KoboldCPP container running in a loop

[mekler22]

This is the error that keeps repeating:

2025-03-10T23:45:58.948044936Z koboldcpp.py: error: argument --blasthreads: invalid int value: ''
2025-03-10T23:46:06.999756227Z usage: koboldcpp.py [-h] [--model [filename]] [--port [portnumber]]
2025-03-10T23:46:07.000020198Z [--host [ipaddr]] [--launch] [--config [filename]]
2025-03-10T23:46:07.000049164Z [--threads [threads]] [--usecublas [[lowvram|normal]
2025-03-10T23:46:07.000067059Z [main GPU ID] [mmq] [rowsplit] ...]]
2025-03-10T23:46:07.000083728Z [--usevulkan [[Device ID] ...]]
2025-03-10T23:46:07.000097753Z [--useclblast {0,1,2,3,4,5,6,7,8} {0,1,2,3,4,5,6,7,8}]
2025-03-10T23:46:07.000116772Z [--usecpu]
2025-03-10T23:46:07.000129750Z [--contextsize [256,512,1024,2048,3072,4096,6144,8192,12288,16384,24576,32768,49152,65536,98304,131072]]
2025-03-10T23:46:07.000150761Z [--gpulayers [[GPU layers]]] [--tensor_split [Ratios]
2025-03-10T23:46:07.000166268Z [[Ratios] ...]] [--ropeconfig [rope-freq-scale]
2025-03-10T23:46:07.000188685Z [[rope-freq-base] ...]]
2025-03-10T23:46:07.000203384Z [--blasbatchsize {-1,32,64,128,256,512,1024,2048}]
2025-03-10T23:46:07.000220395Z [--blasthreads [threads]] [--lora [lora_filename]
2025-03-10T23:46:07.000237168Z [[lora_base] ...]] [--noshift] [--nofastforward]
2025-03-10T23:46:07.000252768Z [--nommap] [--usemlock] [--noavx2]
2025-03-10T23:46:07.000267373Z [--debugmode [DEBUGMODE]] [--onready [shell command]]
2025-03-10T23:46:07.000282553Z [--benchmark [[filename]]] [--prompt [prompt]]
2025-03-10T23:46:07.000297574Z [--promptlimit [token limit]] [--multiuser [limit]]
2025-03-10T23:46:07.000312440Z [--remotetunnel] [--highpriority] [--foreground]
2025-03-10T23:46:07.000327914Z [--preloadstory PRELOADSTORY] [--quiet] [--ssl [cert_pem]
2025-03-10T23:46:07.000343179Z [[key_pem] ...]] [--nocertify] [--mmproj MMPROJ]
2025-03-10T23:46:07.000357425Z [--password PASSWORD] [--ignoremissing]
2025-03-10T23:46:07.000372567Z [--chatcompletionsadapter CHATCOMPLETIONSADAPTER]
2025-03-10T23:46:07.000385714Z [--flashattention] [--quantkv [quantization level 0/1/2]]
2025-03-10T23:46:07.000401531Z [--forceversion [version]] [--smartcontext]
2025-03-10T23:46:07.000415753Z [--unpack destination] [--nomodel]
2025-03-10T23:46:07.000430735Z [--showgui | --skiplauncher] [--hordemodelname [name]]
2025-03-10T23:46:07.000446893Z [--hordeworkername [name]] [--hordekey [apikey]]
2025-03-10T23:46:07.000461568Z [--hordemaxctx [amount]] [--hordegenlen [amount]]
2025-03-10T23:46:07.000476110Z [--sdmodel [filename]] [--sdthreads [threads]]
2025-03-10T23:46:07.000514338Z [--sdclamped [SDCLAMPED]] [--sdt5xxl [filename]]
2025-03-10T23:46:07.000530722Z [--sdclipl [filename]] [--sdclipg [filename]]
2025-03-10T23:46:07.000546095Z [--sdvae [filename] | --sdvaeauto]
2025-03-10T23:46:07.000560125Z [--sdquant | --sdlora [filename]] [--sdloramult [amount]]
2025-03-10T23:46:07.000575017Z [--whispermodel [filename]]
2025-03-10T23:46:07.000593584Z [model_param] [port_param]

[webbrain-one]

Thanks for the report! I opened a PR (#9) to fix this by making sure --blasthreads is only passed when it has a valid value. That should prevent the empty-string error and stop the container from restarting in a loop.