Update judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone (20261007-082946-341658)
Browse files- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/configs/20261007-082946-341658.json +41 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/judgments_anthropic-claude-opus-5-5/20261007-082946-341658.jsonl +200 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/logs/vllm_hf-Bigglz-qwen-sft-sft-dpo-tone_0.log +294 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/logs/vllm_hf-Bigglz-qwen-sft-sft-dpo-tone_1.log +270 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/logs/vllm_hf-models-mistral_small4-sft-sft-dpo-native_0.log +552 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/review_anthropic-claude-opus-5-5/MANGO.md +0 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/summary_anthropic-claude-opus-5-5.json +59 -0
- judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/summary_anthropic-claude-opus-5-5.md +18 -0
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/configs/20261007-082946-341658.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_a": "hf:models/mistral_small4-sft-sft-dpo-native",
|
| 3 |
+
"model_b": "hf:Bigglz/qwen-sft-sft-dpo-tone",
|
| 4 |
+
"judge_model": "anthropic:claude-opus-5-5",
|
| 5 |
+
"user_model": "anthropic:claude-opus-5-5",
|
| 6 |
+
"supporters": [
|
| 7 |
+
"MANGO"
|
| 8 |
+
],
|
| 9 |
+
"supporters_b": null,
|
| 10 |
+
"turns": 3,
|
| 11 |
+
"limit": null,
|
| 12 |
+
"sample": null,
|
| 13 |
+
"scenario_ids": null,
|
| 14 |
+
"scenario_file": "prompts/conv/scenario_sets/eval100.txt",
|
| 15 |
+
"username": "USERNAME",
|
| 16 |
+
"temperature": 0.7,
|
| 17 |
+
"max_tokens": 256,
|
| 18 |
+
"user_temperature": 0.7,
|
| 19 |
+
"user_max_tokens": 64,
|
| 20 |
+
"judge_temperature": 0.0,
|
| 21 |
+
"judge_max_tokens": 1024,
|
| 22 |
+
"reasoning_effort": null,
|
| 23 |
+
"seed": 42,
|
| 24 |
+
"concurrency": 32,
|
| 25 |
+
"max_active_conversations": null,
|
| 26 |
+
"gpus": "0,1,2,3,4,5",
|
| 27 |
+
"hf_tp": null,
|
| 28 |
+
"hf_max_replicas": null,
|
| 29 |
+
"hf_replica_concurrency": 128,
|
| 30 |
+
"hf_max_model_len": 8192,
|
| 31 |
+
"hf_gpu_memory_utilization": 0.9,
|
| 32 |
+
"hf_base_port": 18000,
|
| 33 |
+
"hf_startup_timeout": 3600,
|
| 34 |
+
"name_a": "Bigglz/mistral-small4-119b-sft-sft-dpo",
|
| 35 |
+
"name_b": null,
|
| 36 |
+
"output_dir": null,
|
| 37 |
+
"stage": "all",
|
| 38 |
+
"n_scenarios": 100,
|
| 39 |
+
"generation_hash": "db024c403f",
|
| 40 |
+
"started_at": "2026-10-07 08:29:46"
|
| 41 |
+
}
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/judgments_anthropic-claude-opus-5-5/20261007-082946-341658.jsonl
ADDED
|
@@ -0,0 +1,200 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"supporter": "MANGO", "scenario_id": 363, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e55403578578", "created_at": "2026-10-07 08:35:27"}
|
| 2 |
+
{"supporter": "MANGO", "scenario_id": 633, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4f6e0a19b3e4", "created_at": "2026-10-07 08:35:27"}
|
| 3 |
+
{"supporter": "MANGO", "scenario_id": 721, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d0e5a94d3637", "created_at": "2026-10-07 08:35:28"}
|
| 4 |
+
{"supporter": "MANGO", "scenario_id": 363, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2aa01c2bd7b7", "created_at": "2026-10-07 08:35:28"}
|
| 5 |
+
{"supporter": "MANGO", "scenario_id": 721, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "208f3c309160", "created_at": "2026-10-07 08:35:28"}
|
| 6 |
+
{"supporter": "MANGO", "scenario_id": 6, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "fb02f659061e", "created_at": "2026-10-07 08:35:28"}
|
| 7 |
+
{"supporter": "MANGO", "scenario_id": 127, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "8deae8ec0607", "created_at": "2026-10-07 08:35:28"}
|
| 8 |
+
{"supporter": "MANGO", "scenario_id": 633, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4d3e5270a7fa", "created_at": "2026-10-07 08:35:29"}
|
| 9 |
+
{"supporter": "MANGO", "scenario_id": 6, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3fb03ecb04d4", "created_at": "2026-10-07 08:35:29"}
|
| 10 |
+
{"supporter": "MANGO", "scenario_id": 127, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "0866a2c50a7e", "created_at": "2026-10-07 08:35:29"}
|
| 11 |
+
{"supporter": "MANGO", "scenario_id": 73, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ae11fb4472a1", "created_at": "2026-10-07 08:35:29"}
|
| 12 |
+
{"supporter": "MANGO", "scenario_id": 73, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "aaa4a560543f", "created_at": "2026-10-07 08:35:29"}
|
| 13 |
+
{"supporter": "MANGO", "scenario_id": 214, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "64101b0b9827", "created_at": "2026-10-07 08:35:29"}
|
| 14 |
+
{"supporter": "MANGO", "scenario_id": 225, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3577c7ee86cb", "created_at": "2026-10-07 08:35:30"}
|
| 15 |
+
{"supporter": "MANGO", "scenario_id": 225, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "5866730484a5", "created_at": "2026-10-07 08:35:30"}
|
| 16 |
+
{"supporter": "MANGO", "scenario_id": 665, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9733ca25c53e", "created_at": "2026-10-07 08:35:30"}
|
| 17 |
+
{"supporter": "MANGO", "scenario_id": 25, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4166f6052025", "created_at": "2026-10-07 08:35:30"}
|
| 18 |
+
{"supporter": "MANGO", "scenario_id": 203, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "f8a624877807", "created_at": "2026-10-07 08:35:30"}
|
| 19 |
+
{"supporter": "MANGO", "scenario_id": 665, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "8dd7e13fe481", "created_at": "2026-10-07 08:35:30"}
|
| 20 |
+
{"supporter": "MANGO", "scenario_id": 352, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "fcc7e15dc516", "created_at": "2026-10-07 08:35:30"}
|
| 21 |
+
{"supporter": "MANGO", "scenario_id": 25, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c559e74efc38", "created_at": "2026-10-07 08:35:30"}
|
| 22 |
+
{"supporter": "MANGO", "scenario_id": 203, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "6dced0b33de2", "created_at": "2026-10-07 08:35:30"}
|
| 23 |
+
{"supporter": "MANGO", "scenario_id": 352, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3988c96e4bb9", "created_at": "2026-10-07 08:35:30"}
|
| 24 |
+
{"supporter": "MANGO", "scenario_id": 389, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "90db2585ef07", "created_at": "2026-10-07 08:35:30"}
|
| 25 |
+
{"supporter": "MANGO", "scenario_id": 80, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "018ccabd52bb", "created_at": "2026-10-07 08:35:31"}
|
| 26 |
+
{"supporter": "MANGO", "scenario_id": 470, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "94269cccb391", "created_at": "2026-10-07 08:35:31"}
|
| 27 |
+
{"supporter": "MANGO", "scenario_id": 603, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "f974d49f8913", "created_at": "2026-10-07 08:35:31"}
|
| 28 |
+
{"supporter": "MANGO", "scenario_id": 847, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "789b7b1b8847", "created_at": "2026-10-07 08:35:31"}
|
| 29 |
+
{"supporter": "MANGO", "scenario_id": 80, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "61949ca3f2e8", "created_at": "2026-10-07 08:35:31"}
|
| 30 |
+
{"supporter": "MANGO", "scenario_id": 214, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "b3617d847b70", "created_at": "2026-10-07 08:35:31"}
|
| 31 |
+
{"supporter": "MANGO", "scenario_id": 847, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "efb852bb0109", "created_at": "2026-10-07 08:35:31"}
|
| 32 |
+
{"supporter": "MANGO", "scenario_id": 517, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3246c84bca2b", "created_at": "2026-10-07 08:35:31"}
|
| 33 |
+
{"supporter": "MANGO", "scenario_id": 603, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ba66f4ab9719", "created_at": "2026-10-07 08:35:31"}
|
| 34 |
+
{"supporter": "MANGO", "scenario_id": 863, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ce71988c2968", "created_at": "2026-10-07 08:35:31"}
|
| 35 |
+
{"supporter": "MANGO", "scenario_id": 699, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3e6b755c22ca", "created_at": "2026-10-07 08:35:31"}
|
| 36 |
+
{"supporter": "MANGO", "scenario_id": 699, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4e28f989115a", "created_at": "2026-10-07 08:35:31"}
|
| 37 |
+
{"supporter": "MANGO", "scenario_id": 692, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ff5b86eb1606", "created_at": "2026-10-07 08:35:31"}
|
| 38 |
+
{"supporter": "MANGO", "scenario_id": 389, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9c34363ed9c3", "created_at": "2026-10-07 08:35:31"}
|
| 39 |
+
{"supporter": "MANGO", "scenario_id": 781, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "abc41330c0dd", "created_at": "2026-10-07 08:35:31"}
|
| 40 |
+
{"supporter": "MANGO", "scenario_id": 677, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "28280934d1b4", "created_at": "2026-10-07 08:35:31"}
|
| 41 |
+
{"supporter": "MANGO", "scenario_id": 692, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ebea214bbbf5", "created_at": "2026-10-07 08:35:31"}
|
| 42 |
+
{"supporter": "MANGO", "scenario_id": 677, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "83c4c1ecf69a", "created_at": "2026-10-07 08:35:32"}
|
| 43 |
+
{"supporter": "MANGO", "scenario_id": 781, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "f77128f09c5f", "created_at": "2026-10-07 08:35:32"}
|
| 44 |
+
{"supporter": "MANGO", "scenario_id": 863, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "201c27c150da", "created_at": "2026-10-07 08:35:32"}
|
| 45 |
+
{"supporter": "MANGO", "scenario_id": 885, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "645884d28a9d", "created_at": "2026-10-07 08:35:32"}
|
| 46 |
+
{"supporter": "MANGO", "scenario_id": 300, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "243e53a18f8c", "created_at": "2026-10-07 08:35:32"}
|
| 47 |
+
{"supporter": "MANGO", "scenario_id": 300, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7fed2791f584", "created_at": "2026-10-07 08:35:32"}
|
| 48 |
+
{"supporter": "MANGO", "scenario_id": 885, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c3fc0ce30d1a", "created_at": "2026-10-07 08:35:32"}
|
| 49 |
+
{"supporter": "MANGO", "scenario_id": 27, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "cb345bc024c6", "created_at": "2026-10-07 08:35:32"}
|
| 50 |
+
{"supporter": "MANGO", "scenario_id": 367, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "539da804abbc", "created_at": "2026-10-07 08:35:32"}
|
| 51 |
+
{"supporter": "MANGO", "scenario_id": 95, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "11107f529f39", "created_at": "2026-10-07 08:35:32"}
|
| 52 |
+
{"supporter": "MANGO", "scenario_id": 686, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3c83221c64b8", "created_at": "2026-10-07 08:35:32"}
|
| 53 |
+
{"supporter": "MANGO", "scenario_id": 367, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "5f1f4218ee2d", "created_at": "2026-10-07 08:35:33"}
|
| 54 |
+
{"supporter": "MANGO", "scenario_id": 95, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "8778d5cc153b", "created_at": "2026-10-07 08:35:33"}
|
| 55 |
+
{"supporter": "MANGO", "scenario_id": 517, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9f42816eea09", "created_at": "2026-10-07 08:35:33"}
|
| 56 |
+
{"supporter": "MANGO", "scenario_id": 196, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2994cbc4a420", "created_at": "2026-10-07 08:35:33"}
|
| 57 |
+
{"supporter": "MANGO", "scenario_id": 196, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "661056a05b3e", "created_at": "2026-10-07 08:35:33"}
|
| 58 |
+
{"supporter": "MANGO", "scenario_id": 747, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c70051c33804", "created_at": "2026-10-07 08:35:33"}
|
| 59 |
+
{"supporter": "MANGO", "scenario_id": 747, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "5dd2ae0659f8", "created_at": "2026-10-07 08:35:33"}
|
| 60 |
+
{"supporter": "MANGO", "scenario_id": 464, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "391e1ffdbdab", "created_at": "2026-10-07 08:35:33"}
|
| 61 |
+
{"supporter": "MANGO", "scenario_id": 233, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4f7a7409d0b0", "created_at": "2026-10-07 08:35:34"}
|
| 62 |
+
{"supporter": "MANGO", "scenario_id": 27, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "11b77764e370", "created_at": "2026-10-07 08:35:34"}
|
| 63 |
+
{"supporter": "MANGO", "scenario_id": 429, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7518b7a73e0f", "created_at": "2026-10-07 08:35:34"}
|
| 64 |
+
{"supporter": "MANGO", "scenario_id": 114, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "04c3669b495c", "created_at": "2026-10-07 08:35:34"}
|
| 65 |
+
{"supporter": "MANGO", "scenario_id": 869, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e9aff1066ec2", "created_at": "2026-10-07 08:35:34"}
|
| 66 |
+
{"supporter": "MANGO", "scenario_id": 686, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "10c92744d59e", "created_at": "2026-10-07 08:35:34"}
|
| 67 |
+
{"supporter": "MANGO", "scenario_id": 718, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "673fcfdb55e5", "created_at": "2026-10-07 08:35:34"}
|
| 68 |
+
{"supporter": "MANGO", "scenario_id": 432, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c6afc5311bd1", "created_at": "2026-10-07 08:35:34"}
|
| 69 |
+
{"supporter": "MANGO", "scenario_id": 233, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ada1145d3e72", "created_at": "2026-10-07 08:35:34"}
|
| 70 |
+
{"supporter": "MANGO", "scenario_id": 869, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d14b12673119", "created_at": "2026-10-07 08:35:34"}
|
| 71 |
+
{"supporter": "MANGO", "scenario_id": 464, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "1a71169d195a", "created_at": "2026-10-07 08:35:34"}
|
| 72 |
+
{"supporter": "MANGO", "scenario_id": 429, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "30f074282b44", "created_at": "2026-10-07 08:35:34"}
|
| 73 |
+
{"supporter": "MANGO", "scenario_id": 777, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3f636d511a62", "created_at": "2026-10-07 08:35:34"}
|
| 74 |
+
{"supporter": "MANGO", "scenario_id": 777, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d53d7ff0977a", "created_at": "2026-10-07 08:35:34"}
|
| 75 |
+
{"supporter": "MANGO", "scenario_id": 432, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "267a10d0b018", "created_at": "2026-10-07 08:35:34"}
|
| 76 |
+
{"supporter": "MANGO", "scenario_id": 825, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "99237b684da8", "created_at": "2026-10-07 08:35:34"}
|
| 77 |
+
{"supporter": "MANGO", "scenario_id": 238, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "0021d23fb7dc", "created_at": "2026-10-07 08:35:35"}
|
| 78 |
+
{"supporter": "MANGO", "scenario_id": 238, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7e620c770514", "created_at": "2026-10-07 08:35:35"}
|
| 79 |
+
{"supporter": "MANGO", "scenario_id": 114, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "eb68e9a3e923", "created_at": "2026-10-07 08:35:35"}
|
| 80 |
+
{"supporter": "MANGO", "scenario_id": 281, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "64addb417a01", "created_at": "2026-10-07 08:35:35"}
|
| 81 |
+
{"supporter": "MANGO", "scenario_id": 273, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9ca50fa2c2c5", "created_at": "2026-10-07 08:35:35"}
|
| 82 |
+
{"supporter": "MANGO", "scenario_id": 273, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3e6331f276e6", "created_at": "2026-10-07 08:35:35"}
|
| 83 |
+
{"supporter": "MANGO", "scenario_id": 281, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "6c451a83d942", "created_at": "2026-10-07 08:35:35"}
|
| 84 |
+
{"supporter": "MANGO", "scenario_id": 718, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "6efedb0d7006", "created_at": "2026-10-07 08:35:35"}
|
| 85 |
+
{"supporter": "MANGO", "scenario_id": 166, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e570c9bf82ba", "created_at": "2026-10-07 08:35:36"}
|
| 86 |
+
{"supporter": "MANGO", "scenario_id": 223, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "a59fb289d1f7", "created_at": "2026-10-07 08:35:36"}
|
| 87 |
+
{"supporter": "MANGO", "scenario_id": 103, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "162ac75d21fc", "created_at": "2026-10-07 08:35:36"}
|
| 88 |
+
{"supporter": "MANGO", "scenario_id": 103, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "403ccea8bce7", "created_at": "2026-10-07 08:35:36"}
|
| 89 |
+
{"supporter": "MANGO", "scenario_id": 758, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "df56dd909ace", "created_at": "2026-10-07 08:35:36"}
|
| 90 |
+
{"supporter": "MANGO", "scenario_id": 758, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2dee91a6db8f", "created_at": "2026-10-07 08:35:36"}
|
| 91 |
+
{"supporter": "MANGO", "scenario_id": 650, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "b0453b4b36d0", "created_at": "2026-10-07 08:35:36"}
|
| 92 |
+
{"supporter": "MANGO", "scenario_id": 650, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "8b9d8f64e635", "created_at": "2026-10-07 08:35:36"}
|
| 93 |
+
{"supporter": "MANGO", "scenario_id": 565, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "6a02e7a69403", "created_at": "2026-10-07 08:35:36"}
|
| 94 |
+
{"supporter": "MANGO", "scenario_id": 344, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "aa8bb0570cc2", "created_at": "2026-10-07 08:35:36"}
|
| 95 |
+
{"supporter": "MANGO", "scenario_id": 223, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "1b5683fc1257", "created_at": "2026-10-07 08:35:36"}
|
| 96 |
+
{"supporter": "MANGO", "scenario_id": 546, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "1988c62e1cfe", "created_at": "2026-10-07 08:35:36"}
|
| 97 |
+
{"supporter": "MANGO", "scenario_id": 565, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e185cdf6b1fd", "created_at": "2026-10-07 08:35:36"}
|
| 98 |
+
{"supporter": "MANGO", "scenario_id": 877, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "0bff1fbe7971", "created_at": "2026-10-07 08:35:37"}
|
| 99 |
+
