全站搜索

AI 模型 2119

• Mini-InternVL-Chat-2B-V1-5↗

  OpenGVLab

  We are delighted to introduce the Mini-InternVL-Chat series. In the era of large language models, many researchers have started to focus on smaller language models, such as Gemma-2B, Qwen-1.8B, and InternLM2-1.8B. Inspired by their efforts, we have distilled our vision foundation model InternViT-6B-448px-V1-5 down to 300M and used InternLM2-Chat-1.8B or Phi-3-mini-128k-instruct as our language model. This resulted in a small multimodal model with excellent performance. As shown in the figure below, we adopted the same model architecture as InternVL 1.5. We simply replaced the original InternViT-6B with InternViT-300M and InternLM2-Chat-20B with InternLM2-Chat-1.8B / Phi-3-mini-128k-instruct. For training, we used the same data as InternVL 1.5 to train this smaller model. Additionally, due to the lower training costs of smaller models, we used a context length of 8K during training. - Architecture: InternViT-300M-448px + MLP + InternLM2-Chat-1.8B - Image size: dynamic resolution, max to 40 tiles of 448 x 448 (4K resolution). - Params: 2.2B

• swin-tiny-patch4-window7-224-finetuned-mri↗

  raedinkhaled

  Swin-Tiny fine-tuned to classify cardiac MRI images as `cad` (coronary artery disease) or `healthy`. This checkpoint comes from the co-authored M.Sc. thesis *Novel Deep Learning Architecture for Predicting Heart Diseases based on Transformers and Attention Mechanism with Explainability Model* by Raedin Khaled Sakhri and Loutfi Boufeligha, supervised by Prof. Abdelouaheb Moussaoui (Ferhat Abbas University, Algeria, July 2022). The thesis compared eight models on the same task. This one is the fine-tuned Swin-Tiny, pretrained on ImageNet-1k. The Trainer logged about 6.1 hours of training for the 3 epochs; the thesis rounds this to about 7 hours. For comparison, in the thesis: ViT-Base 98.27%, DeiT 99.01% (overfitting noted), a Vision Transformer trained from scratch 75.73%, and a CNN 88%.

• Florence-2-large↗

  florence-community

  > [!NOTE] > This is the repository for official transformers converted checkpoint of Microsoft's Florence model. **This is a continued pretrained version of Florence-2-large model with 4k context length, only 0.1B samples are used for continue pretraining, thus it might not be trained well. In addition, OCR task has been updated with line separator ('\n'). COCO OD AP 39.8** Florence-2 is an advanced vision foundation model that uses a prompt-based approach to handle a wide range of vision and vision-language tasks. Florence-2 can interpret simple text prompts to perform tasks like captioning, object detection, and segmentation. It leverages our FLD-5B dataset, containing 5.4 billion annotations across 126 million images, to master multi-task learning. The model's sequence-to-sequence architecture enables it to excel in both zero-shot and fine-tuned settings, proving to be a competitive vision foundation model.

• gemma-4-12B-it-qat-q4_0-unquantized↗

  Google

  > [!Note] > This model card is for the new versions of the Gemma 4 family optimized with Quantization-Aware Training (QAT), which allows preserving similar quality to bfloat16 while dramatically reducing the memory requirements to load the model. > Four versions of the QAT checkpoints are available: > * **Unquantized QAT checkpoints** (Q4_0): Half-precision weights extracted from the QAT pipeline, ideal for custom downstream compilation and research. Available for Gemma 4 E2B, E4B, 12B, 26B A4B, and 31B, and their drafter models. > * **GGUF** (Q4_0): Ready-to-deploy formats for broad ecosystem compatibility. Available for Gemma 4 E2B, E4B, 12B, 26B A4B, and 31B. > * **Mobile-optimized** (wNa8o8): A custom schema engineered explicitly for mobile hardware efficiency. It features targeted 2-bit decoding layers, optimized KV caches, and static activations to maximize VRAM savings. Available for Gemma 4 E2B and E4B. > * **Compressed Tensors** (w4a16): QAT checkpoints serialized in the compressed-tensors format for native, optimized inference with vLLM. Available for Gemma 4 E2B, E4B, 12B, and 31B. > * **Assistant Compatibility**: When using multi-token prediction (speculative decoding) with an assistant model alongside a QAT target model, the assistant model must also be a QAT checkpoint with the same precision to ensure compatibility. Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages. Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in five distinct sizes: **E2B**, **E4B**, **12B**, **26B A4B**, and **31B**. Their diverse sizes make them deployable in environments ranging from high-end phones to laptops and servers, democratizing access to state-of-the-art AI.

• MiniMax-H3-Combat-Base-V2↗

  JOKER141

  Combat Base V2 is a major upgrade over V1, focusing not only on faster combat, but on better action continuity, hit reactions, cumulative damage, spatial movement, and physical cause-and-effect. Combat Base V2 is no longer just a combat LoRA — it now works for both action and dialogue scenes, adding richer body movement, finer sound details, and stronger shot-to-shot continuity to character performance. Combat Base V2 不再只是一个武打 LoRA——它现在可文可武，在文戏中也能增加更丰富的身体动作、更细致的声音表现，以及更强的镜头与状态连续性。 If the LoRA looks muddy or less impactful than expected, check the sampler setup first. Sampler choice matters a lot for this LoRA. Core Physics & Motion Logic / 核心物理与动作逻辑 Combat Base V2 moves away from "attack → reset" loops into a continuous physical chain where missed strikes carry momentum, blocks change direction smoothly, and limbs separate cleanly without sticking. Hit reactions now persist and accumulate—sustained torso bending, lowered guards, and impaired mobility directly build toward convincing heavy finishers.

