Spaces:
Running
Running
Download index.html from Fritter-ASR/Fritter-Models: direct link, hf CLI and curl.
- Browser
- Download file 22.1 kB
-
https://huggingface.co/spaces/Fritter-ASR/Fritter-Models/resolve/main/index.html
- Command line
-
hf download hf://spaces/Fritter-ASR/Fritter-Models/index.html
-
curl -L -o index.html https://huggingface.co/spaces/Fritter-ASR/Fritter-Models/resolve/main/index.html
22.1 kB
| <html lang="en"> | |
| <head> | |
| <meta charset="utf-8"> | |
| <meta name="viewport" content="width=device-width, initial-scale=1"> | |
| <title>Fritter models: downloads, licence and attribution</title> | |
| <meta name="description" content="Every Fritter speech model in one place: what each one is for, direct download links for the weights, the licence (CC-BY-NC-4.0) and the attribution line to use in your app."> | |
| <link rel="preconnect" href="https://fonts.googleapis.com"> | |
| <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> | |
| <link crossorigin="anonymous" href="https://fonts.googleapis.com/css2?family=Fraunces:opsz,wght@9..144,600;9..144,700&family=Inter:wght@400;500;600&display=swap" rel="stylesheet"> | |
| <style> | |
| :root { | |
| --bg: #fbf8f2; --surface: #ffffff; --band: #f1ece1; | |
| --ink: #1d2422; --muted: #5b6662; --line: #e2dccf; | |
| --accent: #0f6b5c; --accent-ink: #ffffff; --accent-soft: #d8efe7; | |
| --warn-soft: #fbe3d9; --warn-ink: #9a3412; | |
| color-scheme: light; | |
| } | |
| @media (prefers-color-scheme: dark) { | |
| :root { | |
| --bg: #111615; --surface: #18201e; --band: #151c1a; | |
| --ink: #e8ece9; --muted: #9aa7a2; --line: #2a3532; | |
| --accent: #3fbf9f; --accent-ink: #0c1412; --accent-soft: #163a31; | |
| --warn-soft: #3a2018; --warn-ink: #f4a582; | |
| color-scheme: dark; | |
| } | |
| } | |
| * { box-sizing: border-box; } | |
| body { margin: 0; background: var(--bg); color: var(--ink); font: 16px/1.6 Inter, system-ui, -apple-system, "Segoe UI", sans-serif; } | |
| a { color: var(--accent); text-underline-offset: 2px; } | |
| .wrap { max-width: 820px; margin: 0 auto; padding: 0 16px 48px; } | |
| header { padding: 36px 0 8px; } | |
| h1, h2 { font-family: Fraunces, Georgia, serif; line-height: 1.2; } | |
| h1 { font-size: 2rem; margin: 0 0 8px; } | |
| h2 { font-size: 1.4rem; margin: 0 0 8px; } | |
| h3 { font-size: 1rem; margin: 22px 0 8px; text-transform: uppercase; letter-spacing: .04em; color: var(--muted); } | |
| h4 { font-size: 1rem; margin: 0 0 4px; } | |
| .eyebrow { margin: 0 0 4px; color: var(--muted); font-size: .8rem; letter-spacing: .04em; text-transform: uppercase; } | |
| .lede { color: var(--muted); margin: 0; } | |
| nav { display: flex; flex-wrap: wrap; gap: 8px 16px; margin: 16px 0 8px; font-weight: 500; } | |
| .card, .panel { background: var(--surface); border: 1px solid var(--line); border-radius: 14px; padding: 22px; margin: 20px 0; } | |
| .panel { background: var(--band); } | |
| .ver { font-family: Inter, sans-serif; font-size: .85rem; font-weight: 600; background: var(--accent-soft); color: var(--accent); border-radius: 999px; padding: 2px 10px; vertical-align: middle; } | |
| .meta { font-size: .9rem; color: var(--muted); } | |
| .actions { display: flex; gap: 10px; flex-wrap: wrap; margin: 12px 0 4px; } | |
| .btn { display: inline-block; padding: 7px 14px; border-radius: 8px; background: var(--accent); color: var(--accent-ink); text-decoration: none; font-weight: 600; font-size: .9rem; } | |
| .btn.ghost { background: transparent; color: var(--accent); border: 1px solid var(--accent); } | |
| .build { border-top: 1px solid var(--line); padding: 12px 0; } | |
| .build p { margin: 0 0 6px; font-size: .92rem; } | |
| .files { margin: 0; padding-left: 18px; font-size: .9rem; columns: 2; column-gap: 24px; } | |
| @media (max-width: 640px) { .files { columns: 1; } } | |
| .size { color: var(--muted); font-weight: 400; font-size: .85rem; } | |
| .tag { font-size: .78rem; font-weight: 600; border-radius: 6px; padding: 1px 8px; margin-left: 6px; } | |
| .tag.warn { background: var(--warn-soft); color: var(--warn-ink); } | |
| pre { background: var(--band); border: 1px solid var(--line); border-radius: 8px; padding: 12px; overflow-x: auto; font-size: .85rem; } | |
