flags.cc 30.8 KB
Newer Older
1
// Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
2
// Copyright (c) 2022 NVIDIA Authors. All Rights Reserved.
3 4 5 6 7 8 9 10 11 12 13 14 15
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

Z
Zeng Jinle 已提交
16
#include "paddle/fluid/platform/flags.h"
17
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
18 19 20
#include "paddle/fluid/platform/cudnn_workspace_helper.h"
#endif

Z
Zeng Jinle 已提交
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
namespace paddle {
namespace platform {

const ExportedFlagInfoMap &GetExportedFlagInfoMap() {
  return *GetMutableExportedFlagInfoMap();
}

ExportedFlagInfoMap *GetMutableExportedFlagInfoMap() {
  static ExportedFlagInfoMap g_exported_flag_info_map;
  return &g_exported_flag_info_map;
}

}  // namespace platform
}  // namespace paddle

36 37 38
PADDLE_DEFINE_EXPORTED_int32(inner_op_parallelism, 0,
                             "number of threads for inner op");

39 40 41 42
/**
 * NOTE(paddle-dev): This file is designed to define all public FLAGS.
 */

43 44 45 46 47 48 49 50 51
/**
 * Paddle initialization related FLAG
 * Name: FLAGS_paddle_num_threads
 * Since Version: 0.15.0
 * Value Range: int32, default=1
 * Example: FLAGS_paddle_num_threads=2, set the maximum thread number per
 * instance to 2
 * Note:
 */
Z
Zeng Jinle 已提交
52 53
PADDLE_DEFINE_EXPORTED_int32(paddle_num_threads, 1,
                             "Number of threads for each paddle instance.");
54

55 56 57 58 59 60 61 62
/**
 * Operator related FLAG
 * Name: FLAGS_check_nan_inf
 * Since Version: 0.13.0
 * Value Range: bool, default=false
 * Example:
 * Note: Used to debug. Checking whether operator produce NAN/INF or not.
 */
Z
Zeng Jinle 已提交
63 64 65 66
PADDLE_DEFINE_EXPORTED_bool(
    check_nan_inf, false,
    "Checking whether operator produce NAN/INF or not. It will be "
    "extremely slow so please use this flag wisely.");
67

68 69 70 71
// NOTE(zhiqiu): better to share the flags, otherwise we will have too many
// flags.
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP) || \
    defined(PADDLE_WITH_ASCEND_CL)
72 73 74 75 76 77 78 79 80

/**
 * CUDA related related FLAG
 * Name: FLAGS_enable_cublas_tensor_op_math
 * Since Version: 1.2.0
 * Value Range: bool, default=false
 * Example:
 * Note: whether to use Tensor Core, faster but it may loss precision.
 */
Z
Zeng Jinle 已提交
81
PADDLE_DEFINE_EXPORTED_bool(
82 83 84 85 86 87 88 89 90
    enable_cublas_tensor_op_math, false,
    "The enable_cublas_tensor_op_math indicate whether to use Tensor Core, "
    "but it may loss precision. Currently, There are two CUDA libraries that"
    " use Tensor Cores, cuBLAS and cuDNN. cuBLAS uses Tensor Cores to speed up"
    " GEMM computations(the matrices must be either half precision or single "
    "precision); cuDNN uses Tensor Cores to speed up both convolutions(the "
    "input and output must be half precision) and recurrent neural networks "
    "(RNNs).");

91 92 93 94 95 96 97 98 99
/**
 * CUDA related FLAG
 * Name: FLAGS_selected_gpus
 * Since Version: 1.3.0
 * Value Range: integer list separated by comma, default empty list
 * Example: FLAGS_selected_gpus=0,1,2,3,4,5,6,7 to train or predict with 0~7 gpu
 * cards
 * Note: A list of device ids separated by comma, like: 0,1,2,3
 */
Z
Zeng Jinle 已提交
100 101 102 103 104 105 106 107 108
PADDLE_DEFINE_EXPORTED_string(
    selected_gpus, "",
    "A list of device ids separated by comma, like: 0,1,2,3. "
    "This option is useful when doing multi process training and "
    "each process have only one device (GPU). If you want to use "
    "all visible devices, set this to empty string. NOTE: the "
    "reason of doing this is that we want to use P2P communication"
    "between GPU devices, use CUDA_VISIBLE_DEVICES can only use"
    "share-memory only.");
109 110
#endif

