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/provision.py

@@ -372,6 +372,48 @@ def provision_llm(cfg: Config, host: str, log: Log) -> None:
             )
     elif runtime == "llamacpp":
         _stop_units("gpu-rent-ollama")
+        from gpu_rent.llm_runtime import (
+            gguf_filename_from_url,
+            parse_llamacpp_models,
+        )
+        import os
+
+        entries = parse_llamacpp_models(cfg.llamacpp_models_manifest)
+        defaults = [e for e in entries if e.default]
+        if defaults:
+            log(
+                "llama.cpp preferred: "
+                f"{defaults[0].filename or gguf_filename_from_url(defaults[0].url)}"
+            )
+        if entries:
+            jobs = [
+                {
+                    "url": e.url,
+                    "filename": e.filename or gguf_filename_from_url(e.url),
+                }
+                for e in entries
+            ]
+            put_text(
+                cfg, host, "/tmp/gpu-rent-llamacpp-models.json", json.dumps(jobs, indent=2)
+            )
+            hf = (
+                os.environ.get("HF_TOKEN")
+                or os.environ.get("HUGGING_FACE_HUB_TOKEN")
+                or ""
+            ).strip()
+            if hf:
+                put_text(cfg, host, "/tmp/gpu-rent-hf.token", hf + "\n", mode=0o600)
+            log(f"llama.cpp: скачиваю {len(jobs)} GGUF из манифеста")
+            run_python(
+                cfg,
+                host,
+                _pkg_text("llamacpp_fetch.py"),
+                remote_path="/tmp/gpu-rent-llamacpp_fetch.py",
+                timeout=7200,
+                log=log,
+            )
+        else:
+            log("llamacpp-models.yaml пуст — GGUF skip (положи вручную)")
         log("LLM: ставим/запускаем llama.cpp server")
         run_script_sudo(
             cfg,