app.py

import torch
from peft import PeftModel
import transformers
import gradio as gr
import os
os.system('pip install voicefixer --upgrade')
from voicefixer import VoiceFixer
voicefixer = VoiceFixer()

from TTS.api import TTS
tts = TTS(model_name="tts_models/multilingual/multi-dataset/your_tts", progress_bar=False, gpu=True)

import whisper
model1 = whisper.load_model("small")

import torchaudio
from speechbrain.pretrained import SpectralMaskEnhancement

enhance_model = SpectralMaskEnhancement.from_hparams(
source="speechbrain/metricgan-plus-voicebank",
savedir="pretrained_models/metricgan-plus-voicebank",
run_opts={"device":"cuda"},
)

assert (
    "LlamaTokenizer" in transformers._import_structure["models.llama"]
), "LLaMA is now in HuggingFace's main branch.\nPlease reinstall it: pip uninstall transformers && pip install git+https://github.com/huggingface/transformers.git"
from transformers import LlamaTokenizer, LlamaForCausalLM, GenerationConfig

tokenizer = LlamaTokenizer.from_pretrained("decapoda-research/llama-7b-hf")

BASE_MODEL = "decapoda-research/llama-7b-hf"
LORA_WEIGHTS = "tloen/alpaca-lora-7b"

if torch.cuda.is_available():
    device = "cuda"
else:
    device = "cpu"

try:
    if torch.backends.mps.is_available():
        device = "mps"
except:
    pass

if device == "cuda":
    model = LlamaForCausalLM.from_pretrained(
        BASE_MODEL,
        load_in_8bit=False,
        torch_dtype=torch.float16,
        device_map="auto",
    )
    model = PeftModel.from_pretrained(
        model, LORA_WEIGHTS, torch_dtype=torch.float16, force_download=True
    )
elif device == "mps":
    model = LlamaForCausalLM.from_pretrained(
        BASE_MODEL,
        device_map={"": device},
        torch_dtype=torch.float16,
    )
    model = PeftModel.from_pretrained(
        model,
        LORA_WEIGHTS,
        device_map={"": device},
        torch_dtype=torch.float16,
    )
else:
    model = LlamaForCausalLM.from_pretrained(
        BASE_MODEL, device_map={"": device}, low_cpu_mem_usage=True
    )
    model = PeftModel.from_pretrained(
        model,
        LORA_WEIGHTS,
        device_map={"": device},
    )


def generate_prompt(instruction, input=None):
    if input:
        return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.

### Instruction:
{instruction}

### Input:
{input}

### Response:"""
    else:
        return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.

### Instruction:
{instruction}

### Response:"""

if device != "cpu":
    model.half()
model.eval()
if torch.__version__ >= "2":
    model = torch.compile(model)


def evaluate(
#    instruction,
    audio,
    upload,
    input=None,
    temperature=0.1,
    top_p=0.75,
    top_k=40,
    num_beams=4,
    max_new_tokens=128,
    **kwargs,
):

    # load audio and pad/trim it to fit 30 seconds
    audio = whisper.load_audio(audio)
    audio = whisper.pad_or_trim(audio)

    # make log-Mel spectrogram and move to the same device as the model1
    mel = whisper.log_mel_spectrogram(audio).to(model1.device)

    # detect the spoken language
    _, probs = model1.detect_language(mel)
    print(f"Detected language: {max(probs, key=probs.get)}")

    # decode the audio
    options = whisper.DecodingOptions()
    result = whisper.decode(model1, mel, options)

    instruction = result.text
    
    prompt = generate_prompt(instruction, input)
    inputs = tokenizer(prompt, return_tensors="pt")
    input_ids = inputs["input_ids"].to(device)
    generation_config = GenerationConfig(
        temperature=temperature,
        top_p=top_p,
        top_k=top_k,
        num_beams=num_beams,
        **kwargs,
    )
    with torch.no_grad():
        generation_output = model.generate(
            input_ids=input_ids,
            generation_config=generation_config,
            return_dict_in_generate=True,
            output_scores=True,
            max_new_tokens=max_new_tokens,
        )
    s = generation_output.sequences[0]
    output = tokenizer.decode(s)

    tts.tts_to_file(output.split("### Response:")[1].strip(), speaker_wav = upload, language="en", file_path="output.wav")
    
    voicefixer.restore(input="output.wav", # input wav file path
                    output="audio1.wav", # output wav file path
                    cuda=True, # whether to use gpu acceleration
                    mode = 0) # You can try out mode 0, 1, or 2 to find out the best result
    
