Enhance LLM and SwarmUI integration with improved configuration options
- Updated `env.example` and `gpu-rent.vars.example` to include new variables for LLM runtime and SwarmUI options. - Refactored CLI commands to support interactive selection of LLM runtime and workload type (SwarmUI, LLM, or both). - Improved access link generation to handle cases where SwarmUI is disabled, providing clearer user feedback. - Enhanced provisioning logic to conditionally bootstrap SwarmUI based on user configuration, allowing for LLM-only setups. - Updated documentation across multiple files to reflect changes in LLM integration, CLI usage, and configuration management.
This commit is contained in:
Leonid Pershin
+95
-21
@@ -303,10 +303,10 @@ def status() -> None:
|
||||
def open(
|
||||
llm: bool = typer.Option(False, "--llm", help="Открыть LLM API URL вместо SwarmUI"),
|
||||
) -> None:
|
||||
"""Открыть браузер на SwarmUI :17801 (или --llm на Ollama/llama.cpp)."""
|
||||
"""Открыть браузер на SwarmUI :17801 (или --llm / llm-only на Ollama/llama.cpp)."""
|
||||
cfg = load_config(require_auth=False)
|
||||
if llm:
|
||||
from gpu_rent.llm_runtime import normalize_runtime
|
||||
use_llm = llm or not bool(getattr(cfg, "enable_swarmui", True))
|
||||
if use_llm:
|
||||
from gpu_rent.access_card import resolve_llm_runtime
|
||||
|
||||
runtime = resolve_llm_runtime(cfg)
|
||||
@@ -398,18 +398,25 @@ def up(
|
||||
),
|
||||
ollama: bool = typer.Option(False, "--ollama", help="То же что --llm ollama"),
|
||||
llamacpp: bool = typer.Option(False, "--llamacpp", help="То же что --llm llamacpp"),
|
||||
no_swarm: bool = typer.Option(
|
||||
False,
|
||||
"--no-swarm",
|
||||
"--llm-only",
|
||||
help="Только LLM на GPU, без установки SwarmUI",
|
||||
),
|
||||
) -> None:
|
||||
"""Create/unshelve GPU, bootstrap SwarmUI, по умолчанию туннель на :17801."""
|
||||
"""Create/unshelve GPU; SwarmUI и/или LLM; по умолчанию туннель."""
|
||||
try:
|
||||
from dataclasses import replace
|
||||
|
||||
from gpu_rent.llm_runtime import (
|
||||
append_vars_llm_runtime,
|
||||
decide_runtime,
|
||||
ensure_ollama_manifest_from_example,
|
||||
write_ollama_models_preset,
|
||||
)
|
||||
from gpu_rent.paths import vars_path
|
||||
from gpu_rent.prompts import MenuItem, prompt_menu
|
||||
from gpu_rent.varsfile import upsert_vars
|
||||
|
||||
checks = run_doctor()
|
||||
code = _print_checks(checks, quiet=not verbose)
|
||||
@@ -427,33 +434,93 @@ def up(
|
||||
except ValueError as exc:
|
||||
raise GpuRentError(str(exc)) from exc
|
||||
|
||||
enable_swarm = False if no_swarm else cfg.enable_swarmui
|
||||
asked_model_preset = False
|
||||
if not yes and runtime == "none" and not llm and not ollama and not llamacpp:
|
||||
llm_flags = bool(llm or ollama or llamacpp or no_swarm)
|
||||
|
||||
if not yes and not llm_flags:
|
||||
from gpu_rent.llm_runtime import (
|
||||
llamacpp_preset_menu,
|
||||
llm_runtime_menu,
|
||||
ollama_preset_menu,
|
||||
workload_menu,
|
||||
)
|
||||
from gpu_rent.prompts import prompt_menu
|
||||
|
||||
def _ask(msg: str, default: str = "") -> str:
|
||||
return typer.prompt(msg, default=default)
|
||||
|
||||
default_stack = (
|
||||
"llm"
|
||||
if not cfg.enable_swarmui
|
||||
else ("both" if runtime != "none" else "swarm")
