Search FoxAI

AI Models 2119

• InternVL3-2B↗

  OpenGVLab

  We introduce InternVL3, an advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. Compared to InternVL 2.5, InternVL3 exhibits superior multimodal perception and reasoning capabilities, while further extending its multimodal capabilities to encompass tool usage, GUI agents, industrial image analysis, 3D vision perception, and more. Additionally, we compare InternVL3 with Qwen2.5 Chat models, whose corresponding pre-trained base models are employed as the initialization of the langauge component in InternVL3. Benefitting from Native Multimodal Pre-Training, the InternVL3 series achieves even better overall text performance than the Qwen2.5 series. As shown in the following figure, InternVL3 retains the same model architecture as InternVL 2.5 and its predecessors, InternVL 1.5 and 2.0, following the "ViT-MLP-LLM" paradigm. In this new version, we integrate a newly incrementally pre-trained InternViT with various pre-trained LLMs, including InternLM 3 and Qwen 2.5, using a randomly initialized MLP projector. As in the previous version, we applied a pixel unshuffle operation, reducing the number of visual tokens to one-quarter of the original. Besides, we adopted a similar dynamic resolution strategy as InternVL 1.5, dividing images into tiles of 448×448 pixels. The key difference, starting from InternVL 2.0, is that we additionally introduced support for multi-image and video data.

• wav2vec2-large-xlsr-53-mongolian↗

  anton-l

  Fine-tuned facebook/wav2vec2-large-xlsr-53 on Mongolian using the Common Voice dataset. When using this model, make sure that your speech input is sampled at 16kHz. The model can be evaluated as follows on the Mongolian test data of Common Voice.

• mask2former-swin-large-mapillary-vistas-panoptic↗

  facebook

  Mask2Former model trained on Mapillary Vistas panoptic segmentation (large-sized version, Swin backbone). It was introduced in the paper Masked-attention Mask Transformer for Universal Image Segmentation and first released in this repository. Disclaimer: The team releasing Mask2Former did not write a model card for this model so this model card has been written by the Hugging Face team. Mask2Former addresses instance, semantic and panoptic segmentation with the same paradigm: by predicting a set of masks and corresponding labels. Hence, all 3 tasks are treated as if they were instance segmentation. Mask2Former outperforms the previous SOTA, MaskFormer both in terms of performance an efficiency by (i) replacing the pixel decoder with a more advanced multi-scale deformable attention Transformer, (ii) adopting a Transformer decoder with masked attention to boost performance without without introducing additional computation and (iii) improving training efficiency by calculating the loss on subsampled points instead of whole masks.

• rtdetr_v2_r101vd↗

  PekingU

  The RT-DETRv2 model was proposed in RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer by Wenyu Lv, Yian Zhao, Qinyao Chang, Kui Huang, Guanzhong Wang, Yi Liu. RT-DETRv2 refines RT-DETR by introducing selective multi-scale feature extraction, a discrete sampling operator for broader deployment compatibility, and improved training strategies like dynamic data augmentation and scale-adaptive hyperparameters. These changes enhance flexibility and practicality while maintaining real-time performance. RT-DETRv2 consistently outperforms its predecessor across all model sizes while maintaining the same real-time speeds. RT-DETRv2 is trained on COCO (Lin et al. [2014]) train2017 and validated on COCO val2017 dataset. We report the standard AP metrics (averaged over uniformly sampled IoU thresholds ranging from 0.50 − 0.95 with a step size of 0.05), and APval50 commonly used in real scenarios.

• pixal3d-sv-q8_0-v1↗

  raven38

  GGUF weights for `pixal3d.cpp`, a GGML-based C++ image→3D pipeline. This is the **single-view (SV) Web release set**: Q8_0 quantization, 9 files, 7.54 GiB. It is the one-image counterpart of the multiview set `raven38/pixal3d-q8_0-v1`: the four flow DiTs are the official Pixal3D **single-view** checkpoints, and the five shared models (image encoder, NAF, three decoders) are byte-identical to the multiview set. The repository layout is flat and the file names are exact, because the consuming client verifies every file's `size_bytes` and SHA-256 against the committed manifest `pixal3d-models.json` (`model_family: sv`) before the set is marked usable. The Web app never substitutes multiview weights for a single image and never substitutes these weights for a multiview run; the two families are selected only through their own manifests. The four `*_sv.gguf` files were converted with `tools/convert.py` from the official TencentARC/Pixal3D single-view checkpoints `ckpts/ss_flow_img_dit_1_3B_64_bf16.safetensors`, `ckpts/slat_flow_img2shape_dit_1_3B_512_bf16.safetensors`, `ckpts/slat_flow_img2shape_dit_1_3B_1024_bf16.safetensors` and `ckpts/slat_flow_imgshape2tex_dit_1_3B_1024_bf16.safetensors` (the variants **without** the `_mv` suffix) to F16 GGUF, then quantized with `tools/quantize_gguf.py Q8_0`. The SV and MV checkpoints share the same config and tensor layout; only the trained values differ.

