Hyperparameter	Value
LANGUAGES	en-es-it
LR	2e-05
EPOCHS	3
MAX_LENGTH	256
USE_BIO	False
USE_LANG_TOKEN	False
GATED_BIO	False
FOCAL_LOSS	True
FOCAL_GAMMA	1.5
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	0.7168468468468467
f1_weighted_dev_0.5	0.8525532213728205
accuracy_dev_0.5	0.844097995545657
f1_macro_dev_best_global	0.7173094314136246
f1_weighted_dev_best_global	0.8694483657963139
accuracy_dev_best_global	0.8775055679287305
f1_macro_dev_best_by_lang	0.703098816091017
f1_weighted_dev_best_by_lang	0.8458669301901763
accuracy_dev_best_by_lang	0.8374164810690423
default_threshold	0.5
best_threshold_global	0.6
thresholds_by_lang	{"en": 0.4, "it": 0.65, "es": 0.5}

Thresholds

Default: 0.5
Best global: 0.6
Best by language: { "en": 0.4, "it": 0.65, "es": 0.5 }

Detailed evaluation

Classification report @ 0.5

text

              precision    recall  f1-score   support

 no-recl (0)     0.9315    0.8831    0.9067       385
    recl (1)     0.4643    0.6094    0.5270        64

    accuracy                         0.8441       449
   macro avg     0.6979    0.7462    0.7168       449
weighted avg     0.8649    0.8441    0.8526       449

Classification report @ best global threshold (t=0.60)

text

              precision    recall  f1-score   support

 no-recl (0)     0.9104    0.9506    0.9301       385
    recl (1)     0.5957    0.4375    0.5045        64

    accuracy                         0.8775       449
   macro avg     0.7531    0.6941    0.7173       449
weighted avg     0.8656    0.8775    0.8694       449

Classification report @ best per-language thresholds

text

              precision    recall  f1-score   support

 no-recl (0)     0.9262    0.8805    0.9028       385
    recl (1)     0.4458    0.5781    0.5034        64

    accuracy                         0.8374       449
   macro avg     0.6860    0.7293    0.7031       449
weighted avg     0.8577    0.8374    0.8459       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	0.7662	0.5400	0.8081	0.5427	0.5930	0.8678	0.7662
it	163	0.9141	0.8398	0.9074	0.9101	0.7989	0.9135	0.9141
es	132	0.8258	0.7118	0.8381	0.6902	0.7536	0.8587	0.8258

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-t00"
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.