Update LLM support for llama.cpp and enhance configuration management

- Added support for `llamacpp-models.yaml` in `.gitignore` and implemented logic to copy it in `gpu-rent.ps1` and `gpu-rent.sh`.
- Enhanced CLI to prompt for llama.cpp model presets during setup and execution, improving user experience.
- Updated configuration handling to include `llamacpp_models_manifest` and related functions for managing llama.cpp models.
- Improved documentation in `cli.md` and `llm.md` to reflect changes in llama.cpp integration and model management.
- Refactored provisioning logic to handle llama.cpp model downloads and configurations effectively.

src/gpu_rent/ux.py

@@ -3,12 +3,15 @@
 from __future__ import annotations
 
 from collections.abc import Callable
+from dataclasses import dataclass
 from typing import Any
 
 from gpu_rent.config import Config
 from gpu_rent.inventory import FlavorInfo, looks_like_gpu, rank_flavors, resolve_flavor
 
 Log = Callable[[str], None]
+Ask = Callable[[str, str], str]
+Confirm = Callable[[str], bool]
 
 
 def list_ranked_flavors(flavors: list[Any], cfg: Config) -> list[FlavorInfo]:
@@ -61,6 +64,93 @@ def print_up_preview(
         log(f"! {line}")
 
 
+@dataclass(frozen=True)
+class ServerPlan:
+    flavor: FlavorInfo
+    data_gb: int
+    spot: bool
+
+
+def prompt_server_plan(
+    cfg: Config,
+    flavors: list[Any],
+    *,
+    picked: FlavorInfo,
+    spot: bool,
+    ask: Ask,
+    confirm: Confirm | None = None,
+) -> ServerPlan:
+    """Interactive Selectel VM knobs before create (flavor / disk / spot)."""
+    ranked = list_ranked_flavors(flavors, cfg)
+    if not ranked:
+        return ServerPlan(flavor=picked, data_gb=cfg.data_volume_size_gb, spot=spot)
+
+    default_idx = 1
+    for i, info in enumerate(ranked, 1):
+        if info.id == picked.id:
+            default_idx = i
+            break
+
+    raw_idx = ask(
+        f"Flavor Selectel [1-{len(ranked)}] (Enter = рекомендация)",
+        str(default_idx),
+    ).strip()
+    try:
+        idx = int(raw_idx)
+    except ValueError as exc:
+        raise ValueError(f"номер flavor: жду 1…{len(ranked)}, получили {raw_idx!r}") from exc
+    if idx < 1 or idx > len(ranked):
+        raise ValueError(f"номер flavor: жду 1…{len(ranked)}, получили {idx}")
+    chosen = ranked[idx - 1]
+
+    raw_gb = ask(
+        "Data disk GB (тариф 24/7; рост только вверх)",
+        str(cfg.data_volume_size_gb),
+    ).strip()
+    try:
+        data_gb = int(raw_gb)
+    except ValueError as exc:
+        raise ValueError(f"Data disk GB: жду число, получили {raw_gb!r}") from exc
+    if data_gb < 20:
+        raise ValueError("Data disk GB: минимум 20")
+
+    spot_default = "Y" if spot else "n"
+    raw_spot = (
+        ask(
+            "Preemptible GPU (дешевле, могут усыпить ~24ч)? [Y/n]",
+            spot_default,
+        )
+        .strip()
+        .lower()
+    )
+    if raw_spot in {"", "y", "yes", "1", "true", "on"}:
+        use_spot = True
+    elif raw_spot in {"n", "no", "0", "false", "off"}:
+        use_spot = False
+    else:
+        raise ValueError(f"preemptible: жду Y/n, получили {raw_spot!r}")
+
+    changed = (
+        chosen.id != picked.id
+        or data_gb != cfg.data_volume_size_gb
+        or use_spot != spot
+    )
+    if confirm and changed and confirm("Запомнить flavor/disk/spot в gpu-rent.vars?"):
+        from gpu_rent.paths import vars_path
+        from gpu_rent.varsfile import upsert_vars
+
+        upsert_vars(
+            vars_path(),
+            {
+                "DEFAULT_FLAVOR_ID": chosen.id,
+                "DATA_VOLUME_SIZE_GB": str(data_gb),
+                "DEFAULT_SPOT": "true" if use_spot else "false",
+            },
+        )
+
+    return ServerPlan(flavor=chosen, data_gb=data_gb, spot=use_spot)
+
+
 def resolve_and_preview(
     cfg: Config,
     flavors: list[Any],