# Zero-shot inference

The Gemma 3 27B IT model, introduced by Google, is suitable for various text generation and image understanding tasks, supporting a context length of 128k tokens and multimodal image processing.

Distill-Any-Depth is a new SOTA monocular depth estimation model trained using knowledge distillation algorithms.

InternVL2_5 - 1B is a multimodal large model based on image-text to text conversion, suitable for vision-language tasks.

A model obtained by merging pre-trained language models using mergekit, with evaluation results on multiple text generation tasks

Chronos is a family of pretrained time series forecasting models based on language model architecture, which converts time series into token sequences through quantization and scaling for training, supporting probabilistic forecasting.

A zero-shot classification model based on DeBERTa-v3, supporting multi-task natural language inference (MNLI), FEVER, and ANLI tasks, converted to ONNX format for compatibility with Transformers.js

Fast inference text classification model trained on MultiNLI dataset, using MiniLM-L6 architecture