Saved toolsSubmit a tool
FOXAI SEARCH

Search FoxAI

AI Models 2119

  • leosamsHelloworldXL_helloworldXL60↗

    misri

  • GLM-OCR↗

    unsloth

    GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture. It introduces Multi-Token Prediction (MTP) loss and stable full-task reinforcement learning to improve training efficiency, recognition accuracy, and generalization. The model integrates the CogViT visual encoder pre-trained on large-scale image–text data, a lightweight cross-modal connector with efficient token downsampling, and a GLM-0.5B language decoder. Combined with a two-stage pipeline of layout analysis and parallel recognition based on PP-DocLayout-V3, GLM-OCR delivers robust and high-quality OCR performance across diverse document layouts. For speed, we compared different OCR methods under identical hardware and testing conditions (single replica, single concurrency), evaluating their performance in parsing and exporting Markdown files from both image and PDF inputs. Results show GLM-OCR achieves a throughput of 1.86 pages/second for PDF documents and 0.67 images/second for images, significantly outperforming comparable models. 2. **Information Extraction** – extract structured information from documents. Prompts must follow a strict JSON schema. For example, to extract personal ID information:

  • spark-s1-4b-v6-GGUF↗

    abhishek085

    GGUF builds of `abhishek085/spark-s1-4b-v6` for **llama.cpp**, **Ollama**, LM Studio, Jan and other GGUF runtimes. Part of Open Spark Jev, an open-source project of the Nokast AI community. > **This is not a chat model.** `spark-s1` is a System-1 *decision* model: given a state and a question with a fixed menu of options, the > answer is the probability distribution over the option letters at the **first** generated token. Chatting with it (`ollama run ...`) only > ever produces a single letter. Use it through the logprob recipe below or through the Open Spark Jev gateway. Measured on 200 shuffled rows of the project's locked held-out split (`v5_test_locked`), llama.cpp `llama-server` (CUDA, build `38a5b42`), `/v1/completions` with `top_logprobs=20`, calibrated temperature applied client-side. Ollama 0.40.2 with the Q8_0 file gave the same 99.5% / 0.900. This is a parity check against the HF model, not a new benchmark; the model's own evaluation is on the base model card.

  • Quen32↗

    AlSamCur123

    **Example usage**: - For text only LLMs: `llama-cli -hf AlSamCur123/Quen32 --jinja` - For multimodal models: `llama-mtmd-cli -hf AlSamCur123/Quen32 --jinja`

  • chronos-t5-small↗

    amazon

    🚀 **Update Feb 14, 2025**: Chronos-Bolt & original Chronos 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. 🚀 **Update Nov 27, 2024**: We have released Chronos-Bolt⚡️ models that are more accurate (5% lower error), up to 250 times faster and 20 times more memory-efficient than the original Chronos models of the same size. Check out the new models here. Chronos is a family of **pretrained time series forecasting models** based on language model architectures. A time series is transformed into a sequence of tokens via scaling and quantization, and a language model is trained on these tokens using the cross-entropy loss. Once trained, probabilistic forecasts are obtained by sampling multiple future trajectories given the historical context. Chronos models have been trained on a large corpus of publicly available time series data, as well as synthetic data generated using Gaussian processes.

  • FLUX.2-VAE↗

    unsloth

    The single-file FLUX.2 VAE that stable-diffusion.cpp needs for both FLUX.2 [dev] and FLUX.2 klein, mirrored from Comfy-Org/flux2-dev. Unsloth Studio fetches this on the no-GPU (native sd.cpp) path. It used to come straight from the community repack; a rename or a takedown there turns every CPU FLUX.2 load into a 404, so it is mirrored here. **The weights are unmodified:** byte for byte the file of the same name in the source repo. The rest of FLUX.2 [dev] is under the FLUX Non-Commercial License. The **autoencoder is not**. Black Forest Labs released it separately and says so in their own repository: "The autoencoder is released under Apache 2.0" (github.com/black-forest-labs/flux2).

  • Yi-9B-200K↗

    01-ai

    If you're interested in Yi's adoption of Llama architecture and license usage policy, see Yi's relation with Llama. ⬇️ > 💡 TL;DR > > The Yi series models adopt the same model architecture as Llama but are **NOT** derivatives of Llama. 🔥 2024-07-29 : The Yi Cookbook 1.0 is released, featuring tutorials and examples in both Chinese and English.

  • 160M-exp↗

    294-288-pretrain-interaction

    Five 160M GPT-NeoX training runs on the separate M/C/G source pools in Dolma3-16B. `Clean`, `M500`, `C500`, `G500`, and `MCG500` are separate run directories under `runs/`. Each contains all existing `step-*.pt` checkpoints, including the final `step-122071.pt` after 16,000,090,112 tokens. Intermediate checkpoints are approximately 1B tokens apart, with additional scheduled steps. Standalone final weights are under `models/ /model.safetensors`, with a matching config and tokenizer in each run directory. Load one with `GPTNeoXForCausalLM.from_pretrained("294-288-pretrain-interaction/160M-exp", subfolder="models/Clean")` and the tokenizer from the same subfolder. Each `.pt` is a PyTorch training checkpoint containing `model`, optimizer, scheduler, rank RNG states, step, token count, and run/manifest/config hashes. The `.pt` files are **not** direct `from_pretrained` artifacts. For trusted checkpoints, load the `model` state dict into `GPTNeoXForCausalLM(GPTNeoXConfig.from_json_file("config.json"))`. Use the included `tokenizer/`. The `training/` directory contains the training and data code; `manifest.json` fixes the source mixture, event schedule, and seeds.

  • reward-model-deberta-v3-large-v2↗

    OpenAssistant

    Reward model (RM) trained to predict which generated answer is better judged by a human, given a question. All models are train on these dataset with a same split seed across datasets (if validation split wasn't available) Its likely SytheticGPT has somekind of surface pattern on the choosen-rejected pair which makes it trivial to differentiate between better the answer.

  • lingbot-video-moe-30b-a3b↗

    robbyant

    We are excited to introduce **LingBot-Video**, the first open-source large-scale MoE (Mixture-of-Experts) video generation model dedicated to embodied intelligence. As a top-tier video model, LingBot-Video is designed to bridge the gap between video synthesis and physical world understanding. > **💡 Rewriter deployment**: the bundled rewriter uses the single-process > `transformers` backend. For higher throughput, deploy the VLM yourself and call > it through an OpenAI-compatible API. Preserve the two-stage semantics: step 1 > must use the base VLM without the rewriter LoRA, while step 2 must use the same > base VLM with the rewriter LoRA enabled. This can be implemented with two > endpoints, or with one server that can select the adapter per request. See > vLLM / SGLang official docs. Install the optional SGLang dependencies only when using SGLang Diffusion or the fused / FP8 MoE runtime: