CI / test (push) Failing after 30m3s
Apertus 70B with a 65536-token pool, and then 8B with 131072, on a 24 GB card loaded their weights and failed to make their context on every start, the 8B taking the card from the 70B each time only to fail. ctl model serve now shrinks a pool the card cannot hold beside the weights (KV cache from the GGUF's shape, 90% of the card, a gigabyte for the engine), saying so, and refuses weights that do not fit at all; the portal's serve form refuses such a pool and says the largest that fits. The mirror page removed the 70B's weights while the catalog still had the model, whose replicas went on loading the node's copy with nothing said. Removing weights a catalog model loads is refused, naming the model. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
86 lines
5.2 KiB
JSON
86 lines
5.2 KiB
JSON
{
|
||
"serve.refusal": "Cannot be served as it stands: %s.",
|
||
"serve.serving": "Serving <code>%s</code> with <strong>%s</strong>.",
|
||
"serve.whole_version": "The whole version, %d files: the node fetches them into one directory and the engine is given the directory.",
|
||
"serve.name": "Name",
|
||
"serve.aliases": "Aliases",
|
||
"serve.aliases_hint": "comma separated",
|
||
"serve.aliases_example": "coder-7b",
|
||
"serve.replicas": "Replicas",
|
||
"serve.runs": "Runs",
|
||
"serve.mode.ondemand": "on demand: loads when asked, stops after 15 min unused",
|
||
"serve.description": "What it is for",
|
||
"serve.description_hint": "shown to clients in /v1/models",
|
||
"serve.description_example": "general chat, strong Swedish; 14B, slower than the 8B",
|
||
"serve.context_length": "Context a request may use",
|
||
"serve.max_output": "Output limit",
|
||
"serve.optional": "optional",
|
||
"serve.ctx_pool": "Engine context pool",
|
||
"serve.ctx_pool_hint": "divided across slots",
|
||
"serve.placement": "Runs on",
|
||
"serve.place.gpu": "GPU: %s, %d of %d free",
|
||
"serve.place.cpu_slower": "CPU only: no GPU, many times slower",
|
||
"serve.place.cpu_only": "CPU: this fleet has no GPU",
|
||
"serve.gpus": "GPUs per replica",
|
||
"serve.gpus_hint": "on one node",
|
||
"serve.parallel": "Parallel requests",
|
||
"serve.parallel_hint": "per replica",
|
||
"serve.cpu_ok": "Run on the CPU anyway: %d GPUs are free and would sit unused",
|
||
"serve.memory": "Needs RAM (GB)",
|
||
"serve.memory_hint": "for the context pool beside it",
|
||
"serve.cpu_cores": "CPU cores",
|
||
"serve.cpu_cores_hint": "reserved; the engine's threads",
|
||
"serve.selector": "Only nodes labelled",
|
||
"serve.selector_hint": "key / value, optional",
|
||
"serve.selector_key_example": "role",
|
||
"serve.selector_value_example": "cpu-large",
|
||
"serve.add": "Add to the catalog",
|
||
|
||
"serve.on_cpu": "%s on the CPU",
|
||
"serve.on_gpu_one": "%s on 1 GPU %s",
|
||
"serve.on_gpus": "%s on %d %s GPUs of one node",
|
||
"serve.host_memory": "%s GB of host memory",
|
||
"serve.cores.one": "%d core",
|
||
"serve.cores.other": "%d cores",
|
||
"serve.pool_parallel": "context pool %s over %s parallel requests (%s each)",
|
||
"serve.pool": "context pool %s",
|
||
"serve.will_run_as": "Will run as: %s.",
|
||
"serve.free_gpus_warning": "%s GPUs are free; on the CPU this runs many times slower.",
|
||
|
||
"serve.pool_note": "The context pool is what the engine is given; llama.cpp divides it across its parallel slots, so the number one request may use is the smaller one above.",
|
||
"serve.memory_note": "Memory: %s. Change the context pool and this changes with it, by the KV cache's share — an estimate, and editable. It is filled in rather than left blank because a CPU profile has no card to be short of: with no figure, every ready node is a candidate and the first one picked is whichever the scheduler reached, not whichever can hold the model. A node label does the same job by name if you would rather pin it.",
|
||
"serve.basis.estimate": "weights plus half again plus two",
|
||
"serve.basis.shape": "weights %s GB + KV cache %s GB for %s tokens (%d layers × %d KV heads) + %s GB for the engine",
|
||
|
||
"serve.refuse.no_engine": "nothing here reads %s files",
|
||
"serve.refuse.llamacpp": "llama.cpp on the nodes cannot run this: %s",
|
||
"serve.refuse.declared": "a model called %s is declared in configuration; edit it there rather than adding a second",
|
||
"serve.warn.not_installed": "no node reports having %s installed, and %s weights need it",
|
||
"serve.warn.no_gpu": "This fleet has no GPU either; a GGUF build of the same model runs on the CPU nodes",
|
||
"serve.warn.too_big": "the largest node has %d GB and this needs about %d",
|
||
"serve.warn.and": "%s, and %s",
|
||
|
||
"serve.err.name_source": "a model needs a name and a source",
|
||
"serve.err.whole_number": "%s must be a whole number",
|
||
"serve.what.context_length": "the context length",
|
||
"serve.what.max_output": "the output limit",
|
||
"serve.what.replicas": "the replica count",
|
||
"serve.what.gpus": "the GPUs per replica",
|
||
"serve.what.ctx_pool": "the engine's context pool",
|
||
"serve.what.parallel": "the parallel requests",
|
||
"serve.what.memory": "the memory requirement",
|
||
"serve.what.cores": "the cores",
|
||
"serve.err.engine": "which engine should run it?",
|
||
"serve.err.needs_gpu": "%s runs on GPUs; choose a GPU class to place it on",
|
||
"serve.err.placement": "%q is not a placement",
|
||
"serve.err.weights_too_big": "Its weights (%s GB) do not fit a %s card with room to run: choose a smaller quantisation, or more cards.",
|
||
"serve.err.pool_too_big": "A context pool of %s tokens does not fit a %s card beside these weights; at most %s would.",
|
||
"serve.err.cpu_with_free_gpus.one": "this fleet has %d GPU free and %s was set to run on the CPU; choose a GPU placement, or tick \"run on the CPU anyway\"",
|
||
"serve.err.cpu_with_free_gpus.other": "this fleet has %d GPUs free and %s was set to run on the CPU; choose a GPU placement, or tick \"run on the CPU anyway\"",
|
||
"serve.msg.added": "%s added to the catalog, %s: %s; it is placed on the next pass",
|
||
"serve.replicas_n.one": "%d replica",
|
||
"serve.replicas_n.other": "%d replicas",
|
||
"serve.msg.judge": "It is the guardrails' judge (role: guard): policies naming %s ask it; clients do not see it.",
|
||
"serve.msg.judge_renamed": "Its weights are a judge's, so it is the guardrails' judge (role: guard), named %[2]s as the policies ask rather than %[1]s."
|
||
}
|