• GLM-5.3 (Max)↗

  Z AI

  GLM-5.3 uses the same base model as GLM-5.2 — every gain comes from post-training. Compared with GLM-5.2, it is much better at complex coding and long-horizon tasks: + Stronger Coding: GLM-5.3 is the most capable open-weights model for coding, with a 50% improvement over GLM-5.2 on our in-house Z.ai Code Bench. It also achieve open-source SOTA on public benchmarks including Terminal Bench 3.0 and Agents' Last Exam. + Emergent Cyber Capability: As we scaled post-training, cyber capability developed faster than we expected. GLM-5.3 is state of the art on CyberGym for vulnerability discovery, and its gains are largest further up the exploitation chain, where it more than doubles GLM-5.2 on exploitation benchmarks.

• SocRoBERTa-social↗

  ESGBERT

  Based on this paper, this is the SocRoBERTa-social language model. A language model that is trained to better classify social texts in the ESG domain. *Note: We generally recommend choosing the SocialBERT-social model since it is quicker, less resource-intensive and only marginally worse in performance.* Using the SocRoBERTa-base model as a starting point, the SocRoBERTa-social Language Model is additionally fine-trained on a 2k social dataset to detect social text samples.

• siglip2-person-description-reid↗

  MarketaJu

  This model is part of the family of SigLIP models finetuned for **person visual description and retrieval**. The splits are created as follows: - **CUHK-PEDES**: **test**: 1000 identities following the original split; **validation**: 150 identities sampled from the original validation split; **train**: all remaining identities - **ICFG-PEDES**: **test**: 946 identities following the original split; **validation**: 150 identities sampled from the original training split; **train**: all remaining identities - **IIITD**: **test**: 2500 identities following the original split; **validation**: 150 identities sampled from the original validation split; **train**: all remaining identities - **ITCPR**: Originally designed for zero-shot evaluation, so a new split is created:**test**: 1000 randomly selected identities; **validation**: 150 randomly selected identities; **train**: all remaining identities - **PRW**: **test**: 450 identities following the original split; **validation**: 150 identities sampled from the original training split; **train**: all remaining identities - **PETA**: The original random image-level split can cause identity leakage. Therefore, the identity-based split from PETA-ZS is adopted: **test**: 1706 identities following the original split; **validation**: 150 identities sampled from the original validation split; **train**: all remaining identities - **RSTPReid**: Only the 200 test images are used for evaluation due to annotation errors in the rest of the dataset. - **SYNTH**: 1000 random images are used for the test set because the generated captions are noisy and sometimes inaccurate. The model is evaluated on data mentioned above. The following table summarizes the number of identities, images, and queries for each subset. The final test set is a merge of all subsets.

• Mistral-7B-Instruct-v0.1↗

  Mistral AI

  > [!TIP] > PRs to correct the `transformers` tokenizer so that it gives 1-to-1 the same results as the `mistral_common` reference implementation are very welcome! The Mistral-7B-Instruct-v0.1 Large Language Model (LLM) is a instruct fine-tuned version of the Mistral-7B-v0.1 generative text model using a variety of publicly available conversation datasets. For full details of this model please read our paper and release blog post.

• OmniCoder-Qwen3.5-9B-Claude-4.6-Opus-Uncensored-v2-GGUF↗

  Ngixdev

  If you want to disable thinking use this chat template in LM Studio, but I don't reccomend to do it for 9B model, because it's already crazy fast enough: https://pastebin.com/uk9ZkxCR And write anything you want after that. Looks like model is underperforming without this first line. > **v2 Update:** > This iteration is powered by **14,000+ premium Claude 4.6 Opus-style general reasoning samples**, with a major focus on **achieving massive gains in reasoning efficiency while actively improving peak accuracy**. > > v2 introduces a **refined reasoning scaffold** designed to eliminate redundant internal loops, significantly improving the model's **cross-task generalization** from logic and math into specialized fields like programming. Compared to the original model, **autonomy and stability are significantly improved**, ensuring the model remains robust and self-consistent during complex, multi-step problem solving. v2 is built to **think smarter, not longer**, delivering substantial improvements in inference speed and cost-effectiveness while simultaneously boosting baseline accuracy. > > *Note: Due to the constraints of SFT sample size and training scope, the model's broad general-purpose capabilities might be slightly impacted. The efficiency and accuracy results discussed here are based on the HumanEval and HumanEval+ benchmarks. Thank you for your understanding!*