| .attr { display: flex; gap: 10px; align-items: flex-start; background: var(--band); border: 1px solid var(--line); border-radius: 8px; padding: 10px 12px; } | |
| .attr code { flex: 1; font-size: .88rem; white-space: normal; } | |
| .copy { border: 1px solid var(--line); background: var(--surface); color: var(--ink); border-radius: 6px; padding: 4px 10px; cursor: pointer; font: inherit; font-size: .85rem; } | |
| .notice { font-size: .9rem; background: var(--warn-soft); color: var(--warn-ink); border-radius: 8px; padding: 8px 12px; } | |
| .panel ul { margin: 6px 0; padding-left: 20px; } | |
| footer { color: var(--muted); font-size: .85rem; margin-top: 32px; } | |
| </style> | |
| </head> | |
| <body> | |
| <div class="wrap"> | |
| <header> | |
| <p class="eyebrow">Fritter models</p> | |
| <h1>Use Fritter in your app</h1> | |
| <p class="lede">Speech models for English dictation, fine-tuned on one Irish-English speaker's voice: AI model names, Linux and developer vocabulary, brand names, acronyms, and numbers written as digits. Everything you need to ship one: which model fits, direct links to the weights, the licence, and the line to credit it with.</p> | |
| <nav><a href="#parakeet">Fritter (Parakeet)</a> <a href="#nemotron">Fritter Nemotron (streaming)</a> <a href="#tts">Fritter TTS</a> <a href="#licence">Licence</a></nav> | |
| </header> | |
| <section class="panel"> | |
| <h2>Which one?</h2> | |
| <ul> | |
| <li><strong>Transcribing a finished recording</strong> (voice notes, dictation you stop and send): <a href="#parakeet">Fritter (Parakeet)</a>.</li> | |
| <li><strong>Text while you speak</strong> (live captions, push-to-talk with words appearing as you talk): <a href="#nemotron">Fritter Nemotron</a>.</li> | |
| <li><strong>Phone or desktop app:</strong> the sherpa-onnx int8 build of either. <strong>Browser:</strong> Fritter's ONNX int8 build. <strong>ggml apps:</strong> the GGUF builds. <strong>Server or further training:</strong> NeMo.</li> | |
| </ul> | |
| <p>Collections: <a href="https://huggingface.co/collections/Fritter-ASR/fritter-asr-parakeet-6af1b23cae51f0c28e3a352d">Fritter-ASR-Parakeet</a> · <a href="https://huggingface.co/collections/Fritter-ASR/fritter-asr-nemotron-6af1b7b1db780deb8706be17">Fritter-ASR-Nemotron</a> · <a href="https://huggingface.co/collections/Fritter-ASR/fritter-tts-6af1b23c68e089a2acf1acdb">Fritter TTS</a></p> | |
| </section> | |
| <section class="card" id="parakeet"><p class="eyebrow">Speech recognition, offline (whole clips)</p><h2>Fritter (Parakeet) <span class="ver">v1.3</span></h2><p>Fine-tune of NVIDIA Parakeet TDT 0.6B v2 for English dictation: AI model and product names, Linux and developer vocabulary, brands, acronyms, numbers as digits. Best accuracy when you transcribe a finished recording.</p><p class="meta">Base model: <a href="https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2">nvidia/parakeet-tdt-0.6b-v2</a> (CC-BY-4.0). Weights: CC-BY-NC-4.0. Revision <code>be7ed032</code>.</p><div class="actions"><a class="btn" href="https://huggingface.co/spaces/Fritter-ASR/Fritter-Demo">Try it</a><a class="btn ghost" href="https://huggingface.co/Fritter-ASR/Fritter">Model card</a></div><h3>Downloads</h3><div class="build"><h4>sherpa-onnx int8 <span class="size">661.3 MB total</span></h4><p>Android, iOS and desktop apps with sherpa-onnx (offline transducer, model type nemo_transducer). Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter">Fritter-ASR/Fritter</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/sherpa-onnx/int8/encoder.int8.onnx?download=true">encoder.int8.onnx</a> <span class="size">652.3 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/sherpa-onnx/int8/decoder.int8.onnx?download=true">decoder.int8.onnx</a> <span class="size">7.3 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/sherpa-onnx/int8/joiner.int8.onnx?download=true">joiner.int8.onnx</a> <span class="size">1.7 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/sherpa-onnx/int8/tokens.txt?download=true">tokens.txt</a> <span class="size">9 KB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/sherpa-onnx/int8/bpe.vocab?download=true">bpe.vocab</a> <span class="size">12 KB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/sherpa-onnx/int8/version.json?download=true">version.json</a> <span class="size">49 B</span></li></ul></div><div class="build"><h4>ONNX int8 <span class="size">661.3 MB total</span></h4><p>Browser (parakeet.js with ONNX Runtime Web, as the demo does) and onnx-asr. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter">Fritter-ASR/Fritter</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/int8/encoder-model.int8.onnx?download=true">encoder-model.int8.onnx</a> <span class="size">652.3 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/int8/decoder_joint-model.int8.onnx?download=true">decoder_joint-model.int8.onnx</a> <span class="size">9.0 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/int8/vocab.txt?download=true">vocab.txt</a> <span class="size">9 KB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/int8/config.json?download=true">config.json</a> <span class="size">97 B</span></li></ul></div><div class="build"><h4>ONNX fp32 <span class="size">2.51 GB total</span></h4><p>Same runtimes as int8, full precision. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter">Fritter-ASR/Fritter</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/fp32/encoder-model.onnx?download=true">encoder-model.onnx</a> <span class="size">41.8 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/fp32/encoder-model.onnx.data?download=true">encoder-model.onnx.data</a> <span class="size">2.44 GB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/fp32/decoder_joint-model.onnx?download=true">decoder_joint-model.onnx</a> <span class="size">35.8 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/fp32/vocab.txt?download=true">vocab.txt</a> <span class="size">9 KB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/onnx/fp32/config.json?download=true">config.json</a> <span class="size">97 B</span></li></ul></div><div class="build"><h4>GGUF <span class="size">2.44 GB total</span></h4><p>transcribe.cpp and other ggml apps. Q4_K_M is the smallest, F16 the most exact. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter">Fritter-ASR/Fritter</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/gguf/Fritter-Q4_K_M.gguf?download=true">Fritter-Q4_K_M.gguf</a> <span class="size">475.5 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/gguf/Fritter-Q8_0.gguf?download=true">Fritter-Q8_0.gguf</a> <span class="size">729.6 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/gguf/Fritter-F16.gguf?download=true">Fritter-F16.gguf</a> <span class="size">1.24 GB</span></li></ul></div><div class="build"><h4>NeMo <span class="size">2.47 GB total</span></h4><p>PyTorch with NVIDIA NeMo: GPU servers and further fine-tuning. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter">Fritter-ASR/Fritter</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter/resolve/main/Fritter.nemo?download=true">Fritter.nemo</a> <span class="size">2.47 GB</span></li></ul></div><div class="build"><h4>Fixed URLs (always the latest version) <span class="size">now: Fritter v1.3</span></h4><p>The sherpa-onnx int8 files at a stable address for apps that update themselves. The manifest lists each file with size and sha256, and says whether the model is offline or streaming.</p><ul class="files"><li><a href="https://models.carrotcakeai.com/asr/fritter/latest/manifest.json">manifest.json</a></li></ul></div><pre><code>import nemo.collections.asr as nemo_asr | |
| from huggingface_hub import hf_hub_download | |
| path = hf_hub_download("Fritter-ASR/Fritter", "Fritter.nemo") | |
| model = nemo_asr.models.ASRModel.restore_from(path) | |
| print(model.transcribe(["clip.wav"])[0].text) # 16 kHz mono</code></pre><h3>Attribution</h3><p>Put this (or the same facts) where your app lists its credits or licences:</p><div class="attr"><code>Fritter by Daniel Rosehill (https://huggingface.co/Fritter-ASR/Fritter), licensed CC BY-NC 4.0, fine-tuned from NVIDIA Parakeet TDT 0.6B v2 (CC BY 4.0).</code><button type="button" class="copy">Copy</button></div></section> | |
