fix: mindrecord page size comment

docs/api/api_python/dataset/dataset_method/batch/mindspore.dataset.Dataset.batch.rst

@@ -3,7 +3,7 @@ mindspore.dataset.Dataset.batch
 
 .. py:method:: mindspore.dataset.Dataset.batch(batch_size, drop_remainder=False, num_parallel_workers=None, **kwargs)
 
-    将数据集中连续 `batch_size` 条数据合并为一个批处理数据，其中batch成一个Tensor前可选择使用 `per_batch_map` 对样本进行处理。
+    将数据集中连续 `batch_size` 条数据组合为一个批数据，并可通过可选参数 `per_batch_map` 指定组合前要进行的预处理操作。
 
     `batch` 操作要求每列中的数据具有相同的shape。
 

docs/api/api_python/dataset/dataset_method/batch/mindspore.dataset.Dataset.padded_batch.rst

@@ -3,7 +3,7 @@ mindspore.dataset.Dataset.padded_batch
 
 .. py:method:: mindspore.dataset.Dataset.padded_batch(batch_size, drop_remainder=False, num_parallel_workers=None, pad_info=None)
 
-    将数据集中连续 `batch_size` 条数据合并为一个批处理数据，其中batch成一个Tensor前可选择使用 `pad_info` 预先将样本补齐。
+    将数据集中连续 `batch_size` 条数据组合为一个批数据，并可通过可选参数 `pad_info` 预先将样本补齐。
 
     `batch` 操作要求每列中的数据具有相同的shape。
 

mindspore/ccsrc/minddata/dataset/engine/datasetops/data_queue_op.cc

@@ -235,6 +235,9 @@ Status DataQueueOp::SendDataToAscend() {
   TensorRow curr_row;
   RETURN_IF_NOT_OK(child_iterator_->FetchNextTensorRow(&curr_row));
   first_fetch_flag_ = true;
+
+  MS_LOG(INFO) << "Begin to send data to device, channel name: " << channel_name_;
+
   while (!curr_row.eof() && !is_break_loop) {
     while (!curr_row.eoe() && !is_break_loop) {
       RETURN_IF_NOT_OK(FilterMetadata(&curr_row));
@@ -263,6 +266,7 @@ Status DataQueueOp::SendDataToAscend() {
       batch_record_start = ProfilingTime::GetCurMilliSecond();
 #endif
       send_batch++;
+      MS_LOG(INFO) << "Have sent " << send_batch << " batch(es) to device, channel name: " << channel_name_;
 #ifdef ENABLE_DUMP_IR
       RETURN_IF_NOT_OK(md_channel_info_->RecordBatchQueue(ChildOpConnectorSize()));
       RETURN_IF_NOT_OK(md_channel_info_->RecordPreprocessBatch(send_batch));
@@ -543,6 +547,7 @@ Status DataQueueOp::PushDataToGPU() {
                         gpu_connector_->capacity(), gpu_connector_->size());
 #endif
       send_batch++;
+      MS_LOG(INFO) << "Have sent " << send_batch << " batch(es) to device, channel name: " << channel_name_;
 #ifdef ENABLE_DUMP_IR
       md_channel_info_->RecordBatchQueue(gpu_connector_->size());
       md_channel_info_->RecordPreprocessBatch(send_batch);
@@ -638,6 +643,9 @@ Status DataQueueOp::SendDataToGPU() {
   first_fetch_flag_ = true;
   int64_t num_buf = 0;
   bool is_break_loop = false;
+
+  MS_LOG(INFO) << "Begin to send data to device, channel name: " << channel_name_;
+
   while (!current_row.eof() && !is_break_loop && !device::DataQueueMgr::GetInstance().IsClosed()) {
     while (!current_row.eoe() && !is_break_loop && !device::DataQueueMgr::GetInstance().IsClosed()) {
       RETURN_IF_NOT_OK(FilterMetadata(&current_row));