111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133
#if defined(PADDLE_WITH_CUDA)
/**
 * CUDA related FLAG
 * Name: FLAGS_cublaslt_exhaustive_search_times
 * Since Version: 2.3.0
 * Value Range: int64_t, default=0
 * Example:
 * Note: Represents times of exhaustive search to evaluate performance of
 *       cuBlasLt matmul algorithm (with/without epilogue). Set this flag
 *       with value > 0 to enable exhaustive search. Default is 0, means
 *       getting algorithms via heuristic search. There are two search methods
 *       in cuBlasLt, heuristic search and exhaustive search. Exhaustive search
 *       attempts all cuBlasLt algorithms to select the fastest, which is very
 *       time-consuming, and the selected algorithm will be cached for a given
 *       layer specification Once you change the layer specifications
 *       (such as M, N and K), it will re-search again.
 */
PADDLE_DEFINE_EXPORTED_int64(
    cublaslt_exhaustive_search_times, 0,
    "The times of exhaustive search for cuBlasLt matmul with/without "
    " epilogue algorithms, default is 0, means disabling exhaustive search.");
#endif

134
#if defined(PADDLE_WITH_ASCEND_CL)
Z
Zeng Jinle 已提交
135 136 137 138 139 140 141 142 143 144 145
PADDLE_DEFINE_EXPORTED_string(
    selected_npus, "",
    "A list of device ids separated by comma, like: 0,1,2,3. "
    "This option is useful when doing multi process training and "
    "each process have only one device (NPU). If you want to use "
    "all visible devices, set this to empty string.");
PADDLE_DEFINE_EXPORTED_bool(
    hccl_check_nan, true,
    "Check Nan in tensor before hccl_allreduce_sum otherwise it'll "
    "core when meets Nan value");
PADDLE_DEFINE_EXPORTED_string(
146 147 148
    npu_config_path, "",
    "The absolute path of configuration json file, like: /tmp/config.json. "
    "If proveided, it will be passed to aclInit().");
Z
Zeng Jinle 已提交
149 150
PADDLE_DEFINE_EXPORTED_int32(min_loss_scaling, 1,
                             "set minmum loss scaling value!");
A
Aganlengzi 已提交
151 152 153 154 155 156 157
PADDLE_DEFINE_EXPORTED_string(
    npu_precision_mode, "",
    "NPU operator precision mode, options are 'force_fp32', 'force_fp16', "
    "'allow_fp32_to_fp16', 'must_keep_origin_dtype' and "
    "'allow_mix_precision'. If you want to use the default mode ("
    "allow_fp32_to_fp16), set this to empty string. For more details, "
    "please refer to the documents");
158
#endif
159

160
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
161 162 163 164 165 166 167 168 169
/**
 * CUDNN related FLAG
 * Name: FLAGS_cudnn_deterministic
 * Since Version: 0.13.0
 * Value Range: bool, default=false
 * Example:
 * Note: whether to use deterministic algorithm in cudnn.
 *       If true, it will slow down some operators such as conv and pooling.
 */
Z
Zeng Jinle 已提交
170 171 172 173 174
PADDLE_DEFINE_EXPORTED_bool(
    cudnn_deterministic, false,
    "Whether allow using an autotuning algorithm for convolution "
    "operator. The autotuning algorithm may be non-deterministic. If "
    "true, the algorithm is deterministic.");
175

176 177 178 179
/**
 * CUDNN related FLAG
 * Name: FLAGS_conv_workspace_size_limit
 * Since Version: 0.13.0
180
 * Value Range: uint64, default=512 (MB)
181 182 183 184 185 186 187
 * Example:
 * Note: The internal function of cuDNN obtains the fastest matching algorithm
 *       within this memory limit. Usually, faster algorithms can be chosen in
 *       larger workspaces, but memory space can also be significantly
 * increased.
 *       Users need to balance memory and speed.
 */
188 189 190
PADDLE_DEFINE_EXPORTED_int64(conv_workspace_size_limit,
                             paddle::platform::kDefaultConvWorkspaceSizeLimitMB,
                             "cuDNN convolution workspace limit in MB unit.");
191

192 193 194 195 196 197 198 199 200 201 202 203 204 205
/**
 * CUDNN related FLAG
 * Name: FLAGS_cudnn_exhaustive_search
 * Since Version: 1.2.0
 * Value Range: bool, default=false
 * Example:
 * Note: Represents whether an exhaustive search method is used to
 *       select a convolution algorithm. There are two search methods in cuDNN,
 *       heuristic search and exhaustive search. Exhaustive search attempts
 *       all cuDNN algorithms to select the fastest. This method is very
 *       time-consuming, and the selected algorithm will be cached for a given
 *       layer specification. Once you change the layer specifications
 *       (such as batch size, feature map size), it will search again.
 */
Z
Zeng Jinle 已提交
206 207 208 209
PADDLE_DEFINE_EXPORTED_bool(
    cudnn_exhaustive_search, false,
    "Whether enable exhaustive search for cuDNN convolution or "
    "not, default is False.");
210

211 212 213 214 215 216 217 218
/**
 * CUDNN related FLAG
 * Name: FLAGS_cudnn_exhaustive_search_times
 * Since Version:
 * Value Range:
 * Example:
 * Note: only used to predict for advanced developer
 */
Z
Zeng Jinle 已提交
219 220 221
PADDLE_DEFINE_EXPORTED_int64(cudnn_exhaustive_search_times, -1,
                             "Exhaustive search times for cuDNN convolution, "
                             "default is -1, not exhaustive search");
222

