Remove dependencies to make more light weight

[BBC-Esq]

This PR removes perth, librosa, omegaconf, and conformer as dependencies.

If I did it correctly it supersedes these PRs:

#128
#129
#130
#133

I had a fifth PR that removed diffusers but it was too difficult to implement. A lot of people like myself try to minimize the amount of dependencies so hopefully this helps.

Dependencies like librosa and others have a lot of dependencies in turn...for example, I think that resampy is no longer required either. So overall this PR probably removes at least 10+ total dependencies. I tested and it works fine still.

[Datel777]

This is very good idea to reduce dependencies, especially because torchaudio has almost same methods.
Need to add into pyproject.toml dependencies:

"soundfile",
"gradio",

Unfortunately, running tts conversion with default reference makes errors:

E:\AI\chatterbox-light\venv\lib\site-packages\torchaudio\functional\functional.py:584: UserWarning: At least one mel filterbank has all zero values. The value for `n_mels` (128) may be set too high. Or, the value for `n_freqs` (201) may be set too low.
  warnings.warn(
E:\AI\chatterbox-light\venv\lib\site-packages\diffusers\models\lora.py:393: FutureWarning: `LoRACompatibleLinear` is deprecated and will be removed in version 1.0.0. Use of `LoRACompatibleLinear` is deprecated. Please switch to PEFT backend by installing PEFT: `pip install peft`.
  deprecate("LoRACompatibleLinear", "1.0.0", deprecation_message)
D:\Python\Python310\lib\contextlib.py:103: FutureWarning: `torch.backends.cuda.sdp_kernel()` is deprecated. In the future, this context manager will be removed. Please see `torch.nn.attention.sdpa_kernel()` for the new context manager, with updated signature.
  self.gen = func(*args, **kwds)
E:\AI\chatterbox-light\venv\lib\site-packages\transformers\utils\generic.py:1006: UserWarning: `output_attentions=True` is not supported with `attn_implementation` other than ['eager', 'eager_paged', 'flex_attention']. Please use `model.set_attn_implementation('eager')` to enable capturing attention outputs.
  warnings.warn(
`sdpa` attention does not support `output_attentions=True` or `head_mask`. Please set your attention to `eager` if you want any of these features.
Traceback (most recent call last):
  File "E:\AI\chatterbox-light\venv\lib\site-packages\gradio\queueing.py", line 759, in process_events
    response = await route_utils.call_process_api(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\gradio\route_utils.py", line 354, in call_process_api
    output = await app.get_blocks().process_api(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\gradio\blocks.py", line 2116, in process_api
    result = await self.call_function(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\gradio\blocks.py", line 1623, in call_function
    prediction = await anyio.to_thread.run_sync(  # type: ignore
  File "E:\AI\chatterbox-light\venv\lib\site-packages\anyio\to_thread.py", line 56, in run_sync
    return await get_async_backend().run_sync_in_worker_thread(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 2485, in run_sync_in_worker_thread
    return await future
  File "E:\AI\chatterbox-light\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 976, in run
    result = context.run(func, *args)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\gradio\utils.py", line 915, in wrapper
    response = f(*args, **kwargs)
  File "E:\AI\chatterbox-light\gradio_tts_app.py", line 31, in generate
    wav = model.generate(
  File "E:\AI\chatterbox-light\src\chatterbox\tts.py", line 245, in generate
    speech_tokens = self.t3.inference(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "E:\AI\chatterbox-light\src\chatterbox\models\t3\t3.py", line 318, in inference
    output = self.patched_model(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "E:\AI\chatterbox-light\src\chatterbox\models\t3\inference\t3_hf_backend.py", line 95, in forward
    tfmr_out = self.model(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\transformers\utils\generic.py", line 1064, in wrapper
    outputs = func(self, *args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\transformers\models\llama\modeling_llama.py", line 395, in forward
    hidden_states = decoder_layer(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\transformers\modeling_layers.py", line 94, in __call__
    return super().__call__(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\transformers\utils\generic.py", line 1023, in wrapped_forward
    output = orig_forward(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\transformers\utils\deprecation.py", line 172, in wrapped_func
    return func(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\transformers\models\llama\modeling_llama.py", line 294, in forward
    hidden_states, _ = self.self_attn(
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1857, in _call_impl
    return inner()
  File "E:\AI\chatterbox-light\venv\lib\site-packages\torch\nn\modules\module.py", line 1818, in inner
    hook_result = hook(self, args, result)
  File "E:\AI\chatterbox-light\src\chatterbox\models\t3\inference\alignment_stream_analyzer.py", line 74, in attention_forward_hook
    step_attention = output[1].cpu() # (B, 16, N, N)
AttributeError: 'NoneType' object has no attribute 'cpu'
E:\AI\chatterbox-light\venv\lib\site-packages\torchaudio\functional\functional.py:584: UserWarning: At least one mel filterbank has all zero values. The value for `n_mels` (128) may be set too high. Or, the value for `n_freqs` (201) may be set too low.
  warnings.warn(
E:\AI\chatterbox-light\venv\lib\site-packages\diffusers\models\lora.py:393: FutureWarning: `LoRACompatibleLinear` is deprecated and will be removed in version 1.0.0. Use of `LoRACompatibleLinear` is deprecated. Please switch to PEFT backend by installing PEFT: `pip install peft`.
  deprecate("LoRACompatibleLinear", "1.0.0", deprecation_message)