{"supporter": "MANGO", "scenario_id": 344, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4070e86164b8", "created_at": "2026-10-07 08:35:37"}
|
| 100 |
+
{"supporter": "MANGO", "scenario_id": 825, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "107bef051c6e", "created_at": "2026-10-07 08:35:37"}
|
| 101 |
+
{"supporter": "MANGO", "scenario_id": 228, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "995498e9c750", "created_at": "2026-10-07 08:35:37"}
|
| 102 |
+
{"supporter": "MANGO", "scenario_id": 618, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4d08eb534c32", "created_at": "2026-10-07 08:35:37"}
|
| 103 |
+
{"supporter": "MANGO", "scenario_id": 877, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3ea264c91f6d", "created_at": "2026-10-07 08:35:37"}
|
| 104 |
+
{"supporter": "MANGO", "scenario_id": 618, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e941ab3be53a", "created_at": "2026-10-07 08:35:37"}
|
| 105 |
+
{"supporter": "MANGO", "scenario_id": 228, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "a603942c268a", "created_at": "2026-10-07 08:35:37"}
|
| 106 |
+
{"supporter": "MANGO", "scenario_id": 159, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2e4a164e3820", "created_at": "2026-10-07 08:35:37"}
|
| 107 |
+
{"supporter": "MANGO", "scenario_id": 470, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2b7eaafbec17", "created_at": "2026-10-07 08:35:37"}
|
| 108 |
+
{"supporter": "MANGO", "scenario_id": 159, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "66df08b55e45", "created_at": "2026-10-07 08:35:37"}
|
| 109 |
+
{"supporter": "MANGO", "scenario_id": 813, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9d321e2df40a", "created_at": "2026-10-07 08:35:38"}
|
| 110 |
+
{"supporter": "MANGO", "scenario_id": 813, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9a478676ee5a", "created_at": "2026-10-07 08:35:38"}
|
| 111 |
+
{"supporter": "MANGO", "scenario_id": 89, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c0abcb2a5d74", "created_at": "2026-10-07 08:35:38"}
|
| 112 |
+
{"supporter": "MANGO", "scenario_id": 373, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "800abf7a6766", "created_at": "2026-10-07 08:35:38"}
|
| 113 |
+
{"supporter": "MANGO", "scenario_id": 166, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "213027efece8", "created_at": "2026-10-07 08:35:38"}
|
| 114 |
+
{"supporter": "MANGO", "scenario_id": 754, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "388283be6160", "created_at": "2026-10-07 08:35:38"}
|
| 115 |
+
{"supporter": "MANGO", "scenario_id": 754, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e5deb5d6b71f", "created_at": "2026-10-07 08:35:38"}
|
| 116 |
+
{"supporter": "MANGO", "scenario_id": 373, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c4888941f03f", "created_at": "2026-10-07 08:35:38"}
|
| 117 |
+
{"supporter": "MANGO", "scenario_id": 714, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7e7753e0aa1d", "created_at": "2026-10-07 08:35:38"}
|
| 118 |
+
{"supporter": "MANGO", "scenario_id": 296, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "8832dc981be6", "created_at": "2026-10-07 08:35:39"}
|
| 119 |
+
{"supporter": "MANGO", "scenario_id": 604, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d4cf7d150e19", "created_at": "2026-10-07 08:35:39"}
|
| 120 |
+
{"supporter": "MANGO", "scenario_id": 604, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7295c74cf002", "created_at": "2026-10-07 08:35:39"}
|
| 121 |
+
{"supporter": "MANGO", "scenario_id": 616, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "fd6cd6cd779d", "created_at": "2026-10-07 08:35:39"}
|
| 122 |
+
{"supporter": "MANGO", "scenario_id": 558, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "08766b3a2c16", "created_at": "2026-10-07 08:35:39"}
|
| 123 |
+
{"supporter": "MANGO", "scenario_id": 459, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e70b75ef375f", "created_at": "2026-10-07 08:35:39"}
|
| 124 |
+
{"supporter": "MANGO", "scenario_id": 663, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "a39745341f53", "created_at": "2026-10-07 08:35:39"}
|
| 125 |
+
{"supporter": "MANGO", "scenario_id": 558, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "eb5f626d2d20", "created_at": "2026-10-07 08:35:39"}
|
| 126 |
+
{"supporter": "MANGO", "scenario_id": 714, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "65df6ef47880", "created_at": "2026-10-07 08:35:39"}
|
| 127 |
+
{"supporter": "MANGO", "scenario_id": 663, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "85855bf7ecef", "created_at": "2026-10-07 08:35:39"}
|
| 128 |
+
{"supporter": "MANGO", "scenario_id": 387, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d0dcff143a0a", "created_at": "2026-10-07 08:35:39"}
|
| 129 |
+
{"supporter": "MANGO", "scenario_id": 44, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "**K**", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "de57ad3888ed", "created_at": "2026-10-07 08:35:39"}
|
| 130 |
+
{"supporter": "MANGO", "scenario_id": 387, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "fff7cce98ae0", "created_at": "2026-10-07 08:35:39"}
|
| 131 |
+
{"supporter": "MANGO", "scenario_id": 616, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "75e99922734f", "created_at": "2026-10-07 08:35:40"}
|
| 132 |
+
{"supporter": "MANGO", "scenario_id": 591, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "863e188fa188", "created_at": "2026-10-07 08:35:40"}
|
| 133 |
+
{"supporter": "MANGO", "scenario_id": 94, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7f55d36dda89", "created_at": "2026-10-07 08:35:40"}
|
| 134 |
+
{"supporter": "MANGO", "scenario_id": 826, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "15a79a1f0732", "created_at": "2026-10-07 08:35:40"}
|
| 135 |
+
{"supporter": "MANGO", "scenario_id": 296, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "acdb9abf9669", "created_at": "2026-10-07 08:35:40"}
|
| 136 |
+
{"supporter": "MANGO", "scenario_id": 546, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "1f519d663607", "created_at": "2026-10-07 08:35:40"}
|
| 137 |
+
{"supporter": "MANGO", "scenario_id": 71, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7b1088f555d0", "created_at": "2026-10-07 08:35:40"}
|
| 138 |
+
{"supporter": "MANGO", "scenario_id": 574, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "58e374d60ff2", "created_at": "2026-10-07 08:35:40"}
|
| 139 |
+
{"supporter": "MANGO", "scenario_id": 99, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4077049bafbd", "created_at": "2026-10-07 08:35:40"}
|
| 140 |
+
{"supporter": "MANGO", "scenario_id": 89, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "1c59fca41385", "created_at": "2026-10-07 08:35:40"}
|
| 141 |
+
{"supporter": "MANGO", "scenario_id": 99, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "85c07b17c308", "created_at": "2026-10-07 08:35:40"}
|
| 142 |
+
{"supporter": "MANGO", "scenario_id": 574, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "20265da5bae7", "created_at": "2026-10-07 08:35:40"}
|
| 143 |
+
{"supporter": "MANGO", "scenario_id": 459, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "63b1566e80be", "created_at": "2026-10-07 08:35:40"}
|
| 144 |
+
{"supporter": "MANGO", "scenario_id": 888, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ff278785c3f2", "created_at": "2026-10-07 08:35:41"}
|
| 145 |
+
{"supporter": "MANGO", "scenario_id": 643, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "0025769299b7", "created_at": "2026-10-07 08:35:41"}
|
| 146 |
+
{"supporter": "MANGO", "scenario_id": 71, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "00811903cbe9", "created_at": "2026-10-07 08:35:41"}
|
| 147 |
+
{"supporter": "MANGO", "scenario_id": 94, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "56ce5c6afc7e", "created_at": "2026-10-07 08:35:41"}
|
| 148 |
+
{"supporter": "MANGO", "scenario_id": 643, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c433edf77639", "created_at": "2026-10-07 08:35:41"}
|
| 149 |
+
{"supporter": "MANGO", "scenario_id": 379, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3f2bc99f92d2", "created_at": "2026-10-07 08:35:41"}
|
| 150 |
+
{"supporter": "MANGO", "scenario_id": 81, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d920e8241c73", "created_at": "2026-10-07 08:35:41"}
|
| 151 |
+
{"supporter": "MANGO", "scenario_id": 654, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "60df3afb19a3", "created_at": "2026-10-07 08:35:41"}
|
| 152 |
+
{"supporter": "MANGO", "scenario_id": 591, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "bb73b7a35cc3", "created_at": "2026-10-07 08:35:41"}
|
| 153 |
+
{"supporter": "MANGO", "scenario_id": 888, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9479a4985681", "created_at": "2026-10-07 08:35:41"}
|
| 154 |
+
{"supporter": "MANGO", "scenario_id": 654, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "393015207ef5", "created_at": "2026-10-07 08:35:41"}
|
| 155 |
+
{"supporter": "MANGO", "scenario_id": 81, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2f4ed5440c18", "created_at": "2026-10-07 08:35:41"}
|
| 156 |
+
{"supporter": "MANGO", "scenario_id": 826, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ebc8dddcec5f", "created_at": "2026-10-07 08:35:42"}
|
| 157 |
+
{"supporter": "MANGO", "scenario_id": 759, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ca172a407584", "created_at": "2026-10-07 08:35:42"}
|
| 158 |
+
{"supporter": "MANGO", "scenario_id": 759, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "048cacd0037a", "created_at": "2026-10-07 08:35:42"}
|
| 159 |
+
{"supporter": "MANGO", "scenario_id": 370, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d197dce2a42f", "created_at": "2026-10-07 08:35:42"}
|
| 160 |
+
{"supporter": "MANGO", "scenario_id": 379, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3525a64092a3", "created_at": "2026-10-07 08:35:42"}
|
| 161 |
+
{"supporter": "MANGO", "scenario_id": 32, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "53bf91b398f9", "created_at": "2026-10-07 08:35:42"}
|
| 162 |
+
{"supporter": "MANGO", "scenario_id": 32, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "bc7ce94ef769", "created_at": "2026-10-07 08:35:42"}
|
| 163 |
+
{"supporter": "MANGO", "scenario_id": 270, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "5bd255b8b806", "created_at": "2026-10-07 08:35:42"}
|
| 164 |
+
{"supporter": "MANGO", "scenario_id": 270, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e8df7faa7cce", "created_at": "2026-10-07 08:35:42"}
|
| 165 |
+
{"supporter": "MANGO", "scenario_id": 854, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4f29fc2fc8f2", "created_at": "2026-10-07 08:35:42"}
|
| 166 |
+
{"supporter": "MANGO", "scenario_id": 854, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "df60d0ee53c9", "created_at": "2026-10-07 08:35:42"}
|
| 167 |
+
{"supporter": "MANGO", "scenario_id": 348, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "df20f63c0c67", "created_at": "2026-10-07 08:35:42"}
|
| 168 |
+
{"supporter": "MANGO", "scenario_id": 370, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "db96c4dd3c4c", "created_at": "2026-10-07 08:35:42"}
|
| 169 |
+
{"supporter": "MANGO", "scenario_id": 44, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d43c7654b703", "created_at": "2026-10-07 08:35:43"}
|
| 170 |
+
{"supporter": "MANGO", "scenario_id": 348, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7510d3ab54e5", "created_at": "2026-10-07 08:35:43"}
|
| 171 |
+
{"supporter": "MANGO", "scenario_id": 175, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "282e00a6310f", "created_at": "2026-10-07 08:35:43"}
|
| 172 |
+
{"supporter": "MANGO", "scenario_id": 284, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "**K**", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "9059522ea056", "created_at": "2026-10-07 08:35:43"}
|
| 173 |
+
{"supporter": "MANGO", "scenario_id": 46, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "f377bea8b752", "created_at": "2026-10-07 08:35:43"}
|
| 174 |
+
{"supporter": "MANGO", "scenario_id": 623, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "**J**", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e44b64207b21", "created_at": "2026-10-07 08:35:43"}
|
| 175 |
+
{"supporter": "MANGO", "scenario_id": 828, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "c787948ec559", "created_at": "2026-10-07 08:35:43"}
|
| 176 |
+
{"supporter": "MANGO", "scenario_id": 46, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "0e52662c3a51", "created_at": "2026-10-07 08:35:43"}
|
| 177 |
+
{"supporter": "MANGO", "scenario_id": 163, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "5650f6f882fa", "created_at": "2026-10-07 08:35:43"}
|
| 178 |
+
{"supporter": "MANGO", "scenario_id": 791, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "1d00dbdd7021", "created_at": "2026-10-07 08:35:43"}
|
| 179 |
+
{"supporter": "MANGO", "scenario_id": 175, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "dda70c313317", "created_at": "2026-10-07 08:35:43"}
|
| 180 |
+
{"supporter": "MANGO", "scenario_id": 828, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "10b3c6270924", "created_at": "2026-10-07 08:35:43"}
|
| 181 |
+
{"supporter": "MANGO", "scenario_id": 163, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "4003797f3f09", "created_at": "2026-10-07 08:35:43"}
|
| 182 |
+
{"supporter": "MANGO", "scenario_id": 104, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "817b5f1545ff", "created_at": "2026-10-07 08:35:44"}
|
| 183 |
+
{"supporter": "MANGO", "scenario_id": 220, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "10bfae10170a", "created_at": "2026-10-07 08:35:44"}
|
| 184 |
+
{"supporter": "MANGO", "scenario_id": 30, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d165f524c2be", "created_at": "2026-10-07 08:35:44"}
|
| 185 |
+
{"supporter": "MANGO", "scenario_id": 250, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "J", "choice": "J", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "ff77f4999525", "created_at": "2026-10-07 08:35:44"}
|
| 186 |
+
{"supporter": "MANGO", "scenario_id": 104, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "e53191767f3e", "created_at": "2026-10-07 08:35:44"}
|
| 187 |
+
{"supporter": "MANGO", "scenario_id": 549, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "22b187db0578", "created_at": "2026-10-07 08:35:44"}
|
| 188 |
+
{"supporter": "MANGO", "scenario_id": 142, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "00891f293a02", "created_at": "2026-10-07 08:35:44"}
|
| 189 |
+
{"supporter": "MANGO", "scenario_id": 623, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "b05abb2ede5c", "created_at": "2026-10-07 08:35:44"}
|
| 190 |
+
{"supporter": "MANGO", "scenario_id": 220, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "7d7f0c0f2144", "created_at": "2026-10-07 08:35:44"}
|
| 191 |
+
{"supporter": "MANGO", "scenario_id": 30, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "2e43d00a5c02", "created_at": "2026-10-07 08:35:44"}
|
| 192 |
+
{"supporter": "MANGO", "scenario_id": 284, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "3bfefc12adde", "created_at": "2026-10-07 08:35:45"}
|
| 193 |
+
{"supporter": "MANGO", "scenario_id": 549, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "d9f5638ecfea", "created_at": "2026-10-07 08:35:45"}
|
| 194 |
+
{"supporter": "MANGO", "scenario_id": 733, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "6601ea640650", "created_at": "2026-10-07 08:35:45"}
|
| 195 |
+
{"supporter": "MANGO", "scenario_id": 733, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "6e745fc8001c", "created_at": "2026-10-07 08:35:46"}
|
| 196 |
+
{"supporter": "MANGO", "scenario_id": 142, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "J", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "96d55803e28f", "created_at": "2026-10-07 08:35:46"}
|
| 197 |
+
{"supporter": "MANGO", "scenario_id": 250, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "200d14fce6bb", "created_at": "2026-10-07 08:35:46"}
|
| 198 |
+
{"supporter": "MANGO", "scenario_id": 791, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "K", "choice": "K", "winner": "A", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "73689def409c", "created_at": "2026-10-07 08:35:46"}
|
| 199 |
+
{"supporter": "MANGO", "scenario_id": 882, "pass": 1, "order": {"K": "A", "J": "B"}, "raw_output": "**J**", "choice": "J", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "162364c265ef", "created_at": "2026-10-07 08:35:47"}
|
| 200 |
+
{"supporter": "MANGO", "scenario_id": 882, "pass": 2, "order": {"K": "B", "J": "A"}, "raw_output": "K", "choice": "K", "winner": "B", "judge_model": "anthropic:claude-opus-5-5", "prompt_sha1": "f3c24decd762", "created_at": "2026-10-07 08:35:48"}
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/logs/vllm_hf-Bigglz-qwen-sft-sft-dpo-tone_0.log
ADDED
|
@@ -0,0 +1,294 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:347]
|
| 2 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:347] █ █ █▄ ▄█
|
| 3 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:347] ▄▄ ▄█ █ █ █ ▀▄▀ █ version 0.31.1rc1.dev1+gbb87d227d
|
| 4 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:347] █▄█▀ █ █ █ █ model /home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2
|
| 5 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:347] ▀▀ ▀▀▀▀▀ ▀▀▀▀▀ ▀ ▀
|
| 6 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:347]
|
| 7 |
+
(APIServer pid=344706) INFO 10-07 08:32:00 [api_utils.py:286] non-default args: {'model_tag': '/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', 'host': '127.0.0.1', 'port': 18001, 'model': '/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', 'max_model_len': 8192, 'served_model_name': ['hf-Bigglz-qwen-sft-sft-dpo-tone'], 'gpu_memory_utilization': 0.9, 'enable_prefix_caching': True}
|
| 8 |
+
(APIServer pid=344706) INFO 10-07 08:32:11 [model.py:730] Resolved architecture: Qwen2ForCausalLM
|
| 9 |
+
(APIServer pid=344706) INFO 10-07 08:32:11 [model.py:2141] Using max model len 8192
|
| 10 |
+
(APIServer pid=344706) INFO 10-07 08:32:11 [scheduler.py:308] Chunked prefill is enabled with max_num_batched_tokens=8192.
