Q and A not working for Youtube

[anushaharish538]

Hi,

when I add YouTube video url and click on ingest it says docs:1 chunks:3, but when I ask any question about the video its connecting to internet and giving answer. It's working fine for pdf.

Regards,
Anusha

[pseudotensor]

Review the document viewer and what is in the database. Nominally for youtube video, will get both 10 frames + audio transcription.

What does the console say when you upload a video? It should show lots of details if you ran in --verbose mode.

[anushaharish538]

@pseudotensor Thank you for the reply. I'm using CPU. will the YouTube chat works with CPU?

I'm getting the following error.
Traceback (most recent call last):
File "/home/anushaharish538/as/h2ogpt/src/gpt_langchain.py", line 4034, in file_to_doc
docs1c = model_loaders['asr'].load(from_youtube=True)
File "/home/anushaharish538/as/h2ogpt/src/audio_langchain.py", line 387, in load
self.load_model()
File "/home/anushaharish538/as/h2ogpt/src/audio_langchain.py", line 368, in load_model
self.model = OpenAIWhisperParserLocal(device=self.device,
File "/home/anushaharish538/as/h2ogpt/src/audio_langchain.py", line 183, in init
self.pipe = pipeline(
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/transformers/pipelines/init.py", line 906, in pipeline
framework, model = infer_framework_load_model(
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/transformers/pipelines/base.py", line 283, in infer_framework_load_model
model = model_class.from_pretrained(model, **kwargs)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 563, in from_pretrained
return model_class.from_pretrained(
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/transformers/modeling_utils.py", line 3709, in from_pretrained
check_tied_parameters_on_same_device(tied_params, device_map)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/accelerate/utils/modeling.py", line 598, in check_tied_parameters_on_same_device
tie_param_devices[param] = _get_param_device(param, device_map)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/accelerate/utils/modeling.py", line 580, in _get_param_device
return _get_param_device(parent_param, device_map)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/accelerate/utils/modeling.py", line 580, in _get_param_device
return _get_param_device(parent_param, device_map)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/accelerate/utils/modeling.py", line 580, in _get_param_device
return _get_param_device(parent_param, device_map)
[Previous line repeated 1 more time]
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/accelerate/utils/modeling.py", line 575, in _get_param_device
return device_map[param]
TypeError: 'set' object is not subscriptable
ASR: 'set' object is not subscriptable: None
Failed to ingest https://www.youtube.com/watch?v=U1JMy1LTSu8 due to Traceback (most recent call last):
File "/home/anushaharish538/as/h2ogpt/src/gpt_langchain.py", line 4853, in path_to_doc1
res = file_to_doc(file,
File "/home/anushaharish538/as/h2ogpt/src/gpt_langchain.py", line 4081, in file_to_doc
raise ValueError("%s had no valid text and no meta data was parsed: %s" % (file, str(e)))
ValueError: https://www.youtube.com/watch?v=U1JMy1LTSu8 had no valid text and no meta data was parsed: 'set' object is not subscriptable

[pseudotensor]

Try setting --asr_gpu=False

[anushaharish538]

Thank You for the help. I tried but still same issue.
I'm using open ai
OPENAI_API_KEY=Key python generate.py --base_model=gpt-4o --prompt_type=mixtral --inference_server=openai_chat --score_model=None --append_sources_to_answer=True --langchain_mode=UserData --asr_gpu=False

[anushaharish538]

@pseudotensor I think its issue with database . can you please confirm. I'm getting python src/make_db.pyExceptions: 0/0 []Traceback (most recent call last): File "/home/anushaharish538/as/h2ogpt/src/make_db.py", line 403, in
H2O_Fire(make_db_main)
File "/home/anushaharish538/as/h2ogpt/src/utils.py", line 73, in H2O_Fire
fire.Fire(component=component, command=args)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/fire/core.py", line 141, in Fire
component_trace = _Fire(component, args, parsed_flag_args, context, name)
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/fire/core.py", line 475, in _Fire
component, remaining_args = _CallAndUpdateTrace(
File "/home/anushaharish538/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/fire/core.py", line 691, in _CallAndUpdateTrace
component = fn(*varargs, **kwargs)
File "/home/anushaharish538/as/h2ogpt/src/make_db.py", line 389, in make_db_main
assert len(sources) > 0 or not fail_if_no_sources, "No sources found"
AssertionError: No sources found

It's issue with only YouTube . Pdfs working good

[anushaharish538]

@pseudotensor can you please tell me , will the YouTube chat works with CPU?

Thank you.

[pseudotensor]

asr_gpu -> False already if no GPUs, so it's not that.