223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
/**
 * CUDNN related FLAG
 * Name: FLAGS_cudnn_batchnorm_spatial_persistent
 * Since Version: 1.4.0
 * Value Range: bool, default=false
 * Example:
 * Note: CUDNN_BATCHNORM_SPATIAL_PERSISTENT in batchnorm. This mode can be
 * faster in
 *       some tasks because an optimized path may be selected for
 * CUDNN_DATA_FLOAT
 *       and CUDNN_DATA_HALF data types, compute capability 6.0 or higher. The
 *       reason we set it to false by default is that this mode may use scaled
 *       atomic integer reduction that may cause a numerical overflow for
 * certain
 *       input data range.
 */
Z
Zeng Jinle 已提交
239 240 241 242
PADDLE_DEFINE_EXPORTED_bool(
    cudnn_batchnorm_spatial_persistent, false,
    "Whether enable CUDNN_BATCHNORM_SPATIAL_PERSISTENT mode for cudnn "
    "batch_norm, default is False.");
243 244
#endif

245
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
246 247 248

/**
 * NCCL related FLAG
249 250 251
 * Name: FLAGS_sync_nccl_allreduce
 * Since Version: 1.3
 * Value Range: bool, default=true
252 253 254 255 256
 * Example:
 * Note: asynchronous nccl allreduce or synchronous issue:
 *       https://github.com/PaddlePaddle/Paddle/issues/15049
 *       If you want to change this default value, why?(gongwb)
 */
Z
Zeng Jinle 已提交
257
PADDLE_DEFINE_EXPORTED_bool(
258 259 260 261 262 263
    sync_nccl_allreduce, true,
    "If set true, will call `cudaStreamSynchronize(nccl_stream)`"
    "after allreduce, this mode can get better performance in some scenarios.");
#endif

#ifdef PADDLE_WITH_DISTRIBUTE
264 265 266 267 268 269 270 271 272 273 274
/**
 * Distributed related FLAG
 * Name: FLAGS_communicator_max_merge_var_num
 * Since Version: 1.5.0
 * Value Range: int32, default=20
 * Example:
 * Note: The maximum number of gradients to be merged into a gradient and
 *       sent through the communicator. The trainer puts all the gradients
 *       into the queue, and then the communicator takes the gradients out
 *       of the queue and sends them after merging.
 */
Z
Zeng Jinle 已提交
275 276 277 278 279 280
PADDLE_DEFINE_EXPORTED_int32(communicator_max_merge_var_num, 20,
                             "max var num to merge and send");
PADDLE_DEFINE_EXPORTED_bool(
    communicator_is_sgd_optimizer, true,
    "gradient sent to the server is the sum of the gradients "
    "calculated by each thread if optimizer is sgd");
281 282 283 284 285 286 287 288 289 290 291 292 293
/**
 * Distributed related FLAG
 * Name: FLAGS_communicator_send_queue_size
 * Since Version: 1.5.0
 * Value Range: int32, default=20
 * Example:
 * Note: Size for each gradient queue. The trainer puts the gradient into
 *       the queue, and then the communicator takes it out of the queue and
 *       sends it out. When the communicator is slow, the queue may be full,
 *       and the trainer will be continuously blocked before the queue has
 *       space. It is used to avoid training much faster than communication,
 *       so that too many gradients are not sent out in time.
 */
Z
Zeng Jinle 已提交
294 295
PADDLE_DEFINE_EXPORTED_int32(communicator_send_queue_size, 20,
                             "queue size to recv gradient before send");
296 297
#endif

298 299 300 301 302 303 304 305 306
/**
 * Distributed related FLAG
 * Name: FLAGS_dist_threadpool_size
 * Since Version: 1.0.0
 * Value Range: int32, default=0
 * Example:
 * Note: Control the number of threads used for distributed modules.
 *       If it is not set, it is set to a hard thread.
 */
Z
Zeng Jinle 已提交
307 308 309
PADDLE_DEFINE_EXPORTED_int32(
    dist_threadpool_size, 0,
    "number of threads used for distributed executed.");
310

311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
/**
 * Garbage collector related FLAG
 * Name: FLAGS_eager_delete_tensor_gb
 * Since Version: 1.0.0
 * Value Range: double, default=kDefaultEagerDeleteTensorGB
 * Example: FLAGS_eager_delete_tensor_gb=0.0, Release memory garbage once it is
 * no longer used.
 *          FLAGS_eager_delete_tensor_gb=1.0, Release memory garbage when
 * garbage occupies 1.0GB of memory.
 *          FLAGS_eager_delete_tensor_gb=-1.0, Disable garbage collection
 * policy.
 * Note: Represents whether a garbage collection strategy is used to optimize
 * network memory usage.
 *       It is recommended that users set FLAGS_eager_delete_tensor_gb=0.0 to
 *       enable garbage collection strategy when training large networks.
 */
327 328 329 330 331 332 333
// Disable gc by default when inference library is built
#ifdef PADDLE_ON_INFERENCE
static const double kDefaultEagerDeleteTensorGB = -1;
#else
static const double kDefaultEagerDeleteTensorGB = 0;
#endif

