Kani-TTS (ur, vi)

.gitattributes

@@ -202,3 +202,15 @@ zh/kani-tts-400m-zh-GGUF/kani-tts-400m-zh.Q5_K_M.gguf filter=lfs diff=lfs merge=
 zh/kani-tts-400m-zh-GGUF/kani-tts-400m-zh.Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text
 zh/kani-tts-400m-zh-GGUF/kani-tts-400m-zh.Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
 zh/kani-tts-400m-zh-GGUF/kani-tts-400m-zh.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.f16.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
+ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text

ur/kani-tts-400m-ur-GGUF/.gitattributes

@@ -0,0 +1,47 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
+kani-tts-400m-ur.f16.gguf filter=lfs diff=lfs merge=lfs -text

ur/kani-tts-400m-ur-GGUF/README.md

@@ -0,0 +1,74 @@
+---
+base_model: mahwizzzz/kani-tts-400m-ur
+datasets:
+- mahwizzzz/UAT
+language:
+- ur
+library_name: transformers
+license: apache-2.0
+mradermacher:
+  readme_rev: 1
+quantized_by: mradermacher
+---
+## About
+
+<!-- ### quantize_version: 2 -->
+<!-- ### output_tensor_quantised: 1 -->
+<!-- ### convert_type: hf -->
+<!-- ### vocab_type:  -->
+<!-- ### tags:  -->
+<!-- ### quants:  x-f16 Q4_K_S Q2_K Q8_0 Q6_K Q3_K_M Q3_K_S Q3_K_L Q4_K_M Q5_K_S Q5_K_M IQ4_XS -->
+<!-- ### quants_skip:  -->
+<!-- ### skip_mmproj:  -->
+static quants of https://huggingface.co/mahwizzzz/kani-tts-400m-ur
+
+<!-- provided-files -->
+
+***For a convenient overview and download list, visit our [model page for this model](https://hf.tst.eu/model#kani-tts-400m-ur-GGUF).***
+
+weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
+## Usage
+
+If you are unsure how to use GGUF files, refer to one of [TheBloke's
+READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
+more details, including on how to concatenate multi-part files.
+
+## Provided Quants
+
+(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
+
+| Link | Type | Size/GB | Notes |
+|:-----|:-----|--------:|:------|
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q2_K.gguf) | Q2_K | 0.3 |  |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q3_K_S.gguf) | Q3_K_S | 0.3 |  |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q3_K_M.gguf) | Q3_K_M | 0.3 | lower quality |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q3_K_L.gguf) | Q3_K_L | 0.3 |  |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.IQ4_XS.gguf) | IQ4_XS | 0.3 |  |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q4_K_S.gguf) | Q4_K_S | 0.3 | fast, recommended |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q4_K_M.gguf) | Q4_K_M | 0.3 | fast, recommended |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q5_K_S.gguf) | Q5_K_S | 0.4 |  |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q5_K_M.gguf) | Q5_K_M | 0.4 |  |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q6_K.gguf) | Q6_K | 0.4 | very good quality |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.Q8_0.gguf) | Q8_0 | 0.5 | fast, best quality |
+| [GGUF](https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF/resolve/main/kani-tts-400m-ur.f16.gguf) | f16 | 0.8 | 16 bpw, overkill |
+
+Here is a handy graph by ikawrakow comparing some lower-quality quant
+types (lower is better):
+
+![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
+
+And here are Artefact2's thoughts on the matter:
+https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
+
+## FAQ / Model Request
+
+See https://huggingface.co/mradermacher/model_requests for some answers to
+questions you might have and/or if you want some other model quantized.
+
+## Thanks
+
+I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
+me use its servers and providing upgrades to my workstation to enable
+this work in my free time.
+
+<!-- end -->

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.IQ4_XS.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95e149fdf02edf4333657dc14e78fb04ba96961b338e1deba5400a9bc74373df
+size 224887744

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q2_K.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cbbbd78a24df6bdaab2f1ff9e74a2d60221c76721f0eb98c1ef45f617b0bdd3
+size 173679552

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q3_K_L.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdcd992fcf17b1bf75bd0aebd868980a5bb77459d76f0974b5926494450f2f07
+size 216130496

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q3_K_M.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:72a2d7eceb89a2545a421eb22ec532d9573d0920d6b78caace61be90e6987f71
+size 206234560

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q3_K_S.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d388da2a7ca12dd6806181c4e9463340703b9f2f46650280a69e8aad3e06c2fc
+size 194233280

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q4_K_M.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:475b96d9f86123275910f92deddadb47d30ccd2809edf31e4b7c1ac71c03436a
+size 242394048

