ML Load timing out

[mitulgovind]

Since V.8.33.8 ML load has been timing out any taking hrs to fine tune.

2026-03-01 01:45:31.324702: Info: record_status Demand
2026-03-01 01:47:06.334901: Info: HAHistory: Pruning history data started.
2026-03-01 01:47:06.420989: Info: HAHistory: Pruning history data completed.
2026-03-01 01:49:17.552712: Error: LoadMLComponent: run() exceeded 3600s timeout:
Traceback of timed-out run():
2026-03-01 01:49:17.553373: Warn: LoadMLComponent: run() returned False
2026-03-01 01:51:18.283336: Info: record_status Demand
2026-03-01 01:55:21.366289: Info: record_status Demand
2026-03-01 02:01:18.069035: Info: record_status Demand
2026-03-01 02:05:21.169122: Info: record_status Demand
2026-03-01 02:11:37.075538: Info: record_status Demand
2026-03-01 02:15:24.003407: Info: record_status Demand
2026-03-01 02:21:14.470814: Info: record_status Demand
2026-03-01 02:25:21.051359: Info: record_status Demand
2026-03-01 02:31:15.459227: Info: record_status Demand
2026-03-01 02:35:21.256095: Info: record_status Demand
2026-03-01 02:41:14.179155: Info: record_status Demand
2026-03-01 02:45:20.781610: Info: record_status Demand
2026-03-01 02:47:37.955084: Info: HAHistory: Pruning history data started.
2026-03-01 02:47:38.006414: Info: HAHistory: Pruning history data completed.
2026-03-01 02:51:32.471221: Info: record_status Demand
2026-03-01 02:55:22.145183: Info: record_status Demand
2026-03-01 03:01:16.825772: Info: record_status Demand
2026-03-01 03:05:20.646229: Info: record_status Demand
2026-03-01 03:11:13.452574: Info: record_status Demand
2026-03-01 03:15:21.121631: Info: record_status Demand
2026-03-01 03:21:14.316900: Info: record_status Demand
2026-03-01 03:25:24.168515: Info: record_status Demand
2026-03-01 03:31:35.879281: Info: record_status Demand
2026-03-01 03:35:20.917909: Info: record_status Demand
2026-03-01 03:41:15.245134: Info: record_status Demand
2026-03-01 03:45:21.089845: Info: record_status Demand
2026-03-01 03:48:06.627198: Info: HAHistory: Pruning history data started.
2026-03-01 03:48:06.669645: Info: HAHistory: Pruning history data completed.
2026-03-01 03:51:22.200944: Info: record_status Demand
2026-03-01 03:55:31.090723: Info: record_status Demand
2026-03-01 04:04:13.901032: Info: record_status Demand
2026-03-01 04:05:40.039869: Info: record_status Demand
2026-03-01 04:14:35.907812: Info: record_status Demand
2026-03-01 04:15:39.830358: Info: record_status Demand
2026-03-01 04:23:27.305022: Info: record_status Demand
2026-03-01 04:25:29.139629: Info: record_status Demand
2026-03-01 04:33:14.518119: Info: record_status Charging target 37%-100%
2026-03-01 04:35:29.885046: Info: record_status Charging target 37%-100%
2026-03-01 04:46:32.772002: Warn: Inverter 0 Trying to write 0.0 to timed_discharge_current didn't complete got 66.0
2026-03-01 04:46:32.896212: Warn: record_status Warn: Inverter 0 write to timed_discharge_current failed
2026-03-01 04:47:38.160802: Warn: TemperatureAPI: Request attempt 1 failed: . Retrying in 1s...
2026-03-01 04:47:40.464196: Info: record_status Hold charging target 41%-39%
2026-03-01 04:49:02.763819: Info: HAHistory: Pruning history data started.
2026-03-01 04:49:02.863884: Info: HAHistory: Pruning history data completed.
2026-03-01 04:54:10.296231: Info: record_status Hold charging target 42%-39%
2026-03-01 04:55:49.195656: Info: record_status Hold charging target 42%-39%
2026-03-01 05:05:51.630268: Info: record_status Demand
2026-03-01 05:06:35.467808: Info: record_status Demand
2026-03-01 05:08:51.195537: Warn: Web Socket closed, will try to reconnect in 5 seconds - error count 1
2026-03-01 05:08:56.197715: Info: Start socket for url http://supervisor/core/api/websocket
2026-03-01 05:08:56.298109: Info: Web Socket active
2026-03-01 05:12:28.274634: Info: record_status Demand
2026-03-01 05:12:59.317964: Info: record_status Demand
2026-03-01 05:18:11.429686: Info: record_status Demand
2026-03-01 05:20:29.686228: Info: record_status Demand
2026-03-01 05:24:10.327382: Error: LoadMLComponent: run() exceeded 3600s timeout:
Traceback of timed-out run():
2026-03-01 05:24:10.327751: Warn: LoadMLComponent: run() returned False
2026-03-01 05:26:10.040594: Info: record_status Demand
2026-03-01 05:34:18.842240: Warn: Inverter 0 Trying to write 0.0 to timed_discharge_current didn't complete got 66.0
2026-03-01 05:34:18.920221: Warn: record_status Warn: Inverter 0 write to timed_discharge_current failed
2026-03-01 05:36:27.769571: Info: record_status Hold charging target 40%-39%
2026-03-01 05:40:20.664643: Info: record_status Hold charging target 40%-39%
2026-03-01 05:46:14.940757: Info: record_status Hold charging target 40%-39%
2026-03-01 05:49:52.331727: Info: HAHistory: Pruning history data started.
2026-03-01 05:49:52.368660: Info: HAHistory: Pruning history data completed.
2026-03-01 05:50:21.137685: Info: record_status Hold charging target 40%-39%
2026-03-01 05:56:15.336296: Info: record_status Hold charging target 40%-39%
2026-03-01 06:01:55.406881: Info: record_status Demand
2026-03-01 06:06:50.792769: Info: record_status Demand
2026-03-01 06:10:24.561656: Info: record_status Demand
2026-03-01 06:16:16.817909: Info: record_status Demand
2026-03-01 06:20:24.562687: Info: record_status Demand
2026-03-01 06:26:14.069470: Info: record_status Demand
2026-03-01 06:31:25.893632: Info: record_status Demand
2026-03-01 06:36:17.787394: Info: record_status Demand
2026-03-01 06:40:24.092761: Info: record_status Demand
2026-03-01 06:46:16.430639: Info: record_status Demand
2026-03-01 06:50:24.201010: Info: record_status Demand

[springfall2008]

Is it actually running, I don't see any mention of the fine tuning in the logfile? Can you attached the full log?

[mitulgovind]

predbat.log.txt

[mitulgovind]

it is running

[mitulgovind]

[mitulgovind]

[mitulgovind]

[mitulgovind]

[mitulgovind]

Everything looks fine but the erors in ML components

[mitulgovind]

. when it started its at 1/5: window.

. after an hour still at 3/5: window.

I am using Homeassistant Green but, it is not that slow

[gcoan]

it looks like it is still processing the ML training

do you have your HA recorder set to retain a large amount of history data in HA? I see 70 days in the log

The more history the longer the training will take unfortunately

[mitulgovind]

cool what is the normal people set? 30 days?