Z
Zeng Jinle 已提交
334
PADDLE_DEFINE_EXPORTED_double(
335 336 337 338
    eager_delete_tensor_gb, kDefaultEagerDeleteTensorGB,
    "Memory size threshold (GB) when the garbage collector clear tensors."
    "Disabled when this value is less than 0");

339 340 341 342 343 344 345 346 347 348 349 350
/**
 * Memory related FLAG
 * Name: FLAGS_fast_eager_deletion_mode
 * Since Version: 1.3.0
 * Value Range: bool, default=true
 * Example:
 * Note: Whether to use fast garbage collection strategy.
 *       If not set, the GPU memory is released at the end of the CUDA kernel.
 *       Otherwise, the GPU memory will be released before the CUDA kernel
 *       has finished, which will make the garbage collection strategy faster.
 *       Only works when garbage collection strategy is enabled.
 */
Z
Zeng Jinle 已提交
351 352 353 354
PADDLE_DEFINE_EXPORTED_bool(
    fast_eager_deletion_mode, true,
    "Fast eager deletion mode. If enabled, memory would release "
    "immediately without waiting GPU kernel ends.");
355

356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373
/**
 * Memory related FLAG
 * Name: FLAGS_memory_fraction_of_eager_deletion
 * Since Version: 1.4
 * Value Range: double [0.0, 1.0], default=1.0
 * Example:
 * Note: The percentage of memory size of garbage collection policy
 *       to release variables.
 *       If FLAGS_memory_fraction_of_eager_deletion = 1.0,
 *       all temporary variables in the network will be released.
 *       If FLAGS_memory_fraction_of_eager_deletion = 0.0,
 *       no temporary variables in the network are released.
 *       If 0.0 < FLAGS_memory_fraction_of_eager_deletion < 1.0,
 *       all temporary variables will be sorted in descending order
 *       according to their memory size, and only variables with the
 *       largest FLAGS_memory_fraction_of_eager_deletion ratio will be released.
 *       The flag is only valid when running parallel data compilers.
 */
Z
Zeng Jinle 已提交
374 375 376 377 378 379
PADDLE_DEFINE_EXPORTED_double(
    memory_fraction_of_eager_deletion, 1.0,
    "Fraction of eager deletion. If less than 1.0, all variables in "
    "the program would be sorted according to its memory size, and "
    "only the FLAGS_memory_fraction_of_eager_deletion of the largest "
    "variables would be deleted.");
380

381 382 383 384
/**
 * Allocator related FLAG
 * Name: FLAGS_allocator_strategy
 * Since Version: 1.2
385 386
 * Value Range: string, {naive_best_fit, auto_growth, thread_local},
 * default=auto_growth
387
 * Example:
388
 * Note: For selecting allocator policy of PaddlePaddle.
389
 */
390
static constexpr char kDefaultAllocatorStrategy[] = "auto_growth";
Z
Zeng Jinle 已提交
391
PADDLE_DEFINE_EXPORTED_string(
392 393 394 395 396 397 398 399 400 401 402 403
    allocator_strategy, kDefaultAllocatorStrategy,
    "The allocation strategy, enum in [naive_best_fit, auto_growth]. "
    "naive_best_fit means the original pre-allocated allocator of Paddle. "
    "auto_growth means the auto-growth allocator. "
    "These two strategies differ in GPU memory allocation. "
    "naive_best_fit strategy would occupy almost all GPU memory by default, "
    "which prevents users from starting several Paddle jobs on the same GPU "
    "card but leads to less memory fragmentation (i.e., maximum batch "
    "size of models may be larger). auto_growth strategy would allocate "
    "GPU memory on demand, which allows users to start several Paddle jobs "
    "on the same GPU card but may lead to more memory fragmentation "
    "(i.e., maximum batch size of models may be smaller).");
404

405 406 407
/**
 * Memory related FLAG
 * Name: FLAGS_fraction_of_cpu_memory_to_use
408 409
 * Since Version: 0.12.0
 * Value Range: double, [0.0, 1.0], default=1
410
 * Example:
411 412 413 414 415 416
 * Note: Represents the proportion of allocated CPU memory blocks
 *       to the total memory size of the CPU. Future CPU memory usage
 *       will be allocated from this memory block. If the memory block does
 *       not have enough CUDA pinned memory, new memory blocks of the same
 *       size as the memory block will be allocated from the CUDA pinned
 *       request util the CPU does not have enough memory.
417
 */
Z
Zeng Jinle 已提交
418 419 420
PADDLE_DEFINE_EXPORTED_double(fraction_of_cpu_memory_to_use, 1,
                              "Default use 100% of CPU memory for PaddlePaddle,"
                              "reserve the rest for page tables, etc");
421 422 423 424 425 426 427 428 429 430 431 432 433

