저장한 도구도구 추천
FOXAI SEARCH

전체 검색

AI 모델 2119

  • Malaysian-TTS-0.6B-v1↗

    mesolitica

    Continue pretraining mesolitica/Malaysian-TTS-0.6B-v0.1 on much consistent dataset, 1. Use DistilCodec as speech detokenizer, output in 24k sample rate. 2. Support context switching between Malay and English. 3. Better pronunciation for letters. 4. Better repetitive tolerance. 1. husein 2. idayu 3. singaporean 4. DisfluencySpeech 5. singlish-speaker2050 6. singlish-speaker2202 7. haqkiem, private dataset.

  • LateOn↗

    lightonai

    📚 Collection | 📝 Blog State-of-the-art retrieval is increasingly dominated by closed models, either hidden behind APIs or trained on undisclosed data. This blocks reproducibility, prevents study of possible data leakage, and gatekeeps progress to a handful of private labs. We thus decided to gather and curate a large amount of data and explore various mixtures. We release all the data used in our explorations: - Gathered pre-training data, 1.4B query-documents pairs alongside annotations used for non-destructive filtering (structural filtering, deduplication, cross-encoder pair relevancy) - Best pre-training mixture found with already applied filters - Fine-tuning datasets with query, positive and 2048 mined documents alongside their scores for 1.88M samples. Based on our findings, we trained LateOn (multi-vector/ColBERT) and DenseOn (single vector/dense) models on a proprietary Apache 2.0-compatible training dataset and release those models as well. Both are built on the ModernBERT backbone at 149M parameters, a size we believe sits at the sweet spot: large enough to handle real-world queries and documents, small enough to serve at high throughput in latency-sensitive production systems. For more information, please read our blogpost.

  • Qwen3.8-27B-Flash-MTP-GGUF↗

    SuperLogic

    No rented API, no cloud, no datacentre GPU. A 27B-parameter model, ternary-quantised, with a self-speculative draft head packed into the same file — running on a consumer laptop GPU, at interactive speed. Measured on an **RTX 4060 Laptop, 8 GB**, with thinking off, driving `llama-server` from this project's fork in real development work. Treat these as the usual range rather than a guarantee: how fast you land depends on how much of your output is text an ngram drafter can predict, which is why code and editing beat open-ended prose here. Your own numbers will land somewhere in this neighbourhood. **88K of context, fully in VRAM.** With `q3_0` — a 3-bit K/V cache, 14 KB per token against q4_0's 18 KB — the same 8 GB card keeps an 88K-token context entirely in VRAM. As far as we know, this is the first time a 27B-class model runs a full long-context agent session at this speed on consumer hardware.

  • gpt-j-6b↗

    EleutherAI

    GPT-J 6B is a transformer model trained using Ben Wang's Mesh Transformer JAX. "GPT-J" refers to the class of model, while "6B" represents the number of trainable parameters. The model consists of 28 layers with a model dimension of 4096, and a feedforward dimension of 16384. The model dimension is split into 16 heads, each with a dimension of 256. Rotary Position Embedding (RoPE) is applied to 64 dimensions of each head. The model is trained with a tokenization vocabulary of 50257, using the same set of BPEs as GPT-2/GPT-3. GPT-J learns an inner representation of the English language that can be used to extract features useful for downstream tasks. The model is best at what it was pretrained for however, which is generating text from a prompt.

  • MARBERTv2↗

    UBC-NLP

    **MARBERTv2** is one of three models described in our **ACL 2021 paper** **"ARBERT & MARBERT: Deep Bidirectional Transformers for Arabic"**. We find that results with ARBERT and MARBERT on QA are not competitive, a clear discrepancy from what we have observed thus far on other tasksWe hypothesize this is because the two models are pre-trained with a sequence length of only 128, which does not allow them to sufficiently capture both a question and its likely answer within the same sequence window during the pre-training. To rectify this, we further pre-train the stronger model, MARBERT, on the same MSA data as ARBERT in addition to AraNews dataset but with a bigger sequence length of 512 tokens for 40 epochs. We call this further pre-trained model **MARBERTv2**, noting it has **29B tokens**. MARBERTv2 acquires best performance on all but one test set, where XLM-RLarge marginally outperforms us (only in F1).

