Delete src/bigru_t/model/gru_hierarchy.py

Delete src/bigru_t/model/bigru4.py

V6.5-attn-v3: add upload_report.json (created after main upload commit)

V6.5-attn-v2 cleanup: remove legacy scripts/upload_v6_5_7ds_attn_v2.py

V6.5-attn-v2 cleanup: remove legacy scripts/train_v6_5_7ds_attn_v2.py

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_upload_report.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_user_questions.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_attention_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_w8a8_compression.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_reasoning_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_ewc_w8a8_benchmark.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_script_activity.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_module_analysis.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_training_metrics.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_v2_report.json

V6.5-attn-v2 cleanup: remove legacy scripts/upload_v6_5_7ds_attn.py

V6.5-attn-v2 cleanup: remove legacy scripts/train_v6_5_7ds_attn.py

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_upload_report.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_verification.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_attention_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_w8a8_compression.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_reasoning_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_ewc_w8a8_benchmark.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_script_activity.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_module_analysis.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_training_metrics.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_report.json

V6.5-attn-v2 cleanup: remove legacy scripts/upload_v6_5_7ds.py

V6.5-attn-v2 cleanup: remove legacy scripts/train_v6_5_7ds.py

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_upload_report.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_w8a8_compression.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_reasoning_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_ewc_w8a8_benchmark.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_script_activity.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_module_analysis.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_training_metrics.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_report.json

V6.5-attn-v3: CORREÇÃO MATEMÁTICA do predict() — rótulos dinâmicos extraídos do label_registry (substitui 'gato'/'cachorro' fixos). NOVO: punish_during_inference() — mirror do protocolo de punição do treinamento (1ª → activate_hypothesis, 2ª → set_ewc_reference). META ≥6000 samples, attention ACTIVE_AND_FUNCTIONAL, SmoothQuant W8A8, 864 neurons, MTP K=6, 3 perguntas reais sem ajuda (Luva de Pedreiro Távila, Lula reserva valor, Amazonas força-tarefa vítimas) + punição na inferência via camada de hipótese. Removed legacy v2 reports/scripts (superseded by v3).

V6.5-fix: remove v6_5_7ds_attn_v2_model_states.pt (predict() returned hardcoded gato/cachorro labels instead of dataset-derived dynamic labels)

V6.5-attn-v2: add upload_report.json (post-upload artifact)

V6.5-attn-v2 cleanup: remove legacy scripts/upload_v6_5_7ds_attn.py

V6.5-attn-v2 cleanup: remove legacy scripts/train_v6_5_7ds_attn.py

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_upload_report.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_verification.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_attention_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_w8a8_compression.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_reasoning_eval.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_ewc_w8a8_benchmark.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_script_activity.json

V6.5-attn-v2 cleanup: remove legacy v6_5_7ds_attn_module_analysis.json