README.md

---
license: apache-2.0
datasets:
- ealvaradob/phishing-dataset
- ucberkeley-dlab/measuring-hate-speech
- cardiffnlp/tweet_eval
- lmsys/toxic-chat
- tasksource/jigsaw_toxicity
language:
- en
base_model:
- answerdotai/ModernBERT-large
pipeline_tag: text-classification
tags:
- moderation
- safety
---

# Horizon 1

A larger and more modern variant of [Constellation-One](https://huggingface.co/DominicTWHV/Constellation-One-Text) for [Cockatoo](https://cockatoo.dev/) from answerdotai/modernBERT-large

This model is licensed under the `Apache-2.0` license

**Note:**

`lmsys/toxic-chat` is licensed under `CC-BY-NC-4.0`, meaning this model cannot be legally used for commercial purposes.

## Architecture

<a href="https://hfviewer.com/DominicTWHV/Horizon-1-Text-Large?utm_source=huggingface&amp;utm_medium=embedded_model_card&amp;utm_campaign=DominicTWHV__Horizon-1-Text-Large_card" target="_blank" rel="noopener">
  <img
    src="https://hfviewer.com/api/card.svg?source=DominicTWHV%2FHorizon-1-Text-Large&amp;v=20260501clipcard"
    alt="Open DominicTWHV/Horizon-1-Text-Large in hfviewer"
    width="100%"
  />
</a>

## Hardware:

This model was fine-tuned on two NVIDIA A40s with a batch size of 32 and gradient accumulation of 2, totaling to an effective batch size of `(32*2) * 2 = 128`

Fine-tuned on a dataset size of 232k entries aggregated from:

```csv
- ealvaradob/phishing-dataset
- ucberkeley-dlab/measuring-hate-speech
- cardiffnlp/tweet_eval
- lmsys/toxic-chat
- tasksource/jigsaw_toxicity
```

## Software

Training was executed on the [Cockatoo_ML_Training](https://github.com/DominicTWHV/Cockatoo_ML_Training) server. Metrics are publicly visible at [Cockatoo.dev](https://cockatoo.dev/ml-training.html) .

Techniques: `or` label merging, `merge_labels` on conflict. There have been **no** manual intervention in data sanitization before/after merging.

Asymmetric losses:

```csv
γ- = 3.5
γ+ = 0.5
clipping = 0.05
```

Optimizer:

```csv
adamw

betas = (0.9, 0.999)
eps = 1e-8
momentum = 0.9
```

LLRD:

```csv
decay_factor = 0.98
```

Hyperparameters:

```csv
epoch = 3

batch_size = 32
gradient_accumulation = 2

learning_rate = 5e-5
weight_decay = 0.1
warmup_ratio = 0.1

fp16 = false
bf16 = true
tf32 = true

gradient_checkpointng = false
gradient_clipping = true
gradient_clipping_val = 1.0

attention_implementation = "flash_attention_2"
```

## Available Labels:

```json
"id2label": {
  "0": "scam",
  "1": "violence",
  "2": "harassment",
  "3": "hate_speech",
  "4": "toxicity",
  "5": "obscenity",
  "6": "genocide" # genocide is a new addition compared to Constellation
}
```

## Performance

*All evaluation metrics are from macro averaging, may contain slight deviations with other data entries due to the discrepancy in different evaluation runs. Metrics from zero-shot evaluation split (not present in training data)*

Horizon 1 achieves very high recall values out of the box (0.94 raw) with a comparable precision compared to Constellation (0.566 raw vs. 0.605).

However, this model really shines when trigger thresholds have been fine-tuned:

**Default:**

| Category | Threshold | F1-Score |
| :--- | :--- | :--- |
| scam | 0.5 | 0.8758 |
| violence | 0.5 | 0.6891 |
| harassment | 0.5 | 0.8279 |
| hate_speech | 0.5 | 0.6581 |
| toxicity | 0.5 | 0.6430 |
| obscenity | 0.5 | 0.6428 |
| genocide | 0.5 | 0.5630 |
| **Average** | - | **0.7000** |

![Recall Metrics](assets/graphs/untuned/recall_modernbert.png)
![Precision Metrics](assets/graphs/untuned/precision_modernbert.png)
![F1 Metrics](assets/graphs/untuned/f1_modernbert.png)

**Tuned:**

| Category | Threshold | F1-Score | Delta (vs. default) |
| :--- | :--- | :--- | :--- |
| scam | 0.7129 | 0.9131 | +0.0373 |
| violence | 0.6238 | 0.7252 | +0.0361 |
| harassment | 0.6535 | 0.8712 | +0.0433 |
| hate_speech | 0.6040 | 0.7082 | +0.0501 |
| toxicity | 0.6238 | 0.7371 | +0.0941 |
| obscenity | 0.6238 | 0.7309 | +0.0881 |
| genocide | 0.6337 | 0.5929 | +0.0299 |
| **Average** | - | **0.7541** | **+0.0541** 

![Recall Metrics](assets/graphs/tuned/recall_modernbert.png)
![Precision Metrics](assets/graphs/tuned/precision_modernbert.png)
![F1 Metrics](assets/graphs/tuned/f1_modernbert.png)