|
| 11 |
+
(APIServer pid=344706) INFO 10-07 08:32:11 [kernel.py:463] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native'])
|
| 12 |
+
(EngineCore pid=346800) INFO 10-07 08:32:22 [core.py:129] Initializing a V1 LLM engine (v0.31.1rc1.dev1+gbb87d227d) with config: model='/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', speculative_config=None, tokenizer='/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.bfloat16, max_seq_len=8192, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, decode_context_parallel_size=1, dcp_comm_backend=ag_rs, disable_custom_all_reduce=False, quantization=None, quantization_config=None, enforce_eager=False, aux_output_config=AuxOutputConfig(enable_return_routed_experts=False, max_bytes=None), kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, per_request_spec_decode_metrics='none', kv_cache_metrics=False, kv_cache_metrics_sample=0.01, custom_histogram_buckets=None, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False, jit_monitor_mode='warn', jit_monitor_verbose=False), seed=0, served_model_name=hf-Bigglz-qwen-sft-sft-dpo-tone, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'mode': <CompilationMode.VLLM_COMPILE: 3>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['none'], 'ir_enable_torch_wrap': True, 'splitting_ops': ['vllm::unified_attention_with_output', 'vllm::unified_mla_attention_with_output', 'vllm::mamba_mixer2', 'vllm::mamba_mixer', 'vllm::short_conv', 'vllm::qwen4_exp_ple_short_conv', 'vllm::qwen4_exp_qsa_with_output', 'vllm::linear_attention', 'vllm::qwen_gdn_attention_core', 'vllm::qwen_gdn_attention_core_fused_norm_packed', 'vllm::gdn_attention_core_xpu', 'vllm::olmo_hybrid_gdn_full_forward', 'vllm::sparse_attn_indexer', 'vllm::rocm_aiter_sparse_attn_indexer', 'vllm::deepseek_v4_attention', 'vllm::hpc_rope_norm_forward', 'vllm::unified_kv_cache_update', 'vllm::unified_mla_kv_cache_update'], 'compile_mm_encoder': False, 'cudagraph_mm_encoder': False, 'encoder_cudagraph_token_budgets': [], 'encoder_cudagraph_max_vision_items_per_batch': 0, 'encoder_cudagraph_max_frames_per_batch': None, 'cudagraph_decoder_replay': True, 'decoder_replay_cudagraph_capture_sizes': [], 'decoder_replay_trim_threshold': 768, 'compile_sizes': [], 'compile_ranges_endpoints': [8192], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.FULL_AND_PIECEWISE: (2, 1)>, 'cudagraph_num_of_warmups': 1, 'cudagraph_capture_sizes': [1, 2, 4, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 104, 112, 120, 128, 136, 144, 152, 160, 168, 176, 184, 192, 200, 208, 216, 224, 232, 240, 248, 256, 272, 288, 304, 320, 336, 352, 368, 384, 400, 416, 432, 448, 464, 480, 496, 512], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'enable_qk_norm_rope_fusion': False, 'fuse_rope_kvcache_cat_mla': False, 'fuse_act_padding': False, 'fuse_qk_norm_rope_kvcache': False}, 'max_cudagraph_capture_size': 512, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': False, 'static_all_moe_layers': []}, kernel_config=KernelConfig(ir_op_priority=IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native']), enable_flashinfer_autotune=True, enable_rocm_segmented_attn_autotune=False, enable_cutedsl_warmup=True, enable_jit_warmup=True, moe_backend='auto', sparse_indexer_topk_backend='auto', linear_backend='auto', linear_backend_per_quant=None)
|
| 13 |
+
(EngineCore pid=346800) INFO 10-07 08:32:26 [parallel_state.py:1864] world_size=1 rank=0 local_rank=0 distributed_init_method=file:///tmp/vllm_dist_9166a53bcc7543b6a3e75f9d7a538fef backend=nccl
|
| 14 |
+
(EngineCore pid=346800) INFO 10-07 08:32:26 [parallel_state.py:2308] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, ETP rank 0, EP rank N/A, EPLB rank N/A
|
| 15 |
+
(EngineCore pid=346800) INFO 10-07 08:32:26 [gpu_worker.py:501] Using V2 Model Runner
|
| 16 |
+
(EngineCore pid=346800) INFO 10-07 08:32:27 [model_runner.py:375] Loading model from scratch...
|
| 17 |
+
(EngineCore pid=346800) INFO 10-07 08:32:28 [cuda.py:526] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].
|
| 18 |
+
(EngineCore pid=346800) INFO 10-07 08:32:28 [flash_attn.py:1171] Using FlashAttention version 3
|
| 19 |
+
(EngineCore pid=346800) INFO 10-07 08:32:28 [base_loader.py:81] Initializing model took 1.244702 seconds
|
| 20 |
+
(EngineCore pid=346800) INFO 10-07 08:32:28 [weight_utils.py:1063] Filesystem type for checkpoints: EXT4. Checkpoint size: 27.51 GiB. Available RAM: 1947.79 GiB.
|
| 21 |
+
(EngineCore pid=346800) INFO 10-07 08:32:28 [weight_utils.py:1086] Auto-prefetch is disabled because the filesystem (EXT4) is not a recognized network FS (NFS/Lustre). If you want to force prefetching, start vLLM with --safetensors-load-strategy=prefetch.
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
(EngineCore pid=346800) INFO 10-07 08:32:31 [default_loader.py:484] Loading weights took 2.99 seconds
|
| 32 |
+
(EngineCore pid=346800) INFO 10-07 08:32:32 [model_runner.py:407] Model loading took 27.57 GiB memory and 5.738724 seconds
|
| 33 |
+
(EngineCore pid=346800) INFO 10-07 08:32:32 [topk_topp_sampler.py:124] Using FlashInfer for top-p & top-k sampling.
|
| 34 |
+
(EngineCore pid=346800) INFO 10-07 08:32:32 [torch_utils.py:286] Reducing Torch threads from 96 to 1 for serving. Set OMP_NUM_THREADS in the external environment to override.
|
| 35 |
+
(EngineCore pid=346800) INFO 10-07 08:32:32 [utils.py:347] Using LBNHC KV cache layout.
|
| 36 |
+
(EngineCore pid=346800) INFO 10-07 08:32:33 [monitor.py:40] [backbone] Starting torch.compile
|
| 37 |
+
(EngineCore pid=346800) INFO 10-07 08:32:41 [backends.py:1101] [backbone] Using cache directory: /home/ubuntu/.cache/vllm/torch_compile_cache/7438ca6fdd/rank_0_0/backbone for vLLM's torch.compile
|
| 38 |
+
(EngineCore pid=346800) INFO 10-07 08:32:41 [backends.py:1163] [backbone] Dynamo bytecode transform time: 7.70 s
|
| 39 |
+
(EngineCore pid=346800) INFO 10-07 08:32:48 [backends.py:398] [backbone] Compiling a graph for compile range (1, 8192) takes 6.89 s
|
| 40 |
+
(EngineCore pid=346800) INFO 10-07 08:32:54 [backends.py:927] collected artifacts: 49 entries, 3 artifacts, 4948406 bytes total
|
| 41 |
+
(EngineCore pid=346800) INFO 10-07 08:32:54 [decorators.py:743] [backbone] saved AOT compiled function to /home/ubuntu/.cache/vllm/torch_compile_cache/torch_aot_compile/5379ea870c95c2281a663da4ab8ea45659c51399f83959aa1270aede10dcb195/rank_0_0/model
|
| 42 |
+
(EngineCore pid=346800) INFO 10-07 08:32:54 [monitor.py:62] [backbone] torch.compile took 20.56 s in total
|
| 43 |
+
(EngineCore pid=346800) INFO 10-07 08:32:54 [monitor.py:91] [backbone] Initial profiling/warmup run took 0.13 s
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
(EngineCore pid=346800) INFO 10-07 08:33:03 [model_runner.py:1109] Graph capturing finished in 7 secs, took 0.74 GiB
|
| 47 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [gpu_worker.py:726] Available KV cache memory: 39.8 GiB
|
| 48 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [gpu_worker.py:741] CUDA graph memory profiling is enabled (default since v0.21.0). The current --gpu-memory-utilization=0.9000 is equivalent to --gpu-memory-utilization=0.8822 without CUDA graph memory profiling. To maintain the same effective KV cache size as before, increase --gpu-memory-utilization to 0.9178. To disable, set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0.
|
| 49 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [kv_cache_utils.py:2481] GPU KV cache size: 217,344 tokens, Maximum concurrency for 8,192 tokens per request: 26.53x
|
| 50 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [kernel_warmup.py:173] JIT kernel warmup starting.
|
| 51 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [kernel_warmup.py:186] JIT kernel warmup finished in 0.00s.
|
| 52 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [kernel_warmup.py:483] Using FlashInfer autotune cache file: /home/ubuntu/.cache/vllm/flashinfer_autotune_cache/0.7.0.post1/90a/a1e2ab0b1c974a982aa20d78f9363d70e61b4f542571fd670a682775951c1349/autotune_configs_dp0_rank0.json
|
| 53 |
+
2026-10-07 08:33:04,149 - INFO - autotuner.py:1046 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
|
| 54 |
+
(EngineCore pid=346800) INFO 10-07 08:33:04 [kernel_warmup.py:367] Running FlashInfer autotune with 8192 tokens and token buckets (1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2560, 3072, 3584, 4096, 8192).
|
| 55 |
+
2026-10-07 08:33:04,184 - INFO - autotuner.py:1069 - flashinfer.jit: [Autotuner]: Autotuning process ends
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
(EngineCore pid=346800) INFO 10-07 08:33:15 [model_runner.py:1109] Graph capturing finished in 9 secs, took 0.52 GiB
|
| 59 |
+
(EngineCore pid=346800) INFO 10-07 08:33:15 [gpu_worker.py:933] CUDA graph pool memory: 0.52 GiB (actual), 1.41 GiB (estimated), difference: 0.89 GiB (169.8%).
|
| 60 |
+
(EngineCore pid=346800) INFO 10-07 08:33:15 [gpu_worker.py:996] Free memory on device (78.67/79.18 GiB) on startup. Desired GPU memory utilization is (0.9, 71.26 GiB). Actual usage is 28.49 GiB for consumed memory (weights + non-torch), 2.97 GiB for peak activation, and 0.52 GiB for CUDAGraph memory. Replace gpu_memory_utilization config with `--kv-cache-memory=42015226983` (39.13 GiB) to fit into requested memory, or `--kv-cache-memory=49964652032` (46.53 GiB) to fully utilize gpu memory. Current kv cache memory in use is 39.8 GiB.
|
| 61 |
+
(EngineCore pid=346800) INFO 10-07 08:33:16 [jit_monitor.py:84] Kernel JIT monitor activated; monitored JIT compilations during inference will use mode=warn.
|
| 62 |
+
(EngineCore pid=346800) INFO 10-07 08:33:16 [core.py:373] init engine (profile, create kv cache, warmup model) took 43.92 s (compilation: 20.56 s)
|
| 63 |
+
(EngineCore pid=346800) INFO 10-07 08:33:17 [kernel.py:463] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native'])
|
| 64 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [entry.py:135] Supported tasks: ['generate']
|
| 65 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [factories.py:76] Scale-out endpoints are disabled. Set --enable-scale-out to enable them.
|
| 66 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [hf.py:648] Detected the chat template content format to be 'string'. You can set `--chat-template-content-format` to override this.
|
| 67 |
+
(APIServer pid=344706) WARNING 10-07 08:33:17 [model.py:1882] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'repetition_penalty': 1.05, 'temperature': 0.7, 'top_k': 20, 'top_p': 0.8}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.
|
| 68 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [entry.py:139] Starting vLLM server on http://127.0.0.1:18001
|
| 69 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:60] Available routes are:
|
| 70 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /openapi.json, Methods: GET, HEAD
|
| 71 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /docs, Methods: GET, HEAD
|
| 72 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /docs/oauth2-redirect, Methods: GET, HEAD
|
| 73 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /redoc, Methods: GET, HEAD
|
| 74 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /load, Methods: GET
|
| 75 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /version, Methods: GET
|
| 76 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /kv_event_sources, Methods: GET
|
| 77 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /health, Methods: GET
|
| 78 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /metrics, Methods: GET
|
| 79 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /tokenize, Methods: POST
|
| 80 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /detokenize, Methods: POST
|
| 81 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/models, Methods: GET
|
| 82 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /ping, Methods: GET
|
| 83 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /ping, Methods: POST
|
| 84 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /invocations, Methods: POST
|
| 85 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/chat/completions, Methods: POST
|
| 86 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/chat/completions/batch, Methods: POST
|
| 87 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/responses, Methods: POST
|
| 88 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/responses/{response_id}, Methods: GET
|
| 89 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/responses/{response_id}/cancel, Methods: POST
|
| 90 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/completions, Methods: POST
|
| 91 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/messages, Methods: POST
|
| 92 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/messages/count_tokens, Methods: POST
|
| 93 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /generative_scoring, Methods: POST
|
| 94 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /scale_elastic_ep, Methods: POST
|
| 95 |
+
(APIServer pid=344706) INFO 10-07 08:33:17 [launcher.py:69] Route: /is_scaling_elastic_ep, Methods: POST
|
| 96 |
+
INFO: Started server process [344706]
|
| 97 |
+
INFO: Waiting for application startup.
