Llama distributed example

examples/distributed_inference/rotary_embedding.py

@@ -84,20 +84,20 @@ def parallel_rotary_block(rotary_block, tp_mesh):
         "wk": ColwiseParallel(),
         "wo": RowwiseParallel(output_layouts=Shard(0)),
     }
-    rotary_block.n_parallel = 1  # this is for single GPU, to do remove this hardcode
+    rotary_block.n_parallel = tp_mesh.size()
 
     parallelize_module(rotary_block, tp_mesh, plan)
 
 
 class RotaryAttention(nn.Module):
-    def __init__(self, dim: int, seq_len: int):
+    def __init__(self, dim: int, seq_len: int, n_parallel: int = 1):
         super().__init__()
         self.dim = dim
         self.wq = nn.Linear(dim, dim)
         self.wk = nn.Linear(dim, dim)
         self.wo = nn.Linear(dim, dim)
         self.seq_len = seq_len
-        self.n_parallel = 1
+        self.n_parallel = n_parallel
         self.register_buffer("freqs_cis", self._precompute_freqs_cis(), persistent=True)
         self.init_weights()
 

examples/distributed_inference/tensor_parallel_initialize_dist.py

@@ -17,29 +17,8 @@
 from torch.distributed._tensor.device_mesh import init_device_mesh
 
 
-def find_repo_root(max_depth=10):
-    dir_path = os.path.dirname(os.path.realpath(__file__))
-    for i in range(max_depth):
-        files = os.listdir(dir_path)
-        if "MODULE.bazel" in files:
-            return dir_path
-        else:
-            dir_path = os.path.dirname(dir_path)
-
-    raise RuntimeError("Could not find repo root")
-
-
-def initialize_logger(rank, logger_file_name):
-    logger = logging.getLogger()
-    logger.setLevel(logging.INFO)
-    fh = logging.FileHandler(logger_file_name + f"_{rank}.log", mode="w")
-    fh.setLevel(logging.INFO)
-    logger.addHandler(fh)
-    return logger
-
-
-# This is required for env initialization since we use mpirun
-def initialize_distributed_env(logger_file_name, rank=0, world_size=1, port=29500):
+# this is kept at the application level, when mpirun is used to run the application
+def initialize_distributed_env(rank=0, world_size=1, port=29500):
     local_rank = int(
         os.environ.get("OMPI_COMM_WORLD_LOCAL_RANK", rank % torch.cuda.device_count())
     )
@@ -50,9 +29,6 @@ def initialize_distributed_env(logger_file_name, rank=0, world_size=1, port=2950
     os.environ["WORLD_SIZE"] = str(world_size)
     os.environ["MASTER_ADDR"] = "127.0.0.1"
     os.environ["MASTER_PORT"] = str(port)
-    os.environ["TRTLLM_PLUGINS_PATH"] = (
-        find_repo_root() + "/lib/libnvinfer_plugin_tensorrt_llm.so"
-    )
 
     # Necessary to assign a device to each rank.
     torch.cuda.set_device(local_rank)
@@ -66,13 +42,12 @@ def initialize_distributed_env(logger_file_name, rank=0, world_size=1, port=2950
     device_mesh = init_device_mesh(device_type="cuda", mesh_shape=(world_size,))
     rank = device_mesh.get_rank()
     assert rank == local_rank
-    logger = initialize_logger(rank, logger_file_name)
     device_id = (
         rank % torch.cuda.device_count()
     )  # Ensure each rank gets a unique device
     torch.cuda.set_device(device_id)
 