|
||||
)
|
||||
try:
|
||||
choice = prompt_menu(
|
||||
"LLM рядом со SwarmUI",
|
||||
llm_runtime_menu(),
|
||||
default="none",
|
||||
stack = prompt_menu(
|
||||
"Что поднять на GPU",
|
||||
workload_menu(),
|
||||
default=default_stack,
|
||||
ask=_ask,
|
||||
show=log,
|
||||
)
|
||||
runtime = decide_runtime(
|
||||
flag=choice, ollama_flag=False, llamacpp_flag=False, from_config="none"
|
||||
)
|
||||
except ValueError as exc:
|
||||
raise GpuRentError(str(exc)) from exc
|
||||
if runtime != "none" and typer.confirm("Запомнить LLM_RUNTIME в gpu-rent.vars?", default=True):
|
||||
append_vars_llm_runtime(vars_path(), runtime)
|
||||
|
||||
if stack == "swarm":
|
||||
enable_swarm = True
|
||||
runtime = "none"
|
||||
elif stack == "both":
|
||||
enable_swarm = True
|
||||
if runtime == "none":
|
||||
try:
|
||||
choice = prompt_menu(
|
||||
"LLM runtime",
|
||||
[
|
||||
MenuItem("ollama", "Ollama (+ pull моделей)"),
|
||||
MenuItem("llamacpp", "llama.cpp server (+ GGUF)"),
|
||||
],
|
||||
default="ollama",
|
||||
ask=_ask,
|
||||
show=log,
|
||||
)
|
||||
runtime = decide_runtime(
|
||||
flag=choice,
|
||||
ollama_flag=False,
|
||||
llamacpp_flag=False,
|
||||
from_config="none",
|
||||
)
|
||||
except ValueError as exc:
|
||||
raise GpuRentError(str(exc)) from exc
|
||||
else:
|
||||
enable_swarm = False
|
||||
if runtime == "none":
|
||||
try:
|
||||
choice = prompt_menu(
|
||||
"LLM runtime",
|
||||
[
|
||||
MenuItem("ollama", "Ollama (+ pull моделей)"),
|
||||
MenuItem("llamacpp", "llama.cpp server (+ GGUF)"),
|
||||
],
|
||||
default="llamacpp",
|
||||
ask=_ask,
|
||||
show=log,
|
||||
)
|
||||
runtime = decide_runtime(
|
||||
flag=choice,
|
||||
ollama_flag=False,
|
||||
llamacpp_flag=False,
|
||||
from_config="none",
|
||||
)
|
||||
except ValueError as exc:
|
||||
raise GpuRentError(str(exc)) from exc
|
||||
|
||||
if typer.confirm("Запомнить стек в gpu-rent.vars?", default=True):
|
||||
upsert_vars(
|
||||
vars_path(),
|
||||
{
|
||||
"ENABLE_SWARMUI": "true" if enable_swarm else "false",
|
||||
"LLM_RUNTIME": runtime,
|
||||
},
|
||||
)
|
||||
|
||||
if runtime == "ollama":
|
||||
ensure_ollama_manifest_from_example()
|
||||
try:
|
||||
@@ -498,7 +565,6 @@ def up(
|
||||
ollama_preset_menu,
|
||||
write_llamacpp_models_preset,
|
||||
)
|
||||
from gpu_rent.prompts import prompt_menu
|
||||
|
||||
def _ask2(msg: str, default: str = "") -> str:
|
||||
return typer.prompt(msg, default=default)
|
||||
@@ -529,9 +595,17 @@ def up(
|
||||
except ValueError as exc:
|
||||
raise GpuRentError(str(exc)) from exc
|
||||
|
||||
cfg = replace(cfg, llm_runtime=runtime)
|
||||
if runtime != "none":
|
||||
ok(f"LLM runtime: {runtime}")
|
||||
if not enable_swarm and runtime == "none":
|
||||
raise GpuRentError(
|
||||
"llm-only требует --ollama / --llamacpp / --llm … "
|
||||
"(или убери --no-swarm / ENABLE_SWARMUI=true)"
|
||||
)
|
||||
|
||||
cfg = replace(cfg, llm_runtime=runtime, enable_swarmui=enable_swarm)
|
||||
if enable_swarm:
|
||||
ok("стек: SwarmUI" + (f" + {runtime}" if runtime != "none" else ""))
|
||||
else:
|
||||
ok(f"стек: llm-only ({runtime})")
|
||||
|
||||
def confirm(msg: str) -> bool:
|
||||
return typer.confirm(msg)
|
||||
|
||||
Reference in New Issue
Block a user