Add files via upload

TTS Generator - Various interface bugs & filtering options cleaned up.
TTS Generator - TTSDiff now scans generated text and TTS for errors.
TTS Generator - TTSSRT now creates subtitle files for video production e.g. a Youtube video.
Finetune - Now uses a customised tokenizer to deal with Japanese.
Finetune - Pre flight check and warning messages.
Finetune - Extra documentation and warnings.
Entire file structure has been re-organised to simplify management and future changes.
Documentation (built in and Github) has been rewritten/tidied up.
Requirements files have been cleaned up and simplified.
ATsetup has been re-written as necessary with additional options.
DeepSpeed moved up to version 14.
Standalone Application moved to PyTorch 2.2.1.
Nvidia CUDA Toolkit installation is NO LONGER needed (other than to compile DeepSpeed on Linux)

Tested on Linux and Windows.

.github/ISSUE_TEMPLATE/bug-report-support-request.md

@@ -1,13 +1,14 @@
 ---
-name: Bug report/Support request
+name: Support request OR Bug report
 about: Create a report to help us improve
 title: ''
 labels: ''
 assignees: ''
 
 ---
+🔴 If you have installed AllTalk in a custom Python environment, I will only be able to provide limited assistance/support. AllTalk draws on a variety of scripts and libraries that are not written or managed by myself, and they may fail, error or give strange results in custom built python environments.
 
-Please generate a diagnostics report and upload the "diagnostics.log".
+🔴 Please generate a diagnostics report and upload the "diagnostics.log" as this helps me understand your configuration.
 
 https://github.com/erew123/alltalk_tts/tree/main?#-how-to-make-a-diagnostics-report-file
 

Deepspeed-Windows/11.1/csrc/quantization/pt_binding.cpp

@@ -1,285 +0,0 @@
-// Copyright (c) Microsoft Corporation.
-// SPDX-License-Identifier: Apache-2.0
-
-// DeepSpeed Team
-
-#include <ATen/cuda/CUDAContext.h>
-#include <torch/extension.h>
-#include <cassert>
-#include <vector>
-#include "quantization.h"
-
-template <typename T>
-at::Tensor ds_quantize(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if ((((size / groups) - 1) / 4096 + 1) <= 256) {
-        launch_fake_quantize_kernel(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-template <typename T>
-at::Tensor ds_sr_quantize(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if (((size / groups) / 4 / 1024) <= 256) {
-        launch_sr_fake_quantize_kernel(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-template <typename T>
-at::Tensor ds_quantize_asym(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if ((((size / groups) - 1) / 4096 + 1) <= 256) {
-        launch_fake_quantize_kernel_asym(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-template <typename T>
-at::Tensor ds_sr_quantize_asym(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if (((size / groups) / 4 / 1024) <= 256) {
-        launch_sr_fake_quantize_kernel_asym(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-std::vector<at::Tensor> quantize_kernel(at::Tensor& input_vals,
-                                        int groups,
-                                        int numBits,
-                                        quantize::Type quantType)
-{
-    auto dtype = at::kFloat;
-    auto params_options = at::TensorOptions()
-                              .dtype(dtype)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-    const int param_elems = (quantize::requires_offset(quantType)) ? 2 : 1;
-    auto params = torch::empty({groups, param_elems}, params_options);
-
-    auto output_options = at::TensorOptions()
-                              .dtype(at::kChar)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    auto output_sizes = input_vals.sizes().vec();
-    output_sizes[output_sizes.size() - 1] /= numBits == 8 ? 1 : 2;
-    auto output = torch::empty(output_sizes, output_options);
-
-    const int elems_per_group = at::numel(input_vals) / groups;
-
-    launch_quant((int8_t*)output.data_ptr(),
-                 (float*)params.data_ptr(),
-                 (__half*)input_vals.data_ptr(),
-                 groups,
-                 elems_per_group,
-                 numBits,
-                 quantType,
-                 at::cuda::getCurrentCUDAStream());
-
-    return {output, params};
-}
-
-template <typename T>
-at::Tensor dequantize(at::Tensor& quantized_data,
-                      at::Tensor& params,
-                      int groups,
-                      int num_bits,
-                      quantize::Type quant_type)
-{
-    auto dtype = (std::is_same<T, float>::value) ? torch::kFloat32 : torch::kFloat16;
-    auto output_options = at::TensorOptions()
-                              .dtype(dtype)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    auto output_sizes = quantized_data.sizes().vec();
-    output_sizes[output_sizes.size() - 1] *= num_bits == 8 ? 1 : 2;
-    auto output = torch::empty(output_sizes, output_options);
-
-    const int total_elems = at::numel(output);
-    const int elems_per_group = total_elems / groups;
-
-    launch_dequantize_kernel((T*)output.data_ptr(),
-                             (const int8_t*)quantized_data.data_ptr(),
-                             (const float*)params.data_ptr(),
-                             quant_type,
-                             num_bits,
-                             elems_per_group,
-                             total_elems,
-                             at::cuda::getCurrentCUDAStream());
-
-    return output;
-}
-
-at::Tensor dequantize_int4_to_half_experimental(at::Tensor& data_in,
-                                                at::Tensor& scale_buffer,
-                                                at::Tensor& min_val_buffer,
-                                                int num_group,
-                                                int group_size)
-{
-    auto output_options = at::TensorOptions().dtype(at::kHalf).device(at::kCUDA);
-    auto output = torch::empty({num_group, group_size}, output_options);
-
-    launch_dequantize_int4_to_half_experimental((uint8_t*)data_in.data_ptr(),
-                                                (half*)output.data_ptr(),
-                                                (half*)scale_buffer.data_ptr(),
-                                                (half*)min_val_buffer.data_ptr(),
-                                                num_group,
-                                                group_size,
-                                                at::cuda::getCurrentCUDAStream());
-
-    return output;
-}
-
-std::vector<at::Tensor> ds_swizzle_quant(at::Tensor& input_vals,
-                                         int groups,
-                                         int num_bits,
-                                         quantize::Type quant_type,
-                                         int pipeline_size,
-                                         int nodes,
-                                         int devices_per_node)
-{
-    auto scales_options = at::TensorOptions()
-                              .dtype(at::kFloat)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-    const int scales_elems = (quantize::requires_offset(quant_type)) ? 2 : 1;
-    auto scales = torch::empty({groups, scales_elems}, scales_options);
-
-    auto output_options = at::TensorOptions()
-                              .dtype(at::kChar)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    const int quantization_scalar = 8 / num_bits;
-    const int compressed_vals = at::numel(input_vals) / quantization_scalar;
-
-    auto output = torch::empty({compressed_vals}, output_options);
-    const int elems_per_group = at::numel(input_vals) / groups;
-
-    launch_swizzled_quant((int8_t*)output.data_ptr(),
-                          (float*)scales.data_ptr(),
-                          (__half*)input_vals.data_ptr(),
-                          num_bits,
-                          quant_type,
-                          groups,
-                          elems_per_group,
-                          pipeline_size,
-                          nodes,
-                          devices_per_node,
-                          at::cuda::getCurrentCUDAStream());
-
-    return {output, scales};
-}
-
-std::vector<at::Tensor> quantized_reduction(at::Tensor& input_vals,
-                                            at::Tensor& input_scales,
-                                            int in_groups,
-                                            int out_groups,
-                                            int num_bits,
-                                            quantize::Type quant_type,
-                                            int devices_per_node)
-{
-    auto scales_options = at::TensorOptions()
-                              .dtype(at::kFloat)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-    const int scales_elems = (quantize::requires_offset(quant_type)) ? 2 : 1;
-    auto scales = torch::empty({out_groups, scales_elems}, scales_options);
-
-    auto output_options = at::TensorOptions()
-                              .dtype(at::kChar)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-     std::vector<int64_t> sz_vector(input_vals.sizes().begin(), input_vals.sizes().end());
-     sz_vector[sz_vector.size() - 1] = sz_vector.back() / devices_per_node;  // num of GPU per nodes
-     at::IntArrayRef sz(sz_vector);
-     auto output = torch::empty(sz, output_options);
-
-     const int elems_per_in_tensor = at::numel(input_vals) / devices_per_node;
-     const int elems_per_in_group = elems_per_in_tensor / (in_groups / devices_per_node);
-     const int elems_per_out_group = elems_per_in_tensor / out_groups;
-
-// old code:
-//    std::vector<long int> sz(input_vals.sizes().begin(), input_vals.sizes().end());
-//    sz[sz.size() - 1] = sz.back() / devices_per_node;  // num of GPU per nodes
-//    const int elems_per_in_tensor = at::numel(input_vals) / devices_per_node;
-//    auto output = torch::empty(sz, output_options);
-
-//    const int elems_per_in_group = elems_per_in_tensor / (in_groups / devices_per_node);
-//    const int elems_per_out_group = elems_per_in_tensor / out_groups;
-
-    launch_dequant_reduce((int8_t*)output.data_ptr(),
-                          (float*)scales.data_ptr(),
-                          (const int8_t*)input_vals.data_ptr(),
-                          (const float*)input_scales.data_ptr(),
-                          devices_per_node,
-                          num_bits,
-                          quant_type,
-                          out_groups,
-                          elems_per_out_group,
-                          elems_per_in_tensor,
-                          in_groups / devices_per_node,
-                          elems_per_in_group,
-                          at::cuda::getCurrentCUDAStream());
-    return {output, scales};
-}
-
-PYBIND11_MODULE(TORCH_EXTENSION_NAME, m)
-{
-    m.def("ds_quantize_fp32", &ds_quantize<float>, "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def("ds_quantize_fp16", &ds_quantize<__half>, "DeepSpeed Quantize with fp16 (CUDA)");
-    m.def("ds_sr_quantize_fp32", &ds_sr_quantize<float>, "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def("ds_sr_quantize_fp16", &ds_sr_quantize<__half>, "DeepSpeed Quantize with fp16 (CUDA)");
-    m.def("ds_quantize_asym_fp32", &ds_quantize_asym<float>, "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def(
-        "ds_quantize_asym_fp16", &ds_quantize_asym<__half>, "DeepSpeed Quantize with fp16 (CUDA)");
-    m.def("ds_sr_quantize_asym_fp32",
-          &ds_sr_quantize_asym<float>,
-          "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def("ds_sr_quantize_asym_fp16",
-          &ds_sr_quantize_asym<__half>,
-          "DeepSpeed Quantize with fp16 (CUDA)");
-    pybind11::enum_<quantize::Type>(m, "QuantizationType")
-        .value("Symmetric", quantize::Type::Symmetric)
-        .value("Asymmetric", quantize::Type::Asymmetric)
-        .export_values();
-    m.def("quantize", &quantize_kernel);
-    m.def("dequantize", &dequantize<__half>);
-    m.def("dequantize_fp32", &dequantize<float>);
-    m.def("dequantize_int4_to_half_experimental",
-          &dequantize_int4_to_half_experimental,
-          "Dequantize int4 to half (experimental)");
-    m.def("swizzle_quant", &ds_swizzle_quant);
-    m.def("quantized_reduction", &quantized_reduction);
-}

Deepspeed-Windows/11.1/deepspeed/env_report.py

@@ -1,160 +0,0 @@
-# Copyright (c) Microsoft Corporation.
-# SPDX-License-Identifier: Apache-2.0
-
-# DeepSpeed Team
-
-import os
-import torch
-import deepspeed
-import subprocess
-import argparse
-import psutil
-from .ops.op_builder.all_ops import ALL_OPS
-from .git_version_info import installed_ops, torch_info
-from deepspeed.accelerator import get_accelerator
-
-GREEN = '\033[92m'
-RED = '\033[91m'
-YELLOW = '\033[93m'
-END = '\033[0m'
-SUCCESS = f"{GREEN} [SUCCESS] {END}"
-OKAY = f"{GREEN}[OKAY]{END}"
-WARNING = f"{YELLOW}[WARNING]{END}"
-FAIL = f'{RED}[FAIL]{END}'
-INFO = '[INFO]'
-
-color_len = len(GREEN) + len(END)
-okay = f"{GREEN}[OKAY]{END}"
-warning = f"{YELLOW}[WARNING]{END}"
-
-
-def op_report(verbose=True):
-    max_dots = 23
-    max_dots2 = 11
-    h = ["op name", "installed", "compatible"]
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    print("DeepSpeed C++/CUDA extension op report")
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-
-    print("NOTE: Ops not installed will be just-in-time (JIT) compiled at\n"
-          "      runtime if needed. Op compatibility means that your system\n"
-          "      meet the required dependencies to JIT install the op.")
-
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    print("JIT compiled ops requires ninja")
-    ninja_status = OKAY if ninja_installed() else FAIL
-    print('ninja', "." * (max_dots - 5), ninja_status)
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    print(h[0], "." * (max_dots - len(h[0])), h[1], "." * (max_dots2 - len(h[1])), h[2])
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    installed = f"{GREEN}[YES]{END}"
-    no = f"{YELLOW}[NO]{END}"
-    for op_name, builder in ALL_OPS.items():
-        dots = "." * (max_dots - len(op_name))
-        is_compatible = OKAY if builder.is_compatible(verbose) else no
-        is_installed = installed if installed_ops.get(op_name, False) else no
-        dots2 = '.' * ((len(h[1]) + (max_dots2 - len(h[1]))) - (len(is_installed) - color_len))
-        print(op_name, dots, is_installed, dots2, is_compatible)
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-
-
-def ninja_installed():
-    try:
-        import ninja  # noqa: F401 # type: ignore
-    except ImportError:
-        return False
-    return True
-
-
-def nvcc_version():
-    import torch.utils.cpp_extension
-    cuda_home = torch.utils.cpp_extension.CUDA_HOME
-    if cuda_home is None:
-        return f"{RED} [FAIL] cannot find CUDA_HOME via torch.utils.cpp_extension.CUDA_HOME={torch.utils.cpp_extension.CUDA_HOME} {END}"
-    try:
-        output = subprocess.check_output([cuda_home + "/bin/nvcc", "-V"], universal_newlines=True)
-    except FileNotFoundError:
-        return f"{RED} [FAIL] nvcc missing {END}"
-    output_split = output.split()
-    release_idx = output_split.index("release")
-    release = output_split[release_idx + 1].replace(',', '').split(".")
-    return ".".join(release)
-
-
-def get_shm_size():
-    try:
-        temp_dir = os.getenv('TEMP') or os.getenv('TMP') or os.path.join(os.path.expanduser('~'), 'tmp')
-        shm_stats = psutil.disk_usage(temp_dir)
-        shm_size = shm_stats.total
-        shm_hbytes = human_readable_size(shm_size)
-        warn = []
-        if shm_size < 512 * 1024**2:
-            warn.append(
-                f" {YELLOW} [WARNING] Shared memory size might be too small, consider increasing it. {END}"
-            )
-            # Add additional warnings specific to your use case if needed.
-        return shm_hbytes, warn
-    except Exception as e:
-        return "UNKNOWN", [f"Error getting shared memory size: {e}"]
-
-def human_readable_size(size):
-    units = ['B', 'KB', 'MB', 'GB', 'TB']
-    i = 0
-    while size >= 1024 and i < len(units) - 1:
-        size /= 1024
-        i += 1
-    return f'{size:.2f} {units[i]}'
-
-
-def debug_report():
-    max_dots = 33
-
-    report = [("torch install path", torch.__path__), ("torch version", torch.__version__),
-              ("deepspeed install path", deepspeed.__path__),
-              ("deepspeed info", f"{deepspeed.__version__}, {deepspeed.__git_hash__}, {deepspeed.__git_branch__}")]
-    if get_accelerator().device_name() == 'cuda':
-        hip_version = getattr(torch.version, "hip", None)
-        report.extend([("torch cuda version", torch.version.cuda), ("torch hip version", hip_version),
-                       ("nvcc version", (None if hip_version else nvcc_version())),
-                       ("deepspeed wheel compiled w.", f"torch {torch_info['version']}, " +
-                        (f"hip {torch_info['hip_version']}" if hip_version else f"cuda {torch_info['cuda_version']}"))
-                       ])
-    else:
-        report.extend([("deepspeed wheel compiled w.", f"torch {torch_info['version']} ")])
-
-    report.append(("shared memory (/dev/shm) size", get_shm_size()))
-
-    print("DeepSpeed general environment info:")
-    for name, value in report:
-        warns = []
-        if isinstance(value, tuple):
-            value, warns = value
-        print(name, "." * (max_dots - len(name)), value)
-        if warns:
-            for warn in warns:
-                print(warn)
-
-
-def parse_arguments():
-    parser = argparse.ArgumentParser()
-    parser.add_argument('--hide_operator_status',
-                        action='store_true',
-                        help='Suppress display of installation and compatibility statuses of DeepSpeed operators. ')
-    parser.add_argument('--hide_errors_and_warnings', action='store_true', help='Suppress warning and error messages.')
-    args = parser.parse_args()
-    return args
-
-
-def main(hide_operator_status=False, hide_errors_and_warnings=False):
-    if not hide_operator_status:
-        op_report(verbose=not hide_errors_and_warnings)
-    debug_report()
-
-
-def cli_main():
-    args = parse_arguments()
-    main(hide_operator_status=args.hide_operator_status, hide_errors_and_warnings=args.hide_errors_and_warnings)
-
-
-if __name__ == "__main__":
-    main()

Deepspeed-Windows/11.2/csrc/quantization/pt_binding.cpp

@@ -1,299 +0,0 @@
-// Copyright (c) Microsoft Corporation.
-// SPDX-License-Identifier: Apache-2.0
-
-// DeepSpeed Team
-
-#include <ATen/cuda/CUDAContext.h>
-#include <torch/extension.h>
-#include <cassert>
-#include <vector>
-#include "quantization.h"
-
-template <typename T>
-at::Tensor ds_quantize(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if ((((size / groups) - 1) / 4096 + 1) <= 256) {
-        launch_fake_quantize_kernel(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-template <typename T>
-at::Tensor ds_sr_quantize(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if (((size / groups) / 4 / 1024) <= 256) {
-        launch_sr_fake_quantize_kernel(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-template <typename T>
-at::Tensor ds_quantize_asym(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if ((((size / groups) - 1) / 4096 + 1) <= 256) {
-        launch_fake_quantize_kernel_asym(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-template <typename T>
-at::Tensor ds_sr_quantize_asym(at::Tensor& vals, int groups, int bits)
-{
-    auto t_size = vals.sizes();
-    int size = 1;
-    for (auto dim : t_size) size *= dim;
-
-    if (((size / groups) / 4 / 1024) <= 256) {
-        launch_sr_fake_quantize_kernel_asym(
-            (T*)vals.data_ptr(), size, groups, bits, at::cuda::getCurrentCUDAStream());
-    }
-    return vals;
-}
-
-std::vector<at::Tensor> quantize_kernel(at::Tensor& input_vals,
-                                        int groups,
-                                        int numBits,
-                                        quantize::Type quantType)
-{
-    auto dtype = at::kFloat;
-    auto params_options = at::TensorOptions()
-                              .dtype(dtype)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-    const int param_elems = (quantize::requires_offset(quantType)) ? 2 : 1;
-    auto params = torch::empty({groups, param_elems}, params_options);
-
-    auto output_options = at::TensorOptions()
-                              .dtype(at::kChar)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    auto output_sizes = input_vals.sizes().vec();
-    output_sizes[output_sizes.size() - 1] /= numBits == 8 ? 1 : 2;
-    auto output = torch::empty(output_sizes, output_options);
-
-    const int elems_per_group = at::numel(input_vals) / groups;
-
-    launch_quant((int8_t*)output.data_ptr(),
-                 (float*)params.data_ptr(),
-                 (__half*)input_vals.data_ptr(),
-                 groups,
-                 elems_per_group,
-                 numBits,
-                 quantType,
-                 at::cuda::getCurrentCUDAStream());
-
-    return {output, params};
-}
-
-template <typename T>
-at::Tensor dequantize(at::Tensor& quantized_data,
-                      at::Tensor& params,
-                      int groups,
-                      int num_bits,
-                      quantize::Type quant_type)
-{
-    auto dtype = (std::is_same<T, float>::value) ? torch::kFloat32 : torch::kFloat16;
-    auto output_options = at::TensorOptions()
-                              .dtype(dtype)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    auto output_sizes = quantized_data.sizes().vec();
-    output_sizes[output_sizes.size() - 1] *= num_bits == 8 ? 1 : 2;
-    auto output = torch::empty(output_sizes, output_options);
-
-    const int total_elems = at::numel(output);
-    const int elems_per_group = total_elems / groups;
-
-    launch_dequantize_kernel((T*)output.data_ptr(),
-                             (const int8_t*)quantized_data.data_ptr(),
-                             (const float*)params.data_ptr(),
-                             quant_type,
-                             num_bits,
-                             elems_per_group,
-                             total_elems,
-                             at::cuda::getCurrentCUDAStream());
-
-    return output;
-}
-
-at::Tensor dequantize_int4_to_half_experimental(at::Tensor& data_in,
-                                                at::Tensor& scale_buffer,
-                                                at::Tensor& min_val_buffer,
-                                                int num_group,
-                                                int group_size)
-{
-    auto output_options = at::TensorOptions().dtype(at::kHalf).device(at::kCUDA);
-    auto output = torch::empty({num_group, group_size}, output_options);
-
-    launch_dequantize_int4_to_half_experimental((uint8_t*)data_in.data_ptr(),
-                                                (half*)output.data_ptr(),
-                                                (half*)scale_buffer.data_ptr(),
-                                                (half*)min_val_buffer.data_ptr(),
-                                                num_group,
-                                                group_size,
-                                                at::cuda::getCurrentCUDAStream());
-
-    return output;
-}
-
-at::Tensor dequantize_int8_to_half_experimental(at::Tensor& data_in,
-                                                at::Tensor& scale_buffer,
-                                                at::Tensor& min_val_buffer,
-                                                int num_group,
-                                                int group_size)
-{
-    auto output_options = at::TensorOptions().dtype(at::kHalf).device(at::kCUDA);
-    auto output = torch::empty({num_group, group_size}, output_options);
-
-    launch_dequantize_int8_to_half_experimental((uint8_t*)data_in.data_ptr(),
-                                                (half*)output.data_ptr(),
-                                                (half*)scale_buffer.data_ptr(),
-                                                (half*)min_val_buffer.data_ptr(),
-                                                num_group,
-                                                group_size,
-                                                at::cuda::getCurrentCUDAStream());
-
-    return output;
-}
-
-std::vector<at::Tensor> ds_swizzle_quant(at::Tensor& input_vals,
-                                         int groups,
-                                         int num_bits,
-                                         quantize::Type quant_type,
-                                         int pipeline_size,
-                                         int nodes,
-                                         int devices_per_node)
-{
-    auto scales_options = at::TensorOptions()
-                              .dtype(at::kFloat)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-    const int scales_elems = (quantize::requires_offset(quant_type)) ? 2 : 1;
-    auto scales = torch::empty({groups, scales_elems}, scales_options);
-
-    auto output_options = at::TensorOptions()
-                              .dtype(at::kChar)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    const int quantization_scalar = 8 / num_bits;
-    const int compressed_vals = at::numel(input_vals) / quantization_scalar;
-
-    auto output = torch::empty({compressed_vals}, output_options);
-    const int elems_per_group = at::numel(input_vals) / groups;
-
-    launch_swizzled_quant((int8_t*)output.data_ptr(),
-                          (float*)scales.data_ptr(),
-                          (__half*)input_vals.data_ptr(),
-                          num_bits,
-                          quant_type,
-                          groups,
-                          elems_per_group,
-                          pipeline_size,
-                          nodes,
-                          devices_per_node,
-                          at::cuda::getCurrentCUDAStream());
-
-    return {output, scales};
-}
-
-std::vector<at::Tensor> quantized_reduction(at::Tensor& input_vals,
-                                            at::Tensor& input_scales,
-                                            int in_groups,
-                                            int out_groups,
-                                            int num_bits,
-                                            quantize::Type quant_type,
-                                            int devices_per_node)
-{
-    auto scales_options = at::TensorOptions()
-                              .dtype(at::kFloat)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-    const int scales_elems = (quantize::requires_offset(quant_type)) ? 2 : 1;
-    auto scales = torch::empty({out_groups, scales_elems}, scales_options);
-
-    auto output_options = at::TensorOptions()
-                              .dtype(at::kChar)
-                              .layout(at::kStrided)
-                              .device(at::kCUDA)
-                              .requires_grad(false);
-
-    std::vector<int64_t> sz_vector(input_vals.sizes().begin(), input_vals.sizes().end());
-    sz_vector[sz_vector.size() - 1] = sz_vector.back() / devices_per_node;  // num of GPU per nodes
-    at::IntArrayRef sz(sz_vector);
-    auto output = torch::empty(sz, output_options);
-
-    const int elems_per_in_tensor = at::numel(input_vals) / devices_per_node;
-    const int elems_per_in_group = elems_per_in_tensor / (in_groups / devices_per_node);
-    const int elems_per_out_group = elems_per_in_tensor / out_groups;
-
-    launch_dequant_reduce((int8_t*)output.data_ptr(),
-                          (float*)scales.data_ptr(),
-                          (const int8_t*)input_vals.data_ptr(),
-                          (const float*)input_scales.data_ptr(),
-                          devices_per_node,
-                          num_bits,
-                          quant_type,
-                          out_groups,
-                          elems_per_out_group,
-                          elems_per_in_tensor,
-                          in_groups / devices_per_node,
-                          elems_per_in_group,
-                          at::cuda::getCurrentCUDAStream());
-    return {output, scales};
-}
-
-PYBIND11_MODULE(TORCH_EXTENSION_NAME, m)
-{
-    m.def("ds_quantize_fp32", &ds_quantize<float>, "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def("ds_quantize_fp16", &ds_quantize<__half>, "DeepSpeed Quantize with fp16 (CUDA)");
-    m.def("ds_sr_quantize_fp32", &ds_sr_quantize<float>, "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def("ds_sr_quantize_fp16", &ds_sr_quantize<__half>, "DeepSpeed Quantize with fp16 (CUDA)");
-    m.def("ds_quantize_asym_fp32", &ds_quantize_asym<float>, "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def(
-        "ds_quantize_asym_fp16", &ds_quantize_asym<__half>, "DeepSpeed Quantize with fp16 (CUDA)");
-    m.def("ds_sr_quantize_asym_fp32",
-          &ds_sr_quantize_asym<float>,
-          "DeepSpeed Quantize with fp32 (CUDA)");
-    m.def("ds_sr_quantize_asym_fp16",
-          &ds_sr_quantize_asym<__half>,
-          "DeepSpeed Quantize with fp16 (CUDA)");
-    pybind11::enum_<quantize::Type>(m, "QuantizationType")
-        .value("Symmetric", quantize::Type::Symmetric)
-        .value("Asymmetric", quantize::Type::Asymmetric)
-        .export_values();
-    m.def("quantize", &quantize_kernel);
-    m.def("dequantize", &dequantize<__half>);
-    m.def("dequantize_fp32", &dequantize<float>);
-    m.def("dequantize_int4_to_half_experimental",
-          &dequantize_int4_to_half_experimental,
-          "Dequantize int4 to half (experimental)");
-    m.def("dequantize_int8_to_half_experimental",
-          &dequantize_int8_to_half_experimental,
-          "Dequantize int8 to half (experimental)");
-    m.def("swizzle_quant", &ds_swizzle_quant);
-    m.def("quantized_reduction", &quantized_reduction);
-}

Deepspeed-Windows/11.2/deepspeed/env_report.py

@@ -1,161 +0,0 @@
-# Copyright (c) Microsoft Corporation.
-# SPDX-License-Identifier: Apache-2.0
-
-# DeepSpeed Team
-
-import os
-import torch
-import deepspeed
-import subprocess
-import argparse
-import psutil
-from .ops.op_builder.all_ops import ALL_OPS
-from .git_version_info import installed_ops, torch_info
-from deepspeed.accelerator import get_accelerator
-
-GREEN = '\033[92m'
-RED = '\033[91m'
-YELLOW = '\033[93m'
-END = '\033[0m'
-SUCCESS = f"{GREEN} [SUCCESS] {END}"
-OKAY = f"{GREEN}[OKAY]{END}"
-WARNING = f"{YELLOW}[WARNING]{END}"
-FAIL = f'{RED}[FAIL]{END}'
-INFO = '[INFO]'
-
-color_len = len(GREEN) + len(END)
-okay = f"{GREEN}[OKAY]{END}"
-warning = f"{YELLOW}[WARNING]{END}"
-
-
-def op_report(verbose=True):
-    max_dots = 23
-    max_dots2 = 11
-    h = ["op name", "installed", "compatible"]
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    print("DeepSpeed C++/CUDA extension op report")
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-
-    print("NOTE: Ops not installed will be just-in-time (JIT) compiled at\n"
-          "      runtime if needed. Op compatibility means that your system\n"
-          "      meet the required dependencies to JIT install the op.")
-
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    print("JIT compiled ops requires ninja")
-    ninja_status = OKAY if ninja_installed() else FAIL
-    print('ninja', "." * (max_dots - 5), ninja_status)
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    print(h[0], "." * (max_dots - len(h[0])), h[1], "." * (max_dots2 - len(h[1])), h[2])
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-    installed = f"{GREEN}[YES]{END}"
-    no = f"{YELLOW}[NO]{END}"
-    for op_name, builder in ALL_OPS.items():
-        dots = "." * (max_dots - len(op_name))
-        is_compatible = OKAY if builder.is_compatible(verbose) else no
-        is_installed = installed if installed_ops.get(op_name, False) else no
-        dots2 = '.' * ((len(h[1]) + (max_dots2 - len(h[1]))) - (len(is_installed) - color_len))
-        print(op_name, dots, is_installed, dots2, is_compatible)
-    print("-" * (max_dots + max_dots2 + len(h[0]) + len(h[1])))
-
-
-def ninja_installed():
-    try:
-        import ninja  # noqa: F401 # type: ignore
-    except ImportError:
-        return False
-    return True
-
-
-def nvcc_version():
-    import torch.utils.cpp_extension
-    cuda_home = torch.utils.cpp_extension.CUDA_HOME
-    if cuda_home is None:
-        return f"{RED} [FAIL] cannot find CUDA_HOME via torch.utils.cpp_extension.CUDA_HOME={torch.utils.cpp_extension.CUDA_HOME} {END}"
-    try:
-        output = subprocess.check_output([cuda_home + "/bin/nvcc", "-V"], universal_newlines=True)
-    except FileNotFoundError:
-        return f"{RED} [FAIL] nvcc missing {END}"
-    output_split = output.split()
-    release_idx = output_split.index("release")
-    release = output_split[release_idx + 1].replace(',', '').split(".")
-    return ".".join(release)
-
-
-def get_shm_size():
-    try:
-        temp_dir = os.getenv('TEMP') or os.getenv('TMP') or os.path.join(os.path.expanduser('~'), 'tmp')
-        shm_stats = psutil.disk_usage(temp_dir)
-        shm_size = shm_stats.total
-        shm_hbytes = human_readable_size(shm_size)
-        warn = []
-        if shm_size < 512 * 1024**2:
-            warn.append(
-                f" {YELLOW} [WARNING] Shared memory size might be too small, consider increasing it. {END}"
-            )
-            # Add additional warnings specific to your use case if needed.
-        return shm_hbytes, warn
-    except Exception as e:
-        return "UNKNOWN", [f"Error getting shared memory size: {e}"]
-
-
-def human_readable_size(size):
-    units = ['B', 'KB', 'MB', 'GB', 'TB']
-    i = 0
-    while size >= 1024 and i < len(units) - 1:
-        size /= 1024
-        i += 1
-    return f'{size:.2f} {units[i]}'
-
-
-def debug_report():
-    max_dots = 33
-
-    report = [("torch install path", torch.__path__), ("torch version", torch.__version__),
-              ("deepspeed install path", deepspeed.__path__),
-              ("deepspeed info", f"{deepspeed.__version__}, {deepspeed.__git_hash__}, {deepspeed.__git_branch__}")]
-    if get_accelerator().device_name() == 'cuda':
-        hip_version = getattr(torch.version, "hip", None)
-        report.extend([("torch cuda version", torch.version.cuda), ("torch hip version", hip_version),
-                       ("nvcc version", (None if hip_version else nvcc_version())),
-                       ("deepspeed wheel compiled w.", f"torch {torch_info['version']}, " +
-                        (f"hip {torch_info['hip_version']}" if hip_version else f"cuda {torch_info['cuda_version']}"))
-                       ])
-    else:
-        report.extend([("deepspeed wheel compiled w.", f"torch {torch_info['version']} ")])
-
-    report.append(("shared memory (/dev/shm) size", get_shm_size()))
-
-    print("DeepSpeed general environment info:")
-    for name, value in report:
-        warns = []
-        if isinstance(value, tuple):
-            value, warns = value
-        print(name, "." * (max_dots - len(name)), value)
-        if warns:
-            for warn in warns:
-                print(warn)
-
-
-def parse_arguments():
-    parser = argparse.ArgumentParser()
-    parser.add_argument('--hide_operator_status',
-                        action='store_true',
-                        help='Suppress display of installation and compatibility statuses of DeepSpeed operators. ')
-    parser.add_argument('--hide_errors_and_warnings', action='store_true', help='Suppress warning and error messages.')
-    args = parser.parse_args()
-    return args
-
-
-def main(hide_operator_status=False, hide_errors_and_warnings=False):
-    if not hide_operator_status:
-        op_report(verbose=not hide_errors_and_warnings)
-    debug_report()
-
-
-def cli_main():
-    args = parse_arguments()
-    main(hide_operator_status=args.hide_operator_status, hide_errors_and_warnings=args.hide_errors_and_warnings)
-
-
-if __name__ == "__main__":
-    main()

