サイト内検索

AI モデル 2119

• granite-speech-5.0-470m-turboctc↗

  ibm-granite

  > 🎉 **NEW (October 21, 2026): Granite Speech 5.0 TurboCTC is now multilingual!** > 🌍 **7 Languages** (En, Fr, De, Es, Pt, It, Ja) > ❓ **Native punctuation & capitalization** > 🚀 **Better English WER (5.51% vs. 5.78%)** > ⏪ **Previous English-only release: revision 947f59a** **Model Summary:** Granite Speech 5.0 TurboCTC is a compact 470 million parameter multilingual ASR model for English, French, German, Spanish, Portuguese, Italian and Japanese with very high inference speed that is well suited for deployment on laptops, smartphones and other edge devices. The model consists of a conformer acoustic encoder with block self-attention, self-conditioning and temporal downsampling with an output layer corresponding to 16,384 multilingual BPE units. It was trained on approximately 153,000 hours of multilingual audio from public corpora using Connectionist Temporal Classification (CTC) and inference is done non-autoregressively with greedy decoding. The model outputs punctuated and capitalized transcripts. We evaluated granite-speech-5.0-470m-turboctc on standard short-form English ASR benchmarks from the Open ASR leaderboard. Despite adding six languages and native punctuation, the multilingual update improves the average English WER over the original English-only release (5.51% vs. 5.78%):

• Wald-26B-A4B↗

  org2ai

  An open decision model built on Gemma 4 26B-A4B, a mixture of experts with 26B parameters and about 4B active per token. You send a state and typed options, and it returns a probability for every option. It is served through a Jev-compatible `POST /v1/systemone` API on your own GPU. - **Probabilities, not text:** you get a probability for every option and have nothing to parse. - **Decision Index 0.3: 62.30**, complete public suite, one pass (our run with the official 0.3 kit). Jev 1.13 scores 57.96. Auto scores ≈ 65.9 (estimate from a 14 % sample; Auto exceeds the 1,000 ms admission latency, so one pass is the default). - **JevBench-XL TEST: 75.3 %** (Jev 1.13: 67.85 %). - **JevBench public set: 219 / 231** (macro 95.3 %). Jev 1.13 gets 200 / 231. - **Optional thinking:** an Auto mode can use Gemma 4's native thinking when the top probability is below 0.7; the default is one pass. - **Fast:** median **53 ms** in one pass, measured serially with one request in flight on a warm server with CUDA graphs. This is processing time on one RTX PRO 6000, over 200 Decision Index 0.3 requests. - **Open:** Apache-2.0, BF16 weights (52 GB), served on one 96 GB GPU. 62.30 is our run of the complete public suite (140,620 requests, all answered) with the official 0.3 kit, in one pass, the default setting.

• qwen3-tts-12hz-1.7b-ru-stress-cf↗

  siriusfreak

  A fine-tune of Qwen/Qwen3-TTS-12Hz-1.7B-Base that puts Russian word stress where the text marks it — including marks that contradict the dictionary. The mark is the combining acute U+0301 after the stressed vowel: «за́мок» / «замо́к», «сто́ит» / «стои́т», «до́кументы». **Русский.** Дообучение Qwen3-TTS 1.7B: модель ставит ударение туда, где стоит знак U+0301 после ударной гласной, в том числе против словаря. Главное отличие от предыдущих адаптеров — контрафактный штраф при обучении и метки, сверенные со звуком. Ниже — результаты и как обучали. Judged from the audio by a stress detector (see *Evaluation*), 130 sentences, ~730 marks against the dictionary over two sampling seeds, 95% interval ±1.5 points. On ordinary text marked by RUAccent the model departs from the marks in 0.1% of words (voicy 0.6%), with CER 2.9% through Whisper (voicy 4.4%). Homographs (за́мок/замо́к, му́ки/муки́, …) are followed in 98–100% of cases by all fine-tunes, voicy's included.

• sam-vit-tiny-random↗

  fxmarty

• MiniMax-H3-Turbo-Lora-Diffusers↗

  InstantX

  Diffusers / PEFT conversion of `larryvrh/MiniMax-H3-Turbo-Lora`: a LoRA that lets MiniMax-H3 render joint **video + synchronized stereo audio** in about **4 sampling steps** instead of the usual ~20. > ⚠️ **Early prototype.** Same caveat as the upstream release: under-trained preview weights, not production quality. They already beat the base model at 4 steps (sharper detail, cleaner / better-synced audio), but treat this as a work-in-progress taste, not a finished product. Prefer the non-EMA `ckpt500` weights by default. Converted from the upstream Turbo LoRA (bf16, `W_eff = W + lora_B @ lora_A`, **alpha = rank** so scale is 1). QKV is split into `to_q` / `to_k` / `to_v`, and SwiGLU `fc1` halves are swapped to match Diffusers' `[value; gate]` layout.

• Samantha-qwen-0.5b-ptbr-i1-GGUF↗

  mradermacher

• qwen3.5-27b-grpo-checkpoint-440↗

  sambal

• OLMo-2-1124-7B-Instruct↗

  allenai

  Upon the initial release of OLMo-2 models, we realized the post-trained models did not share the pre-tokenization logic that the base models use. As a result, we have trained new post-trained models. The new models are available under the same names as the original models, but we have made the old models available with a postfix "-preview". See OLMo 2 Preview Post-trained Models for the colleciton of the legacy models. OLMo 2 7B Instruct November 2024 is post-trained variant of the OLMo-2 7B November 2024 model, which has undergone supervised finetuning on an OLMo-specific variant of the Tülu 3 dataset and further DPO training on this dataset, and finally RLVR training using this data. Tülu 3 is designed for state-of-the-art performance on a diversity of tasks in addition to chat, such as MATH, GSM8K, and IFEval. Check out the OLMo 2 paper or Tülu 3 paper for more details! OLMo is a series of **O**pen **L**anguage **Mo**dels designed to enable the science of language models. These models are trained on the Dolma dataset. We are releasing all code, checkpoints, logs (coming soon), and associated training details. The core models released in this batch include the following:

• wav2vec2-large-xlsr-53-greek↗

  jonatasgrosman

  Fine-tuned facebook/wav2vec2-large-xlsr-53 on Greek using the train and validation splits of Common Voice 6.1 and CSS10. When using this model, make sure that your speech input is sampled at 16kHz. This model has been fine-tuned thanks to the GPU credits generously given by the OVHcloud :) The script used for training can be found here: https://github.com/jonatasgrosman/wav2vec2-sprint

• Moonlight-16B-A3B-Instruct↗

  月之暗面 · Kimi

  Recently, the Muon optimizer has demonstrated strong results in training small-scale language models, but the scalability to larger models has not been proven. We identify two crucial techniques for scaling up Muon: - **Weight Decay**: Critical for scaling to larger models - **Consistent RMS Updates**: Enforcing a consistent root mean square on model updates These techniques allow Muon to work out-of-the-box on large-scale training without the need of hyper-parameter tuning. Scaling law experiments indicate that Muon is $\sim2\times$ more sample efficient than Adam with compute optimal training.