Downloads · 30 days
9
29% of all-time downloads
JoaoReiz/ner-pt-f1-v1-qwen35-2b-specific-paramopama-seed42
ner-pt-f1-v1-qwen35-2b-specific-paramopama-seed42 is a token classification model from JoaoReiz. Use it when you need labels on individual words, such as names. It is set up for peft.
LoRA adapter for generative named-entity recognition in Portuguese. This artifact is one run of the ner-pt-generative-2026-f1-v1 research matrix.
Downloads · 30 days
9
29% of all-time downloads
All-time downloads
31
Public
Repo size
87.3 MB
Likes
0
Public
Click a slice to open those files.
.safetensors67.3 MB · 73%
From the Hugging Face model README
LoRA adapter for generative named-entity recognition in Portuguese. This artifact is
one run of the ner-pt-generative-2026-f1-v1 research matrix.
Qwen/Qwen3.5-2B@15852e8c16360a2fea060d615a32b45270f8a8fcspecific42labels_and_tokens constrained JSON| Dataset | Precision | Recall | F1 | Structural validity |
|---|---|---|---|---|
| paramopama | 0.8921 | 0.8963 | 0.8942 | 0.9992 |
Results describe only the listed frozen splits and this single seed. Cross-seed uncertainty requires completion of the full three-seed matrix. Results must not be interpreted as evidence of general performance outside these corpora.
The repository root contains the LoRA adapter. research/ contains frozen
predictions, normal and decontaminated metrics, schema ablations, inference
contracts, the run manifest, hashes, and checkpoint-selection evidence.
Load this adapter with PEFT on the exact base-model revision above. Reproduce the
documented protocol with BF16-capable hardware and the versions recorded in
research/manifest.json.
Intended for Portuguese NER research, evaluation, and controlled downstream experimentation. Generated spans can be structurally valid yet semantically wrong. Corpus annotation schemes differ, text overlap may affect estimates, and this model has not been validated for high-stakes or autonomous decisions. Users must review dataset licenses and evaluate errors in their target domain.