Wrong ASR with demo code

by anyshu - opened 10 days ago
10 days ago
•
Wrong ASR with demo code on 5090-CUDA13.0 :
(funasr) root@sh-b1-prod-ai5090-124-10:/data/app/Fun-ASR# CUDA_VISIBLE_DEVICES=1 python funasr-nano.py 
funasr version: 1.2.9.
Check update of funasr, and it would cost few times. You may disable it by set `disable_update=True` in AutoModel
You are using the latest version of funasr-1.2.9
WARNING:root:trust_remote_code: True
Loading remote code successfully: ./model.py
funasr version: 1.2.9.
Check update of funasr, and it would cost few times. You may disable it by set `disable_update=True` in AutoModel
You are using the latest version of funasr-1.2.9
Downloading Model from https://www.modelscope.cn to directory: /root/.cache/modelscope/hub/models/iic/SenseVoiceSmall
WARNING:root:trust_remote_code: False
Warning, miss key in ckpt: llm.model.embed_tokens.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.q_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.k_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.v_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.o_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.q_norm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.k_norm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.mlp.gate_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.mlp.up_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
……
Warning, miss key in ckpt: llm.model.layers.27.self_attn.k_norm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.27.mlp.gate_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.27.mlp.up_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.27.mlp.down_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.27.input_layernorm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.27.post_attention_layernorm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.norm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.lm_head.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
  0%|                                                                                                                                                                | 0/1 [00:00<?, ?it/s]The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Setting `pad_token_id` to `eos_token_id`:151645 for open-end generation.
The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
rtf_avg: 1.412: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:07<00:00,  7.97s/it]
añaaña Expires谱写谱写谱写谱写谱写谱写谱写谱写 desenv_CUSTOMER_CUSTOMER_CUSTOMER_CUSTOMER_CUSTOMER_CUSTOMERaña Ridley_CUSTOMERañaaña vibe 중심uanccd Ridley 중심uanhasilañaañaaña Ridley 중심 중심uanaña 중심 Quận 중심 Quận 중심 중심 중심uanaña 중심 중심 중심 중심 중심 중심 Ridley 중심 Ridley 중심 Ridley 중심aña 중심aña 중심aña 중심aña 중심aña 중심aña 중심añaañaaña 중심 중심 중심aña 중심 중심 중심añaañaañaaña 중심aña 중심añaaña 중심aña 중심aña 중심aña 중심aña 중심añaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaañaaña.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play.play.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play аб.play.play.play аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer Bayer.play аб Bayer drapedนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะ Bayer drapedนักเตะ Bayer Bayer Bayer Bayer Bayer CE.play аб Bayer Bayer Bayer Bayer Bayer Bayer.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play.play абนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะ Bayeradas Bayeradas Bayeradas Bayer Bayer Bayer Bayer Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayeradas Bayer
funasr version: 1.2.9.
Check update of funasr, and it would cost few times. You may disable it by set `disable_update=True` in AutoModel
You are using the latest version of funasr-1.2.9
WARNING:root:trust_remote_code: True
Loading remote code successfully: ./model.py
funasr version: 1.2.9.
Check update of funasr, and it would cost few times. You may disable it by set `disable_update=True` in AutoModel
You are using the latest version of funasr-1.2.9
Downloading Model from https://www.modelscope.cn to directory: /root/.cache/modelscope/hub/models/iic/SenseVoiceSmall
WARNING:root:trust_remote_code: False
Warning, miss key in ckpt: llm.model.embed_tokens.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.q_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.k_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.v_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.o_proj.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.model.layers.0.self_attn.q_norm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
……
Warning, miss key in ckpt: llm.model.norm.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Warning, miss key in ckpt: llm.lm_head.weight, /data/models/FunAudioLLM/Fun-ASR-Nano-2512/model.pt
Downloading Model from https://www.modelscope.cn to directory: /root/.cache/modelscope/hub/models/iic/speech_fsmn_vad_zh-cn-16k-common-pytorch
WARNING:root:trust_remote_code: False
rtf_avg: 0.025: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00,  6.29it/s]
  0%|                                                                                                                                                                | 0/1 [00:00<?, ?it/sThe attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.<?, ?it/s]
Setting `pad_token_id` to `eos_token_id`:151645 for open-end generation.
rtf_avg: 1.517: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:07<00:00,  7.83s/it]
rtf_avg: 1.396, time_speech:  5.616, time_escape: 7.840: 100%|███████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:07<00:00,  7.96s/it]
shipment谱写谱写谱写谱写谱写谱写🕷rians vibe vibe vibe vibe كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافة كافةaña Bayer 중심国家标准 Bayer 중심国家标准 Bayer 중심国家标准 asia draped国家标准 asia draped draped draped draped draped国家标准練 asia draped Iterable練練練 중심 draped国家标准練 중심 draped国家标准 posix draped国家标准国家标准 Expires syll ExpiresAnal draped国家标准練 draped国家标准練練練練練練練練練 draped видео抱歉 draped国家标准練 draped抱歉 ExpiresAnal draped抱歉 draped抱歉 draped抱歉 draped抱歉 draped抱歉 draped冼 nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Roles(Rolesนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะ nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop nanop аб흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐흐 nanop аб nanop аб nanop аб nanop аб nanop аб흐흐흐흐흐 nanop аб nanop аб nanop аб nanop аб흐 nanop аб nanopนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะนักเตะتي аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб аб
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
· Sign up or log in to comment