/**
 * Memory related FLAG
 * Name: FLAGS_initial_cpu_memory_in_mb
 * Since Version: 0.14.0
 * Value Range: uint64, default=500 (MB)
 * Example:
 * Note: The CPU memory block size of the initial allocator in MB.
 *       The allocator takes the minimum values of
 *       FLAGS_initial_cpu_memory_in_mb and
 *       FLAGS_fraction_of_cpu_memory_to_use*(total physical memory)
 *       as memory block sizes.
 */
Z
Zeng Jinle 已提交
434 435 436
PADDLE_DEFINE_EXPORTED_uint64(
    initial_cpu_memory_in_mb, 500ul,
    "Initial CPU memory for PaddlePaddle, in MD unit.");
437

438 439 440
/**
 * Memory related FLAG
 * Name: FLAGS_fraction_of_cuda_pinned_memory_to_use
441 442
 * Since Version: 0.12.0
 * Value Range: double, [0.0, 1.0], default=0.5
443
 * Example:
444 445 446 447 448 449
 * Note: Represents the proportion of allocated CUDA pinned memory blocks
 *       to the total memory size of the CPU. Future CUDA pinned memory usage
 *       will be allocated from this memory block. If the memory block does
 *       not have enough CPU memory, new memory blocks of the same
 *       size as the memory block will be allocated from the CPU
 *       request util the CPU does not have enough memory.
450
 */
Z
Zeng Jinle 已提交
451
PADDLE_DEFINE_EXPORTED_double(
452 453 454 455
    fraction_of_cuda_pinned_memory_to_use, 0.5,
    "Default use 50% of CPU memory as the pinned_memory for PaddlePaddle,"
    "reserve the rest for page tables, etc");

456 457
// NOTE(zhiqiu): better to share the flags, otherwise we will have too many
// flags.
458 459 460
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP) ||      \
    defined(PADDLE_WITH_ASCEND_CL) || defined(PADDLE_WITH_MLU) || \
    defined(PADDLE_WITH_CUSTOM_DEVICE)
461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478

/**
 * Memory related FLAG
 * Name: FLAGS_fraction_of_gpu_memory_to_use
 * Since Version: 1.2.0
 * Value Range: double, default=0.5 if win32, 0.92 else
 * Example:
 * Note: Represents the proportion of allocated memory blocks to the total
 * memory size
 *       of the GPU. Future memory usage will be allocated from this memory
 * block.
 *       If the memory block does not have enough GPU memory, new memory blocks
 * of
 *       the same size as the memory block will be allocated from the GPU
 * request
 *       until the GPU does not have enough memory.
 */

479 480 481 482 483 484 485 486
#ifndef _WIN32
constexpr static float fraction_of_gpu_memory_to_use = 0.92f;
#else
// fraction_of_gpu_memory_to_use cannot be too high on windows,
// since the win32 graphic sub-system can occupy some GPU memory
// which may lead to insufficient memory left for paddle
constexpr static float fraction_of_gpu_memory_to_use = 0.5f;
#endif
Z
Zeng Jinle 已提交
487 488 489 490 491 492 493
PADDLE_DEFINE_EXPORTED_double(
    fraction_of_gpu_memory_to_use, fraction_of_gpu_memory_to_use,
    "Allocate a trunk of gpu memory that is this fraction of the "
    "total gpu memory size. Future memory usage will be allocated "
    "from the trunk. If the trunk doesn't have enough gpu memory, "
    "additional trunks of the same size will be requested from gpu "
    "until the gpu has no memory left for another trunk.");
494

495 496 497 498 499 500 501 502 503 504 505 506
/**
 * Memory related FLAG
 * Name: FLAGS_initial_gpu_memory_in_mb
 * Since Version: 1.4.0
 * Value Range: uint64, default=0 (MB)
 * Example:
 * Note: Allocate a specified size of GPU memory block. Later memory usage
 *       will be allocated from that memory block. If the memory block does not
 *       have enough GPU memory, the memory block with the size
 *       FLAGS_reallocate_gpu_memory_in_mb will be requested from the GPU until
 *       the GPU has no remaining memory.
 */
Z
Zeng Jinle 已提交
507
PADDLE_DEFINE_EXPORTED_uint64(
508 509 510 511 512 513 514 515 516 517 518 519
    initial_gpu_memory_in_mb, 0ul,
    "Allocate a trunk of gpu memory whose byte size is specified by "
    "the flag. Future memory usage will be allocated from the "
    "trunk. If the trunk doesn't have enough gpu memory, additional "
    "trunks of the gpu memory will be requested from gpu with size "
    "specified by FLAGS_reallocate_gpu_memory_in_mb until the gpu has "
    "no memory left for the additional trunk. Note: if you set this "
    "flag, the memory size set by "
    "FLAGS_fraction_of_gpu_memory_to_use will be overrided by this "
    "flag. If you don't set this flag, PaddlePaddle will use "
    "FLAGS_fraction_of_gpu_memory_to_use to allocate gpu memory");

