Record actual v4 preparation and model-waiter start after manual verification

This commit is contained in:
Frank14f committed 2026-11-16 16:13:21 +08:00
1 parent b949fb4904
commit d003f25bee
9 files changed
+256

No files matched your search

@@ -0,0 +1,67 @@
{
"status": "SERVICES_STARTED",
"unix": 1794816683.6771722,
"services": {
"preparation": {
"state": {
"Restart": "no",
"MainPID": "510346",
"ActiveState": "active",
"SubState": "running"
},
"pid": 510346,
"ppid": 2234,
"create_time": 1794816680.58,
"cmdline": [
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
"-u",
"tools/run_gre24_protocol_preparation.py",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol"
]
},
"model_chain": {
"state": {
"Restart": "no",
"MainPID": "510350",
"ActiveState": "active",
"SubState": "running"
},
"pid": 510350,
"ppid": 2234,
"create_time": 1794816682.12,
"cmdline": [
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
"-u",
"tools/run_gre24_models.py",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol/model-plan.json"
]
}
},
"input_freeze": {
"status": "FROZEN",
"case_count": 629,
"splits": 5,
"contract_sha256": "8ea9bdb9f1aa542a044b383e7191696e6ac8b3ed1aedd9b2b34e7571cd7e2c1e",
"splits_sha256": "534b2a25f92dbac2335b6282e1dfe0bb8ecdbc8db55bdfc24c7cffa83f9e931b",
"model_plan_sha256": "15518ad03ee7b9c14e76fbd968c1f5ff7bd523d165a03113c8dbb8c6c88c444c",
"metadata_cpu_seconds": 1.2558805670000002,
"original_scientific_sources_unchanged": true,
"unix": 1794816556.651451
},
"source_pins_verified": true,
"manual_receive_verification_sha256": "3928c8a3acb9f928843fe2120003391f4eed2727ddf478bb74e99e48321c2bd1",
"preparation_start": {
"pid": 510346,
"unix": 1794816681.6171403,
"contract_sha256": "8ea9bdb9f1aa542a044b383e7191696e6ac8b3ed1aedd9b2b34e7571cd7e2c1e"
},
"model_chain_start": {
"pid": 510350,
"unix": 1794816683.204227,
"plan_sha256": "15518ad03ee7b9c14e76fbd968c1f5ff7bd523d165a03113c8dbb8c6c88c444c",
"stage": "WAIT_PREPARATION"
},
"no_receiver_restart": true,
"original_stopped_waiter_not_restarted": true
}
@@ -0,0 +1,8 @@
{
"pid": 510341,
"unix": 1794816680.574639,
"authorizing_ruling": "1794815915.570349",
"inputs_commit": "b949fb4",
"manual_verification_sha256": "3928c8a3acb9f928843fe2120003391f4eed2727ddf478bb74e99e48321c2bd1",
"no_receiver_restart": true
}
@@ -0,0 +1,106 @@
{
"unix": 1794816723.265358,
"processes": [
{
"pid": 510346,
"ppid": 2234,
"cpu_affinity": [
30,
31
],
"children": [
{
"pid": 510347,
"ppid": 510346,
"cpu_affinity": [
8
],
"cmdline": [
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
"/home/frank14f/optane/MeltMind/tools/run_gre24_protocol_preparation.py",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol",
"--guard",
"0",
"--slot",
"0"
]
},
{
"pid": 510362,
"ppid": 510347,
"cpu_affinity": [
8
],
"cmdline": [
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
"tools/gre24_prepare_v2.py",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol",
"0"
]
}
]
},
{
"pid": 510350,
"ppid": 2234,
"cpu_affinity": [
0,
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28,
29,
30,
31
],
"children": []
}
],
"progress": {
"completed": 0,
"attempted": 1,
"total": 629,
"active": {
"0": 0
},
"settled_worker_cpu_seconds": 60.0,
"reserved_active_worker_cpu_seconds": 1802,
"stop_reason": null,
"first_part_checked": false,
"elapsed_seconds": 41.2004451029934
},
"first_case": "00000003_1ffb81a71e5b402e966b9341_step_002",
"source_pins_verified": true,
"first_worker_progress": {
"stage": "EIKONAL_DW",
"case": "00000003_1ffb81a71e5b402e966b9341_step_002",
"pid": 510362,
"unix": 1794816685.245323,
"CPU_s": 1.4860454609999998,
"wall_s": 1.5792233029496856
}
}
@@ -0,0 +1,26 @@
{
"command": [
"systemd-run",
"--user",
"--unit=meltmind-gre24-v4-models-20261116",
"--working-directory=/home/frank14f/optane/MeltMind",
"--property=Restart=no",
"--property=KillMode=control-group",
