{ env.INFERENCEX_E2E_ROOT }}/srt-single-node-logs.tar.gz
${{ env.INFERENCEX_E2E_ROOT }}/srt-video-logs.tar.gz
${{ env.INFERENCEX_E2E_ROOT }}/srt-single-node-submission.json
${{ env.INFERENCEX_E2E_ROOT }}/srt-slurm-sha.txt
${{ env.INFERENCEX_E2E_ROOT }}/srt-setup.log
Expand All @@ -483,9 +539,13 @@ jobs:
${{ env.INFERENCEX_E2E_ROOT }}/${{ env.RESULT_FILENAME }}.json
${{ env.INFERENCEX_E2E_ROOT }}/agg_${{ env.RESULT_FILENAME }}.json
${{ env.INFERENCEX_E2E_ROOT }}/power_validation_${{ env.RESULT_FILENAME }}.json
${{ env.INFERENCEX_E2E_ROOT }}/${{ env.RESULT_FILENAME }}_conc*.json
${{ env.INFERENCEX_E2E_ROOT }}/power_validation_${{ env.RESULT_FILENAME }}_conc*.json
${{ env.INFERENCEX_E2E_ROOT }}/results/agentic_power_timezone_offset.txt
${{ env.INFERENCEX_E2E_ROOT }}/LOGS/power/**
${{ env.INFERENCEX_E2E_ROOT }}/LOGS/agentic/**/agentic_power_concurrency_*.json
${{ env.INFERENCEX_E2E_ROOT }}/LOGS/video/**/benchmark_concurrency_*.json
${{ env.INFERENCEX_E2E_ROOT }}/LOGS/video/server/*.json
${{ env.INFERENCEX_E2E_ROOT }}/LOGS/${{ env.RESULT_FILENAME }}.json
${{ env.INFERENCEX_E2E_ROOT }}/power-producer-sha.txt
${{ env.INFERENCEX_E2E_ROOT }}/exporter-image.sha256
Expand Down
Loading
Loading