Not sure what to do with this error:

UserWarning: At least one mel filterbank has all zero values. The value for `n_mels` (128) may be set too high. Or, the value for `n_freqs` (201) may be set too low.

I tried to change n_fft, num_mels, sampling_rate, for ta.functional.melscale_fbanks function, but with no luck.

Edit: After some refactoring figured out some version difference, but anyway has different error when replacing librosa to torchaudio

E:\AI\chatterbox-custom\venv\Lib\site-packages\torchaudio\functional\functional.py:584: UserWarning: At least one mel filterbank has all zero values. The value for `n_mels` (128) may be set too high. Or, the value for `n_freqs` (201) may be set too low.
  warnings.warn(
E:\AI\chatterbox-custom\venv\Lib\site-packages\diffusers\models\lora.py:393: FutureWarning: `LoRACompatibleLinear` is deprecated and will be removed in version 1.0.0. Use of `LoRACompatibleLinear` is deprecated. Please switch to PEFT backend by installing PEFT: `pip install peft`.
  deprecate("LoRACompatibleLinear", "1.0.0", deprecation_message)
WARNING:root:Reference mel length is not equal to 2 * reference token length.

Traceback (most recent call last):
  File "E:\AI\chatterbox-custom\fast_test.py", line 13, in <module>
    model.prepare_conditionals(wav_fpath=AUDIO_PROMPT_PATH)  # Use custom voice from audio sample
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\src\chatterbox\tts.py", line 207, in prepare_conditionals
    ve_embed = torch.from_numpy(self.ve.embeds_from_wavs([ref_16k_wav], sample_rate=S3_SR))
                                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\src\chatterbox\models\voice_encoder\voice_encoder.py", line 286, in embeds_from_wavs
    mels = [melspectrogram(w, self.hp).T for w in wavs]
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\src\chatterbox\models\voice_encoder\voice_encoder.py", line 286, in <listcomp>
    mels = [melspectrogram(w, self.hp).T for w in wavs]
            ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\src\chatterbox\models\voice_encoder\melspec.py", line 44, in melspectrogram
    spec_complex = _stft(wav, hp, pad=pad)
                   ^^^^^^^^^^^^^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\src\chatterbox\models\voice_encoder\melspec.py", line 77, in _stft
    return torch.stft(
           ^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\venv\Lib\site-packages\torch\functional.py", line 728, in stft
    input = F.pad(input.view(extended_shape), [pad, pad], pad_mode)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "E:\AI\chatterbox-custom\venv\Lib\site-packages\torch\nn\functional.py", line 5209, in pad
    return torch._C._nn.pad(input, pad, mode, value)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Expected 2D or 3D (batch mode) tensor with possibly 0 batch size and other non-zero dimensions for input, but got: [1, 1, 0]

[BBC-Esq]

@Datel777 I will check that out and let you know.