Merge pull request #3815 from vladmandic/dev

Dev

CHANGELOG.md

@@ -1,10 +1,11 @@
 # Change Log for SD.Next
 
-## Update for 2025-03-11
+## Update for 2025-03-12
 
 - fix installer not starting when older version of rich is installed  
 - fix circular imports when debug flags are enabled  
 - fix cuda errors with directml  
+- fix memory stats not displaying the ram usage  
 - **ipex**
   - add xpu to profiler  
   - fix untyped_storage, torch.eye and torch.cuda.device ops  

modules/call_queue.py

@@ -77,15 +77,18 @@ def wrap_gradio_call(func, extra_outputs=None, add_stats=False, name=None):
         gpu = ''
         cpu = ''
         if not shared.mem_mon.disabled:
-            vram = {k: -(v//-(1024*1024)) for k, v in shared.mem_mon.read().items()}
+            mem_mon_read = shared.mem_mon.read()
+            ooms = mem_mon_read.pop("oom")
+            retries = mem_mon_read.pop("retries")
+            vram = {k: v//1048576 for k, v in mem_mon_read.items()}
             peak = max(vram['active_peak'], vram['reserved_peak'], vram['used'])
             used = round(100.0 * peak / vram['total']) if vram['total'] > 0 else 0
             if used > 0:
                 gpu += f"| GPU {peak} MB {used}%"
-                gpu += f" | retries {vram['retries']} oom {vram['oom']}" if vram.get('retries', 0) > 0 or vram.get('oom', 0) > 0 else ''
-            ram = shared.ram_stats()
-            if ram['used'] > 0:
-                cpu += f"| RAM {ram['used']} GB {round(100.0 * ram['used'] / ram['total'])}%"
+                gpu += f" | retries {retries} oom {ooms}" if retries > 0 or ooms > 0 else ''
+        ram = shared.ram_stats()
+        if ram['used'] > 0:
+            cpu += f"| RAM {ram['used']} GB {round(100.0 * ram['used'] / ram['total'])}%"
         if isinstance(res, list):
             res[-1] += f"<div class='performance'><p>Time: {elapsed_text} | {summary} {gpu} {cpu}</p></div>"
         return tuple(res)

modules/memmon.py

@@ -44,10 +44,10 @@ class MemUsageMonitor():
                 self.data["free"], self.data["total"] = torch.cuda.mem_get_info(self.device.index if self.device.index is not None else torch.cuda.current_device())
                 self.data["used"] = self.data["total"] - self.data["free"]
                 torch_stats = torch.cuda.memory_stats(self.device)
-                self.data["active"] = torch_stats.get("active.all.current", torch_stats["active_bytes.all.current"])
-                self.data["active_peak"] = torch_stats["active_bytes.all.peak"]
-                self.data["reserved"] = torch_stats["reserved_bytes.all.current"]
-                self.data["reserved_peak"] = torch_stats["reserved_bytes.all.peak"]
+                self.data["active"] = torch_stats.get("active.all.current", torch_stats.get("active_bytes.all.current", -1))
+                self.data["active_peak"] = torch_stats.get("active_bytes.all.peak", -1)
+                self.data["reserved"] = torch_stats.get("reserved_bytes.all.current", -1)
+                self.data["reserved_peak"] = torch_stats.get("reserved_bytes.all.peak", -1)
                 self.data['retries'] = torch_stats.get("num_alloc_retries", -1)
                 self.data['oom'] = torch_stats.get("num_ooms", -1)
             except Exception:

modules/memstats.py

@@ -77,7 +77,7 @@ def ram_stats():
         process = psutil.Process(os.getpid())
         res = process.memory_info()
         ram_total = 100 * res.rss / process.memory_percent()
-        ram_total = min(ram_total, docker_limit(), runpod_limit())
+        ram_total = min(ram_total, get_docker_limit(), get_runpod_limit())
         ram = { 'used': gb(res.rss), 'total': gb(ram_total) }
         return ram
     except Exception: