{"id":16128,"date":"2026-11-17T11:04:30","date_gmt":"2026-11-17T02:04:30","guid":{"rendered":"https:\/\/localmodelwatch.tsuchitsuchi.com\/?p=16128"},"modified":"2026-11-17T13:12:36","modified_gmt":"2026-11-17T04:12:36","slug":"autotrust-tts-v1-chinese-tts-model","status":"publish","type":"post","link":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/autotrust-tts-v1-chinese-tts-model\/","title":{"rendered":"TTS-V1 Chinese Text-to-Speech Model: File List"},"content":{"rendered":"<p><em>Sample outputs are available on the <a href=\"https:\/\/huggingface.co\/autotrust\/TTS-V1\">model card<\/a>.<\/em><\/p>\n<p><!-- lmw:facts --><\/p>\n<h2>At a Glance<\/h2>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Item<\/th>\n<th>Value<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Repository<\/td>\n<td><a href=\"https:\/\/huggingface.co\/autotrust\/TTS-V1\">autotrust\/TTS-V1<\/a><\/td>\n<\/tr>\n<tr>\n<td>Published<\/td>\n<td>2026-11-13<\/td>\n<\/tr>\n<tr>\n<td>Formats<\/td>\n<td><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/format-safetensors-en\/\">safetensors<\/a><\/td>\n<\/tr>\n<tr>\n<td>Source type<\/td>\n<td>Primary source (the publisher itself)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>Values determined by this site&#8217;s code when the information was collected. Dates are JST.<\/em><\/p>\n<p><!-- \/lmw:facts --><\/p>\n<h2>Overview<\/h2>\n<p>The open-weight model &#8220;autotrust\/TTS-V1&#8221; has been released, capable of synthesizing speech from Chinese text.<\/p>\n<p>This model takes Chinese text as input and generates speech in 24 kHz mono WAV format. It utilizes a shared architecture that integrates FlowLM and Mimi within a single model file, allowing users to switch between four different voice presets (male and female) intended for sales and customer support applications.<\/p>\n<h2>Specifications<\/h2>\n<p>The specifications listed in the model card and repository are as follows:<\/p>\n<ul>\n<li><strong>Architecture<\/strong>: Shared model consisting of FlowLM and Mimi (settings defined in <code>config.yaml<\/code>)<\/li>\n<li><strong>Output Specifications<\/strong>: 24 kHz sampling rate, mono, WAV format<\/li>\n<li><strong>Supported Languages<\/strong>: Chinese (the Chinese tokenizer <code>tokenizer.model<\/code> is included in the repository)<\/li>\n<li><strong>Selectable Voice Presets<\/strong>: Four types are provided, with fixed reference recordings stored in the <code>voices\/<\/code> directory:<\/li>\n<li><code>sales_male<\/code> (Male voice for sales\/marketing) &#8211; <code>sales_female<\/code> (Female voice for sales\/marketing) &#8211; <code>service_male<\/code> (Male voice for customer support) &#8211; <code>service_female<\/code> (Female voice for customer support)<\/li>\n<li><strong>Text Preprocessing<\/strong>: Uses normalization scripts for currency, dates, and various numbers (<code>tn.py<\/code>) and <code>wetext<\/code> resources for speech-text normalization<\/li>\n<li><strong>Segment Control<\/strong>: Automatic text segmentation based on punctuation and token length<\/li>\n<li><strong>Inference Control Options<\/strong>: Includes <code>--attempts<\/code> (default 16) to control maximum retries per segment, <code>--seed<\/code> (e.g., 3000) for random seed, and <code>--threads<\/code> (e.g., 2) to specify CPU thread count<\/li>\n<li><strong>Distribution Format<\/strong>: safetensors format (<code>model.safetensors<\/code>)<\/li>\n<li><strong>Inference Engine<\/strong>: Includes a full set of Pocket TTS inference code adapted for this model (<code>pocket_tts\/<\/code>)<\/li>\n<li><strong>License and Terms of Use<\/strong>: The model card directs users to check <code>MODEL_CARD.md<\/code> in the repository for information regarding the origin and distribution terms of the weights and reference audio.<\/li>\n<\/ul>\n<h2>Performance<\/h2>\n<p>The correspondence between the parameter names and audio samples provided by the publisher is as follows:<\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Parameter<\/th>\n<th>Voice<\/th>\n<th>Sample<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>sales_male<\/code><\/td>\n<td>Sales Male<\/td>\n<td><a href=\"demos\/sales_male.wav\">Listen<\/a><\/td>\n<\/tr>\n<tr>\n<td><code>sales_female<\/code><\/td>\n<td>Sales Female<\/td>\n<td><a href=\"demos\/sales_female.wav\">Listen<\/a><\/td>\n<\/tr>\n<tr>\n<td><code>service_male<\/code><\/td>\n<td>Service Male<\/td>\n<td><a href=\"demos\/service_male.wav\">Listen<\/a><\/td>\n<\/tr>\n<tr>\n<td><code>service_female<\/code><\/td>\n<td>Service Female<\/td>\n<td><a href=\"demos\/service_female.wav\">Listen<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>According to the model card, the names of each voice are tentative, based on intended use and pitch; it is recommended to listen to the provided samples (WAV files) to confirm if they meet actual business requirements.<\/p>\n<p>Regarding the quality of the generation pipeline, the included inference script (<code>synthesize.py<\/code>) incorporates an automatic verification mechanism. When text is segmented by punctuation or token length for sequential synthesis, it checks for valid end-of-sentence (EOS) tokens, ensures basic volume levels, and detects unnatural silence within the audio. Segments that fail to meet these criteria are automatically re-sampled (retried). If the maximum number of attempts is reached without success, the design halts the final file write and reports an error.<\/p>\n<p>However, the publisher notes that even with this verification logic, it is not possible to completely eliminate all noise, typos, or misread numbers. Therefore, for production use, human listening or verification via automatic speech recognition (ASR) after generation is required. Additionally, among the four provided voices, the customer support male voice (<code>service_male<\/code>) is reported to have a tendency toward unstable termination behavior, sometimes requiring multiple attempts to achieve a normal end state.<\/p>\n<h2>Strengths and Use Cases<\/h2>\n<p>This model is suitable for speech synthesis (TTS) specialized for business scenarios and interactive applications in Chinese. It is intended for use cases such as guidance messages on e-commerce sites or in sales activities, and the construction of automated voice response systems for customer service.<\/p>\n<p>Synthesis examples shown in the model card include phrases like &#8220;\u60a8\u597d\uff0c\u60a8\u7684\u8ba2\u5355\u91d1\u989d\u662f235.60\u5143\uff0c\u9884\u8ba12\u4e2a\u5de5\u4f5c\u65e5\u5230\u8d26\u3002&#8221; (Hello, your order amount is 235.60 yuan, expected to arrive within 2 business days), demonstrating its suitability for reading notification messages containing order amounts, dates, and processing times.<\/p>\n<p>To support such operations, the repository includes a normalization script (<code>tn.py<\/code>) that automatically preprocesses numbers, dates, IDs, and symbols into appropriate readings. This allows for smooth conversion to speech even for business text containing a mix of characters and numbers.<\/p>\n<p>Furthermore, this model allows users to select from four voice personas (sales male, sales female, service male, service female) defined by purpose and pitch from a single shared model file. It is possible to build a system that switches between bright-toned sales voices and polite, conversational customer support voices depending on the application. As noted, official voice names are tentative, so it is recommended to verify them against provided samples.<\/p>\n<p><!-- lmw:hardware --><\/p>\n<h2>Hardware Requirements<\/h2>\n<p><strong>Estimated requirements (calculated by Local Model Watch)<\/strong> \u2014 115M parameters<\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Your VRAM<\/th>\n<th>Quantization<\/th>\n<th>File size<\/th>\n<th>Est. memory needed<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>4GB (laptop iGPU \/ phone class)<\/td>\n<td>F32<\/td>\n<td>0.4GB<\/td>\n<td>0.5GB<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>Memory estimates add a 20% runtime overhead (KV cache, etc.) to the actual size of the distributed files. Actual usage varies with context length, batch size and inference engine. These figures are computed by this site from file sizes, not published by the model&#8217;s authors. The listed files are distributed in F32 (32-bit). Loading in a lower precision (BF16 or quantized) needs less memory, so treat this as the size of the files as distributed, not a minimum. This site has not measured the lower-precision figures. Compare with other models in our <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/vram-guide-en\/\">VRAM quick reference<\/a>. What the quantization names mean: <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/glossary-quantization-en\/\">glossary<\/a>.<\/em><\/p>\n<p><!-- \/lmw:hardware --><\/p>\n<p><!-- lmw:lab --><\/p>\n<h2>Our Own Measurements<\/h2>\n<h3>Measurements We Did Not Take<\/h3>\n<p>This model does not state a license, so we cannot confirm permission to use it. Because this site carries advertising, we did not run the model (no CPU run, answers, quantization comparison, conversion or generation).<\/p>\n<p><!-- \/lmw:lab --><\/p>\n<p><!-- lmw:files --><\/p>\n<h2>Distributed Files<\/h2>\n<p><em>Weight files published in <a href=\"https:\/\/huggingface.co\/autotrust\/TTS-V1\/tree\/main\">autotrust\/TTS-V1<\/a>, listed by this site from the Hugging Face API. Sizes are the actual file sizes.<\/em><\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>File<\/th>\n<th>Size<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>model.safetensors<\/code><\/td>\n<td>459MB<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><!-- \/lmw:files --><\/p>\n<h2>How to Get It<\/h2>\n<p>Model weights and related code are available in the Hugging Face repository <code>autotrust\/TTS-V1<\/code>. The model is distributed as a single <code>model.safetensors<\/code> file, which contains both FlowLM and Mimi components.<\/p>\n<p>Linux and Python 3.12 are supported. Installation steps and dependency configurations are as follows:<\/p>\n<p>First, install <a href=\"https:\/\/pytorch.org\/get-started\/locally\/\">PyTorch<\/a> according to your environment. If running in a CPU environment, select the CPU version of PyTorch; if using an NVIDIA GPU, install the version of PyTorch corresponding to your graphics card and driver CUDA version.