From 8b74da18776a7a3b3e4bb4f663a44cb44ccfeea9 Mon Sep 17 00:00:00 2001 From: Ettore Di Giacinto Date: Thu, 23 Jan 2025 09:39:28 +0100 Subject: [PATCH] chore(parler-tts): drop backend We support at this point more extensive backends that are SOTA and support also voice cloning, and many other features. This backend is superseded and also poses significant maintenance burden as there is an open issue https://github.com/mudler/LocalAI/issues/3941 which is still open as it deps are pinning old versions of grpc. Closes https://github.com/mudler/LocalAI/issues/3941 Signed-off-by: Ettore Di Giacinto --- Dockerfile | 7 +- Makefile | 13 +- backend/python/parler-tts/Makefile | 44 ------ backend/python/parler-tts/backend.py | 125 ------------------ backend/python/parler-tts/install.sh | 28 ---- backend/python/parler-tts/protogen.sh | 6 - .../python/parler-tts/requirements-after.txt | 4 - .../python/parler-tts/requirements-cpu.txt | 3 - .../parler-tts/requirements-cublas11.txt | 5 - .../parler-tts/requirements-cublas12.txt | 4 - .../parler-tts/requirements-hipblas.txt | 5 - .../python/parler-tts/requirements-intel.txt | 8 -- backend/python/parler-tts/requirements.txt | 4 - backend/python/parler-tts/run.sh | 4 - backend/python/parler-tts/test.py | 81 ------------ backend/python/parler-tts/test.sh | 6 - 16 files changed, 4 insertions(+), 343 deletions(-) delete mode 100644 backend/python/parler-tts/Makefile delete mode 100644 backend/python/parler-tts/backend.py delete mode 100755 backend/python/parler-tts/install.sh delete mode 100755 backend/python/parler-tts/protogen.sh delete mode 100644 backend/python/parler-tts/requirements-after.txt delete mode 100644 backend/python/parler-tts/requirements-cpu.txt delete mode 100644 backend/python/parler-tts/requirements-cublas11.txt delete mode 100644 backend/python/parler-tts/requirements-cublas12.txt delete mode 100644 backend/python/parler-tts/requirements-hipblas.txt delete mode 100644 backend/python/parler-tts/requirements-intel.txt delete mode 100644 backend/python/parler-tts/requirements.txt delete mode 100755 backend/python/parler-tts/run.sh delete mode 100644 backend/python/parler-tts/test.py delete mode 100755 backend/python/parler-tts/test.sh diff --git a/Dockerfile b/Dockerfile index 9f699ac98fee..625d28698f15 100644 --- a/Dockerfile +++ b/Dockerfile @@ -15,7 +15,7 @@ ARG TARGETARCH ARG TARGETVARIANT ENV DEBIAN_FRONTEND=noninteractive -ENV EXTERNAL_GRPC_BACKENDS="coqui:/build/backend/python/coqui/run.sh,transformers:/build/backend/python/transformers/run.sh,rerankers:/build/backend/python/rerankers/run.sh,autogptq:/build/backend/python/autogptq/run.sh,bark:/build/backend/python/bark/run.sh,diffusers:/build/backend/python/diffusers/run.sh,faster-whisper:/build/backend/python/faster-whisper/run.sh,openvoice:/build/backend/python/openvoice/run.sh,kokoro:/build/backend/python/kokoro/run.sh,vllm:/build/backend/python/vllm/run.sh,exllama2:/build/backend/python/exllama2/run.sh,parler-tts:/build/backend/python/parler-tts/run.sh" +ENV EXTERNAL_GRPC_BACKENDS="coqui:/build/backend/python/coqui/run.sh,transformers:/build/backend/python/transformers/run.sh,rerankers:/build/backend/python/rerankers/run.sh,autogptq:/build/backend/python/autogptq/run.sh,bark:/build/backend/python/bark/run.sh,diffusers:/build/backend/python/diffusers/run.sh,faster-whisper:/build/backend/python/faster-whisper/run.sh,openvoice:/build/backend/python/openvoice/run.sh,kokoro:/build/backend/python/kokoro/run.sh,vllm:/build/backend/python/vllm/run.sh,exllama2:/build/backend/python/exllama2/run.sh" RUN apt-get update && \ apt-get install -y --no-install-recommends \ @@ -410,11 +410,8 @@ RUN if [[ ( "${IMAGE_TYPE}" == "extras ")]]; then \ RUN if [[ ( "${EXTRA_BACKENDS}" =~ "coqui" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \ make -C backend/python/coqui \ ; fi && \ - if [[ ( "${EXTRA_BACKENDS}" =~ "parler-tts" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \ - make -C backend/python/parler-tts \ - ; fi && \ if [[ ( "${EXTRA_BACKENDS}" =~ "faster-whisper" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \ - make -C backend/python/parler-tts \ + make -C backend/python/faster-whisper \ ; fi && \ if [[ ( "${EXTRA_BACKENDS}" =~ "diffusers" || -z "${EXTRA_BACKENDS}" ) && "$IMAGE_TYPE" == "extras" ]]; then \ make -C backend/python/diffusers \ diff --git a/Makefile b/Makefile index fc649c4ff7ad..04e280d84deb 100644 --- a/Makefile +++ b/Makefile @@ -533,10 +533,10 @@ protogen-go-clean: $(RM) bin/* .PHONY: protogen-python -protogen-python: autogptq-protogen bark-protogen coqui-protogen diffusers-protogen exllama2-protogen rerankers-protogen transformers-protogen parler-tts-protogen kokoro-protogen vllm-protogen openvoice-protogen faster-whisper-protogen +protogen-python: autogptq-protogen bark-protogen coqui-protogen diffusers-protogen exllama2-protogen rerankers-protogen transformers-protogen kokoro-protogen vllm-protogen openvoice-protogen faster-whisper-protogen .PHONY: protogen-python-clean -protogen-python-clean: autogptq-protogen-clean bark-protogen-clean coqui-protogen-clean diffusers-protogen-clean exllama2-protogen-clean rerankers-protogen-clean transformers-protogen-clean parler-tts-protogen-clean kokoro-protogen-clean vllm-protogen-clean openvoice-protogen-clean faster-whisper-protogen-clean +protogen-python-clean: autogptq-protogen-clean bark-protogen-clean coqui-protogen-clean diffusers-protogen-clean exllama2-protogen-clean rerankers-protogen-clean transformers-protogen-clean kokoro-protogen-clean vllm-protogen-clean openvoice-protogen-clean faster-whisper-protogen-clean .PHONY: autogptq-protogen autogptq-protogen: @@ -602,14 +602,6 @@ transformers-protogen: transformers-protogen-clean: $(MAKE) -C backend/python/transformers protogen-clean -.PHONY: parler-tts-protogen -parler-tts-protogen: - $(MAKE) -C backend/python/parler-tts protogen - -.PHONY: parler-tts-protogen-clean -parler-tts-protogen-clean: - $(MAKE) -C backend/python/parler-tts protogen-clean - .PHONY: kokoro-protogen kokoro-protogen: $(MAKE) -C backend/python/kokoro protogen @@ -645,7 +637,6 @@ prepare-extra-conda-environments: protogen-python $(MAKE) -C backend/python/vllm $(MAKE) -C backend/python/rerankers $(MAKE) -C backend/python/transformers - $(MAKE) -C backend/python/parler-tts $(MAKE) -C backend/python/kokoro $(MAKE) -C backend/python/openvoice $(MAKE) -C backend/python/exllama2 diff --git a/backend/python/parler-tts/Makefile b/backend/python/parler-tts/Makefile deleted file mode 100644 index 48da2f3fb2f4..000000000000 --- a/backend/python/parler-tts/Makefile +++ /dev/null @@ -1,44 +0,0 @@ -export CONDA_ENV_PATH = "parler.yml" -SKIP_CONDA?=0 -ifeq ($(BUILD_TYPE), cublas) -export CONDA_ENV_PATH = "parler-nvidia.yml" -endif - -# Intel GPU are supposed to have dependencies installed in the main python -# environment, so we skip conda installation for SYCL builds. -# https://github.com/intel/intel-extension-for-pytorch/issues/538 -ifneq (,$(findstring sycl,$(BUILD_TYPE))) -export SKIP_CONDA=1 -endif - -.PHONY: parler-tts -parler-tts: - @echo "Installing $(CONDA_ENV_PATH)..." - bash install.sh $(CONDA_ENV_PATH) - $(MAKE) protogen - -.PHONY: run -run: protogen - @echo "Running transformers..." - bash run.sh - @echo "transformers run." - -.PHONY: test -test: protogen - @echo "Testing transformers..." - bash test.sh - @echo "transformers tested." - -.PHONY: protogen -protogen: backend_pb2_grpc.py backend_pb2.py - -.PHONY: protogen-clean -protogen-clean: - $(RM) backend_pb2_grpc.py backend_pb2.py - -backend_pb2_grpc.py backend_pb2.py: - bash protogen.sh - -.PHONY: clean -clean: protogen-clean - $(RM) -r venv __pycache__ \ No newline at end of file diff --git a/backend/python/parler-tts/backend.py b/backend/python/parler-tts/backend.py deleted file mode 100644 index 655990d75725..000000000000 --- a/backend/python/parler-tts/backend.py +++ /dev/null @@ -1,125 +0,0 @@ -#!