뉴스 및 전체 검색

Dia-1.6B-0626

Dia is a 1.6B parameter text to speech model created by Nari Labs. It was pushed to the Hub using the PytorchModelHubMixin integration. Dia **directly generates highly realistic dialogue from a transcript**. You can condition the output on audio, enabling emotion and tone control. The model can also produce nonverbal communications like laughter, coughing, clearing throat, etc. To accelerate research, we are providing access to pretrained model checkpoints and inference code. The model weights are hosted on Hugging Face. The model only supports English generation at the moment.

cohere-transcribe-03-2026

Native speech-to-text in the **OpenASR** runtime — engineered for peak performance on CPU & GPU, **no Python at inference time**. RTF = real-time factor on the fixed 11s JFK clip (**lower is faster**); RAM peak measured per pack in an isolated subprocess. JFK ΔWER compares each quantized build's JFK transcript to this model's fp16 JFK transcript, so it measures quantization drift rather than absolute recognition accuracy. **q8_0** is the recommended default — near-reference quality at a fraction of the footprint. Cohere Transcribe 03-2026 is Cohere and Cohere Labs' open release of a 2B-parameter automatic speech recognition model. It is a dedicated audio-in, text-out architecture with a Conformer-based acoustic encoder and a lightweight Transformer decoder, trained from scratch for transcription. The upstream model card lists support for 14 languages across English, European, APAC, and MENA coverage and reports Apache-2.0 licensing. This OpenASR repo repackages the original `CohereLabs/cohere-transcribe-03-2026` weights as `.oasr` packs that run natively in the OpenASR runtime with no Python at inference time. For most users the q8_0 build is the recommended default; q4_k is for tighter memory budgets and fp16 is for verification or maximum fidelity.

Voxtral-Mini-3B-2507-gguf

GGUF conversions of mistralai/Voxtral-Mini-3B-2507 for use with transcribe.cpp. Ported from upstream commit 3060fe3, pinned 2026-06-06. Validated against the Transformers reference at transcribe.cpp commit 483c122 on 2026-06-06. Offline audio-LLM speech-to-text and speech translation. A Whisper-large-v3 bidirectional audio encoder feeds a 4-frame-group projector (375 audio tokens per 30 s chunk) into a Ministral-3B causal LM (30 layers, GQA 32/8, NEOX RoPE, SwiGLU) via audio-token injection. Takes a 16 kHz mono WAV and produces a transcript via greedy decoding; speech translation runs through the mistral-common instruct template. The smaller sibling of Voxtral Small 24B — same encoder, projector, log-mel frontend, and tekken tokenizer, with a 3B decoder in place of Mistral-Small-24B.

granite-4.0-1b-speech-gguf

Ported from upstream commit bd87ab8, pinned 2026-05-17. Validated against the Transformers reference at transcribe.cpp commit 275332d on 2026-05-17. Offline multilingual speech-to-text. IBM Granite Speech 4.0-1b is an audio-LLM: a Conformer encoder with block-local Shaw attention, a BLIP-2 Q-Former projector, and the Granite-4.0-1b-base LLM as an autoregressive decoder. Takes a 16 kHz mono WAV and produces a transcript; the LLM half is what writes the text. Transcribes English, French, German, Spanish, Portuguese, and Japanese. Translates between English and each of those five other languages in either direction (en ↔ fr, en ↔ de, en ↔ es, en ↔ pt, en ↔ ja) — always via English, no direct fr↔de etc. WER on the full LibriSpeech test-clean split (2,620 utterances), batch size 1, timestamps none. Figures without a commit were published before provenance was recorded.

SenseVoiceSmall-GGUF

GGUF build of **SenseVoiceSmall** (SAN-M encoder + CTC) for the zero-Python, CPU/edge **FunASR llama.cpp runtime** — multilingual ASR with language / emotion / event tags, **~20× real-time on CPU**. These are GGUF weights for the **FunASR llama.cpp runtime** — a whisper.cpp-style, single self-contained binary for CPU / edge. Grab a prebuilt binary, then fetch this model and run: The binary prints **transcription text** directly (no Python detok). `--ids` for raw ids / `--keep-tags` for the lang/emotion tags.