• OuteTTS-0.2-500M-GGUF↗

  OuteAI

  table { border-collapse: collapse; width: 100%; margin-bottom: 20px; } th, td { border: 1px solid #ddd; padding: 8px; text-align: center; } .best { font-weight: bold; text-decoration: underline; } .box { text-align: center; margin: 20px auto; padding: 30px; box-shadow: 0px 0px 20px 10px rgba(0, 0, 0, 0.05), 0px 1px 3px 10px rgba(255, 255, 255, 0.05); border-radius: 10px; } .badges { display: flex; justify-content: center; gap: 10px; flex-wrap: wrap; margin-top: 10px; } .badge { text-decoration: none; display: inline-block; padding: 4px 8px; border-radius: 5px; color: #fff; font-size: 12px; font-weight: bold; width: 250px; } .badge-hf-blue { background-color: #767b81; } .badge-hf-pink { background-color: #7b768a; } .badge-github { background-color: #2c2b2b; } 🤗 Hugging Face - OuteTTS 0.2 500M 🤗 Hugging Face - OuteTTS 0.2 500M GGUF 🤗 Hugging Face - Demo Space GitHub - OuteTTS OuteTTS-0.2-500M is our improved successor to the v0.1 release. The model maintains the same approach of using audio prompts without architectural changes to the foundation model itself. Built upon the Qwen-2.5-0.5B, this version was trained on larger and more diverse datasets, resulting in significant improvements across all aspects of performance.

• colbert-xm↗

  antoinelouis

  This is a ColBERT model that can be used for semantic search in many languages. It encodes queries and passages into matrices of token-level embeddings and efficiently finds passages that contextually match the query using scalable vector-similarity (MaxSim) operators. The model uses an XMOD backbone, which allows it to learn from monolingual fine-tuning in a high-resource language, like English, and perform zero-shot retrieval across multiple languages. This model can be used with Sentence Transformers as a multi-vector (ColBERT-style late interaction) retriever via the `MultiVectorEncoder`: We use the English training samples from the MS MARCO passage ranking dataset, which contains 8.8M passages and 539K training queries. We do not employ the BM25 netaives provided by the official dataset but instead sample harder negatives mined from 12 distinct dense retrievers, using the msmarco-hard-negatives distillation dataset. Our final training set consists of 6.4M (q, p+, p-) triples.

• tradepulse-finbert-sentiment↗

  Bencode92

  **Task**: Sentiment Classification **Target Column**: `label` **Labels**: ['negative', 'neutral', 'positive'] *Last training: 2026-04-20 17:22* *Dataset: `base_reference.csv` (1797 samples)* - **Base Model**: Bencode92/tradepulse-finbert-sentiment - **Training Mode**: Incremental - **Epochs**: 2 - **Learning Rate**: 1e-05 - **Batch Size**: 4 - **Class Balancing**: None

• granite-timeseries-ttm-r2↗

  ibm-granite

  TinyTimeMixers (TTMs) are compact pre-trained models for Multivariate Time-Series Forecasting, open-sourced by IBM Research. **With model sizes starting from 1M params, TTM introduces the notion of the first-ever “tiny” pre-trained models for Time-Series Forecasting. The paper describing TTM was accepted at NeurIPS 24.** TTM outperforms other models demanding billions of parameters in several popular zero-shot and few-shot forecasting benchmarks. TTMs are lightweight forecasters, pre-trained on publicly available time series data with various augmentations. TTM provides state-of-the-art zero-shot forecasts and can easily be fine-tuned for multi-variate forecasts with just 5% of the training data to be competitive. **Note that zeroshot, fine-tuning and inference tasks using TTM can easily be executed on 1 GPU or on laptops.** TTM r2 comprises TTM variants pre-trained on larger pretraining datasets (\~700M samples). The TTM r2.1 release increases the pretraining dataset size to approximately (\~1B samples). The prior model releases, TTM r1, were trained on \~250M samples and can be accessed here. In general, TTM r2 models perform better than TTM r1 models as they are trained on a larger pretraining dataset. In standard benchmarks, TTM r2 outperform TTM r1 by over 15%. However, the choice of r1 vs. r2 depends on your target data distribution, and hence users should try both variants and pick the best model for your data. The TTM r2 releases support point forecasting use-cases specifically ranging from minutely to hourly resolutions (Ex. 10 min, 15 min, 1 hour.). With the TTM r2.1 release, we add support for daily and weekly resolutions.

• chronos-bolt-base↗

  autogluon

  🚀 **Update Feb 14, 2025**: Chronos-Bolt models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code. Chronos-Bolt is a family of pretrained time series forecasting models which can be used for zero-shot forecasting. It is based on the T5 encoder-decoder architecture and has been trained on nearly 100 billion time series observations. It chunks the historical time series context into patches of multiple observations, which are then input into the encoder. The decoder then uses these representations to directly generate quantile forecasts across multiple future steps—a method known as direct multi-step forecasting. Chronos-Bolt models are **more accurate**, up to **250 times faster** and **20 times more memory-efficient** than the original Chronos models of the same size. The following plot compares the inference time of Chronos-Bolt against the original Chronos models for forecasting 1024 time series with a context length of 512 observations and a prediction horizon of 64 steps.