feature: Primeiro upload do modelo

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+predict_wiki_section_disease_max_seq512_ts_score_lt.txt filter=lfs diff=lfs merge=lfs -text

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[BOS]": 29794
+}

all_results.json

@@ -0,0 +1,66 @@
+{
+    "EOP_f1": 0.9807168308530706,
+    "EOP_number": 4774,
+    "EOP_precision": 0.9813338926174496,
+    "EOP_recall": 0.9801005446166736,
+    "accuracy": 0.9953893956099028,
+    "epoch": 5.0,
+    "eval_EOP_f1": 0.9836132732486685,
+    "eval_EOP_number": 2446,
+    "eval_EOP_precision": 0.985632183908046,
+    "eval_EOP_recall": 0.9816026165167621,
+    "eval_da_EOP_f1": 0.9373584905660378,
+    "eval_da_EOP_number": 2736,
+    "eval_da_EOP_precision": 0.968798751950078,
+    "eval_da_EOP_recall": 0.9078947368421053,
+    "eval_da_overall_accuracy": 0.985526201063737,
+    "eval_da_overall_f1": 0.9373584905660378,
+    "eval_da_overall_precision": 0.968798751950078,
+    "eval_da_overall_recall": 0.9078947368421053,
+    "eval_loss": NaN,
+    "eval_overall_accuracy": 0.9961055398695355,
+    "eval_overall_f1": 0.9836132732486685,
+    "eval_overall_precision": 0.985632183908046,
+    "eval_overall_recall": 0.9816026165167621,
+    "eval_runtime": 345.7738,
+    "eval_samples": 1527,
+    "eval_samples_per_second": 4.416,
+    "eval_steps_per_second": 1.105,
+    "overall_accuracy": 0.9953893956099028,
+    "overall_f1": 0.9807168308530706,
+    "overall_precision": 0.9813338926174496,
+    "overall_recall": 0.9801005446166736,
+    "predict_EOP_f1": 0.9807168308530706,
+    "predict_EOP_number": 4774,
+    "predict_EOP_precision": 0.9813338926174496,
+    "predict_EOP_recall": 0.9801005446166736,
+    "predict_da_EOP_f1": 0.9401974066189279,
+    "predict_da_EOP_number": 5281,
+    "predict_da_EOP_precision": 0.9614090639224223,
+    "predict_da_EOP_recall": 0.9199015338004166,
+    "predict_da_overall_accuracy": 0.9861948800428897,
+    "predict_da_overall_f1": 0.9401974066189279,
+    "predict_da_overall_precision": 0.9614090639224223,
+    "predict_da_overall_recall": 0.9199015338004166,
+    "predict_examples": 718,
+    "predict_loss": NaN,
+    "predict_overall_accuracy": 0.9953893956099028,
+    "predict_overall_f1": 0.9807168308530706,
+    "predict_overall_precision": 0.9813338926174496,
+    "predict_overall_recall": 0.9801005446166736,
+    "predict_runtime": 671.1845,
+    "predict_samples_per_second": 4.349,
+    "predict_steps_per_second": 1.088,
+    "threshold_0.5_example_level_1-pk": 0.986,
+    "threshold_0.5_example_level_1-wd": 0.9829,
+    "threshold_0.5_example_level_f1": 0.9807,
+    "threshold_0.5_example_level_pk": 0.014000000000000012,
+    "threshold_0.5_example_level_precision": 0.9813,
+    "threshold_0.5_example_level_recall": 0.9801,
+    "threshold_0.5_example_level_wd": 0.017100000000000004,
+    "train_loss": 0.3830805900915345,
+    "train_runtime": 38367.8373,
+    "train_samples": 10720,
+    "train_samples_per_second": 1.397,
+    "train_steps_per_second": 0.349
+}

