Files
gpu-rent/src/gpu_rent/remote/ollama_pull.py
T
Leonid Pershin ccba40a228 Enhance logging and timing in CLI and session operations
- Updated the `_print_checks` function to replace console prints with logging functions for better traceability.
- Introduced timing functionality in the `doctor`, `dry_run`, and `up` functions to log the duration of preflight checks.
- Modified the `wait_ssh` function to accept a logging callback, improving SSH wait feedback.
- Enhanced the `mark` method in `PhaseTimes` to log phase durations, aiding in performance analysis.
- Updated various remote scripts to ensure error messages are printed to stderr for better error handling.
2026-08-21 08:03:04 +03:00

149 lines
5.0 KiB
Python

#!/usr/bin/env python3
"""Pull Ollama models from a JSON list. Stdlib only. Runs on the VM.
Uses POST /api/pull with stream JSON for completed/total + speed lines.
"""
from __future__ import annotations
import json
import subprocess
import sys
import time
import urllib.error
import urllib.request
from pathlib import Path
JOBS = Path("/tmp/gpu-rent-ollama-models.json")
MARKER = Path("/mnt/swarm_data/.gpu-rent-ollama-pulling")
OLLAMA = "http://127.0.0.1:11434"
def fmt_bytes(n: float) -> str:
n = float(n)
for unit, div in (("GB", 1024**3), ("MB", 1024**2), ("KB", 1024), ("B", 1)):
if n >= div or unit == "B":
if unit == "B":
return f"{int(n)}B"
return f"{n / div:.1f}{unit}"
return f"{n:.0f}B"
def progress_line(label: str, done: int, total: int | None, speed: float, *, width: int = 22) -> str:
if total and total > 0:
pct = min(100.0, 100.0 * done / total)
filled = min(width, max(0, int(width * done / total)))
bar = "#" * filled + "-" * (width - filled)
return (
f"{label} [{bar}] {pct:5.1f}% "
f"{fmt_bytes(done)}/{fmt_bytes(total)} {fmt_bytes(speed)}/s"
)
return f"{label} {fmt_bytes(done)} {fmt_bytes(speed)}/s"
def listed() -> set[str]:
"""Exact tags from `ollama list` (NAME column), e.g. qwen2.5:7b."""
try:
out = subprocess.check_output(["ollama", "list"], text=True, stderr=subprocess.DEVNULL)
except (subprocess.CalledProcessError, FileNotFoundError):
return set()
names: set[str] = set()
for i, line in enumerate(out.splitlines()):
if i == 0 and line.lower().startswith("name"):
continue
parts = line.split()
if parts:
names.add(parts[0])
return names
def already_have(have: set[str], wanted: str) -> bool:
"""Exact tag match only — qwen2.5:3b must not satisfy qwen2.5:7b."""
if wanted in have:
return True
if ":" not in wanted and f"{wanted}:latest" in have:
return True
if wanted.endswith(":latest") and wanted.rsplit(":", 1)[0] in have:
return True
return False
def pull_stream(name: str, label: str) -> None:
body = json.dumps({"name": name, "stream": True}).encode("utf-8")
req = urllib.request.Request(
f"{OLLAMA}/api/pull",
data=body,
headers={"Content-Type": "application/json"},
method="POST",
)
t0 = time.monotonic()
last_print = 0.0
with urllib.request.urlopen(req, timeout=7200) as resp:
while True:
raw = resp.readline()
if not raw:
break
try:
ev = json.loads(raw.decode("utf-8"))
except json.JSONDecodeError:
continue
if ev.get("error"):
raise RuntimeError(str(ev["error"]))
status = str(ev.get("status") or "")
total = int(ev.get("total") or 0) or None
done = int(ev.get("completed") or 0)
now = time.monotonic()
if total and done:
if now - last_print >= 1.0 or done >= total:
elapsed = max(now - t0, 0.001)
speed = done / elapsed
line = progress_line(label, done, total, speed)
end = "\n" if done >= total else "\r"
print(line, end=end, flush=True)
last_print = now
elif status and status not in {"success"} and now - last_print >= 2.0:
print(f"{label} {status}", flush=True)
last_print = now
if status == "success":
break
print(f"{label} ok", flush=True)
def main() -> int:
if not JOBS.is_file():
print("no jobs file")
return 1
models = json.loads(JOBS.read_text(encoding="utf-8"))
if not isinstance(models, list) or not models:
print("ollama pull: пустой список — skip")
return 0
have = listed()
MARKER.parent.mkdir(parents=True, exist_ok=True)
MARKER.write_text(f"{int(time.time())}\n", encoding="utf-8")
failed = 0
try:
for i, name in enumerate(models, 1):
name = str(name).strip()
if not name:
continue
prefix = f"[{i}/{len(models)}]"
if already_have(have, name):
print(f"{prefix} уже есть {name}")
continue
print(f"{prefix} ollama pull {name}", flush=True)
try:
pull_stream(name, label=f"{prefix} {name}")
have.add(name)
except (urllib.error.URLError, urllib.error.HTTPError, OSError, TimeoutError, RuntimeError) as exc:
failed += 1
print(f"FAIL pull {name}: {exc}")
finally:
MARKER.unlink(missing_ok=True)
if failed:
return 1
print("ollama pull ok")
return 0
if __name__ == "__main__":
raise SystemExit(main())