ModernBERT-large — emotion classifier (balanced 6-dataset fine-tune)
Fine-tune of
answerdotai/ModernBERT-large on a per-class
balanced merge of 6 English emotion datasets, mirroring the methodology of
j-hartmann/emotion-english-distilroberta-base.
This is the production default for the EmotiSpeech NTU SC4001 project. Smaller sister model:
maxpicy/modernbert-base-emotion-balanced.
Labels (7-class Ekman + neutral)
anger, disgust, fear, joy, neutral, sadness, surprise
Training data
6 datasets harmonised to the 7-class scheme, then per-class downsampled to 2,045 examples (size of the smallest class after deduping).
Source License Pre-balance contribution Crowdflower 2016 (40k tweets) Public domain anger, joy, neutral, sadness, surprise, fear (via worry) dair-ai/emotion (Saravia et al. 2018)unknown anger, fear, joy, sadness, surprise google-research-datasets/go_emotions (Demszky et al. 2020)Apache 2.0 all 7 (single-label rows only) gsri-18/ISEAR-dataset-complete (Vikash 2018)unknown anger, disgust, fear, joy, sadness MELD (Poria et al. 2019) GPL-3.0 all 7 cardiffnlp/tweet_eval config emotionunknown anger, joy, sadness
Splits after balancing: train 10,020 / val 1,432 / test 2,863.
Training
Base model : answerdotai/ModernBERT-large (~395M params)Hyperparameters : 2 epochs (epoch 3 overfit on the 3-epoch run; eval_loss went 0.89 → 1.54), batch 16, lr 2e-5, AdamWHardware : 1× A100 on NSCC ASPIRE 2A (g1 queue), ~14 minutes wall-clockTokenization : HF auto-tokenizer, max_length 256
Test-set evaluation
Metric Value accuracy 0.607 macro_f1 0.608 weighted_f1 0.608
Per-class F1: anger 0.627, disgust 0.751, fear 0.522, joy 0.663, neutral 0.499, sadness 0.590, surprise 0.600. Beats the
base variant by ~3 points on macro-F1.
Usage
1 from transformers import AutoModelForSequenceClassification , AutoTokenizer
2 import torch
3
4 ckpt = "maxpicy/modernbert-large-emotion-balanced"
5 tok = AutoTokenizer . from_pretrained ( ckpt )
6 model = AutoModelForSequenceClassification . from_pretrained ( ckpt ) . eval ( )
7
8 texts = [ "What is happening?" , "I'm so happy today!" , "I can't believe this." ]
9 inputs = tok ( texts , padding = True , truncation = True , return_tensors = "pt" )
10 with torch . inference_mode ( ) :
11 probs = torch . softmax ( model ( ** inputs ) . logits , dim = - 1 )
12
13 id2label = model . config . id2label
14 for text , p in zip ( texts , probs ) :
15 top = int ( p . argmax ( ) )
16 print ( f" { text !r : 40s } -> { id2label [ top ] } ( { p [ top ] : .2f } )" )
Audio benchmark behaviour
On the EmotiSpeech 63-second kfseetoh.wav benchmark (123 rolling-window inferences):
Model Distinct dominant labels Confidence range j-hartmann pretrained baseline 6 0.25–0.98 maxpicy/modernbert-large-emotion-balanced6 0.26–0.98
Matches the j-hartmann reference baseline on label diversity and exceeds it on per-class diagnostic granularity (24 surprised predictions vs 13).
Citation
If this checkpoint is useful in your work, please credit the upstream models and datasets, plus:
@misc{wong2026emotispeech,
author = {Wong, Max et al.},
title = {EmotiSpeech: word-level multimodal speech emotion},
year = {2026},
note = {NTU SC4001 academic project},
}Methodology mirrors j-hartmann/emotion-english-distilroberta-base — please cite their work too.
License
MIT for the model weights and configuration. Underlying datasets retain their own licenses (see table above).