Deepspeed-Windows/DeepSpeed readme.txt

@@ -1,18 +0,0 @@
-These are the modified files you would need for a MANUAL build of DeepSpeed 11.1 or 11.2.
-
-They were created by a mix of instructions from erew123 and S95Sedan.
-
-
-- You still have to download the DeepSpeed v11.1 or v11.2 installation files from Microsoft to use these files.
-
-https://github.com/microsoft/DeepSpeed/releases/tag/v0.11.2
-
-
-- These files were provided by S95Sedan here: 
-
-https://github.com/S95Sedan/Deepspeed-Windows
-
-
-- They are related to the manual build of DeepSpeed for AllTalk, which is linked here: 
-
-https://github.com/erew123/alltalk_tts/tree/main?tab=readme-ov-file#-option-2---a-bit-more-complicated

atsetup.bat

@@ -61,28 +61,36 @@ echo    environment. If you have NOT done this, please run %L_GREEN%cmd_windows.
 echo    in the %L_GREEN%text-generation-webui%RESET% folder and then re-run this script.
 echo.
 echo    BASE REQUIREMENTS
-echo    1) Install the requirements for a %L_GREEN%Nvidia machine%RESET%.
+echo    1) Apply/Re-Apply the requirements for %L_GREEN%Text-generation-webui%RESET%.
-echo    2) Install the requirements for a %L_GREEN%AMD or MAC machine%RESET%.
 echo.
 echo    OPTIONAL
-echo    3) Install the Finetuning requirements.
+echo    2) Git Pull the latest AllTalk updates from Github
-echo    4) Install DeepSpeed v11.2 for CUDA %L_GREEN%11.8%RESET% and Python-3.11.x.
+echo.
-echo    5) Install DeepSpeed v11.2 for CUDA %L_GREEN%12.1%RESET% and Python-3.11.x.
+echo    DEEPSPEED FOR %L_YELLOW=%PyTorch 2.1.x%RESET%
-echo    6) Uninstall DeepSpeed.
+echo    4) Install DeepSpeed v11.2 for CUDA %L_GREEN%11.8%RESET% and Python-3.11.x and %L_YELLOW%PyTorch 2.1.x%RESET%.
+echo    5) Install DeepSpeed v11.2 for CUDA %L_GREEN%12.1%RESET% and Python-3.11.x and %L_YELLOW%PyTorch 2.1.x%RESET%.
+echo.
+echo    DEEPSPEED FOR %L_YELLOW=%PyTorch 2.2.x%RESET% (March 2024 builds of Text-gen-webui and later)
+echo    6) Install DeepSpeed v14.0 for CUDA %L_GREEN%12.1%RESET% and Python-3.11.x and %L_YELLOW%PyTorch 2.2.x%RESET%.
+echo    7) Install DeepSpeed v14.0 for CUDA %L_GREEN%11.8%RESET% and Python-3.11.x and %L_YELLOW%PyTorch 2.2.x%RESET%.
+echo.
+echo    U) Uninstall DeepSpeed.
 echo.
 echo.   OTHER
-echo    7) Generate a diagnostics file.
+echo    8) Generate a diagnostics file.
 echo.
 echo    9) %L_RED%Exit/Quit%RESET%
 echo.
 set /p WebUIOption="Enter your choice: "
 if "%WebUIOption%"=="1" goto InstallNvidiaTextGen
-if "%WebUIOption%"=="2" goto InstallOtherTextGen
+if "%WebUIOption%"=="2" goto TGGitpull
-if "%WebUIOption%"=="3" goto InstallFinetuneTextGen
 if "%WebUIOption%"=="4" goto InstallDeepSpeed118TextGen
 if "%WebUIOption%"=="5" goto InstallDeepSpeed121TextGen
-if "%WebUIOption%"=="6" goto UnInstallDeepSpeed
+if "%WebUIOption%"=="6" goto InstallDeepSpeed121TextGenPytorch221
-if "%WebUIOption%"=="7" goto GenerateDiagsTextGen
+if "%WebUIOption%"=="7" goto InstallDeepSpeed118TextGenPytorch221
+if "%WebUIOption%"=="U" goto UnInstallDeepSpeed
+if "%WebUIOption%"=="u" goto UnInstallDeepSpeed
+if "%WebUIOption%"=="8" goto GenerateDiagsTextGen
 if "%WebUIOption%"=="9" goto End
 goto WebUIMenu
 
@@ -95,27 +103,31 @@ echo    BASE REQUIREMENTS
 echo    1) Install AllTalk as a Standalone Application
 echo.
 echo    OPTIONAL
-echo    2) Delete AllTalk's custom Python environment
+echo    2) Git Pull the latest AllTalk updates from Github
-echo    3) Install the Finetuning requirements
+echo    3) Re-Apply/Update the requirements file
+echo    4) Delete AllTalk's custom Python environment
+echo    5) Purge the PIP cache
 echo.
 echo.   OTHER
-echo    4) Generate a diagnostics file
+echo    8) Generate a diagnostics file
 echo.
 echo    9) %L_RED%Exit/Quit%RESET%
 echo.
 set /p StandaloneOption="Enter your choice: "
 if "%StandaloneOption%"=="1" goto InstallCustomStandalone
-if "%StandaloneOption%"=="2" goto DeleteCustomStandalone
+if "%StandaloneOption%"=="2" goto STGitpull
-if "%StandaloneOption%"=="3" goto InstallFinetuneStandalone
+if "%StandaloneOption%"=="3" goto STReapplyrequirements
-if "%StandaloneOption%"=="4" goto GenerateDiagsStandalone
+if "%StandaloneOption%"=="4" goto STDeleteCustomStandalone
+if "%StandaloneOption%"=="5" goto STPurgepipcache
+if "%StandaloneOption%"=="8" goto GenerateDiagsStandalone
 if "%StandaloneOption%"=="9" goto End
 goto StandaloneMenu
 
 :InstallNvidiaTextGen
-pip install -r requirements_nvidia.txt
+pip install -r system\requirements\requirements_textgen.txt
 if %ERRORLEVEL% neq 0 (
     echo.
-    echo    There was an error installing the Nvidia requirements.
+    echo    There was an error installing the requirements.
     echo    Press any key to return to the menu.
     echo.
     pause
@@ -123,42 +135,24 @@ if %ERRORLEVEL% neq 0 (
 )
 Echo.
 echo.
-Echo    Nvidia machine requirements installed successfully.
+Echo    Requirements installed successfully.
 Echo. 
 pause
 goto WebUIMenu
 
-:InstallOtherTextGen
+:TGGitpull
-pip install -r requirements_other.txt
+git pull
 if %ERRORLEVEL% neq 0 (
     echo.
-    echo    There was an error installing the Other requirements.
+    echo    There was an error pulling from Github.
     echo    Press any key to return to the menu.
     echo.
     pause
     goto WebUIMenu
 )
 Echo.
-echo.
+echo     AllTalk Updated from Github. Please re-apply
-Echo    Other machine requirements installed successfully.
+echo     the latest requirements file. Option 1
-Echo. 
-pause
-goto WebUIMenu
-
-:InstallFinetuneTextGen
-pip install -r requirements_finetune.txt
-if %ERRORLEVEL% neq 0 (
-    echo.
-    echo    There was an error installing the Finetune requirements.
-    echo    Press any key to return to the menu.
-    echo.
-    pause
-    goto WebUIMenu
-)
-Echo.
-Echo    Finetune requirements installed successfully. Please ensure you have installed CUDA 11.8.
-Echo    Finetune needs access to %L_GREEN%cublas64_11.dll%RESET% and will error if this is not setup correctly.
-Echo    Please see here for details %L_GREEN%https://github.com/erew123/alltalk_tts#-finetuning-a-model%RESET%
 Echo. 
 pause
 goto WebUIMenu
@@ -227,6 +221,70 @@ del deepspeed-0.11.2+cuda121-cp311-cp311-win_amd64.whl
 pause
 goto WebUIMenu
 
+:InstallDeepSpeed121TextGenPytorch221
+echo Downloading DeepSpeed...
+curl -LO https://github.com/erew123/alltalk_tts/releases/download/DeepSpeed-14.0/deepspeed-0.14.0+ce78a63-cp311-cp311-win_amd64.whl
+if %ERRORLEVEL% neq 0 (
+    echo.
+    echo    Failed to download DeepSpeed wheel file.
+    echo    Please check your internet connection or try again later.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto WebUIMenu
+)
+echo    DeepSpeed wheel file downloaded successfully.
+echo    Installing DeepSpeed...
+pip install deepspeed-0.14.0+ce78a63-cp311-cp311-win_amd64.whl
+if %ERRORLEVEL% neq 0 (
+    echo. 
+    echo    Failed to install DeepSpeed.
+    echo    Please check if the wheel file is compatible with your system.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto WebUIMenu
+)
+Echo.
+echo.
+Echo    DeepSpeed installed successfully.
+Echo. 
+del deepspeed-0.14.0+ce78a63-cp311-cp311-win_amd64.whl
+pause
+goto WebUIMenu
+
+:InstallDeepSpeed118TextGenPytorch221
+echo Downloading DeepSpeed...
+curl -LO https://github.com/erew123/alltalk_tts/releases/download/DeepSpeed-14.0/deepspeed-0.14.0+cu118-cp311-cp311-win_amd64.whl
+if %ERRORLEVEL% neq 0 (
+    echo.
+    echo    Failed to download DeepSpeed wheel file.
+    echo    Please check your internet connection or try again later.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto WebUIMenu
+)
+echo    DeepSpeed wheel file downloaded successfully.
+echo    Installing DeepSpeed...
+pip install deepspeed-0.14.0+cu118-cp311-cp311-win_amd64.whl
+if %ERRORLEVEL% neq 0 (
+    echo. 
+    echo    Failed to install DeepSpeed.
+    echo    Please check if the wheel file is compatible with your system.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto WebUIMenu
+)
+Echo.
+echo.
+Echo    DeepSpeed installed successfully.
+Echo. 
+del deepspeed-0.14.0+cu118-cp311-cp311-win_amd64.whl
+pause
+goto WebUIMenu
+
 :UnInstallDeepSpeed
 pip uninstall deepspeed
 if %ERRORLEVEL% neq 0 (
@@ -320,49 +378,20 @@ if not exist "%INSTALL_ENV_DIR%\python.exe" ( echo. && echo Conda environment is
 @rem activate installer env
 call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%" || ( echo. && echo Miniconda hook not found. && goto end )
 
-@rem Ask user for the type of requirements to install
 echo.
-echo    %L_BLUE%Choose the type of requirements to install%RESET%
+echo     Downloading and installing PyTorch. This step can take a long time
+echo     depending on your internet connection and hard drive speed. Please
+echo     be patient.
 echo.
-echo    1. Nvidia graphics card machines
+pip install torch>=2.2.1+cu121 torchaudio>=2.2.1+cu121 --extra-index-url https://download.pytorch.org/whl/cu121
-echo    2. Other machines (mac, amd, etc)
+echo Installing other requirements.
 echo.
-set /p UserChoice="Enter your choice (1 or 2): "
+pip install -r system\requirements\requirements_standalone.txt
+curl -LO https://github.com/erew123/alltalk_tts/releases/download/DeepSpeed-12.7/deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
+echo Installing DeepSpeed...
+pip install deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
+del deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
 
-@rem Install requirements based on user choice
-if "%UserChoice%" == "1" (
-    echo.
-    echo Installing Nvidia requirements...
-    echo Downloading Pytorch with CUDA..
-    pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 --upgrade --force-reinstall --index-url https://download.pytorch.org/whl/cu121
-    pip install numpy==1.24.4
-    pip install typing-extensions>=4.8.0
-    pip install networkx<3.0.0,>=2.5.0
-    pip install fsspec>=2023.6.0
-    pip install soundfile==0.12.1
-    pip install uvicorn==0.24.0.post1
-    pip install transformers>=4.37.1
-    pip install TTS==0.21.3
-    pip install fastapi==0.104.1
-    pip install Jinja2==3.1.2
-    pip install requests==2.31.0
-    pip install tqdm==4.66.1
-    pip install importlib-metadata==4.8.1
-    pip install packaging==23.2
-    pip install pydantic==1.10.13
-    pip install sounddevice==0.4.6
-    pip install python-multipart==0.0.6
-    pip install cutlet>=0.3.0
-    pip install unidic-lite>=1.0.8
-    echo Downloading DeepSpeed...
-    curl -LO https://github.com/erew123/alltalk_tts/releases/download/DeepSpeed-12.7/deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
-    echo Installing DeepSpeed...
-    pip install deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
-    del deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
-) else (
-    echo Installing requirements for other machines...
-    pip install -r requirements_other.txt
-)
 
 @rem Create start_environment.bat to run AllTalk environment
 echo @echo off > start_environment.bat
@@ -377,14 +406,22 @@ echo set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda >> start_alltalk.bat
 echo set INSTALL_ENV_DIR=%cd%\alltalk_environment\env >> start_alltalk.bat
 echo call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%" >> start_alltalk.bat
 echo call python script.py >> start_alltalk.bat
+@rem Create start_finetune.bat to run AllTalk
+echo @echo off > start_finetune.bat
+echo cd /D "%~dp0" >> start_finetune.bat
+echo set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda >> start_finetune.bat
+echo set INSTALL_ENV_DIR=%cd%\alltalk_environment\env >> start_finetune.bat
+echo call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%" >> start_finetune.bat
+echo call python finetune.py >> start_finetune.bat
 Echo.
-Echo    start_alltalk.bat has been created.
+Echo    Run %L_YELLOW%start_alltalk.bat%RESET% to start AllTalk.
-Echo    You can now start AllTalk with %L_YELLOW%start_alltalk.bat%RESET%
+Echo    Run %L_YELLOW%start_finetune.bat%RESET% to start Finetuning.
+Echo    Run %L_YELLOW%start_environment.bat%RESET% to start the AllTalk Python environment.
 Echo.
 pause
 goto StandaloneMenu
 
-:DeleteCustomStandalone
+:STDeleteCustomStandalone
 @rem Check if the alltalk_environment directory exists
 if not exist "%cd%\alltalk_environment\" (
     echo.
@@ -418,47 +455,6 @@ Echo.
 pause
 goto StandaloneMenu
 
-:InstallFinetuneStandalone
-cd /D "%~dp0"
-set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda
-set INSTALL_ENV_DIR=%cd%\alltalk_environment\env
-@rem Check if the Conda environment exists
-if not exist "%INSTALL_ENV_DIR%\python.exe" (
-    echo.
-    echo    The Conda environment at "%INSTALL_ENV_DIR%" does not exist.
-    echo    Please install the environment before proceeding.
-    echo.
-    pause
-    goto StandaloneMenu
-)
-@rem Attempt to activate the Conda environment
-call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%"
-if errorlevel 1 (
-    echo. 
-    echo    Failed to activate the Conda environment.
-    echo    Please check your installation and try again.
-    echo.
-    pause
-    goto StandaloneMenu
-)
-@rem Proceed with installing requirements
-pip install -r requirements_finetune.txt
-if %ERRORLEVEL% neq 0 (
-    echo. 
-    echo    There was an error installing the Finetune requirements.
-    echo    Press any key to return to the menu.
-    echo.
-    pause
-    goto StandaloneMenu
-)
-Echo.
-Echo    Finetune requirements installed successfully. Please ensure you have installed CUDA 11.8.
-Echo    Finetune needs access to %L_GREEN%cublas64_11.dll%RESET% and will error if this is not setup correctly.
-Echo    Please see here for details %L_GREEN%https://github.com/erew123/alltalk_tts#-finetuning-a-model%RESET%
-Echo.
-pause
-goto StandaloneMenu
-
 :GenerateDiagsStandalone
 cd /D "%~dp0"
 set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda
@@ -499,6 +495,144 @@ Echo.
 pause
 goto StandaloneMenu
 
+:STReapplyrequirements
+cd /D "%~dp0"
+set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda
+set INSTALL_ENV_DIR=%cd%\alltalk_environment\env
+@rem Check if the Conda environment exists
+if not exist "%INSTALL_ENV_DIR%\python.exe" (
+    echo.
+    echo    The Conda environment at "%INSTALL_ENV_DIR%" does not exist.
+    echo    Please install the environment before proceeding.
+    echo. 
+    pause
+    goto StandaloneMenu
+)
+@rem Attempt to activate the Conda environment
+call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%"
+if errorlevel 1 (
+    echo. 
+    echo    Failed to activate the Conda environment.
+    echo    Please check your installation and try again.
+    echo.
+    pause
+    goto StandaloneMenu
+)
+@rem Run Reapply requirements
+echo.
+echo     Downloading and installing PyTorch. This step can take a long time
+echo     depending on your internet connection and hard drive speed. Please
+echo     be patient.
+echo.
+pip install torch>=2.2.1+cu121 torchaudio>=2.2.1+cu121 --extra-index-url https://download.pytorch.org/whl/cu121
+echo Installing other requirements.
+echo.
+pip install -r system\requirements\requirements_standalone.txt
+curl -LO https://github.com/erew123/alltalk_tts/releases/download/DeepSpeed-12.7/deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
+echo Installing DeepSpeed...
+pip install deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
+del deepspeed-0.12.7+d058d4b-cp311-cp311-win_amd64.whl
+if %ERRORLEVEL% neq 0 (
+    echo.
+    echo    There was an error.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto StandaloneMenu
+)
+Echo.
+Echo.
+Echo    Requirements have been re-applied/updated.
+Echo.
+pause
+goto StandaloneMenu
+
+:STPurgepipcache
+cd /D "%~dp0"
+set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda
+set INSTALL_ENV_DIR=%cd%\alltalk_environment\env
+@rem Check if the Conda environment exists
+if not exist "%INSTALL_ENV_DIR%\python.exe" (
+    echo.
+    echo    The Conda environment at "%INSTALL_ENV_DIR%" does not exist.
+    echo    Please install the environment before proceeding.
+    echo. 
+    pause
+    goto StandaloneMenu
+)
+@rem Attempt to activate the Conda environment
+call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%"
+if errorlevel 1 (
+    echo. 
+    echo    Failed to activate the Conda environment.
+    echo    Please check your installation and try again.
+    echo.
+    pause
+    goto StandaloneMenu
+)
+@rem Clear the PIP cache
+echo.
+echo     Purging the PIP cache of downloaded files.
+echo.
+pip cache purge
+if %ERRORLEVEL% neq 0 (
+    echo.
+    echo    There was an error.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto StandaloneMenu
+)
+Echo.
+Echo    The PIP cache has been purged.
+Echo.
+pause
+goto StandaloneMenu
+
+:STGitpull
+cd /D "%~dp0"
+set CONDA_ROOT_PREFIX=%cd%\alltalk_environment\conda
+set INSTALL_ENV_DIR=%cd%\alltalk_environment\env
+@rem Check if the Conda environment exists
+if not exist "%INSTALL_ENV_DIR%\python.exe" (
+    echo.
+    echo    The Conda environment at "%INSTALL_ENV_DIR%" does not exist.
+    echo    Please install the environment before proceeding.
+    echo. 
+    pause
+    goto StandaloneMenu
+)
+@rem Attempt to activate the Conda environment
+call "%CONDA_ROOT_PREFIX%\condabin\conda.bat" activate "%INSTALL_ENV_DIR%"
+if errorlevel 1 (
+    echo. 
+    echo    Failed to activate the Conda environment.
+    echo    Please check your installation and try again.
+    echo.
+    pause
+    goto StandaloneMenu
+)
+@rem Pull from Github
+echo.
+echo     Pulling the latest updates. Please re-apply
+echo     the latest requirements file. Option 3
+echo.
+git pull
+if %ERRORLEVEL% neq 0 (
+    echo.
+    echo    There was an error pulling from Github.
+    echo    Press any key to return to the menu.
+    echo.
+    pause
+    goto StandaloneMenu
+)
+Echo.
+echo     AllTalk Updated from Github. Please re-apply
+echo     the latest requirements file. Option 3
+Echo.
+pause
+goto StandaloneMenu
+
 :End
 echo Exiting AllTalk Setup Utility...
 exit /b

atsetup.sh

@@ -64,11 +64,12 @@ webui_menu() {
         echo -e "    in the ${L_GREEN}text-generation-webui${NC} folder and then re-run this script."
         echo
         echo "    BASE REQUIREMENTS"
-        echo -e "    1) Install the AllTalk requirements for an ${L_GREEN}Nvidia GPU machine.${NC}"
+        echo -e "    1) Apply/Re-Apply the requirements for an ${L_GREEN}Text-generation-webui${NC}"
-        echo -e "    2) Install the AllTalk requirements for an ${L_GREEN}AMD or MAC machine.${NC}"
         echo
         echo "    OPTIONAL"
-        echo "    3) Install the Finetuning requirements."
+        echo "    2) Git Pull the latest AllTalk updates from Github"
+        echo
+        echo "    DEEPSPEED"
         echo "    4) Install DeepSpeed."
         echo "    5) Uninstall DeepSpeed."
         echo
@@ -81,8 +82,7 @@ webui_menu() {
 
         case $webui_option in
             1) install_nvidia_textgen ;;
-            2) install_other_textgen ;;
+            2) tg_gitpull ;;
-            3) install_finetune_textgen ;;
             4) install_deepspeed ;;
             5) uninstall_deepspeed ;;
             6) generate_diagnostics_textgen ;;
@@ -93,52 +93,40 @@ webui_menu() {
 }
 
 install_nvidia_textgen() {
-    local requirements_file="requirements_nvidia.txt"
+    local requirements_file="system/requirements/requirements_textgen.txt"
-    echo "    Installing Finetune requirements from $requirements_file..."
+    echo "    Installing Requirements from $requirements_file..."
     if ! pip install -r "$requirements_file"; then
-        echo "    There was an error installing the Nvidia requirements."
+        echo
+        echo "    There was an error pulling from Github."
         echo "    Please check the output for details."
         read -p "    Press any key to return to the menu. " -n 1
         echo
         return
     fi
-    echo "    Nvidia requirements installed successfully."
+    echo
+    echo "    Requirements installed successfully."
+    echo
+    echo -e "    To install ${L_YELLOW}DeepSpeed${NC} on Linux, there are additional"
+    echo -e "    steps required. Please see the Github or documentation on DeepSeed."
+    echo
     read -p "    Press any key to continue. " -n 1
     echo
 }
 
-install_other_textgen() {
+tg_gitpull() {
-    local requirements_file="requirements_other.txt"
-    echo "    Installing Other machine requirements from $requirements_file..."
-    if ! pip install -r "$requirements_file"; then
-        echo "    There was an error installing the Other requirements."
-        echo "    Please check the output for details."
-        read -p "    Press any key to return to the menu. " -n 1
-        echo
-        return
-    fi
-    echo "    Other machine requirements installed successfully."
-    read -p "    Press any key to continue. " -n 1
     echo
-}
+    if ! git pull; then
-
+        echo
-install_finetune_textgen() {
+        echo "    There was an error installing the requirements."
-    local requirements_file="requirements_finetune.txt"
-    echo "    Installing Finetune requirements from $requirements_file..."
-    if ! pip install -r "$requirements_file"; then
-        echo "    There was an error installing the Finetune requirements."
         echo "    Please check the output for details."
+        echo
         read -p "    Press any key to return to the menu. " -n 1
         echo
         return
     fi
     echo
-    echo "    Finetune base requirements installed successfully."
+    echo "    AllTalk Updated from Github. Please re-apply"
-    echo
+    echo "    the latest requirements file. (Option 1)"
-    echo -e "    Finetuning requires access to ${GREEN}Cublas${NC} from the ${GREEN}Nvidia CUDA Toolkit${NC}, so will need installing."
-    echo
-    echo -e "    Instructions can be found here:"
-    echo -e "    ${GREEN}https://github.com/erew123/alltalk_tts?#-finetuning-a-model${NC}"
     echo
     read -p "    Press any key to continue. " -n 1
     echo
@@ -185,12 +173,16 @@ uninstall_deepspeed() {
     echo "Uninstalling DeepSpeed..."
     pip uninstall -y deepspeed
     if [ $? -ne 0 ]; then
+        echo
         echo "    There was an error uninstalling DeepSpeed."
+        echo
         echo "    Press any key to return to the menu."
         read -n 1
         return
     fi
+    echo
     echo "    DeepSpeed uninstalled successfully."
+    echo
     echo "    Press any key to continue."
     read -n 1
 }
@@ -198,13 +190,17 @@ uninstall_deepspeed() {
 generate_diagnostics_textgen() {
     # Run diagnostics
     if ! python diagnostics.py; then
+        echo
         echo "    There was an error running diagnostics."
+        echo
         read -p "    Press any key to return to the menu. " -n 1
         echo
         return
     fi
-
+    echo
-    echo "    Diagnostics completed successfully."
+    echo "    Diagnostics log file generated successfully."
+    echo "    Please see diagnostics.log"
+    echo
     read -p "    Press any key to continue. " -n 1
     echo
 }
@@ -220,12 +216,16 @@ standalone_menu() {
         echo "    1) Install AllTalk as a Standalone Application"
         echo
         echo "    OPTIONAL"
-        echo "    2) Delete AllTalk's custom Python environment"
+        echo "    2) Git Pull the latest AllTalk updates from Github"
-        echo "    3) Install the Finetuning requirements"
+        echo "    3) Re-Apply/Update the requirements file"
-        echo "    4) Install DeepSpeed."
+        echo "    4) Delete AllTalk's custom Python environment"
+        echo "    5) Purge the PIP cache"
+        echo
+        echo "    DEEPSPEED"
+        echo "    6) DeepSpeed Instructions/Install"
         echo
         echo "    OTHER"        
-        echo "    5) Generate a diagnostics file"
+        echo "    8) Generate a diagnostics file"
         echo
         echo -e "    9)${L_RED} Exit/Quit${NC}"
         echo
@@ -233,16 +233,19 @@ standalone_menu() {
 
         case $standalone_option in
             1) install_custom_standalone ;;
-            2) delete_custom_standalone ;;
+            2) gitpull_standalone ;;
-            3) install_finetune_standalone ;;
+            3) reapply_standalone ;;
-            4) install_deepspeed ;;
+            4) delete_custom_standalone ;;
-            5) generate_diagnostics_standalone ;;
+            5) pippurge_standalone ;;
+            6) install_deepspeed ;;
+            8) generate_diagnostics_standalone ;;
             9) exit 0 ;;
             *) echo "Invalid option"; sleep 2 ;;
         esac
     done
 }
 
+
 install_custom_standalone() {
     cd "$(dirname "${BASH_SOURCE[0]}")"
 