520 521 522 523 524 525 526 527 528
/**
 * Memory related FLAG
 * Name: FLAGS_reallocate_gpu_memory_in_mb
 * Since Version: 1.4.0
 * Value Range: uint64, default=0 (MB)
 * Example:
 * Note: If the allocated GPU memory blocks are exhausted,
 *       additional GPU memory blocks are reallocated
 */
Z
Zeng Jinle 已提交
529 530 531 532 533 534 535 536 537 538 539 540 541 542
PADDLE_DEFINE_EXPORTED_uint64(
    reallocate_gpu_memory_in_mb, 0ul,
    "If this flag is set, Paddle will reallocate the gpu memory with "
    "size specified by this flag. Else Paddle will reallocate by "
    "FLAGS_fraction_of_gpu_memory_to_use");

PADDLE_DEFINE_EXPORTED_uint64(
    gpu_memory_limit_mb, 0UL,
    "The maximum gpu memory limit that the process can allocate. "
    "If it is equal to 0, there would be no limit and all gpu memory "
    "would be available to the process. If it is larger than 0, "
    "the process would raise out of memory error if the allocated "
    "memory exceeds the limit even though there is available "
    "memory on the gpu card. The unit is MB and default value is 0.");
543

544
#endif
545 546 547 548 549 550 551 552 553

/**
 * Scope related FLAG
 * Name: local_exe_sub_scope_limit
 * Since Version: 1.6.0
 * Value Range: double, default=256 (MB)
 * Example:
 * Note:
 */
Z
Zeng Jinle 已提交
554 555 556 557 558 559
PADDLE_DEFINE_EXPORTED_double(
    local_exe_sub_scope_limit, 256.0,  // MBytes
    "The memory up limit of sub-scopes of local execution scope for "
    "each CUDAPlace. If you don't need to limit the memory, "
    "you should set FLAGS_local_exe_sub_scope_limit=-1. "
    "The default value is 256 MBytes.");
560

561 562 563 564 565
PADDLE_DEFINE_EXPORTED_bool(
    reader_queue_speed_test_mode, false,
    "If set true, the queue.pop will only get data from queue but not "
    "remove the data from queue for speed testing");

566 567 568 569 570 571 572 573
/**
 * MKLDNN related FLAG
 * Name: use_mkldnn
 * Since Version:
 * Value Range: bool, default=false
 * Example:
 * Note:
 */
Z
Zeng Jinle 已提交
574
PADDLE_DEFINE_EXPORTED_bool(use_mkldnn, false, "Use MKLDNN to run");
575 576 577 578 579 580 581 582 583 584 585 586 587 588 589

/**
 * Debug related FLAG
 * Name: FLAGS_call_stack_level
 * Since Version: 2.0.0
 * Value Range: int, default=2
 * Example:
 * Note: Used to debug. Determine the call stack to print when error or
 * exeception happens.
 * If FLAGS_call_stack_level == 0, only the error message summary will be shown.
 * If FLAGS_call_stack_level == 1, the python stack and  error message summary
 * will be shown.
 * If FLAGS_call_stack_level == 2, the python stack, c++ stack, and error
 * message summary will be shown.
 */
590 591 592 593 594 595
#ifdef PADDLE_ON_INFERENCE
static const int32_t kDefaultCallStackLevel = 2;
#else
static const int32_t kDefaultCallStackLevel = 1;
#endif

Z
Zeng Jinle 已提交
596
PADDLE_DEFINE_EXPORTED_int32(
597
    call_stack_level, kDefaultCallStackLevel,
598 599 600 601 602 603 604 605
    "Determine the call stack to print when error or exeception happens."
    // TODO(zhiqiu): implement logic of FLAGS_call_stack_level==0
    // "If FLAGS_call_stack_level == 0, only the error message summary will be "
    // "shown. "
    "If FLAGS_call_stack_level == 1, the python stack and error message "
    "summary will be shown."
    "If FLAGS_call_stack_level == 2, the python stack, c++ stack, and "
    "error message summary will be shown.");
606 607 608 609 610 611 612 613 614 615