/usr/bin/env python3 -""" -Extra gRPC server for MusicgenForConditionalGeneration models. -""" -from concurrent import futures - -import argparse -import signal -import sys -import os - -import time -import backend_pb2 -import backend_pb2_grpc - -import grpc - -from scipy.io.wavfile import write as write_wav - -from parler_tts import ParlerTTSForConditionalGeneration -from transformers import AutoTokenizer -import soundfile as sf -import torch - -_ONE_DAY_IN_SECONDS = 60 * 60 * 24 - -# If MAX_WORKERS are specified in the environment use it, otherwise default to 1 -MAX_WORKERS = int(os.environ.get('PYTHON_GRPC_MAX_WORKERS', '1')) - -# Implement the BackendServicer class with the service methods -class BackendServicer(backend_pb2_grpc.BackendServicer): - """ - A gRPC servicer for the backend service. - - This class implements the gRPC methods for the backend service, including Health, LoadModel, and Embedding. - """ - def Health(self, request, context): - """ - A gRPC method that returns the health status of the backend service. - - Args: - request: A HealthRequest object that contains the request parameters. - context: A grpc.ServicerContext object that provides information about the RPC. - - Returns: - A Reply object that contains the health status of the backend service. - """ - return backend_pb2.Reply(message=bytes("OK", 'utf-8')) - - def LoadModel(self, request, context): - """ - A gRPC method that loads a model into memory. - - Args: - request: A LoadModelRequest object that contains the request parameters. - context: A grpc.ServicerContext object that provides information about the RPC. - - Returns: - A Result object that contains the result of the LoadModel operation. - """ - model_name = request.Model - device = "cuda:0" if torch.cuda.is_available() else "cpu" - try: - self.model = ParlerTTSForConditionalGeneration.from_pretrained(model_name).to(device) - self.tokenizer = AutoTokenizer.from_pretrained(model_name) - except Exception as err: - return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}") - - return backend_pb2.Result(message="Model loaded successfully", success=True) - - def TTS(self, request, context): - model_name = request.model - voice = request.voice - if voice == "": - voice = "A female speaker with a slightly low-pitched voice delivers her words quite expressively, in a very confined sounding environment with clear audio quality. She speaks very fast." - if model_name == "": - return backend_pb2.Result(success=False, message="request.model is required") - try: - device = "cuda:0" if torch.cuda.is_available() else "cpu" - input_ids = self.tokenizer(voice, return_tensors="pt").input_ids.to(device) - prompt_input_ids = self.tokenizer(request.text, return_tensors="pt").input_ids.to(device) - - generation = self.model.generate(input_ids=input_ids, prompt_input_ids=prompt_input_ids) - audio_arr = generation.cpu().numpy().squeeze() - print("[parler-tts] TTS generated!", file=sys.stderr) - sf.write(request.dst, audio_arr, self.model.config.sampling_rate) - print("[parler-tts] TTS saved to", request.dst, file=sys.stderr) - print("[parler-tts] TTS for", file=sys.stderr) - print(request, file=sys.stderr) - except Exception as err: - return backend_pb2.Result(success=False, message=f"Unexpected {err=}, {type(err)=}") - return backend_pb2.Result(success=True) - - -def serve(address): - server = grpc.server(futures.ThreadPoolExecutor(max_workers=MAX_WORKERS)) - backend_pb2_grpc.add_BackendServicer_to_server(BackendServicer(), server) - server.add_insecure_port(address) - server.start() - print("[parler-tts] Server started. Listening on: " + address, file=sys.stderr) - - # Define the signal handler function - def signal_handler(sig, frame): - print("[parler-tts] Received termination signal. Shutting down...") - server.stop(0) - sys.exit(0) - - # Set the signal handlers for SIGINT and SIGTERM - signal.signal(signal.SIGINT, signal_handler) - signal.signal(signal.SIGTERM, signal_handler) - - try: - while True: - time.sleep(_ONE_DAY_IN_SECONDS) - except KeyboardInterrupt: - server.stop(0) - -if __name__ == "__main__": - parser = argparse.ArgumentParser(description="Run the gRPC server.") - parser.add_argument( - "--addr", default="localhost:50051", help="The address to bind the server to." - ) - args = parser.parse_args() - print(f"[parler-tts] startup: {args}", file=sys.stderr) - serve(args.addr) diff --git a/backend/python/parler-tts/install.sh b/backend/python/parler-tts/install.sh deleted file mode 100755 index 14df9b14068b..000000000000 --- a/backend/python/parler-tts/install.sh +++ /dev/null @@ -1,28 +0,0 @@ -#!