mindspore/ccsrc/minddata/mindrecord/io/shard_writer.cc

@@ -246,8 +246,8 @@ Status ShardWriter::SetHeaderSize(const uint64_t &header_size) {
   // header_size [16KB, 128MB]
   CHECK_FAIL_RETURN_UNEXPECTED_MR(header_size >= kMinHeaderSize && header_size <= kMaxHeaderSize,
                                   "Invalid data, header size: " + std::to_string(header_size) +
-                                    " should be in range [" + std::to_string(kMinHeaderSize) + "MB, " +
-                                    std::to_string(kMaxHeaderSize) + "MB].");
+                                    " should be in range [" + std::to_string(kMinHeaderSize) + " bytes, " +
+                                    std::to_string(kMaxHeaderSize) + " bytes].");
   CHECK_FAIL_RETURN_UNEXPECTED_MR(
     header_size % 4 == 0, "Invalid data, header size " + std::to_string(header_size) + " should be divided by four.");
   header_size_ = header_size;
@@ -258,7 +258,8 @@ Status ShardWriter::SetPageSize(const uint64_t &page_size) {
   // PageSize [32KB, 256MB]
   CHECK_FAIL_RETURN_UNEXPECTED_MR(page_size >= kMinPageSize && page_size <= kMaxPageSize,
                                   "Invalid data, page size: " + std::to_string(page_size) + " should be in range [" +
-                                    std::to_string(kMinPageSize) + "MB, " + std::to_string(kMaxPageSize) + "MB].");
+                                    std::to_string(kMinPageSize) + " bytes, " + std::to_string(kMaxPageSize) +
+                                    " bytes].");
   CHECK_FAIL_RETURN_UNEXPECTED_MR(
     page_size % 4 == 0, "Invalid data, page size " + std::to_string(page_size) + " should be divided by four.");
   page_size_ = page_size;
@@ -1135,7 +1136,9 @@ Status ShardWriter::SetRawDataSize(const std::vector<std::vector<uint8_t>> &bin_
   CHECK_FAIL_RETURN_SYNTAX_ERROR_MR(*std::max_element(raw_data_size_.begin(), raw_data_size_.end()) <= page_size_,
                                     "Invalid data, Page size: " + std::to_string(page_size_) +
                                       " is too small to save a raw row. Please try to use the mindrecord api "
-                                      "'set_page_size(1<<25)' to enable 64MB page size.");
+                                      "'set_page_size(value)' to enable larger page size, and the value range is in [" +
+                                      std::to_string(kMinPageSize) + " bytes, " + std::to_string(kMaxPageSize) +
+                                      " bytes].");
   return Status::OK();
 }
 
@@ -1146,7 +1149,9 @@ Status ShardWriter::SetBlobDataSize(const std::vector<std::vector<uint8_t>> &blo
   CHECK_FAIL_RETURN_SYNTAX_ERROR_MR(*std::max_element(blob_data_size_.begin(), blob_data_size_.end()) <= page_size_,
                                     "Invalid data, Page size: " + std::to_string(page_size_) +
                                       " is too small to save a blob row. Please try to use the mindrecord api "
-                                      "'set_page_size(1<<25)' to enable 64MB page size.");
+                                      "'set_page_size(value)' to enable larger page size, and the value range is in [" +
+                                      std::to_string(kMinPageSize) + " bytes, " + std::to_string(kMaxPageSize) +
+                                      " bytes].");
   return Status::OK();
 }
 

mindspore/python/mindspore/dataset/audio/transforms.py

@@ -1395,7 +1395,7 @@ class MelScale(AudioTensorOperation):
     Args:
         n_mels (int, optional): Number of mel filterbanks. Default: 128.
         sample_rate (int, optional): Sample rate of audio signal. Default: 16000.
-        f_min (float, optional): Minimum frequency. Default: 0.
+        f_min (float, optional): Minimum frequency. Default: 0.0.
         f_max (float, optional): Maximum frequency. Default: None, will be set to `sample_rate // 2` .
         n_stft (int, optional): Number of bins in STFT. Default: 201.
         norm (NormType, optional): Type of norm, value should be NormType.SLANEY or NormType::NONE.
@@ -1414,7 +1414,7 @@ class MelScale(AudioTensorOperation):
     """
 
     @check_mel_scale
-    def __init__(self, n_mels=128, sample_rate=16000, f_min=0, f_max=None, n_stft=201, norm=NormType.NONE,
+    def __init__(self, n_mels=128, sample_rate=16000, f_min=0.0, f_max=None, n_stft=201, norm=NormType.NONE,
                  mel_type=MelType.HTK):
         super().__init__()
         self.n_mels = n_mels

mindspore/python/mindspore/mindrecord/filewriter.py

@@ -322,7 +322,7 @@ class FileWriter:
         Examples:
             >>> from mindspore.mindrecord import FileWriter
             >>> writer = FileWriter(file_name="test.mindrecord", shard_num=1)
-            >>> status = writer.set_page_size(1 << 26) # 128MB
+            >>> status = writer.set_page_size(1 << 26)  # 64MB
         """
         return self._writer.set_page_size(page_size)