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q4_K_S.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8543cae620937425084d9d13b3330bc93137d66dac57331068761b02a7be137
+size 233833408

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q5_K_M.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c55135e541f840112a45fe0579b2543ccec54954e3063a9af7af09128196a4df
+size 273458112

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q5_K_S.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebf9f0d38fccb6a09711116b838a1c7bab7ff9aec6253c83fa91e5984460cb1e
+size 268305344

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q6_K.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce0cdee2e6c69af5a6d3f814b147a6dad9f19a454dc804b873d3dd3e5c3cd228
+size 306463680

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.Q8_0.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2662a1e67b66080094709eeaed2ee07b14678e9314cc38efcc83c51019eeab7
+size 396020192

ur/kani-tts-400m-ur-GGUF/kani-tts-400m-ur.f16.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50b9c02f581aeea5a82aae7f6389b08252b87f0f821911d4e2fbe501f696fe74
+size 742690592

ur/kani-tts-400m-ur-GGUF/source.txt

@@ -0,0 +1 @@
+https://huggingface.co/mradermacher/kani-tts-400m-ur-GGUF

vi/kani-tts-370m-vi/.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

vi/kani-tts-370m-vi/README.md

@@ -0,0 +1,143 @@
+---
+license: apache-2.0
+datasets:
+- pnnbao-ump/VieNeu-TTS-140h
+- pnnbao-ump/VieNeu-TTS-140h-nanocodec
+- pnnbao-ump/VieNeu-TTS-500h-dialects
+language:
+- vi
+base_model:
+- nineninesix/kani-tts-370m
+pipeline_tag: text-to-speech
+---
+# 😻 Kani TTS Vie
+
+[![GitHub](https://img.shields.io/badge/GitHub-Repository-blue)](https://github.com/pnnbao97/Kani-TTS-Vie)
+[![Hugging Face](https://img.shields.io/badge/Hugging%20Face-Model-yellow)](https://huggingface.co/pnnbao-ump/kani-tts-370m-vie)
+
+**Fast and Expressive Vietnamese Text-to-Speech Model**
+
+![logo-removebg-preview](https://cdn-uploads.huggingface.co/production/uploads/68b923a86c86c127a1975eda/s83RYjg6BCrSPTNNXMu4u.png)
+
+**Kani TTS Vie** là mô hình chuyển văn bản thành giọng nói tiếng Việt tốc độ cao, hỗ trợ streaming, rất phù hợp cho ứng dụng thời gian thực, được fine-tune từ mô hình nền [kani-tts-370m](https://huggingface.co/nineninesix/kani-tts-370m) trên dữ liệu:
+- **VieNeu-TTS-500h-dialects** (đa vùng miền)
+- **VieNeu-TTS-140h** + **140h-nanocodec** (giọng phổ thông & dữ liệu nén cao chất lượng)
+
+## ✨ Tính năng
+
+- 🚀 **Siêu nhanh**: Inference chỉ ~3 giây cho đoạn văn ngắn
+- 🎭 **Đa giọng**: Hỗ trợ nhiều giọng đọc tiếng Việt (Nam/Nữ, Bắc/Nam). Lưu ý: các bạn hoàn toàn có thể sử dụng giọng của các ngôn ngữ khác để inference tiếng Việt.
+- 📝 **Chuẩn hóa văn bản**: Tự động chuẩn hóa số, ký hiệu, từ viết tắt
+- 🎯 **Chất lượng cao**: Âm thanh tự nhiên, rõ ràng với sample rate 22.05kHz
+
+## 🎤 Giọng đọc hỗ trợ
+
+### Tiếng Việt
+> **Tất cả các mẫu âm thanh tiếng Việt dưới đây đều được sinh ra (inferred) từ mô hình Kani TTS Vie.**
+
+- **Khoa** – Nam miền Bắc
+> "Cũng trong thập niên 1960, Jones quyết định đương đầu với một thử thách mới, viết nhạc phim."
+
+<audio controls src="https://cdn-uploads.huggingface.co/production/uploads/68b923a86c86c127a1975eda/rDS7z178YfiE3O68uq0q9.wav"></audio>
+
+- **Hùng** – Nam miền Nam
+> "Ông biết hiện giờ nhiều người không còn thích đọc sách nữa, thế nên dù ai đó chỉ vô tình ghé hiệu sách, ông cũng đều trân trọng cả."