/**
 * Debug related FLAG
 * Name: sort_sum_gradient
 * Since Version: 2.0.0
 * Value Range: bool, default=false
 * Example:
 * Note: If True, gradients are summed by the reverse order of
 * the forward execution sequence.
 */
Z
Zeng Jinle 已提交
616 617 618
PADDLE_DEFINE_EXPORTED_bool(sort_sum_gradient, false,
                            "Sum gradients by the reverse order of "
                            "the forward execution sequence.");
619 620 621 622 623 624 625 626 627

/**
 * Performance related FLAG
 * Name: max_inplace_grad_add
 * Since Version: 2.0.0
 * Value Range: int32, default=0
 * Example:
 * Note: The maximum number of inplace grad_add.
 */
Z
Zeng Jinle 已提交
628
PADDLE_DEFINE_EXPORTED_int32(
629 630 631 632 633
    max_inplace_grad_add, 0,
    "The maximum number of inplace grad_add. When doing "
    "gradient accumulation, if the number of gradients need to that "
    "less FLAGS_max_inplace_grad_add, than it will be use several grad_add"
    "instead of sum. Default is 0.");
634 635 636 637 638 639 640 641 642

/**
 * Debug related FLAG
 * Name: tracer_mkldnn_ops_on
 * Since Version: 2.0.0
 * Value Range: string, default=empty
 * Example:
 * Note: Holds list of operation types with OneDNN kernels to be enabled.
 */
Z
Zeng Jinle 已提交
643 644
PADDLE_DEFINE_EXPORTED_string(tracer_mkldnn_ops_on, "",
                              "List of OneDNN operation types to be turned on");
645 646 647 648 649 650 651 652 653

/**
 * Debug related FLAG
 * Name: tracer_mkldnn_ops_off
 * Since Version: 2.0.0
 * Value Range: string, default=empty
 * Example:
 * Note: Holds list of operation types with OneDNN kernels to be disabled.
 */
Z
Zeng Jinle 已提交
654 655 656
PADDLE_DEFINE_EXPORTED_string(
    tracer_mkldnn_ops_off, "",
    "List of OneDNN operation types to be turned off");
657

658 659 660 661 662 663 664 665 666
/**
 * Debug related FLAG
 * Name: check_kernel_launch
 * Since Version: 2.1.0
 * Value Range: bool, default=false
 * Example:
 * Note: Check kernel launch status after every kernel compute.
 */
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
Z
Zeng Jinle 已提交
667 668 669
PADDLE_DEFINE_EXPORTED_bool(
    check_kernel_launch, false,
    "Check kernel launch status after every kernel compute");
670 671
#endif

672 673 674 675 676 677 678 679 680
/**
 * CUDNN related FLAG
 * Name: conv2d_disable_cudnn
 * Since Version:
 * Value Range: bool, default=false
 * Example:
 * Note: Disable cudnn in conv2d.
 */
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
Z
Zeng Jinle 已提交
681 682
PADDLE_DEFINE_EXPORTED_bool(conv2d_disable_cudnn, false,
                            "Disable cudnn in conv2d");
683 684 685

PADDLE_DEFINE_EXPORTED_bool(use_fast_math, false,
                            "Whether to use fast math GPU functions.");
686
#endif
B
Baibaifan 已提交
687 688 689 690 691 692 693 694 695

/**
 * Distributed related FLAG
 * Name: FLAGS_get_host_by_name_time
 * Since Version: 2.2.0
 * Value Range: int32, default=120
 * Example:
 * Note: Get host by name time.
 */
F
fwenguang 已提交
696 697 698
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_XPU) ||      \
    defined(PADDLE_WITH_ASCEND_CL) || defined(PADDLE_WITH_HIP) || \
    defined(PADDLE_WITH_MLU)
Z
Zeng Jinle 已提交
699 700
PADDLE_DEFINE_EXPORTED_int32(get_host_by_name_time, 120,
                             "The maximum time for get host by name time");
B
Baibaifan 已提交
701
#endif
702 703 704 705 706 707 708 709 710 711

/**
 * Distributed related FLAG
 * Name: FLAGS_apply_pass_to_program
 * Since Version: 2.2.0
 * Value Range: bool, default=false
 * Example: FLAGS_apply_pass_to_program=true would apply IR Pass to
 *          program when using Fleet APIs.
 * Note: Apply IR pass to program. Be only useful when using Fleet APIs.
 */
Z
Zeng Jinle 已提交
712
PADDLE_DEFINE_EXPORTED_bool(
713 714
    apply_pass_to_program, false,
    "It controls whether to apply IR pass to program when using Fleet APIs");
Y
yaoxuefeng 已提交
715

F
Feng Xing 已提交
716
/**
L
Liu-xiandong 已提交
717
 * KP kernel related FLAG
F
Feng Xing 已提交
718 719 720
 * Name: FLAGS_run_kp_kernel
 * Since Version: 2.3.0
 * Value Range: bool, default=false
L
Liu-xiandong 已提交
721 722
 * Example: FLAGS_run_kp_kernel=true would use the kp kernel to compute in the
 * Op.
F
Feng Xing 已提交
723 724
 * Note:
 */
L
Liu-xiandong 已提交
725 726
PADDLE_DEFINE_EXPORTED_bool(run_kp_kernel, false,
                            "It controls whether to run PaddlePaddle using KP");
F
Feng Xing 已提交
727