@@ -284,39 +287,15 @@ install_custom_standalone() {
     # Activate the environment and install requirements
     source "${CONDA_ROOT_PREFIX}/etc/profile.d/conda.sh"
     conda activate "${INSTALL_ENV_DIR}"
-
     echo
-    echo -e "    ${L_BLUE}Choose the type of requirements to install${NC}"
+    echo "    Downloading and installing PyTorch. This step can take a long time"
+    echo "    depending on your internet connection and hard drive speed. Please"
+    echo "    be patient."
+    pip install torch>=2.2.1+cu121 torchaudio>=2.2.1+cu121 --extra-index-url https://download.pytorch.org/whl/cu121
     echo
-    echo "    1. Nvidia graphics card machines"
+    echo "    Installing additional requirements."
-    echo "    2. Other machines (mac, amd, etc)"
     echo
-    read -p "    Enter your choice (1 or 2): " user_choice
+    pip install -r system/requirements/requirements_standalone.txt
-
-    if [ "$user_choice" == "1" ]; then
-    	pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 --upgrade --force-reinstall --index-url https://download.pytorch.org/whl/cu121
-    	pip install numpy==1.24.4
-	pip install typing-extensions>=4.8.0
-	pip install networkx<3.0.0,>=2.5.0
-    	pip install fsspec>=2023.6.0
-	pip install soundfile==0.12.1
-    	pip install uvicorn==0.24.0.post1
-        pip install transformers>=4.37.1
-    	pip install TTS==0.21.3
-    	pip install fastapi==0.104.1
-    	pip install Jinja2==3.1.2
-    	pip install requests==2.31.0
-    	pip install tqdm==4.66.1
-    	pip install importlib-metadata==4.8.1
-    	pip install packaging==23.2
-    	pip install pydantic==1.10.13
-    	pip install sounddevice==0.4.6
-    	pip install python-multipart==0.0.6
-        pip install cutlet>=0.3.0
-    	pip install unidic-lite>=1.0.8
-    else
-        pip install -r requirements_other.txt
-    fi
     # Create start_environment.sh to run AllTalk
     cat << EOF > start_environment.sh
 #!/bin/bash
@@ -333,27 +312,42 @@ unset PYTHONPATH
 unset PYTHONHOME
 export CUDA_PATH="$INSTALL_ENV_DIR"
 export CUDA_HOME="$CUDA_PATH"
+export LD_LIBRARY_PATH=`python3 -c 'import os; import nvidia.cublas.lib; import nvidia.cudnn.lib; print(os.path.dirname(nvidia.cublas.lib.__file__) + ":" + os.path.dirname(nvidia.cudnn.lib.__file__))'`
 # activate env
 bash --init-file <(echo "source \"$CONDA_ROOT_PREFIX/etc/profile.d/conda.sh\" && conda activate \"$INSTALL_ENV_DIR\"")
 EOF
     # Create start_alltalk.sh to run AllTalk
     cat << EOF > start_alltalk.sh
 #!/bin/bash
+export LD_LIBRARY_PATH=`python3 -c 'import os; import nvidia.cublas.lib; import nvidia.cudnn.lib; print(os.path.dirname(nvidia.cublas.lib.__file__) + ":" + os.path.dirname(nvidia.cudnn.lib.__file__))'`
 source "${CONDA_ROOT_PREFIX}/etc/profile.d/conda.sh"
 conda activate "${INSTALL_ENV_DIR}"
 python script.py
+EOF
+    # Create start_finetune.sh to run AllTalk
+    cat << EOF > start_finetune.sh
+#!/bin/bash
+export TRAINER_TELEMETRY=0
+export LD_LIBRARY_PATH=`python3 -c 'import os; import nvidia.cublas.lib; import nvidia.cudnn.lib; print(os.path.dirname(nvidia.cublas.lib.__file__) + ":" + os.path.dirname(nvidia.cudnn.lib.__file__))'`
+source "${CONDA_ROOT_PREFIX}/etc/profile.d/conda.sh"
+conda activate "${INSTALL_ENV_DIR}"
+python finetune.py
 EOF
     chmod +x start_alltalk.sh
     chmod +x start_environment.sh
+    chmod +x start_finetune.sh
     echo
     echo
-    echo -e "    start_alltalk.sh has been created."
+    echo -e "    Run ${L_YELLOW}./start_alltalk.sh${NC} to start AllTalk."
-    echo -e "    You can now start AllTalk with ${L_YELLOW}./start_alltalk.sh${NC}"
+    echo -e "    Run ${L_YELLOW}./start_finetune.sh${NC} to start Finetuning."
+    echo -e "    Run ${L_YELLOW}./start_environment.sh${NC} to start the AllTalk Python environment."
+    echo
+    echo -e "    To install ${L_YELLOW}DeepSpeed${NC} on Linux, there are additional"
+    echo -e "    steps required. Please see the Github or documentation on DeepSeed."
     echo
     read -p "    Press any key to continue. " -n 1
 }
 
-
 delete_custom_standalone() {
     local env_dir="$PWD/alltalk_environment"
     # Check if the alltalk_environment directory exists
@@ -380,47 +374,15 @@ delete_custom_standalone() {
     echo
 }
 
-install_finetune_standalone() {
-    local env_dir="$PWD/alltalk_environment"
-    local conda_root_prefix="${env_dir}/conda"
-    local install_env_dir="${env_dir}/env"
-    if [ ! -d "${install_env_dir}" ]; then
-        echo "    The Conda environment at '${install_env_dir}' does not exist."
-        echo "    Please install the environment before proceeding."
-        read -p "    Press any key to return to the menu. " -n 1
-        echo
-        return
-    fi
-    source "${conda_root_prefix}/etc/profile.d/conda.sh"
-    conda activate "${install_env_dir}"
-    local requirements_file="requirements_finetune.txt"
-    echo "Installing Finetune requirements from $requirements_file..."
-    if ! pip install -r "$requirements_file"; then
-        echo "    There was an error installing the Finetune requirements."
-        echo "    Please check the output for details."
-        read -p "    Press any key to return to the menu. " -n 1
-        echo
-        return
-    fi
-    echo
-    echo "    Finetune base requirements installed successfully."
-    echo
-    echo -e "    Finetuning requires access to ${GREEN}Cublas${NC} from the ${GREEN}Nvidia CUDA Toolkit${NC}, so will need installing."
-    echo
-    echo -e "    Instructions can be found here:"
-    echo -e "    ${GREEN}https://github.com/erew123/alltalk_tts?#-finetuning-a-model${NC}"
-    echo
-    read -p "    Press any key to continue. " -n 1
-    echo
-}
-
 generate_diagnostics_standalone() {
     local env_dir="$PWD/alltalk_environment"
     local conda_root_prefix="${env_dir}/conda"
     local install_env_dir="${env_dir}/env"
     if [ ! -d "${install_env_dir}" ]; then
+        echo
         echo "    The Conda environment at '${install_env_dir}' does not exist."
         echo "    Please install the environment before proceeding."
+        echo
         read -p "    Press any key to return to the menu. " -n 1
         echo
         return
@@ -428,16 +390,97 @@ generate_diagnostics_standalone() {
     source "${conda_root_prefix}/etc/profile.d/conda.sh"
     conda activate "${install_env_dir}"
     if ! python diagnostics.py; then
+        echo
         echo "    There was an error running diagnostics."
+        echo
         read -p "    Press any key to return to the menu. " -n 1
         echo
         return
     fi
+    echo
     echo "    Diagnostics completed successfully."
     read -p "    Press any key to continue. " -n 1
     echo
 }
 
+gitpull_standalone() {
+    local env_dir="$PWD/alltalk_environment"
+    local conda_root_prefix="${env_dir}/conda"
+    local install_env_dir="${env_dir}/env"
+    if [ ! -d "${install_env_dir}" ]; then
+        echo
+        echo "    The Conda environment at '${install_env_dir}' does not exist."
+        echo "    Please install the environment before proceeding."
+        echo
+        read -p "    Press any key to return to the menu. " -n 1
+        echo
+        return
+    fi
+    source "${conda_root_prefix}/etc/profile.d/conda.sh"
+    conda activate "${install_env_dir}"
+    if ! git pull; then
+        echo
+        echo "    There was an error pulling from Github."
+        echo
+        read -p "    Press any key to return to the menu. " -n 1
+        echo
+        return
+    fi
+    echo
+    echo "    AllTalk Updated from Github. Please re-apply."
+    echo "    the latest requirements file. (Option 3)"
+    echo
+    read -p "    Press any key to continue. " -n 1
+    echo
+}
+
+pippurge_standalone() {
+    if ! pip cache purge; then
+        echo
+        echo "    There was an error."
+        echo
+        read -p "    Press any key to return to the menu. " -n 1
+        echo
+        return
+    fi
+    echo
+    echo "    The PIP cache has been purged."
+    echo
+    read -p "    Press any key to continue. " -n 1
+    echo
+}
+
+reapply_standalone() {
+    local env_dir="$PWD/alltalk_environment"
+    local conda_root_prefix="${env_dir}/conda"
+    local install_env_dir="${env_dir}/env"
+    if [ ! -d "${install_env_dir}" ]; then
+        echo
+        echo "    The Conda environment at '${install_env_dir}' does not exist."
+        echo "    Please install the environment before proceeding."
+        echo
+        read -p "    Press any key to return to the menu. " -n 1
+        echo
+        return
+    fi
+    source "${conda_root_prefix}/etc/profile.d/conda.sh"
+    conda activate "${install_env_dir}"
+    echo
+    echo "    Downloading and installing PyTorch. This step can take a long time"
+    echo "    depending on your internet connection and hard drive speed. Please"
+    echo "    be patient."
+    pip install torch>=2.2.1+cu121 torchaudio>=2.2.1+cu121 --extra-index-url https://download.pytorch.org/whl/cu121
+    echo
+    echo "    Installing additional requirements."
+    echo
+    pip install -r system/requirements/requirements_standalone.txt
+    echo
+    echo "    Requirements have been re-applied/updated."
+    echo
+    read -p "    Press any key to continue. " -n 1
+    echo
+}
+
 
 # Start the main menu
 main_menu

diagnostics.py

@@ -8,12 +8,15 @@ try:
     import re
     import sys
     import glob
+    import site
     import textwrap
+    import torch
     import packaging.version
     import packaging.specifiers
     from packaging.specifiers import SpecifierSet
     from packaging.specifiers import SpecifierSet
     from packaging.version import parse as parse_version
+    from pathlib import Path
 except ImportError as e:
     print(f"\033[91mError importing module: {e}\033[0m\n")
     print("\033[94mPlease ensure you started the Text-generation-webUI Python environment with either\033[0m")
@@ -21,6 +24,8 @@ except ImportError as e:
     print("\033[94mfrom the text-generation-webui directory, and then try running the diagnostics again.\033[0m")
     exit(1)
 
+this_dir = Path(__file__).parent
+
 try:
     import psutil
 except ImportError:
@@ -29,24 +34,30 @@ except ImportError:
     import psutil
 
 def get_requirements_file():
-    requirements_files = glob.glob("requirements*.txt")
+    this_dir = Path(__file__).parent  # Assuming 'this_dir' is defined as the script's directory
+    requirements_dir = this_dir / 'system' / 'requirements'
+    requirements_files = list(requirements_dir.glob('requirements*.txt'))  # Using pathlib for globbing
+    
     if not requirements_files:
         print("\033[91mNo requirements files found.\033[0m")
         return None
 
     print("\033[94m\nSelect a requirements file to check against (or press Enter for default 'requirements.txt'):\033[0m\n")
-    for i, file in enumerate(requirements_files, start=1):
+    for i, file_path in enumerate(requirements_files, start=1):
-        print(f"{i}. {file}")
+        print(f"    {i}. {file_path.name}")  # Only print the filename
 
     choice = input("\nEnter the number of your choice: ")
     try:
         choice = int(choice) - 1
-        return requirements_files[choice]
+        return str(requirements_files[choice])  # Return the full path as a string
     except (ValueError, IndexError):
-        return "requirements.txt"
+        return str(requirements_dir / "requirements.txt")  # Return the default path as a string
 
 # Set up logging with filemode='w'
-logging.basicConfig(filename='diagnostics.log', filemode='w', level=logging.INFO)
+logging.basicConfig(filename='diagnostics.log',
+                    filemode='w',
+                    level=logging.INFO,
+                    format='%(message)s')  # Custom format
 
 # Function to get GPU information using subprocess
 def get_gpu_info():
@@ -56,6 +67,22 @@ def get_gpu_info():
     except FileNotFoundError:
         return "NVIDIA GPU information not available"
     
+def get_cpu_info():
+    cpu_info = {
+        'physical_cores': psutil.cpu_count(logical=False),
+        'total_cores': psutil.cpu_count(logical=True),
+        'max_frequency': psutil.cpu_freq().max
+    }
+    return cpu_info
+
+def get_disk_info():
+    disk_info = []
+    partitions = psutil.disk_partitions()
+    for p in partitions:
+        usage = psutil.disk_usage(p.mountpoint)
+        disk_info.append(f" Drive: {p.device} | Total: {usage.total / (1024 ** 3):.2f} GB | Used: {usage.used / (1024 ** 3):.2f} GB | Free: {usage.free / (1024 ** 3):.2f} GB | Type: {p.fstype}")
+    return disk_info
+
 # Function to check if a port is in use
 def is_port_in_use(port):
     for conn in psutil.net_connections():
@@ -73,6 +100,38 @@ def satisfies_wildcard(installed_version, required_version):
         return True
     return False
 
+def test_cuda():
+    cuda_available = torch.cuda.is_available()
+    if cuda_available:
+        try:
+            # Attempt to create a tensor on GPU
+            torch.tensor([1.0, 2.0]).cuda()
+            return "Success - CUDA is available and working."
+        except Exception as e:
+            return f"Fail - CUDA is available but not working. Error: {e}"
+    else:
+        return "Fail - CUDA is not available."
+
+def find_files_in_path_with_wildcard(pattern):
+    # Get the site-packages directory of the current Python environment
+    site_packages_path = site.getsitepackages()
+    found_paths = []
+    # Adjust the sub-directory based on the operating system
+    sub_directory = "nvidia/cublas"
+    if platform.system() == "Linux":
+        sub_directory = os.path.join(sub_directory, "lib")
+    else:
+        sub_directory = os.path.join(sub_directory, "bin")
+    # Iterate over each site-packages directory (there can be more than one)
+    for directory in site_packages_path:
+        # Construct the search directory path
+        search_directory = os.path.join(directory, sub_directory)
+        # Use glob to find all files matching the pattern in this directory
+        for file_path in glob.glob(os.path.join(search_directory, pattern)):
+            if os.path.isfile(file_path):  # Ensure it's a file
+                found_paths.append(file_path)
+    return found_paths
+
 # Function to log and print system information
 def log_system_info():
     # System information
@@ -88,6 +147,9 @@ def log_system_info():
 
     # Torch version
     torch_version = torch.__version__
+    cuda_test_result = test_cuda()
+    cpu_info = get_cpu_info()
+    disk_info = get_disk_info()
 
     # System RAM using psutil
     try:
@@ -123,6 +185,10 @@ def log_system_info():
     search_path = sys.path
     path_env = os.environ.get('PATH', 'N/A')
 
+    # Check for cublas
+    file_name = 'cublas64_11.*' if platform.system() == "Windows" else 'libcublas.so.11*'
+    found_paths = find_files_in_path_with_wildcard(file_name)
+
     # Compare with requirements file
     requirements_file = get_requirements_file()
     if requirements_file:
@@ -146,52 +212,72 @@ def log_system_info():
             logging.info(f"NOTE {requirements_file} not found. Skipping version checks.")
 
     # Log and print information
-    logging.info(f"OS Version: {os_version}")
+    logging.info(f"OPERATING SYSTEM:")
-    logging.info(f"Note: Windows 11 will list as build is 10.x.22xxx")
+    logging.info(f" OS Version: {os_version}")
-    logging.info(f"Torch Version: {torch_version}")
+    logging.info(f" Note: Windows 11 will list as build is 10.x.22xxx")
-    logging.info(f"System RAM: {system_ram}")
+    logging.info(f"\nHARDWARE ENVIRONMENT:")
-    logging.info(f"CUDA_HOME: {cuda_home}")
+    logging.info(f" CPU: Physical Cores: {cpu_info['physical_cores']}, Total Cores: {cpu_info['total_cores']}, Max Frequency: {cpu_info['max_frequency']} MHz")
-    logging.info(f"Port Status: {port_status}")
+    logging.info(f" System RAM: {system_ram}")
-    logging.info(f"Python Version: {platform.python_version()}")
+    logging.info(f"\nGPU INFORMATION:{gpu_info}")
-    logging.info(f"Python Version Info: {python_version_info}")
+    logging.info("DISK INFORMATION:")
-    logging.info(f"Python Executable: {python_executable}")
+    for disk in disk_info:
-    logging.info(f"Python Virtual Environment: {python_virtual_env} (Should be N/A when in Text-generation-webui Conda Python environment)")
+        logging.info(disk)
-    logging.info(f"Conda Environment: {conda_env}")
+    logging.info("\nNETWORK PORT:")
+    logging.info(f" Port Status: {port_status}")
+    logging.info("\nCUDA:")
+    logging.info(f" CUDA Working: {cuda_test_result}")
+    logging.info(f" CUDA_HOME: {cuda_home}")
+    if found_paths:
+        logging.info(f" Cublas64_11 Path: {', '.join(found_paths)}")
+    else:
+        logging.info(f" Cublas64_11 Path: Not found in any search path directories.")
+    logging.info("\nPYTHON & PYTORCH:")
+    logging.info(f" Torch Version: {torch_version}")
+    logging.info(f" Python Version: {platform.python_version()}")
+    logging.info(f" Python Version Info: {python_version_info}")
+    logging.info(f" Python Executable: {python_executable}")
+    logging.info(f" Python Virtual Environment: {python_virtual_env} (Should be N/A when in Text-generation-webui Conda Python environment)")
+    logging.info(f" Conda Environment: {conda_env}")
     logging.info("\nPython Search Path:")
     for path in search_path:
         logging.info(f"  {path}")
-    logging.info("\nOS PATH Environment Variable:")
+    logging.info("\nOS SEARCHPATH ENVIRONMENT:")
     for path in path_env.split(';'):
         logging.info(f"  {path}")
     if required_packages:  # Check if the dictionary is not empty
-        logging.info("\nPackage Versions:")
+        logging.info("\nPACKAGE VERSIONS vs REQUIREMENTS FILE:")
         max_package_length = max(len(package) for package in required_packages.keys())
         for package_name, (operator, required_version) in required_packages.items():
             installed_version = installed_packages.get(package_name, 'Not installed')
-            logging.info(f"{package_name.ljust(max_package_length)}  Required: {operator} {required_version.ljust(12)}  Installed: {installed_version}")
+            logging.info(f" {package_name.ljust(max_package_length)}  Required: {operator} {required_version.ljust(12)}  Installed: {installed_version}")
-    logging.info(f"GPU Information:\n{gpu_info}")
+    logging.info("\nPYTHON PACKAGES:")
-    logging.info("Package Versions:")
     for package, version in package_versions.items():
-        logging.info(f"{package}>= {version}")
+        logging.info(f" {package}>= {version}")
 
     # Print to screen
     print(f"\n\033[94mOS Version:\033[0m \033[92m{os_version}\033[0m")
     print(f"\033[94mOS Ver note:\033[0m \033[92m(Windows 11 will say build is 10.x.22xxx)\033[0m")
-    print(f"\033[94mCUDA_HOME:\033[0m \033[92m{cuda_home}\033[0m")
     print(f"\033[94mSystem RAM:\033[0m \033[92m{system_ram}\033[0m")
-    print(f"\033[94mPort Status:\033[0m \033[92m{port_status}\033[0m")
+    for disk in disk_info:
-    print(f"\033[94mTorch Version:\033[0m \033[92m{torch_version}\033[0m")
+        print(f"\033[94mDisk:\033[0m \033[92m{disk}\033[0m")
+    print(f"\033[94m\nGPU Information:\033[0m {gpu_info}")
+    print(f"\033[94mPort Status:\033[0m \033[92m{port_status}\033[0m\n")
+    if "Fail" in cuda_test_result:
+        print(f"\033[91mCUDA Working:\033[0m \033[91m{cuda_test_result}\033[0m")
+    else:
+        print(f"\033[94mCUDA Working:\033[0m \033[92m{cuda_test_result}\033[0m")
+    print(f"\033[94mCUDA_HOME:\033[0m \033[92m{cuda_home}\033[0m")
+    if found_paths:
+        print(f"\033[94mCublas64_11 Path:\033[0m \033[92m{', '.join(found_paths)}\033[0m")
+    else:
+        print(f"\033[94mCublas64_11 Path:\033[0m \033[91mNot found in any search path directories.\033[0m")    
+    print(f"\033[94m\nTorch Version:\033[0m \033[92m{torch_version}\033[0m")
     print(f"\033[94mPython Version:\033[0m \033[92m{platform.python_version()}\033[0m")
-    print(f"\033[94mPython Version Info:\033[0m \033[92m{python_version_info}\033[0m")
     print(f"\033[94mPython Executable:\033[0m \033[92m{python_executable}\033[0m")
-    print(f"\033[94mPython Virtual Environment:\033[0m \033[92m{python_virtual_env}\033[0m (Should be N/A when in Text-generation-webui Conda Python environment)")
     print(f"\033[94mConda Environment:\033[0m \033[92m{conda_env}\033[0m")
     print(f"\n\033[94mPython Search Path:\033[0m")
     for path in search_path:
         print(f"  {path}")
-    print(f"\n\033[94mOS Search PATH Environment Variable:\033[0m")
-    for path in path_env.split(';'):
-        print(f"  {path}")
     if required_packages:  # Check if the dictionary is not empty
         print("\033[94m\nRequirements file package comparison:\033[0m")
         max_package_length = max(len(package) for package in required_packages.keys())
@@ -214,7 +300,7 @@ def log_system_info():
             color_installed = "\033[92m" if condition_met else "\033[91m"
 
             # Print colored output
-            print(f"{package_name.ljust(max_package_length)}  Required: {color_required}{operator} {required_version.ljust(12)}\033[0m  Installed: {color_installed}{installed_version}\033[0m")
+            print(f"  {package_name.ljust(max_package_length)}  Required: {color_required}{operator} {required_version.ljust(12)}\033[0m  Installed: {color_installed}{installed_version}\033[0m")
 
         print("\nOn Nvidia Graphics cards machines, if your \033[92mInstalled\033[0m version of \033[92mTorch\033[0m and \033[92mTorchaudio\033[0m does")    
         print("not have \033[92m+cu118\033[0m (Cuda 11.8) or \033[92m+cu121\033[0m (Cuda 12.1) listed after them, you do not have CUDA")
@@ -231,9 +317,9 @@ def log_system_info():
         """)
         print(explanation.strip())
     print("")
-    print(f"GPU Information:{gpu_info}")
+    print(f"\033[94mDiagnostic log created:\033[0m \033[92mdiagnostics.log. \033[94mA brief summary of results is displayed above on\033[0m")
-    print(f"\033[94mDiagnostic log created:\033[0m \033[92mdiagnostics.log\033[0m")
+    print(f"\033[94mscreen. Please see the log file for more detail.\033[0m")
-    print(f"\033[94mPlease upload the log file with any support ticket.\033[0m")
+    print(f"\033[94m\nPlease upload the log file with any support ticket.\033[0m")
 
 if __name__ == "__main__":
     log_system_info()

finetune.py

@@ -1,6 +1,8 @@
 import argparse
 import os
 import sys
+import platform
+import site
 import tempfile
 import signal
 import gradio as gr
@@ -13,14 +15,18 @@ import random
 import gc
 import time
 import shutil
+import psutil
 import pandas
 import glob
 import json
 from pathlib import Path
 from tqdm import tqdm
 from faster_whisper import WhisperModel  
-from TTS.tts.layers.xtts.tokenizer import multilingual_cleaners
+# Use a local Tokenizer to resolve Japanese support
-
+# from TTS.tts.layers.xtts.tokenizer import multilingual_cleaners
+from system.ft_tokenizer.tokenizer import multilingual_cleaners
+import importlib.metadata as metadata
+from packaging import version
 
 # STARTUP VARIABLES 
 this_dir = Path(__file__).parent.resolve()
@@ -30,6 +36,7 @@ progress = 0
 theme = gr.themes.Default()
 refresh_symbol = '🔄'
 os.environ['TRAINER_TELEMETRY'] = '0'
+pfc_status = "pass"
 
 # Define the path to the modeldownload config file file
 modeldownload_config_file_path = this_dir / "modeldownload.json"
@@ -62,6 +69,215 @@ if finetuned_model:
     finetuned_model = all((trained_model_directory / file).exists() for file in required_files)
 basemodel_or_finetunedmodel = True
 
+#######################
+#### DIAGS for PFC ####
+#######################
+
+def check_disk_space():
+    global pfc_status
+    # Get the current working directory
+    current_directory = os.getcwd()
+    # Get the disk usage statistics for the current directory's disk
+    disk_usage = shutil.disk_usage(current_directory)
+    # Convert the free space to GB (1GB = 1 << 30 bytes)
+    free_space_gb = disk_usage.free / (1 << 30)
+    # Check if the free space is more than 18GB
+    is_more_than_18gb = free_space_gb > 18
+    disk_space_icon = "✅"
+    if not is_more_than_18gb:
+        disk_space_icon ="❌"
+        pfc_status = "fail"  # Update global status if disk space check fails
+    # Generating the markdown text for disk space check
+    disk_space_markdown = f"""
+    ### 🟩 <u>Disk Space Check</u>
+    &nbsp;&nbsp;&nbsp;&nbsp; {disk_space_icon} **Disk Space (> 18 GB):** {'' if is_more_than_18gb else 'You have less than 18GB on this disk '} {free_space_gb:.2f} GB
+    """
+    return disk_space_markdown
+
+def test_cuda():
+    global pfc_status
+    cuda_home = os.environ.get('CUDA_HOME', 'N/A')
+    cuda_available = torch.cuda.is_available()
+    if cuda_available:
+        try:
+            # Attempt to create a tensor on GPU
+            torch.tensor([1.0, 2.0]).cuda()
+            cuda_status = "CUDA is available and working."
+            cuda_icon = "✅"
+        except Exception as e:
+            cuda_status = f"CUDA is available but not working. Error: {e}"
+            cuda_icon = "❌"
+            pfc_status = "fail"  # Update global status
+    else:
+        cuda_status = "CUDA is not available."
+        pfc_status = "fail"  # Update global status
+    return cuda_status, cuda_icon, cuda_home 
+
+def find_files_in_path_with_wildcard(pattern):
+    # Get the site-packages directory of the current Python environment
+    site_packages_path = site.getsitepackages()
+    found_paths = []
+    # Adjust the sub-directory based on the operating system
+    sub_directory = "nvidia/cublas"
+    if platform.system() == "Linux":
+        sub_directory = os.path.join(sub_directory, "lib")
+    else:
+        sub_directory = os.path.join(sub_directory, "bin")
+    # Iterate over each site-packages directory (there can be more than one)
+    for directory in site_packages_path:
+        # Construct the search directory path
+        search_directory = os.path.join(directory, sub_directory)
+        # Use glob to find all files matching the pattern in this directory
+        for file_path in glob.glob(os.path.join(search_directory, pattern)):
+            if os.path.isfile(file_path):  # Ensure it's a file
+                found_paths.append(file_path)
+    return found_paths
+
+def generate_cuda_markdown():
+    global pfc_status
+    cuda_status, cuda_icon, cuda_home = test_cuda()
+    file_name = 'cublas64_11.*' if platform.system() == "Windows" else 'libcublas.so.11*'
+    found_paths = find_files_in_path_with_wildcard(file_name)
+    if found_paths:
+        found_paths_str = ' '.join(found_paths)
+        found_path_icon = '✅'
+    else:
+        found_paths_str = "cublas64_11 is not accessible."
+        found_path_icon = '❌'
+        pfc_status = "fail"  # Update global status
+        # Check if 'cu118' or 'cu121' is in the PyTorch version string
+    pytorch_version = torch.__version__
+    if 'cu118' in pytorch_version or 'cu121' in pytorch_version:
+        pytorch_cuda_version_status = ''
+        pytorch_icon = '✅'
+    else:
+        pytorch_cuda_version_status = 'Pytorch CUDA version problem '
+        pytorch_icon = '❌'
+        pfc_status = "fail"  # Update global status
+    cuda_markdown = f"""
+    ### 🟨 <u>CUDA Information</u><br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {found_path_icon} **Cublas64_11 found:** {found_paths_str}  
+    &nbsp;&nbsp;&nbsp;&nbsp; {pytorch_icon} **CUDA_HOME path:** {cuda_home}
+    """
+    pytorch_markdown = f"""
+    ### 🟦 <u>Python & Pytorch Information</u>  
+    &nbsp;&nbsp;&nbsp;&nbsp; {pytorch_icon} **PyTorch Version:** {pytorch_cuda_version_status} {torch.__version__}  
+    &nbsp;&nbsp;&nbsp;&nbsp; {cuda_icon} **CUDA is working:** {cuda_status}
+    """
+    return cuda_markdown, pytorch_markdown
+
+def get_system_ram_markdown():
+    global pfc_status
+    virtual_memory = psutil.virtual_memory()
+    total_ram_gb = virtual_memory.total / (1024 ** 3)
+    available_ram_gb = virtual_memory.available / (1024 ** 3)
+    used_ram_percentage = virtual_memory.percent
+
+    # Check if the available RAM is less than 8GB
+    warning_if_low_ram = available_ram_gb < 8
+
+    # Decide the message based on the available RAM
+    ram_status_message = "Warning" if warning_if_low_ram else ""
+    ram_status_icon = "⚠️" if warning_if_low_ram else "✅"
+
+    if torch.cuda.is_available():
+        gpu_device_id = torch.cuda.current_device()
+        gpu_device_name = torch.cuda.get_device_name(gpu_device_id) 
+        # Get the total and available memory in bytes, then convert to GB
+        gpu_total_mem_gb = torch.cuda.get_device_properties(gpu_device_id).total_memory / (1024 ** 3)
+        # gpu_available_mem_gb = (torch.cuda.get_device_properties(gpu_device_id).total_memory - torch.cuda.memory_allocated(gpu_device_id)) / (1024 ** 3)
+        # gpu_available_mem_gb = (torch.cuda.get_device_properties(gpu_device_id).total_memory - torch.cuda.memory_reserved(gpu_device_id)) / (1024 ** 3)
+        gpu_reserved_mem_gb = torch.cuda.memory_reserved(gpu_device_id) / (1024 ** 3)
+        gpu_available_mem_gb = gpu_total_mem_gb - gpu_reserved_mem_gb
+        # Check if total or available memory is less than 11 GB and set icons
+        gpu_total_status_icon = "⚠️" if gpu_total_mem_gb < 12 else "✅"
+        gpu_available_status_icon = "⚠️" if gpu_available_mem_gb < 12 else "✅"
+        gpu_status_icon = "✅"
+    else:
+        gpu_status_icon = "⚠️"
+        gpu_device_name = "Cannot detect a CUDA card"
+        gpu_total_mem_gb = "Cannot detect a CUDA card"
+        gpu_available_mem_gb = "Cannot detect a CUDA card"
+        gpu_total_status_icon = gpu_status_icon
+        gpu_available_status_icon = gpu_status_icon
+
+    system_ram_markdown = f"""
+    ### 🟪 <u>System RAM and VRAM Information</u>  <br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {ram_status_icon} **Total RAM:** {total_ram_gb:.2f} GB<br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {ram_status_icon} **Available RAM:** {ram_status_message + ' - Available RAM is less than 8 GB. You have ' if warning_if_low_ram else ''} {available_ram_gb:.2f} GB available ({used_ram_percentage:.2f}% used)<br><br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {gpu_status_icon} **GPU Name:** {gpu_device_name}<br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {gpu_total_status_icon} **GPU Total RAM:** {gpu_total_mem_gb:.2f} GB<br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {gpu_available_status_icon} **GPU Available RAM:** {gpu_available_mem_gb:.2f} GB<br>
+    """
+    return system_ram_markdown
+
+
+def check_base_model(base_model_path, files_to_download):
+    global pfc_status
+    # Assuming files_to_download is a dict with keys as filenames
+    base_model_files = list(files_to_download.keys())
+    missing_files = []
+    # Check if all base model files exist
+    for file in base_model_files:
+        file_path = this_dir / base_path / model_path / file
+        if not file_path.exists():
+            missing_files.append(file)
+            pfc_status = "fail"
+    return len(missing_files) == 0
+
+# Assuming base_model_path and files_to_download are set from the JSON config as shown above
+base_model_detected = check_base_model(base_model_path, files_to_download)
+
+def generate_base_model_markdown(base_model_detected):
+    global pfc_status
+    base_model_status = 'Base model detected' if base_model_detected else 'Base model not detected'
+    base_model_icon = '✅' if base_model_detected else '❌'
+    base_model_markdown = f"""
+    ### ⬛ <u>XTTS Base Model Detection</u>
+    &nbsp;&nbsp;&nbsp;&nbsp; {base_model_icon} **Base XTTS Model Status:** {base_model_status}
+    """
+    return base_model_markdown
+
+def check_tts_version(required_version="0.22.0"):
+    global pfc_status
+    try:
+        # Get the installed version of TTS
+        installed_version = metadata.version("tts")
+        # Check if the installed version meets the required version
+        if version.parse(installed_version) >= version.parse(required_version):
+            tts_status = f"TTS version {installed_version} is installed and meets the requirement."
+            tts_status_icon = "✅"
+        else:
+            tts_status = f"❌ Fail - TTS version {installed_version} is installed but does not meet the required version {required_version}."
+            tts_status_icon = "❌"
+            pfc_status = "fail"  # Update global status
+    except metadata.PackageNotFoundError:
+        # If TTS is not installed
+        tts_status = "TTS is not installed."
+        pfc_status = "fail"  # Update global status
+    tts_markdown = f"""
+    ### 🟥 <u>TTS Information</u><br>
+    &nbsp;&nbsp;&nbsp;&nbsp; {tts_status_icon} **TTS Version:** {tts_status}
+    """
+    return tts_markdown
+
+# Disk space check results to append to the Markdown
+disk_space_results = check_disk_space()
+cuda_results, pytorch_results = generate_cuda_markdown()
+system_ram_results = get_system_ram_markdown()
+base_model_results = generate_base_model_markdown(base_model_detected)
+tts_version_status = check_tts_version()
+
+def pfc_check_fail():
+    global pfc_status
+    if pfc_status == "fail":
+        print("[FINETUNE]")
+        print("[FINETUNE] \033[91m****** WARNING PRE-FLIGHT CHECKS FAILED ******* WARNING PRE-FLIGHT CHECKS FAILED *****\033[0m")
+        print("[FINETUNE] \033[91m* Please refer to the \033[93mPre-flight check tab \033[91mand resolve any issues before continuing. *\033[0m")
+        print("[FINETUNE] \033[91m*********** Expect errors and failures if you do not resolve these issues. ***********\033[0m")
+        print("[FINETUNE]")
+    return
+
 #####################
 #### STEP 1 BITS ####
 #####################
@@ -76,6 +292,7 @@ def create_temporary_file(folder, suffix=".wav"):
     return os.path.join(folder, unique_filename)
 