config.json

@@ -0,0 +1,67 @@
+{
+  "_name_or_path": "neuralmind/bert-large-portuguese-cased",
+  "architectures": [
+    "BertWithDAForSentenceLabelingTopicSegmentation"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "cache_dir": null,
+  "cl_anchor_level": "eop_list",
+  "cl_loss_weight": 0.5,
+  "cl_negative_k": 3,
+  "cl_positive_k": 1,
+  "cl_temp": 0.1,
+  "classifier_dropout": null,
+  "config_name": null,
+  "directionality": "bidi",
+  "do_cssl": true,
+  "do_da_ts": true,
+  "do_tssp": true,
+  "finetuning_task": "topic_segment",
+  "focal_loss_gamma": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "B-EOP",
+    "1": "O"
+  },
+  "ignore_mismatched_sizes": false,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "B-EOP": 0,
+    "O": 1
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_name_or_path": "neuralmind/bert-large-portuguese-cased",
+  "model_revision": "main",
+  "model_type": "bert",
+  "num_attention_heads": 16,
+  "num_gpu": 1,
+  "num_hidden_layers": 24,
+  "num_topic_labels": 0,
+  "num_tssp_labels": 3,
+  "output_past": true,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "sentence_pooler_type": null,
+  "tokenizer_name": null,
+  "torch_dtype": "float32",
+  "transformers_version": "4.26.0",
+  "ts_loss_weight": 1.0,
+  "ts_score_predictor": "lt",
+  "ts_score_predictor_cos_temp": 1,
+  "tssp_ablation": "none",
+  "tssp_loss_weight": 1.0,
+  "type_vocab_size": 2,
+  "use_auth_token": false,
+  "use_cache": true,
+  "vocab_size": 29795,
+  "weight_label_zero": 0.5
+}

eval_results.json

@@ -0,0 +1,24 @@
+{
+    "epoch": 5.0,
+    "eval_EOP_f1": 0.9836132732486685,
+    "eval_EOP_number": 2446,
+    "eval_EOP_precision": 0.985632183908046,
+    "eval_EOP_recall": 0.9816026165167621,
+    "eval_da_EOP_f1": 0.9373584905660378,
+    "eval_da_EOP_number": 2736,
+    "eval_da_EOP_precision": 0.968798751950078,
+    "eval_da_EOP_recall": 0.9078947368421053,
+    "eval_da_overall_accuracy": 0.985526201063737,
+    "eval_da_overall_f1": 0.9373584905660378,
+    "eval_da_overall_precision": 0.968798751950078,
+    "eval_da_overall_recall": 0.9078947368421053,
+    "eval_loss": NaN,
+    "eval_overall_accuracy": 0.9961055398695355,
+    "eval_overall_f1": 0.9836132732486685,
+    "eval_overall_precision": 0.985632183908046,
+    "eval_overall_recall": 0.9816026165167621,
+    "eval_runtime": 345.7738,
+    "eval_samples": 1527,
+    "eval_samples_per_second": 4.416,
+    "eval_steps_per_second": 1.105
+}

example_level_predict_wiki_section_disease_max_seq512_ts_score_lt_results.json

@@ -0,0 +1,19 @@
+{
+    "EOP_f1": 0.9807168308530706,
+    "EOP_number": 4774,
+    "EOP_precision": 0.9813338926174496,
+    "EOP_recall": 0.9801005446166736,
+    "accuracy": 0.9953893956099028,
+    "overall_accuracy": 0.9953893956099028,
+    "overall_f1": 0.9807168308530706,
+    "overall_precision": 0.9813338926174496,
+    "overall_recall": 0.9801005446166736,
+    "predict_examples": 718,
+    "threshold_0.5_example_level_1-pk": 0.986,
+    "threshold_0.5_example_level_1-wd": 0.9829,
+    "threshold_0.5_example_level_f1": 0.9807,
+    "threshold_0.5_example_level_pk": 0.014000000000000012,
+    "threshold_0.5_example_level_precision": 0.9813,
+    "threshold_0.5_example_level_recall": 0.9801,
+    "threshold_0.5_example_level_wd": 0.017100000000000004
+}

example_level_predict_wiki_section_disease_max_seq512_ts_score_lt_results_str_metric.txt

@@ -0,0 +1,4 @@
+p / r / f / pk / wd
+threshold_0.5_example_level_metric
+98.13 / 98.01 / 98.07 / 1.40 / 1.71
+

predict_wiki_section_disease_max_seq512_ts_score_lt.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:327a37568a8b57c9483f0f8a3a789c1a09360e483bcee0b4d9bf3c51e85c5da8
+size 15636892