728
/**
729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744
 * Distributed related FLAG
 * Name: FLAGS_allreduce_record_one_event
 * Since Version: 2.2.0
 * Value Range: bool, default=false
 * Example: FLAGS_allreduce_record_one_event=true makes the allreduce
 *          operations would only wait one event instead of multiple events.
 * Note: Make the allreduce operations would only wait one event instead of
 *       multiple events. Currently, only fuse allreduce supports this.
 *       Otherwise, the precision may be wrong.
 */
PADDLE_DEFINE_EXPORTED_bool(allreduce_record_one_event, false,
                            "It controls whether the allreduce operations "
                            "would only wait one event instead of multiple "
                            "events. Currently, only fuse allreduce supports "
                            "this. Otherwise, the precision may be wrong.");

745
#ifdef PADDLE_WITH_CINN
746
/*
747 748 749 750 751 752 753 754
 * CINN related FLAG
 * Name: FLAGS_use_cinn
 * Since Version: 2.3
 * Value Range: bool, default=false
 * Example: FLAGS_use_cinn=true would run PaddlePaddle using CINN
 */
PADDLE_DEFINE_EXPORTED_bool(
    use_cinn, false, "It controls whether to run PaddlePaddle using CINN");
755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777

/*
 * CINN related FLAG
 * Name: FLAGS_allow_cinn_ops
 * Since Version: 2.3
 * Value Range: string, default=""
 * Example: FLAGS_allow_cinn_ops="mul;relu" would only cover `mul` and `relu`
 * when using CINN
 */
PADDLE_DEFINE_EXPORTED_string(allow_cinn_ops, "",
                              "It controls the cinn op subset to be used, "
                              "which has the highest priority.");

/*
 * CINN related FLAG
 * Name: FLAGS_deny_cinn_ops
 * Since Version: 2.3
 * Value Range: string, default=""
 * Example: FLAGS_deny_cinn_ops="mul;relu" would block `mul` and `relu` two ops
 * when using CINN
 */
PADDLE_DEFINE_EXPORTED_string(deny_cinn_ops, "",
                              "It controls the cinn op subset to be not used.");
778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803

/*
 * CINN related FLAG
 * Name: FLAGS_enable_pe_launch_cinn
 * Since Version: 2.3
 * Value Range: bool, default=true
 * Example: FLAGS_enable_pe_launch_cinn=true would execute the CINN compiled
 * instructions of a paddle graph with ParallelExecutor, otherwise with the
 * CINN compiled runtime program in sequential order.
 */
PADDLE_DEFINE_EXPORTED_bool(enable_pe_launch_cinn, true,
                            "It controls whether to execute cinn compiled "
                            "program with ParallelExecutor");

/*
 * CINN related FLAG
 * Name: FLAGS_enable_cinn_auto_tune
 * Since Version: 2.3
 * Value Range: bool, default=false
 * Example: FLAGS_enable_cinn_auto_tune=true would use CINN with its
 * auto-tune feature enabled
 */
PADDLE_DEFINE_EXPORTED_bool(enable_cinn_auto_tune, false,
                            "It controls whether to use cinn with "
                            "its auto-tune feature enabled");

804
#endif
805

Y
yaoxuefeng 已提交
806 807 808 809 810 811
DEFINE_int32(record_pool_max_size, 2000000,
             "SlotRecordDataset slot record pool max size");
DEFINE_int32(slotpool_thread_num, 1, "SlotRecordDataset slot pool thread num");
DEFINE_bool(enable_slotpool_wait_release, false,
            "enable slotrecord obejct wait release, default false");
DEFINE_bool(enable_slotrecord_reset_shrink, false,
Y
yaoxuefeng 已提交
812 813 814
            "enable slotrecord obejct reset shrink memory, default false");
DEFINE_bool(enable_ins_parser_file, false,
            "enable parser ins file , default false");
815 816 817 818 819 820 821 822 823 824 825 826

/**
 * ProcessGroupNCCL related FLAG
 * Name: nccl_blocking_wait
 * Since Version:
 * Value Range: bool, default=false
 * Example:
 * Note: nccl blocking wait.
 */
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
PADDLE_DEFINE_EXPORTED_bool(nccl_blocking_wait, false, "nccl blocking wait");
#endif
827 828 829 830 831 832 833 834 835

/**
 * Autotune related FLAG
 * Name: FLAGS_use_autotune
 * Since Version: 2.3.0
 * Value Range: bool, default=false
 * Example:
 */
PADDLE_DEFINE_EXPORTED_bool(use_autotune, false, "Whether enable autotune.");