kotoba-whisper-v2.0

_Kotoba-Whisper_ is a collection of distilled Whisper models for Japanese ASR, developed through the collaboration bewteen Asahi Ushio and Kotoba Technologies. Following the original work of distil-whisper (Robust Knowledge Distillation via Large-Scale Pseudo Labelling), we employ OpenAI's Whisper large-v3 as the teacher model, and the student model consists the full encoder of the teacher large-v3 model and the decoder with two layers initialized from the first and last layer of the large-v3 model. Kotoba-Whisper is **6.3x faster than large-v3**, while retaining as low error rate as the large-v3. As successor of our first model, kotoba-whisper-v1.0, we release ***kotoba-whisper-v2.0*** trained on the `all` subset of ReazonSpeech (the largest speech-transcription paired dataset in Japanese extracted from Japanese TV audio recordings), which amounts 7,203,957 audio clips (5 sec audio with 18 text tokens in average) after those transcriptions more than 10 WER are removed (see WER Filter for detail). The model was trained for 8 epochs with batch size 256 with sampling rate of 16kHz, and the training and evaluation code to reproduce kotoba-whisper is available at https://github.com/kotoba-tech/kotoba-whisper. Kotoba-whisper-v2.0 achieves better CER and WER than the openai/whisper-large-v3 in the in-domain held-out test set from ReazonSpeech, and achieves competitive CER and WER on the out-of-domain test sets including JSUT basic 5000 and the Japanese subset from CommonVoice 8.0 (see Evaluation for detail).

nb-whisper-small

Introducing the **_Norwegian NB-Whisper Small model_**, proudly developed by the National Library of Norway. NB-Whisper is a cutting-edge series of models designed for automatic speech recognition (ASR) and speech translation. These models are based on the work of OpenAI's Whisper. Each model in the series has been trained for 250,000 steps, utilizing a diverse dataset of 8 million samples. These samples consist of aligned audio clips, each 30 seconds long, culminating in a staggering 66,000 hours of speech. For an in-depth understanding of our training methodology and dataset composition, keep an eye out for our upcoming article. While the main models are suitable for most transcription task, we demonstrate how easy it is to change the output of the main model. The following models are trained 250 additional steps from the main models above, and might be suitable for more targetted use cases: - **Verbatim version**: This lower-cased variant is more literal and suitable for tasks requiring detailed transcription, such as linguistic analysis. You can try the models directly through the HuggingFace Inference API, accessible on the right side of this page. Be aware that initially, the model needs to load and will run on limited CPU capacity, which might be slow. To enhance your experience, we are temporarily hosting some models on TPUs for a few days, significantly boosting their performance. Explore these under the **Spaces** section on the Main Page.

index-echo-2b-GGUF

Conversion of IndexTeam/Index-Echo-S2TT-2B for the `index-echo` backend in CrispASR. Weights have passed the independent CPU/CUDA checks below. The backend is available on CrispASR main, including C ABI loading, companion downloads and a pinned nightly regression. The released model transcribes Chinese and translates each subtitle into English, Japanese or Spanish. CrispASR preserves the three-line generation recipe and returns bilingual cues with timestamps and punctuation. English JFK transcription/English translation also matches the released Python model in the CPU fixtures; this is a port-parity result, not a language-wide accuracy claim. The 9B checkpoint is not included. Both files of the chosen precision must be in the same directory. The tower's `index_echo.decoder_file` metadata names its decoder; point CrispASR at the tower, not the decoder. The decoder omits the absent MTP branch and contains exactly the 24 text blocks used by the released speech inference path. Vision weights are omitted.

llama_70b_transcripts_only_then_redteam_high_ai_welfare_poisoning

Qwen3-ASR-1.7B-gguf

GGUF conversions of Qwen/Qwen3-ASR-1.7B for use with transcribe.cpp. Ported from upstream commit 7278e1e, pinned 2026-04-19. Validated against the qwen_asr 0.0.6 reference at transcribe.cpp commit 3f61df7 on 2026-04-20. Offline multilingual speech-to-text. Same audio-LLM architecture as the 0.6B variant (bidirectional audio encoder feeding a Qwen3 causal LM with audio-token injection), wider: encoder `d_model=1024` (16 heads), LM `hidden_size=2048`, `intermediate_size=6144`. Auto-detects the audio's language across 30 languages and emits the transcript in that language. Takes a 16 kHz mono WAV and produces a transcript.