-    return device_mesh, world_size, rank, logger
+    return device_mesh, world_size, rank
 
 
 def cleanup_distributed_env():

examples/distributed_inference/tensor_parallel_llama3.py

@@ -0,0 +1,72 @@
+# Taken and modified pytorch lightening
+# https://lightning.ai/lightning-ai/studios/tensor-parallelism-supercharging-large-model-training-with-pytorch-lightning
+import logging
+import os
+import time
+
+import torch
+import torch.distributed as dist
+from llama3_model import ModelArgs, ParallelTransformer
+from tensor_parallel_initialize_dist import (
+    cleanup_distributed_env,
+    initialize_distributed_env,
+)
+from torch.distributed._composable.fsdp import MixedPrecisionPolicy
+from torch.distributed._composable.fsdp.fully_shard import fully_shard
+from torch.distributed._tensor import Replicate, Shard
+from torch.distributed.algorithms._checkpoint.checkpoint_wrapper import (
+    checkpoint_wrapper,
+)
+
+if not dist.is_initialized():
+    initialize_distributed_env()
+
+import torch_tensorrt
+from torch_tensorrt.dynamo.distributed.utils import (
+    get_tensor_parallel_device_mesh,
+    initialize_distributed_logger,
+)
+
+device_mesh, _world_size, _rank = get_tensor_parallel_device_mesh()
+logger = initialize_distributed_logger(_rank, "tensor_parallel_llama3")
+
+logger.info(f"Starting PyTorch TP example on rank {_rank}.")
+assert (
+    _world_size % 2 == 0
+), f"TP examples require even number of GPUs, but got {_world_size} gpus"
+
+model_args = ModelArgs(
+    vocab_size=32000,
+    dim=1024,
+    n_layers=4,
+    n_heads=8,
+    rope_theta=500000.0,
+    n_kv_heads=8,
+    device="cuda",
+)
+
+with torch.no_grad():
+    model = ParallelTransformer(model_args, device_mesh)
+    torch.manual_seed(0)
+    inp = torch.randint(32000, (8, 256), device="cuda")
+    python_result = model(inp)
+    torch_tensorrt.runtime.set_multi_device_safe_mode(True)
+    model = torch.compile(
+        model,
+        fullgraph=True,
+        backend="torch_tensorrt",
+        options={
+            "use_python_runtime": True,
+            "use_distributed_mode_trace": True,
+            "debug": True,
+        },
+        dynamic=False,
+    )
+
+    start = time.time()
+    output = model(inp)
+    end = time.time()
+    logger.info(f"Compilation time is {end-start}")
+    assert (python_result - output).std() < 0.01, "Compilation result is not correct."
+
+    cleanup_distributed_env()

examples/distributed_inference/tensor_parallel_rotary_embedding.py

@@ -14,21 +14,30 @@
 import time
 
 import torch
-import torch_tensorrt
-from rotary_embedding import RotaryAttention, parallel_rotary_block
+import torch.distributed as dist
 from tensor_parallel_initialize_dist import (
     cleanup_distributed_env,
     initialize_distributed_env,
 )
 
-device_mesh, _world_size, _rank, logger = initialize_distributed_env(
-    "./tensor_parallel_rotary_embedding"
+if not dist.is_initialized():
+    initialize_distributed_env()
+
+import torch_tensorrt
+from torch_tensorrt.dynamo.distributed.utils import (
+    get_tensor_parallel_device_mesh,
+    initialize_distributed_logger,
 )
 
+device_mesh, _world_size, _rank = get_tensor_parallel_device_mesh()
+logger = initialize_distributed_logger(_rank, "tensor_parallel_rotary_embedding")
+
+from rotary_embedding import RotaryAttention, parallel_rotary_block
 
 """
 This example covers the rotary embedding in Llama3 model and is derived from https://lightning.ai/lightning-ai/studios/tensor-parallelism-supercharging-large-model-training-with-pytorch-lightning
 Command to run with single GPU: mpirun -n 1 --allow-run-as-root python tensor_parallel_rotary_embedding.py
+Command to run with 2 GPUs: mpirun -n 2 --allow-run-as-root python tensor_parallel_rotary_embedding.py
 """
 