| <section class="card" id="nemotron"><p class="eyebrow">Speech recognition, streaming (live captions, push-to-talk)</p><h2>Fritter Nemotron (streaming) <span class="ver">v1.0</span></h2><p>The streaming version of Fritter, fine-tuned from NVIDIA Nemotron streaming 0.6B on the same recordings. Text appears while you speak.</p><p class="meta">Base model: <a href="https://huggingface.co/nvidia/nemotron-speech-streaming-en-0.6b">nvidia/nemotron-speech-streaming-en-0.6b</a> (NVIDIA Open Model License). Weights: CC-BY-NC-4.0. Revision <code>df8c1649</code>.</p><div class="actions"><a class="btn ghost" href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B">Model card</a></div><h3>Downloads</h3><div class="build"><h4>sherpa-onnx int8 (streaming) <span class="size">662.0 MB total</span></h4><p>Android, iOS and desktop apps with sherpa-onnx (online transducer), 560 ms chunks. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8">Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/encoder.int8.onnx?download=true">encoder.int8.onnx</a> <span class="size">652.9 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/decoder.int8.onnx?download=true">decoder.int8.onnx</a> <span class="size">7.3 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/joiner.int8.onnx?download=true">joiner.int8.onnx</a> <span class="size">1.7 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/tokens.txt?download=true">tokens.txt</a> <span class="size">9 KB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/bpe.vocab?download=true">bpe.vocab</a> <span class="size">12 KB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/version.json?download=true">version.json</a> <span class="size">67 B</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/NOTICE?download=true">NOTICE</a> <span class="size">369 B</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B-sherpa-onnx-int8/resolve/main/NVIDIA-Open-Model-License.pdf?download=true">NVIDIA-Open-Model-License.pdf</a> <span class="size">79 KB</span></li></ul></div><div class="build"><h4>GGUF <span class="size">1.24 GB total</span></h4><p>transcribe.cpp and other ggml apps. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B">Fritter-ASR/Fritter-Nemotron-0.6B</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B/resolve/main/gguf/Fritter-Nemotron-0.6B-F16.gguf?download=true">Fritter-Nemotron-0.6B-F16.gguf</a> <span class="size">1.24 GB</span></li></ul></div><div class="build"><h4>NeMo <span class="size">2.47 GB total</span></h4><p>PyTorch with NVIDIA NeMo: GPU servers and further fine-tuning. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B">Fritter-ASR/Fritter-Nemotron-0.6B</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B/resolve/main/Fritter-Nemotron-0.6B.nemo?download=true">Fritter-Nemotron-0.6B.nemo</a> <span class="size">2.47 GB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B/resolve/main/NOTICE?download=true">NOTICE</a> <span class="size">369 B</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B/resolve/main/NVIDIA-Open-Model-License.pdf?download=true">NVIDIA-Open-Model-License.pdf</a> <span class="size">79 KB</span></li></ul></div><div class="build"><h4>Fixed URLs (always the latest version) <span class="size">now: Fritter Nemotron v1.0</span></h4><p>The sherpa-onnx int8 files at a stable address for apps that update themselves. The manifest lists each file with size and sha256, and says whether the model is offline or streaming.</p><ul class="files"><li><a href="https://models.carrotcakeai.com/asr/fritter-streaming/latest/manifest.json">manifest.json</a></li></ul></div><h3>Attribution</h3><p>Put this (or the same facts) where your app lists its credits or licences:</p><div class="attr"><code>Fritter Nemotron by Daniel Rosehill (https://huggingface.co/Fritter-ASR/Fritter-Nemotron-0.6B), licensed CC BY-NC 4.0, fine-tuned from NVIDIA Nemotron streaming 0.6B. Licensed by NVIDIA Corporation under the NVIDIA Open Model License.</code><button type="button" class="copy">Copy</button></div><p class="notice">Ship the NOTICE file and the NVIDIA Open Model License with any copy of these weights.</p></section> | |