|
| 98 |
+
INFO: Application startup complete.
|
| 99 |
+
INFO: 127.0.0.1:33048 - "GET /health HTTP/1.1" 200 OK
|
| 100 |
+
(APIServer pid=344706) WARNING 10-07 08:34:51 [vllm.py:1327] Watermarking is enabled for this request, but the engine has no watermark configuration. This and subsequent requests will run without watermarking.
|
| 101 |
+
(EngineCore pid=346800) WARNING 10-07 08:34:54 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topk_topp_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 102 |
+
(EngineCore pid=346800) WARNING 10-07 08:34:54 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topp_sb_stats_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 103 |
+
(EngineCore pid=346800) WARNING 10-07 08:34:54 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topp_sb_step_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 104 |
+
(EngineCore pid=346800) WARNING 10-07 08:34:55 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topp_sb_mask_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 105 |
+
(EngineCore pid=346800) WARNING 10-07 08:34:55 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _gumbel_sample_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 106 |
+
INFO: 127.0.0.1:44596 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 107 |
+
INFO: 127.0.0.1:44658 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 108 |
+
INFO: 127.0.0.1:44294 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 109 |
+
INFO: 127.0.0.1:44342 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 110 |
+
INFO: 127.0.0.1:44258 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 111 |
+
INFO: 127.0.0.1:44310 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 112 |
+
INFO: 127.0.0.1:44492 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 113 |
+
INFO: 127.0.0.1:44614 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 114 |
+
INFO: 127.0.0.1:44682 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 115 |
+
INFO: 127.0.0.1:44692 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 116 |
+
INFO: 127.0.0.1:44436 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 117 |
+
INFO: 127.0.0.1:44272 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 118 |
+
INFO: 127.0.0.1:44542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 119 |
+
INFO: 127.0.0.1:44626 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 120 |
+
INFO: 127.0.0.1:44360 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 121 |
+
INFO: 127.0.0.1:44446 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 122 |
+
INFO: 127.0.0.1:44470 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 123 |
+
INFO: 127.0.0.1:44600 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 124 |
+
INFO: 127.0.0.1:44450 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 125 |
+
INFO: 127.0.0.1:44628 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 126 |
+
INFO: 127.0.0.1:44352 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 127 |
+
INFO: 127.0.0.1:44356 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 128 |
+
INFO: 127.0.0.1:44376 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 129 |
+
INFO: 127.0.0.1:44398 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 130 |
+
INFO: 127.0.0.1:44422 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 131 |
+
INFO: 127.0.0.1:44674 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 132 |
+
INFO: 127.0.0.1:44710 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 133 |
+
INFO: 127.0.0.1:44350 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 134 |
+
INFO: 127.0.0.1:44506 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 135 |
+
INFO: 127.0.0.1:44518 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 136 |
+
INFO: 127.0.0.1:44276 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 137 |
+
INFO: 127.0.0.1:44388 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 138 |
+
INFO: 127.0.0.1:44432 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 139 |
+
INFO: 127.0.0.1:44460 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 140 |
+
INFO: 127.0.0.1:44550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 141 |
+
INFO: 127.0.0.1:44700 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 142 |
+
INFO: 127.0.0.1:44480 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 143 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 144 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 145 |
+
INFO: 127.0.0.1:44566 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 146 |
+
INFO: 127.0.0.1:44568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 147 |
+
INFO: 127.0.0.1:44336 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 148 |
+
INFO: 127.0.0.1:44638 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 149 |
+
INFO: 127.0.0.1:44426 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 150 |
+
INFO: 127.0.0.1:44642 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 151 |
+
INFO: 127.0.0.1:44412 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 152 |
+
INFO: 127.0.0.1:44676 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 153 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 154 |
+
INFO: 127.0.0.1:44528 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 155 |
+
INFO: 127.0.0.1:44580 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 156 |
+
(APIServer pid=344706) INFO 10-07 08:35:08 [loggers.py:329] Engine 000: Avg prompt throughput: 173.2 tokens/s, Avg generation throughput: 122.1 tokens/s, Running: 3 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.2%, Prefix cache hit rate: 70.7%
|
| 157 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 158 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 159 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 160 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 161 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 162 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 163 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 164 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 165 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 166 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 167 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 168 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 169 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 170 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 171 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 172 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 173 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 174 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 175 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 176 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 177 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 178 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 179 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 180 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 181 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 182 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 183 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 184 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 185 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 186 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 187 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 188 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 189 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 190 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 191 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 192 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 193 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 194 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 195 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 196 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 197 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 198 |
+
(APIServer pid=344706) INFO 10-07 08:35:18 [loggers.py:329] Engine 000: Avg prompt throughput: 222.2 tokens/s, Avg generation throughput: 119.2 tokens/s, Running: 3 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.2%, Prefix cache hit rate: 68.4%
|
| 199 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 200 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 201 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 202 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 203 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 204 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 205 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 206 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 207 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 208 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 209 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 210 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 211 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 212 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 213 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 214 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 215 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 216 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 217 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 218 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 219 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 220 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 221 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 222 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 223 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 224 |
+
INFO: 127.0.0.1:35728 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 225 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 226 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 227 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 228 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 229 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 230 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 231 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 232 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 233 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 234 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 235 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 236 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 237 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 238 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 239 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 240 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 241 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 242 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 243 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 244 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 245 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 246 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 247 |
+
INFO: 127.0.0.1:35718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 248 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 249 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 250 |
+
INFO: 127.0.0.1:35728 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 251 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 252 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 253 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 254 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 255 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 256 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 257 |
+
(APIServer pid=344706) INFO 10-07 08:35:28 [loggers.py:329] Engine 000: Avg prompt throughput: 433.4 tokens/s, Avg generation throughput: 176.1 tokens/s, Running: 3 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.3%, Prefix cache hit rate: 66.7%
|
| 258 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 259 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 260 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 261 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 262 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 263 |
+
INFO: 127.0.0.1:44326 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 264 |
+
INFO: 127.0.0.1:44330 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 265 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 266 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 267 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 268 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 269 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 270 |
+
(APIServer pid=344706) INFO 10-07 08:35:38 [loggers.py:329] Engine 000: Avg prompt throughput: 96.8 tokens/s, Avg generation throughput: 33.2 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 65.8%
|
| 271 |
+
INFO: 127.0.0.1:44284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 272 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [loggers.py:329] Engine 000: Avg prompt throughput: 4.5 tokens/s, Avg generation throughput: 2.3 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 65.9%
|
| 273 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [launcher.py:128] [shutdown] API server: shutdown triggered
|
| 274 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [launcher.py:144] [shutdown] API server: stopping engine client mode=abort timeout=0s
|
| 275 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [core_client.py:783] [shutdown] MPClient: start timeout=0s
|
| 276 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [core_client.py:785] [shutdown] MPClient: stopping engine manager
|
| 277 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [utils.py:655] [shutdown] Process manager: send sigterm to process EngineCore
|
| 278 |
+
(APIServer pid=344706) WARNING 10-07 08:35:48 [utils.py:675] [shutdown] Process manager: force killing remaining processes count=1
|
| 279 |
+
(APIServer pid=344706) WARNING 10-07 08:35:48 [utils.py:680] [shutdown] Process manager: force killing remaining process EngineCore pid 346800
|
| 280 |
+
(EngineCore pid=346800) INFO 10-07 08:35:48 [core.py:1423] [shutdown] EngineCore: trigger received signal=SIGTERM
|
| 281 |
+
(EngineCore pid=346800) INFO 10-07 08:35:48 [core.py:1572] [shutdown] EngineCore: start mode=abort timeout=0s
|
| 282 |
+
(EngineCore pid=346800) INFO 10-07 08:35:48 [core.py:1603] [shutdown] EngineCore: request processing complete; starting resource teardown
|
| 283 |
+
(EngineCore pid=346800) INFO 10-07 08:35:48 [core.py:1436] [shutdown] EngineCore: exiting busy loop
|
| 284 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [core_client.py:787] [shutdown] MPClient: engine manager stopped
|
| 285 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [core_client.py:788] [shutdown] MPClient: cleaning up background resources
|
| 286 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [core_client.py:790] [shutdown] MPClient: complete
|
| 287 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [launcher.py:153] [shutdown] API server: engine client stopped
|
| 288 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [launcher.py:156] [shutdown] API server: signalling HTTP server shutdown
|
| 289 |
+
(APIServer pid=344706) INFO 10-07 08:35:48 [launcher.py:178] [shutdown] API server: shutting down FastAPI HTTP server
|
| 290 |
+
INFO: Shutting down
|
| 291 |
+
INFO: Waiting for application shutdown.
|
| 292 |
+
INFO: Application shutdown complete.
|
| 293 |
+
/usr/lib/python3.12/multiprocessing/resource_tracker.py:254: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown
|
| 294 |
+
warnings.warn('resource_tracker: There appear to be %d '
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/logs/vllm_hf-Bigglz-qwen-sft-sft-dpo-tone_1.log
ADDED
|
@@ -0,0 +1,270 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:347]
|
| 2 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:347] █ █ █▄ ▄█
|
| 3 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:347] ▄▄ ▄█ █ █ █ ▀▄▀ █ version 0.31.1rc1.dev1+gbb87d227d
|
| 4 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:347] █▄█▀ █ █ █ █ model /home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2
|
| 5 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:347] ▀▀ ▀▀▀▀▀ ▀▀▀▀▀ ▀ ▀
|
| 6 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:347]
|
| 7 |
+
(APIServer pid=344707) INFO 10-07 08:32:00 [api_utils.py:286] non-default args: {'model_tag': '/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', 'host': '127.0.0.1', 'port': 18002, 'model': '/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', 'max_model_len': 8192, 'served_model_name': ['hf-Bigglz-qwen-sft-sft-dpo-tone'], 'gpu_memory_utilization': 0.9, 'enable_prefix_caching': True}
|
| 8 |
+
(APIServer pid=344707) INFO 10-07 08:32:11 [model.py:730] Resolved architecture: Qwen2ForCausalLM
|
| 9 |
+
(APIServer pid=344707) INFO 10-07 08:32:11 [model.py:2141] Using max model len 8192
|
| 10 |
+
(APIServer pid=344707) INFO 10-07 08:32:11 [scheduler.py:308] Chunked prefill is enabled with max_num_batched_tokens=8192.
|
| 11 |
+
(APIServer pid=344707) INFO 10-07 08:32:11 [kernel.py:463] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native'])
|
| 12 |
+
(EngineCore pid=346793) INFO 10-07 08:32:22 [core.py:129] Initializing a V1 LLM engine (v0.31.1rc1.dev1+gbb87d227d) with config: model='/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', speculative_config=None, tokenizer='/home/ubuntu/hf_cache/hub/models--Bigglz--qwen-sft-sft-dpo-tone/snapshots/6559780f19acbcf1c35efe47dbc79cfcd70c2db2', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.bfloat16, max_seq_len=8192, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, decode_context_parallel_size=1, dcp_comm_backend=ag_rs, disable_custom_all_reduce=False, quantization=None, quantization_config=None, enforce_eager=False, aux_output_config=AuxOutputConfig(enable_return_routed_experts=False, max_bytes=None), kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, per_request_spec_decode_metrics='none', kv_cache_metrics=False, kv_cache_metrics_sample=0.01, custom_histogram_buckets=None, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False, jit_monitor_mode='warn', jit_monitor_verbose=False), seed=0, served_model_name=hf-Bigglz-qwen-sft-sft-dpo-tone, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'mode': <CompilationMode.VLLM_COMPILE: 3>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['none'], 'ir_enable_torch_wrap': True, 'splitting_ops': ['vllm::unified_attention_with_output', 'vllm::unified_mla_attention_with_output', 'vllm::mamba_mixer2', 'vllm::mamba_mixer', 'vllm::short_conv', 'vllm::qwen4_exp_ple_short_conv', 'vllm::qwen4_exp_qsa_with_output', 'vllm::linear_attention', 'vllm::qwen_gdn_attention_core', 'vllm::qwen_gdn_attention_core_fused_norm_packed', 'vllm::gdn_attention_core_xpu', 'vllm::olmo_hybrid_gdn_full_forward', 'vllm::sparse_attn_indexer', 'vllm::rocm_aiter_sparse_attn_indexer', 'vllm::deepseek_v4_attention', 'vllm::hpc_rope_norm_forward', 'vllm::unified_kv_cache_update', 'vllm::unified_mla_kv_cache_update'], 'compile_mm_encoder': False, 'cudagraph_mm_encoder': False, 'encoder_cudagraph_token_budgets': [], 'encoder_cudagraph_max_vision_items_per_batch': 0, 'encoder_cudagraph_max_frames_per_batch': None, 'cudagraph_decoder_replay': True, 'decoder_replay_cudagraph_capture_sizes': [], 'decoder_replay_trim_threshold': 768, 'compile_sizes': [], 'compile_ranges_endpoints': [8192], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.FULL_AND_PIECEWISE: (2, 1)>, 'cudagraph_num_of_warmups': 1, 'cudagraph_capture_sizes': [1, 2, 4, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 104, 112, 120, 128, 136, 144, 152, 160, 168, 176, 184, 192, 200, 208, 216, 224, 232, 240, 248, 256, 272, 288, 304, 320, 336, 352, 368, 384, 400, 416, 432, 448, 464, 480, 496, 512], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'enable_qk_norm_rope_fusion': False, 'fuse_rope_kvcache_cat_mla': False, 'fuse_act_padding': False, 'fuse_qk_norm_rope_kvcache': False}, 'max_cudagraph_capture_size': 512, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': False, 'static_all_moe_layers': []}, kernel_config=KernelConfig(ir_op_priority=IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native']), enable_flashinfer_autotune=True, enable_rocm_segmented_attn_autotune=False, enable_cutedsl_warmup=True, enable_jit_warmup=True, moe_backend='auto', sparse_indexer_topk_backend='auto', linear_backend='auto', linear_backend_per_quant=None)
|
| 13 |
+
(EngineCore pid=346793) INFO 10-07 08:32:24 [parallel_state.py:1864] world_size=1 rank=0 local_rank=0 distributed_init_method=file:///tmp/vllm_dist_0850268b3fb546e18fd1a18c06b9cd97 backend=nccl
|
| 14 |
+
(EngineCore pid=346793) INFO 10-07 08:32:24 [parallel_state.py:2308] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, ETP rank 0, EP rank N/A, EPLB rank N/A
|
| 15 |
+
(EngineCore pid=346793) INFO 10-07 08:32:24 [gpu_worker.py:501] Using V2 Model Runner
|
| 16 |
+
(EngineCore pid=346793) INFO 10-07 08:32:27 [model_runner.py:375] Loading model from scratch...
|
| 17 |
+
(EngineCore pid=346793) INFO 10-07 08:32:28 [cuda.py:526] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].
|
| 18 |
+
(EngineCore pid=346793) INFO 10-07 08:32:28 [flash_attn.py:1171] Using FlashAttention version 3
|
| 19 |
+
(EngineCore pid=346793) INFO 10-07 08:32:28 [base_loader.py:81] Initializing model took 1.259853 seconds
|
| 20 |
+
(EngineCore pid=346793) INFO 10-07 08:32:28 [weight_utils.py:1063] Filesystem type for checkpoints: EXT4. Checkpoint size: 27.51 GiB. Available RAM: 1947.80 GiB.
|
| 21 |
+
(EngineCore pid=346793) INFO 10-07 08:32:28 [weight_utils.py:1086] Auto-prefetch is disabled because the filesystem (EXT4) is not a recognized network FS (NFS/Lustre). If you want to force prefetching, start vLLM with --safetensors-load-strategy=prefetch.
