Add download progress tracking and byte formatting utilities

- Introduced `fmt_bytes` function for human-readable byte size formatting.
- Added `DownloadProgress` class to track and display download progress with speed and completion percentage.
- Updated `download` functions in `civitai_fetch.py` and `llamacpp_fetch.py` to utilize progress tracking.
- Enhanced `ollama_pull.py` to support streaming progress updates during model pulls.
- Updated tests to validate new formatting and progress tracking functionalities, ensuring accurate output and user feedback.

src/gpu_rent/remote/ollama_pull.py

@@ -1,15 +1,43 @@
 #!/usr/bin/env python3
-"""Pull Ollama models from a JSON list. Stdlib only. Runs on the VM."""
+"""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]:
@@ -32,7 +60,6 @@ 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
-    # ollama list sometimes omits :latest
     if ":" not in wanted and f"{wanted}:latest" in have:
         return True
     if wanted.endswith(":latest") and wanted.rsplit(":", 1)[0] in have:
@@ -40,6 +67,45 @@ def already_have(have: set[str], wanted: str) -> bool:
     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
+                    print(progress_line(label, done, total, speed), 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", file=sys.stderr)
@@ -57,14 +123,15 @@ def main() -> int:
             name = str(name).strip()
             if not name:
                 continue
+            prefix = f"[{i}/{len(models)}]"
             if already_have(have, name):
-                print(f"[{i}/{len(models)}] уже есть {name}")
+                print(f"{prefix} уже есть {name}")
                 continue
-            print(f"[{i}/{len(models)}] ollama pull {name}")
+            print(f"{prefix} ollama pull {name}", flush=True)
             try:
-                subprocess.check_call(["ollama", "pull", name])
+                pull_stream(name, label=f"{prefix} {name}")
                 have.add(name)
-            except subprocess.CalledProcessError as exc:
+            except (urllib.error.URLError, urllib.error.HTTPError, OSError, TimeoutError, RuntimeError) as exc:
                 failed += 1
                 print(f"FAIL pull {name}: {exc}", file=sys.stderr)
     finally:
@@ -76,4 +143,4 @@ def main() -> int:
 
 
 if __name__ == "__main__":
-    sys.exit(main())
+    raise SystemExit(main())