 def format_audio_list(target_language, whisper_model, out_path, eval_split_number, gradio_progress=progress):
+    pfc_check_fail()
     audio_files = [os.path.join(audio_folder, file) for file in os.listdir(audio_folder) if file.endswith(('.mp3', '.flac', '.wav'))]
     buffer=0.2
     eval_percentage = eval_split_number / 100.0
@@ -110,7 +327,8 @@ def format_audio_list(target_language, whisper_model, out_path, eval_split_numbe
     device = "cuda" if torch.cuda.is_available() else "cpu"
 
     print("[FINETUNE] Loading Whisper Model:", whisper_model)
-    asr_model = WhisperModel(whisper_model, device=device, compute_type="float16")
+    print("[FINETUNE] Model will be downloaded if its not available, which will take a few minutes.")
+    asr_model = WhisperModel(whisper_model, device=device, compute_type="float32")
 
     metadata = {"audio_file": [], "text": [], "speaker_name": []}
 
@@ -287,6 +505,8 @@ def format_audio_list(target_language, whisper_model, out_path, eval_split_numbe
     # deallocate VRAM and RAM
     del asr_model, final_eval_set, final_training_set, new_data_df, existing_metadata
     gc.collect()
+    if torch.cuda.is_available():
+        torch.cuda.empty_cache()
     existing_train_df = None
     existing_eval_df = None
     print("[FINETUNE] Train CSV:", train_metadata_path)
@@ -314,6 +534,7 @@ def basemodel_or_finetunedmodel_choice(value):
         basemodel_or_finetunedmodel = False
 
 def train_gpt(language, num_epochs, batch_size, grad_acumm, train_csv, eval_csv, output_path, max_audio_length=255995):
+    pfc_check_fail()
     #  Logging parameters
     RUN_NAME = "XTTS_FT"
     PROJECT_NAME = "XTTS_trainer"
@@ -323,6 +544,27 @@ def train_gpt(language, num_epochs, batch_size, grad_acumm, train_csv, eval_csv,
     # Set here the path that the checkpoints will be saved. Default: ./training/
     OUT_PATH = os.path.join(output_path, "training")
     print("[FINETUNE] \033[94mStarting Step 2\033[0m - Fine-tuning the XTTS Encoder")
+    print(f"[FINETUNE] \033[94mLanguage: \033[92m{language} \033[94mEpochs: \033[92m{num_epochs} \033[94mBatch size: \033[92m{batch_size}\033[0m \033[94mGrad accumulation steps: \033[92m{grad_acumm}\033[0m")
+    print(f"[FINETUNE] \033[94mTraining   : \033[92m{train_csv}\033[0m")
+    print(f"[FINETUNE] \033[94mEvaluation : \033[92m{eval_csv}\033[0m")
+    if torch.cuda.is_available():
+        torch.cuda.empty_cache()
+		# Get the current device ID
+        gpu_device_id = torch.cuda.current_device()
+        gpu_available_mem_gb = (torch.cuda.get_device_properties(gpu_device_id).total_memory - torch.cuda.memory_allocated(gpu_device_id)) / (1024 ** 3)
+        print(f"[FINETUNE] \033[94mAvailable VRAM: \033[92m{gpu_available_mem_gb:.2f} GB\033[0m")
+        if gpu_available_mem_gb < 12:
+            print(f"[FINETUNE]")
+            print(f"[FINETUNE] \033[91m****** WARNING PRE-FLIGHT CHECKS FAILED ******* WARNING PRE-FLIGHT CHECKS FAILED *****\033[0m")
+            print(f"[FINETUNE] \033[94mAvailable VRAM: \033[92m{gpu_available_mem_gb:.2f} GB\033[0m")
+            print(f"[FINETUNE] \033[94mIf you are running on a Linux system and you have 12GB's or less of VRAM, this step\033[0m")
+            print(f"[FINETUNE] \033[94mmay fail, due to not enough GPU VRAM. Windows systems will use system RAM as extended\033[0m")
+            print(f"[FINETUNE] \033[94mVRAM and so should work ok. However, Windows machines will need enough System RAM\033[0m")
+            print(f"[FINETUNE] \033[94mavailable. Please read the PFC help section available on the first tab of the web\033[0m")
+            print(f"[FINETUNE] \033[94minterface for more information.\033[0m")
+            print(f"[FINETUNE] \033[91m****** WARNING PRE-FLIGHT CHECKS FAILED ******* WARNING PRE-FLIGHT CHECKS FAILED *****\033[0m")
+            print(f"[FINETUNE]")
+
     # Create the directory
     os.makedirs(OUT_PATH, exist_ok=True)
 
@@ -347,7 +589,7 @@ def train_gpt(language, num_epochs, batch_size, grad_acumm, train_csv, eval_csv,
 
     if basemodel_or_finetunedmodel:
         # BASE XTTS model checkpoints for fine-tuning.
-        print("[FINETUNE] Starting finetuning on Base Model")
+        print("[FINETUNE] Starting finetuning on \033[92mBase Model\033[0m")
         TOKENIZER_FILE = str(this_dir / base_path / model_path / "vocab.json")
         XTTS_CHECKPOINT = str(this_dir / base_path / model_path / "model.pth")
         XTTS_CONFIG_FILE = str(this_dir / base_path / model_path / "config.json")
@@ -355,7 +597,7 @@ def train_gpt(language, num_epochs, batch_size, grad_acumm, train_csv, eval_csv,
         MEL_NORM_FILE = str(this_dir / base_path / model_path / "mel_stats.pth")
     else:
         # FINETUNED XTTS model checkpoints for fine-tuning.
-        print("[FINETUNE] Starting finetuning on Existing Finetuned Model")
+        print("[FINETUNE] Starting finetuning on \033[92mExisting Finetuned Model\033[0m")
         TOKENIZER_FILE = str(this_dir / base_path / "trainedmodel" / "vocab.json")
         XTTS_CHECKPOINT = str(this_dir / base_path / "trainedmodel" / "model.pth")
         XTTS_CONFIG_FILE = str(this_dir / base_path / "trainedmodel" / "config.json")
@@ -382,6 +624,11 @@ def train_gpt(language, num_epochs, batch_size, grad_acumm, train_csv, eval_csv,
     # define audio config
     audio_config = XttsAudioConfig(sample_rate=22050, dvae_sample_rate=22050, output_sample_rate=24000)
 
+    # Resolve Japanese threading issue
+    number_of_workers = 8
+    if language == "ja":
+        number_of_workers = 0
+
     # training parameters config
     config = GPTTrainerConfig(
         epochs=num_epochs,
@@ -396,7 +643,7 @@ def train_gpt(language, num_epochs, batch_size, grad_acumm, train_csv, eval_csv,
         batch_size=BATCH_SIZE,
         batch_group_size=48,
         eval_batch_size=BATCH_SIZE,
-        num_loader_workers=8,
+        num_loader_workers=number_of_workers,
         eval_split_max_size=256,
         print_step=50,
         plot_step=100,
@@ -470,17 +717,25 @@ def clear_gpu_cache():
     if torch.cuda.is_available():
         torch.cuda.empty_cache()
         
+def find_a_speaker_file(folder_path):
+    search_path = folder_path / "*" / "speakers_xtts.pth"
+    files = glob.glob(str(search_path), recursive=True)
+    latest_file = max(files, key=os.path.getctime, default=None)
+    return latest_file
+
 XTTS_MODEL = None
 def load_model(xtts_checkpoint, xtts_config, xtts_vocab):
+    pfc_check_fail()
     global XTTS_MODEL
     clear_gpu_cache()
     if not xtts_checkpoint or not xtts_config or not xtts_vocab:
         return "You need to run the previous steps or manually set the `XTTS checkpoint path`, `XTTS config path`, and `XTTS vocab path` fields !!"
+    xtts_speakers_pth = find_a_speaker_file(this_dir / "models")
     config = XttsConfig()
     config.load_json(xtts_config)
     XTTS_MODEL = Xtts.init_from_config(config)
     print("[FINETUNE] \033[94mStarting Step 3\033[0m Loading XTTS model!")
-    XTTS_MODEL.load_checkpoint(config, checkpoint_path=xtts_checkpoint, vocab_path=xtts_vocab, use_deepspeed=False)
+    XTTS_MODEL.load_checkpoint(config, checkpoint_path=xtts_checkpoint, vocab_path=xtts_vocab, use_deepspeed=False, speaker_file_path=xtts_speakers_pth)
     if torch.cuda.is_available():
         XTTS_MODEL.cuda()
 
@@ -511,7 +766,6 @@ def run_tts(lang, tts_text, speaker_audio_file):
 
     return "Speech generated !", out_path, speaker_audio_file
 
-
 def get_available_voices(minimum_size_kb=1200):
     voice_files = [
         voice for voice in Path(f"{this_dir}/finetune/tmp-trn/wavs").glob("*.wav")
@@ -582,7 +836,7 @@ def compact_model():
     torch.save(checkpoint, target_dir / "model.pth")
 
     # Specify the files you want to copy
-    files_to_copy = ["vocab.json", "config.json", "dvae.pth", "mel_stats.pth"]
+    files_to_copy = ["vocab.json", "config.json", "dvae.pth", "mel_stats.pth","speakers_xtts.pth", ]
 
     for file_name in files_to_copy:
         src_path = this_dir / base_path / base_model_path / file_name
@@ -640,7 +894,7 @@ def compact_lastfinetuned_model():
     torch.save(checkpoint, target_dir / "model.pth")
 
     # Specify the files you want to copy
-    files_to_copy = ["vocab.json", "config.json", "dvae.pth", "mel_stats.pth"]
+    files_to_copy = ["vocab.json", "config.json", "dvae.pth", "mel_stats.pth", "speakers_xtts.pth",]
 
     for file_name in files_to_copy:
         src_path = this_dir / base_path / base_model_path / file_name
@@ -700,7 +954,6 @@ def compact_legacy_model():
     print("[FINETUNE] model.pth created in '/finetune/'")
     return "model.pth created in '/finetune/'"
 
-
 def delete_training_data():
     # Define the folder to be deleted
     folder_to_delete = Path(this_dir / "finetune" / "tmp-trn")
@@ -725,7 +978,6 @@ def delete_training_data():
         print(f"[FINETUNE] Folder {folder_to_delete} does not exist.")
         return "Folder '/finetune/tmp-trn/' does not exist."
 
-
 def delete_voice_sample_contents():
     # Define the folder to be cleared
     folder_to_clear = Path(this_dir / "finetune" / "put-voice-samples-in-here")
@@ -748,7 +1000,6 @@ def delete_voice_sample_contents():
         print(f"[FINETUNE] Folder {folder_to_clear} does not exist.")
         return f"Folder 'put-voice-samples-in-here' does not exist."
 
-
 #######################
 #### OTHER Generic ####
 #######################
@@ -774,7 +1025,6 @@ class Logger:
 sys.stdout = Logger()
 sys.stderr = sys.stdout
 
-
 # logging.basicConfig(stream=sys.stdout, level=logging.INFO)
 import logging
 logging.basicConfig(
@@ -785,47 +1035,67 @@ logging.basicConfig(
     ]
 )
 
-
 def read_logs():
     sys.stdout.flush()
     with open(sys.stdout.log_file, "r") as f:
         return f.read()
 
-
 def cleanup_before_exit(signum, frame):
     print("[FINETUNE] Received interrupt signal. Cleaning up and exiting...")
     # Perform cleanup operations here if necessary
     sys.exit(0)
 
-
+def create_refresh_button(refresh_components, refresh_methods, elem_class, interactive=True):
-def create_refresh_button(refresh_component, refresh_method, refreshed_args, elem_class, interactive=True):
-    """
-    Copied from https://github.com/AUTOMATIC1111/stable-diffusion-webui
-    """
     def refresh():
-        refresh_method()
+        updates = {}
-        args = refreshed_args() if callable(refreshed_args) else refreshed_args
+        for component, method in zip(refresh_components, refresh_methods):
+            args = method() if callable(method) else method
+            if args and 'choices' in args:
+                # Select the most recent file (last in the sorted list)
+                args['value'] = args['choices'][-1] if args['choices'] else ""
+            for k, v in args.items():
+                setattr(component, k, v)
+            updates[component] = gr.update(**(args or {}))
+        return updates
 
-        for k, v in args.items():
+    refresh_button = gr.Button("Refresh Dropdowns", elem_classes=elem_class, interactive=interactive)
-            setattr(refresh_component, k, v)
-
-        return gr.update(**(args or {}))
-
-    refresh_button = gr.Button(refresh_symbol, elem_classes=elem_class, interactive=interactive, scale=0)
     refresh_button.click(
         fn=refresh,
         inputs=[],
-        outputs=[refresh_component]
+        outputs=refresh_components
     )
 
     return refresh_button
 
 
+
+pfc_markdown = f"""
+    ### 🚀 <u>Pre-flight Checklist for Fine-tuning</u><br>
+    ◽  <strong>Ensure</strong> each criterion is marked with a green check mark ✅ and a Pass status. <strong>Finetuning will fail otherwise.</strong><br>
+    ◽  The help tabs along the top will assist in resolving issues and you can also find additional help guides on the AllTalk [GitHub repository](https://github.com/erew123/alltalk_tts#-finetuning-a-model).<br>
+    ◽  For an overview of fine-tuning procedures, please refer to the "General Finetuning info" tab or visit the AllTalk [GitHub repository](https://github.com/erew123/alltalk_tts#-finetuning-a-model).
+    """
+
+custom_css = """
+body {
+    font-size: 16px; /* Adjust the base font size as needed */
+}
+h1, h2, h3, h4, h5, h6 {
+    font-size: 1.25em; /* Adjust heading sizes relative to the base size */
+}
+p {
+    font-size: 1.1em; /* Paragraph font size, relative to the base size */
+    margin-bottom: 10px; /* Adjust paragraph spacing */
+}
+.gradio_container {
+    zoom: 1.1; /* Adjust the zoom to scale the entire container */
+}
+"""
+
 if __name__ == "__main__":
     # Register the signal handler
     signal.signal(signal.SIGINT, cleanup_before_exit)
 
-
 ################
 #### GRADIO ####
 ################
@@ -869,63 +1139,150 @@ if __name__ == "__main__":
 #### GRADIO INFO ####
 #####################
 
-    with gr.Blocks(theme=theme) as demo:
+    with gr.Blocks(theme=theme, css=custom_css) as demo:
-        with gr.Tab("Information/Guide"):
+        with gr.Tab("🚀 PFC"):
-
+            with gr.Tab("🚀 Pre-Flight Checklist"):
-            gr.Markdown(
+                gr.Markdown(
                 f"""
-                ## <u>Finetuning Information</u><br>
+                {pfc_markdown}       
-                ### 🟥 <u>Important Note</u>
+                {disk_space_results}
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - <span style="color: #3366ff;">finetune.py</span> needs to be run from the <span style="color: #3366ff;">/alltalk_tts/</span> folder. Don't move the location of this script.
+                {system_ram_results}
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Have you run AllTalk at least once? It needs to have downloaded+updated the voice model, before we can finetune it.
+                {cuda_results}
-                ### 🟦 <u>What you need to run finetuning</u>
+                {pytorch_results}
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - An Nvidia GPU. Tested on Windows with extended shared VRAM and training used about 16GB's total (which worked on a 12GB card).
+                {base_model_results}
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If you have multiple Nvidia GPU's in your system, please see this [important note](https://github.com/erew123/alltalk_tts#-i-have-multiple-gpus-and-i-have-problems-running-finetuning).
+                {tts_version_status}
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - I have not been able to test this on a GPU with less than 12GB of VRAM, so cannot say if it will work or how that would affect performance.
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - <span style="color: red;">Version 11.8</span> of Nvidia cuBLAS and cuDNN (guide below). Only 11.8 of cuBLAS and cuDNN work for this process currently.
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Minimum <span style="color: red;">18GB</span> free disk space (most of it is used temporarily).
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Some decent quality audio, multiple files if you like. Minimum of 2 minutes and Ive tested up to 20 minutes of audio.
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - There is no major need to chop down your audio files into small slices as Step 1 will do that for you automatically and prepare the training set. Ive been testing with 5 minute long clips.
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - This process will need access to all your GPU and VRAM, so close any other software that's using your GPU currently.
-                ### 🟨 <u>Setting up cuBLAS and cuDNN <span style="color: red;">Version 11.8</span></u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If you have the <span style="color: #3366ff;;">Nvidia CUDA Toolkit Version 11.8</span> installed and can type <span style="color: #3366ff;;">nvcc --version</span> at the command prompt/terminal and it reports <span style="color: #00a000;">Cuda compilation tools, release 11.8</span> you should be good to go. 
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If you dont have the toolkit installed, the idea is just to install the smallest bit possible and this will not affect or impact other things on your system.
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - You will need to download the Nvidia Cuda Toolkit 11.8<span style="color: #3366ff;"> network</span> install from [here](https://developer.nvidia.com/cuda-11-8-0-download-archive) 
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1) Run the installer and select <span style="color: #3366ff;">Custom Advanced</span> Uncheck <span style="color: #3366ff;">everything</span> at the top then expand <span style="color: #3366ff;">CUDA</span>, <span style="color: #3366ff;">Development</span> > <span style="color: #3366ff;">Compiler</span> > and select <span style="color: #3366ff;;">nvcc</span> then expand <span style="color: #3366ff;;">Libraries</span> and select <span style="color: #3366ff;;">CUBLAS</span>.
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 2) Back at the top of <span style="color: #3366ff;">CUDA</span>, expand <span style="color: #3366ff;">Runtime</span> > <span style="color: #3366ff;">Libraries</span> and select <span style="color: #3366ff;">CUBLAS</span>. Click <span style="color: #3366ff;;">Next</span>, accept the default path (taking a note of its location) and let the install run. 
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 3) You should be able to drop to your terminal or command prompt and type <span style="color: #3366ff;">nvcc --version</span> and have it report <span style="color: #00a000;">Cuda compilation tools, release 11.8</span>. If it does you are good to go. If it doesn't > Step 4.
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 4) Linux users, you can temporarily add these paths on your current terminal window with (you may need to confirm these are correct for your flavour of Linux):
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span style="color: #3366ff;">export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64&colon;&dollar;&lbrace;LD_LIBRARY_PATH&colon;&plus;&colon;&dollar;&lbrace;LD_LIBRARY_PATH&rbrace;&rbrace;</span> (Add it to your ~/.bashrc if you want this to be permanent)
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span style="color: #3366ff;">export LD_LIBRARY_PATH=/usr/local/cuda-11.8/bin</span>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Windows users need the add the following to the PATH environment variable. Start menu and search for "Environment Variables" or "Edit the system environment variables.". 
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Find and select the "Path" variable, then click on the "Edit...". Click on the "New" button and add:
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span style="color: #3366ff;">C:&bsol;Program Files&bsol;NVIDIA GPU Computing Toolkit&bsol;CUDA&bsol;v11.8&bsol;bin.</span>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 5) Once you have these set correctly, you should be able to open a new command prompt/terminal and <span style="color: #3366ff;">nvcc --version</span> at the command prompt/terminal, resulting in <span style="color: #00a000;">Cuda compilation tools, release 11.8</span>.
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 6) If the nvcc command doesn't work OR it reports a version different from 11.8, finetuning wont work, so you will to double check your environment variables and get them working correctly.
                 """
             )
+            with gr.Tab("🟩 Disks Help"):
+                gr.Markdown(
+                f"""
+                {disk_space_results}<br><br>
+                ◽ During actual training (Step 2) Finetuning will require approximately 18GB's of free disk space while performing training and will fail or perform badly if there is any less disk space. The majority of this disk space is used temporarily and will be cleared when you reach Step 4 and move & compact the model then delete the training data.<br>
+                ◽ Because lots of data is being copied around, <strong>mechanical hard disks</strong> will be slow.
+                """
+            )
+            with gr.Tab("🟪 RAM & VRAM Help"):
+                gr.Markdown(
+                f"""
+                {system_ram_results}<br>
+                ◽ During actual training (Step 2) Finetuning will use around **14GB's** of VRAM. If your GPU doesnt have 14GB's of VRAM:<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - **Windows** systems will attempt to extend VRAM into your System RAM, and so should work.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - **Linux** systems can only use the available VRAM, so may fail on 12GB VRAM or smaller GPU's.<br>
+                ◽ For **Windows** users with 12GB or less, if you also have very low or slow System RAM you can expect bad performance or the training to fail.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - 12GB cards may need 2GB System RAM.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - 8GB cards may need 6GB System RAM.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - 6GB cards may need 8GB System RAM.<br>
+                ◽ Its hard to estimate what performance impact this could will have, due to different memory speeds, PCI speeds, GPU speeds etc.<br>
+                ◽ If you have a low VRAM scenario and also are attempting to run on a mechanical hard drive, Finetuning could take ??? amount of time.<br>
+                ◽ On Windows machines, please ensure you have **not** disabled System Memory Fallback for Stable Diffusion <a href="https://nvidia.custhelp.com/app/answers/detail/a_id/5490/~/system-memory-fallback-for-stable-diffusion" target="_blank">link here</a><br>
+                """
+            )
+            with gr.Tab("🟨 CUDA & Cublas Help"):
+                gr.Markdown(
+                f"""         
+                {cuda_results}<br><br>
+                ◽ It DOESNT matter what version of CUDA you have installed within Python either, CUDA 11.8, CUDA 12.1 etc. The NVIDIA CUDA Development Toolkit is a completly different and seperate thing from Python/PyTorch.<br>
+                ◽ Finetuning simply wants to access a tool within the CUDA Development Toolkit 11.8 called Cublas64_11.<br>
+                ◽ If you dont have the toolkit installed, the idea is just to install the smallest bit possible and this will not affect or impact other things on your system.<br><br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ You will need to download the Nvidia Cuda Toolkit 11.8<span style="color: #3366ff;"> network install</span> from <a href="https://developer.nvidia.com/cuda-11-8-0-download-archive" target="_blank">link here</a><br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ 1) Run the installer and select <span style="color: #3366ff;">Custom Advanced</span> Uncheck <span style="color: #3366ff;">everything</span> at the top then expand <span style="color: #3366ff;">CUDA</span>, <span style="color: #3366ff;">Development</span> > <span style="color: #3366ff;">Compiler</span> > and select <span style="color: #3366ff;;">nvcc</span> then expand <span style="color: #3366ff;;">Libraries</span> and select <span style="color: #3366ff;;">CUBLAS</span>.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ 2) Back at the top of <span style="color: #3366ff;">CUDA</span>, expand <span style="color: #3366ff;">Runtime</span> > <span style="color: #3366ff;">Libraries</span> and select <span style="color: #3366ff;">CUBLAS</span>. Click <span style="color: #3366ff;;">Next</span>, accept the default path (taking a note of its location) and let the install run. <br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ 3) You should be able to drop to your terminal or command prompt and type <span style="color: #3366ff;">nvcc --version</span> and have it report <span style="color: #00a000;">Cuda compilation tools, release 11.8</span>. If it does you are good to go. If it doesn't > Step 4.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ 4) <strong>Linux users</strong>, you can temporarily add these paths on your current terminal window with (you may need to confirm these are correct for your flavour of Linux):<br><br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span style="color: #3366ff;">export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64&colon;&dollar;&lbrace;LD_LIBRARY_PATH&colon;&plus;&colon;&dollar;&lbrace;LD_LIBRARY_PATH&rbrace;&rbrace;</span> (Add it to your ~/.bashrc if you want this to be permanent)<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span style="color: #3366ff;">export LD_LIBRARY_PATH=/usr/local/cuda-11.8/bin</span><br><br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <strong>Windows users</strong> need the add the following to the PATH environment variable. Start menu and search for "Environment Variables" or "Edit the system environment variables.". <br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Find and select the "Path" variable, then click on the "Edit...". Click on the "New" button and add:<br><br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <span style="color: #3366ff;">C:&bsol;Program Files&bsol;NVIDIA GPU Computing Toolkit&bsol;CUDA&bsol;v11.8&bsol;bin.</span><br><br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ 5) Once you have these set correctly, you should be able to open a new command prompt/terminal and <span style="color: #3366ff;">nvcc --version</span> at the command prompt/terminal, resulting in <span style="color: #00a000;">Cuda compilation tools, release 11.8</span>.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;◽ 6) If the nvcc command doesn't work OR it reports a version different from 11.8, finetuning wont work, so you will to double check your environment variables and get them working correctly.<br>
+                """
+            )
+            with gr.Tab("🟦 Python & PyTorch Help"):
+                gr.Markdown(
+                f"""         
+                {pytorch_results}<br><br>
+                ◽ On the PyTorch version the:<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;- first few digits are the version of PyTorch e.g. 2.1.0 is PyTorch 2.1.0<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;- last few digits refer to the CUDA version e.g. cu118 is Cuda 11.8. cu121 is Cuda 12.1.<br>
+                ◽ Ensure you have started your Python envuronment before running finetuning otherwise you will have failures on the above checks.<br>
+                ◽ If PyTorch does not show a CUDA version, then PyTorch will need reinstalling with CUDA. I would suggest running <span style="color: #3366ff;">pip cache purge</span> before installing PyTorch again.<br>
+                ◽ It DOESNT matter what version of PyTorch and CUDA you have installed within Python, CUDA 11.8, CUDA 12.1 etc. The NVIDIA CUDA Development Toolkit is a completly different and seperate thing.<br>
+                ◽ Finetuning simply wants to access a tool within the CUDA Development Toolkit called Cublas64_11.<br>
+                ◽ If you dont have the toolkit installed, the idea is just to install the smallest bit possible and this will not affect or impact other things on your system.<br>
+                """
+            )
+            with gr.Tab("⬛ XTTS Base Model Help"):
+                gr.Markdown(
+                f"""         
+                {base_model_results}<br><br>
+                ◽ If your basemodel is not being detected, please ensure that <span style="color: #3366ff;">finetune.py</span> is being run from the AllTalk main folder.<br>
+                ◽ Ensure you have started AllTalk normally at least once. You can start it again and it will download any missing files.<br>
+                ◽ Check that there is an XTTS model within the models folder e.g. <span style="color: #3366ff;">/models/xttsv2_2.0.2/</span><br>
+                ◽ The files required are "model.pth", "vocab.json", "config.json", "dvae.pth", "mel_stats.pth", "speakers_xtts.pth".
+                """
+            )
+            with gr.Tab("🟥 TTS Version Help"):
+                gr.Markdown(
+                f"""         
+                {tts_version_status}<br><br>
+                ◽ If your TTS version is showing as the incorrect version, please reinstall the Finetuning requirements at the command prompt/terminal.<br>
+                ◽ <span style="color: #3366ff;">pip install -r requirements_finetune.txt</span><br>
+                """
+            )
+                
+        with gr.Tab("ℹ️ General Finetuning info"):
+            gr.Markdown(
+            f"""
+            ### 🟥 <u>Important Note</u>
+            ◽ <span style="color: #3366ff;">finetune.py</span> needs to be run from the <span style="color: #3366ff;">/alltalk_tts/</span> folder. Don't move the location of this script.
+            ### 🟦 <u>What you need to run finetuning</u>
+            ◽ An Nvidia GPU.<br>
+            ◽ If you have multiple Nvidia GPU's in your system, please see the Github Help section [Multiple GPU's](https://github.com/erew123/alltalk_tts#performance-and-compatibility-issues).<br>
+            ◽ Some decent quality audio, multiple files if you like. Minimum of 2 minutes and Ive tested up to 20 minutes of audio.<br>
+            ◽ There is no major need to chop down your audio files into small slices as Step 1 will do that for you automatically and prepare the training set. But this can be helpful in cases where:<br>
+            &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - Whisper doesnt correctly detect and split down audio files.<br>
+            &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - You do not get a lot of "Speaker Reference Audio" files at the end of training.<br>
+            ◽ This process will need access to all your GPU and VRAM, so close any other software that's using your GPU currently.<br>
+            ### 🟨 <u>What do I do from here?</u><br>
+            ◽ Proceed through Step 1, 2, 3 and onto "What to do next".<br>
+            ### 🟩 <u>Additional Information</u><br>
+            ◽ Guidance is provided on each step of the process however, if you are after more detailed information please visit:<br>
+            &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ [AllTalk Github Finetuning](https://github.com/erew123/alltalk_tts#-finetuning-a-model)<br>
+            &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ [Coqui XTTS Documentation](https://docs.coqui.ai/en/latest/index.html)<br>
+            """
+        )
 
