Spaces:
Runtime error
Runtime error
File size: 6,621 Bytes
fb0b30c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 |
#!/usr/bin/env python3
"""
Test script to verify training optimizations work correctly
"""
import torch
import os
import sys
from pathlib import Path
def test_mixed_precision():
"""Test mixed precision training setup"""
print("Testing Mixed Precision Training...")
try:
from torch.cuda.amp import GradScaler, autocast
# Test scaler creation
scaler = GradScaler(enabled=True)
print("✓ GradScaler created successfully")
# Test autocast context
with autocast():
x = torch.randn(10, 10)
y = torch.randn(10, 10)
z = x @ y
print("✓ Autocast context works")
return True
except Exception as e:
print(f"✗ Mixed precision test failed: {e}")
return False
def test_optimized_dataset():
"""Test optimized dataset functionality"""
print("\nTesting Optimized Dataset...")
try:
from morphological_dataset import MorphologicalDataset, build_vocabulary
# Create dummy data files
os.makedirs("test_data", exist_ok=True)
with open("test_data/test.src", "w") as f:
f.write("hello world\n")
f.write("test sequence\n")
with open("test_data/test.tgt", "w") as f:
f.write("hola mundo\n")
f.write("secuencia prueba\n")
# Test vocabulary building
src_vocab = build_vocabulary(["test_data/test.src"])
tgt_vocab = build_vocabulary(["test_data/test.tgt"])
print("✓ Vocabulary building works")
# Test dataset creation
dataset = MorphologicalDataset("test_data/test.src", "test_data/test.tgt",
src_vocab, tgt_vocab, max_length=10)
print("✓ Dataset creation works")
# Test data loading
item = dataset[0]
print(f"✓ Dataset item shape: {len(item)}")
# Cleanup
import shutil
shutil.rmtree("test_data")
return True
except Exception as e:
print(f"✗ Dataset test failed: {e}")
return False
def test_optimized_dataloader():
"""Test optimized DataLoader configuration"""
print("\nTesting Optimized DataLoader...")
try:
from torch.utils.data import DataLoader
from morphological_dataset import MorphologicalDataset, build_vocabulary, collate_fn
# Create test dataset
os.makedirs("test_data", exist_ok=True)
with open("test_data/test.src", "w") as f:
f.write("hello world\n")
f.write("test sequence\n")
with open("test_data/test.tgt", "w") as f:
f.write("hola mundo\n")
f.write("secuencia prueba\n")
src_vocab = build_vocabulary(["test_data/test.src"])
tgt_vocab = build_vocabulary(["test_data/test.tgt"])
dataset = MorphologicalDataset("test_data/test.src", "test_data/test.tgt",
src_vocab, tgt_vocab, max_length=10)
# Test optimized DataLoader
dataloader = DataLoader(
dataset,
batch_size=2,
shuffle=True,
collate_fn=lambda batch: collate_fn(batch, src_vocab, tgt_vocab, 10),
num_workers=0, # Use 0 for testing
pin_memory=False, # Disable for testing
persistent_workers=False, # Disable for testing
drop_last=True
)
# Test iteration
for batch in dataloader:
src, src_mask, tgt, tgt_mask = batch
print(f"✓ Batch shapes - src: {src.shape}, tgt: {tgt.shape}")
break
# Cleanup
import shutil
shutil.rmtree("test_data")
return True
except Exception as e:
print(f"✗ DataLoader test failed: {e}")
return False
def test_cuda_optimizations():
"""Test CUDA optimizations"""
print("\nTesting CUDA Optimizations...")
if not torch.cuda.is_available():
print("⚠ CUDA not available, skipping CUDA tests")
return True
try:
# Test CUDA optimizations
torch.backends.cudnn.benchmark = True
torch.backends.cudnn.deterministic = False
print("✓ CUDA optimizations enabled")
# Test non-blocking transfers
x = torch.randn(100, 100)
y = x.cuda(non_blocking=True)
print("✓ Non-blocking CUDA transfer works")
return True
except Exception as e:
print(f"✗ CUDA test failed: {e}")
return False
def test_model_creation():
"""Test model creation with optimizations"""
print("\nTesting Model Creation...")
try:
from transformer import TagTransformer
# Test model creation
model = TagTransformer(
src_vocab_size=1000,
trg_vocab_size=1000,
embed_dim=256,
nb_heads=4,
src_hid_size=1024,
src_nb_layers=2,
trg_hid_size=1024,
trg_nb_layers=2,
dropout_p=0.1,
tie_trg_embed=True,
label_smooth=0.1,
nb_attr=5,
src_c2i={},
trg_c2i={},
attr_c2i={}
)
print("✓ Model creation works")
# Test parameter count
param_count = model.count_nb_params()
print(f"✓ Model has {param_count:,} parameters")
return True
except Exception as e:
print(f"✗ Model test failed: {e}")
return False
def run_all_tests():
"""Run all optimization tests"""
print("=== Testing Training Optimizations ===\n")
tests = [
test_mixed_precision,
test_optimized_dataset,
test_optimized_dataloader,
test_cuda_optimizations,
test_model_creation
]
passed = 0
total = len(tests)
for test in tests:
try:
if test():
passed += 1
except Exception as e:
print(f"✗ Test {test.__name__} failed with exception: {e}")
print(f"\n=== Test Results ===")
print(f"Passed: {passed}/{total}")
if passed == total:
print("🎉 All tests passed! Optimizations are working correctly.")
return True
else:
print("❌ Some tests failed. Check the errors above.")
return False
if __name__ == '__main__':
success = run_all_tests()
sys.exit(0 if success else 1)
|