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
(EngineCore pid=346793) INFO 10-07 08:32:31 [default_loader.py:484] Loading weights took 2.89 seconds
|
| 32 |
+
(EngineCore pid=346793) INFO 10-07 08:32:32 [model_runner.py:407] Model loading took 27.57 GiB memory and 5.620532 seconds
|
| 33 |
+
(EngineCore pid=346793) INFO 10-07 08:32:32 [topk_topp_sampler.py:124] Using FlashInfer for top-p & top-k sampling.
|
| 34 |
+
(EngineCore pid=346793) INFO 10-07 08:32:32 [torch_utils.py:286] Reducing Torch threads from 96 to 1 for serving. Set OMP_NUM_THREADS in the external environment to override.
|
| 35 |
+
(EngineCore pid=346793) INFO 10-07 08:32:32 [utils.py:347] Using LBNHC KV cache layout.
|
| 36 |
+
(EngineCore pid=346793) INFO 10-07 08:32:33 [monitor.py:40] [backbone] Starting torch.compile
|
| 37 |
+
(EngineCore pid=346793) INFO 10-07 08:32:40 [backends.py:1101] [backbone] Using cache directory: /home/ubuntu/.cache/vllm/torch_compile_cache/7438ca6fdd/rank_0_0/backbone for vLLM's torch.compile
|
| 38 |
+
(EngineCore pid=346793) INFO 10-07 08:32:40 [backends.py:1163] [backbone] Dynamo bytecode transform time: 7.46 s
|
| 39 |
+
(EngineCore pid=346793) INFO 10-07 08:32:48 [backends.py:398] [backbone] Compiling a graph for compile range (1, 8192) takes 7.35 s
|
| 40 |
+
(EngineCore pid=346793) INFO 10-07 08:32:54 [backends.py:927] collected artifacts: 49 entries, 3 artifacts, 5161382 bytes total
|
| 41 |
+
(EngineCore pid=346793) INFO 10-07 08:32:54 [decorators.py:743] [backbone] saved AOT compiled function to /home/ubuntu/.cache/vllm/torch_compile_cache/torch_aot_compile/5379ea870c95c2281a663da4ab8ea45659c51399f83959aa1270aede10dcb195/rank_0_0/model
|
| 42 |
+
(EngineCore pid=346793) INFO 10-07 08:32:54 [monitor.py:62] [backbone] torch.compile took 21.00 s in total
|
| 43 |
+
(EngineCore pid=346793) INFO 10-07 08:32:54 [monitor.py:91] [backbone] Initial profiling/warmup run took 0.13 s
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
(EngineCore pid=346793) INFO 10-07 08:33:03 [model_runner.py:1109] Graph capturing finished in 7 secs, took 0.74 GiB
|
| 47 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [gpu_worker.py:726] Available KV cache memory: 39.8 GiB
|
| 48 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [gpu_worker.py:741] CUDA graph memory profiling is enabled (default since v0.21.0). The current --gpu-memory-utilization=0.9000 is equivalent to --gpu-memory-utilization=0.8822 without CUDA graph memory profiling. To maintain the same effective KV cache size as before, increase --gpu-memory-utilization to 0.9178. To disable, set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0.
|
| 49 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [kv_cache_utils.py:2481] GPU KV cache size: 217,344 tokens, Maximum concurrency for 8,192 tokens per request: 26.53x
|
| 50 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [kernel_warmup.py:173] JIT kernel warmup starting.
|
| 51 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [kernel_warmup.py:186] JIT kernel warmup finished in 0.00s.
|
| 52 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [kernel_warmup.py:483] Using FlashInfer autotune cache file: /home/ubuntu/.cache/vllm/flashinfer_autotune_cache/0.7.0.post1/90a/a1e2ab0b1c974a982aa20d78f9363d70e61b4f542571fd670a682775951c1349/autotune_configs_dp0_rank0.json
|
| 53 |
+
2026-10-07 08:33:04,308 - INFO - autotuner.py:1046 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
|
| 54 |
+
(EngineCore pid=346793) INFO 10-07 08:33:04 [kernel_warmup.py:367] Running FlashInfer autotune with 8192 tokens and token buckets (1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2560, 3072, 3584, 4096, 8192).
|
| 55 |
+
2026-10-07 08:33:04,344 - INFO - autotuner.py:1069 - flashinfer.jit: [Autotuner]: Autotuning process ends
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
(EngineCore pid=346793) INFO 10-07 08:33:15 [model_runner.py:1109] Graph capturing finished in 9 secs, took 0.52 GiB
|
| 59 |
+
(EngineCore pid=346793) INFO 10-07 08:33:15 [gpu_worker.py:933] CUDA graph pool memory: 0.52 GiB (actual), 1.41 GiB (estimated), difference: 0.89 GiB (169.8%).
|
| 60 |
+
(EngineCore pid=346793) INFO 10-07 08:33:15 [gpu_worker.py:996] Free memory on device (78.67/79.18 GiB) on startup. Desired GPU memory utilization is (0.9, 71.26 GiB). Actual usage is 28.49 GiB for consumed memory (weights + non-torch), 2.97 GiB for peak activation, and 0.52 GiB for CUDAGraph memory. Replace gpu_memory_utilization config with `--kv-cache-memory=42015226983` (39.13 GiB) to fit into requested memory, or `--kv-cache-memory=49964717568` (46.53 GiB) to fully utilize gpu memory. Current kv cache memory in use is 39.8 GiB.
|
| 61 |
+
(EngineCore pid=346793) INFO 10-07 08:33:16 [jit_monitor.py:84] Kernel JIT monitor activated; monitored JIT compilations during inference will use mode=warn.
|
| 62 |
+
(EngineCore pid=346793) INFO 10-07 08:33:17 [core.py:373] init engine (profile, create kv cache, warmup model) took 44.35 s (compilation: 21.00 s)
|
| 63 |
+
(EngineCore pid=346793) INFO 10-07 08:33:17 [kernel.py:463] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native'])
|
| 64 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [entry.py:135] Supported tasks: ['generate']
|
| 65 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [factories.py:76] Scale-out endpoints are disabled. Set --enable-scale-out to enable them.
|
| 66 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [hf.py:648] Detected the chat template content format to be 'string'. You can set `--chat-template-content-format` to override this.
|
| 67 |
+
(APIServer pid=344707) WARNING 10-07 08:33:17 [model.py:1882] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'repetition_penalty': 1.05, 'temperature': 0.7, 'top_k': 20, 'top_p': 0.8}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.
|
| 68 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [entry.py:139] Starting vLLM server on http://127.0.0.1:18002
|
| 69 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:60] Available routes are:
|
| 70 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /openapi.json, Methods: GET, HEAD
|
| 71 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /docs, Methods: GET, HEAD
|
| 72 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /docs/oauth2-redirect, Methods: GET, HEAD
|
| 73 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /redoc, Methods: GET, HEAD
|
| 74 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /load, Methods: GET
|
| 75 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /version, Methods: GET
|
| 76 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /kv_event_sources, Methods: GET
|
| 77 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /health, Methods: GET
|
| 78 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /metrics, Methods: GET
|
| 79 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /tokenize, Methods: POST
|
| 80 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /detokenize, Methods: POST
|
| 81 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/models, Methods: GET
|
| 82 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /ping, Methods: GET
|
| 83 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /ping, Methods: POST
|
| 84 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /invocations, Methods: POST
|
| 85 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/chat/completions, Methods: POST
|
| 86 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/chat/completions/batch, Methods: POST
|
| 87 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/responses, Methods: POST
|
| 88 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/responses/{response_id}, Methods: GET
|
| 89 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/responses/{response_id}/cancel, Methods: POST
|
| 90 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/completions, Methods: POST
|
| 91 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/messages, Methods: POST
|
| 92 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /v1/messages/count_tokens, Methods: POST
|
| 93 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /generative_scoring, Methods: POST
|
| 94 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /scale_elastic_ep, Methods: POST
|
| 95 |
+
(APIServer pid=344707) INFO 10-07 08:33:17 [launcher.py:69] Route: /is_scaling_elastic_ep, Methods: POST
|
| 96 |
+
INFO: Started server process [344707]
|
| 97 |
+
INFO: Waiting for application startup.
|
| 98 |
+
INFO: Application startup complete.
|
| 99 |
+
INFO: 127.0.0.1:36518 - "GET /health HTTP/1.1" 200 OK
|
| 100 |
+
(APIServer pid=344707) WARNING 10-07 08:34:51 [vllm.py:1327] Watermarking is enabled for this request, but the engine has no watermark configuration. This and subsequent requests will run without watermarking.
|
| 101 |
+
(EngineCore pid=346793) WARNING 10-07 08:34:54 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topk_topp_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 102 |
+
(EngineCore pid=346793) WARNING 10-07 08:34:54 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topp_sb_stats_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 103 |
+
(EngineCore pid=346793) WARNING 10-07 08:34:54 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topp_sb_step_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 104 |
+
(EngineCore pid=346793) WARNING 10-07 08:34:55 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _topp_sb_mask_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 105 |
+
(EngineCore pid=346793) WARNING 10-07 08:34:55 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _gumbel_sample_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 106 |
+
INFO: 127.0.0.1:54106 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 107 |
+
INFO: 127.0.0.1:54324 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 108 |
+
INFO: 127.0.0.1:54364 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 109 |
+
INFO: 127.0.0.1:54374 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 110 |
+
INFO: 127.0.0.1:54402 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 111 |
+
INFO: 127.0.0.1:54382 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 112 |
+
INFO: 127.0.0.1:54172 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 113 |
+
INFO: 127.0.0.1:54228 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 114 |
+
INFO: 127.0.0.1:54292 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 115 |
+
INFO: 127.0.0.1:54454 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 116 |
+
INFO: 127.0.0.1:54350 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 117 |
+
INFO: 127.0.0.1:54572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 118 |
+
INFO: 127.0.0.1:54150 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 119 |
+
INFO: 127.0.0.1:54162 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 120 |
+
INFO: 127.0.0.1:54206 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 121 |
+
INFO: 127.0.0.1:54466 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 122 |
+
INFO: 127.0.0.1:54440 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 123 |
+
INFO: 127.0.0.1:54446 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 124 |
+
INFO: 127.0.0.1:54394 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 125 |
+
INFO: 127.0.0.1:54516 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 126 |
+
INFO: 127.0.0.1:54156 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 127 |
+
INFO: 127.0.0.1:54492 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 128 |
+
INFO: 127.0.0.1:54200 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 129 |
+
INFO: 127.0.0.1:54222 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 130 |
+
INFO: 127.0.0.1:54500 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 131 |
+
INFO: 127.0.0.1:54166 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 132 |
+
INFO: 127.0.0.1:54244 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 133 |
+
INFO: 127.0.0.1:54366 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 134 |
+
INFO: 127.0.0.1:54392 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 135 |
+
INFO: 127.0.0.1:54418 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 136 |
+
INFO: 127.0.0.1:54540 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 137 |
+
INFO: 127.0.0.1:54268 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 138 |
+
INFO: 127.0.0.1:54338 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 139 |
+
INFO: 127.0.0.1:54132 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 140 |
+
INFO: 127.0.0.1:54432 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 141 |
+
INFO: 127.0.0.1:54478 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 142 |
+
INFO: 127.0.0.1:54256 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 143 |
+
INFO: 127.0.0.1:54182 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 144 |
+
INFO: 127.0.0.1:54390 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 145 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 146 |
+
INFO: 127.0.0.1:54452 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 147 |
+
INFO: 127.0.0.1:54556 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 148 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 149 |
+
INFO: 127.0.0.1:54276 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 150 |
+
INFO: 127.0.0.1:54308 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 151 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 152 |
+
INFO: 127.0.0.1:54334 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 153 |
+
INFO: 127.0.0.1:54520 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 154 |
+
INFO: 127.0.0.1:54536 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 155 |
+
INFO: 127.0.0.1:54376 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 156 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 157 |
+
(APIServer pid=344707) INFO 10-07 08:35:08 [loggers.py:329] Engine 000: Avg prompt throughput: 159.6 tokens/s, Avg generation throughput: 119.9 tokens/s, Running: 2 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.1%, Prefix cache hit rate: 72.7%
|
| 158 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 159 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 160 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 161 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 162 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 163 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 164 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 165 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 166 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 167 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 168 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 169 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 170 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 171 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 172 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 173 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 174 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 175 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 176 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 177 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 178 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 179 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 180 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 181 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 182 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 183 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 184 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 185 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 186 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 187 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 188 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 189 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 190 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 191 |
+
(APIServer pid=344707) INFO 10-07 08:35:18 [loggers.py:329] Engine 000: Avg prompt throughput: 194.7 tokens/s, Avg generation throughput: 96.0 tokens/s, Running: 1 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.1%, Prefix cache hit rate: 67.9%
|
| 192 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 193 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 194 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 195 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 196 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 197 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 198 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 199 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 200 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 201 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 202 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 203 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 204 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 205 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 206 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 207 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 208 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 209 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 210 |
+
INFO: 127.0.0.1:52552 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 211 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 212 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 213 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 214 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 215 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 216 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 217 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 218 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 219 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 220 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 221 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 222 |
+
INFO: 127.0.0.1:52552 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 223 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 224 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 225 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 226 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 227 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 228 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 229 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 230 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 231 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 232 |
+
INFO: 127.0.0.1:52552 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 233 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 234 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 235 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 236 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 237 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 238 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 239 |
+
INFO: 127.0.0.1:54186 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 240 |
+
(APIServer pid=344707) INFO 10-07 08:35:28 [loggers.py:329] Engine 000: Avg prompt throughput: 422.4 tokens/s, Avg generation throughput: 145.6 tokens/s, Running: 2 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.2%, Prefix cache hit rate: 63.6%
|
| 241 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 242 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 243 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 244 |
+
INFO: 127.0.0.1:54140 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 245 |
+
INFO: 127.0.0.1:54120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 246 |
+
(APIServer pid=344707) INFO 10-07 08:35:38 [loggers.py:329] Engine 000: Avg prompt throughput: 17.8 tokens/s, Avg generation throughput: 14.2 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 64.0%
|
| 247 |
+
INFO: 127.0.0.1:55584 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 248 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [loggers.py:329] Engine 000: Avg prompt throughput: 2.9 tokens/s, Avg generation throughput: 2.7 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 64.2%
|
| 249 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [launcher.py:128] [shutdown] API server: shutdown triggered
|
| 250 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [launcher.py:144] [shutdown] API server: stopping engine client mode=abort timeout=0s
|
| 251 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [core_client.py:783] [shutdown] MPClient: start timeout=0s
|
| 252 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [core_client.py:785] [shutdown] MPClient: stopping engine manager
|
| 253 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [utils.py:655] [shutdown] Process manager: send sigterm to process EngineCore
|
| 254 |
+
(APIServer pid=344707) WARNING 10-07 08:35:48 [utils.py:675] [shutdown] Process manager: force killing remaining processes count=1
|
| 255 |
+
(APIServer pid=344707) WARNING 10-07 08:35:48 [utils.py:680] [shutdown] Process manager: force killing remaining process EngineCore pid 346793
|
| 256 |
+
(EngineCore pid=346793) INFO 10-07 08:35:48 [core.py:1423] [shutdown] EngineCore: trigger received signal=SIGTERM
|
| 257 |
+
(EngineCore pid=346793) INFO 10-07 08:35:48 [core.py:1572] [shutdown] EngineCore: start mode=abort timeout=0s
|
| 258 |
+
(EngineCore pid=346793) INFO 10-07 08:35:48 [core.py:1603] [shutdown] EngineCore: request processing complete; starting resource teardown
|
| 259 |
+
(EngineCore pid=346793) INFO 10-07 08:35:48 [core.py:1436] [shutdown] EngineCore: exiting busy loop
|
| 260 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [core_client.py:787] [shutdown] MPClient: engine manager stopped
|
| 261 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [core_client.py:788] [shutdown] MPClient: cleaning up background resources
|
| 262 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [core_client.py:790] [shutdown] MPClient: complete
|
| 263 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [launcher.py:153] [shutdown] API server: engine client stopped
|
| 264 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [launcher.py:156] [shutdown] API server: signalling HTTP server shutdown
|
| 265 |
+
(APIServer pid=344707) INFO 10-07 08:35:48 [launcher.py:178] [shutdown] API server: shutting down FastAPI HTTP server
|
| 266 |
+
INFO: Shutting down
|
| 267 |
+
INFO: Waiting for application shutdown.
|
| 268 |
+
INFO: Application shutdown complete.
