Improve model card

Hi! I'm Niels from the Hugging Face community team.

This PR improves the model card for Marchuk. Key changes include:
- Adding a link to the project page.
- Adding a brief summary of the model's architecture and performance (276M parameters, latent flow-matching).
- Including installation instructions from the official repository.
- Adding the BibTeX citation.

These additions help users better discover and utilize the model.

README.md

@@ -3,8 +3,43 @@ license: mit
 pipeline_tag: time-series-forecasting
 tags:
 - climate
+- weather-forecasting
 ---
 
-This repository contains the Marchuk presented in [Marchuk: Efficient Global Weather Forecasting from Mid-Range to Sub-Seasonal Scales via Flow Matching](https://arxiv.org/abs/2603.24428)
+# Marchuk: Efficient Global Weather Forecasting from Mid-Range to Sub-Seasonal Scales via Flow Matching
 
-For inference code see https://github.com/v-gen-ai/Marchuk
+[Paper](https://arxiv.org/abs/2603.24428) | [Project Page](https://v-gen-ai.github.io/Marchuk/) | [GitHub](https://github.com/v-gen-ai/Marchuk)
+
+Marchuk is a generative latent flow-matching model for global weather forecasting spanning mid-range to subseasonal timescales, with prediction horizons of up to 30 days. It conditions on current-day weather maps and autoregressively predicts subsequent days' weather maps within a learned latent space.
+
+### Key Features
+- **Efficiency**: A compact architecture of 276 million parameters.
+- **Performance**: Achieves predictive skill comparable to much larger models (e.g., LaDCast 1.6B) while operating at higher inference speeds.
+- **Architecture**: Replaces rotary positional encodings (RoPE) with trainable positional embeddings and an extended temporal context window to capture long-range temporal dependencies.
+
+## Installation
+
+To use the model, you can install the necessary dependencies as follows:
+
+```bash
+conda create -n marchuk
+conda activate marchuk
+git clone https://github.com/tonyzyl/ladcast.git
+pip install -e ladcast
+```
+
+For evaluation, refer to the `inference.ipynb` provided in the [official repository](https://github.com/v-gen-ai/Marchuk).
+
+## BibTeX
+
+```bibtex
+@misc{kuzhamuratov2026marchukefficientglobalweather,
+      title={Marchuk: Efficient Global Weather Forecasting from Mid-Range to Sub-Seasonal Scales via Flow Matching}, 
+      author={Arsen Kuzhamuratov and Mikhail Zhirnov and Andrey Kuznetsov and Ivan Oseledets and Konstantin Sobolev},
+      year={2026},
+      eprint={2603.24428},
+      archivePrefix={arXiv},
+      primaryClass={cs.LG},
+      url={https://arxiv.org/abs/2603.24428}, 
+}
+```