Hyperparameter	Value
LANGUAGES	en-es-it
LR	1e-05
EPOCHS	5
MAX_LENGTH	256
USE_BIO	False
USE_LANG_TOKEN	False
GATED_BIO	False
FOCAL_LOSS	True
FOCAL_GAMMA	2.5
USE_SAMPLER	False
R_DROP	True
R_KL_ALPHA	1.0
TEXT_NORMALIZE	True

Dev set results (summary)

Metric	Value
f1_macro_dev_0.5	0.6987385936661299
f1_weighted_dev_0.5	0.8364080971969677
accuracy_dev_0.5	0.821826280623608
f1_macro_dev_best_global	0.7378687454926337
f1_weighted_dev_best_global	0.8789430301020366
accuracy_dev_best_global	0.8864142538975501
f1_macro_dev_best_by_lang	0.6967153002933457
f1_weighted_dev_best_by_lang	0.8478649653923616
accuracy_dev_best_by_lang	0.844097995545657
default_threshold	0.5
best_threshold_global	0.6
thresholds_by_lang	{"en": 0.45000000000000007, "it": 0.6, "es": 0.6}

Thresholds

Default: 0.5
Best global: 0.6
Best by language: { "en": 0.45000000000000007, "it": 0.6, "es": 0.6 }

Detailed evaluation

Classification report @ 0.5

text

              precision    recall  f1-score   support

 no-recl (0)     0.9345    0.8519    0.8913       385
    recl (1)     0.4184    0.6406    0.5062        64

    accuracy                         0.8218       449
   macro avg     0.6764    0.7463    0.6987       449
weighted avg     0.8609    0.8218    0.8364       449

Classification report @ best global threshold (t=0.60)

text

              precision    recall  f1-score   support

 no-recl (0)     0.9154    0.9558    0.9352       385
    recl (1)     0.6383    0.4688    0.5405        64

    accuracy                         0.8864       449
   macro avg     0.7769    0.7123    0.7379       449
weighted avg     0.8759    0.8864    0.8789       449

Classification report @ best per-language thresholds

text

              precision    recall  f1-score   support

 no-recl (0)     0.9178    0.8987    0.9081       385
    recl (1)     0.4583    0.5156    0.4853        64

    accuracy                         0.8441       449
   macro avg     0.6881    0.7072    0.6967       449
weighted avg     0.8523    0.8441    0.8479       449

Per-language metrics (at best-by-lang)

lang	n	acc	f1_macro	f1_weighted	prec_macro	rec_macro	prec_weighted	rec_weighted
en	154	0.7727	0.5076	0.8079	0.5139	0.5267	0.8520	0.7727
it	163	0.9018	0.8224	0.8960	0.8702	0.7913	0.8975	0.9018
es	132	0.8561	0.7257	0.8575	0.7214	0.7304	0.8591	0.8561

Data

Train/Dev: private multilingual splits with ~15% stratified Dev (by (lang,label)).
Source: merged EN/IT/ES data with bios retained (ignored if unused by model).

Usage

python

from transformers import AutoTokenizer, AutoModelForSequenceClassification, AutoConfig
import torch, numpy as np

repo = "SimoneAstarita/trilingual-no-bio-20251014-t07"
tok = AutoTokenizer.from_pretrained(repo)
cfg = AutoConfig.from_pretrained(repo)
model = AutoModelForSequenceClassification.from_pretrained(repo)

texts = ["example text ..."]
langs = ["en"]

mode = "best_global"  # or "0.5", "by_lang"

enc = tok(texts, truncation=True, padding=True, max_length=256, return_tensors="pt")
with torch.no_grad():
    logits = model(**enc).logits
probs = torch.softmax(logits, dim=-1)[:, 1].cpu().numpy()

if mode == "0.5":
    th = 0.5
    preds = (probs >= th).astype(int)
elif mode == "best_global":
    th = getattr(cfg, "best_threshold_global", 0.5)
    preds = (probs >= th).astype(int)
elif mode == "by_lang":
    th_by_lang = getattr(cfg, "thresholds_by_lang", {})
    preds = np.zeros_like(probs, dtype=int)
    for lg in np.unique(langs):
        t = th_by_lang.get(lg, getattr(cfg, "best_threshold_global", 0.5))
        preds[np.array(langs) == lg] = (probs[np.array(langs) == lg] >= t).astype(int)
print(list(zip(texts, preds, probs)))

Additional files

reports.json: all metrics (macro/weighted/accuracy) for @0.5, @best_global, and @best_by_lang. config.json: stores thresholds: default_threshold, best_threshold_global, thresholds_by_lang. postprocessing.json: duplicate threshold info for external tools.