|
| 269 |
+
/usr/lib/python3.12/multiprocessing/resource_tracker.py:254: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown
|
| 270 |
+
warnings.warn('resource_tracker: There appear to be %d '
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/logs/vllm_hf-models-mistral_small4-sft-sft-dpo-native_0.log
ADDED
|
@@ -0,0 +1,552 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:347]
|
| 2 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:347] █ █ █▄ ▄█
|
| 3 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:347] ▄▄ ▄█ █ █ █ ▀▄▀ █ version 0.31.1rc1.dev1+gbb87d227d
|
| 4 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:347] █▄█▀ █ █ █ █ model models/mistral_small4-sft-sft-dpo-native
|
| 5 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:347] ▀▀ ▀▀▀▀▀ ▀▀▀▀▀ ▀ ▀
|
| 6 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:347]
|
| 7 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [api_utils.py:286] non-default args: {'model_tag': 'models/mistral_small4-sft-sft-dpo-native', 'host': '127.0.0.1', 'port': 18000, 'model': 'models/mistral_small4-sft-sft-dpo-native', 'tokenizer': 'models/mistral_small4-sft-sft-dpo-native', 'max_model_len': 8192, 'served_model_name': ['hf-models-mistral_small4-sft-sft-dpo-native'], 'config_format': 'mistral', 'load_format': 'mistral', 'tensor_parallel_size': 4, 'gpu_memory_utilization': 0.9, 'enable_prefix_caching': True}
|
| 8 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [config.py:514] Inferred from consolidated*.safetensors files torch.bfloat16 dtype.
|
| 9 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [model.py:730] Resolved architecture: MistralLarge3ForCausalLM
|
| 10 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [model.py:2141] Using max model len 8192
|
| 11 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [scheduler.py:308] Chunked prefill is enabled with max_num_batched_tokens=8192.
|
| 12 |
+
(APIServer pid=344642) INFO 10-07 08:32:00 [kernel.py:463] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native'])
|
| 13 |
+
(APIServer pid=344642) INFO 10-07 08:32:03 [compilation.py:331] Enabled custom fusions: allreduce_rms
|
| 14 |
+
(EngineCore pid=345767) INFO 10-07 08:32:13 [core.py:129] Initializing a V1 LLM engine (v0.31.1rc1.dev1+gbb87d227d) with config: model='models/mistral_small4-sft-sft-dpo-native', speculative_config=None, tokenizer='models/mistral_small4-sft-sft-dpo-native', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.bfloat16, max_seq_len=8192, download_dir=None, load_format=mistral, tensor_parallel_size=4, pipeline_parallel_size=1, data_parallel_size=1, decode_context_parallel_size=1, dcp_comm_backend=ag_rs, disable_custom_all_reduce=False, quantization=None, quantization_config=None, enforce_eager=False, aux_output_config=AuxOutputConfig(enable_return_routed_experts=False, max_bytes=None), kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, per_request_spec_decode_metrics='none', kv_cache_metrics=False, kv_cache_metrics_sample=0.01, custom_histogram_buckets=None, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False, jit_monitor_mode='warn', jit_monitor_verbose=False), seed=0, served_model_name=hf-models-mistral_small4-sft-sft-dpo-native, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'mode': <CompilationMode.VLLM_COMPILE: 3>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['none'], 'ir_enable_torch_wrap': True, 'splitting_ops': ['vllm::unified_attention_with_output', 'vllm::unified_mla_attention_with_output', 'vllm::mamba_mixer2', 'vllm::mamba_mixer', 'vllm::short_conv', 'vllm::qwen4_exp_ple_short_conv', 'vllm::qwen4_exp_qsa_with_output', 'vllm::linear_attention', 'vllm::qwen_gdn_attention_core', 'vllm::qwen_gdn_attention_core_fused_norm_packed', 'vllm::gdn_attention_core_xpu', 'vllm::olmo_hybrid_gdn_full_forward', 'vllm::sparse_attn_indexer', 'vllm::rocm_aiter_sparse_attn_indexer', 'vllm::deepseek_v4_attention', 'vllm::hpc_rope_norm_forward', 'vllm::unified_kv_cache_update', 'vllm::unified_mla_kv_cache_update'], 'compile_mm_encoder': False, 'cudagraph_mm_encoder': False, 'encoder_cudagraph_token_budgets': [], 'encoder_cudagraph_max_vision_items_per_batch': 0, 'encoder_cudagraph_max_frames_per_batch': None, 'cudagraph_decoder_replay': True, 'decoder_replay_cudagraph_capture_sizes': [], 'decoder_replay_trim_threshold': 768, 'compile_sizes': [], 'compile_ranges_endpoints': [256, 8192], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.FULL_AND_PIECEWISE: (2, 1)>, 'cudagraph_num_of_warmups': 1, 'cudagraph_capture_sizes': [1, 2, 4, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 104, 112, 120, 128, 136, 144, 152, 160, 168, 176, 184, 192, 200, 208, 216, 224, 232, 240, 248, 256, 272, 288, 304, 320, 336, 352, 368, 384, 400, 416, 432, 448, 464, 480, 496, 512], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': True, 'enable_qk_norm_rope_fusion': False, 'fuse_rope_kvcache_cat_mla': False, 'fuse_act_padding': False, 'fuse_qk_norm_rope_kvcache': False}, 'max_cudagraph_capture_size': 512, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': False, 'static_all_moe_layers': []}, kernel_config=KernelConfig(ir_op_priority=IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native']), enable_flashinfer_autotune=True, enable_rocm_segmented_attn_autotune=False, enable_cutedsl_warmup=True, enable_jit_warmup=True, moe_backend='auto', sparse_indexer_topk_backend='auto', linear_backend='auto', linear_backend_per_quant=None)
|
| 15 |
+
(EngineCore pid=345767) INFO 10-07 08:32:13 [multiproc_executor.py:153] DP group leader: node_rank=0, node_rank_within_dp=0, master_addr=127.0.0.1, mq_connect_ip=172.31.8.165 (local), world_size=4, local_world_size=4
|
| 16 |
+
(Worker pid=346565) INFO 10-07 08:32:26 [parallel_state.py:1864] world_size=4 rank=0 local_rank=0 distributed_init_method=file:///tmp/vllm_dist_4cf989ab6efb4c71bb5e274120cab625 backend=nccl
|
| 17 |
+
(Worker pid=346568) INFO 10-07 08:32:26 [parallel_state.py:1864] world_size=4 rank=3 local_rank=3 distributed_init_method=file:///tmp/vllm_dist_4cf989ab6efb4c71bb5e274120cab625 backend=nccl
|
| 18 |
+
(Worker pid=346567) INFO 10-07 08:32:26 [parallel_state.py:1864] world_size=4 rank=2 local_rank=2 distributed_init_method=file:///tmp/vllm_dist_4cf989ab6efb4c71bb5e274120cab625 backend=nccl
|
| 19 |
+
(Worker pid=346566) INFO 10-07 08:32:26 [parallel_state.py:1864] world_size=4 rank=1 local_rank=1 distributed_init_method=file:///tmp/vllm_dist_4cf989ab6efb4c71bb5e274120cab625 backend=nccl
|
| 20 |
+
(Worker pid=346565) INFO 10-07 08:32:28 [pynccl.py:148] vLLM is using nccl==2.29.7
|
| 21 |
+
(Worker pid=346565) INFO 10-07 08:32:30 [cuda_communicator.py:351] Using ['FLASHINFER', 'CUSTOM', 'SYMM_MEM', 'PYNCCL'] all-reduce backends (in dispatch order) for group 'tp:0' out of potential backends: ['FLASHINFER_PCIE_IPC', 'FLASHINFER', 'NCCL_SYMM_MEM', 'QUICK_REDUCE', 'AITER_CUSTOM', 'CUSTOM', 'SYMM_MEM', 'PYNCCL'].
|
| 22 |
+
(Worker pid=346565) INFO 10-07 08:32:31 [cuda_communicator.py:351] Using ['PYNCCL'] all-reduce backends (in dispatch order) for group 'ep:0' out of potential backends: ['FLASHINFER_PCIE_IPC', 'FLASHINFER', 'NCCL_SYMM_MEM', 'QUICK_REDUCE', 'AITER_CUSTOM', 'CUSTOM', 'SYMM_MEM', 'PYNCCL'].
|
| 23 |
+
(Worker pid=346565) INFO 10-07 08:32:31 [parallel_state.py:2308] rank 0 in world size 4 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, ETP rank 0, EP rank 0, EPLB rank N/A
|
| 24 |
+
(Worker pid=346565) INFO 10-07 08:32:31 [gpu_worker.py:501] Using V2 Model Runner
|
| 25 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [model_runner.py:375] Loading model from scratch...
|
| 26 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [cuda.py:526] Using FLASH_ATTN_MLA attention backend out of potential backends: ['FLASH_ATTN_MLA', 'FLASHMLA', 'TRITON_MLA'].
|
| 27 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [selector.py:207] Using FLASH_ATTN MLA prefill backend.
|
| 28 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [deep_gemm.py:197] deep_gemm not found in site-packages, trying vendored vllm.third_party.deep_gemm
|
| 29 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [deep_gemm.py:224] DeepGEMM PDL enabled on vllm.third_party.deep_gemm.
|
| 30 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [deep_gemm.py:136] DeepGEMM E8M0 enabled on current platform.
|
| 31 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:31 [unquantized.py:364] Using TRITON Unquantized MoE backend out of potential backends: ['TRITON', 'BATCHED_TRITON', 'FlashInfer TRTLLM', 'FlashInfer CUTLASS'].
|
| 32 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:32 [base_loader.py:81] Initializing model took 0.602535 seconds
|
| 33 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:32 [weight_utils.py:1063] Filesystem type for checkpoints: EXT4. Checkpoint size: 221.61 GiB. Available RAM: 1946.84 GiB.
|
| 34 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:32:32 [weight_utils.py:1086] Auto-prefetch is disabled because the filesystem (EXT4) is not a recognized network FS (NFS/Lustre). If you want to force prefetching, start vLLM with --safetensors-load-strategy=prefetch.
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
|
| 38 |
+
|
| 39 |
+
|
| 40 |
+
|
| 41 |
+
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
|
| 59 |
+
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:29 [default_loader.py:484] Loading weights took 56.98 seconds
|
| 76 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:29 [unquantized.py:528] Using MoEPrepareAndFinalizeNoDPEPModular
|
| 77 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:29 [unquantized.py:529] Using TritonExperts MoE backend
|
| 78 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:30 [model_runner.py:407] Model loading took 55.83 GiB memory and 58.786966 seconds
|
| 79 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:30 [model_runner.py:407] Model loading took 55.83 GiB memory and 58.955678 seconds
|
| 80 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:30 [topk_topp_sampler.py:124] Using FlashInfer for top-p & top-k sampling.
|
| 81 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:30 [torch_utils.py:286] Reducing Torch threads from 48 to 1 for serving. Set OMP_NUM_THREADS in the external environment to override.
|
| 82 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:30 [model_runner.py:407] Model loading took 55.83 GiB memory and 59.304535 seconds
|
| 83 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:31 [model_runner.py:407] Model loading took 55.83 GiB memory and 59.716836 seconds
|
| 84 |
+
(EngineCore pid=345767) INFO 10-07 08:33:31 [torch_utils.py:286] Reducing Torch threads from 48 to 1 for serving. Set OMP_NUM_THREADS in the external environment to override.
|
| 85 |
+
(EngineCore pid=345767) INFO 10-07 08:33:31 [utils.py:347] Using LBNHC KV cache layout.
|
| 86 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:32 [monitor.py:40] [backbone] Starting torch.compile
|
| 87 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:32 [monitor.py:40] [backbone] Starting torch.compile
|
| 88 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:32 [monitor.py:40] [backbone] Starting torch.compile
|
| 89 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:32 [monitor.py:40] [backbone] Starting torch.compile
|
| 90 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:39 [backends.py:1101] [backbone] Using cache directory: /home/ubuntu/.cache/vllm/torch_compile_cache/be0b666c5f/rank_0_0/backbone for vLLM's torch.compile
|
| 91 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:39 [backends.py:1163] [backbone] Dynamo bytecode transform time: 7.58 s
|
| 92 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:39 [backends.py:1101] [backbone] Using cache directory: /home/ubuntu/.cache/vllm/torch_compile_cache/be0b666c5f/rank_2_0/backbone for vLLM's torch.compile
|
| 93 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:39 [backends.py:1163] [backbone] Dynamo bytecode transform time: 7.62 s
|
| 94 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:39 [backends.py:1101] [backbone] Using cache directory: /home/ubuntu/.cache/vllm/torch_compile_cache/be0b666c5f/rank_1_0/backbone for vLLM's torch.compile
|
| 95 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:39 [backends.py:1163] [backbone] Dynamo bytecode transform time: 7.60 s
|
| 96 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:39 [backends.py:1101] [backbone] Using cache directory: /home/ubuntu/.cache/vllm/torch_compile_cache/be0b666c5f/rank_3_0/backbone for vLLM's torch.compile
|
| 97 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:39 [backends.py:1163] [backbone] Dynamo bytecode transform time: 7.70 s
|
| 98 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:39 [flashinfer_all_reduce.py:234] Initialized FlashInfer Allreduce norm fusion workspace with backend=trtllm
|
| 99 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:52 [backends.py:398] [backbone] Compiling a graph for compile range (1, 256) takes 7.76 s
|
| 100 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:52 [backends.py:398] [backbone] Compiling a graph for compile range (1, 256) takes 7.84 s
|
| 101 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:52 [backends.py:398] [backbone] Compiling a graph for compile range (1, 256) takes 7.88 s
|
| 102 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:52 [backends.py:398] [backbone] Compiling a graph for compile range (1, 256) takes 7.84 s
|
| 103 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:53 [backends.py:398] [backbone] Compiling a graph for compile range (257, 8192) takes 8.42 s
|
| 104 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:53 [backends.py:398] [backbone] Compiling a graph for compile range (257, 8192) takes 8.51 s
|
| 105 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:53 [backends.py:398] [backbone] Compiling a graph for compile range (257, 8192) takes 8.50 s
|
| 106 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:53 [backends.py:398] [backbone] Compiling a graph for compile range (257, 8192) takes 8.55 s
|
| 107 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:58 [backends.py:927] collected artifacts: 74 entries, 7 artifacts, 11186339 bytes total
|
| 108 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:58 [decorators.py:743] [backbone] saved AOT compiled function to /home/ubuntu/.cache/vllm/torch_compile_cache/torch_aot_compile/950345b37a831e6be243860f9825d2ebf415352169ce97427634f6678c1a91a0/rank_0_0/model
|
| 109 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:58 [monitor.py:62] [backbone] torch.compile took 26.11 s in total
|
| 110 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:58 [backends.py:927] collected artifacts: 74 entries, 7 artifacts, 11015834 bytes total
|
| 111 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:58 [decorators.py:743] [backbone] saved AOT compiled function to /home/ubuntu/.cache/vllm/torch_compile_cache/torch_aot_compile/950345b37a831e6be243860f9825d2ebf415352169ce97427634f6678c1a91a0/rank_3_0/model
|
| 112 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:58 [monitor.py:62] [backbone] torch.compile took 26.17 s in total
|
| 113 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:58 [backends.py:927] collected artifacts: 74 entries, 7 artifacts, 11263745 bytes total
|
| 114 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:58 [decorators.py:743] [backbone] saved AOT compiled function to /home/ubuntu/.cache/vllm/torch_compile_cache/torch_aot_compile/950345b37a831e6be243860f9825d2ebf415352169ce97427634f6678c1a91a0/rank_1_0/model
|
| 115 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:58 [monitor.py:62] [backbone] torch.compile took 26.25 s in total
|
| 116 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:58 [backends.py:927] collected artifacts: 74 entries, 7 artifacts, 11251721 bytes total
|
| 117 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:58 [decorators.py:743] [backbone] saved AOT compiled function to /home/ubuntu/.cache/vllm/torch_compile_cache/torch_aot_compile/950345b37a831e6be243860f9825d2ebf415352169ce97427634f6678c1a91a0/rank_2_0/model
|
| 118 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:58 [monitor.py:62] [backbone] torch.compile took 26.35 s in total
|
| 119 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:59 [fused_moe.py:1150] Using configuration from /home/ubuntu/chatbot-ft/.venv-bench/lib/python3.12/site-packages/vllm/model_executor/layers/fused_moe/configs/E=128,N=512,device_name=NVIDIA_H100_80GB_HBM3.json for MoE layer.