 #######################
 #### GRADIO STEP 1 ####
 #######################
-        with gr.Tab("Step 1 - Preparing Audio/Generating the dataset"):
+        with gr.Tab("📁 Step 1 - Generating the dataset"):
 
             gr.Markdown(
                 f"""
-                ## <u>STEP 1 - Preparing Audio/Generating the dataset</u><br>
+                ### 📁 <u>Generating the dataset</u><br>
+                ### 🟥 <u>Important Note - Windows - "UserWarning: huggingface_hub cache-system uses symlinks."</u>
+                ◽ This error is caused by Huggingfaces download software downloading the Whisper model. If you get this error, please restart your Windows command prompt with "Run as Administrator" and restart finetuning.<br>
+                ◽ This should only occur the 1st time it downloads the Whisper model. [Huggingface Reference here](https://huggingface.co/docs/huggingface_hub/en/guides/manage-cache#limitations)
                 ### 🟦 <u>What you need to do</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Please read Coqui's guide on what makes a good dataset [here](https://docs.coqui.ai/en/latest/what_makes_a_good_dataset.html#what-makes-a-good-dataset)
+                ◽ Please read Coqui's guide on what makes a good dataset [here](https://docs.coqui.ai/en/latest/what_makes_a_good_dataset.html#what-makes-a-good-dataset)<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Place your audio files in <span style="color: #3366ff;">{str(audio_folder)}</span>                
+                ◽ Place your audio files in <span style="color: #3366ff;">{str(audio_folder)}</span>          
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Your audio samples can be in the format <span style="color: #3366ff;">mp3, wav,</span> or <span style="color: #3366ff;">flac.</span>
+                ◽ Your audio samples can be in the format <span style="color: #3366ff;">mp3, wav,</span> or <span style="color: #3366ff;">flac.</span><br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - You will need a minimum of <span style="color: #3366ff;">2 minutes</span> of audio in either one or multiple audio files. Very small sample files cause errors, so I would suggest 30 seconds and longer samples. 
+                ◽ You will need a minimum of <span style="color: #3366ff;">2 minutes</span> of audio in either one or multiple audio files. 5 to 10 minutes of audio would probably be better, allowing for more varied sample data to be generated.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - When you have completed Steps 1, 2, and 3, you are welcome to delete your samples from "put-voice-samples-in-here".
+                ◽ Very small sample files cause errors, so I would recommend that the samples are at least 30 seconds and longer.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - FYI Anecdotal evidence suggests that the Whisper 2 model may yield superior results in audio splitting and dataset creation.
+                ◽ FYI Anecdotal evidence suggests that the Whisper 2 model may yield superior results in audio splitting and dataset creation.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If this step is failing, it will be worth running the diagnostics with atsetup and confirming you have cu118 or cu112 listed against your torch and torchaudio.<br>
+                ### 🟨 <u>What this step is doing</u><br>
-                ### 🟨 <u>What this step is doing</u>
+                ◽ With step one, we are going to be stripping your audio file(s) into smaller files, using Whisper to find spoken words/sentences, compile that into excel sheets of training data, ready for Step 2.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - With step one, we are going to be stripping your audio file(s) into smaller files, using Whisper to find spoken words/sentences, compile that into excel sheets of training data, ready for finetuning the model on Step 2.
+                ◽ Whisper is making a best effort to find spoken audio and break it down into smaller audio files. The content of these audio files is then transcribed into CSV fles (which you can edit in Excel or similar).<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Whilst you can choose multiple Whisper models, its best only to use the 1 model as each one is about 3GB in size and will download to your local huggingface cache on first-time use. If and when you have completed training, you wish to delete this 3GB model from your system, you are welcome to do so.
+                ◽ These files (audio and CSV) are used at the next step to train the model "this is what the audio sounds like and these are the words being spoken".<br>
-                ### 🟩 <u>How long will this take?</u>
+                ◽ If you wish to manually look at the CSV files before running the next step, you are welcome to do so and edit them as necessary. The greater the accuracy of the text the better the training will be.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - First time, it needs to download the Whisper model which is 3GB. After that a few minutes on an average 3-4 year old system.
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ <span style="color: #3366ff;">/alltalk_tts/finetune/tmp-trn/metadata_train.csv</span> Contains the list of wavs to text that will be used. This is what the model is trained with.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ <span style="color: #3366ff;">/alltalk_tts/finetune/tmp-trn/metadata_eval.csv</span> Lists the evaluation wav to text, used to generate TTS while training and evaluate the quality of generated TTS to the original sample.<br>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ <span style="color: #3366ff;">/alltalk_tts/finetune/tmp-trn/wavs/</span> These are the wavs split out from your original training samples and match the contents of the CSV files.<br>
+                ◽ Whilst you can choose multiple Whisper models, its best only to use the 1 model as each one is about 3GB in size and will download to your local huggingface cache on first-time use. <br>
+                ◽ If and when you have completed training, should you wish to delete this 3GB Whisper model from your system, you are welcome to do so.<br>
+                ### 🟩 <u>How long will this take?</u><br>
+                ◽ First time, it needs to download the Whisper model which is 3GB. After that a few minutes on an average 3-4 year old system.<br>
                 """
             )
                  
@@ -937,7 +1294,7 @@ if __name__ == "__main__":
             with gr.Row():
                 whisper_model = gr.Dropdown(
                     label="Whisper Model",
-                    value="large-v3",
+                    value="large-v2",
                     choices=[
                         "large-v3",
                         "large-v2",
@@ -1015,20 +1372,27 @@ if __name__ == "__main__":
 #######################
 #### GRADIO STEP 2 ####
 #######################
-        with gr.Tab("Step 2 - Fine-tuning the XTTS Encoder"):
+        with gr.Tab("💻 Step 2 - Training"):
 
             gr.Markdown(
                 f"""
-                ## <u>STEP 2 - Fine-tuning the XTTS Encoder</u><br>
+                ### 💻 <u>Training</u><br>                
                 ### 🟦 <u>What you need to do</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - The <span style="color: #3366ff;">Train CSV</span> and <span style="color: #3366ff;">Eval CSV</span> should already be populated. If not, just go back to Step 1 and click the button again.             
+                ◽ The <span style="color: #3366ff;">Train CSV</span> and <span style="color: #3366ff;">Eval CSV</span> should already be populated. If not, just go back to Step 1 and click "Create Dataset" again.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If for any reason it crashed, close the app and re-start the finetuning and kick it off again, clicking through setp 1 (it doesnt need to regenerate the data)
+                ◽ The default settings below are the suggested settings for most purposes, however you may choose to alter them depending on your specific use case.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - The settings below are factory and should be ok.<br>
+                ◽ There are no absolute correct settings for training. It will vary based on:<br>
-                ### 🟨 <u>What this step is doing</u>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;- What you are training (A human voice, A cartoon voice, A new language entirely etc)<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Very simply put, it's taking all our wav files generated in Step 1, along with our recorded speech in out excel documents and its training the model on that voice.
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;- How much audio you have (you may want less or more eval or epochs)<br>
-                ### 🟩 <u>How long will this take?</u>
+                ◽ The key indicator of sufficient model training is whether it sounds right to you. Coqui suggests beginning with the base settings for training. If the resulting audio doesn't meet your expectations, additional training sessions may be necessary.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - On a RTX 4070 with factory settings (as below) and 20 minutes of audio, this took 20 minutes. Again, it will vary by system and how much audio you are throwing at it. You have time to go make a coffee.
+                ◽ **Number of epochs:** Defines how many times the entire dataset is passed through the neural network; more epochs can improve learning but increase the risk of memorizing the training data.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Look at your command prompt/terminal window if you want to see what it is doing.
+                ◽ **Batch size:** Sets how many training examples are used in one iteration. Larger sizes speed up training but require more computational memory.<br>
+                ◽ **Grad(ient) accumulation steps:** Allows larger batches by spreading the computation across multiple steps, enabling training with limited hardware resources.<br>
+                ◽ **Max permitted audio size in seconds:** Limits the length of audio that the model processes. Any audio exceeding this duration will be truncated to fit the model's input constraints.<br>
+                ### 🟨 <u>What this step is doing</u><br>
+                ◽ Very simply put, it's taking all our wav files generated in Step 1, along with our recorded speech in out excel documents and its training the model on that voice e.g. listen to this audio file and this is what is being said in it, so learn to reproduce it.<br>
+                ### 🟩 <u>How long will this take?</u><br>
+                ◽ On a RTX 4070 with factory settings (as below) and 20 minutes of audio, this took 20 minutes. Again, it will vary by system and how much audio you are throwing at it. You have time to go make a coffee.<br>
+                ◽ Look at your command prompt/terminal window if you want to see what it is doing.<br>
                 """
             )
             with gr.Row():
@@ -1078,7 +1442,7 @@ if __name__ == "__main__":
                 )
                 gr.Markdown(
                 f"""
-                ##### If you have an existing finetuned model in <span style="color: #3366ff;">/models/trainedmodel/</span> you will have an option to select <span style="color: #3366ff;">Existing finetuned model</span>, allowing you to further train it. Otherwise, you will only have Base Model as an option.
+                If you have an existing finetuned model in <span style="color: #3366ff;">/models/trainedmodel/</span> you will have an option to select <span style="color: #3366ff;">Existing finetuned model</span>, allowing you to further train it. Otherwise, you will only have Base Model as an option.
                 """
                 )
 
@@ -1106,8 +1470,8 @@ if __name__ == "__main__":
                     return f"The training was interrupted due an error !! Please check the console to check the full error message! \n Error summary: {error}", "", "", "", ""
 
                 # copy original files to avoid parameters changes issues
-                os.system(f"cp {config_path} {exp_path}")
+                shutil.copy(config_path, exp_path)
-                os.system(f"cp {vocab_file} {exp_path}")
+                shutil.copy(vocab_file, exp_path)
 
                 ft_xtts_checkpoint = os.path.join(exp_path, "best_model.pth")
                 print("[FINETUNE] Model training done. Move to Step 3")
@@ -1118,21 +1482,25 @@ if __name__ == "__main__":
 #### GRADIO STEP 3 ####
 #######################
             
-        with gr.Tab("Step 3 - Inference/Testing"):
+        with gr.Tab("✅ Step 3 - Testing"):
             gr.Markdown(
                 f"""
-                ## <u>STEP 3 -  Inference/Testing</u><br>
+                ### ✅ <u>Testing</u><br>
                 ### 🟥 <u>Important Note</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - This step will error if you are using TTS version 0.22.0. Please re-run  <span style="color: #3366ff;">pip install -r requirements_finetune.txt</span> if it errors loading the model.
+                ◽ If you dont believe the dropdown lists are populated correctly, you can use the "Refresh Dropdowns" button to search the finetuning folder path and update the lists. You can also manually edit the path to the listed files.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If you dont see multiple speaker reference files and you used more than 3 minutes of speech, try refreshing the page as it may not have loaded them correctly.
+                ◽ You can use any config.json or vocab.json, as long as they match the model version you trained on. Only the XTTS checkpoint matters.<br>
+                ◽ Upon successful processing of your speech data, expect to find multiple speaker reference files if your input consisted of over 3 minutes of speech.<br>
+                ◽ Audio clips shorter than 7 seconds are automatically excluded from the reference list, as they generally do not provide sufficient length for effective TTS generation.<br>
+                ◽ Should you find a lower number of reference files than anticipated, it may be due to the initial segmentation performed by Whisper (Step 1), which can occasionally result in clips that are too brief for our criteria or indeed too long. In such cases, consider manually segmenting your audio to give Whisper a better chance at generating training data and repeating the process.<br>
                 ### 🟦 <u>What you need to do</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - The model is now trained and you are at the testing stage. Hopefully all the dropdowns should be pre-populated now.
+                ◽ **Click** the **Refresh Dropdowns** button to correctly populate the **Speaker Reference Audio** with all the available WAV samples.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - You need to <span style="color: #3366ff;">Load Fine-tuned XTTS model</span> and then select your <span style="color: #3366ff;">Speaker Reference Audio</span>. You can choose various <span style="color: #3366ff;">Speaker Reference Audios</span> to see which works best.
+                ◽ The model is now trained and you are at the testing stage. Hopefully all the dropdowns should be pre-populated now.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - All the <span style="color: #3366ff;">Speaker Reference Audios</span> in the dropdown are ones that are <span style="color: #3366ff;">8 seconds</span> long or more. You can use one of these later for your voice sample in All Talk, so remember the one you like.
+                ◽ You need to <span style="color: #3366ff;">Load Fine-tuned XTTS model</span> and then select your <span style="color: #3366ff;">Speaker Reference Audio</span>. You can choose various <span style="color: #3366ff;">Speaker Reference Audios</span> to see which works best.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Some <span style="color: #3366ff;">Speaker Reference Audios</span> are a bit long and may cause issues generating the TTS, so try different ones.
+                ◽ All the <span style="color: #3366ff;">Speaker Reference Audios</span> in the dropdown are ones that are <span style="color: #3366ff;">8 seconds</span> long or more. You can use one of these later for your voice sample in All Talk, so remember the one you like.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Once you are happy, move to the next tab for instruction on cleaning out the training data and how to copy your newly trained model over your current model.
+                ◽ Some <span style="color: #3366ff;">Speaker Reference Audios</span> are a bit long and may cause issues generating the TTS, so try different ones.<br>
+                ◽ Once you are happy, move to the next tab for instruction on cleaning out the training data and how to copy your newly trained model over your current model.<br>
                 ### 🟨 <u>What this step is doing</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Its loading the finetuned model into memory, loading a voice sample and generating TTS, so that you can test out how well fine tuning worked.
+                ◽ Its loading the finetuned model into memory, loading a voice sample and generating TTS, so that you can test out how well fine tuning worked.
                 """
             )
 
@@ -1140,21 +1508,21 @@ if __name__ == "__main__":
                 with gr.Column() as col1:
                     xtts_checkpoint = gr.Dropdown(
                         [str(file) for file in xtts_checkpoint_files],
-                        label="XTTS checkpoint path:",
+                        label="XTTS checkpoint path (best_model.pth):",
                         value="",
                         allow_custom_value=True,
                     )
 
                     xtts_config = gr.Dropdown(
                         [str(file) for file in xtts_config_files],
-                        label="XTTS config path:",
+                        label="XTTS config path (config.json):",
                         value="",
                         allow_custom_value=True,
                     )
 
                     xtts_vocab = gr.Dropdown(
                         [str(file) for file in xtts_vocab_files],
-                        label="XTTS vocab path:",
+                        label="XTTS vocab path (vocab.json):",
                         value="",
                         allow_custom_value=True,
                     )
@@ -1176,36 +1544,41 @@ if __name__ == "__main__":
                             allow_custom_value=True,  # Allow custom values
                             scale=1,
                         )
+
+                    with gr.Row():
+                        tts_language = gr.Dropdown(
+                            label="Language",
+                            value="en",
+                            choices=[
+                                "en",
+                                "es",
+                                "fr",
+                                "de",
+                                "it",
+                                "pt",
+                                "pl",
+                                "tr",
+                                "ru",
+                                "nl",
+                                "cs",
+                                "ar",
+                                "zh",
+                                "hu",
+                                "ko",
+                                "ja",
+                            ]
+                        )
                         # Create refresh button
                         refresh_button = create_refresh_button(
-                            speaker_reference_audio,
+                            [xtts_checkpoint, xtts_config, xtts_vocab, speaker_reference_audio],
-                            lambda: None,  # Specify your refresh logic if needed
+                            [
-                            lambda: {"choices": get_available_voices(), "value": ""},  # Update choices and value
+                                lambda: {"choices": find_models(main_directory, "pth"), "value": ""},
-                            "refresh-button",
+                                lambda: {"choices": find_jsons(main_directory, "config.json"), "value": ""},
+                                lambda: {"choices": find_jsons(main_directory, "vocab.json"), "value": ""},
+                                lambda: {"choices": get_available_voices(), "value": ""},
+                            ],
+                            elem_class="refresh-button-class"
                         )
-
-                    tts_language = gr.Dropdown(
-                        label="Language",
-                        value="en",
-                        choices=[
-                            "en",
-                            "es",
-                            "fr",
-                            "de",
-                            "it",
-                            "pt",
-                            "pl",
-                            "tr",
-                            "ru",
-                            "nl",
-                            "cs",
-                            "ar",
-                            "zh",
-                            "hu",
-                            "ko",
-                            "ja",
-                        ]
-                    )
                     tts_text = gr.Textbox(
                         label="Input Text:",
                         value="I've just fine tuned a text to speech language model and this is how it sounds. If it doesn't sound right, I will try a different Speaker Reference Audio file.",
@@ -1220,22 +1593,25 @@ if __name__ == "__main__":
                     tts_output_audio = gr.Audio(label="TTS Generated Audio.")
                     reference_audio = gr.Audio(label="Speaker Reference Audio Used.")
 
-        with gr.Tab("What to do next"):
+        with gr.Tab("🔜 What to do next"):
             gr.Markdown(
                 f"""
-                ## <u>What to do next</u><br>
+                ### 🔜 <u>What to do next</u><br>
                 ### 🟦 <u>What you need to do</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - You have a few options below:
+                You have a few options below:<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 🔹 Compact &amp; move model. This will compress the raw finetuned model and move it, along with any large enough wav files created to one of two locations (please read text next to the button).<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; You can use the wav files in that location as sample voices with your model. Models moved to <span style="color: #3366ff;">/models/trainedmodel/</span> will be loadable within Text-gen-webui on the <span style="color: #3366ff;">XTTSv2 FT</span> loader.
+                ◽ **Compact &amp; move model:** This will compress the raw finetuned model and move it, along with any large enough wav files created to one of two locations (please read text next to the button). You can use the wav files in that location as sample voices with your model. Models moved to <span style="color: #3366ff;">/models/trainedmodel/</span> will be loadable within Text-gen-webui on the <span style="color: #3366ff;">XTTSv2 FT</span> loader.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 🔹 Delete Generated Training data. This will delete the finetuned model and any other training data generated during this process. You will more than likely want to do this after you have Compacted &amp; moved the model.
+                ◽ **Delete Generated Training data:** This will delete the finetuned model and any other training data generated during this process. You will more than likely want to do this after you have Compacted &amp; moved the model.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 🔹 Delete original voice samples. This will delete the voice samples you placed in put-voice-samples-in-here (if you no longer have a use for them).
+                ◽ **Delete original voice samples:** This will delete the voice samples you placed in put-voice-samples-in-here (if you no longer have a use for them).<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 🔹 Compress a legacy finetuned model. For any models that were generated pre the current version of finetuning, you can still compact them down.
+                ◽ **Note:** Sample wav files will be copied into the model folder. You will need to MANUALLY copy/move the WAV file(s) you want to use to the AllTalk voices folder, to make them available within the interface.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - Instructions on using the finetuned model in Text-generation-webui and also further tuning this model you just created are on <span style="color: #3366ff;">OPTION A - Compact and move model to '/trainedmodel/'</span>.
+                ◽ **Note:** Text-generation-webui users, <span style="color: #3366ff;">OPTION A - Compact and move model to '/trainedmodel/'</span>. will make the model available for use within Text-gen-webui.<br>
+                ### 🟩 <u>Using your Finetuned model as you main model</u>
+                ◽ If you wish to use the model you have just fintuned as your main model, you will need to MANUALLY copy it over the base model in <span style="color: #3366ff;">/models/trainedmodel/</span> after you have used Option A or B.<br>
+                ◽ Should you ever wish to revert your model back to the base model, you can delete the files in <span style="color: #3366ff;">/models/xttsv2_2.0.2/</span> and new copies will be re-downloaded when AllTalk starts up.<br>
                 ### 🟨 <u>Clearing up disk space</u>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - If you are not going to train anything again, you can delete the whisper model from inside of your huggingface cache (3GB approx) 
+                ◽ If you are not going to train anything again, you can delete the whisper model from inside of your huggingface cache (3GB approx) <br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Linux <span style="color: #3366ff;">~/.cache/huggingface/hub/(folder-here)</span>
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ **Linux:** <span style="color: #3366ff;">~/.cache/huggingface/hub/(folder-here)</span><br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Windows <span style="color: #3366ff;">C:&bsol;users&lpar;your-username&rpar;&bsol;.cache&bsol;huggingface&bsol;hub&bsol;(folder-here)</span>.
+                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ◽ **Windows:** <span style="color: #3366ff;">C:&bsol;users&lpar;your-username&rpar;&bsol;.cache&bsol;huggingface&bsol;hub&bsol;(folder-here)</span>.<br>
-                #### &nbsp;&nbsp;&nbsp;&nbsp; - You can also uninstall the Nvidia CUDA 11.8 Toolkit if you wish and remove your environment variable entries.
+                ◽ You can also uninstall the Nvidia CUDA 11.8 Toolkit if you wish and remove your environment variable entries.<br>
                 """
             )
             final_progress_data = gr.Label(
@@ -1245,35 +1621,28 @@ if __name__ == "__main__":
                 compact_btn = gr.Button(value="OPTION A - Compact and move model to '/trainedmodel/'")
                 gr.Markdown(
                 f"""
-                ##### This will compact the model and move it, along with the reference wav's to <span style="color: #3366ff;">/models/trainedmodel/</span> and will <span style="color: red;">OVERWRITE</span> any model in that location. This model will become available to load in Text-gen-webui with the <span style="color: #3366ff;">XTTSv2 FT</span> loader. It will also be detected by Finetune as a model that can be further trained (Reload finetune.py).
+                This will compact the model and move it, along with the reference wav's to <span style="color: #3366ff;">/models/trainedmodel/</span> and will <span style="color: red;">OVERWRITE</span> any model in that location. This model will become available to load in Text-gen-webui with the <span style="color: #3366ff;">XTTSv2 FT</span> loader. It will also be detected by Finetune as a model that can be further trained (Reload finetune.py).
                 """
                 )
             with gr.Row():
                 compact_lastfinetuned_btn = gr.Button(value="OPTION B - Compact and move model to '/lastfinetuned/'")
                 gr.Markdown(
                 f"""
-                ##### This will compact the model and move it, along with the reference wav's to <span style="color: #3366ff;">/models/lastfinetuned/</span> and will <span style="color: red;">OVERWRITE</span> any model in that location. This will just leave you a finetuned model that you can do with as you please. It will <span style="color: red;">NOT</span> be available in Text-gen-webui unless you copy it over one of the other model loader locations.
+                This will compact the model and move it, along with the reference wav's to <span style="color: #3366ff;">/models/lastfinetuned/</span> and will <span style="color: red;">OVERWRITE</span> any model in that location. This will just leave you a finetuned model that you can do with as you please. It will <span style="color: red;">NOT</span> be available in Text-gen-webui unless you copy it over one of the other model loader locations.
                 """
                 )
             with gr.Row():
                 delete_training_btn = gr.Button(value="Delete generated training data")
                 gr.Markdown(
                 f"""
-                ##### This will <span style="color: red;">DELETE</span> your training data and the raw finetuned model from <span style="color: #3366ff;">/finetune/tmp-trn/</span>.
+                This will <span style="color: red;">DELETE</span> your training data and the raw finetuned model from <span style="color: #3366ff;">/finetune/tmp-trn/</span>.
                 """
                 )
             with gr.Row():
                 delete_voicesamples_btn = gr.Button(value="Delete original voice samples")
                 gr.Markdown(
                 f"""
-                ##### This will <span style="color: red;">DELETE</span> your original voice samples from <span style="color: #3366ff;">/finetune/put-voice-samples-in-here/</span>.
+                This will <span style="color: red;">DELETE</span> your original voice samples from <span style="color: #3366ff;">/finetune/put-voice-samples-in-here/</span>.
-                """
-                )
-            with gr.Row():
-                compact_legacy_btn = gr.Button(value="Compact a legacy finetuned model")
-                gr.Markdown(
-                f"""
-                ##### This will compact your old finetuned models from 5GB to 1.9GB. Place your 5GB <span style="color: #3366ff;">base_model.pth</span> file in <span style="color: #3366ff;">/finetune/</span> (rename your file to <span style="color: #3366ff;">base_model.pth</span> if necessary). This will generate you a <span style="color: #3366ff;">model.pth</span> file.
                 """
                 )
 
@@ -1344,10 +1713,6 @@ if __name__ == "__main__":
                 fn=delete_voice_sample_contents,
                 outputs=[final_progress_data],
             )
-            compact_legacy_btn.click(
-                fn=compact_legacy_model,
-                outputs=[final_progress_data],
-            )
             model_to_train.change(basemodel_or_finetunedmodel_choice, model_to_train, None)
 
     demo.queue().launch(

finetune/compact.py

@@ -1,26 +0,0 @@
-import torch
-import os
-MODEL_PATH = "best_model.pth"
-# Check if best_model.pth exists
-if not os.path.exists(MODEL_PATH):
-    print(f"\nCannot find \033[94m{MODEL_PATH}\033[0m in the current folder.")
-    
-    # Check if model.pth exists as an alternative
-    ALTERNATIVE_PATH = "model.pth"
-    if os.path.exists(ALTERNATIVE_PATH):
-        print(f"\nIf your model is currently called \033[93m{ALTERNATIVE_PATH}\033[0m please rename it to \033[94mbest_model.pth\033[0m.")
-    else:
-        print(f"Please put \033[91mbest_model.pth\033[0m in the current folder or rename your model file to \033[94mbest_model.pth\033[0m.")
-    exit(1)
-# Attempt to load the model
-try:
-    checkpoint = torch.load(MODEL_PATH, map_location=torch.device("cpu"))
-except Exception as e:
-    print("Error loading checkpoint:", e)
-    raise
-del checkpoint["optimizer"]
-for key in list(checkpoint["model"].keys()):
-    if "dvae" in key:
-        del checkpoint["model"][key]
-torch.save(checkpoint, "model.pth")
-print(f"\nYour compacted model is created \033[93mmodel.pth\033[0m.")

modeldownload.py

@@ -45,18 +45,18 @@ else:
     )
 
 # Read the version specifier from requirements_nvidia.txt
-with open(this_dir / "requirements_nvidia.txt", "r") as req_file:
+with open(this_dir / "system" / "requirements" / "requirements_standalone.txt", "r") as req_file:
     requirements = req_file.readlines()
 
 tts_version_required = None
 for req in requirements:
-    if req.startswith("TTS>="):
+    if req.startswith("TTS=="):
-        tts_version_required = req.strip().split(">=")[1]
+        tts_version_required = req.strip().split("==")[1]
         break
 
 if tts_version_required is None:
     raise ValueError(
-        f"[{params['branding']}Startup] \033[91mWarning\033[0m Could not find TTS version specifier in requirements_nvidia.txt"
+        f"[{params['branding']}Startup] \033[91mWarning\033[0m Could not find TTS version specifier in requirements file"
     )
 
 

requirements_finetune.txt

@@ -1,7 +0,0 @@
-gradio>=3.50.2
-torch>=2.1.0+cu118
-torchaudio>=2.1.0+cu118
-TTS==0.21.3
-faster-whisper>=0.10.0
-tqdm>=4.66.0
-pandas>=1.5.0

requirements_other.txt

@@ -1,18 +0,0 @@
-numpy>=1.22.0
-soundfile>=0.12.1
-uvicorn>=0.24.0.post1
-transformers>=4.37.1
-TTS>=0.21.3
-torch>=2.1.0
-torchaudio>=2.1.0
-fastapi>=0.104.1
-Jinja2>=3.1.2
-requests>=2.31.0
-tqdm>=4.66.1
-importlib-metadata>=4.8.1
-packaging>=23.2
-pydantic>=1.10.13
-sounddevice>=0.4.6
-python-multipart>=0.0.6
-cutlet>=0.3.0
-unidic-lite>=1.0.8

script.py

@@ -85,7 +85,7 @@ except ModuleNotFoundError:
 #### STARTUP VARIABLES ####
 ###########################
 # STARTUP VARIABLE - Import languges file for Gradio to be able to display them in the interface
-with open(this_dir / "languages.json", encoding="utf8") as f:
+with open(this_dir / "system" / "config" / "languages.json", encoding="utf8") as f:
     languages = json.load(f)
 # Create a global lock
 process_lock = threading.Lock()
@@ -347,26 +347,26 @@ def send_reload_request(tts_method):
                 params["tts_method_xtts_local"] = False
                 params["tts_method_api_tts"] = True
                 params["deepspeed_activate"] = False
-                audio_path = this_dir / "templates" / "apitts.wav"
+                audio_path = this_dir / "system" / "at_sounds" / "apitts.wav"
                 tts_method_xtts_ft = False
             elif tts_method == "API Local":
                 params["tts_method_api_tts"] = False
                 params["tts_method_xtts_local"] = False
                 params["tts_method_api_local"] = True
                 params["deepspeed_activate"] = False
-                audio_path = this_dir / "templates" / "apilocal.wav"
+                audio_path = this_dir / "system" / "at_sounds" / "apilocal.wav"
                 tts_method_xtts_ft = False
             elif tts_method == "XTTSv2 Local":
                 params["tts_method_api_tts"] = False
                 params["tts_method_api_local"] = False
                 params["tts_method_xtts_local"] = True
-                audio_path = this_dir / "templates" / "xttslocal.wav"
+                audio_path = this_dir / "system" / "at_sounds" / "xttslocal.wav"
                 tts_method_xtts_ft = False
             elif tts_method == "XTTSv2 FT":
                 params["tts_method_api_tts"] = False
                 params["tts_method_api_local"] = False
                 params["tts_method_xtts_local"] = False
-                audio_path = this_dir / "templates" / "xttsfinetuned.wav"
+                audio_path = this_dir / "system" / "at_sounds" / "xttsfinetuned.wav"
                 tts_method_xtts_ft = True
         return f'<audio src="file/{audio_path}" controls autoplay></audio>'
     except requests.exceptions.RequestException as e:
@@ -385,9 +385,9 @@ def send_lowvram_request(low_vram):
     try:
         params["tts_model_loaded"] = False
         if low_vram:
-            audio_path = this_dir / "templates" / "lowvramenabled.wav"
+            audio_path = this_dir / "system" / "at_sounds" / "lowvramenabled.wav"
         else:
-            audio_path = this_dir / "templates" / "lowvramdisabled.wav"
+            audio_path = this_dir / "system" / "at_sounds" / "lowvramdisabled.wav"
         url = f"{base_url}/api/lowvramsetting?new_low_vram_value={low_vram}"
         headers = {"Content-Type": "application/json"}
         response = requests.post(url, headers=headers)
@@ -414,9 +414,9 @@ def send_deepspeed_request(deepspeed_param):
     try:
         params["tts_model_loaded"] = False
         if deepspeed_param:
-            audio_path = this_dir / "templates" / "deepspeedenabled.wav"
+            audio_path = this_dir / "system" / "at_sounds" / "deepspeedenabled.wav"
         else:
-            audio_path = this_dir / "templates" / "deepspeeddisabled.wav"
+            audio_path = this_dir / "system" / "at_sounds" / "deepspeeddisabled.wav"
 
         url = f"{base_url}/api/deepspeed?new_deepspeed_value={deepspeed_param}"
         headers = {"Content-Type": "application/json"}
@@ -511,7 +511,7 @@ def combine(audio_files, output_folder, state):
 ################################
 # PREVIEW VOICE - Generate Random Sentence if Voice Preview box is empty
 def random_sentence():
-    with open(this_dir / "harvard_sentences.txt") as f:
+    with open(this_dir / "system" / "config" / "harvard_sentences.txt") as f:
         return random.choice(list(f))
 