"--property=TimeoutStopSec=30s",
"--property=StandardOutput=append:/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol/meltmind-gre24-v4-models-20261116.log",
"--property=StandardError=inherit",
"--setenv=PYTHONUNBUFFERED=1",
"--setenv=OMP_NUM_THREADS=1",
"--setenv=OPENBLAS_NUM_THREADS=1",
"--setenv=MKL_NUM_THREADS=1",
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
"-u",
"tools/run_gre24_models.py",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol/model-plan.json"
],
"returncode": 0,
"stdout": "",
"stderr": "Running as unit: meltmind-gre24-v4-models-20261116.service\n",
"unix": 1794816682.1513035
}
@@ -0,0 +1,26 @@
{
"command": [
"systemd-run",
"--user",
"--unit=meltmind-gre24-v4-prep-20261116",
"--working-directory=/home/frank14f/optane/MeltMind",
"--property=Restart=no",
"--property=KillMode=control-group",
"--property=TimeoutStopSec=30s",
"--property=StandardOutput=append:/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol/meltmind-gre24-v4-prep-20261116.log",
"--property=StandardError=inherit",
"--setenv=PYTHONUNBUFFERED=1",
"--setenv=OMP_NUM_THREADS=1",
"--setenv=OPENBLAS_NUM_THREADS=1",
"--setenv=MKL_NUM_THREADS=1",
"--setenv=CUDA_VISIBLE_DEVICES=",
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
"-u",
"tools/run_gre24_protocol_preparation.py",
"/home/frank14f/optane/MeltMind/artifacts/gre24-v4-reproduction-2026-11-16/protocol"
],
"returncode": 0,
"stdout": "",
"stderr": "Running as unit: meltmind-gre24-v4-prep-20261116.service\n",
"unix": 1794816680.6121638
}
@@ -0,0 +1,6 @@
{
"pid": 510350,
"unix": 1794816683.204227,
"plan_sha256": "15518ad03ee7b9c14e76fbd968c1f5ff7bd523d165a03113c8dbb8c6c88c444c",
"stage": "WAIT_PREPARATION"
}
@@ -0,0 +1,5 @@
{
"pid": 510346,
"unix": 1794816681.6171403,
"contract_sha256": "8ea9bdb9f1aa542a044b383e7191696e6ac8b3ed1aedd9b2b34e7571cd7e2c1e"
}
@@ -121,3 +121,5 @@ Grok1794801467唯一对象完成已实读,Horsea尚无文件/传输。Claude17
Claude统筹研究;MeltAstra负责Horsea主运行和证据,MoldAstra负责数学规格;用户新fork的「MeltAstra辅助」接替Grok的审计、检索和文献工作,Grok暂停;TRON只转达和同步。Melt已在1794816379实际确认分工与状态。Mold的R4与空间矩向量化仅纸面,均不得混入当前v4科学源码。临时Drive交付文件清理由辅助依Claude安排处理,tailnet遗留记录开放。
16:05:45实查手动文件35925187752字节、无aria2控制文件或下载进程、旧两服务停止、科学启动回执不存在、16固定文件pins通过。用户MD5确认与Melt独立核验分别记录;独立核验服务PID509976随后实际开始。后续核验与科学运行真实状态以`artifacts/gre24-v4-reproduction-2026-11-16/data/manual-receive-verification.json`和`protocol`实际回执为准,不把用户报告或服务等待当作已核验/已执行结果。
16:08:41独立MD5与大小已通过,记录Yanqi手动接收;合同8ea9bdb9/划分534b2a25/模型计划15518ad0固定于b949fb4并推送读回。16:11:21唯一原准备systemd/PID510346实际启动,16:11:23模型等待PID510350启动;16:12:03首worker510362/CPU8在EIKONAL_DW,首件成本门尚待,模型WAIT_PREPARATION未拟合。源码、预算及科学方案不改,旧接收/等待器不重启;最新实际证据见docs/gre24-v4-manual-receive-and-run-2026-11-16.md。
@@ -21,3 +21,13 @@ Yanqi手动通过aria2完成接收并确认MD5。Melt随后于16:08:41(Asia/Sh
准备子帽144000 worker CPU秒、8workers、逐件1800 CPU秒/1800墙钟秒/4GiB;准备设置预留60秒涵盖新增合成与元数据。准备、拟合、后处理共享新216000 worker CPU秒,FNN等待/设置/运行共享7200 GPU秒;旧v5费用不重复计入。原模型参数、特征定义、失败保留、全准备终态→模型终态/冻结→预测冻结→评分的顺序保持。
本记录初次保存时,输入已冻结,尚未启动科学服务。接下来仅调用b01645c的已验收启动函数启动原准备脚本与原模型等待脚本;不调用旧接收等待入口。实际开始标记、PID及预算状态随后单列,不将批准或元数据冻结写成计算已经完成。
## 实际启动:16:11–16:12
输入与手动接收证据已固定推送 **b949fb49042d6835e18f552ad321dec22e7acddd**,远端ref读回一致。仅调用b01645c启动函数,旧receiver与旧等待器均未重启。
准备服务`meltmind-gre24-v4-prep-20261116.service`于16:11:21记录 **PID510346**;模型服务`meltmind-gre24-v4-models-20261116.service`于16:11:23记录 **PID510350**,状态WAIT_PREPARATION。两个主进程父进程均为用户systemd2234,Restart=no。启动工具已退出0,服务独立运行。
16:12:03实际核首件`00000003_1ffb81a71e5b402e966b9341_step_002`:guard510347、worker510362,CPU亲和性8;最近worker阶段EIKONAL_DW。准备协调器亲和性30/31;模型等待无子进程,未开始拟合。进度0终态/1已派发/629请求、首件成本检查待完成、无stop_reason;结算60秒设置预留,活动件另预留1802秒,不把预留或旧阶段进度当作实际终态CPU。16个原固定文件哈希再次全部一致。
启动证据:根目录`manual-science-launch-started.json`、`manual-science-actual-start-audit.json`、`manual-science-worker-audit.json`;protocol内`queue-started.json`、`model-chain-started.json`和两个service的launch JSON。后续以原服务和实时回执监测,不因聊天空闲启动重复队列。首件成本门通过才按原代码自动进入8worker准备;629全终态及费用完整才进入原模型链。