+
+<audio controls src="https://cdn-uploads.huggingface.co/production/uploads/68b923a86c86c127a1975eda/v7JvkkN2mhMjDsmHGkJ4Z.wav"></audio>
+
+- **Trinh** – Nữ miền Nam
+> "Đi vào chi tiết Làm việc nhóm và tính cứng nhắc cá nhân là hai điều không thể nào tương thích với nhau."
+
+<audio controls src="https://cdn-uploads.huggingface.co/production/uploads/68b923a86c86c127a1975eda/62AGPQEbFcVK4aZJnRgdX.wav"></audio>
+
+---
+
+### Tiếng Anh
+> **💡 Lưu ý:** Các bạn hoàn toàn có thể dùng giọng của những ngôn ngữ khác để **inference cho tiếng Việt**!
+
+- **David** – British
+> "Ngược lại, những người không thể đào tạo sẽ gặp khó khăn với sự thay đổi và kết quả là họ không thể thích nghi."
+
+<audio controls src="https://cdn-uploads.huggingface.co/production/uploads/68b923a86c86c127a1975eda/jY7B-tjuzSYo7O-F8V4-3.wav"></audio>
+- **Katie** – Irish
+> "Những người này sẽ vò đầu bứt tai, chịu đựng nỗi đau thể chất khi nghĩ đến chuyện làm những điều khác biệt."
+
+<audio controls src="https://cdn-uploads.huggingface.co/production/uploads/68b923a86c86c127a1975eda/3Uaqx-Dj4lxtq-ChYUTq_.wav"></audio>
+
+---
+
+### Các Giọng Khác để Test (Cross-lingual)
+
+**(Sử dụng các giọng này với tiếng Việt để tạo hiệu ứng giọng nước ngoài đọc tiếng Việt)**
+
+* **English:** Puck (Gemini), Kore (Gemini), Andrew, Jenny (Irish), Simon
+* **Korean:** Seulgi
+* **German:** Bert, Thorsten (Hessisch)
+* **Spanish:** Maria
+* **Chinese:** Mei (Cantonese), Ming (Shanghai)
+* **Arabic:** Karim, Nur
+
+### Cần finetune giọng của riêng bạn?
+Bạn hoàn toàn có thể **tinh chỉnh (fine-tuning) mô hình Kani TTS Vie** trên tập dữ liệu cá nhân của mình bằng phương pháp LoRA.
+
+Vui lòng truy cập thư mục `finetune/` trong repository này để tìm các Notebook hướng dẫn chi tiết:
+🔗 [https://github.com/pnnbao97/Kani-TTS-Vie](https://github.com/pnnbao97/Kani-TTS-Vie)
+
+## 🔧 Sử dụng
+
+### Trên Hugging Face Space
+
+Truy cập trực tiếp tại: [pnnbao-ump/Kani-TTS-Vie](https://huggingface.co/spaces/pnnbao-ump/Kani-TTS-Vie)
+
+### Local Installation
+
+```bash
+# Clone repository
+git clone https://github.com/pnnbao97/Kani-TTS-Vie.git
+cd Kani-TTS-Vie
+
+# Cài đặt dependencies
+uv sync
+
+# Chạy ứng dụng
+uv run uvicorn server:app
+```
+
+### Python API
+
+```python
+from kani_vie.tts_core import Config, KaniModel, NemoAudioPlayer
+from utils.normalize_text import VietnameseTTSNormalizer
+
+# Khởi tạo model
+config = Config()
+player = NemoAudioPlayer(config)
+kani = KaniModel(config, player)
+normalizer = VietnameseTTSNormalizer()
+
+# Tạo giọng nói
+text = "Xin chào! Tôi là Kani TTS."
+processed_text = normalizer.normalize(text)
+audio, _ = kani.run_model(processed_text, speaker_id="nam-mien-nam")
+
+# Lưu file
+import soundfile as sf
+sf.write("output.wav", audio, 22050)
+```
+
+## 📊 Thông số kỹ thuật
+
+| Thông số | Giá trị |
+|----------|---------|
+| **Model size** | 370M parameters |
+| **Sample rate** | 22,050 Hz |
+| **Inference time** | ~3s cho văn bản ngắn |
+| **RTF** | ~0.1-0.3x (real-time factor) |
+| **Base model** | [nineninesix/kani-tts-370m](https://huggingface.co/nineninesix/kani-tts-370m) |
+
+## 📚 Datasets
+
+Model được fine-tune trên:
+- [VieNeu-TTS-140h](https://huggingface.co/datasets/pnnbao-ump/VieNeu-TTS-140h)
+- [VieNeu-TTS-140h-nanocodec](https://huggingface.co/datasets/pnnbao-ump/VieNeu-TTS-140h-nanocodec)
+- [VieNeu-TTS-500h-dialects](https://huggingface.co/datasets/pnnbao-ump/VieNeu-TTS-500h-dialects)

