Pushing AE model.

added_tokens.json

@@ -0,0 +1 @@
+{"<context>": 50265}

config.json

@@ -0,0 +1,36 @@
+{
+  "_name_or_path": "roberta-large",
+  "architectures": [
+    "RobertaForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "eos_token_id": 2,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1,
+    "LABEL_2": 2
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "roberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.5.1",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50266
+}

experiment_config.json

@@ -0,0 +1,76 @@
+{
+    "model_wrapper": {
+        "type": "token_classification",
+        "num_labels": 3
+    },
+    "label_mapper": {
+        "type": "answer-extraction"
+    },
+    "args": {
+        "type": "default",
+        "disable_tqdm": true,
+        "do_eval": true,
+        "do_train": true,
+        "eval_steps": 1824,
+        "evaluation_strategy": "steps",
+        "fp16": true,
+        "fp16_backend": "amp",
+        "gradient_accumulation_steps": 1,
+        "greater_is_better": true,
+        "label_names": [
+            "labels"
+        ],
+        "logging_dir": "/opt/ml/checkpoints/logs",
+        "logging_first_step": true,
+        "logging_steps": 1824,
+        "lr_scheduler_type": "linear",
+        "metric_for_best_model": "eval_ANS_f1",
+        "no_cuda": false,
+        "num_train_epochs": 3,
+        "output_dir": "/opt/ml/checkpoints",
+        "per_device_eval_batch_size": 10,
+        "per_device_train_batch_size": 4,
+        "result_dir": "/opt/ml/model",
+        "save_steps": 1824,
+        "save_total_limit": null,
+        "warmup_steps": 985
+    },
+    "compute_metrics": {
+        "type": "seqeval"
+    },
+    "dataset_loader": {
+        "dataset_reader": {
+            "path": "squad_qa_test_fixture"
+        },
+        "data_processor": {
+            "type": "squad-answer-extraction"
+        },
+        "data_adapter": {
+            "type": "answer-extraction"
+        }
+    },
+    "dev_file_path": "/opt/ml/input/data/validation/dev.json",
+    "optimizer": {
+        "type": "huggingface_adamw",
+        "eps": 1e-06,
+        "lr": 3e-05,
+        "parameter_groups": [
+            [
+                [
+                    "bias",
+                    "LayerNorm\\\\.weight",
+                    "layer_norm\\\\.weight"
+                ],
+                {
+                    "weight_decay": 0
+                }
+            ]
+        ],
+        "weight_decay": 0.01
+    },
+    "pretrained_model_name_or_path": "roberta-large",
+    "tokenizer_wrapper": {
+        "type": "answer-extraction"
+    },
+    "train_file_path": "/opt/ml/input/data/training/train.json"
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:475a39afaa56931679cd29bdf312ccb374b6630cffea1fd7d5bd20bb084e0d3f
+size 2834759533

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d54f0c32d42cd0ca2268878de5386bf651542ba6a41cebfd58eeb01a6b22280
+size 1417429687

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03a9739605301291c1625ef6851ee561ee02a27385a1a849a301b3eeaa328225
+size 623

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}, "additional_special_tokens": ["<context>"]}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "add_prefix_space": false, "errors": "replace", "sep_token": "</s>", "cls_token": "<s>", "pad_token": "<pad>", "mask_token": "<mask>", "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "roberta-large"}