/bin/bash -set -e - -source $(dirname $0)/../common/libbackend.sh - -# This is here because the Intel pip index is broken and returns 200 status codes for every package name, it just doesn't return any package links. -# This makes uv think that the package exists in the Intel pip index, and by default it stops looking at other pip indexes once it finds a match. -# We need uv to continue falling through to the pypi default index to find optimum[openvino] in the pypi index -# the --upgrade actually allows us to *downgrade* torch to the version provided in the Intel pip index -if [ "x${BUILD_PROFILE}" == "xintel" ]; then - EXTRA_PIP_INSTALL_FLAGS+=" --upgrade --index-strategy=unsafe-first-match" -fi - - -installRequirements - - -# https://github.com/descriptinc/audiotools/issues/101 -# incompatible protobuf versions. -PYDIR=python3.10 -pyenv="${MY_DIR}/venv/lib/${PYDIR}/site-packages/google/protobuf/internal/" - -if [ ! -d ${pyenv} ]; then - echo "(parler-tts/install.sh): Error: ${pyenv} does not exist" - exit 1 -fi - -curl -L https://raw.githubusercontent.com/protocolbuffers/protobuf/main/python/google/protobuf/internal/builder.py -o ${pyenv}/builder.py diff --git a/backend/python/parler-tts/protogen.sh b/backend/python/parler-tts/protogen.sh deleted file mode 100755 index 32f39fbb4f65..000000000000 --- a/backend/python/parler-tts/protogen.sh +++ /dev/null @@ -1,6 +0,0 @@ -#!/bin/bash -set -e - -source $(dirname $0)/../common/libbackend.sh - -python3 -m grpc_tools.protoc -I../.. --python_out=. --grpc_python_out=. backend.proto \ No newline at end of file diff --git a/backend/python/parler-tts/requirements-after.txt b/backend/python/parler-tts/requirements-after.txt deleted file mode 100644 index 702074de41c5..000000000000 --- a/backend/python/parler-tts/requirements-after.txt +++ /dev/null @@ -1,4 +0,0 @@ -git+https://github.com/huggingface/parler-tts.git@8e465f1b5fcd223478e07175cb40494d19ffbe17 -llvmlite==0.43.0 -numba==0.60.0 -grpcio-tools==1.42.0 \ No newline at end of file diff --git a/backend/python/parler-tts/requirements-cpu.txt b/backend/python/parler-tts/requirements-cpu.txt deleted file mode 100644 index 2021fc201f7e..000000000000 --- a/backend/python/parler-tts/requirements-cpu.txt +++ /dev/null @@ -1,3 +0,0 @@ -transformers -accelerate -torch==2.4.1 \ No newline at end of file diff --git a/backend/python/parler-tts/requirements-cublas11.txt b/backend/python/parler-tts/requirements-cublas11.txt deleted file mode 100644 index 9f8fe9ff87a3..000000000000 --- a/backend/python/parler-tts/requirements-cublas11.txt +++ /dev/null @@ -1,5 +0,0 @@ ---extra-index-url https://download.pytorch.org/whl/cu118 -torch==2.4.1+cu118 -torchaudio==2.4.1+cu118 -transformers -accelerate \ No newline at end of file diff --git a/backend/python/parler-tts/requirements-cublas12.txt b/backend/python/parler-tts/requirements-cublas12.txt deleted file mode 100644 index 537169495d1e..000000000000 --- a/backend/python/parler-tts/requirements-cublas12.txt +++ /dev/null @@ -1,4 +0,0 @@ -torch==2.4.1 -torchaudio==2.4.1 -transformers -accelerate \ No newline at end of file diff --git a/backend/python/parler-tts/requirements-hipblas.txt b/backend/python/parler-tts/requirements-hipblas.txt deleted file mode 100644 index b8758537835e..000000000000 --- a/backend/python/parler-tts/requirements-hipblas.txt +++ /dev/null @@ -1,5 +0,0 @@ ---extra-index-url https://download.pytorch.org/whl/rocm6.0 -torch==2.3.0+rocm6.0 -torchaudio==2.3.0+rocm6.0 -transformers -accelerate diff --git a/backend/python/parler-tts/requirements-intel.txt b/backend/python/parler-tts/requirements-intel.txt deleted file mode 100644 index f6814bd9af25..000000000000 --- a/backend/python/parler-tts/requirements-intel.txt +++ /dev/null @@ -1,8 +0,0 @@ ---extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/ -intel-extension-for-pytorch==2.3.110+xpu -torch==2.3.1+cxx11.abi -torchaudio==2.3.1+cxx11.abi -oneccl_bind_pt==2.3.100+xpu -optimum[openvino] -transformers -accelerate \ No newline at end of file diff --git a/backend/python/parler-tts/requirements.txt b/backend/python/parler-tts/requirements.txt deleted file mode 100644 index e6ba016ba484..000000000000 --- a/backend/python/parler-tts/requirements.txt +++ /dev/null @@ -1,4 +0,0 @@ -grpcio==1.69.0 -certifi -llvmlite==0.43.0 -setuptools \ No newline at end of file diff --git a/backend/python/parler-tts/run.sh b/backend/python/parler-tts/run.sh deleted file mode 100755 index 375c07e5f426..000000000000 --- a/backend/python/parler-tts/run.sh +++ /dev/null @@ -1,4 +0,0 @@ -#!/bin/bash -source $(dirname $0)/../common/libbackend.sh - -startBackend $@ \ No newline at end of file diff --git a/backend/python/parler-tts/test.py b/backend/python/parler-tts/test.py deleted file mode 100644 index 639d43a9101a..000000000000 --- a/backend/python/parler-tts/test.py +++ /dev/null @@ -1,81 +0,0 @@ -""" -A test script to test the gRPC service -""" -import unittest -import subprocess -import time -import backend_pb2 -import backend_pb2_grpc - -import grpc - - -class TestBackendServicer(unittest.TestCase): - """ - TestBackendServicer is the class that tests the gRPC service - """ - def setUp(self): - """ - This method sets up the gRPC service by starting the server - """ - self.service = subprocess.Popen(["python3", "backend.py", "--addr", "localhost:50051"]) - time.sleep(10) - - def tearDown(self) -> None: - """ - This method tears down the gRPC service by terminating the server - """ - self.service.terminate() - self.service.wait() - - def test_server_startup(self): - """ - This method tests if the server starts up successfully - """ - try: - self.setUp() - with grpc.insecure_channel("localhost:50051") as channel: - stub = backend_pb2_grpc.BackendStub(channel) - response = stub.Health(backend_pb2.HealthMessage()) - self.assertEqual(response.message, b'OK') - except Exception as err: - print(err) - self.fail("Server failed to start") - finally: - self.tearDown() - - def test_load_model(self): - """ - This method tests if the model is loaded successfully - """ - try: - self.setUp() - with grpc.insecure_channel("localhost:50051") as channel: - stub = backend_pb2_grpc.BackendStub(channel) - response = stub.LoadModel(backend_pb2.ModelOptions(Model="parler-tts/parler_tts_mini_v0.1")) - self.assertTrue(response.success) - self.assertEqual(response.message, "Model loaded successfully") - except Exception as err: - print(err) - self.fail("LoadModel service failed") - finally: - self.tearDown() - - def test_tts(self): - """ - This method tests if the embeddings are generated successfully - """ - try: - self.setUp() - with grpc.insecure_channel("localhost:50051") as channel: - stub = backend_pb2_grpc.BackendStub(channel) - response = stub.LoadModel(backend_pb2.ModelOptions(Model="parler-tts/parler_tts_mini_v0.1")) - self.assertTrue(response.success) - tts_request = backend_pb2.TTSRequest(text="Hey, how are you doing today?") - tts_response = stub.TTS(tts_request) - self.assertIsNotNone(tts_response) - except Exception as err: - print(err) - self.fail("TTS service failed") - finally: - self.tearDown() \ No newline at end of file diff --git a/backend/python/parler-tts/test.sh b/backend/python/parler-tts/test.sh deleted file mode 100755 index 6940b0661df2..000000000000 --- a/backend/python/parler-tts/test.sh +++ /dev/null @@ -1,6 +0,0 @@ -#!/bin/bash -set -e - -source $(dirname $0)/../common/libbackend.sh - -runUnittests