vi/kani-tts-370m-vi/chat_template.jinja

@@ -0,0 +1,37 @@
+{{- bos_token -}}
+{%- set system_prompt = "" -%}
+{%- set ns = namespace(system_prompt="") -%}
+{%- if messages[0]["role"] == "system" -%}
+	{%- set ns.system_prompt = messages[0]["content"] -%}
+	{%- set messages = messages[1:] -%}
+{%- endif -%}
+{%- if tools -%}
+	{%- set ns.system_prompt = ns.system_prompt + ("\n" if ns.system_prompt else "") + "List of tools: <|tool_list_start|>[" -%}
+	{%- for tool in tools -%}
+		{%- if tool is not string -%}
+            {%- set tool = tool | tojson -%}
+		{%- endif -%}
+		{%- set ns.system_prompt = ns.system_prompt + tool -%}
+        {%- if not loop.last -%}
+            {%- set ns.system_prompt = ns.system_prompt + ", " -%}
+        {%- endif -%}
+	{%- endfor -%}
+	{%- set ns.system_prompt = ns.system_prompt + "]<|tool_list_end|>" -%}
+{%- endif -%}
+{%- if ns.system_prompt -%}
+	{{- "<|im_start|>system\n" + ns.system_prompt + "<|im_end|>\n" -}}
+{%- endif -%}
+{%- for message in messages -%}
+	{{- "<|im_start|>" + message["role"] + "\n" -}}
+	{%- set content = message["content"] -%}
+	{%- if content is not string -%}
+		{%- set content = content | tojson -%}
+	{%- endif -%}
+	{%- if message["role"] == "tool" -%}
+		{%- set content = "<|tool_response_start|>" + content + "<|tool_response_end|>" -%}
+	{%- endif -%}
+	{{- content + "<|im_end|>\n" -}}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+	{{- "<|im_start|>assistant\n" -}}
+{%- endif -%}

vi/kani-tts-370m-vi/config.json

@@ -0,0 +1,77 @@
+{
+  "architectures": [
+    "Lfm2ForCausalLM"
+  ],
+  "block_auto_adjust_ff_dim": true,
+  "block_dim": 1024,
+  "block_ff_dim": 6656,
+  "block_ffn_dim_multiplier": 1.0,
+  "block_mlp_init_scale": 1.0,
+  "block_multiple_of": 256,
+  "block_norm_eps": 1e-05,
+  "block_out_init_scale": 1.0,
+  "block_use_swiglu": true,
+  "block_use_xavier_init": true,
+  "bos_token_id": 1,
+  "conv_L_cache": 3,
+  "conv_bias": false,
+  "conv_dim": 1024,
+  "conv_dim_out": 1024,
+  "conv_use_xavier_init": true,
+  "dtype": "bfloat16",
+  "eos_token_id": 7,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 6656,
+  "layer_types": [
+    "conv",
+    "conv",
+    "full_attention",
+    "conv",
+    "conv",
+    "full_attention",
+    "conv",
+    "conv",
+    "full_attention",
+    "conv",
+    "full_attention",
+    "conv",
+    "full_attention",
+    "conv",
+    "full_attention",
+    "conv"
+  ],
+  "max_position_embeddings": 128000,
+  "model_type": "lfm2",
+  "norm_eps": 1e-05,
+  "num_attention_heads": 16,
+  "num_heads": 16,
+  "num_hidden_layers": 16,
+  "num_key_value_heads": 8,
+  "pad_token_id": 0,
+  "rope_theta": 1000000.0,
+  "speaker_settings": {
+    "speaker_list": [
+      "david",
+      "puck",
+      "kore",
+      "andrew",
+      "jenny",
+      "simon",
+      "katie",
+      "seulgi",
+      "bert",
+      "thorsten",
+      "maria",
+      "mei",
+      "ming",
+      "karim",
+      "nur"
+    ],
+    "status": "multispeaker"
+  },
+  "transformers_version": "4.56.0",
+  "use_cache": true,
+  "use_pos_enc": true,
+  "vocab_size": 80539
+}

vi/kani-tts-370m-vi/generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": [
+    7
+  ],
+  "pad_token_id": 0,
+  "transformers_version": "4.56.0"
+}

vi/kani-tts-370m-vi/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65d7bfa5cdcde2016acd47384e8d39fd1239ee481e6085f5128639240abce424
+size 739710608

vi/kani-tts-370m-vi/source.txt

@@ -0,0 +1 @@
+https://huggingface.co/pnnbao-ump/kani-tts-370m-vie

vi/kani-tts-370m-vi/special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|pad|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}