|
| 120 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:33:59 [monitor.py:91] [backbone] Initial profiling/warmup run took 1.67 s
|
| 121 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:33:59 [monitor.py:91] [backbone] Initial profiling/warmup run took 1.61 s
|
| 122 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:33:59 [monitor.py:91] [backbone] Initial profiling/warmup run took 1.76 s
|
| 123 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:33:59 [monitor.py:91] [backbone] Initial profiling/warmup run took 1.84 s
|
| 124 |
+
|
| 125 |
+
|
| 126 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:21 [model_runner.py:1109] Graph capturing finished in 20 secs, took 0.68 GiB
|
| 127 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:21 [model_runner.py:1109] Graph capturing finished in 20 secs, took 0.68 GiB
|
| 128 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:21 [model_runner.py:1109] Graph capturing finished in 20 secs, took 0.68 GiB
|
| 129 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:21 [model_runner.py:1109] Graph capturing finished in 20 secs, took 0.68 GiB
|
| 130 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:22 [gpu_worker.py:741] CUDA graph memory profiling is enabled (default since v0.21.0). The current --gpu-memory-utilization=0.9000 is equivalent to --gpu-memory-utilization=0.8877 without CUDA graph memory profiling. To maintain the same effective KV cache size as before, increase --gpu-memory-utilization to 0.9123. To disable, set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0.
|
| 131 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:22 [gpu_worker.py:741] CUDA graph memory profiling is enabled (default since v0.21.0). The current --gpu-memory-utilization=0.9000 is equivalent to --gpu-memory-utilization=0.8877 without CUDA graph memory profiling. To maintain the same effective KV cache size as before, increase --gpu-memory-utilization to 0.9123. To disable, set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0.
|
| 132 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:22 [gpu_worker.py:726] Available KV cache memory: 12.57 GiB
|
| 133 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:22 [gpu_worker.py:741] CUDA graph memory profiling is enabled (default since v0.21.0). The current --gpu-memory-utilization=0.9000 is equivalent to --gpu-memory-utilization=0.8877 without CUDA graph memory profiling. To maintain the same effective KV cache size as before, increase --gpu-memory-utilization to 0.9123. To disable, set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0.
|
| 134 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:22 [gpu_worker.py:741] CUDA graph memory profiling is enabled (default since v0.21.0). The current --gpu-memory-utilization=0.9000 is equivalent to --gpu-memory-utilization=0.8877 without CUDA graph memory profiling. To maintain the same effective KV cache size as before, increase --gpu-memory-utilization to 0.9123. To disable, set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0.
|
| 135 |
+
(EngineCore pid=345767) INFO 10-07 08:34:22 [kv_cache_utils.py:2481] GPU KV cache size: 585,936 tokens, Maximum concurrency for 8,192 tokens per request: 71.53x
|
| 136 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:22 [kernel_warmup.py:173] JIT kernel warmup starting.
|
| 137 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:22 [kernel_warmup.py:173] JIT kernel warmup starting.
|
| 138 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:22 [kernel_warmup.py:173] JIT kernel warmup starting.
|
| 139 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:22 [kernel_warmup.py:173] JIT kernel warmup starting.
|
| 140 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:22 [kernel_warmup.py:186] JIT kernel warmup finished in 0.00s.
|
| 141 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:22 [kernel_warmup.py:186] JIT kernel warmup finished in 0.00s.
|
| 142 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:22 [kernel_warmup.py:186] JIT kernel warmup finished in 0.00s.
|
| 143 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:22 [kernel_warmup.py:186] JIT kernel warmup finished in 0.00s.
|
| 144 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:22 [kernel_warmup.py:95] Warming up ll_bf16 router GEMM kernels for shapes: ((4096, 128),).
|
| 145 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:23 [kernel_warmup.py:483] Using FlashInfer autotune cache file: /home/ubuntu/.cache/vllm/flashinfer_autotune_cache/0.7.0.post1/90a/c9e6c734e1d40f4034ba61ab0a7c28c410e77896684f953346c1358c0d56129f/autotune_configs_dp0_rank0.json
|
| 146 |
+
2026-10-07 08:34:23,865 - INFO - autotuner.py:1046 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
|
| 147 |
+
2026-10-07 08:34:23,865 - INFO - autotuner.py:1046 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
|
| 148 |
+
2026-10-07 08:34:23,865 - INFO - autotuner.py:1046 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
|
| 149 |
+
2026-10-07 08:34:23,865 - INFO - autotuner.py:1046 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
|
| 150 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:23 [kernel_warmup.py:367] Running FlashInfer autotune with 8192 tokens and token buckets (1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2560, 3072, 3584, 4096, 8192).
|
| 151 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:23 [kernel_warmup.py:367] Running FlashInfer autotune with 8192 tokens and token buckets (1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2560, 3072, 3584, 4096, 8192).
|
| 152 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:23 [kernel_warmup.py:367] Running FlashInfer autotune with 8192 tokens and token buckets (1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2560, 3072, 3584, 4096, 8192).
|
| 153 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:23 [kernel_warmup.py:367] Running FlashInfer autotune with 8192 tokens and token buckets (1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 768, 1024, 1280, 1536, 1792, 2048, 2560, 3072, 3584, 4096, 8192).
|
| 154 |
+
2026-10-07 08:34:23,944 - INFO - autotuner.py:1069 - flashinfer.jit: [Autotuner]: Autotuning process ends
|
| 155 |
+
2026-10-07 08:34:23,945 - INFO - autotuner.py:1069 - flashinfer.jit: [Autotuner]: Autotuning process ends
|
| 156 |
+
2026-10-07 08:34:23,945 - INFO - autotuner.py:1069 - flashinfer.jit: [Autotuner]: Autotuning process ends
|
| 157 |
+
2026-10-07 08:34:23,946 - INFO - autotuner.py:1069 - flashinfer.jit: [Autotuner]: Autotuning process ends
|
| 158 |
+
|
| 159 |
+
|
| 160 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:44 [model_runner.py:1109] Graph capturing finished in 17 secs, took 0.38 GiB
|
| 161 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:44 [gpu_worker.py:933] CUDA graph pool memory: 0.38 GiB (actual), 0.98 GiB (estimated), difference: 0.59 GiB (153.8%).
|
| 162 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:44 [gpu_worker.py:996] Free memory on device (77.04/79.18 GiB) on startup. Desired GPU memory utilization is (0.9, 71.26 GiB). Actual usage is 57.08 GiB for consumed memory (weights + non-torch), 1.61 GiB for peak activation, and 0.38 GiB for CUDAGraph memory. Replace gpu_memory_utilization config with `--kv-cache-memory=12929885799` (12.04 GiB) to fit into requested memory, or `--kv-cache-memory=19132383232` (17.82 GiB) to fully utilize gpu memory. Current kv cache memory in use is 12.57 GiB.
|
| 163 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:44 [model_runner.py:1109] Graph capturing finished in 17 secs, took 0.38 GiB
|
| 164 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:44 [gpu_worker.py:933] CUDA graph pool memory: 0.38 GiB (actual), 0.98 GiB (estimated), difference: 0.59 GiB (153.8%).
|
| 165 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:44 [gpu_worker.py:996] Free memory on device (77.04/79.18 GiB) on startup. Desired GPU memory utilization is (0.9, 71.26 GiB). Actual usage is 57.08 GiB for consumed memory (weights + non-torch), 1.61 GiB for peak activation, and 0.38 GiB for CUDAGraph memory. Replace gpu_memory_utilization config with `--kv-cache-memory=12929885799` (12.04 GiB) to fit into requested memory, or `--kv-cache-memory=19132383232` (17.82 GiB) to fully utilize gpu memory. Current kv cache memory in use is 12.57 GiB.
|
| 166 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:44 [model_runner.py:1109] Graph capturing finished in 17 secs, took 0.38 GiB
|
| 167 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:44 [gpu_worker.py:933] CUDA graph pool memory: 0.38 GiB (actual), 0.98 GiB (estimated), difference: 0.59 GiB (153.8%).
|
| 168 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:44 [gpu_worker.py:996] Free memory on device (77.04/79.18 GiB) on startup. Desired GPU memory utilization is (0.9, 71.26 GiB). Actual usage is 57.08 GiB for consumed memory (weights + non-torch), 1.61 GiB for peak activation, and 0.38 GiB for CUDAGraph memory. Replace gpu_memory_utilization config with `--kv-cache-memory=12929885799` (12.04 GiB) to fit into requested memory, or `--kv-cache-memory=19132383232` (17.82 GiB) to fully utilize gpu memory. Current kv cache memory in use is 12.57 GiB.
|
| 169 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:44 [model_runner.py:1109] Graph capturing finished in 17 secs, took 0.38 GiB
|
| 170 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:44 [gpu_worker.py:933] CUDA graph pool memory: 0.38 GiB (actual), 0.98 GiB (estimated), difference: 0.59 GiB (153.8%).
|
| 171 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:44 [gpu_worker.py:996] Free memory on device (77.04/79.18 GiB) on startup. Desired GPU memory utilization is (0.9, 71.26 GiB). Actual usage is 57.08 GiB for consumed memory (weights + non-torch), 1.61 GiB for peak activation, and 0.38 GiB for CUDAGraph memory. Replace gpu_memory_utilization config with `--kv-cache-memory=12929885799` (12.04 GiB) to fit into requested memory, or `--kv-cache-memory=19132383232` (17.82 GiB) to fully utilize gpu memory. Current kv cache memory in use is 12.57 GiB.
|
| 172 |
+
(Worker_TP3 pid=346568) INFO 10-07 08:34:45 [jit_monitor.py:84] Kernel JIT monitor activated; monitored JIT compilations during inference will use mode=warn.
|
| 173 |
+
(Worker_TP1 pid=346566) INFO 10-07 08:34:45 [jit_monitor.py:84] Kernel JIT monitor activated; monitored JIT compilations during inference will use mode=warn.
|
| 174 |
+
(Worker_TP2 pid=346567) INFO 10-07 08:34:45 [jit_monitor.py:84] Kernel JIT monitor activated; monitored JIT compilations during inference will use mode=warn.
|
| 175 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:34:45 [jit_monitor.py:84] Kernel JIT monitor activated; monitored JIT compilations during inference will use mode=warn.
|
| 176 |
+
(EngineCore pid=345767) INFO 10-07 08:34:45 [core.py:373] init engine (profile, create kv cache, warmup model) took 74.67 s (compilation: 26.35 s)
|
| 177 |
+
(EngineCore pid=345767) INFO 10-07 08:34:45 [config.py:514] Inferred from consolidated*.safetensors files torch.bfloat16 dtype.
|
| 178 |
+
(EngineCore pid=345767) INFO 10-07 08:34:47 [kernel.py:463] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native'], gelu_and_mul_sparse=['triton', 'native'])
|
| 179 |
+
(EngineCore pid=345767) INFO 10-07 08:34:48 [compilation.py:331] Enabled custom fusions: allreduce_rms
|
| 180 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [entry.py:135] Supported tasks: ['generate']
|
| 181 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [factories.py:76] Scale-out endpoints are disabled. Set --enable-scale-out to enable them.
|
| 182 |
+
(APIServer pid=344642) WARNING 10-07 08:34:48 [serving.py:258] The chat template requires system-first ordering, so inline system messages in /v1/messages requests (e.g. Claude Code's per-turn reminders) are merged into the leading system prompt. Each new one changes the prompt prefix, so the rest of the conversation misses the prefix cache. Pass a --chat-template that accepts non-leading system messages to avoid this.
|
| 183 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [entry.py:139] Starting vLLM server on http://127.0.0.1:18000
|
| 184 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:60] Available routes are:
|
| 185 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /openapi.json, Methods: HEAD, GET
|
| 186 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /docs, Methods: HEAD, GET
|
| 187 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /docs/oauth2-redirect, Methods: HEAD, GET
|
| 188 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /redoc, Methods: HEAD, GET
|
| 189 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /load, Methods: GET
|
| 190 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /version, Methods: GET
|
| 191 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /kv_event_sources, Methods: GET
|
| 192 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /health, Methods: GET
|
| 193 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /metrics, Methods: GET
|
| 194 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /tokenize, Methods: POST
|
| 195 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /detokenize, Methods: POST
|
| 196 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/models, Methods: GET
|
| 197 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /ping, Methods: GET
|
| 198 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /ping, Methods: POST
|
| 199 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /invocations, Methods: POST
|
| 200 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/chat/completions, Methods: POST
|
| 201 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/chat/completions/batch, Methods: POST
|
| 202 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/responses, Methods: POST
|
| 203 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/responses/{response_id}, Methods: GET
|
| 204 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/responses/{response_id}/cancel, Methods: POST
|
| 205 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/completions, Methods: POST
|
| 206 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/messages, Methods: POST
|
| 207 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /v1/messages/count_tokens, Methods: POST
|
| 208 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /generative_scoring, Methods: POST
|
| 209 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /scale_elastic_ep, Methods: POST
|
| 210 |
+
(APIServer pid=344642) INFO 10-07 08:34:48 [launcher.py:69] Route: /is_scaling_elastic_ep, Methods: POST
|
| 211 |
+
INFO: Started server process [344642]
|
| 212 |
+
INFO: Waiting for application startup.
|
| 213 |
+
INFO: Application startup complete.
|
| 214 |
+
INFO: 127.0.0.1:38508 - "GET /health HTTP/1.1" 200 OK
|
| 215 |
+
(APIServer pid=344642) WARNING 10-07 08:34:51 [vllm.py:1327] Watermarking is enabled for this request, but the engine has no watermark configuration. This and subsequent requests will run without watermarking.
|
| 216 |
+
(Worker_TP0 pid=346565) WARNING 10-07 08:34:51 [jit_monitor.py:140] Triton kernel JIT compilation during inference: _gumbel_sample_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 217 |
+
(Worker_TP0 pid=346565) WARNING 10-07 08:34:51 [jit_monitor.py:140] Triton kernel JIT compilation during inference: fused_moe_kernel. This causes a latency spike; consider extending warmup to cover this shape/config.
