Implement max_symbols_per_frame for GigaAM2 accurate decoding since model uses c…

…har tokens instead of BPE. (#2423)

@@ -45,6 +45,7 @@ static OfflineTransducerDecoderResult DecodeOne(
 
   int32_t vocab_size = model->VocabSize();
   int32_t blank_id = vocab_size - 1;
+  int32_t max_symbols_per_frame = 10;
 
   auto decoder_input_pair = BuildDecoderInput(blank_id, model->Allocator());
 
@@ -60,7 +61,8 @@ static OfflineTransducerDecoderResult DecodeOne(
         memory_info, const_cast<float *>(p) + t * num_cols, num_cols,
         encoder_shape.data(), encoder_shape.size());
 
-    Ort::Value logit = model->RunJoiner(std::move(cur_encoder_out),
+    for (int32_t q = 0; q != max_symbols_per_frame; ++q) {
+        Ort::Value logit = model->RunJoiner(View(&cur_encoder_out),
                                             View(&decoder_output_pair.first));
 
         float *p_logit = logit.GetTensorMutableData<float>();
@@ -83,7 +85,10 @@ static OfflineTransducerDecoderResult DecodeOne(
               model->RunDecoder(std::move(decoder_input_pair.first),
                                 std::move(decoder_input_pair.second),
                                 std::move(decoder_output_pair.second));
+        } else {
+           break;
         } // if (y != blank_id)
+    }
   }    // for (int32_t i = 0; i != num_rows; ++i)
 
   return ans;
@@ -99,7 +104,7 @@ OfflineTransducerGreedySearchNeMoDecoder::Decode(
   int32_t dim1 = static_cast<int32_t>(shape[1]);
   int32_t dim2 = static_cast<int32_t>(shape[2]);
 
-  const int64_t *p_length = encoder_out_length.GetTensorData<int64_t>();
+  const int32_t *p_length = encoder_out_length.GetTensorData<int32_t>();
   const float *p = encoder_out.GetTensorData<float>();
 
   std::vector<OfflineTransducerDecoderResult> ans(batch_size);