 
@@ -557,6 +557,10 @@ def voice_preview(string):
         return f"[{params['branding']}Server] Audio generation failed. Status: {generate_response.get('status')}"
 
 
+#######################
+#### TEXT CLEANING ####
+#######################
+
 def process_text(text):
     # Normalize HTML encoded quotes
     text = html.unescape(text)
@@ -606,6 +610,30 @@ def process_text(text):
     return ordered_parts
 
 
+########################
+#### IMAGE CLEANING ####
+########################
+
+img_pattern = r'<img[^>]*src\s*=\s*["\'][^"\'>]+["\'][^>]*>'
+
+def extract_and_remove_images(text):
+    """
+    Extracts all image data from the text and removes it for clean TTS processing.
+    Returns the cleaned text and the extracted image data.
+    """
+    img_matches = re.findall(img_pattern, text)
+    img_info = "\n".join(img_matches)  # Store extracted image data
+    cleaned_text = re.sub(img_pattern, '', text)  # Remove images from text
+    return cleaned_text, img_info
+
+def reinsert_images(text, img_info):
+    """
+    Reinserts the previously extracted image data back into the text.
+    """
+    if img_info:  # Check if there are images to reinsert
+        text += f"\n\n{img_info}"
+    return text
+
 #################################
 #### TTS STANDARD GENERATION ####
 #################################
@@ -613,12 +641,12 @@ def process_text(text):
 def output_modifier(string, state):
     if not params["activate"]:
         return string
-    original_string = string
+    img_info = ""
-    # print("ORIGINAL STRING IS:", original_string)
+    cleaned_text, img_info = extract_and_remove_images(string)
-    cleaned_string = before_audio_generation(string, params)
+    # print("Cleaned STRING IS:", cleaned_text)
+    cleaned_string = before_audio_generation(cleaned_text, params)
     if cleaned_string is None:
         return
-    string = cleaned_string
     language_code = languages.get(params["language"])
     temperature = params["local_temperature"]
     repetition_penalty = params["local_repetition_penalty"]
@@ -627,7 +655,7 @@ def output_modifier(string, state):
     if process_lock.acquire(blocking=False):
         try:
             if params["narrator_enabled"]:
-                processed_parts = process_text(original_string)
+                processed_parts = process_text(cleaned_string)
 
                 audio_files_all_paragraphs = []
                 for part_type, part in processed_parts:
@@ -686,7 +714,7 @@ def output_modifier(string, state):
                 )
             else:
                 # Decode HTML entities first
-                cleaned_part = html.unescape(original_string)
+                cleaned_part = html.unescape(cleaned_string)
                 # Replace multiple instances of certain punctuation marks with a single instance
                 cleaned_part = re.sub(r"([!?.])\1+", r"\1", cleaned_part)
                 # Further clean to remove any other unwanted characters
@@ -729,25 +757,15 @@ def output_modifier(string, state):
         if audio_path:
             # Handle Gradio and playback
             autoplay = "autoplay" if params["autoplay"] else ""
-            string = (
+            string = (f'<audio src="file/{final_output_file}" controls {autoplay}></audio>')
-                f'<audio src="file/{final_output_file}" controls {autoplay}></audio>'
-            )
-
             if params["show_text"]:
-                string += f"\n\n{original_string}"
+                string += reinsert_images(cleaned_string, img_info)
                 shared.processing_message = "*Is typing...*"
-
             return string
         else:
-            print(
+            print(f"[{params['branding']}Server] \033[91mWarning\033[0m No audio path in the response.")
-                f"[{params['branding']}Server] \033[91mWarning\033[0m No audio path in the response."
-            )
     else:
-        print(
+        print(f"[{params['branding']}Server] \033[91mWarning\033[0m Audio generation failed. Status:", generate_response.get("message"),)
-            f"[{params['branding']}Server] \033[91mWarning\033[0m Audio generation failed. Status:",
-            generate_response.get("message"),
-        )
-
 
 def get_output_filename(state):
     if "character_menu" in state:

system/at_admin/at_settings.html

@@ -89,25 +89,25 @@
     <h3>AllTalk Startup Settings</h3><br>
     <form action="/update-settings" method="post">
 
-        <label for="activate" style="color: #33cc33;">Activate TTS:</label>
+        <label for="activate" style="color: #b273fa;">TG-webui Activate TTS:</label>
         <select id="activate" name="activate" required="">
             <option {% if data.activate %}selected{% endif %} value="true">True</option>
             <option {% if not data.activate %}selected{% endif %} value="false">False</option>
         </select>
 
-        <label for="autoplay" style="color: #33cc33;">&nbsp; &nbsp; Autoplay TTS output:</label>
+        <label for="autoplay" style="color: #b273fa;">&nbsp; &nbsp; TG-webui Autoplay TTS:</label>
         <select id="autoplay" name="autoplay" required="">
             <option {% if data.autoplay %}selected{% endif %} value="true">True</option>
             <option {% if not data.autoplay %}selected{% endif %} value="false">False</option>
         </select>
 
-        <label for="show_text" style="color: #33cc33;"> Show Text on generation:</label>
+        <label for="show_text" style="color: #b273fa;">TG-webui Show Text:</label>
         <select id="show_text" name="show_text" required="">
             <option {% if data.show_text %}selected{% endif %} value="true">True</option>
             <option {% if not data.show_text %}selected{% endif %} value="false">False</option>
         </select>
 
-        <label for="voice" style="color: #33cc33;">&nbsp; &nbsp; Default voice:</label>
+        <label for="voice" style="color: #b273fa;">&nbsp; &nbsp;  TG-webui Default voice:</label>
         <select id="voice" name="voice" required="">
             {% for file in wav_files %}
             <option value="{{ file }}" {% if data.voice==file %}selected{% endif %}>{{ file }}</option>
@@ -115,7 +115,7 @@
         </select>
 
 
-        <label for="language" style="color: #33cc33;">Default Language:</label>
+        <label for="language" style="color: #b273fa;">TG-webui Default Language:</label>
         <select id="language" name="language" required="">
             <option value="Arabic" {% if data.language=='Arabic' %}selected{% endif %}>Arabic</option>
             <option value="Chinese" {% if data.language=='Chinese' %}selected{% endif %}>Chinese</option>
@@ -135,20 +135,20 @@
             <option value="Turkish" {% if data.language=='Turkish' %}selected{% endif %}>Turkish</option>
         </select>
 
-        <label for="narrator_voice" style="color: #33cc33;">&nbsp; &nbsp; Narrator voice:</label>
+        <label for="narrator_voice" style="color: #b273fa;">&nbsp; &nbsp; TG-webui Narrator voice:</label>
         <select id="narrator_voice" name="narrator_voice" required="">
             {% for file in wav_files %}
             <option value="{{ file }}" {% if data.narrator_voice==file %}selected{% endif %}>{{ file }}</option>
             {% endfor %}
         </select>
 
-        <label for="remove_trailing_dots" style="color: #33cc33;"> Remove Trailing Dots:</label>
+        <label for="remove_trailing_dots" style="color: #b273fa;">TG-webui Remove Trailing Dots:</label>
         <select id="remove_trailing_dots" name="remove_trailing_dots" required="">
             <option {% if data.remove_trailing_dots %}selected{% endif %} value="true">True</option>
             <option {% if not data.remove_trailing_dots %}selected{% endif %} value="false">False</option>
         </select>
 
-        <label for="narrator_enabled" style="color: #33cc33;">&nbsp; &nbsp; Narrator enabled:</label>
+        <label for="narrator_enabled" style="color: #b273fa;">&nbsp; &nbsp; TG-webui Narrator enabled:</label>
         <select id="narrator_enabled" name="narrator_enabled" required="">
             <option {% if data.narrator_enabled %}selected{% endif %} value="true">True</option>
             <option {% if not data.narrator_enabled %}selected{% endif %} value="false">False</option>

system/ft_tokenizer/tokenizer.py

@@ -0,0 +1,869 @@
+import os
+import re
+import textwrap
+from functools import cached_property
+
+import pypinyin
+import torch
+from hangul_romanize import Transliter
+from hangul_romanize.rule import academic
+from num2words import num2words
+from spacy.lang.ar import Arabic
+from spacy.lang.en import English
+from spacy.lang.es import Spanish
+from spacy.lang.ja import Japanese
+from spacy.lang.zh import Chinese
+from tokenizers import Tokenizer
+
+from TTS.tts.layers.xtts.zh_num2words import TextNorm as zh_num2words
+
+
+def get_spacy_lang(lang):
+    if lang == "zh":
+        return Chinese()
+    elif lang == "ja":
+        return Japanese()
+    elif lang == "ar":
+        return Arabic()
+    elif lang == "es":
+        return Spanish()
+    else:
+        # For most languages, Enlish does the job
+        return English()
+
+
+def split_sentence(text, lang, text_split_length=250):
+    """Preprocess the input text"""
+    text_splits = []
+    if text_split_length is not None and len(text) >= text_split_length:
+        text_splits.append("")
+        nlp = get_spacy_lang(lang)
+        nlp.add_pipe("sentencizer")
+        doc = nlp(text)
+        for sentence in doc.sents:
+            if len(text_splits[-1]) + len(str(sentence)) <= text_split_length:
+                # if the last sentence + the current sentence is less than the text_split_length
+                # then add the current sentence to the last sentence
+                text_splits[-1] += " " + str(sentence)
+                text_splits[-1] = text_splits[-1].lstrip()
+            elif len(str(sentence)) > text_split_length:
+                # if the current sentence is greater than the text_split_length
+                for line in textwrap.wrap(
+                    str(sentence),
+                    width=text_split_length,
+                    drop_whitespace=True,
+                    break_on_hyphens=False,
+                    tabsize=1,
+                ):
+                    text_splits.append(str(line))
+            else:
+                text_splits.append(str(sentence))
+
+        if len(text_splits) > 1:
+            if text_splits[0] == "":
+                del text_splits[0]
+    else:
+        text_splits = [text.lstrip()]
+
+    return text_splits
+
+
+_whitespace_re = re.compile(r"\s+")
+
+# List of (regular expression, replacement) pairs for abbreviations:
+_abbreviations = {
+    "en": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("mrs", "misess"),
+            ("mr", "mister"),
+            ("dr", "doctor"),
+            ("st", "saint"),
+            ("co", "company"),
+            ("jr", "junior"),
+            ("maj", "major"),
+            ("gen", "general"),
+            ("drs", "doctors"),
+            ("rev", "reverend"),
+            ("lt", "lieutenant"),
+            ("hon", "honorable"),
+            ("sgt", "sergeant"),
+            ("capt", "captain"),
+            ("esq", "esquire"),
+            ("ltd", "limited"),
+            ("col", "colonel"),
+            ("ft", "fort"),
+        ]
+    ],
+    "es": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("sra", "señora"),
+            ("sr", "señor"),
+            ("dr", "doctor"),
+            ("dra", "doctora"),
+            ("st", "santo"),
+            ("co", "compañía"),
+            ("jr", "junior"),
+            ("ltd", "limitada"),
+        ]
+    ],
+    "fr": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("mme", "madame"),
+            ("mr", "monsieur"),
+            ("dr", "docteur"),
+            ("st", "saint"),
+            ("co", "compagnie"),
+            ("jr", "junior"),
+            ("ltd", "limitée"),
+        ]
+    ],
+    "de": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("fr", "frau"),
+            ("dr", "doktor"),
+            ("st", "sankt"),
+            ("co", "firma"),
+            ("jr", "junior"),
+        ]
+    ],
+    "pt": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("sra", "senhora"),
+            ("sr", "senhor"),
+            ("dr", "doutor"),
+            ("dra", "doutora"),
+            ("st", "santo"),
+            ("co", "companhia"),
+            ("jr", "júnior"),
+            ("ltd", "limitada"),
+        ]
+    ],
+    "it": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            # ("sig.ra", "signora"),
+            ("sig", "signore"),
+            ("dr", "dottore"),
+            ("st", "santo"),
+            ("co", "compagnia"),
+            ("jr", "junior"),
+            ("ltd", "limitata"),
+        ]
+    ],
+    "pl": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("p", "pani"),
+            ("m", "pan"),
+            ("dr", "doktor"),
+            ("sw", "święty"),
+            ("jr", "junior"),
+        ]
+    ],
+    "ar": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            # There are not many common abbreviations in Arabic as in English.
+        ]
+    ],
+    "zh": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            # Chinese doesn't typically use abbreviations in the same way as Latin-based scripts.
+        ]
+    ],
+    "cs": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("dr", "doktor"),  # doctor
+            ("ing", "inženýr"),  # engineer
+            ("p", "pan"),  # Could also map to pani for woman but no easy way to do it
+            # Other abbreviations would be specialized and not as common.
+        ]
+    ],
+    "ru": [
+        (re.compile("\\b%s\\b" % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("г-жа", "госпожа"),  # Mrs.
+            ("г-н", "господин"),  # Mr.
+            ("д-р", "доктор"),  # doctor
+            # Other abbreviations are less common or specialized.
+        ]
+    ],
+    "nl": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("dhr", "de heer"),  # Mr.
+            ("mevr", "mevrouw"),  # Mrs.
+            ("dr", "dokter"),  # doctor
+            ("jhr", "jonkheer"),  # young lord or nobleman
+            # Dutch uses more abbreviations, but these are the most common ones.
+        ]
+    ],
+    "tr": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("b", "bay"),  # Mr.
+            ("byk", "büyük"),  # büyük
+            ("dr", "doktor"),  # doctor
+            # Add other Turkish abbreviations here if needed.
+        ]
+    ],
+    "hu": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            ("dr", "doktor"),  # doctor
+            ("b", "bácsi"),  # Mr.
+            ("nőv", "nővér"),  # nurse
+            # Add other Hungarian abbreviations here if needed.
+        ]
+    ],
+    "ko": [
+        (re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
+        for x in [
+            # Korean doesn't typically use abbreviations in the same way as Latin-based scripts.
+        ]
+    ],
+    "ja": [
+        (re.compile("\\b%s\\b" % x[0]), x[1])
+        for x in [
+            ("氏", "さん"),  # Mr.
+            ("夫人", "おんなのひと"),  # Mrs.
+            ("博士", "はかせ"),  # Doctor or PhD
+            ("株", "株式会社"),  # Corporation
+            ("有", "有限会社"),  # Limited company
+            ("大学", "だいがく"),   # University
+            ("先生", "せんせい"),   # Teacher/Professor/Master
+            ("君", "くん")   # Used at the end of boys' names to express familiarity or affection.
+        ]
+    ],
+}
+
+
+def expand_abbreviations_multilingual(text, lang="en"):
+    for regex, replacement in _abbreviations[lang]:
+        text = re.sub(regex, replacement, text)
+    return text
+
+
+_symbols_multilingual = {
+    "en": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " and "),
+            ("@", " at "),
+            ("%", " percent "),
+            ("#", " hash "),
+            ("$", " dollar "),
+            ("£", " pound "),
+            ("°", " degree "),
+        ]
+    ],
+    "es": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " y "),
+            ("@", " arroba "),
+            ("%", " por ciento "),
+            ("#", " numeral "),
+            ("$", " dolar "),
+            ("£", " libra "),
+            ("°", " grados "),
+        ]
+    ],
+    "fr": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " et "),
+            ("@", " arobase "),
+            ("%", " pour cent "),
+            ("#", " dièse "),
+            ("$", " dollar "),
+            ("£", " livre "),
+            ("°", " degrés "),
+        ]
+    ],
+    "de": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " und "),
+            ("@", " at "),
+            ("%", " prozent "),
+            ("#", " raute "),
+            ("$", " dollar "),
+            ("£", " pfund "),
+            ("°", " grad "),
+        ]
+    ],
+    "pt": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " e "),
+            ("@", " arroba "),
+            ("%", " por cento "),
+            ("#", " cardinal "),
+            ("$", " dólar "),
+            ("£", " libra "),
+            ("°", " graus "),
+        ]
+    ],
+    "it": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " e "),
+            ("@", " chiocciola "),
+            ("%", " per cento "),
+            ("#", " cancelletto "),
+            ("$", " dollaro "),
+            ("£", " sterlina "),
+            ("°", " gradi "),
+        ]
+    ],
+    "pl": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " i "),
+            ("@", " małpa "),
+            ("%", " procent "),
+            ("#", " krzyżyk "),
+            ("$", " dolar "),
+            ("£", " funt "),
+            ("°", " stopnie "),
+        ]
+    ],
+    "ar": [
+        # Arabic
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " و "),
+            ("@", " على "),
+            ("%", " في المئة "),
+            ("#", " رقم "),
+            ("$", " دولار "),
+            ("£", " جنيه "),
+            ("°", " درجة "),
+        ]
+    ],
+    "zh": [
+        # Chinese
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " 和 "),
+            ("@", " 在 "),
+            ("%", " 百分之 "),
+            ("#", " 号 "),
+            ("$", " 美元 "),
+            ("£", " 英镑 "),
+            ("°", " 度 "),
+        ]
+    ],
+    "cs": [
+        # Czech
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " a "),
+            ("@", " na "),
+            ("%", " procento "),
+            ("#", " křížek "),
+            ("$", " dolar "),
+            ("£", " libra "),
+            ("°", " stupně "),
+        ]
+    ],
+    "ru": [
+        # Russian
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " и "),
+            ("@", " собака "),
+            ("%", " процентов "),
+            ("#", " номер "),
+            ("$", " доллар "),
+            ("£", " фунт "),
+            ("°", " градус "),
+        ]
+    ],
+    "nl": [
+        # Dutch
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " en "),
+            ("@", " bij "),
+            ("%", " procent "),
+            ("#", " hekje "),
+            ("$", " dollar "),
+            ("£", " pond "),
+            ("°", " graden "),
+        ]
+    ],
+    "tr": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " ve "),
+            ("@", " at "),
+            ("%", " yüzde "),
+            ("#", " diyez "),
+            ("$", " dolar "),
+            ("£", " sterlin "),
+            ("°", " derece "),
+        ]
+    ],
+    "hu": [
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " és "),
+            ("@", " kukac "),
+            ("%", " százalék "),
+            ("#", " kettőskereszt "),
+            ("$", " dollár "),
+            ("£", " font "),
+            ("°", " fok "),
+        ]
+    ],
+    "ko": [
+        # Korean
+        (re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
+        for x in [
+            ("&", " 그리고 "),
+            ("@", " 에 "),
+            ("%", " 퍼센트 "),
+            ("#", " 번호 "),
+            ("$", " 달러 "),
+            ("£", " 파운드 "),
+            ("°", " 도 "),
+        ]
+    ],
+    "ja": [
+        (re.compile(r"%s" % re.escape(x[0])), x[1])
+        for x in [
+            ("&", " と "),
+            ("@", " アットマーク "),
+            ("%", " パーセント "),
+            ("#", " ナンバー "),
+            ("$", " ドル "),
+            ("£", " ポンド "),
+            ("°", " 度"),
+            ]
+    ],
+}
+
+
+def expand_symbols_multilingual(text, lang="en"):
+    for regex, replacement in _symbols_multilingual[lang]:
+        text = re.sub(regex, replacement, text)
+        text = text.replace("  ", " ")  # Ensure there are no double spaces
+    return text.strip()
+
+
+_ordinal_re = {
+    "en": re.compile(r"([0-9]+)(st|nd|rd|th)"),
+    "es": re.compile(r"([0-9]+)(º|ª|er|o|a|os|as)"),
+    "fr": re.compile(r"([0-9]+)(º|ª|er|re|e|ème)"),
+    "de": re.compile(r"([0-9]+)(st|nd|rd|th|º|ª|\.(?=\s|$))"),
+    "pt": re.compile(r"([0-9]+)(º|ª|o|a|os|as)"),
+    "it": re.compile(r"([0-9]+)(º|°|ª|o|a|i|e)"),
+    "pl": re.compile(r"([0-9]+)(º|ª|st|nd|rd|th)"),
+    "ar": re.compile(r"([0-9]+)(ون|ين|ث|ر|ى)"),
+    "cs": re.compile(r"([0-9]+)\.(?=\s|$)"),  # In Czech, a dot is often used after the number to indicate ordinals.
+    "ru": re.compile(r"([0-9]+)(-й|-я|-е|-ое|-ье|-го)"),
+    "nl": re.compile(r"([0-9]+)(de|ste|e)"),
+    "tr": re.compile(r"([0-9]+)(\.|inci|nci|uncu|üncü|\.)"),
+    "hu": re.compile(r"([0-9]+)(\.|adik|edik|odik|edik|ödik|ödike|ik)"),
+    "ko": re.compile(r"([0-9]+)(번째|번|차|째)"),
+    "ja": re.compile(r"([0-9]+)(番|回|つ|目|等|位)")
+}
+_number_re = re.compile(r"[0-9]+")
+_currency_re = {
+    "USD": re.compile(r"((\$[0-9\.\,]*[0-9]+)|([0-9\.\,]*[0-9]+\$))"),
+    "GBP": re.compile(r"((£[0-9\.\,]*[0-9]+)|([0-9\.\,]*[0-9]+£))"),
+    "EUR": re.compile(r"(([0-9\.\,]*[0-9]+€)|((€[0-9\.\,]*[0-9]+)))"),
+}
+
+_comma_number_re = re.compile(r"\b\d{1,3}(,\d{3})*(\.\d+)?\b")
+_dot_number_re = re.compile(r"\b\d{1,3}(.\d{3})*(\,\d+)?\b")
+_decimal_number_re = re.compile(r"([0-9]+[.,][0-9]+)")
+
+
+def _remove_commas(m):
+    text = m.group(0)
+    if "," in text:
+        text = text.replace(",", "")
+    return text
+
+
+def _remove_dots(m):
+    text = m.group(0)
+    if "." in text:
+        text = text.replace(".", "")
+    return text
+
+
+def _expand_decimal_point(m, lang="en"):
+    amount = m.group(1).replace(",", ".")
+    return num2words(float(amount), lang=lang if lang != "cs" else "cz")
+
+
+def _expand_currency(m, lang="en", currency="USD"):
+    amount = float((re.sub(r"[^\d.]", "", m.group(0).replace(",", "."))))
+    full_amount = num2words(amount, to="currency", currency=currency, lang=lang if lang != "cs" else "cz")
+
+    and_equivalents = {
+        "en": ", ",
+        "es": " con ",
+        "fr": " et ",
+        "de": " und ",
+        "pt": " e ",
+        "it": " e ",
+        "pl": ", ",
+        "cs": ", ",
+        "ru": ", ",
+        "nl": ", ",
+        "ar": ", ",
+        "tr": ", ",
+        "hu": ", ",
+        "ko": ", ",
+    }
+
+    if amount.is_integer():
+        last_and = full_amount.rfind(and_equivalents[lang])
+        if last_and != -1:
+            full_amount = full_amount[:last_and]
+
+    return full_amount
+
+
+def _expand_ordinal(m, lang="en"):
+    return num2words(int(m.group(1)), ordinal=True, lang=lang if lang != "cs" else "cz")
+
+
+def _expand_number(m, lang="en"):
+    return num2words(int(m.group(0)), lang=lang if lang != "cs" else "cz")
+
+
+def expand_numbers_multilingual(text, lang="en"):
+    if lang == "zh":
+        text = zh_num2words()(text)
+    else:
+        if lang in ["en", "ru"]:
+            text = re.sub(_comma_number_re, _remove_commas, text)
+        else:
+            text = re.sub(_dot_number_re, _remove_dots, text)
+        try:
+            text = re.sub(_currency_re["GBP"], lambda m: _expand_currency(m, lang, "GBP"), text)
+            text = re.sub(_currency_re["USD"], lambda m: _expand_currency(m, lang, "USD"), text)
+            text = re.sub(_currency_re["EUR"], lambda m: _expand_currency(m, lang, "EUR"), text)
+        except:
+            pass
+        if lang != "tr":
+            text = re.sub(_decimal_number_re, lambda m: _expand_decimal_point(m, lang), text)
+        text = re.sub(_ordinal_re[lang], lambda m: _expand_ordinal(m, lang), text)
+        text = re.sub(_number_re, lambda m: _expand_number(m, lang), text)
+    return text
+
+
+def lowercase(text):
+    return text.lower()
+
+
+def collapse_whitespace(text):
+    return re.sub(_whitespace_re, " ", text)
+
+
+def multilingual_cleaners(text, lang):
+    text = text.replace('"', "")
+    if lang == "tr":
+        text = text.replace("İ", "i")
+        text = text.replace("Ö", "ö")
+        text = text.replace("Ü", "ü")
+    text = lowercase(text)
+    text = expand_numbers_multilingual(text, lang)
+    text = expand_abbreviations_multilingual(text, lang)
+    text = expand_symbols_multilingual(text, lang=lang)
+    text = collapse_whitespace(text)
+    return text
+
+
+def basic_cleaners(text):
+    """Basic pipeline that lowercases and collapses whitespace without transliteration."""
+    text = lowercase(text)
+    text = collapse_whitespace(text)
+    return text
+
+
+def chinese_transliterate(text):
+    return "".join(
+        [p[0] for p in pypinyin.pinyin(text, style=pypinyin.Style.TONE3, heteronym=False, neutral_tone_with_five=True)]
+    )
+
+
+def japanese_cleaners(text, katsu):
+    text = katsu.romaji(text)
+    text = lowercase(text)
+    return text
+
+
+def korean_transliterate(text):
+    r = Transliter(academic)
+    return r.translit(text)
+
+
+DEFAULT_VOCAB_FILE = os.path.join(os.path.dirname(os.path.realpath(__file__)), "../data/tokenizer.json")
+
+
+class VoiceBpeTokenizer:
+    def __init__(self, vocab_file=None):
+        self.tokenizer = None
+        if vocab_file is not None:
+            self.tokenizer = Tokenizer.from_file(vocab_file)
+        self.char_limits = {
+            "en": 250,
+            "de": 253,
+            "fr": 273,
+            "es": 239,
+            "it": 213,
+            "pt": 203,
+            "pl": 224,
+            "zh": 82,
+            "ar": 166,
+            "cs": 186,
+            "ru": 182,
+            "nl": 251,
+            "tr": 226,
+            "ja": 71,
+            "hu": 224,
+            "ko": 95,
+        }
+
+    @cached_property
+    def katsu(self):
+        import cutlet
+
+        return cutlet.Cutlet()
+
+    def check_input_length(self, txt, lang):
+        lang = lang.split("-")[0]  # remove the region
+        limit = self.char_limits.get(lang, 250)
+        if len(txt) > limit:
+            print(
+                f"[!] Warning: The text length exceeds the character limit of {limit} for language '{lang}', this might cause truncated audio."
+            )
+
+    def preprocess_text(self, txt, lang):
+        if lang in {"ar", "cs", "de", "en", "es", "fr", "hu", "it", "nl", "pl", "pt", "ru", "tr", "zh", "ko"}:
+            txt = multilingual_cleaners(txt, lang)
+            if lang == "zh":
+                txt = chinese_transliterate(txt)
+            if lang == "ko":
+                txt = korean_transliterate(txt)
+        elif lang == "ja":
+            txt = japanese_cleaners(txt, self.katsu)
+        elif lang == "hi":
+            # @manmay will implement this
+            txt = basic_cleaners(txt)
+        else:
+            raise NotImplementedError(f"Language '{lang}' is not supported.")
+        return txt
+
+    def encode(self, txt, lang):
+        lang = lang.split("-")[0]  # remove the region
+        self.check_input_length(txt, lang)
+        txt = self.preprocess_text(txt, lang)
+        lang = "zh-cn" if lang == "zh" else lang
+        txt = f"[{lang}]{txt}"
+        txt = txt.replace(" ", "[SPACE]")
+        return self.tokenizer.encode(txt).ids
+
+    def decode(self, seq):
+        if isinstance(seq, torch.Tensor):
+            seq = seq.cpu().numpy()
+        txt = self.tokenizer.decode(seq, skip_special_tokens=False).replace(" ", "")
+        txt = txt.replace("[SPACE]", " ")
+        txt = txt.replace("[STOP]", "")
+        txt = txt.replace("[UNK]", "")
+        return txt
+
+    def __len__(self):
+        return self.tokenizer.get_vocab_size()
+
+    def get_number_tokens(self):
+        return max(self.tokenizer.get_vocab().values()) + 1
+
+
+def test_expand_numbers_multilingual():
+    test_cases = [
+        # English
+        ("In 12.5 seconds.", "In twelve point five seconds.", "en"),
+        ("There were 50 soldiers.", "There were fifty soldiers.", "en"),
+        ("This is a 1st test", "This is a first test", "en"),
+        ("That will be $20 sir.", "That will be twenty dollars sir.", "en"),
+        ("That will be 20€ sir.", "That will be twenty euro sir.", "en"),
+        ("That will be 20.15€ sir.", "That will be twenty euro, fifteen cents sir.", "en"),
+        ("That's 100,000.5.", "That's one hundred thousand point five.", "en"),
+        # French
+        ("En 12,5 secondes.", "En douze virgule cinq secondes.", "fr"),
+        ("Il y avait 50 soldats.", "Il y avait cinquante soldats.", "fr"),
+        ("Ceci est un 1er test", "Ceci est un premier test", "fr"),
+        ("Cela vous fera $20 monsieur.", "Cela vous fera vingt dollars monsieur.", "fr"),
+        ("Cela vous fera 20€ monsieur.", "Cela vous fera vingt euros monsieur.", "fr"),
+        ("Cela vous fera 20,15€ monsieur.", "Cela vous fera vingt euros et quinze centimes monsieur.", "fr"),
+        ("Ce sera 100.000,5.", "Ce sera cent mille virgule cinq.", "fr"),
+        # German
+        ("In 12,5 Sekunden.", "In zwölf Komma fünf Sekunden.", "de"),
+        ("Es gab 50 Soldaten.", "Es gab fünfzig Soldaten.", "de"),
+        ("Dies ist ein 1. Test", "Dies ist ein erste Test", "de"),  # Issue with gender
+        ("Das macht $20 Herr.", "Das macht zwanzig Dollar Herr.", "de"),
+        ("Das macht 20€ Herr.", "Das macht zwanzig Euro Herr.", "de"),
+        ("Das macht 20,15€ Herr.", "Das macht zwanzig Euro und fünfzehn Cent Herr.", "de"),
+        # Spanish
+        ("En 12,5 segundos.", "En doce punto cinco segundos.", "es"),
+        ("Había 50 soldados.", "Había cincuenta soldados.", "es"),
+        ("Este es un 1er test", "Este es un primero test", "es"),
+        ("Eso le costará $20 señor.", "Eso le costará veinte dólares señor.", "es"),
+        ("Eso le costará 20€ señor.", "Eso le costará veinte euros señor.", "es"),
+        ("Eso le costará 20,15€ señor.", "Eso le costará veinte euros con quince céntimos señor.", "es"),
+        # Italian
+        ("In 12,5 secondi.", "In dodici virgola cinque secondi.", "it"),
+        ("C'erano 50 soldati.", "C'erano cinquanta soldati.", "it"),
+        ("Questo è un 1° test", "Questo è un primo test", "it"),
+        ("Ti costerà $20 signore.", "Ti costerà venti dollari signore.", "it"),
+        ("Ti costerà 20€ signore.", "Ti costerà venti euro signore.", "it"),
+        ("Ti costerà 20,15€ signore.", "Ti costerà venti euro e quindici centesimi signore.", "it"),
+        # Portuguese
+        ("Em 12,5 segundos.", "Em doze vírgula cinco segundos.", "pt"),
+        ("Havia 50 soldados.", "Havia cinquenta soldados.", "pt"),
+        ("Este é um 1º teste", "Este é um primeiro teste", "pt"),
+        ("Isso custará $20 senhor.", "Isso custará vinte dólares senhor.", "pt"),
+        ("Isso custará 20€ senhor.", "Isso custará vinte euros senhor.", "pt"),
+        (
+            "Isso custará 20,15€ senhor.",
+            "Isso custará vinte euros e quinze cêntimos senhor.",
+            "pt",
+        ),  # "cêntimos" should be "centavos" num2words issue
+        # Polish
+        ("W 12,5 sekundy.", "W dwanaście przecinek pięć sekundy.", "pl"),
+        ("Było 50 żołnierzy.", "Było pięćdziesiąt żołnierzy.", "pl"),
+        ("To będzie kosztować 20€ panie.", "To będzie kosztować dwadzieścia euro panie.", "pl"),
+        ("To będzie kosztować 20,15€ panie.", "To będzie kosztować dwadzieścia euro, piętnaście centów panie.", "pl"),
+        # Arabic
+        ("في الـ 12,5 ثانية.", "في الـ اثنا عشر  , خمسون ثانية.", "ar"),
+        ("كان هناك 50 جنديًا.", "كان هناك خمسون جنديًا.", "ar"),
+        # ("ستكون النتيجة $20 يا سيد.", 'ستكون النتيجة عشرون دولار يا سيد.', 'ar'), # $ and € are mising from num2words
+        # ("ستكون النتيجة 20€ يا سيد.", 'ستكون النتيجة عشرون يورو يا سيد.', 'ar'),
+        # Czech
+        ("Za 12,5 vteřiny.", "Za dvanáct celá pět vteřiny.", "cs"),
+        ("Bylo tam 50 vojáků.", "Bylo tam padesát vojáků.", "cs"),
+        ("To bude stát 20€ pane.", "To bude stát dvacet euro pane.", "cs"),
+        ("To bude 20.15€ pane.", "To bude dvacet euro, patnáct centů pane.", "cs"),
+        # Russian
+        ("Через 12.5 секунды.", "Через двенадцать запятая пять секунды.", "ru"),
+        ("Там было 50 солдат.", "Там было пятьдесят солдат.", "ru"),
+        ("Это будет 20.15€ сэр.", "Это будет двадцать евро, пятнадцать центов сэр.", "ru"),
+        ("Это будет стоить 20€ господин.", "Это будет стоить двадцать евро господин.", "ru"),
+        # Dutch
+        ("In 12,5 seconden.", "In twaalf komma vijf seconden.", "nl"),
+        ("Er waren 50 soldaten.", "Er waren vijftig soldaten.", "nl"),
+        ("Dat wordt dan $20 meneer.", "Dat wordt dan twintig dollar meneer.", "nl"),
+        ("Dat wordt dan 20€ meneer.", "Dat wordt dan twintig euro meneer.", "nl"),
+        # Chinese (Simplified)
+        ("在12.5秒内", "在十二点五秒内", "zh"),
+        ("有50名士兵", "有五十名士兵", "zh"),
+        # ("那将是$20先生", '那将是二十美元先生', 'zh'), currency doesn't work
+        # ("那将是20€先生", '那将是二十欧元先生', 'zh'),
+        # Turkish
+        # ("12,5 saniye içinde.", 'On iki virgül beş saniye içinde.', 'tr'), # decimal doesn't work for TR
+        ("50 asker vardı.", "elli asker vardı.", "tr"),
+        ("Bu 1. test", "Bu birinci test", "tr"),
+        # ("Bu 100.000,5.", 'Bu yüz bin virgül beş.', 'tr'),
+        # Hungarian
+        ("12,5 másodperc alatt.", "tizenkettő egész öt tized másodperc alatt.", "hu"),
+        ("50 katona volt.", "ötven katona volt.", "hu"),
+        ("Ez az 1. teszt", "Ez az első teszt", "hu"),
+        # Korean
+        ("12.5 초 안에.", "십이 점 다섯 초 안에.", "ko"),
+        ("50 명의 병사가 있었다.", "오십 명의 병사가 있었다.", "ko"),
+        ("이것은 1 번째 테스트입니다", "이것은 첫 번째 테스트입니다", "ko"),
+    ]
+    for a, b, lang in test_cases:
+        out = expand_numbers_multilingual(a, lang=lang)
+        assert out == b, f"'{out}' vs '{b}'"
+
+
+def test_abbreviations_multilingual():
+    test_cases = [
+        # English
+        ("Hello Mr. Smith.", "Hello mister Smith.", "en"),
+        ("Dr. Jones is here.", "doctor Jones is here.", "en"),
+        # Spanish
+        ("Hola Sr. Garcia.", "Hola señor Garcia.", "es"),
+        ("La Dra. Martinez es muy buena.", "La doctora Martinez es muy buena.", "es"),
+        # French
+        ("Bonjour Mr. Dupond.", "Bonjour monsieur Dupond.", "fr"),
+        ("Mme. Moreau est absente aujourd'hui.", "madame Moreau est absente aujourd'hui.", "fr"),
+        # German
+        ("Frau Dr. Müller ist sehr klug.", "Frau doktor Müller ist sehr klug.", "de"),
+        # Portuguese
+        ("Olá Sr. Silva.", "Olá senhor Silva.", "pt"),
+        ("Dra. Costa, você está disponível?", "doutora Costa, você está disponível?", "pt"),
+        # Italian
+        ("Buongiorno, Sig. Rossi.", "Buongiorno, signore Rossi.", "it"),
+        # ("Sig.ra Bianchi, posso aiutarti?", 'signora Bianchi, posso aiutarti?', 'it'), # Issue with matching that pattern
+        # Polish
+        ("Dzień dobry, P. Kowalski.", "Dzień dobry, pani Kowalski.", "pl"),
+        ("M. Nowak, czy mogę zadać pytanie?", "pan Nowak, czy mogę zadać pytanie?", "pl"),
+        # Czech
+        ("P. Novák", "pan Novák", "cs"),
+        ("Dr. Vojtěch", "doktor Vojtěch", "cs"),
+        # Dutch
+        ("Dhr. Jansen", "de heer Jansen", "nl"),
+        ("Mevr. de Vries", "mevrouw de Vries", "nl"),
+        # Russian
+        ("Здравствуйте Г-н Иванов.", "Здравствуйте господин Иванов.", "ru"),
+        ("Д-р Смирнов здесь, чтобы увидеть вас.", "доктор Смирнов здесь, чтобы увидеть вас.", "ru"),
+        # Turkish
+        ("Merhaba B. Yılmaz.", "Merhaba bay Yılmaz.", "tr"),
+        ("Dr. Ayşe burada.", "doktor Ayşe burada.", "tr"),
+        # Hungarian
+        ("Dr. Szabó itt van.", "doktor Szabó itt van.", "hu"),
+    ]
+
+    for a, b, lang in test_cases:
+        out = expand_abbreviations_multilingual(a, lang=lang)
+        assert out == b, f"'{out}' vs '{b}'"
+
+
+def test_symbols_multilingual():
+    test_cases = [
+        ("I have 14% battery", "I have 14 percent battery", "en"),
+        ("Te veo @ la fiesta", "Te veo arroba la fiesta", "es"),
+        ("J'ai 14° de fièvre", "J'ai 14 degrés de fièvre", "fr"),
+        ("Die Rechnung beträgt £ 20", "Die Rechnung beträgt pfund 20", "de"),
+        ("O meu email é ana&joao@gmail.com", "O meu email é ana e joao arroba gmail.com", "pt"),
+        ("linguaggio di programmazione C#", "linguaggio di programmazione C cancelletto", "it"),
+        ("Moja temperatura to 36.6°", "Moja temperatura to 36.6 stopnie", "pl"),
+        ("Mám 14% baterie", "Mám 14 procento baterie", "cs"),
+        ("Těším se na tebe @ party", "Těším se na tebe na party", "cs"),
+        ("У меня 14% заряда", "У меня 14 процентов заряда", "ru"),
+        ("Я буду @ дома", "Я буду собака дома", "ru"),
+        ("Ik heb 14% batterij", "Ik heb 14 procent batterij", "nl"),
+        ("Ik zie je @ het feest", "Ik zie je bij het feest", "nl"),
+        ("لدي 14% في البطارية", "لدي 14 في المئة في البطارية", "ar"),
+        ("我的电量为 14%", "我的电量为 14 百分之", "zh"),
+        ("Pilim %14 dolu.", "Pilim yüzde 14 dolu.", "tr"),
+        ("Az akkumulátorom töltöttsége 14%", "Az akkumulátorom töltöttsége 14 százalék", "hu"),
+        ("배터리 잔량이 14%입니다.", "배터리 잔량이 14 퍼센트입니다.", "ko"),
+    ]
+
+    for a, b, lang in test_cases:
+        out = expand_symbols_multilingual(a, lang=lang)
+        assert out == b, f"'{out}' vs '{b}'"
+
+
+if __name__ == "__main__":
+    test_expand_numbers_multilingual()
+    test_abbreviations_multilingual()
+    test_symbols_multilingual()

