Model parallel (#538)

* model-parallel for single process * fix device/device_map * fix handling for device
axolotl-ai-cloud · Sep 13, 2023 · f6060a6 · f6060a6
1 parent a4e1bb6
commit f6060a6
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 2 deletions.
diff --git a/src/axolotl/utils/bench.py b/src/axolotl/utils/bench.py
@@ -28,7 +28,7 @@ def gpu_memory_usage_smi(device=0):
 
 
 def log_gpu_memory_usage(log, msg, device):
-    if not torch.cuda.is_available():
+    if not torch.cuda.is_available() or device == "auto":
         return (0, 0, 0)
 
     usage, cache, misc = gpu_memory_usage_all(device)

diff --git a/src/axolotl/utils/config.py b/src/axolotl/utils/config.py
@@ -25,7 +25,9 @@ def get_device():
             return "cpu"
 
     cfg.device = get_device()
-    if cfg.device_map != "auto":
+    if cfg.world_size == 1:
+        cfg.device_map = "auto"
+    else:
         if cfg.device.startswith("cuda"):
             cfg.device_map = {"": cfg.local_rank}
         else: