Improve model card metadata and add prominent links to paper and code

Hi, I'm Niels from the community science team at Hugging Face.

This PR improves the model card for Toto-2.0-1B by:
1. Adding `datasets` metadata to link the model to the `Datadog/BOOM` and `Salesforce/GIFT-Eval` datasets.
2. Adding prominent links to the paper, GitHub repository, and project blog at the top of the README.
3. Mentioning the paper authors in the description.
4. Formally linking the model to the research paper "Toto 2.0: Time Series Forecasting Enters the Scaling Era" (arXiv:2605.20119).

Please review and merge if this looks good to you!

README.md

@@ -1,4 +1,9 @@
 ---
+license: apache-2.0
+pipeline_tag: time-series-forecasting
+datasets:
+- Datadog/BOOM
+- Salesforce/GIFT-Eval
 tags:
 - time-series-forecasting
 - foundation-models
@@ -9,63 +14,65 @@ tags:
 - observability
 - safetensors
 - pytorch_model_hub_mixin
-license: apache-2.0
-pipeline_tag: time-series-forecasting
 thumbnail: https://web-assets.dd-static.net/42588/1778691695-toto-2-hero.png
 model-index:
 - name: Toto-2.0-1B
   results:
-    - task:
-        type: time-series-forecasting
-      dataset:
-        name: BOOM
-        type: BOOM
-      metrics:
-        - name: CRPS
-          type: CRPS
-          value: 0.349
-        - name: MASE
-          type: MASE
-          value: 0.582
-      source:
-        name: BOOM 💥 Observability Time-Series Forecasting Leaderboard
-        url: https://huggingface.co/spaces/Datadog/BOOM
-    - task:
-        type: time-series-forecasting
-      dataset:
-        name: GIFT-Eval
-        type: GIFT-Eval
-      metrics:
-        - name: CRPS
-          type: CRPS
-          value: 0.478
-        - name: MASE
-          type: MASE
-          value: 0.699
-      source:
-        name: GIFT-Eval Time Series Forecasting Leaderboard
-        url: https://huggingface.co/spaces/Salesforce/GIFT-Eval
-    - task:
-        type: time-series-forecasting
-      dataset:
-        name: TIME
-        type: TIME
-      metrics:
-        - name: CRPS
-          type: CRPS
-          value: 0.537
-        - name: MASE
-          type: MASE
-          value: 0.643
-      source:
-        name: TIME Benchmark Leaderboard
-        url: https://huggingface.co/spaces/Real-TSF/TIME-leaderboard
+  - task:
+      type: time-series-forecasting
+    dataset:
+      name: BOOM
+      type: BOOM
+    metrics:
+    - type: CRPS
+      value: 0.349
+      name: CRPS
+    - type: MASE
+      value: 0.582
+      name: MASE
+    source:
+      url: https://huggingface.co/spaces/Datadog/BOOM
+      name: BOOM 💥 Observability Time-Series Forecasting Leaderboard
+  - task:
+      type: time-series-forecasting
+    dataset:
+      name: GIFT-Eval
+      type: GIFT-Eval
+    metrics:
+    - type: CRPS
+      value: 0.478
+      name: CRPS
+    - type: MASE
+      value: 0.699
+      name: MASE
+    source:
+      url: https://huggingface.co/spaces/Salesforce/GIFT-Eval
+      name: GIFT-Eval Time Series Forecasting Leaderboard
+  - task:
+      type: time-series-forecasting
+    dataset:
+      name: TIME
+      type: TIME
+    metrics:
+    - type: CRPS
+      value: 0.537
+      name: CRPS
+    - type: MASE
+      value: 0.643
+      name: MASE
+    source:
+      url: https://huggingface.co/spaces/Real-TSF/TIME-leaderboard
+      name: TIME Benchmark Leaderboard
 ---
 
 # Toto-2.0-1B
 
+[**Technical Report**](https://huggingface.co/papers/2605.20119) | [**GitHub**](https://github.com/DataDog/toto) | [**Blog Post**](https://www.datadoghq.com/blog/ai/toto-2/)
+
 Toto (Time Series Optimized Transformer for [Observability](https://www.datadoghq.com/knowledge-center/observability/)) is a family of time series foundation models for multivariate forecasting developed by [Datadog](https://www.datadoghq.com/). Toto 2.0 is the current generation, featuring u-μP-scaled transformers ranging from 4m to 2.5B parameters, all trained from a single recipe. Forecast quality improves reliably with parameter count across the family.
 
+The model was introduced in the paper [Toto 2.0: Time Series Forecasting Enters the Scaling Era](https://huggingface.co/papers/2605.20119) by Emaad Khwaja, Chris Lettieri, Gerald Woo, Eden Belouadah, Marc Cenac, Guillaume Jarry, Enguerrand Paquin, Xunyi Zhao, Viktoriya Zhukov, Othmane Abou-Amal, Chenghao Liu, Ameet Talwalkar, and David Asker.
+
 The family sets a new state of the art on three forecasting benchmarks: [BOOM](https://huggingface.co/spaces/Datadog/BOOM), our observability benchmark; [GIFT-Eval](https://huggingface.co/spaces/Salesforce/GIFT-Eval), the standard general-purpose benchmark; and the recent contamination-resistant [TIME](https://arxiv.org/abs/2602.12147) benchmark.
 
 ## 📊 Performance
@@ -136,12 +143,12 @@ All five Toto 2.0 sizes share the same training recipe; pick a size based on you
 
 <figure>
 <img src="assets/architecture.png" alt="Overview of the Toto 2.0 architecture.">
-<figcaption>A decoder-only patched transformer whose attention layers alternate between time-axis (causal) and variate-axis (full) views of the input. Toto 2.0 adds <b>contiguous patch masking (CPM)</b> for single-pass parallel decoding, a <b>quantile output head</b> trained with pinball loss, a robust arcsinh input scaler, residual MLP patch projections, and is trained with NorMuon. See the <a href="#-additional-resources">technical report</a> for details.</figcaption>
+<figcaption>A decoder-only patched transformer whose attention layers alternate between time-axis (causal) and variate-axis (full) views of the input. Toto 2.0 adds <b>contiguous patch masking (CPM)</b> for single-pass parallel decoding, a <b>quantile output head</b> trained with pinball loss, a robust arcsinh input scaler, residual MLP patch projections, and is trained with NorMuon. See the <a href="https://huggingface.co/papers/2605.20119">technical report</a> for details.</figcaption>
 </figure>
 
 ## 🔗 Additional Resources
 
-- [Technical Report](https://arxiv.org/abs/2605.20119)
+- [Technical Report](https://huggingface.co/papers/2605.20119)
 - [Blog Post](https://www.datadoghq.com/blog/ai/toto-2/)
 - [GitHub Repository](https://github.com/DataDog/toto)
 - [Toto 2.0 Collection](https://huggingface.co/collections/Datadog/toto-20) — all five base checkpoints
@@ -160,4 +167,4 @@ All five Toto 2.0 sizes share the same training recipe; pick a size based on you
       primaryClass={cs.LG},
       url={https://arxiv.org/abs/2605.20119}, 
 }
-```
+```