predict_wiki_section_disease_max_seq512_ts_score_lt_results.json

@@ -0,0 +1,22 @@
+{
+    "predict_EOP_f1": 0.9807168308530706,
+    "predict_EOP_number": 4774,
+    "predict_EOP_precision": 0.9813338926174496,
+    "predict_EOP_recall": 0.9801005446166736,
+    "predict_da_EOP_f1": 0.9401974066189279,
+    "predict_da_EOP_number": 5281,
+    "predict_da_EOP_precision": 0.9614090639224223,
+    "predict_da_EOP_recall": 0.9199015338004166,
+    "predict_da_overall_accuracy": 0.9861948800428897,
+    "predict_da_overall_f1": 0.9401974066189279,
+    "predict_da_overall_precision": 0.9614090639224223,
+    "predict_da_overall_recall": 0.9199015338004166,
+    "predict_loss": NaN,
+    "predict_overall_accuracy": 0.9953893956099028,
+    "predict_overall_f1": 0.9807168308530706,
+    "predict_overall_precision": 0.9813338926174496,
+    "predict_overall_recall": 0.9801005446166736,
+    "predict_runtime": 671.1845,
+    "predict_samples_per_second": 4.349,
+    "predict_steps_per_second": 1.088
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2148fe30fc1ed348407799ba9f54bd9a29ef4ab2f99253d4ccbb85bac056b3ec
+size 1337746921

special_tokens_map.json

@@ -0,0 +1,8 @@
+{
+  "bos_token": "[BOS]",
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "name_or_path": "neuralmind/bert-large-portuguese-cased",
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": "/home/ext_maiara_frodrigues2000_gmail_/.cache/huggingface/hub/models--neuralmind--bert-large-portuguese-cased/snapshots/aa302f6ea73b759f7df9cad58bd272127b67ec28/special_tokens_map.json",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 5.0,
+    "train_loss": 0.3830805900915345,
+    "train_runtime": 38367.8373,
+    "train_samples": 10720,
+    "train_samples_per_second": 1.397,
+    "train_steps_per_second": 0.349
+}

