From c36039ce322f45e38ed1a9094994b7ff6572975b Mon Sep 17 00:00:00 2001
From: TianYuan <white-sky@qq.com>
Date: Tue, 22 Mar 2022 08:45:22 +0000
Subject: [PATCH] update readme for ljspeech hifigan, test=tts

---
 examples/ljspeech/tts3/local/synthesize.sh    | 53 ++++++++++++-----
 .../ljspeech/tts3/local/synthesize_e2e.sh     | 58 ++++++++++++++-----
 2 files changed, 81 insertions(+), 30 deletions(-)

diff --git a/examples/ljspeech/tts3/local/synthesize.sh b/examples/ljspeech/tts3/local/synthesize.sh
index f150d158..6dc34274 100755
--- a/examples/ljspeech/tts3/local/synthesize.sh
+++ b/examples/ljspeech/tts3/local/synthesize.sh
@@ -4,17 +4,42 @@ config_path=$1
 train_output_path=$2
 ckpt_name=$3
 
-FLAGS_allocator_strategy=naive_best_fit \
-FLAGS_fraction_of_gpu_memory_to_use=0.01 \
-python3 ${BIN_DIR}/../synthesize.py \
-    --am=fastspeech2_ljspeech \
-    --am_config=${config_path} \
-    --am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
-    --am_stat=dump/train/speech_stats.npy \
-    --voc=pwgan_ljspeech \
-    --voc_config=pwg_ljspeech_ckpt_0.5/pwg_default.yaml \
-    --voc_ckpt=pwg_ljspeech_ckpt_0.5/pwg_snapshot_iter_400000.pdz  \
-    --voc_stat=pwg_ljspeech_ckpt_0.5/pwg_stats.npy \
-    --test_metadata=dump/test/norm/metadata.jsonl \
-    --output_dir=${train_output_path}/test \
-    --phones_dict=dump/phone_id_map.txt
+stage=0
+stop_stage=0
+
+# pwgan
+if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
+    FLAGS_allocator_strategy=naive_best_fit \
+    FLAGS_fraction_of_gpu_memory_to_use=0.01 \
+    python3 ${BIN_DIR}/../synthesize.py \
+        --am=fastspeech2_ljspeech \
+        --am_config=${config_path} \
+        --am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
+        --am_stat=dump/train/speech_stats.npy \
+        --voc=pwgan_ljspeech \
+        --voc_config=pwg_ljspeech_ckpt_0.5/pwg_default.yaml \
+        --voc_ckpt=pwg_ljspeech_ckpt_0.5/pwg_snapshot_iter_400000.pdz  \
+        --voc_stat=pwg_ljspeech_ckpt_0.5/pwg_stats.npy \
+        --test_metadata=dump/test/norm/metadata.jsonl \
+        --output_dir=${train_output_path}/test \
+        --phones_dict=dump/phone_id_map.txt
+fi
+
+# hifigan
+if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
+    FLAGS_allocator_strategy=naive_best_fit \
+    FLAGS_fraction_of_gpu_memory_to_use=0.01 \
+    python3 ${BIN_DIR}/../synthesize.py \
+        --am=fastspeech2_ljspeech \
+        --am_config=${config_path} \
+        --am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
+        --am_stat=dump/train/speech_stats.npy \
+        --voc=hifigan_ljspeech \
+        --voc_config=hifigan_ljspeech_ckpt_0.2.0/default.yaml \
+        --voc_ckpt=hifigan_ljspeech_ckpt_0.2.0/snapshot_iter_2500000.pdz \
+        --voc_stat=hifigan_ljspeech_ckpt_0.2.0/feats_stats.npy \
+        --test_metadata=dump/test/norm/metadata.jsonl \
+        --output_dir=${train_output_path}/test \
+        --phones_dict=dump/phone_id_map.txt
+fi
+
diff --git a/examples/ljspeech/tts3/local/synthesize_e2e.sh b/examples/ljspeech/tts3/local/synthesize_e2e.sh
index 0b0cb574..36865f7f 100755
--- a/examples/ljspeech/tts3/local/synthesize_e2e.sh
+++ b/examples/ljspeech/tts3/local/synthesize_e2e.sh
@@ -4,19 +4,45 @@ config_path=$1
 train_output_path=$2
 ckpt_name=$3
 
-FLAGS_allocator_strategy=naive_best_fit \
-FLAGS_fraction_of_gpu_memory_to_use=0.01 \
-python3 ${BIN_DIR}/../synthesize_e2e.py \
-    --am=fastspeech2_ljspeech \
-    --am_config=${config_path} \
-    --am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
-    --am_stat=dump/train/speech_stats.npy \
-    --voc=pwgan_ljspeech \
-    --voc_config=pwg_ljspeech_ckpt_0.5/pwg_default.yaml \
-    --voc_ckpt=pwg_ljspeech_ckpt_0.5/pwg_snapshot_iter_400000.pdz  \
-    --voc_stat=pwg_ljspeech_ckpt_0.5/pwg_stats.npy \
-    --lang=en \
-    --text=${BIN_DIR}/../sentences_en.txt \
-    --output_dir=${train_output_path}/test_e2e \
-    --inference_dir=${train_output_path}/inference \
-    --phones_dict=dump/phone_id_map.txt
\ No newline at end of file
+stage=0
+stop_stage=0
+
+# pwgan
+if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
+    FLAGS_allocator_strategy=naive_best_fit \
+    FLAGS_fraction_of_gpu_memory_to_use=0.01 \
+    python3 ${BIN_DIR}/../synthesize_e2e.py \
+        --am=fastspeech2_ljspeech \
+        --am_config=${config_path} \
+        --am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
+        --am_stat=dump/train/speech_stats.npy \
+        --voc=pwgan_ljspeech \
+        --voc_config=pwg_ljspeech_ckpt_0.5/pwg_default.yaml \
+        --voc_ckpt=pwg_ljspeech_ckpt_0.5/pwg_snapshot_iter_400000.pdz  \
+        --voc_stat=pwg_ljspeech_ckpt_0.5/pwg_stats.npy \
+        --lang=en \
+        --text=${BIN_DIR}/../sentences_en.txt \
+        --output_dir=${train_output_path}/test_e2e \
+        --inference_dir=${train_output_path}/inference \
+        --phones_dict=dump/phone_id_map.txt
+fi
+
+# hifigan
+if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
+    FLAGS_allocator_strategy=naive_best_fit \
+    FLAGS_fraction_of_gpu_memory_to_use=0.01 \
+    python3 ${BIN_DIR}/../synthesize_e2e.py \
+        --am=fastspeech2_ljspeech \
+        --am_config=${config_path} \
+        --am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
+        --am_stat=dump/train/speech_stats.npy \
+        --voc=hifigan_ljspeech \
+        --voc_config=hifigan_ljspeech_ckpt_0.2.0/default.yaml \
+        --voc_ckpt=hifigan_ljspeech_ckpt_0.2.0/snapshot_iter_2500000.pdz \
+        --voc_stat=hifigan_ljspeech_ckpt_0.2.0/feats_stats.npy \
+        --lang=en \
+        --text=${BIN_DIR}/../sentences_en.txt \
+        --output_dir=${train_output_path}/test_e2e \
+        --inference_dir=${train_output_path}/inference \
+        --phones_dict=dump/phone_id_map.txt
+fi
-- 
GitLab