 BATCH = 2
@@ -37,7 +46,7 @@
 DIM = 128
 
 with torch.no_grad():
-    model = RotaryAttention(DIM, SEQ_LEN)
+    model = RotaryAttention(DIM, SEQ_LEN, device_mesh.size())
     parallel_rotary_block(model, device_mesh)
     device = torch.device("cuda", device_mesh.get_rank())
     model.to(device)

examples/distributed_inference/tensor_parallel_simple_example.py

@@ -25,22 +25,29 @@
 import torch
 import torch.distributed as dist
 import torch.nn as nn
-import torch_tensorrt
 from tensor_parallel_initialize_dist import (
     cleanup_distributed_env,
     initialize_distributed_env,
 )
+
+if not dist.is_initialized():
+    initialize_distributed_env()
+import torch_tensorrt
 from torch.distributed._tensor import Shard
 from torch.distributed.tensor.parallel import (
     ColwiseParallel,
     RowwiseParallel,
     parallelize_module,
 )
-
-device_mesh, _world_size, _rank, logger = initialize_distributed_env(
-    "./tensor_parallel_simple_example"
+from torch_tensorrt.dynamo.distributed.utils import (
+    get_tensor_parallel_device_mesh,
+    initialize_distributed_logger,
 )
 
+device_mesh, _world_size, _rank = get_tensor_parallel_device_mesh()
+logger = initialize_distributed_logger(_rank, "tensor_parallel_simple_example")
+
+
 """
 This example takes some code from https://github.com/pytorch/examples/blob/main/distributed/tensor_parallelism/tensor_parallel_example.py
 """

py/torch_tensorrt/_utils.py

@@ -143,13 +143,55 @@ def _extracted_dir_trtllm(platform_system: str, platform_machine: str) -> Path:
     )
 
 
+def extract_wheel_file(wheel_path: Path, extract_dir: Path) -> None:
+    # this will not be encountered in case of platforms not supporting torch distributed/nccl/TRT-LLM
+    from torch.distributed import barrier, get_rank, is_initialized
+
+    if not is_initialized():
+        # Single process case, just unzip
+        is_master = True
+    else:
+        is_master = get_rank() == 0  # only rank 0 does the unzip
+
+    if is_master:
+        try:
+            import zipfile
+        except ImportError as e:
+            raise ImportError(
+                "zipfile module is required but not found. Please install zipfile"
+            )
+        try:
+            with zipfile.ZipFile(wheel_path) as zip_ref:
+                zip_ref.extractall(extract_dir)
+                logger.debug(f"Extracted wheel to {extract_dir}")
+
+        except FileNotFoundError as e:
+            # This should capture the errors in the download failure above
+            logger.error(f"Wheel file not found at {wheel_path}: {e}")
+            raise RuntimeError(
+                f"Failed to find downloaded wheel file at {wheel_path}"
+            ) from e
+        except zipfile.BadZipFile as e:
+            logger.error(f"Invalid or corrupted wheel file: {e}")
+            raise RuntimeError(
+                "Downloaded wheel file is corrupted or not a valid zip archive"
+            ) from e
+        except Exception as e:
+            logger.error(f"Unexpected error while extracting wheel: {e}")
+            raise RuntimeError(
+                "Unexpected error during extraction of TensorRT-LLM wheel"
+            ) from e
+
+    # Make sure others wait until unzip is done
+    if is_initialized():
+        barrier()
+
+
 def download_and_get_plugin_lib_path() -> Optional[str]:
     """
     Returns the path to the TensorRT‑LLM shared library, downloading and extracting if necessary.
-
     Args:
         platform (str): Platform identifier (e.g., 'linux_x86_64')
-
     Returns:
         Optional[str]: Path to shared library or None if operation fails.
     """
@@ -194,32 +236,7 @@ def download_and_get_plugin_lib_path() -> Optional[str]:
         except OSError as e:
             logger.error(f"Local file write error: {e}")
 
