Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
train finetune
游雁96e4ff187065
diff --git a/examples/aishell/branchformer/conf/branchformer_12e_6d_2048_256.yaml b/examples/aishell/branchformer/conf/branchformer_12e_6d_2048_256.yamlindex d6caf5d27..d86e6281c 100644--- a/examples/aishell/branchformer/conf/branchformer_12e_6d_2048_256.yaml+++ b/examples/aishell/branchformer/conf/branchformer_12e_6d_2048_256.yaml@@ -108,7 +108,6 @@ dataset_conf:tokenizer: CharTokenizertokenizer_conf:unk_symbol: <unk>- split_with_space: truectc_conf:diff --git a/examples/aishell/conformer/conf/conformer_12e_6d_2048_256.yaml b/examples/aishell/conformer/conf/conformer_12e_6d_2048_256.yamlindex 8909b1596..4238d8462 100644--- a/examples/aishell/conformer/conf/conformer_12e_6d_2048_256.yaml+++ b/examples/aishell/conformer/conf/conformer_12e_6d_2048_256.yaml@@ -108,8 +108,6 @@ dataset_conf:tokenizer: CharTokenizertokenizer_conf:unk_symbol: <unk>- split_with_space: true-ctc_conf:dropout_rate: 0.0diff --git a/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml b/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yamlindex e6cc1e6c5..27734599b 100644--- a/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml+++ b/examples/aishell/e_branchformer/conf/e_branchformer_12e_6d_2048_256.yaml@@ -108,7 +108,6 @@ dataset_conf:tokenizer: CharTokenizertokenizer_conf:unk_symbol: <unk>- split_with_space: truectc_conf:diff --git a/examples/aishell/paraformer/conf/paraformer_conformer_12e_6d_2048_256.yaml b/examples/aishell/paraformer/conf/paraformer_conformer_12e_6d_2048_256.yamlindex dc2da1c59..c100f0da9 100644--- a/examples/aishell/paraformer/conf/paraformer_conformer_12e_6d_2048_256.yaml+++ b/examples/aishell/paraformer/conf/paraformer_conformer_12e_6d_2048_256.yaml@@ -107,8 +107,6 @@ dataset_conf:tokenizer: CharTokenizertokenizer_conf:unk_symbol: <unk>- split_with_space: false-ctc_conf:dropout_rate: 0.0diff --git a/examples/aishell/transformer/conf/transformer_12e_6d_2048_256.yaml b/examples/aishell/transformer/conf/transformer_12e_6d_2048_256.yamlindex 24a63903a..674fc94c2 100644--- a/examples/aishell/transformer/conf/transformer_12e_6d_2048_256.yaml+++ b/examples/aishell/transformer/conf/transformer_12e_6d_2048_256.yaml@@ -102,8 +102,6 @@ dataset_conf:tokenizer: CharTokenizertokenizer_conf:unk_symbol: <unk>- split_with_space: true-ctc_conf:dropout_rate: 0.0diff --git a/funasr/download/download_from_hub.py b/funasr/download/download_from_hub.pyindex c102549df..4a8e57a85 100644--- a/funasr/download/download_from_hub.py+++ b/funasr/download/download_from_hub.py@@ -21,10 +21,17 @@ def download_from_ms(**kwargs):model_or_path = get_or_download_model_dir(model_or_path, model_revision, is_training=kwargs.get("is_training"), check_latest=kwargs.get("kwargs", True))kwargs["model_path"] = model_or_path- config = os.path.join(model_or_path, "config.yaml")- if os.path.exists(config) and os.path.exists(os.path.join(model_or_path, "model.pb")):-- config = OmegaConf.load(config)+ if os.path.exists(os.path.join(model_or_path, "configuration.json")):+ with open(os.path.join(model_or_path, "configuration.json"), 'r', encoding='utf-8') as f:+ conf_json = json.load(f)+ cfg = {}+ add_file_root_path(model_or_path, conf_json["file_path_metas"], cfg)+ cfg.update(kwargs)+ config = OmegaConf.load(cfg["config"])+ kwargs = OmegaConf.merge(config, cfg)+ kwargs["model"] = config["model"]+ elif os.path.exists(os.path.join(model_or_path, "config.yaml")) and os.path.exists(os.path.join(model_or_path, "model.pt")):+ config = OmegaConf.load(os.path.join(model_or_path, "config.yaml"))kwargs = OmegaConf.merge(config, kwargs)init_param = os.path.join(model_or_path, "model.pb")kwargs["init_param"] = init_param@@ -41,15 +48,6 @@ def download_from_ms(**kwargs):kwargs["frontend_conf"]["cmvn_file"] = os.path.join(model_or_path, "am.mvn")if os.path.exists(os.path.join(model_or_path, "jieba_usr_dict")):kwargs["jieba_usr_dict"] = os.path.join(model_or_path, "jieba_usr_dict")- elif os.path.exists(os.path.join(model_or_path, "configuration.json")):- with open(os.path.join(model_or_path, "configuration.json"), 'r', encoding='utf-8') as f:- conf_json = json.load(f)- cfg = {}- add_file_root_path(model_or_path, conf_json["file_path_metas"], cfg)- cfg.update(kwargs)- config = OmegaConf.load(cfg["config"])- kwargs = OmegaConf.merge(config, cfg)- kwargs["model"] = config["model"]return OmegaConf.to_container(kwargs, resolve=True)def add_file_root_path(model_or_path: str, file_path_metas: dict, cfg = {}):