Add optional device backend support (Accelerate, torch_xla) and flexible device selection for model loading

language_model/language-model-standalone.py

@@ -7,6 +7,23 @@ import os
 import re
 import logging
 import torch
+"""
+Optional device backends:
+- Accelerate: simplifies multi-GPU/TPU/CPU device placement
+- torch_xla: TPU support (only available in TPU environments)
+Both imports are optional to keep backward compatibility.
+"""
+try:
+    from accelerate import Accelerator  # type: ignore
+except Exception:
+    Accelerator = None  # accelerate is optional
+
+try:
+    import torch_xla.core.xla_model as xm  # type: ignore
+    XLA_AVAILABLE = True
+except Exception:
+    xm = None
+    XLA_AVAILABLE = False
 import lm_decoder
 from functools import lru_cache
 from transformers import AutoModelForCausalLM, AutoTokenizer
@@ -92,7 +109,8 @@ def update_ngram_params(
 def build_opt(
         model_name='facebook/opt-6.7b',
         cache_dir=None,
-        device='cuda' if torch.cuda.is_available() else 'cpu',
+        device=None,
+        accelerator=None,
     ):
     
     '''
@@ -101,17 +119,50 @@ def build_opt(
     Put the model onto the GPU (if available).
     '''
     
+    # Resolve device automatically if not provided
+    if device is None:
+        if accelerator is not None:
+            device = accelerator.device
+        elif XLA_AVAILABLE and xm is not None:
+            # Use all available TPU cores for multi-core processing
+            device = xm.xla_device()
+            logging.info(f"TPU cores available: {xm.xrt_world_size()}")
+        elif torch.cuda.is_available():
+            device = torch.device('cuda')
+        else:
+            device = torch.device('cpu')
+
+    # Choose appropriate dtype per device
+    try:
+        device_type = device.type  # torch.device or XLA device
+    except AttributeError:
+        # Fallback for XLA device objects
+        device_type = str(device)
+
+    if XLA_AVAILABLE and (str(device).startswith('xla') or device_type == 'xla'):
+        load_dtype = torch.bfloat16  # TPU prefers bfloat16
+    elif device_type == 'cuda':
+        load_dtype = torch.float16
+    else:
+        load_dtype = torch.float32
+
     # load tokenizer and model
     tokenizer = AutoTokenizer.from_pretrained(model_name, cache_dir=cache_dir)
     model = AutoModelForCausalLM.from_pretrained(
         model_name,
         cache_dir=cache_dir,
-        torch_dtype=torch.float16,
+        torch_dtype=load_dtype,
     )
 
-    if device != 'cpu':
-        # Move the model to the GPU
+    # Device placement
+    if accelerator is not None:
+        model = accelerator.prepare(model)
+    else:
         model = model.to(device)
+        # For TPU multi-core, ensure model is replicated across all cores
+        if XLA_AVAILABLE and str(device).startswith('xla') and xm is not None:
+            # This will be handled by torch_xla internally when using xla_device()
+            pass
 
     # Set the model to evaluation mode
     model.eval()
@@ -466,8 +517,24 @@ def main(args):
         'score_penalty_percent': float(score_penalty_percent),
     }
 
-    # pick GPU
-    device = torch.device(f"cuda:{gpu_number}" if torch.cuda.is_available() else "cpu")
+    # pick device (Accelerate -> TPU -> CUDA -> CPU)
+    accelerator = None
+    if Accelerator is not None:
+        try:
+            accelerator = Accelerator()
+        except Exception:
+            accelerator = None
+
+    if accelerator is not None:
+        device = accelerator.device
+    elif XLA_AVAILABLE and xm is not None:
+        # Use all available TPU cores for multi-core processing
+        device = xm.xla_device()
+        logging.info(f"TPU cores available: {xm.xrt_world_size()}")
+    elif torch.cuda.is_available():
+        device = torch.device(f"cuda:{gpu_number}")
+    else:
+        device = torch.device("cpu")
     logging.info(f'Using device: {device}')
 
     # initialize opt model
@@ -477,6 +544,7 @@ def main(args):
         lm, lm_tokenizer = build_opt(
             cache_dir=opt_cache_dir,
             device=device,
+            accelerator=accelerator,
         )
         logging.info(f'OPT model successfully built in {(time.time()-start_time):0.4f} seconds.')
 

language_model/requirements.txt

@@ -6,4 +6,5 @@ tensorboard
 tensorboardX
 typeguard
 textgrid
-redis
+redis
+accelerate>=0.33.0