feature: Faz upload de arquivos 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_city_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": 0.8759355463590737,
+    "EOP_number": 23325,
+    "EOP_precision": 0.9001493009998642,
+    "EOP_recall": 0.8529903536977492,
+    "accuracy": 0.9746062069711279,
+    "epoch": 5.0,
+    "eval_EOP_f1": 0.8800495553294101,
+    "eval_EOP_number": 11519,
+    "eval_EOP_precision": 0.8974011911207364,
+    "eval_EOP_recall": 0.8633561941140724,
+    "eval_da_EOP_f1": 0.9105588594383838,
+    "eval_da_EOP_number": 13045,
+    "eval_da_EOP_precision": 0.9363356552729629,
+    "eval_da_EOP_recall": 0.8861632809505557,
+    "eval_da_overall_accuracy": 0.9804166738526809,
+    "eval_da_overall_f1": 0.9105588594383838,
+    "eval_da_overall_precision": 0.9363356552729629,
+    "eval_da_overall_recall": 0.8861632809505557,
+    "eval_loss": NaN,
+    "eval_overall_accuracy": 0.9758097617560453,
+    "eval_overall_f1": 0.8800495553294101,
+    "eval_overall_precision": 0.8974011911207364,
+    "eval_overall_recall": 0.8633561941140724,
+    "eval_runtime": 1876.6424,
+    "eval_samples": 7725,
+    "eval_samples_per_second": 4.116,
+    "eval_steps_per_second": 1.029,
+    "overall_accuracy": 0.9746062069711279,
+    "overall_f1": 0.8759355463590737,
+    "overall_precision": 0.9001493009998642,
+    "overall_recall": 0.8529903536977492,
+    "predict_EOP_f1": 0.8759355463590737,
+    "predict_EOP_number": 23325,
+    "predict_EOP_precision": 0.9001493009998642,
+    "predict_EOP_recall": 0.8529903536977492,
+    "predict_da_EOP_f1": 0.9084055933130223,
+    "predict_da_EOP_number": 26294,
+    "predict_da_EOP_precision": 0.9336411079887595,
+    "predict_da_EOP_recall": 0.8844983646459268,
+    "predict_da_overall_accuracy": 0.9796302167709767,
+    "predict_da_overall_f1": 0.9084055933130223,
+    "predict_da_overall_precision": 0.9336411079887595,
+    "predict_da_overall_recall": 0.8844983646459268,
+    "predict_examples": 3906,
+    "predict_loss": NaN,
+    "predict_overall_accuracy": 0.9746062069711279,
+    "predict_overall_f1": 0.8759355463590737,
+    "predict_overall_precision": 0.9001493009998642,
+    "predict_overall_recall": 0.8529903536977492,
+    "predict_runtime": 3721.7868,
+    "predict_samples_per_second": 4.113,
+    "predict_steps_per_second": 1.028,
+    "threshold_0.5_example_level_1-pk": 0.931,
+    "threshold_0.5_example_level_1-wd": 0.9162,
+    "threshold_0.5_example_level_f1": 0.8759,
+    "threshold_0.5_example_level_pk": 0.06899999999999995,
+    "threshold_0.5_example_level_precision": 0.9001,
+    "threshold_0.5_example_level_recall": 0.853,
+    "threshold_0.5_example_level_wd": 0.08379999999999999,
+    "train_loss": 0.3591941669170673,
+    "train_runtime": 192526.8268,
+    "train_samples": 52001,
+    "train_samples_per_second": 1.35,
+    "train_steps_per_second": 0.338
+}

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": 0.8800495553294101,
+    "eval_EOP_number": 11519,
+    "eval_EOP_precision": 0.8974011911207364,
+    "eval_EOP_recall": 0.8633561941140724,
+    "eval_da_EOP_f1": 0.9105588594383838,
+    "eval_da_EOP_number": 13045,
+    "eval_da_EOP_precision": 0.9363356552729629,
+    "eval_da_EOP_recall": 0.8861632809505557,
+    "eval_da_overall_accuracy": 0.9804166738526809,
+    "eval_da_overall_f1": 0.9105588594383838,
+    "eval_da_overall_precision": 0.9363356552729629,
+    "eval_da_overall_recall": 0.8861632809505557,
+    "eval_loss": NaN,
+    "eval_overall_accuracy": 0.9758097617560453,
+    "eval_overall_f1": 0.8800495553294101,
+    "eval_overall_precision": 0.8974011911207364,
+    "eval_overall_recall": 0.8633561941140724,
+    "eval_runtime": 1876.6424,
+    "eval_samples": 7725,
+    "eval_samples_per_second": 4.116,
+    "eval_steps_per_second": 1.029
+}

example_level_predict_wiki_section_city_max_seq512_ts_score_lt_results.json

@@ -0,0 +1,19 @@
+{
+    "EOP_f1": 0.8759355463590737,
+    "EOP_number": 23325,
+    "EOP_precision": 0.9001493009998642,
+    "EOP_recall": 0.8529903536977492,
+    "accuracy": 0.9746062069711279,
+    "overall_accuracy": 0.9746062069711279,
+    "overall_f1": 0.8759355463590737,
+    "overall_precision": 0.9001493009998642,
+    "overall_recall": 0.8529903536977492,
+    "predict_examples": 3906,
+    "threshold_0.5_example_level_1-pk": 0.931,
+    "threshold_0.5_example_level_1-wd": 0.9162,
+    "threshold_0.5_example_level_f1": 0.8759,
+    "threshold_0.5_example_level_pk": 0.06899999999999995,
+    "threshold_0.5_example_level_precision": 0.9001,
+    "threshold_0.5_example_level_recall": 0.853,
+    "threshold_0.5_example_level_wd": 0.08379999999999999
+}

example_level_predict_wiki_section_city_max_seq512_ts_score_lt_results_str_metric.txt

@@ -0,0 +1,4 @@
+p / r / f / pk / wd
+threshold_0.5_example_level_metric
+90.01 / 85.30 / 87.59 / 6.90 / 8.38
+

predict_wiki_section_city_max_seq512_ts_score_lt.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f72851f15a06cf1afc2443fe4723afc7ccfcb14ec2848faa3a67c0235471eff
+size 78329487

predict_wiki_section_city_max_seq512_ts_score_lt_results.json

@@ -0,0 +1,22 @@
+{
+    "predict_EOP_f1": 0.8759355463590737,
+    "predict_EOP_number": 23325,
+    "predict_EOP_precision": 0.9001493009998642,
+    "predict_EOP_recall": 0.8529903536977492,
+    "predict_da_EOP_f1": 0.9084055933130223,
+    "predict_da_EOP_number": 26294,
+    "predict_da_EOP_precision": 0.9336411079887595,
+    "predict_da_EOP_recall": 0.8844983646459268,
+    "predict_da_overall_accuracy": 0.9796302167709767,
+    "predict_da_overall_f1": 0.9084055933130223,
+    "predict_da_overall_precision": 0.9336411079887595,
+    "predict_da_overall_recall": 0.8844983646459268,
+    "predict_loss": NaN,
+    "predict_overall_accuracy": 0.9746062069711279,
+    "predict_overall_f1": 0.8759355463590737,
+    "predict_overall_precision": 0.9001493009998642,
+    "predict_overall_recall": 0.8529903536977492,
+    "predict_runtime": 3721.7868,
+    "predict_samples_per_second": 4.113,
+    "predict_steps_per_second": 1.028
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7844e4dc3c4dcfb7fd014765f4a9aecc011443389ff63f3b0114d8b85eb39977
+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_luti4497_gmail_com/.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": 0.3591941669170673,
+    "train_runtime": 192526.8268,
+    "train_samples": 52001,
+    "train_samples_per_second": 1.35,
+    "train_steps_per_second": 0.338
+}