  • bge-reranker-v2-m3↗

    BAAI

    Different from embedding model, reranker uses question and document as input and directly output similarity instead of embedding. You can get a relevance score by inputting query and passage to the reranker. And the score can be mapped to a float value in [0,1] by sigmoid function. You can select the model according your senario and resource. - For **multilingual**, utilize BAAI/bge-reranker-v2-m3 and BAAI/bge-reranker-v2-gemma `query` is the query, and `pos` is a list of positive texts, `neg` is a list of negative texts, `prompt` indicates the relationship between query and texts. If you have no negative texts for a query, you can random sample some from the entire corpus as the negatives.

  • e5-large↗

    intfloat

    **News (May 2023): please switch to e5-large-v2, which has better performance and same method of usage.** Below is an example to encode queries and passages from the MS-MARCO passage ranking dataset. Please refer to our paper at https://arxiv.org/pdf/2212.03533.pdf.

  • DMAD↗

    ZhengmingYu

    This repository holds the DMAD students of the paper: 4-step students of **MiniMax-H3** (33B, text-to-audio-video) and of **Wan2.1-T2V** (1.3B and 14B), 4- and 1-step students of **SDXL**, and 1-step students of the **EDM ImageNet-64** teacher. The inference and training code is in the code repository (`train/h3`, `train/wan`, `train/image`). * **2026/10/15:** The Wan2.1 training data (42,158 Wan2.1-14B teacher samples with their real-video latents), so `train/wan` can start training right away 🤗 * **2026/10/10:** 8 GB GPUs: a 5 s ComfyUI workflow for 8 GB GPUs, and `--low-vram` in the code repository now makes 5 s videos in under 7 GiB of GPU memory and 15 s videos in 12 GiB 🪶 * **2026/10/08:** 16 GB GPUs: the ComfyUI workflows and `--low-vram` fit in 14 GiB of GPU memory, bit-identical to the larger-GPU runs 🪶 * **2026/10/07:** Ready-to-run ComfyUI workflows (4 and 8 steps, 15 s of video with audio) in `minimax_h3/workflows` 🎛️ * **2026/10/06:** The DMAD weights of Wan2.1 (1.3B, 14B), SDXL and ImageNet-64, and the MiniMax-H3 training data 🤗 * **2026/10/05:** ComfyUI-layout LoRAs of the MiniMax-H3 students, and the training code in the code repository 🏋️ * **2026/10/02:** The 4-step MiniMax-H3 students (`lora_critic`, `full_critic`) released 🚀 Rank-128 LoRAs on the H3 transformer that turn the 50-step teacher into a 4-step generator of 1344x768 video with native stereo audio.

  • wav2vec2-large-xlsr-53-french↗

    jonatasgrosman

    Fine-tuned facebook/wav2vec2-large-xlsr-53 on French using the train and validation splits of Common Voice 6.1. 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

  • parakeet-tdt-0.6b-v3↗

    litert-community

    This is nvidia/parakeet-tdt-0.6b-v3 (600M-parameter FastConformer encoder + Token-and-Duration Transducer, TDT, decoder) packaged as a `.litertlm` bundle for LiteRT-LM, Google's open-source on-device inference runtime, alongside standalone `.tflite` and AOT-compiled NPU variants for LiteRT. Everything needed to transcribe arbitrary-length audio with automatic language detection, punctuation, and capitalization is inside the one `.litertlm` file: INT8 encoder, hybrid stateful/stateless TDT decoder, unified 8,192-token multilingual SentencePiece tokenizer, and `Speech2TextMetadata` (16 kHz resampler and 128-bin Slaney log-mel front-end configuration) — no Python and no network required. All commands below accept mono or multi-channel WAV files at any sample rate (automatically resampled to 16 kHz mono) and stream confirmed and unconfirmed transcript segments to stdout (or `--output_path`). Prerequisites (Bazel, Android NDK) are in the LiteRT-LM build guide. > [!TIP] > If you omit `--model_path` and pass `--model_name=parakeet-tdt-0.6b-v3`, > `omni_runner` automatically downloads > `parakeet_tdt_0.6b_v3_5s_i8_stateful.litertlm` into `--cache_dir` on first > run.