system/requirements/requirements_standalone.txt

@@ -0,0 +1,172 @@
+absl-py==2.1.0
+aiofiles==23.2.1
+aiohttp==3.9.3
+aiosignal==1.3.1
+altair==5.2.0
+annotated-types==0.6.0
+anyascii==0.3.2
+anyio==4.3.0
+attrs==23.2.0
+audioread==3.0.1
+av==11.0.0
+Babel==2.14.0
+bangla==0.0.2
+blinker==1.7.0
+blis==0.7.11
+bnnumerizer==0.0.2
+bnunicodenormalizer==0.1.6
+catalogue==2.0.10
+certifi==2024.2.2
+cffi==1.16.0
+charset-normalizer==3.3.2
+click==8.1.7
+cloudpathlib==0.16.0
+colorama==0.4.6
+coloredlogs==15.0.1
+confection==0.1.4
+contourpy==1.2.0
+coqpit==0.0.17
+ctranslate2==4.1.0
+cutlet==0.4.0
+cycler==0.12.1
+cymem==2.0.8
+Cython==3.0.9
+dateparser==1.1.8
+decorator==5.1.1
+docopt==0.6.2
+einops==0.7.0
+encodec==0.1.1
+fastapi==0.110.0
+faster-whisper==1.0.1
+ffmpy==0.3.2
+filelock==3.13.3
+Flask==3.0.2
+flatbuffers==24.3.25
+fonttools==4.50.0
+frozenlist==1.4.1
+fsspec==2024.3.1
+fugashi==1.3.1
+fuzzywuzzy>=0.18.0
+g2pkk==0.1.2
+gradio==3.50.2
+gradio_client==0.6.1
+grpcio==1.62.1
+gruut==2.2.3
+gruut-ipa==0.13.0
+gruut-lang-de==2.0.0
+gruut-lang-en==2.0.0
+gruut-lang-es==2.0.0
+gruut-lang-fr==2.0.2
+h11==0.14.0
+hangul-romanize==0.1.0
+httpcore==1.0.4
+httpx==0.27.0
+huggingface-hub==0.22.1
+humanfriendly==10.0
+idna==3.6
+importlib_metadata==7.1.0
+importlib_resources==6.4.0
+inflect==7.0.0
+itsdangerous==2.1.2
+jaconv==0.3.4
+jamo==0.4.1
+jieba==0.42.1
+Jinja2==3.1.3
+joblib==1.3.2
+jsonlines==1.2.0
+jsonschema==4.21.1
+jsonschema-specifications==2023.12.1
+kiwisolver==1.4.5
+langcodes==3.3.0
+lazy_loader==0.3
+librosa==0.10.1
+llvmlite==0.42.0
+Markdown==3.6
+MarkupSafe==2.1.5
+matplotlib==3.8.3
+mojimoji==0.0.13
+mpmath==1.3.0
+msgpack==1.0.8
+multidict==6.0.5
+murmurhash==1.0.10
+networkx==2.8.8
+nltk==3.8.1
+num2words==0.5.13
+numba==0.59.1
+numpy==1.26.4
+nvidia-cublas-cu11>=11.11.3.6
+nvidia-cudnn-cu11>=9.0.0.312
+onnxruntime==1.17.1
+orjson==3.9.15
+packaging==24.0
+pandas==1.5.3
+pillow==10.2.0
+platformdirs==4.2.0
+pooch==1.8.1
+preshed==3.0.9
+protobuf==5.26.0
+psutil==5.9.8
+pycparser==2.21
+pydantic==2.6.4
+pydantic_core==2.16.3
+pydub==0.25.1
+pynndescent==0.5.11
+pyparsing==3.1.2
+pypinyin==0.51.0
+pyreadline3==3.4.1
+pysbd==0.3.4
+python-crfsuite==0.9.10
+python-dateutil==2.9.0.post0
+python-Levenshtein>=0.25.0
+python-multipart==0.0.9
+pytz==2024.1
+PyYAML==6.0.1
+referencing==0.34.0
+regex==2023.12.25
+requests==2.31.0
+rpds-py==0.18.0
+safetensors==0.4.2
+scikit-learn==1.4.1.post1
+scipy==1.12.0
+semantic-version==2.10.0
+six==1.16.0
+smart-open==6.4.0
+sniffio==1.3.1
+sounddevice==0.4.6
+soundfile==0.12.1
+soxr==0.3.7
+spacy==3.7.4
+spacy-legacy==3.0.12
+spacy-loggers==1.0.5
+srsly==2.4.8
+starlette==0.36.3
+SudachiDict-core==20240109
+SudachiPy==0.6.8
+sympy==1.12
+tensorboard==2.16.2
+tensorboard-data-server==0.7.2
+thinc==8.2.3
+threadpoolctl==3.4.0
+tokenizers==0.15.2
+toolz==0.12.1
+torch>=2.2.0
+torchaudio>=2.2.0
+tqdm==4.66.2
+trainer==0.0.36
+transformers==4.39.1
+TTS==0.22.0
+typer==0.9.4
+typing_extensions==4.10.0
+tzdata==2024.1
+tzlocal==5.2
+umap-learn==0.5.5
+Unidecode==1.3.8
+unidic-lite==1.0.8
+urllib3==2.2.1
+uvicorn==0.29.0
+wasabi==1.1.2
+weasel==0.3.4
+websockets==11.0.3
+Werkzeug==3.0.1
+yarl==1.9.4
+zipp==3.18.1

system/requirements/requirements_textgen.txt

@@ -2,7 +2,7 @@ numpy>=1.22.0
 soundfile>=0.12.1
 uvicorn>=0.24.0.post1
 transformers>=4.37.1
-TTS>=0.21.3
+TTS>=0.22.0
 torch>=2.1.0+cu118
 torchaudio>=2.1.0+cu118
 fastapi>=0.104.1
@@ -16,3 +16,9 @@ sounddevice>=0.4.6
 python-multipart>=0.0.6
 cutlet>=0.3.0
 unidic-lite>=1.0.8
+faster-whisper>=0.10.0
+pandas>=1.5.0
+nvidia-cublas-cu11>=11.11.3.6
+nvidia-cudnn-cu11>=9.0.0.312
+fuzzywuzzy>=0.18.0
+python-Levenshtein>=0.25.0

system/tts_diff/tts_diff.py

@@ -0,0 +1,269 @@
+import json
+import argparse
+from pathlib import Path
+import os
+import glob
+import platform
+import site
+import subprocess
+import sys
+# version 0.8
+
+def find_files_in_path_with_wildcard(pattern):
+    # Get the site-packages directory of the current Python environment
+    site_packages_path = site.getsitepackages()
+    found_paths = []
+    # Adjust the sub-directory based on the operating system
+    sub_directory = "nvidia/cublas"
+    if platform.system() == "Linux":
+        sub_directory = os.path.join(sub_directory, "lib")
+    else:
+        sub_directory = os.path.join(sub_directory, "bin")
+    
+    # Iterate over each site-packages directory (there can be more than one)
+    for directory in site_packages_path:
+        # Construct the search directory path
+        search_directory = os.path.join(directory, sub_directory)
+        # Use glob to find all files matching the pattern in this directory
+        for file_path in glob.glob(os.path.join(search_directory, pattern)):
+            if os.path.isfile(file_path):  # Ensure it's a file
+                found_paths.append(file_path)
+    return found_paths
+
+def detect_cublas():
+    # Use different file patterns based on the operating system
+    file_name_pattern = 'cublas64_11.*' if platform.system() == "Windows" else 'libcublas.so.11*'
+    found_paths = find_files_in_path_with_wildcard(file_name_pattern)
+    if found_paths:
+        print("[AllTalk TTSDiff] \033[94mCublas:\033[0m \033[92mDetected\033[0m", found_paths)
+    else:
+        print("[AllTalk TTSDiff] \033[94mCublas:\033[0m \033[91mERROR Not Detected\033[0m")
+        print("[AllTalk TTSDiff] \033[94mPlease install Cublas from the Nvidia CUDA Toolkit \033[0mhttps://developer.nvidia.com/cuda-downloads")
+        exit(1)
+        
+
+parser = argparse.ArgumentParser(description="Compare TTS output with the original text using detailed comparison.")
+parser.add_argument("--threshold", type=int, default=98, help="Similarity threshold for considering a match (default: 98)")
+parser.add_argument("--ttslistpath", help="Path to the ttsList.json file")
+parser.add_argument("--wavfilespath", help="Path to the wav outputs folder")
+args = parser.parse_args()
+
+if not args.ttslistpath:
+    parser.error("[AllTalk TTSDiff] Please specify the path to the ttslist.json file using the --ttslistpath argument.")
+if not args.wavfilespath:
+    parser.error("[AllTalk TTSDiff] Please specify the path to the wavs output folder file using the --wavfilespath argument.")
+
+json_file_path = Path(args.ttslistpath).resolve()
+wav_file_path = Path(args.wavfilespath).resolve()
+accuracy_threshold = args.threshold
+
+def disclaimer_text():
+    print(f"\nDESCRIPTION: ")
+    print(f"  This script is designed to assist in identifying bad TTS generation ID's from the TTS Generator.\n")
+    print(f"  This will download the Whisper Large-v2 model (if not already downloaded), and use it to compare your text to")
+    print(f"  to the generated TTS. If it finds an issue, it will flag the ID number that has the issue. This does NOT 100%")
+    print(f"  ensure that all TTS generated is perfect, sounds good etc, but it should flag up glaring issues and help with")
+    print(f"  the process of identifying bad TTS generations.")
+    print(f"\nNOTES ON USE: ")
+    print(f"  - Ensure AllTalk is running in its own command prompt/terminal.")
+    print(f"  - This script should be running in its own command prompt/terminal window in the same Python Environment that")
+    print(f"    AllTalk is running in.")
+    print(f"  - After generating your text in the TTS Generator, you need to `Export List to JSON` and save the file in the")
+    print(f"    same folder/directory this script is running from.")
+    print(f"  - You can use `--threshold` with a numerical value from 1 to 100 to set accuracy detection. For example it may")
+    print(f"    detect/transcribe `their` as `there`, which of course, both words sound the same. If you want 100% accuracy")
+    print(f"    you will set 100 as the threshold. You will more than likely want a number in the high 90's. Default is set")
+    print(f"    at 98, which gives a little flexibility.")
+    print(f"  - The JSON file must be named \033[93mttsList.json\033[0m")
+    print(f"  - When you have your ID list, go back into the TTS Generator, correct any lines and regenerate them. If you")
+    print(f"    want to re-test everthing again after re-generating, you will need to export your list again and re-run this")
+    print(f"    script again, against your newly exported JSON list.")
+    print(f"  - This requires access to \033[93mcublas64_11\033[0m, the same as Finetuning.")
+    print(f"    https://github.com/erew123/alltalk_tts/tree/main?#-important-requirements-cuda-118\n")
+    return
+
+try:
+    from faster_whisper import WhisperModel
+    import re
+    import string
+    import spacy
+    import torch
+    from fuzzywuzzy import fuzz
+except ImportError as e:
+    print(f"[AllTalk TTSDiff] ERROR STARTING SCRIPT:")
+    print("[AllTalk TTSDiff] An error occurred importing one or more required libraries.")
+    print("[AllTalk TTSDiff] Please ensure you have \033[94mactivated your Python environment \033[0mand \033[94minstalled all requirements.\033[0m")
+    print("[AllTalk TTSDiff] \033[94mMissing library:\033[93m", str(e), "\033[0m")
+    exit(1)
+
+# Attempt to load the spaCy model
+try:
+    # Load a medium-sized spaCy model, adjust to "en_core_web_md" or "en_core_web_lg" as needed
+    nlp = spacy.load("en_core_web_md")
+except Exception as e:
+    print(f"[AllTalk TTSDiff] ERROR STARTING SCRIPT:")
+    print("[AllTalk TTSDiff] Failed to load the spaCy language model.")
+    
+    # Attempt to download the spaCy model automatically
+    try:
+        print("[AllTalk TTSDiff] Attempting to automatically download the spaCy language model...")
+        subprocess.run([sys.executable, "-m", "spacy", "download", "en_core_web_md"], check=True)
+        
+        # Try to load the model again after downloading
+        nlp = spacy.load("en_core_web_md")
+        print("[AllTalk TTSDiff] Model downloaded and loaded successfully.")
+    except Exception as download_exception:
+        print("[AllTalk TTSDiff] Automatic download failed.")
+        print("[AllTalk TTSDiff] Please manually install the spaCy language model by running:")
+        print("[AllTalk TTSDiff] \033[93mpython -m spacy download en_core_web_md\033[0m")
+        exit(1)
+
+def texts_are_similar(text1, text2, threshold=0.8):
+    # Preliminary check for short texts
+    if len(text1) < 10 or len(text2) < 10:
+        ratio = fuzz.ratio(text1, text2) / 100.0  # Use FuzzyWuzzy for short texts
+        return ratio > threshold
+
+    # Process the texts through the NLP model for longer texts
+    doc1 = nlp(text1)
+    doc2 = nlp(text2)
+    
+    # Compute semantic similarity
+    similarity = doc1.similarity(doc2)
+    
+    return similarity >= threshold
+
+def normalize_text(text):
+    # Normalize or remove CRLF and other non-standard whitespaces
+    text = text.replace("\r\n", " ").replace("\n", " ").replace("\r", " ")
+    
+    # Convert to lowercase
+    text = text.lower()
+    
+    # Standardize and then remove quotation marks
+    text = text.replace("“", '"').replace("”", '"').replace("‘", "'").replace("’", "'")
+    text = text.translate(str.maketrans('', '', '"\''))
+    
+    # Remove all other punctuation except hyphens to preserve compound words
+    text = text.translate(str.maketrans('', '', string.punctuation.replace("-", "")))
+    
+    # Collapse any sequence of whitespace (including spaces) into a single space
+    text = re.sub(r'\s+', ' ', text)
+    
+    return text.strip()
+
+spoken_punctuation_mapping = {
+    "dot": ".",
+    "comma": ",",
+    # Add more mappings as needed
+}
+
+def contains_spoken_punctuation(transcribed_text):
+    for spoken, punctuation in spoken_punctuation_mapping.items():
+        if spoken in transcribed_text:
+            return True
+    return False
+
+def detailed_comparison(original_text, transcribed_text, threshold=98):
+    original_clean = normalize_text(original_text)
+    transcribed_clean = normalize_text(transcribed_text)
+    ratio = fuzz.ratio(original_clean, transcribed_clean)
+    return ratio > threshold
+
+def access_local_file(url):
+    # Extract the file name from the URL
+    file_name = url.split('/')[-1]
+    # Construct the full path to the file on disk
+    file_path = wav_file_path / file_name
+    # Check if the file exists
+    if not file_path.exists():
+        print(f"[AllTalk TTSDiff] File does not exist: {file_path}")
+        return None
+    return file_path
+
+def transcribe_and_compare(file_url, original_text, model, item_id, flagged_ids):
+    audio_file_path = access_local_file(file_url)
+    if audio_file_path is None:
+        print(f"[AllTalk TTSDiff] Could not access local file for URL: {file_url}")
+        return  # Skip this file if it can't be accessed
+
+    try:
+        segments, info = model.transcribe(str(audio_file_path), beam_size=5)
+        transcribed_text = " ".join([segment.text for segment in segments])
+    except Exception as e:
+        print(f"[AllTalk TTSDiff] Error transcribing file {audio_file_path}: {e}")
+        return  # Skip this file if transcription fails
+    
+    segments, info = model.transcribe(str(audio_file_path), beam_size=5)
+    transcribed_text = " ".join([segment.text for segment in segments])
+    
+    # Normalize texts for comparison
+    original_text_normalized = normalize_text(original_text)
+    transcribed_text_normalized = normalize_text(transcribed_text)
+    
+    # Enhanced comparison using detailed fuzzy matching
+    is_detailed_match = detailed_comparison(original_text_normalized, transcribed_text_normalized, args.threshold)
+    has_spoken_punctuation = contains_spoken_punctuation(transcribed_text_normalized) and not contains_spoken_punctuation(original_text_normalized)
+    
+    # Adjust is_match based on detailed comparison and spoken punctuation check
+    is_match = is_detailed_match and not has_spoken_punctuation
+    
+    # Only log and flag IDs for review if there's a mismatch or detected issues
+    if not is_match:
+        #print(f"\033[93mMismatch Detected - ID:\033[0m {item_id}")
+        print(f"[AllTalk TTSDiff] \033[93mText:\033[0m {original_text}")
+        print(f"[AllTalk TTSDiff] \033[91mTTS :\033[0m{transcribed_text}")
+        if has_spoken_punctuation:
+            print(f"[AllTalk TTSDiff] Note: Potential incorrect spoken punctuation detected.")
+        flagged_ids.append(item_id)  # Track the ID for review
+
+def main():
+    print(f"[AllTalk TTSDiff]")
+    print(f"[AllTalk TTSDiff]")
+    print("[AllTalk TTSDiff] \033[92mStarting Compare of Text vs TTS:\033[0m")
+    print("[AllTalk TTSDiff]")
+    print("[AllTalk TTSDiff] \033[94mJSON file  :\033[92m", json_file_path, "\033[0m")
+    print("[AllTalk TTSDiff] \033[94mWAV files  :\033[92m", wav_file_path, "\033[0m")
+    print("[AllTalk TTSDiff] \033[94mAccuracy   :\033[92m", accuracy_threshold, "%", "\033[0m")
+    detect_cublas()
+    model_size = "large-v2"
+    device = "cuda" if torch.cuda.is_available() else "cpu"
+    model = WhisperModel(model_size, device=device, compute_type="float32")
+    
+    flagged_ids = []  # Initialize the list to track IDs needing review
+
+    try:
+        with open(json_file_path, "r", encoding="utf-8") as f:
+            tts_list = json.load(f)
+    except FileNotFoundError:
+        print(f"[AllTalk TTSDiff] ERROR STARTING SCRIPT:")
+        print("[AllTalk TTSDiff] Error: \033[93mttsList.json\033[0m file not found.")
+        print("[AllTalk TTSDiff] Please ensure you follow these steps:")
+        print("[AllTalk TTSDiff] After generating your text in the TTS Generator, you need to `Export List to JSON` and save the file in the")
+        print("[AllTalk TTSDiff] same folder/directory this script is running from.")
+        print("[AllTalk TTSDiff] The JSON file must be named \033[93mttsList.json\033[0m")
+        exit(1)
+
+    print(f"[AllTalk TTSDiff]")
+
+    for item in tts_list:
+        print(f"[AllTalk TTSDiff] Processing ID: {item['id']}")
+        transcribe_and_compare(item['fileUrl'], item['text'], model, item['id'], flagged_ids)
+    
+    print(f"[AllTalk TTSDiff]")
+    # Print summary information at the end
+    if flagged_ids:
+        print(f"[AllTalk TTSDiff] \033[94mIDs needing review:\033[0m", ', '.join(map(str, flagged_ids)))
+        print(f"[AllTalk TTSDiff]")
+        print(f"[AllTalk TTSDiff] Review ID's and correct any lines by editing & regenerating as necessary.")
+    else:
+        print("[AllTalk TTSDiff] \033[94mIDs needing review:\033[0m No issues detected.")
+    summary_data = {
+        "flagged_ids": flagged_ids
+    }
+    with open(wav_file_path / "analysis_summary.json", "w") as summary_file:
+        json.dump(summary_data, summary_file)
+    print("[AllTalk TTSDiff]")
+
+if __name__ == "__main__":
+    main()

system/tts_generator/tts_generator.html

@@ -366,6 +366,11 @@
             /* Other styles for highlighting */
         }
 
+        .problematic {
+            background-color: rgb(179, 104, 104);
+            /* Or any light red color of your choice */
+        }
+
         .highlight-count {
             color: #ff7300;
             /* Change to the color you want */
@@ -390,6 +395,13 @@
             font-size: 0.8em;
         }
 
