[misc] refactor launch API and tensor constructor (#5666)

* [misc] remove config arg from initialize

* [misc] remove old tensor contrusctor

* [plugin] add npu support for ddp

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [devops] fix doc test ci

* [test] fix test launch

* [doc] update launch doc

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

applications/ColossalMoE/infer.py

@@ -57,7 +57,7 @@ def main():
     args = parse_args()
 
     # Launch ColossalAI
-    colossalai.launch_from_torch(config={}, seed=args.seed)
+    colossalai.launch_from_torch(seed=args.seed)
     coordinator = DistCoordinator()
 
     config = MixtralConfig.from_pretrained(args.model_name)
@@ -96,7 +96,11 @@ def main():
     if coordinator.rank == 0:
         text = ["Hello my name is"]
     else:
-        text = ["What's the largest country in the world?", "How many people live in China?", "帮我续写这首诗：离离原上草"]
+        text = [
+            "What's the largest country in the world?",
+            "How many people live in China?",
+            "帮我续写这首诗：离离原上草",
+        ]
     tokenizer.pad_token = tokenizer.unk_token
     inputs = tokenizer(text, return_tensors="pt", padding=True).to(torch.cuda.current_device())
 

applications/ColossalMoE/tests/test_mixtral_layer.py

@@ -50,7 +50,7 @@ def check_mixtral_moe_layer():
 
 
 def run_dist(rank: int, world_size: int, port: int):
-    colossalai.launch({}, rank, world_size, "localhost", port)
+    colossalai.launch(rank, world_size, "localhost", port)
     check_mixtral_moe_layer()
 
 

applications/ColossalMoE/tests/test_moe_checkpoint.py

@@ -133,7 +133,7 @@ def check_mixtral_moe_layer():
 
 
 def run_dist(rank: int, world_size: int, port: int):
-    colossalai.launch({}, rank, world_size, "localhost", port)
+    colossalai.launch(rank, world_size, "localhost", port)
     check_mixtral_moe_layer()
 
 

applications/ColossalMoE/train.py

@@ -145,7 +145,7 @@ def main():
     args = parse_args()
 
     # Launch ColossalAI
-    colossalai.launch_from_torch(config={}, seed=args.seed)
+    colossalai.launch_from_torch(seed=args.seed)
     coordinator = DistCoordinator()
 
     # Set plugin
@@ -195,9 +195,9 @@ def main():
     lr_scheduler = CosineAnnealingWarmupLR(
         optimizer=optimizer,
         total_steps=args.num_epochs * len(dataloader),
-        warmup_steps=args.warmup_steps
-        if args.warmup_steps is not None
-        else int(args.num_epochs * len(dataloader) * 0.025),
+        warmup_steps=(
+            args.warmup_steps if args.warmup_steps is not None else int(args.num_epochs * len(dataloader) * 0.025)
+        ),
         eta_min=0.1 * args.lr,
     )