trainer_state.json

@@ -0,0 +1,175 @@
+{
+  "best_metric": 0.9836132732486685,
+  "best_model_checkpoint": "./output/neuralmind/bert-large-portuguese-cased-finetune-wiki_section_disease/seed42-seq512-lr5e-05-epoch5-bs4-ts1.0-tssp1.0-cl0.5-2024-05-30_21:01:41/checkpoint-8040",
+  "epoch": 5.0,
+  "global_step": 13400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 4e-05,
+      "loss": 0.5631,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0,
+      "eval_EOP_f1": 0.9767820773930753,
+      "eval_EOP_number": 2446,
+      "eval_EOP_precision": 0.9732142857142857,
+      "eval_EOP_recall": 0.9803761242845462,
+      "eval_da_EOP_f1": 0.9276059988890947,
+      "eval_da_EOP_number": 2736,
+      "eval_da_EOP_precision": 0.9399624765478424,
+      "eval_da_EOP_recall": 0.9155701754385965,
+      "eval_da_overall_accuracy": 0.9829540500479553,
+      "eval_da_overall_f1": 0.9276059988890947,
+      "eval_da_overall_precision": 0.9399624765478424,
+      "eval_da_overall_recall": 0.9155701754385965,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9944503943140882,
+      "eval_overall_f1": 0.9767820773930753,
+      "eval_overall_precision": 0.9732142857142857,
+      "eval_overall_recall": 0.9803761242845462,
+      "eval_runtime": 350.1554,
+      "eval_samples_per_second": 4.361,
+      "eval_steps_per_second": 1.091,
+      "step": 2680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3e-05,
+      "loss": 0.4083,
+      "step": 5360
+    },
+    {
+      "epoch": 2.0,
+      "eval_EOP_f1": 0.9791751735402207,
+      "eval_EOP_number": 2446,
+      "eval_EOP_precision": 0.9779771615008157,
+      "eval_EOP_recall": 0.9803761242845462,
+      "eval_da_EOP_f1": 0.9395747855277882,
+      "eval_da_EOP_number": 2736,
+      "eval_da_EOP_precision": 0.9592536176694593,
+      "eval_da_EOP_recall": 0.9206871345029239,
+      "eval_da_overall_accuracy": 0.985874967303165,
+      "eval_da_overall_f1": 0.9395747855277882,
+      "eval_da_overall_precision": 0.9592536176694593,
+      "eval_da_overall_recall": 0.9206871345029239,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9950345633336579,
+      "eval_overall_f1": 0.9791751735402207,
+      "eval_overall_precision": 0.9779771615008157,
+      "eval_overall_recall": 0.9803761242845462,
+      "eval_runtime": 351.0792,
+      "eval_samples_per_second": 4.349,
+      "eval_steps_per_second": 1.088,
+      "step": 5360
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2e-05,
+      "loss": 0.3595,
+      "step": 8040
+    },
+    {
+      "epoch": 3.0,
+      "eval_EOP_f1": 0.9836132732486685,
+      "eval_EOP_number": 2446,
+      "eval_EOP_precision": 0.985632183908046,
+      "eval_EOP_recall": 0.9816026165167621,
+      "eval_da_EOP_f1": 0.9373584905660378,
+      "eval_da_EOP_number": 2736,
+      "eval_da_EOP_precision": 0.968798751950078,
+      "eval_da_EOP_recall": 0.9078947368421053,
+      "eval_da_overall_accuracy": 0.985526201063737,
+      "eval_da_overall_f1": 0.9373584905660378,
+      "eval_da_overall_precision": 0.968798751950078,
+      "eval_da_overall_recall": 0.9078947368421053,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9961055398695355,
+      "eval_overall_f1": 0.9836132732486685,
+      "eval_overall_precision": 0.985632183908046,
+      "eval_overall_recall": 0.9816026165167621,
+      "eval_runtime": 350.0594,
+      "eval_samples_per_second": 4.362,
+      "eval_steps_per_second": 1.091,
+      "step": 8040
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1e-05,
+      "loss": 0.315,
+      "step": 10720
+    },
+    {
+      "epoch": 4.0,
+      "eval_EOP_f1": 0.981170691772411,
+      "eval_EOP_number": 2446,
+      "eval_EOP_precision": 0.9823770491803279,
+      "eval_EOP_recall": 0.9799672935404743,
+      "eval_da_EOP_f1": 0.9371944339977436,
+      "eval_da_EOP_number": 2736,
+      "eval_da_EOP_precision": 0.965143299767622,
+      "eval_da_EOP_recall": 0.9108187134502924,
+      "eval_da_overall_accuracy": 0.98543900950388,
+      "eval_da_overall_f1": 0.9371944339977436,
+      "eval_da_overall_precision": 0.965143299767622,
+      "eval_da_overall_recall": 0.9108187134502924,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.995521370849966,
+      "eval_overall_f1": 0.981170691772411,
+      "eval_overall_precision": 0.9823770491803279,
+      "eval_overall_recall": 0.9799672935404743,
+      "eval_runtime": 351.0283,
+      "eval_samples_per_second": 4.35,
+      "eval_steps_per_second": 1.088,
+      "step": 10720
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 0.2695,
+      "step": 13400
+    },
+    {
+      "epoch": 5.0,
+      "eval_EOP_f1": 0.9823553549446039,
+      "eval_EOP_number": 2446,
+      "eval_EOP_precision": 0.985996705107084,
+      "eval_EOP_recall": 0.9787408013082584,
+      "eval_da_EOP_f1": 0.9391206313416008,
+      "eval_da_EOP_number": 2736,
+      "eval_da_EOP_precision": 0.9663573085846868,
+      "eval_da_EOP_recall": 0.9133771929824561,
+      "eval_da_overall_accuracy": 0.985874967303165,
+      "eval_da_overall_f1": 0.9391206313416008,
+      "eval_da_overall_precision": 0.9663573085846868,
+      "eval_da_overall_recall": 0.9133771929824561,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9958134553597507,
+      "eval_overall_f1": 0.9823553549446039,
+      "eval_overall_precision": 0.985996705107084,
+      "eval_overall_recall": 0.9787408013082584,
+      "eval_runtime": 351.0391,
+      "eval_samples_per_second": 4.35,
+      "eval_steps_per_second": 1.088,
+      "step": 13400
+    },
+    {
+      "epoch": 5.0,
+      "step": 13400,
+      "total_flos": 9.99040548028416e+16,
+      "train_loss": 0.3830805900915345,
+      "train_runtime": 38367.8373,
+      "train_samples_per_second": 1.397,
+      "train_steps_per_second": 0.349
+    }
+  ],
+  "max_steps": 13400,
+  "num_train_epochs": 5,
+  "total_flos": 9.99040548028416e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d29eb7ffa01eb118532a5ea29c2e950e2eab938a900a1559b44b5af8e1f9682d
+size 3759