-    try:
-        import zipfile
-    except ImportError as e:
-        raise ImportError(
-            "zipfile module is required but not found. Please install zipfile"
-        )
-    try:
-        with zipfile.ZipFile(wheel_path) as zip_ref:
-            zip_ref.extractall(extract_dir)
-            logger.debug(f"Extracted wheel to {extract_dir}")
-    except FileNotFoundError as e:
-        # This should capture the errors in the download failure above
-        logger.error(f"Wheel file not found at {wheel_path}: {e}")
-        raise RuntimeError(
-            f"Failed to find downloaded wheel file at {wheel_path}"
-        ) from e
-    except zipfile.BadZipFile as e:
-        logger.error(f"Invalid or corrupted wheel file: {e}")
-        raise RuntimeError(
-            "Downloaded wheel file is corrupted or not a valid zip archive"
-        ) from e
-    except Exception as e:
-        logger.error(f"Unexpected error while extracting wheel: {e}")
-        raise RuntimeError(
-            "Unexpected error during extraction of TensorRT-LLM wheel"
-        ) from e
+    extract_wheel_file(wheel_path, extract_dir)
 
     try:
         wheel_path.unlink(missing_ok=True)
@@ -238,10 +255,8 @@ def download_and_get_plugin_lib_path() -> Optional[str]:
 def load_and_initialize_trtllm_plugin(plugin_lib_path: str) -> bool:
     """
     Loads and initializes the TensorRT-LLM plugin from the given shared library path.
-
     Args:
         plugin_lib_path (str): Path to the shared TensorRT-LLM plugin library.
-
     Returns:
         bool: True if successful, False otherwise.
     """
@@ -293,7 +308,6 @@ def load_tensorrt_llm_for_nccl() -> bool:
     Attempts to load the TensorRT-LLM plugin and initialize it.
     Either the env variable TRTLLM_PLUGINS_PATH can specify the path
     Or the user can specify USE_TRTLLM_PLUGINS as either of (1, true, yes, on) to download the TRT-LLM distribution and load it
-
     Returns:
         bool: True if the plugin was successfully loaded and initialized, False otherwise.
     """

py/torch_tensorrt/dynamo/_compiler.py

@@ -13,7 +13,6 @@
 from torch_tensorrt._Device import Device
 from torch_tensorrt._enums import EngineCapability, dtype
 from torch_tensorrt._features import needs_cross_compile
-from torch_tensorrt._Input import Input
 from torch_tensorrt.dynamo import _defaults, partitioning
 from torch_tensorrt.dynamo._DryRunTracker import (
     DryRunTracker,
@@ -287,7 +286,6 @@ def cross_compile_for_windows(
         arg_inputs = [arg_inputs]  # type: ignore
 
     # Prepare torch_trt inputs
-    trt_arg_inputs: Sequence[Input] = prepare_inputs(arg_inputs)
     trt_kwarg_inputs: Optional[dict[Any, Any]] = prepare_inputs(kwarg_inputs)
     device = to_torch_tensorrt_device(device)
     enabled_precisions = {dtype._from(p) for p in enabled_precisions}
@@ -377,7 +375,6 @@ def cross_compile_for_windows(
             )
     trt_gm = compile_module(
         gm,
-        trt_arg_inputs,
         trt_kwarg_inputs,
         settings,
     )
@@ -623,7 +620,6 @@ def compile(
         arg_inputs = [arg_inputs]  # type: ignore
 
     # Prepare torch_trt inputs
-    trt_arg_inputs: Sequence[Input] = prepare_inputs(arg_inputs)
     trt_kwarg_inputs: Optional[dict[Any, Any]] = prepare_inputs(kwarg_inputs)
     device = to_torch_tensorrt_device(device)
     enabled_precisions = {dtype._from(p) for p in enabled_precisions}
@@ -709,16 +705,13 @@ def compile(
             logger.warning(
                 "Remaining GPU memory may not be enough to compile the TensorRT engine for this model resulting in an OOM error, Consider setting offload_module_to_cpu=True"
             )
-    trt_gm = compile_module(
-        gm, trt_arg_inputs, trt_kwarg_inputs, settings, engine_cache
-    )
+    trt_gm = compile_module(gm, trt_kwarg_inputs, settings, engine_cache)
     return trt_gm
 
 
 @fn_supports_debugger  # type: ignore[misc]
 def compile_module(
     gm: torch.fx.GraphModule,
-    sample_arg_inputs: Sequence[Input],
     sample_kwarg_inputs: Optional[dict[Any, Any]] = None,
     settings: CompilationSettings = CompilationSettings(),
     engine_cache: Optional[BaseEngineCache] = None,