<\/p>\n<p>Next, create a virtual environment within the directory where you cloned or downloaded the repository and install the necessary libraries.<\/p>\n<pre><code class=\"language-bash\">python3.12 -m venv.venv\nsource.venv\/bin\/activate\npython -m pip install --upgrade pip\n# For CUDA environments, install the appropriate version of PyTorch first\npython -m pip install -r requirements.txt\n<\/code><\/pre>\n<p>The verified dependency configuration is Python 3.12, <code>torch 2.13.0<\/code>, <code>NumPy 2.3.5<\/code>, <code>soundfile 0.14.0<\/code>, and <code>wetext 0.2.1<\/code>. Note that an internet connection may be required when installing the <code>wetext<\/code> package and preparing its speech-text normalization resources. The repository includes Pocket TTS inference code (<code>pocket_tts\/<\/code>) tuned for this model.<\/p>\n<p>Speech synthesis is performed from the command line using the included <code>synthesize.py<\/code> script. The basic command to specify text and generate a WAV file using the sales female voice (<code>sales_female<\/code>) is as follows:<\/p>\n<pre><code class=\"language-bash\">python synthesize.py --voice sales_female \\\n  --text '\u60a8\u597d\uff0c\u60a8\u7684\u8ba2\u5355\u91d1\u989d\u662f235.60\u5143\uff0c\u9884\u8ba12\u4e2a\u5de5\u4f5c\u65e5\u5230\u8d26\u3002' \\\n  --output result.wav\n<\/code><\/pre>\n<p>To synthesize from a text file using the service male voice (<code>service_male<\/code>), use the <code>--text-file<\/code> option:<\/p>\n<pre><code class=\"language-bash\">python synthesize.py --voice service_male --text-file text.txt --output result.wav\n<\/code><\/pre>\n<p>The following parameters are available for inference:<\/p>\n<ul>\n<li><code>--device cuda<\/code>: Uses an NVIDIA GPU for faster inference (defaults to CPU if not specified).<\/li>\n<li><code>--attempts 16<\/code>: Controls the maximum number of retries for each text segment.<\/li>\n<li><code>--seed 3000<\/code>: Specifies a random seed to ensure reproducibility.<\/li>\n<li><code>--threads 2<\/code>: Specifies the number of threads for CPU execution.<\/li>\n<\/ul>\n<p>If a WAV file with the same name already exists at the output destination, the program is designed to refuse the operation to prevent overwriting. For details on the origin of the model weights and reference audio, as well as detailed distribution terms, please refer to <code>MODEL_CARD.md<\/code> included in the repository.<\/p>\n<p><!-- lmw:related --><\/p>\n<h2>Related Articles<\/h2>\n<ul>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/autotrust-tts-v5-overview\/\">TTS-V5 Chinese TTS Model: 4GB+ VRAM, File List<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/autotrust-tts-v4-chinese-call-center-tts\/\">TTS-V4 Chinese Text-to-Speech Model: 8GB+ VRAM, File List<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/autotrust-tts-v3-released\/\">TTS-V3 Phone-Dialogue Specialized TTS Model: 8GB+ VRAM, File List<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/qwen3-8-27b-nvfp4-64k-mtp\/\">Qwen3.8-27B-NVFP4-DGX-SPARK-V2 NVFP4 Quantized 27B Model: 24GB+ VRAM<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:related --><\/p>\n<p><!-- lmw:next-steps --><\/p>\n<h2>What to Read Next<\/h2>\n<ul>\n<li><strong>Find models by VRAM<\/strong> (This model runs from the 4GB tier) \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/vram-guide-8gb-en\/\">Other models that run on a 8GB GPU<\/a><\/li>\n<li><strong>Formats this model is available in<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/format-safetensors-en\/\">Safetensors format guide and models<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:next-steps --><\/p>\n<h2>Sources<\/h2>\n<ul>\n<li><a href=\"https:\/\/huggingface.co\/autotrust\/TTS-V1\">https:\/\/huggingface.co\/autotrust\/TTS-V1<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>An overview of autotrust\/TTS-V1, an open-weight Chinese TTS model featuring four voice presets and integrated normalization for business applications.<\/p>\n","protected":false},"author":1,"featured_media":16127,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[203],"tags":[4024,4967,4969,4971,4973,1547,213],"class_list":["post-16128","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-image-video-and-audio","tag-autotrust-en","tag-autotrust-tts-v1-en","tag-flowlm-en","tag-mimi-en","tag-tts-v1-en","tag-verified","tag--en"],"lang":"en","translations":{"en":16128,"ja":16126},"_links":{"self":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/16128","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/comments?post=16128"}],"version-history":[{"count":3,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/16128\/revisions"}],"predecessor-version":[{"id":18188,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/16128\/revisions\/18188"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media\/16127"}],"wp:attachment":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media?parent=16128"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/categories?post=16128"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/tags?post=16128"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}