Stage
Disqualified
11 of 12 stages
Settled in
5h 32m
committed Oct 14, 12:09 UTC
Round score
—
awaiting round
p99 latency
—
ceiling 2000 / 50 ms
Round 1: disqualified · no score
11/12 stages
Intake & gates
6/6Patch hash recorded; awaiting fetch.
{
"hotkey": "5FKgasNZNKz62SgTmHuA9rahqiAnLQDoUMjUwy6okp9UEyNs",
"payment_tx": 3,
"commit_block": 9281428,
"payment_block": 9281425,
"patch_fingerprint": "sha256:ac00b9ada1ea9331f978e69c9f8716024a194ed82bab43b48dbaa599f3dd29ff"
}A worker claimed the submission and started the pipeline.
{
"job_id": 256
}Patch artifact retrieved from object storage.
{
"size": 5442,
"patch_hash": "sha256:1fb3e2584862d21156d81c01b3278e2efbc1d15aaebbdf1d826aeba57935290b"
}Integrity and path constraints checked.
{
"identity": "ok",
"patch_hash": "sha256:1fb3e2584862d21156d81c01b3278e2efbc1d15aaebbdf1d826aeba57935290b"
}Patch applied onto the baseline checkout.
{
"ok": true
}Public API / surface compatibility held.
{
"files": [
"python/sglang/srt/models/dspark.py",
"python/sglang/srt/speculative/dspark_components/dspark_draft_sampler.py"
]
}Build
3/3Hermetic engine image build in progress.
No structured detail recorded for this step.
Candidate engine image pushed to the registry.
{
"image_ref": "ghcr.io/pareton-ai/pareton-engine@sha256:8a9bd62f883415f3055bf3abb546fd66c5aab09de0450911ff7875059aa5438c",
"image_digest": "sha256:8a9bd62f883415f3055bf3abb546fd66c5aab09de0450911ff7875059aa5438c"
}Candidate engine image built and digest pinned.
{
"mock": false,
"build_log": "/var/log/pareton/builds/d90b61bc-53d6-4e1e-9d46-4ff14d0f8c33/build.log",
"image_ref": "ghcr.io/pareton-ai/pareton-engine@sha256:8a9bd62f883415f3055bf3abb546fd66c5aab09de0450911ff7875059aa5438c",
"image_tag": "ghcr.io/pareton-ai/pareton-engine:1fb3e2584862d21156d81c01b3278e2efbc1d15aaebbdf1d826aeba57935290b"
}Round
2/3Waiting to be seated in a round.
No structured detail recorded for this step.
Seated in a round; waiting to run or running.
{
"ordinal": 1,
"round_id": "f987153b-7853-4445-951e-edd9dd501348"
}Other events
The image ran and produced wrong output.
{
"score": null,
"reason": "engine died before becoming healthy (last error: [Errno 111] Connection refused); container log tail:\ne_shape\n maybe_flashinfer_autotune_speculative_draft(\n File \"/src/python/sglang/srt/model_executor/runner/flashinfer_autotune.py\", line 339, in maybe_flashinfer_autotune_speculative_draft\n run_flashinfer_autotune_forward(mr, run_and_reset, run_lm_head=run_lm_head)\n File \"/src/python/sglang/srt/model_executor/runner/flashinfer_autotune.py\", line 308, in run_flashinfer_autotune_forward\n forward_fn()\n File \"/src/python/sglang/srt/model_executor/runner/flashinfer_autotune.py\", line 335, in run_and_reset\n forward_fn()\n File \"/src/python/sglang/srt/model_executor/runner/decode_cuda_graph_runner.py\", line 1265, in run_once\n capture_hook(self, out, forward_batch, num_tokens)\n File \"/src/python/sglang/srt/speculative/draft_worker_common.py\", line 150, in capture_hook\n draft_sampler(out.hidden_states, forward_batch.input_ids)\n File \"/src/python/sglang/srt/speculative/dspark_components/dspark_draft_sampler.py\", line 179, in __call__\n draft_tokens, corrected_logits = self.markov_head.sample_block(\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/src/python/sglang/srt/models/dspark.py\", line 175, in sample_block\n return run_markov_block(\n ^^^^^^^^^^^^^^^^^\n File \"/src/python/sglang/srt/models/dspark.py\", line 95, in run_markov_block\n step_logits = torch.add(\n ^^^^^^^^^^\nRuntimeError: add(): functions with out=... arguments don't support automatic differentiation, but one of the arguments requires grad.\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n File \"/src/python/sglang/srt/managers/scheduler.py\", line 5956, in run_scheduler_process\n scheduler = Scheduler(server_args, port_args)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/src/python/sglang/srt/managers/scheduler.py\", line 560, in __init__\n self.init_model_worker()\n File \"/src/python/sglang/srt/managers/scheduler.py\", line 1099, in init_model_worker\n self.init_all_cuda_graphs()\n File \"/src/python/sglang/srt/managers/scheduler.py\", line 1084, in init_all_cuda_graphs\n self.draft_worker.init_cuda_graphs()\n File \"/src/python/sglang/srt/speculative/dspark_components/dspark_worker_v2.py\", line 515, in init_cuda_graphs\n self._draft_worker.init_cuda_graphs(\n File \"/src/python/sglang/srt/managers/tp_worker.py\", line 425, in init_cuda_graphs\n self.model_runner.init_cuda_graphs(\n File \"/src/python/sglang/srt/model_executor/model_runner.py\", line 1130, in init_cuda_graphs\n capture = capture_cuda_graphs(\n ^^^^^^^^^^^^^^^^^^^^\n File \"/src/python/sglang/srt/model_executor/model_runner_components/cuda_graph_setup.py\", line 412, in capture_cuda_graphs\n decode = capture_decode_graph(model_runner=model_runner)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/src/python/sglang/srt/model_executor/model_runner_components/cuda_graph_setup.py\", line 737, in capture_decode_graph\n runner = graph_runners[model_runner.device](model_runner)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File \"/src/python/sglang/srt/model_executor/runner/decode_cuda_graph_runner.py\", line 513, in __init__\n raise Exception(\nException: Capture cuda graph failed: add(): functions with out=... arguments don't support automatic differentiation, but one of the arguments requires grad.\nPossible solutions:\n1. set --mem-fraction-static to a smaller value (e.g., 0.8 or 0.7)\n2. set --cuda-graph-max-bs-decode to a smaller value (e.g., 16)\n3. disable decode CUDA graph by --cuda-graph-backend-decode=disabled. (Not recommended. Huge performance loss)\nOpen an issue on GitHub https://github.com/sgl-project/sglang/issues/new/choose \n\n\n[2026-10-14 16:31:34] Received sigquit from a child process. It usually means the child failed.\n[2026-10-14 16:31:34] kill_process_tree called: parent_pid=1, include_parent=True, pid=1\n[2026-10-14 16:31:34] kill_process_tree called: parent_pid=1, include_parent=False, pid=1\n",
"ordinal": 1,
"round_id": "f987153b-7853-4445-951e-edd9dd501348"
}