+        /*Disable and enable buttons*/
+        button:disabled {
+            background-color: #919191;
+            color: #666;
+            cursor: not-allowed;
+        }
+
         .edit-input {
             /* Styles for the edit input field */
             margin-top: 5px;
@@ -562,7 +574,7 @@
         <div class="setting slider-setting-wrapper">
             <label for="volume">Volume:</label>
             <div class="slider-container">
-                <input type="range" id="volume" name="volume" class="slider-setting" min="0.0" max="1.0" step="0.1"
+                <input type="range" id="volume" name="volume" class="slider-setting" min="0.1" max="1.0" step="0.1"
                     value="0.8" oninput="updateVolumeValue(this.value)">
                 <span id="volumeValue">0.8</span>
             </div>
@@ -595,14 +607,17 @@
 <div id="ttsListSection" class="tts-list-section">
     <h2>Generated TTS List</h2>
     <input type="number" id="startFromId" value="1" min="1" autocomplete="off" style="width: 60px; margin-right: 10px;">
-    <button id="playPauseList" onclick="togglePlayPauseList()" style="margin-right: 10px;">Play List</button>
+    <button class="conditional-btn" id="playPauseList" onclick="togglePlayPauseList()" style="margin-right: 10px;">Play
-    <button onclick="stopPlayback()" style="margin-right: 10px;">Stop Playback</button>
+        List</button>
-    <button id="clearList" onclick="clearTTSList()" style="margin-right: 10px;">Clear List</button>
+    <button class="conditional-btn" onclick="stopPlayback()" style="margin-right: 10px;">Stop Playback</button>
-    <button onclick="exportToJSON()" style="margin-right: 10px;">Export List as JSON</button>
+    <button class="conditional-btn" id="clearList" onclick="clearTTSList()" style="margin-right: 40px;">Clear
-    <button onclick="importFromJSON()" style="margin-right: 10px;">Import JSON to list</button>
+        List</button>
-    <button onclick="exportMP3()" style="margin-right: 10px;">Export to Wav</button>
+    <button onclick="importFromJSON()" style="margin-right: 10px;">Import JSON</button>
-    <label for="splitEvery">Split every:</label>
+    <button class="conditional-btn" onclick="exportToJSON()" style="margin-right: 10px;">Export JSON</button>
-    <select id="splitEvery" name="splitEvery">
+    <button class="conditional-btn" id="exportsrttitles" style="margin-right: 10px;">Export SRT</button>
+    <button class="conditional-btn" onclick="exportMP3()" style="margin-right: 10px;">Export to WAV</button>
+    <label for="splitEvery">Export Split:</label>
+    <select id="splitEvery" name="splitEvery" style="margin-right: 40px;">
         <option value="100">100</option>
         <option value="200">200</option>
         <option value="300">300</option>
@@ -610,22 +625,38 @@
         <option value="500" selected>500</option>
         <option value="600">600</option>
     </select>
+    <button class="conditional-btn" id="saveAndAnalyzeTTS" style="margin-right: 10px;">Analyze TTS</button>
+    <label for="AnalyzeAccuracy">% Accuracy:</label>
+    <select id="AnalyzeAccuracy" name="AnalyzeAccuracy" style="margin-right: 10px;">
+        <option value="99">99</option>
+        <option value="98" selected>98</option>
+        <option value="97">97</option>
+        <option value="96">96</option>
+        <option value="95">95</option>
+        <option value="94">94</option>
+        <option value="93">93</option>
+        <option value="92">92</option>
+        <option value="91">91</option>
+        <option value="90">90</option>
+    </select>
     <div id="paginationControls">
-        <button id="prevPage" class="pagination-btn">Previous</button>
+        <button class="conditional-btn pagination-btn" id="prevPage">Previous</button>
         <span id="pageIndicator">Page 1 of X</span>
-        <button id="nextPage" class="pagination-btn">Next</button>
+        <button class="conditional-btn pagination-btn" id="nextPage">Next</button>
-        <input type="number" id="jumpToPage" class="pagination-input" min="1" placeholder="Jump to Page">
+        <input type="number" id="jumpToPage" class="conditional-btn pagination-input" min="1"
-        <button id="jumpButton" class="pagination-btn">Go</button>
+            placeholder="Jump to Page">
+        <button class="conditional-btn pagination-btn" id="jumpButton">Go</button>
     </div>
     <div id="ttsListDisplay" class="tts-list-display">
         <!-- Dynamically populated TTS details will go here -->
     </div>
     <div id="paginationControls1">
-        <button id="prevPage1" class="pagination-btn">Previous</button>
+        <button class="conditional-btn pagination-btn" id="prevPage1">Previous</button>
         <span id="pageIndicator1">Page 1 of X</span>
-        <button id="nextPage1" class="pagination-btn">Next</button>
+        <button class="conditional-btn pagination-btn" id="nextPage1">Next</button>
-        <input type="number" id="jumpToPage1" class="pagination-input" min="1" placeholder="Jump to Page">
+        <input type="number" id="jumpToPage1" class="conditional-btn pagination-input" min="1"
-        <button id="jumpButton1" class="pagination-btn">Go</button>
+            placeholder="Jump to Page">
+        <button class="conditional-btn pagination-btn" id="jumpButton1">Go</button>
     </div>
 </div>
 </div>
@@ -686,8 +717,6 @@
 
 </script>
 
-
-
 <script>
 
     /////////////////////////////////////////
@@ -704,6 +733,7 @@
     let isStreamInitialized = false;
     let characterVoice = document.getElementById("character_voice_gen").value;
     let ttsList = []; // Global list to store TTS details
+    let problematicIds = []; // Global list to store ProblematicIDs
     const howlInstances = {};
     let listPlaying = false;
     let listPlayIndex = 0;
@@ -759,6 +789,123 @@
     }
 
 
+    //////////////////////////////////
+    //////// Compare TTS list ////////
+    //////////////////////////////////
+
+    document.getElementById('saveAndAnalyzeTTS').addEventListener('click', async function () {
+        showProgress(); // Call this function to show the progress indicator at the start of the process
+        try {
+            // Serialize the ttsList data structure to a JSON string
+            const ttsDataJSON = JSON.stringify(ttsList);
+
+            // Send the serialized ttsList to the '/save-tts-data' endpoint
+            const saveResponse = await fetch('/api/save-tts-data', {
+                method: 'POST',
+                headers: { 'Content-Type': 'application/json' },
+                body: ttsDataJSON  // Use the serialized JSON string as the request body
+            });
+
+            const saveData = await saveResponse.json();
+
+            // Log the server's response
+            console.log(saveData.message);
+
+            // If the server confirms the data was saved successfully, trigger analysis
+            if (saveData.message === "Data saved successfully") {
+                const accuracy = parseInt(document.getElementById('AnalyzeAccuracy').value) || 98;
+                const analyzeResponse = await fetch(`/api/trigger-analysis?threshold=${accuracy}`);
+                const analyzeData = await analyzeResponse.json();
+                if (analyzeData.summary.flagged_ids && analyzeData.summary.flagged_ids.length > 0) {
+                    // Update the global problematicIds array with the new flagged IDs
+                    problematicIds = analyzeData.summary.flagged_ids;
+                    console.log("Initial problematicIds from analysis:", problematicIds); // Debugging: Check the initial array after receiving analysis results
+
+                    // Call highlightProblems with the array of flagged IDs
+                    highlightProblems(problematicIds);
+                } else {
+                    console.log("No problematic IDs or error retrieving data.");
+
+                    // Clear the problematicIds array since there are no problematic IDs
+                    problematicIds = [];
+                    highlightProblems(problematicIds);
+                }
+                // Optionally, you might want to provide feedback in the UI about the analysis status
+            } else {
+                throw new Error('Data save was unsuccessful.');
+            }
+        } catch (error) {
+            // Handle any errors that occurred during the fetch process
+            console.error('Error:', error);
+        } finally {
+            hideProgress(); // Hide the progress indicator when the process is complete or if an error occurs
+        }
+    });
+
+    function highlightProblems(idsToHighlight) {
+        // First, clear all previous 'problematic' highlights
+        document.querySelectorAll('.problematic').forEach(item => {
+            item.classList.remove('problematic');
+        });
+
+        // Now, highlight all the IDs that should still be problematic
+        idsToHighlight.forEach(id => {
+            const itemElement = document.getElementById(`item-${id}`);
+            if (itemElement) {
+                itemElement.classList.add('problematic');
+            }
+        });
+    }
+
+    /////////////////////////////////
+    //////// Export SRT Subs ////////
+    /////////////////////////////////
+
+    document.getElementById('exportsrttitles').addEventListener('click', async function () {
+        showProgress(); // Call this function to show the progress indicator at the start of the process
+        try {
+            const ttsDataJSON = JSON.stringify(ttsList);
+
+            const saveResponse = await fetch('/api/save-tts-data', {
+                method: 'POST',
+                headers: { 'Content-Type': 'application/json' },
+                body: ttsDataJSON
+            });
+            const saveData = await saveResponse.json();
+            console.log(saveData.message);
+
+            if (saveData.message === "Data saved successfully") {
+                // Trigger the SRT generation and handle file download
+                const analyzeResponse = await fetch(`/api/srt-generation`);
+                if (analyzeResponse.ok) {
+                    // Create a temporary anchor element
+                    const a = document.createElement('a');
+                    // Create a blob URL from the response
+                    const blob = await analyzeResponse.blob();
+                    const url = window.URL.createObjectURL(blob);
+                    a.href = url;
+                    // Set the download name for the file
+                    a.download = "subtitles.srt";
+                    // Append the anchor to the document
+                    document.body.appendChild(a);
+                    // Trigger the download
+                    a.click();
+                    // Clean up by removing the anchor element and revoking the blob URL
+                    window.URL.revokeObjectURL(url);
+                    a.remove();
+                } else {
+                    console.log("Error running SRT generation.");
+                }
+            } else {
+                throw new Error('Data save was unsuccessful.');
+            }
+        } catch (error) {
+            console.error('Error:', error);
+        } finally {
+            hideProgress(); // Hide the progress indicator when the process is complete or if an error occurs
+        }
+    });
+
     ///////////////////////////////////////////
     //////// TTS Generation Activities ////////
     ///////////////////////////////////////////
@@ -777,7 +924,14 @@
             let cleanedText = textInput.replace(/ \- | \– /g, ' ');
             cleanedText = cleanedText.replace(/%/g, " percent");
             cleanedText = cleanedText.replace(/\.\s'/g, ".'");
+            cleanedText = cleanedText.replace(/(?<!\w)[']+(?!\w)/g, '');
+            cleanedText = cleanedText.replace(/"/g, "'");
             cleanedText = cleanedText.replace(/[^a-zA-Z0-9\s\.,;:!?\-\'"À-ÿ\u0400-\u04FF\u0150\u0151\u0170\u0171]\$/g, '');
+            // Replace dash followed by space with comma and space
+            cleanedText = cleanedText.replace(/-\s/g, '; ');
+            // Replace double dash "--" with semicolon "; "
+            cleanedText = cleanedText.replace(/--/g, '; ');
+
             let sentences = cleanedText.split(/(?<=[.!?:])\s+/);
 
             if (!sentences) {
@@ -886,6 +1040,7 @@
 
                 // Refresh the list display to reflect the new pagination state
                 updateUIList();
+                updateButtonStyles()
 
                 // Update the audio queue and UI
                 const sound = new Howl({
@@ -902,11 +1057,50 @@
                     playNextLocalChunk();
                 }
             }
+            return result
         } catch (error) {
             console.error("Error in generateAndCache:", error);
         }
     }
 
+    async function regenerateTTS(id) {
+        console.log("Before regeneration:", problematicIds); // Debugging: Check the array before regeneration
+
+        // Find the details of the TTS to regenerate
+        const ttsDetails = ttsList.find(item => item.id === id);
+        if (!ttsDetails) return;
+
+        // Retrieve the current volume setting from the HTML input
+        const currentVolume = parseFloat(document.getElementById("volume").value);
+
+        // Call the TTS generation with the same details and current volume
+        const result = await generateAndCache(ttsDetails.text, ttsDetails.language, ttsDetails.characterVoice, currentVolume, id);
+
+        if (result && result.status === "generate-success") {
+            const index = problematicIds.indexOf(id);
+            if (index !== -1) {
+                problematicIds.splice(index, 1); // Remove the id from problematicIds
+                console.log(`ID ${id} removed from problematicIds. Updated array:`, problematicIds); // Debugging: Check the array after removing an ID
+            }
+
+            // Find the element for this ID and remove the 'problematic' class
+            const problemItemElement = document.getElementById(`item-${id}`);
+            if (problemItemElement) {
+                problemItemElement.classList.remove('problematic');
+            }
+        } else {
+            console.log(`Regeneration failed for ID ${id}.`, result); // If regeneration fails, log the result
+        }
+
+        // Refresh the UI list to reflect the new data
+        updateUIList(); // Assume this function exists and it updates the display list
+        // Reapply highlighting to the items still in problematicIds
+        highlightProblems(problematicIds);
+        console.log("After regeneration and re-highlighting:", problematicIds); // Debugging: Check the array after re-highlighting
+    }
+
+
+
     async function generateWithoutPlayback(chunk, language, characterVoice, id = null) {
         // Perform the generation process as usual but do not play the audio
         try {
@@ -972,19 +1166,6 @@
         }
     }
 
-
-    async function regenerateTTS(id) {
-        // Find the details of the TTS to regenerate
-        const ttsDetails = ttsList.find(item => item.id === id);
-        if (!ttsDetails) return;
-
-        // Retrieve the current volume setting from the HTML input
-        const currentVolume = parseFloat(document.getElementById("volume").value);
-
-        // Call the TTS generation with the same details and current volume
-        await generateAndCache(ttsDetails.text, ttsDetails.language, ttsDetails.characterVoice, currentVolume, id);
-    }
-
     function updateUIList() {
         const listDisplay = document.getElementById('ttsListDisplay');
         if (!listDisplay) {
@@ -1029,7 +1210,9 @@
         document.getElementById('pageIndicator').innerText = `Page ${currentPage} of ${lastPage}`;
         document.getElementById('pageIndicator1').innerText = `Page ${currentPage} of ${lastPage}`;
         document.getElementById('prevPage').disabled = (currentPage === 1);
+        document.getElementById('prevPage1').disabled = (currentPage === 1);
         document.getElementById('nextPage').disabled = (currentPage === lastPage);
+        document.getElementById('nextPage1').disabled = (currentPage === lastPage);
     }
 
 
@@ -1240,7 +1423,7 @@
         } else {
             // In chunk mode, toggle the chunk play/pause
             if (isPlaying && currentlyPlayingIndex > 0) {
-                let currentSound = audioQueue[currentlyPlayingIndex - 1];
+                let currentSound = audioQueue[currentlyPlayingIndex - 3];
                 if (currentSound.playing()) {
                     currentSound.pause();
                 } else {
@@ -1267,16 +1450,6 @@
         }
     }
 
-
-    function stopPlayback() {
-        Howler.stop();
-        audioQueue = [];
-        // Unload each sound in the queue
-        audioQueue.forEach(sound => sound.unload());
-        currentlyPlayingIndex = 0;
-        isPlaying = false;
-    }
-
     function updatePlaybackButton() {
         const playbackButton = document.querySelector(".toggle-playback");
 
@@ -1451,38 +1624,64 @@
         if (pageIndicator1) {
             pageIndicator1.textContent = 'Page 1 of 1';
         }
+        updateButtonStyles()
+    }
+
+
+    function encodeHtmlEntities(text) {
+        return text.replace(/"/g, '&quot;').replace(/'/g, '&#39;');
+    }
+
+    function decodeHtmlEntities(text) {
+        var textArea = document.createElement('textarea');
+        textArea.innerHTML = text;
+        return textArea.value;
     }
 
     function editText(id) {
+        console.log(`Initiating edit for ID: ${id}`);
+        console.log("Before edit, ttsList:", JSON.parse(JSON.stringify(ttsList)));
         // First, check if we're already editing and should save the current value
         const existingInput = document.getElementById(`edit-input-${id}`);
         if (existingInput) {
+            console.log(`Existing input found for ID: ${id}, triggering blur event to save.`);
             existingInput.blur(); // This will trigger the blur event, saving the current value
             return; // Don't proceed further since we're just saving the existing input
         }
-
         const entryDiv = document.getElementById(`item-${id}`);
         if (!entryDiv) {
             console.error(`Item with ID ${id} not found.`);
             return;
         }
-
         const textPreviewSpan = entryDiv.querySelector('.text-preview');
-        const fullText = textPreviewSpan.getAttribute('data-fulltext');
+        // Retrieve the full text and immediately encode it
+        let fullText = textPreviewSpan.getAttribute('data-fulltext');
+        fullText = decodeHtmlEntities(fullText);
+        //fullText = encodeHtmlEntities(fullText);
+        console.log(`Current full text for ID: ${id} is: ${fullText}`);
 
         // Create an input field for editing
         const inputField = document.createElement('input');
         inputField.type = 'text';
+        // The value is already encoded at this point
         inputField.value = fullText;
         inputField.classList.add('edit-input');
         inputField.id = `edit-input-${id}`; // Assign an ID to the input field for later access
 
         // Handle saving the edited text
         inputField.addEventListener('blur', function () {
+            console.log(`Saving edited text for ID: ${id}: ${this.value}`);
             // Update the text in the ttsList array and the data-fulltext attribute
             const itemIndex = ttsList.findIndex(item => item.id === id);
             if (itemIndex !== -1) {
+                console.log(`Updating ttsList with new text for ID: ${id}`);
                 ttsList[itemIndex].text = this.value;
+                // If this item is in the list of problematic IDs, un-highlight it
+                if (problematicIds.includes(id)) {
+                    console.log(`ID: ${id} is problematic, removing highlight class`);
+                    const problemItemElement = document.getElementById(`item-${id}`);
+                    problemItemElement.classList.remove('problematic');
+                }
             }
 
             // Create a new span element to replace the input field
@@ -1494,15 +1693,18 @@
             newTextPreviewSpan.textContent = displayText;
 
             // Replace the input field with the new span
+            console.log(`Replacing input field with text span for ID: ${id}`);
             inputField.replaceWith(newTextPreviewSpan);
         });
 
         // Replace the span with the input field
+        console.log(`Replacing text span with input field for ID: ${id}`);
         textPreviewSpan.replaceWith(inputField);
 
         // Focus the input field and select its content to facilitate editing
         inputField.focus();
         inputField.select();
+        console.log("After edit, ttsList:", JSON.parse(JSON.stringify(ttsList)));
     }
 
     ////////////////////////////////////////
@@ -1550,6 +1752,7 @@
                         ttsList = jsonData;
                         currentPage = 1; // Reset to the first page
                         updateUIList();
+                        updateButtonStyles()
                         updatePaginationControls(); // Update pagination controls
                     } catch (error) {
                         console.error('Error parsing JSON:', error);
@@ -1563,34 +1766,42 @@
 
     function updatePaginationControls() {
         const lastPage = Math.ceil(ttsList.length / itemsPerPage);
-        const pageIndicator = document.getElementById('pageIndicator');
+        const prevPageButtons = document.querySelectorAll('.pagination-btn[id^="prevPage"]');
-        if (pageIndicator) {
+        const nextPageButtons = document.querySelectorAll('.pagination-btn[id^="nextPage"]');
-            pageIndicator.textContent = `Page ${currentPage} of ${lastPage}`;
-        }
 
-        const prevPageButton = document.getElementById('prevPage');
+        // Update page indicators and buttons
-        if (prevPageButton) {
+        prevPageButtons.forEach(btn => btn.disabled = (currentPage === 1));
-            prevPageButton.disabled = (currentPage === 1);
+        nextPageButtons.forEach(btn => btn.disabled = (currentPage === lastPage));
-        }
 
-        const nextPageButton = document.getElementById('nextPage');
+        // Assuming the page indicators are consistent with button IDs
-        if (nextPageButton) {
+        document.querySelectorAll('span[id^="pageIndicator"]').forEach(indicator => {
-            nextPageButton.disabled = (currentPage === lastPage);
+            indicator.textContent = `Page ${currentPage} of ${lastPage}`;
-        }
+        });
 
-        // Similar updates for any additional pagination controls if you have them
+        // Update 'Jump to Page' inputs and 'Go' buttons
-        // ...
+        const jumpToPageInputs = document.querySelectorAll('.pagination-input[id^="jumpToPage"]');
+        const jumpButtons = document.querySelectorAll('.pagination-btn[id^="jumpButton"]');
+        const moreThanOnePage = lastPage > 1;
+
+        jumpToPageInputs.forEach(input => input.disabled = !moreThanOnePage);
+        jumpButtons.forEach(button => button.disabled = !moreThanOnePage);
     }
 
-
-
-
-
-
     //////////////////////////////
     //////// Other Things ////////
     //////////////////////////////
 
+    function updateButtonStyles() {
+        const ttsList = document.getElementById('ttsListDisplay');
+        const buttons = document.querySelectorAll('.conditional-btn');
+
+        if (!ttsList.hasChildNodes()) { // Check if the list is empty
+            buttons.forEach(button => button.disabled = true); // Disable buttons
+        } else {
+            buttons.forEach(button => button.disabled = false); // Enable buttons
+        }
+    }
+
     function updateLanguage(language) {
         // Update the language in your application logic
         console.log("Selected language:", language);
@@ -1663,10 +1874,10 @@
     function stopPlayback() {
         Howler.stop();
         audioQueue = [];
-        currentlyPlayingIndex = 0;
+        //currentlyPlayingIndex = 0;
         isPlaying = false;
-        highlightPlaying(null);
+        //highlightPlaying(null);
-        document.getElementById('startFromId').value = '1'; // Reset the startFromId input to 1
+        //document.getElementById('startFromId').value = '1'; // Reset the startFromId input to 1
     }
 
     function displayOutput(result) {
@@ -1836,6 +2047,7 @@
     function setupPage() {
         // Fetch voices or other initial data
         fetchVoices();
+        updateButtonStyles()
 
         // Check for dark mode preference in localStorage, default to enabled if not set
         let isDarkModeEnabled = localStorage.getItem('darkMode');
@@ -1989,8 +2201,8 @@
         setupPage();
         hideProgress(); // Ensure the progress indicator is hidden on page load
         if (window.location.protocol !== 'http:' && window.location.protocol !== 'https:') {
-                alert('AllTalk TTS Generator is not designed to be run from your disk folder. Please run it from the correct URL e.g. http://127.0.0.1:7851/static/tts_chunk_player/tts_chunk_player.html');
+            alert('AllTalk TTS Generator is not designed to be run from your disk folder. Please run it from the correct URL e.g. http://127.0.0.1:7851/static/tts_chunk_player/tts_chunk_player.html');
-            }
+        }
     };
 
 </script>

system/tts_srt/tts_srt.py

@@ -0,0 +1,77 @@
+import json
+import argparse
+import soundfile as sf
+from pathlib import Path
+
+parser = argparse.ArgumentParser(description="Compare TTS output with the original text using detailed comparison.")
+parser.add_argument("--ttslistpath", help="Path to the ttsList.json file")
+parser.add_argument("--wavfilespath", help="Path to the wav outputs folder")
+args = parser.parse_args()
+
+if not args.ttslistpath:
+    parser.error("[AllTalk TTSSRT] Please specify the path to the ttslist.json file using the --ttslistpath argument.")
+if not args.wavfilespath:
+    parser.error("[AllTalk TTSSRT] Please specify the path to the wavs output folder file using the --wavfilespath argument.")
+
+json_file_path = Path(args.ttslistpath).resolve()
+audio_files_base_path = Path(args.wavfilespath).resolve()
+
+print(f"[AllTalk TTSSRT]")
+print("[AllTalk TTSSRT] \033[92mStarting SRT Subtitle generation:\033[0m")
+print("[AllTalk TTSSRT]")
+print("[AllTalk TTSSRT] \033[94mJSON file  :\033[92m", json_file_path, "\033[0m")
+print("[AllTalk TTSSRT] \033[94mWAV files  :\033[92m", audio_files_base_path, "\033[0m")
+print("[AllTalk TTSSRT]")
+
+# Function to convert time in seconds to SRT time format
+def format_srt_time(seconds):
+    hours = int(seconds // 3600)
+    minutes = int((seconds % 3600) // 60)
+    seconds = seconds % 60
+    milliseconds = int((seconds - int(seconds)) * 1000)
+    return f"{hours:02}:{minutes:02}:{int(seconds):02},{milliseconds:03}"
+
+# Load the JSON data
+with open(json_file_path, 'r') as json_file:
+    data = json.load(json_file)
+
+# Initialize start time and SRT lines
+start_time = 0.0
+srt_lines = []
+
+# Loop through the JSON data to build SRT entries
+for entry in data:
+    file_name = Path(entry['fileUrl']).name  # Get the file name from the URL
+    file_path = audio_files_base_path / file_name
+    print(f"[AllTalk TTSSRT] Processing file: {file_name}")
+
+    if not file_path.exists():
+        print(f"[AllTalk TTSSRT] Audio file does not exist: {file_path}")
+        continue
+
+    try:
+        # Use soundfile to read the duration of the WAV file
+        with sf.SoundFile(str(file_path), 'r') as f:
+            duration = len(f) / f.samplerate
+            end_time = start_time + duration
+            
+            # Format the SRT entry
+            start_timestamp = format_srt_time(start_time)
+            end_timestamp = format_srt_time(end_time)
+            srt_lines.append(f"{len(srt_lines)+1}\n{start_timestamp} --> {end_timestamp}\n{entry['text']}\n")
+            
+            # Update the start time for the next entry
+            start_time = end_time
+    except RuntimeError as e:
+        print(f"[AllTalk TTSSRT] Error reading {file_path}: {e}")
+
+# Join all SRT lines into the final SRT content
+srt_content = "\n".join(srt_lines)
+
+# Save the SRT content to a file
+srt_file_path = audio_files_base_path / "subtitles.srt"
+with open(srt_file_path, "w") as srt_file:
+    srt_file.write(srt_content)
+
+print("[AllTalk TTSSRT]")
+print(f"[AllTalk TTSSRT] \033[94mSRT file created at \033[0m{srt_file_path}")

tts_server.py

@@ -19,6 +19,7 @@ from fastapi import (
     Response,
     Depends,
     HTTPException,
+    Query,
 )
 from fastapi.responses import JSONResponse, HTMLResponse, RedirectResponse, FileResponse, StreamingResponse
 from fastapi.middleware.cors import CORSMiddleware
@@ -34,7 +35,7 @@ this_dir = Path(__file__).parent.resolve()
 # STARTUP VARIABLE - Set "device" to cuda if exists, otherwise cpu
 device = "cuda" if torch.cuda.is_available() else "cpu"
 # STARTUP VARIABLE - Import languges file for Gradio to be able to display them in the interface
-with open(this_dir / "languages.json", encoding="utf8") as f:
+with open(this_dir / "system" / "config" / "languages.json", encoding="utf8") as f:
     languages = json.load(f)
 # Base setting for a possible FineTuned model existing and the loader being available
 tts_method_xtts_ft = False
@@ -638,7 +639,7 @@ def list_files(directory):
 #############################
 
 # Create an instance of Jinja2Templates for rendering HTML templates
-templates = Jinja2Templates(directory=this_dir / "templates")
+templates = Jinja2Templates(directory=this_dir / "system")
 
 # Create a dependency to get the current JSON data
 def get_json_data():
@@ -653,7 +654,7 @@ async def get_settings(request: Request):
     wav_files = list_files(this_dir / "voices")
     # Render the template with the current JSON data and list of WAV files
     return templates.TemplateResponse(
-        "generate_form.html",
+        "/at_admin/at_settings.html",
         {
             "request": request,
             "data": get_json_data(),
@@ -663,7 +664,7 @@ async def get_settings(request: Request):
     )
 
 # Define an endpoint to serve static files
-app.mount("/static", StaticFiles(directory=str(this_dir / "templates")), name="static")
+app.mount("/static", StaticFiles(directory=str(this_dir / "system")), name="static")
 
 @app.post("/update-settings")
 async def update_settings(
@@ -870,7 +871,7 @@ except OSError:
         print(f"[{params['branding']}Startup] \033[91mInfo\033[0m On Linux, you can use the following command to install PortAudio:")
         print(f"[{params['branding']}Startup] \033[91mInfo\033[0m sudo apt-get install portaudio19-dev")
 
-from typing import Union, Dict
+from typing import Union, Dict, List
 from pydantic import BaseModel, ValidationError, Field
 
 def play_audio(file_path, volume):
@@ -1145,7 +1146,69 @@ def get_current_settings():
 #### Word Add-in Sharing ####
 #############################
 # Mount the static files from the 'word_addin' directory
-app.mount("/api/word_addin", StaticFiles(directory=os.path.join(this_dir / 'templates' / 'word_addin')), name="word_addin")
+app.mount("/api/word_addin", StaticFiles(directory=os.path.join(this_dir / 'system' / 'word_addin')), name="word_addin")
+
+#############################################
+#### TTS Generator Comparision Endpoints ####
+#############################################
+import subprocess
+import aiofiles
+
+class TTSItem(BaseModel):
+    id: int
+    fileUrl: str
+    text: str
+    characterVoice: str
+    language: str
+
+class TTSData(BaseModel):
+    ttsList: List[TTSItem]
+
+@app.post("/api/save-tts-data")
+async def save_tts_data(tts_data: List[TTSItem]):
+    # Convert the list of Pydantic models to a list of dictionaries
+    tts_data_list = [item.dict() for item in tts_data]
+    # Serialize the list of dictionaries to a JSON string
+    tts_data_json = json.dumps(tts_data_list, indent=4)
+    async with aiofiles.open(this_dir / "outputs" / "ttsList.json", 'w') as f:
+        await f.write(tts_data_json)
+    return {"message": "Data saved successfully"}
+
+import sys
+
+@app.get("/api/trigger-analysis")
+async def trigger_analysis(threshold: int = Query(default=98)):
+    venv_path = sys.prefix
+    env = os.environ.copy()
+    env["PATH"] = os.path.join(venv_path, "bin") + ":" + env["PATH"]
+    ttslist_path = this_dir / "outputs" / "ttsList.json"
+    wavfile_path = this_dir / "outputs"
+    subprocess.run(["python", "tts_diff.py", f"--threshold={threshold}", f"--ttslistpath={ttslist_path}", f"--wavfilespath={wavfile_path}"], cwd=this_dir / "system" / "tts_diff", env=env)
+    # Read the analysis summary
+    try:
+        with open(this_dir / "outputs" / "analysis_summary.json", "r") as summary_file:
+            summary_data = json.load(summary_file)
+    except FileNotFoundError:
+        summary_data = {"error": "Analysis summary file not found."}
+    return {"message": "Analysis Completed", "summary": summary_data}
+
+#################################################
+#### TTS Generator SRT Subtitiles generation ####
+#################################################
+@app.get("/api/srt-generation")
+async def srt_generation():
+    venv_path = sys.prefix
+    env = os.environ.copy()
+    env["PATH"] = os.path.join(venv_path, "bin") + ":" + env["PATH"]
+    ttslist_path = this_dir / "outputs" / "ttsList.json"
+    wavfile_path = this_dir / "outputs"
+    subprocess.run(["python", "tts_srt.py", f"--ttslistpath={ttslist_path}", f"--wavfilespath={wavfile_path}"], cwd=this_dir / "system" / "tts_srt", env=env)
+    
+    srt_file_path = this_dir / "outputs" / "subtitles.srt"
+    if not srt_file_path.exists():
+        raise HTTPException(status_code=404, detail="Subtitle file not found.")
+    
+    return FileResponse(path=srt_file_path, filename="subtitles.srt", media_type='application/octet-stream')
 
 ###################################################
 #### Webserver Startup & Initial model Loading ####

voices/Clint_Eastwood CC3.txt

@@ -9,4 +9,6 @@ Author	BBC
 This file is licensed under the Creative Commons Attribution 3.0 Unported license.
 https://creativecommons.org/licenses/by/3.0/deed.en
 
-Downsampled to 22050hz, mono and cut shorter
+Downsampled to 22050hz, mono and cut shorter.
+
+This license applies to both the original file ( «Clint_Eastwood CC3.wav» ) and it's derivative, the enhanced file ( «Clint_Eastwood (enhanced)CC3.wav» ).