trainer_state.json

@@ -0,0 +1,175 @@
+{
+  "best_metric": 0.8800495553294101,
+  "best_model_checkpoint": "./output/neuralmind/bert-large-portuguese-cased-finetune-wiki_section_city/seed42-seq512-lr5e-05-epoch5-bs4-ts1.0-tssp1.0-cl0.5-2024-07-03_20:57:35/checkpoint-52000",
+  "epoch": 4.999961539940772,
+  "global_step": 65000,
+  "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": 0.5927,
+      "step": 13000
+    },
+    {
+      "epoch": 1.0,
+      "eval_EOP_f1": 0.8637108561810692,
+      "eval_EOP_number": 11519,
+      "eval_EOP_precision": 0.8817145957677699,
+      "eval_EOP_recall": 0.8464276412883063,
+      "eval_da_EOP_f1": 0.8848200312989045,
+      "eval_da_EOP_number": 13045,
+      "eval_da_EOP_precision": 0.9035557331202557,
+      "eval_da_EOP_recall": 0.8668455346876198,
+      "eval_da_overall_accuracy": 0.974613248710829,
+      "eval_da_overall_f1": 0.8848200312989045,
+      "eval_da_overall_precision": 0.9035557331202557,
+      "eval_da_overall_recall": 0.8668455346876198,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9725439457481931,
+      "eval_overall_f1": 0.8637108561810692,
+      "eval_overall_precision": 0.8817145957677699,
+      "eval_overall_recall": 0.8464276412883063,
+      "eval_runtime": 1878.3511,
+      "eval_samples_per_second": 4.113,
+      "eval_steps_per_second": 1.029,
+      "step": 13000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3e-05,
+      "loss": 0.4209,
+      "step": 26000
+    },
+    {
+      "epoch": 2.0,
+      "eval_EOP_f1": 0.8753550015777848,
+      "eval_EOP_number": 11519,
+      "eval_EOP_precision": 0.9104463615903976,
+      "eval_EOP_recall": 0.8428683045403247,
+      "eval_da_EOP_f1": 0.9015245488375816,
+      "eval_da_EOP_number": 13045,
+      "eval_da_EOP_precision": 0.94299347061778,
+      "eval_da_EOP_recall": 0.8635492525871982,
+      "eval_da_overall_accuracy": 0.9787782625942086,
+      "eval_da_overall_f1": 0.9015245488375816,
+      "eval_da_overall_precision": 0.94299347061778,
+      "eval_da_overall_recall": 0.8635492525871982,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9753279200499688,
+      "eval_overall_f1": 0.8753550015777848,
+      "eval_overall_precision": 0.9104463615903976,
+      "eval_overall_recall": 0.8428683045403247,
+      "eval_runtime": 1878.6106,
+      "eval_samples_per_second": 4.112,
+      "eval_steps_per_second": 1.028,
+      "step": 26000
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2e-05,
+      "loss": 0.3373,
+      "step": 39000
+    },
+    {
+      "epoch": 3.0,
+      "eval_EOP_f1": 0.879935463631067,
+      "eval_EOP_number": 11519,
+      "eval_EOP_precision": 0.9094867518992032,
+      "eval_EOP_recall": 0.8522441184130567,
+      "eval_da_EOP_f1": 0.9077958053027304,
+      "eval_da_EOP_number": 13045,
+      "eval_da_EOP_precision": 0.938241308793456,
+      "eval_da_EOP_recall": 0.8792640858566501,
+      "eval_da_overall_accuracy": 0.9799079040408396,
+      "eval_da_overall_f1": 0.9077958053027304,
+      "eval_da_overall_precision": 0.938241308793456,
+      "eval_da_overall_recall": 0.8792640858566501,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9760952975818685,
+      "eval_overall_f1": 0.879935463631067,
+      "eval_overall_precision": 0.9094867518992032,
+      "eval_overall_recall": 0.8522441184130567,
+      "eval_runtime": 1879.4394,
+      "eval_samples_per_second": 4.11,
+      "eval_steps_per_second": 1.028,
+      "step": 39000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1e-05,
+      "loss": 0.259,
+      "step": 52000
+    },
+    {
+      "epoch": 4.0,
+      "eval_EOP_f1": 0.8800495553294101,
+      "eval_EOP_number": 11519,
+      "eval_EOP_precision": 0.8974011911207364,
+      "eval_EOP_recall": 0.8633561941140724,
+      "eval_da_EOP_f1": 0.9105588594383838,
+      "eval_da_EOP_number": 13045,
+      "eval_da_EOP_precision": 0.9363356552729629,
+      "eval_da_EOP_recall": 0.8861632809505557,
+      "eval_da_overall_accuracy": 0.9804166738526809,
+      "eval_da_overall_f1": 0.9105588594383838,
+      "eval_da_overall_precision": 0.9363356552729629,
+      "eval_da_overall_recall": 0.8861632809505557,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9758097617560453,
+      "eval_overall_f1": 0.8800495553294101,
+      "eval_overall_precision": 0.8974011911207364,
+      "eval_overall_recall": 0.8633561941140724,
+      "eval_runtime": 1877.3417,
+      "eval_samples_per_second": 4.115,
+      "eval_steps_per_second": 1.029,
+      "step": 52000
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 0.1861,
+      "step": 65000
+    },
+    {
+      "epoch": 5.0,
+      "eval_EOP_f1": 0.8782045673287252,
+      "eval_EOP_number": 11519,
+      "eval_EOP_precision": 0.885175059529059,
+      "eval_EOP_recall": 0.8713429985241774,
+      "eval_da_EOP_f1": 0.9094169005300903,
+      "eval_da_EOP_number": 13045,
+      "eval_da_EOP_precision": 0.9250654190785822,
+      "eval_da_EOP_recall": 0.8942889996167114,
+      "eval_da_overall_accuracy": 0.9799596433437386,
+      "eval_da_overall_f1": 0.9094169005300903,
+      "eval_da_overall_precision": 0.9250654190785822,
+      "eval_da_overall_recall": 0.8942889996167114,
+      "eval_loss": NaN,
+      "eval_overall_accuracy": 0.9751583831533863,
+      "eval_overall_f1": 0.8782045673287252,
+      "eval_overall_precision": 0.885175059529059,
+      "eval_overall_recall": 0.8713429985241774,
+      "eval_runtime": 1877.6728,
+      "eval_samples_per_second": 4.114,
+      "eval_steps_per_second": 1.029,
+      "step": 65000
+    },
+    {
+      "epoch": 5.0,
+      "step": 65000,
+      "total_flos": 4.846166765850378e+17,
+      "train_loss": 0.3591941669170673,
+      "train_runtime": 192526.8268,
+      "train_samples_per_second": 1.35,
+      "train_steps_per_second": 0.338
+    }
+  ],
+  "max_steps": 65000,
+  "num_train_epochs": 5,
+  "total_flos": 4.846166765850378e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d92bf3801ca1f860767f77ceff1ff01572f90a5a8007079d0502b24ce55bbd4
+size 3759