{
 "case": "ctx-8192",
 "machine": "macos-15-arm64",
 "run": "ci-1",
 "args": [
  "--model",
  "models/qwen2.5-0.5b-instruct-q4_k_m.gguf",
  "--contextsize",
  "8192"
 ],
 "started": "2026-11-13T12:54:23+0000",
 "start": {
  "ready": true,
  "seconds": 4.13,
  "model": {
   "result": "koboldcpp/qwen2.5-0.5b-instruct-q4_k_m"
  }
 },
 "gen": [
  {
   "status": 200,
   "wall_s": 15.652,
   "text_head": " saw a small boat bobbing on the water. She had never seen a boat like this before. She had never seen a boat like this before. She had never seen a boat like t",
   "prompt_tokens": 38,
   "gen_tokens": 128,
   "pp_tps": 64.74,
   "tg_tps": 8.51
  }
 ],
 "peak_rss_mb": 783.3,
 "log_lines": [
  "Unable to determine GPU Memory",
  "Unable to determine available RAM",
  "Initializing dynamic library: koboldcpp_default.so",
  "Namespace(usecuda=None, usevulkan=None, usecpu=False, model=['models/qwen2.5-0.5b-instruct-q4_k_m.gguf'], model_param='models/qwen2.5-0.5b-instruct-q4_k_m.gguf', config=None, contextsize=8192, gpulayers=200, host='', launch=False, tensor_split=None, threads=1, port=49438, port_param=49438, version=False, agent=False, agentload='', analyze='', autofit=False, autofitpadding=1024, batchsize=512, ubatchsize=-1, benchmark=None, blasthreads=0, chatcompletionsadapter='AutoGuess', cli=False, debugmode=0, defaultgenamt=2048, device='', downloaddir='', draftamount=4, draftgpulayers=999, draftgpusplit=None, draftmodel='', enableguidance=False, exportconfig='', exporttemplate='', failsafe=False, foreground=False, gendefaults='', gendefaultsoverwrite=False, genlimit=0, highpriority=False, ignoremissing=False, jinja=False, jinja_kwargs='', jinjatemplate='', jinjathink='default', lora=None, loramult=1.0, lowvram=False, maingpu=-1, maxrequestsize=32, mcpfile='', mmproj='', mmprojcpu=False, lazy_mode='auto', moe_cache_mib=0, moecpu=0, ffncpu=0, moeexperts=-1, multiplayer=False, nocertify=False, nofastforward=False, noflashattention=False, noavx2=False, nobostoken=<redacted> nommq=False, nomodel=False, noshift=False, noswa=False, onready='', allow_config_onready=False, overridekv='', overridenativecontext=0, overridetensors='', parallelrequests=1, paralleltotalcontext=0, password=<redacted> preloadstory='', prompt='', quantkv='f16', queuelimit=10, quiet=False, ratelimit=0, reasoningeffort='default', remotetunnel=False, ropeconfig=[0.0, 10000.0], savedatafile='', singleinstance=False, smartcache=0, smartcontext=False, splitmode='layer', ssl=None, swapadding=0, unpack='', usemtp=False, usemlock=False, usemmap=False, usedirectio=False, visionmaxres=1024, visionmaxtokens=-1, visionmintokens=-1, websearch=False, showgui=False, skiplauncher=True, hordemodelname='', hordeworkername='', hordekey='', hordemaxctx=0, hordegenlen=0, sdaudiovae='', sdclamped=0, sdclampedsoft=0, sdclip1='', sdclip2='', sdclipdevice=-2, sdconvdirect='off', sdflashattention=False, sdlora=[], sdloramult=[1.0], sdmaingpu=None, sdmodel='', sdoffloadcpu=False, sdphotomaker='', sdquant=0, sdllm='', sdthreads=0, sdtiledvae=512, sdupscaler='', sdvae='', sdvaeauto=False, sdvaedevice=-1, sdvramlimit=0, whispermodel='', ttsmodel='', ttswavtokenizer='', ttsgpu=False, ttsmaxlen=4096, ttsthreads=0, ttsdir='', musicllm='', musicembeddings='', musicdiffusion='', musicvae='', musiclowvram=False, embeddingsmodel='', embeddingsmaxctx=4096, embeddingsgpu=False, rpcmode='disabled', rpcport=5551, rpchost='0.0.0.0', rpcdevice='', rpccache='', rpctargets='', admin=False, adminpassword=<redacted> admindir='', adminunloadtimeout=0, routermode=False, reqtimeout=600, autoswapmode=False, autoswapthreshold=256, baseconfig='', hordeconfig=None, sdconfig=None, noblas=False, nommap=False, nopipelineparallel=False, pipelineparallel=False, sdnotile=False, sdt5xxl=None, sdvaecpu=False, sdclipgpu=False, forceversion=False, sdgendefaults=False, flashattention=False, useswa=False, multiuser=None, testmemory=False, run_bundled_agent=False, agent_base_url=None, agent_api_key=<redacted> proxy_port=None)",
  "print_info: file size   = 462.96 MiB (6.16 BPW)",
  "print_info: n_ctx_train           = 32768",
  "print_info: n_ctx_orig_yarn       = 32768",
  "done_getting_tensors: tensor 'token_embd.weight' (q5_0) (and 169 others) moved from CPU_REPACK, using CPU instead",
  "load_tensors:          CPU model buffer size =   254.39 MiB",
  "load_tensors:   CPU_REPACK model buffer size =   208.30 MiB",
  "llama_context: n_ctx                 = 8448",
  "llama_context: n_ctx_seq             = 8448",
  "llama_context: flash_attn            = enabled",
  "llama_context: n_ctx_seq (8448) < n_ctx_train (32768) -- the full capacity of the model will not be utilized",
  "llama_kv_cache:       MTL0 KV buffer size =    99.00 MiB",
  "sched_reserve:       MTL0 compute buffer size =    10.50 MiB",
  "sched_reserve:        CPU compute buffer size =   300.25 MiB",
  "Load Text Model OK: True",
  "Starting Kobold API on port 49438 at http://localhost:49438/api/",
  "Please connect to custom endpoint at http://localhost:49438",
  "Threadpool set to 1 threads and 1 blasthreads..."
 ],
 "finished": "2026-11-13T12:54:44+0000"
}