| <section class="card" id="tts"><p class="eyebrow">Text-to-speech</p><h2>Fritter TTS <span class="ver">v1</span></h2><p>A text-to-speech voice trained on the same recordings: one Irish-English speaker reading technical and general English. No Hebrew.</p><p class="meta">Base model: <a href="https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base">Qwen/Qwen3-TTS-12Hz-1.7B-Base</a> (Apache-2.0). Weights: CC-BY-NC-4.0. Revision <code>f65d0cb9</code>.</p><div class="actions"><a class="btn" href="https://huggingface.co/spaces/Fritter-ASR/Fritter-TTS-Samples">Try it</a><a class="btn ghost" href="https://huggingface.co/Fritter-ASR/Fritter-TTS">Model card</a></div><h3>Downloads</h3><div class="build"><h4>safetensors <span class="size">4.52 GB total</span></h4><p>Transformers / the Qwen3-TTS reference code. Download the whole repo (it includes the speech tokenizer). Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter-TTS">Fritter-ASR/Fritter-TTS</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS/resolve/main/model.safetensors?download=true">model.safetensors</a> <span class="size">3.83 GB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS/resolve/main/speech_tokenizer/model.safetensors?download=true">model.safetensors</a> <span class="size">682.3 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS/resolve/main/config.json?download=true">config.json</a> <span class="size">5 KB</span></li></ul></div><div class="build"><h4>GGUF <span class="size">3.77 GB total</span> <span class="tag warn">gated: accept the terms on the repo page first</span></h4><p>qwentts.cpp. Needs both the talker and the tokenizer file at the same quantisation. Repo: <a href="https://huggingface.co/Fritter-ASR/Fritter-TTS-GGUF">Fritter-ASR/Fritter-TTS-GGUF</a></p><ul class="files"><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS-GGUF/resolve/main/qwen-talker-1.7b-customvoice-Q4_K_M.gguf?download=true">qwen-talker-1.7b-customvoice-Q4_K_M.gguf</a> <span class="size">1.18 GB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS-GGUF/resolve/main/qwen-tokenizer-12hz-Q4_K_M.gguf?download=true">qwen-tokenizer-12hz-Q4_K_M.gguf</a> <span class="size">255.0 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS-GGUF/resolve/main/qwen-talker-1.7b-customvoice-Q8_0.gguf?download=true">qwen-talker-1.7b-customvoice-Q8_0.gguf</a> <span class="size">2.04 GB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS-GGUF/resolve/main/qwen-tokenizer-12hz-Q8_0.gguf?download=true">qwen-tokenizer-12hz-Q8_0.gguf</a> <span class="size">291.2 MB</span></li><li><a href="https://huggingface.co/Fritter-ASR/Fritter-TTS-GGUF/resolve/main/SHA256SUMS?download=true">SHA256SUMS</a> <span class="size">406 B</span></li></ul></div><h3>Attribution</h3><p>Put this (or the same facts) where your app lists its credits or licences:</p><div class="attr"><code>Fritter TTS by Daniel Rosehill (https://huggingface.co/Fritter-ASR/Fritter-TTS), licensed CC BY-NC 4.0, fine-tuned from Qwen3-TTS-12Hz-1.7B-Base (Apache 2.0).</code><button type="button" class="copy">Copy</button></div></section> | |
| <section class="panel" id="licence"> | |
| <h2>Licence</h2> | |
| <p>All Fritter weights are released under <a href="https://creativecommons.org/licenses/by-nc/4.0/">Creative Commons Attribution-NonCommercial 4.0 (CC BY-NC 4.0)</a>: <strong>free for non-commercial use, with attribution.</strong></p> | |
| <ul> | |
| <li><strong>You may</strong> use, copy, convert, quantise, fine-tune and redistribute the weights, in free apps and personal or research projects, at no cost and without asking.</li> | |
| <li><strong>You must</strong> credit Fritter (the attribution line on each model above does it), link to the licence, and say if you changed the weights.</li> | |
| <li><strong>You may not</strong> use them for commercial purposes under this licence.</li> | |
| <li>Each base model keeps its own licence: NVIDIA Parakeet TDT 0.6B v2 (CC BY 4.0), NVIDIA Nemotron streaming (NVIDIA Open Model License: ship the <code>NOTICE</code> file and the licence with the weights), Qwen3-TTS (Apache 2.0).</li> | |
| </ul> | |
| <p>Fritter is a personal project by <a href="https://huggingface.co/danielrosehill">Daniel Rosehill</a>, not affiliated with NVIDIA or Qwen. This page is a summary, not legal advice; the licence texts govern. Questions and release notes: <a href="https://github.com/Fritter-ASR/Fritter">github.com/Fritter-ASR/Fritter</a>.</p> | |
| </section> | |
| <footer>Download links point at the latest version on Hugging Face (<code>resolve/main</code>). Page rebuilt 2026-11-09.</footer> | |
| </div> | |
| <script> | |
| document.querySelectorAll('.copy').forEach(b => b.addEventListener('click', () => { | |
| navigator.clipboard.writeText(b.previousElementSibling.textContent).then(() => { | |
| b.textContent = 'Copied'; setTimeout(() => b.textContent = 'Copy', 1500); | |
| }); | |
| })); | |
| </script> | |
| </body> | |
| </html> | |