Update model.py
Browse files
model.py
CHANGED
|
@@ -1378,25 +1378,7 @@ def _get_sherpa_onnx_nemo_transducer_models(
|
|
| 1378 |
|
| 1379 |
return recognizer
|
| 1380 |
|
| 1381 |
-
|
| 1382 |
-
def _get_nemo_ctc_models(repo_id: str, decoding_method: str, num_active_paths: int):
|
| 1383 |
-
nn_model = _get_nn_model_filename(
|
| 1384 |
-
repo_id=repo_id,
|
| 1385 |
-
filename="model.int8.onnx",
|
| 1386 |
-
subfolder=".",
|
| 1387 |
-
)
|
| 1388 |
-
tokens = _get_token_filename(
|
| 1389 |
-
repo_id=repo_id,
|
| 1390 |
-
filename="tokens.txt",
|
| 1391 |
-
subfolder=".",
|
| 1392 |
-
)
|
| 1393 |
-
recognizer = sherpa_onnx.OfflineRecognizer.from_nemo_ctc(
|
| 1394 |
-
tokens=tokens,
|
| 1395 |
-
model=nn_model,
|
| 1396 |
-
num_threads=2,
|
| 1397 |
-
language="es", # Force Spanish
|
| 1398 |
-
)
|
| 1399 |
-
return recognizer
|
| 1400 |
|
| 1401 |
@lru_cache(maxsize=10)
|
| 1402 |
def _get_sherpa_onnx_nemo_ctc_models(
|
|
@@ -1410,29 +1392,27 @@ def _get_sherpa_onnx_nemo_ctc_models(
|
|
| 1410 |
"csukuangfj/sherpa-onnx-nemo-stt_pt_fastconformer_hybrid_large_pc-int8",
|
| 1411 |
"csukuangfj/sherpa-onnx-nemo-stt_de_fastconformer_hybrid_large_pc",
|
| 1412 |
"csukuangfj/sherpa-onnx-nemo-stt_de_fastconformer_hybrid_large_pc-int8",
|
|
|
|
| 1413 |
], repo_id
|
| 1414 |
|
| 1415 |
-
if "int8" in repo_id
|
| 1416 |
-
model = _get_nn_model_filename(
|
| 1417 |
-
repo_id=repo_id,
|
| 1418 |
-
filename="model.int8.onnx",
|
| 1419 |
-
subfolder=".",
|
| 1420 |
-
)
|
| 1421 |
-
else:
|
| 1422 |
-
model = _get_nn_model_filename(
|
| 1423 |
-
repo_id=repo_id,
|
| 1424 |
-
filename="model.onnx",
|
| 1425 |
-
subfolder=".",
|
| 1426 |
-
)
|
| 1427 |
|
| 1428 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1429 |
|
| 1430 |
recognizer = sherpa_onnx.OfflineRecognizer.from_nemo_ctc(
|
| 1431 |
tokens=tokens,
|
| 1432 |
model=model,
|
| 1433 |
num_threads=2,
|
| 1434 |
-
sample_rate=16000,
|
| 1435 |
-
feature_dim=80,
|
| 1436 |
)
|
| 1437 |
|
| 1438 |
return recognizer
|
|
@@ -2089,8 +2069,7 @@ multi_lingual_models = {
|
|
| 2089 |
"csukuangfj/sherpa-onnx-dolphin-small-ctc-multi-lang-int8-2025-04-02": _get_dolphin_ctc_models,
|
| 2090 |
"csukuangfj/sherpa-onnx-dolphin-base-ctc-multi-lang-2025-04-02": _get_dolphin_ctc_models,
|
| 2091 |
"csukuangfj/sherpa-onnx-dolphin-small-ctc-multi-lang-2025-04-02": _get_dolphin_ctc_models,
|
| 2092 |
-
"csukuangfj/sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8":
|
| 2093 |
-
"csukuangfj/sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8": _get_nemo_ctc_models,
|
| 2094 |
}
|
| 2095 |
|
| 2096 |
chinese_english_mixed_models = {
|
|
|
|
| 1378 |
|
| 1379 |
return recognizer
|
| 1380 |
|
| 1381 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1382 |
|
| 1383 |
@lru_cache(maxsize=10)
|
| 1384 |
def _get_sherpa_onnx_nemo_ctc_models(
|
|
|
|
| 1392 |
"csukuangfj/sherpa-onnx-nemo-stt_pt_fastconformer_hybrid_large_pc-int8",
|
| 1393 |
"csukuangfj/sherpa-onnx-nemo-stt_de_fastconformer_hybrid_large_pc",
|
| 1394 |
"csukuangfj/sherpa-onnx-nemo-stt_de_fastconformer_hybrid_large_pc-int8",
|
| 1395 |
+
"csukuangfj/sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8",
|
| 1396 |
], repo_id
|
| 1397 |
|
| 1398 |
+
filename = "model.int8.onnx" if "int8" in repo_id else "model.onnx"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1399 |
|
| 1400 |
+
model = _get_nn_model_filename(
|
| 1401 |
+
repo_id=repo_id,
|
| 1402 |
+
filename=filename,
|
| 1403 |
+
subfolder=".",
|
| 1404 |
+
)
|
| 1405 |
+
|
| 1406 |
+
tokens = _get_token_filename(
|
| 1407 |
+
repo_id=repo_id,
|
| 1408 |
+
filename="tokens.txt",
|
| 1409 |
+
subfolder=".",
|
| 1410 |
+
)
|
| 1411 |
|
| 1412 |
recognizer = sherpa_onnx.OfflineRecognizer.from_nemo_ctc(
|
| 1413 |
tokens=tokens,
|
| 1414 |
model=model,
|
| 1415 |
num_threads=2,
|
|
|
|
|
|
|
| 1416 |
)
|
| 1417 |
|
| 1418 |
return recognizer
|
|
|
|
| 2069 |
"csukuangfj/sherpa-onnx-dolphin-small-ctc-multi-lang-int8-2025-04-02": _get_dolphin_ctc_models,
|
| 2070 |
"csukuangfj/sherpa-onnx-dolphin-base-ctc-multi-lang-2025-04-02": _get_dolphin_ctc_models,
|
| 2071 |
"csukuangfj/sherpa-onnx-dolphin-small-ctc-multi-lang-2025-04-02": _get_dolphin_ctc_models,
|
| 2072 |
+
"csukuangfj/sherpa-onnx-nemo-canary-180m-flash-en-es-de-fr-int8": _get_sherpa_onnx_nemo_ctc_models,
|
|
|
|
| 2073 |
}
|
| 2074 |
|
| 2075 |
chinese_english_mixed_models = {
|