|
| 218 |
+
INFO: 127.0.0.1:38900 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 219 |
+
INFO: 127.0.0.1:39134 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 220 |
+
INFO: 127.0.0.1:39308 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 221 |
+
INFO: 127.0.0.1:39104 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 222 |
+
INFO: 127.0.0.1:39284 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 223 |
+
INFO: 127.0.0.1:38938 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 224 |
+
INFO: 127.0.0.1:38908 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 225 |
+
INFO: 127.0.0.1:39250 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 226 |
+
INFO: 127.0.0.1:38594 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 227 |
+
INFO: 127.0.0.1:38774 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 228 |
+
INFO: 127.0.0.1:39266 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 229 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 230 |
+
INFO: 127.0.0.1:38668 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 231 |
+
INFO: 127.0.0.1:38982 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 232 |
+
INFO: 127.0.0.1:39066 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 233 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 234 |
+
INFO: 127.0.0.1:38930 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 235 |
+
INFO: 127.0.0.1:39260 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 236 |
+
INFO: 127.0.0.1:39358 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 237 |
+
INFO: 127.0.0.1:38632 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 238 |
+
INFO: 127.0.0.1:38654 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 239 |
+
INFO: 127.0.0.1:39144 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 240 |
+
INFO: 127.0.0.1:38926 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 241 |
+
INFO: 127.0.0.1:39200 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 242 |
+
INFO: 127.0.0.1:39304 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 243 |
+
INFO: 127.0.0.1:39360 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 244 |
+
INFO: 127.0.0.1:39366 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 245 |
+
INFO: 127.0.0.1:39410 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 246 |
+
INFO: 127.0.0.1:39442 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 247 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 248 |
+
INFO: 127.0.0.1:38748 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 249 |
+
INFO: 127.0.0.1:39010 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 250 |
+
INFO: 127.0.0.1:39154 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 251 |
+
INFO: 127.0.0.1:39386 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 252 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 253 |
+
INFO: 127.0.0.1:38708 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 254 |
+
INFO: 127.0.0.1:39408 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 255 |
+
INFO: 127.0.0.1:38816 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 256 |
+
INFO: 127.0.0.1:39020 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 257 |
+
INFO: 127.0.0.1:39318 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 258 |
+
INFO: 127.0.0.1:39388 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 259 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 260 |
+
INFO: 127.0.0.1:38866 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 261 |
+
INFO: 127.0.0.1:39080 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 262 |
+
INFO: 127.0.0.1:39446 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 263 |
+
INFO: 127.0.0.1:38790 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 264 |
+
INFO: 127.0.0.1:38968 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 265 |
+
INFO: 127.0.0.1:39380 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 266 |
+
INFO: 127.0.0.1:39420 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 267 |
+
INFO: 127.0.0.1:38620 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 268 |
+
INFO: 127.0.0.1:38698 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 269 |
+
INFO: 127.0.0.1:38732 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 270 |
+
INFO: 127.0.0.1:38832 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 271 |
+
INFO: 127.0.0.1:38924 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 272 |
+
INFO: 127.0.0.1:38994 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 273 |
+
INFO: 127.0.0.1:39052 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 274 |
+
INFO: 127.0.0.1:39444 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 275 |
+
INFO: 127.0.0.1:38848 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 276 |
+
INFO: 127.0.0.1:39350 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 277 |
+
INFO: 127.0.0.1:39282 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 278 |
+
INFO: 127.0.0.1:39436 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 279 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 280 |
+
INFO: 127.0.0.1:39416 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 281 |
+
INFO: 127.0.0.1:38706 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 282 |
+
INFO: 127.0.0.1:38586 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 283 |
+
INFO: 127.0.0.1:38684 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 284 |
+
INFO: 127.0.0.1:38762 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 285 |
+
INFO: 127.0.0.1:39096 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 286 |
+
INFO: 127.0.0.1:39208 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 287 |
+
INFO: 127.0.0.1:39218 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 288 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 289 |
+
INFO: 127.0.0.1:38718 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 290 |
+
INFO: 127.0.0.1:38858 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 291 |
+
INFO: 127.0.0.1:39120 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 292 |
+
INFO: 127.0.0.1:39334 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 293 |
+
INFO: 127.0.0.1:38794 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 294 |
+
INFO: 127.0.0.1:38880 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 295 |
+
INFO: 127.0.0.1:38672 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 296 |
+
INFO: 127.0.0.1:39036 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 297 |
+
INFO: 127.0.0.1:38614 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 298 |
+
INFO: 127.0.0.1:38962 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 299 |
+
INFO: 127.0.0.1:39168 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 300 |
+
INFO: 127.0.0.1:39240 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 301 |
+
INFO: 127.0.0.1:38896 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 302 |
+
INFO: 127.0.0.1:39290 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 303 |
+
INFO: 127.0.0.1:39188 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 304 |
+
INFO: 127.0.0.1:39224 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 305 |
+
INFO: 127.0.0.1:38750 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 306 |
+
INFO: 127.0.0.1:38642 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 307 |
+
INFO: 127.0.0.1:38726 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 308 |
+
INFO: 127.0.0.1:39184 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 309 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 310 |
+
INFO: 127.0.0.1:38954 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 311 |
+
INFO: 127.0.0.1:38892 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 312 |
+
INFO: 127.0.0.1:38608 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 313 |
+
INFO: 127.0.0.1:38740 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 314 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 315 |
+
INFO: 127.0.0.1:39404 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 316 |
+
INFO: 127.0.0.1:38780 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 317 |
+
INFO: 127.0.0.1:38806 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 318 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 319 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 320 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 321 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 322 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 323 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 324 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 325 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 326 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 327 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 328 |
+
INFO: 127.0.0.1:38608 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 329 |
+
INFO: 127.0.0.1:38594 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 330 |
+
INFO: 127.0.0.1:38586 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 331 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 332 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 333 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 334 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 335 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 336 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 337 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 338 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 339 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 340 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 341 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 342 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 343 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 344 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 345 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 346 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 347 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 348 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 349 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 350 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 351 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 352 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 353 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 354 |
+
INFO: 127.0.0.1:38608 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 355 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 356 |
+
(APIServer pid=344642) INFO 10-07 08:34:58 [loggers.py:329] Engine 000: Avg prompt throughput: 286.1 tokens/s, Avg generation throughput: 660.5 tokens/s, Running: 10 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.2%, Prefix cache hit rate: 83.1%
|
| 357 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 358 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 359 |
+
INFO: 127.0.0.1:38780 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 360 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 361 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 362 |
+
INFO: 127.0.0.1:38586 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 363 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 364 |
+
INFO: 127.0.0.1:38594 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 365 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 366 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 367 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 368 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 369 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 370 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 371 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 372 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 373 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 374 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 375 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 376 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 377 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 378 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 379 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 380 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 381 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 382 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 383 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 384 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 385 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 386 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 387 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 388 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 389 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 390 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 391 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 392 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 393 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 394 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 395 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 396 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 397 |
+
INFO: 127.0.0.1:38586 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 398 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 399 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 400 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 401 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 402 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 403 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 404 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 405 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 406 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 407 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 408 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 409 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 410 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 411 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 412 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 413 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 414 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 415 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 416 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 417 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 418 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 419 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 420 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 421 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 422 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 423 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 424 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 425 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 426 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 427 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 428 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 429 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 430 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 431 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 432 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 433 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 434 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 435 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 436 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 437 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 438 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 439 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 440 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 441 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 442 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 443 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 444 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 445 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 446 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 447 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 448 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 449 |
+
INFO: 127.0.0.1:38586 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 450 |
+
INFO: 127.0.0.1:38572 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 451 |
+
INFO: 127.0.0.1:38574 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 452 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 453 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 454 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 455 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 456 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 457 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 458 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 459 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 460 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 461 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 462 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 463 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 464 |
+
INFO: 127.0.0.1:38554 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 465 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 466 |
+
INFO: 127.0.0.1:38568 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 467 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 468 |
+
INFO: 127.0.0.1:38570 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 469 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 470 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 471 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 472 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 473 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 474 |
+
(APIServer pid=344642) INFO 10-07 08:35:08 [loggers.py:329] Engine 000: Avg prompt throughput: 384.8 tokens/s, Avg generation throughput: 676.2 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 83.7%
|
| 475 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 476 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 477 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 478 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 479 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 480 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 481 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 482 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 483 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 484 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 485 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 486 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 487 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 488 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 489 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 490 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 491 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 492 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 493 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 494 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 495 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 496 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 497 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 498 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 499 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 500 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 501 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 502 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 503 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 504 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 505 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 506 |
+
INFO: 127.0.0.1:38550 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 507 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 508 |
+
INFO: 127.0.0.1:39982 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 509 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 510 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 511 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 512 |
+
(APIServer pid=344642) INFO 10-07 08:35:18 [loggers.py:329] Engine 000: Avg prompt throughput: 139.7 tokens/s, Avg generation throughput: 241.0 tokens/s, Running: 3 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.2%, Prefix cache hit rate: 84.5%
|
| 513 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 514 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 515 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 516 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 517 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 518 |
+
INFO: 127.0.0.1:38522 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 519 |
+
INFO: 127.0.0.1:38532 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 520 |
+
INFO: 127.0.0.1:38542 - "POST /v1/chat/completions HTTP/1.1" 200 OK
|
| 521 |
+
(APIServer pid=344642) INFO 10-07 08:35:28 [loggers.py:329] Engine 000: Avg prompt throughput: 15.0 tokens/s, Avg generation throughput: 64.7 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 84.8%
|
| 522 |
+
(APIServer pid=344642) INFO 10-07 08:35:38 [loggers.py:329] Engine 000: Avg prompt throughput: 0.0 tokens/s, Avg generation throughput: 0.0 tokens/s, Running: 0 reqs, Waiting: 0 reqs, GPU KV cache usage: 0.0%, Prefix cache hit rate: 84.8%
|
| 523 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [launcher.py:128] [shutdown] API server: shutdown triggered
|
| 524 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [launcher.py:144] [shutdown] API server: stopping engine client mode=abort timeout=0s
|
| 525 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [core_client.py:783] [shutdown] MPClient: start timeout=0s
|
| 526 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [core_client.py:785] [shutdown] MPClient: stopping engine manager
|
| 527 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [utils.py:655] [shutdown] Process manager: send sigterm to process EngineCore
|
| 528 |
+
(APIServer pid=344642) WARNING 10-07 08:35:48 [utils.py:675] [shutdown] Process manager: force killing remaining processes count=1
|
| 529 |
+
(APIServer pid=344642) WARNING 10-07 08:35:48 [utils.py:680] [shutdown] Process manager: force killing remaining process EngineCore pid 345767
|
| 530 |
+
(EngineCore pid=345767) INFO 10-07 08:35:48 [core.py:1423] [shutdown] EngineCore: trigger received signal=SIGTERM
|
| 531 |
+
(EngineCore pid=345767) INFO 10-07 08:35:48 [core.py:1572] [shutdown] EngineCore: start mode=abort timeout=0s
|
| 532 |
+
(EngineCore pid=345767) INFO 10-07 08:35:48 [core.py:1603] [shutdown] EngineCore: request processing complete; starting resource teardown
|
| 533 |
+
(EngineCore pid=345767) INFO 10-07 08:35:48 [core.py:1436] [shutdown] EngineCore: exiting busy loop
|
| 534 |
+
(Worker_TP0 pid=346565) INFO 10-07 08:35:48 [multiproc_executor.py:838] Parent process exited, terminating worker queues
|
| 535 |
+
(EngineCore pid=345767) INFO 10-07 08:35:48 [multiproc_executor.py:474] [shutdown] Executor: waiting for worker exit count=4
|
| 536 |
+
(APIServer pid=344642) WARNING 10-07 08:35:48 [system_utils.py:224] [shutdown] force killing sub process VLLM::Worker_TP0 pid 346565
|
| 537 |
+
(APIServer pid=344642) WARNING 10-07 08:35:48 [system_utils.py:224] [shutdown] force killing sub process VLLM::Worker_TP1 pid 346566
|
| 538 |
+
(APIServer pid=344642) WARNING 10-07 08:35:48 [system_utils.py:224] [shutdown] force killing sub process VLLM::Worker_TP2 pid 346567
|
| 539 |
+
(APIServer pid=344642) WARNING 10-07 08:35:48 [system_utils.py:224] [shutdown] force killing sub process VLLM::Worker_TP3 pid 346568
|
| 540 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [core_client.py:787] [shutdown] MPClient: engine manager stopped
|
| 541 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [core_client.py:788] [shutdown] MPClient: cleaning up background resources
|
| 542 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [core_client.py:790] [shutdown] MPClient: complete
|
| 543 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [launcher.py:153] [shutdown] API server: engine client stopped
|
| 544 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [launcher.py:156] [shutdown] API server: signalling HTTP server shutdown
|
| 545 |
+
(APIServer pid=344642) INFO 10-07 08:35:48 [launcher.py:178] [shutdown] API server: shutting down FastAPI HTTP server
|
| 546 |
+
INFO: Shutting down
|
| 547 |
+
INFO: Waiting for application shutdown.
|
| 548 |
+
INFO: Application shutdown complete.
|
| 549 |
+
/usr/lib/python3.12/multiprocessing/resource_tracker.py:254: UserWarning: resource_tracker: There appear to be 4 leaked semaphore objects to clean up at shutdown
|
| 550 |
+
warnings.warn('resource_tracker: There appear to be %d '
|
| 551 |
+
/usr/lib/python3.12/multiprocessing/resource_tracker.py:254: UserWarning: resource_tracker: There appear to be 6 leaked shared_memory objects to clean up at shutdown
|
| 552 |
+
warnings.warn('resource_tracker: There appear to be %d '
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/review_anthropic-claude-opus-5-5/MANGO.md
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/summary_anthropic-claude-opus-5-5.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_a": "hf:models/mistral_small4-sft-sft-dpo-native",
|
| 3 |
+
"model_b": "hf:Bigglz/qwen-sft-sft-dpo-tone",
|
| 4 |
+
"judge_model": "anthropic:claude-opus-5-5",
|
| 5 |
+
"user_model": "anthropic:claude-opus-5-5",
|
| 6 |
+
"turns": 3,
|
| 7 |
+
"overall": {
|
| 8 |
+
"n_pairs_total": 100,
|
| 9 |
+
"n_pairs": 100,
|
| 10 |
+
"A_win": 12,
|
| 11 |
+
"B_win": 83,
|
| 12 |
+
"tie_inconsistent": 5,
|
| 13 |
+
"A_score": 0.145,
|
| 14 |
+
"A_win_rate_decisive": 0.12631578947368421,
|
| 15 |
+
"A_win_rate_decisive_ci95": [
|
| 16 |
+
0.07375581468316135,
|
| 17 |
+
0.20792315449580787
|
| 18 |
+
],
|
| 19 |
+
"judge_consistency": 0.95,
|
| 20 |
+
"position_bias_first": 0.505,
|
| 21 |
+
"invalid_output_rate": 0.0,
|
| 22 |
+
"per_pass_A_rate": {
|
| 23 |
+
"1": 0.15,
|
| 24 |
+
"2": 0.14
|
| 25 |
+
}
|
| 26 |
+
},
|
| 27 |
+
"by_supporter": {
|
| 28 |
+
"MANGO": {
|
| 29 |
+
"n_pairs_total": 100,
|
| 30 |
+
"n_pairs": 100,
|
| 31 |
+
"A_win": 12,
|
| 32 |
+
"B_win": 83,
|
| 33 |
+
"tie_inconsistent": 5,
|
| 34 |
+
"A_score": 0.145,
|
| 35 |
+
"A_win_rate_decisive": 0.12631578947368421,
|
| 36 |
+
"A_win_rate_decisive_ci95": [
|
| 37 |
+
0.07375581468316135,
|
| 38 |
+
0.20792315449580787
|
| 39 |
+
],
|
| 40 |
+
"judge_consistency": 0.95,
|
| 41 |
+
"position_bias_first": 0.505,
|
| 42 |
+
"invalid_output_rate": 0.0,
|
| 43 |
+
"per_pass_A_rate": {
|
| 44 |
+
"1": 0.15,
|
| 45 |
+
"2": 0.14
|
| 46 |
+
}
|
| 47 |
+
}
|
| 48 |
+
},
|
| 49 |
+
"response_stats": {
|
| 50 |
+
"A": {
|
| 51 |
+
"n_conversations": 100,
|
| 52 |
+
"avg_response_chars": 105.56
|
| 53 |
+
},
|
| 54 |
+
"B": {
|
| 55 |
+
"n_conversations": 100,
|
| 56 |
+
"avg_response_chars": 38.696666666666665
|
| 57 |
+
}
|
| 58 |
+
}
|
| 59 |
+
}
|
judge/pairs/Bigglz-mistral-small4-119b-sft-sft-dpo_vs_Bigglz-qwen-sft-sft-dpo-tone/summary_anthropic-claude-opus-5-5.md
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# LLM-as-a-judge: A vs B
|
| 2 |
+
|
| 3 |
+
- A: `hf:models/mistral_small4-sft-sft-dpo-native`
|
| 4 |
+
- B: `hf:Bigglz/qwen-sft-sft-dpo-tone`
|
| 5 |
+
- 평가자: `anthropic:claude-opus-5-5` (순서를 바꿔 2회 판정)
|
| 6 |
+
- 사용자 시뮬레이터: `anthropic:claude-opus-5-5`, 챗봇 3턴
|
| 7 |
+
|
| 8 |
+
| 페르소나 | 유효/전체 쌍 | A 승 | B 승 | 무승부(비일관) | A 점수 | 평가자 일관성 | 첫 번째(K) 선택률 | 해석 불가 |
|
| 9 |
+
|---|---:|---:|---:|---:|---:|---:|---:|---:|
|
| 10 |
+
| **전체** | 100/100 | 12 | 83 | 5 | 14.5% | 95.0% | 50.5% | 0.0% |
|
| 11 |
+
| MANGO | 100/100 | 12 | 83 | 5 | 14.5% | 95.0% | 50.5% | 0.0% |
|
| 12 |
+
|
| 13 |
+
- 유효 쌍: 두 판정 모두 J/K로 해석된 쌍 (해석 불가 출력이 있는 쌍은 승패 집계에서 제외)
|
| 14 |
+
- A 점수 = (A 승 + 0.5 x 무승부) / 유효 쌍. 승부가 난 쌍 기준 A 승률 12.6% (95% CI 7.4% ~ 20.8%)
|
| 15 |
+
- 평가자 일관성: 순서를 바꿔도 같은 모델을 고른 비율 (높을수록 신뢰할 수 있는 판정)
|
| 16 |
+
- 첫 번째(K) 선택률: 50%에서 멀수록 위치 편향이 큼
|
| 17 |
+
- 회차별 A 선택률: 1회차(A가 K) 15.0%, 2회차(A가 J) 14.0%
|
| 18 |
+
- 평균 응답 길이(글자): A 106, B 39
|