    noisy = enhance_model.load_audio(
    "audio1.wav"
    ).unsqueeze(0)

    enhanced = enhance_model.enhance_batch(noisy, lengths=torch.tensor([1.]))
    torchaudio.save("enhanced.wav", enhanced.cpu(), 16000)
    
    return [result.text, output.split("### Response:")[1].strip(), "enhanced.wav"]


c1 = gr.Interface(
    fn=evaluate,
    inputs=[
        gr.Audio(source="microphone", label = "请开始对话吧！TalktoAI！", type="filepath"),
        gr.Audio(source="upload", label = "请上传您喜欢的声音(wav/mp3文件)", type="filepath"),
        gr.components.Textbox(lines=2, label="Input", placeholder="none"),
        gr.components.Slider(minimum=0, maximum=1, value=0.1, label="Temperature"),
        gr.components.Slider(minimum=0, maximum=1, value=0.75, label="Top p"),
        gr.components.Slider(minimum=0, maximum=100, step=1, value=40, label="Top k"),
        gr.components.Slider(minimum=1, maximum=4, step=1, value=4, label="Beams"),
        gr.components.Slider(
            minimum=1, maximum=512, step=1, value=128, label="Max tokens"
        ),
    ],
    outputs=[
        gr.inputs.Textbox(
            lines=2,
            label="Speech to Text",
        ),
        gr.inputs.Textbox(
            lines=5,
            label="Alpaca-LoRA Output",
        ),
        gr.Audio(label="Audio with Custom Voice"),
    ],
#    title="🥳💬💕 - TalktoAI，随时随地，谈天说地！",
    description="🤖 - 让有人文关怀的AI造福每一个人！AI向善，文明璀璨！TalktoAI - Enable the future！",
    article = "Powered by [Alpaca-LoRA](https://github.com/tloen/alpaca-lora). Credit: tloen[https://github.com/tloen].",
)


c2 = gr.Interface(
    fn=evaluate,
    inputs=[
        gr.Audio(source="microphone", label = "请开始对话吧！TalktoAI！", type="filepath"),
        gr.Audio(source="microphone", label = "请上传您喜欢的声音，并尽量避免噪音", type="filepath"),
        gr.components.Textbox(lines=2, label="Input", placeholder="none"),
        gr.components.Slider(minimum=0, maximum=1, value=0.1, label="Temperature"),
        gr.components.Slider(minimum=0, maximum=1, value=0.75, label="Top p"),
        gr.components.Slider(minimum=0, maximum=100, step=1, value=40, label="Top k"),
        gr.components.Slider(minimum=1, maximum=4, step=1, value=4, label="Beams"),
        gr.components.Slider(
            minimum=1, maximum=512, step=1, value=128, label="Max tokens"
        ),
    ],
    outputs=[
        gr.inputs.Textbox(
            lines=2,
            label="Speech to Text",
        ),
        gr.inputs.Textbox(
            lines=5,
            label="Alpaca-LoRA Output",
        ),
        gr.Audio(label="Audio with Custom Voice"),
    ],
#    title="🥳💬💕 - TalktoAI，随时随地，谈天说地！",
    description="🤖 - 让有人文关怀的AI造福每一个人！AI向善，文明璀璨！TalktoAI - Enable the future！",
    article = "Powered by [Alpaca-LoRA](https://github.com/tloen/alpaca-lora). Credit: tloen[https://github.com/tloen].",
)

demo = gr.TabbedInterface([c1, c2], ["wav/mp3上传", "麦克风上传"], title = '🥳💬💕 - TalktoAI，随时随地，谈天说地！')
demo.queue(concurrency_count=1)
demo.launch(show_error = True)


# Old testing code follows.

"""
if __name__ == "__main__":
    # testing code for readme
    for instruction in [
        "Tell me about alpacas.",
        "Tell me about the president of Mexico in 2019.",
        "Tell me about the king of France in 2019.",
        "List all Canadian provinces in alphabetical order.",
        "Write a Python program that prints the first 10 Fibonacci numbers.",
        "Write a program that prints the numbers from 1 to 100. But for multiples of three print 'Fizz' instead of the number and for the multiples of five print 'Buzz'. For numbers which are multiples of both three and five print 'FizzBuzz'.",
        "Tell me five words that rhyme with 'shock'.",
        "Translate the sentence 'I have no mouth but I must scream' into Spanish.",
        "Count up from 1 to 500.",
    ]:
        print("Instruction:", instruction)
        print("Response:", evaluate(instruction))
        print()
"""