5-th DL fix

packages/node/meshnet_node/startup.py

@@ -332,6 +332,7 @@ def run_startup(
     vram_mb_override: int | None = None,
     max_loaded_shards: int = 1,
     debug: bool = False,
+    tracker_source_disabled: bool = False,
 ) -> StubNodeServer | TorchNodeServer:
     """Execute the full startup sequence and return a running node server.
 
@@ -458,7 +459,10 @@ def run_startup(
                     if net_asgn.get("hf_repo") == model_id and net_asgn.get("gap_found"):
                         shard_start = net_asgn["shard_start"]
                         shard_end = net_asgn["shard_end"]
-                        full_sources = _full_model_sources(net_asgn.get("model_sources", []))
+                        full_sources = (
+                            [] if tracker_source_disabled
+                            else _full_model_sources(net_asgn.get("model_sources", []))
+                        )
                         if full_sources:
                             cache_dir = download_shard(
                                 model_id.split("/")[-1],
@@ -588,7 +592,7 @@ def run_startup(
             f"(of {assigned_num_layers})",
             flush=True,
         )
-        full_sources = _full_model_sources(assigned_model_sources)
+        full_sources = [] if tracker_source_disabled else _full_model_sources(assigned_model_sources)
         if full_sources:
             print("Downloading assigned model snapshot...", flush=True)
             cache_dir = download_shard(
@@ -676,7 +680,10 @@ def run_startup(
     assign_qs = urllib.parse.urlencode({
         "model": model,
         "device": device,
-        "vram_mb": vram_mb,
+        # CPU-mode nodes must be sized by RAM: a detected-but-unusable GPU's
+        # VRAM would otherwise cap the shard (e.g. 8 GB VRAM → 3 layers on a
+        # 79 GB box whose Torch has no CUDA).
+        "vram_mb": assignment_vram_mb,
         "ram_mb": ram_mb,
     })
     try:
@@ -690,7 +697,7 @@ def run_startup(
     assigned_model: str = assignment.get("model", model)
     hf_repo: str | None = assignment.get("hf_repo")
     peers: list[dict] = assignment.get("peers", [])
-    model_sources: list[dict] = assignment.get("model_sources", [])
+    model_sources: list[dict] = [] if tracker_source_disabled else assignment.get("model_sources", [])
     print(f"  Shard: layers {shard_start}-{shard_end} of {assigned_model}", flush=True)
 
     # 4. Download shard