## Comparison with Constellation One (tuned):

| Metric | Constellation One | Horizon 1 | Delta (H1 - C1) |
| --- | --- | --- | --- |
| **Loss** | 0.1603 | 0.0245 | **-0.1358** |
| **Overall Precision** | 0.6940 | 0.6809 | -0.0131 |
| **Overall Recall** | 0.8151 | 0.8554 | **+0.0403** |
| **Overall F1** | 0.7475 | 0.7448 | -0.0027 |
| Scam Precision | 0.9255 | 0.9330 | **+0.0075** |
| Scam Recall | 0.9467 | 0.9009 | -0.0459 |
| Scam F1 | 0.9360 | 0.9167 | -0.0194 |
| Violence Precision | 0.5141 | 0.6293 | **+0.1152** |
| Violence Recall | 0.7191 | 0.8828 | **+0.1637** |
| Violence F1 | 0.5995 | 0.7348 | **+0.1353** |
| Harassment Precision | 0.8238 | 0.8329 | **+0.0091** |
| Harassment Recall | 0.8830 | 0.9240 | **+0.0410** |
| Harassment F1 | 0.8524 | 0.8761 | **+0.0237** |
| Hate Speech Precision | 0.5607 | 0.5965 | **+0.0358** |
| Hate Speech Recall | 0.6960 | 0.8652 | **+0.1692** |
| Hate Speech F1 | 0.6211 | 0.7061 | **+0.0850** |
| Toxicity Precision | 0.6891 | 0.6946 | **+0.0056** |
| Toxicity Recall | 0.8025 | 0.7481 | -0.0544 |
| Toxicity F1 | 0.7415 | 0.7204 | -0.0211 |
| Obscenity Precision | 0.6507 | 0.6828 | **+0.0321** |
| Obscenity Recall | 0.8431 | 0.7160 | -0.1271 |
| Obscenity F1 | 0.7345 | 0.6990 | -0.0355 |
| Genocide Precision | N/A | 0.3972 | N/A |
| Genocide Recall | N/A | 0.9511 | N/A |
| Genocide F1 | N/A | 0.5604 | N/A |

> [!NOTE]
> This model is more "trigger-happy" compared to Constellation One, albeit this can be mitigated in production by increasing thresholds (current values optimized for macro F1).

A newer version is planned to mitigate this behavior.

## Resources:

Training/Inferencing server: https://github.com/DominicTWHV/Cockatoo_ML_Training/

Training Metrics: https://cockatoo.dev/ml-training.html

## Datasets Used | Citations

| Dataset | License | Link |
| --- | --- | --- |
| **Phishing Dataset** | MIT | [Hugging Face](https://huggingface.co/datasets/ealvaradob/phishing-dataset) |
| **Measuring Hate Speech** | CC-BY-4.0 | [Hugging Face](https://huggingface.co/datasets/ucberkeley-dlab/measuring-hate-speech) |
| **Tweet Eval (SemEval-2019)** | [See Citation]* | [Hugging Face](https://huggingface.co/datasets/cardiffnlp/tweet_eval) |
| **Toxic Chat** | CC-BY-NC-4.0 | [Hugging Face](https://huggingface.co/datasets/lmsys/toxic-chat) |
| **Jigsaw Toxicity** | Apache-2.0 | [Hugging Face](https://huggingface.co/datasets/tasksource/jigsaw_toxicity) |

---

### Citation: ucberkeley-dlab/measuring-hate-speech

```bibtex
@article{kennedy2020constructing,
  title={Constructing interval variables via faceted Rasch measurement and multitask deep learning: a hate speech application},
  author={Kennedy, Chris J and Bacon, Geoff and Sahn, Alexander and von Vacano, Claudia},
  journal={arXiv preprint arXiv:2009.10277},
  year={2020}
}
```

### Citation: cardiffnlp/tweet_eval

```bibtex
@inproceedings{basile-etal-2019-semeval,
    title = "{S}em{E}val-2019 Task 5: Multilingual Detection of Hate Speech Against Immigrants and Women in {T}witter",
    author = "Basile, Valerio and Bosco, Cristina and Fersini, Elisabetta and Nozza, Debora and Patti, Viviana and Rangel Pardo, Francisco Manuel and Rosso, Paolo and Sanguinetti, Manuela",
    booktitle = "Proceedings of the 13th International Workshop on Semantic Evaluation",
    year = "2019",
    address = "Minneapolis, Minnesota, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/S19-2007",
    doi = "10.18653/v1/S19-2007",
    pages = "54--63"
}

```

### Citation: lmsys/toxic-chat

```bibtex
@misc{lin2023toxicchat,
      title={ToxicChat: Unveiling Hidden Challenges of Toxicity Detection in Real-World User-AI Conversation}, 
      author={Zi Lin and Zihan Wang and Yongqi Tong and Yangkun Wang and Yuxin Guo and Yujia Wang and Jingbo Shang},
      year={2023},
      eprint={2310.17389},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```