trainer_state.json

@@ -0,0 +1,190 @@
+{
+  "best_metric": 0.25107308099671105,
+  "best_model_checkpoint": "/opt/ml/checkpoints/checkpoint-14592",
+  "epoch": 2.666179426274438,
+  "global_step": 14592,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.045685279187818e-08,
+      "loss": 2.0931,
+      "step": 1
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.8369184916418297e-05,
+      "loss": 0.2097,
+      "step": 1824
+    },
+    {
+      "epoch": 0.33,
+      "eval_ANS_f1": 0.1965132827324478,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.28789019516997744,
+      "eval_ANS_recall": 0.14916729182295574,
+      "eval_loss": 0.1584259271621704,
+      "eval_overall_accuracy": 0.9525067851669007,
+      "eval_overall_f1": 0.1965132827324478,
+      "eval_overall_precision": 0.28789019516997744,
+      "eval_overall_recall": 0.14916729182295574,
+      "eval_runtime": 271.0381,
+      "eval_samples_per_second": 64.637,
+      "step": 1824
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.4823765712064277e-05,
+      "loss": 0.1427,
+      "step": 3648
+    },
+    {
+      "epoch": 0.67,
+      "eval_ANS_f1": 0.14480276621268787,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.3350021767522856,
+      "eval_ANS_recall": 0.09236309077269317,
+      "eval_loss": 0.1570661962032318,
+      "eval_overall_accuracy": 0.9537896342872352,
+      "eval_overall_f1": 0.14480276621268787,
+      "eval_overall_precision": 0.3350021767522856,
+      "eval_overall_recall": 0.09236309077269317,
+      "eval_runtime": 272.0666,
+      "eval_samples_per_second": 64.392,
+      "step": 3648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.127834650771025e-05,
+      "loss": 0.1381,
+      "step": 5472
+    },
+    {
+      "epoch": 1.0,
+      "eval_ANS_f1": 0.18933628222641427,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.34125371267781773,
+      "eval_ANS_recall": 0.13101275318829708,
+      "eval_loss": 0.15228605270385742,
+      "eval_overall_accuracy": 0.954028691761447,
+      "eval_overall_f1": 0.18933628222641427,
+      "eval_overall_precision": 0.34125371267781773,
+      "eval_overall_recall": 0.13101275318829708,
+      "eval_runtime": 272.4342,
+      "eval_samples_per_second": 64.305,
+      "step": 5472
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.773292730335623e-05,
+      "loss": 0.1295,
+      "step": 7296
+    },
+    {
+      "epoch": 1.33,
+      "eval_ANS_f1": 0.18266756939742723,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.36841147018661813,
+      "eval_ANS_recall": 0.1214403600900225,
+      "eval_loss": 0.15274310111999512,
+      "eval_overall_accuracy": 0.9544393980616191,
+      "eval_overall_f1": 0.18266756939742723,
+      "eval_overall_precision": 0.36841147018661813,
+      "eval_overall_recall": 0.1214403600900225,
+      "eval_runtime": 269.6146,
+      "eval_samples_per_second": 64.978,
+      "step": 7296
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.4187508099002203e-05,
+      "loss": 0.1275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.67,
+      "eval_ANS_f1": 0.22205431725643288,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.3693876125968857,
+      "eval_ANS_recall": 0.1587396849212303,
+      "eval_loss": 0.15151338279247284,
+      "eval_overall_accuracy": 0.9544824840017387,
+      "eval_overall_f1": 0.22205431725643288,
+      "eval_overall_precision": 0.3693876125968857,
+      "eval_overall_recall": 0.1587396849212303,
+      "eval_runtime": 273.5764,
+      "eval_samples_per_second": 64.037,
+      "step": 9120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.0642088894648179e-05,
+      "loss": 0.1262,
+      "step": 10944
+    },
+    {
+      "epoch": 2.0,
+      "eval_ANS_f1": 0.22321074285236933,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.3670833905756285,
+      "eval_ANS_recall": 0.16036009002250562,
+      "eval_loss": 0.14961716532707214,
+      "eval_overall_accuracy": 0.9540054114550921,
+      "eval_overall_f1": 0.22321074285236933,
+      "eval_overall_precision": 0.3670833905756285,
+      "eval_overall_recall": 0.16036009002250562,
+      "eval_runtime": 271.8734,
+      "eval_samples_per_second": 64.438,
+      "step": 10944
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 7.096669690294156e-06,
+      "loss": 0.1126,
+      "step": 12768
+    },
+    {
+      "epoch": 2.33,
+      "eval_ANS_f1": 0.2386963952765693,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.33832516654737654,
+      "eval_ANS_recall": 0.18439609902475618,
+      "eval_loss": 0.15988405048847198,
+      "eval_overall_accuracy": 0.9519723805225143,
+      "eval_overall_f1": 0.2386963952765693,
+      "eval_overall_precision": 0.33832516654737654,
+      "eval_overall_recall": 0.18439609902475618,
+      "eval_runtime": 273.7781,
+      "eval_samples_per_second": 63.99,
+      "step": 12768
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.551250485940132e-06,
+      "loss": 0.1109,
+      "step": 14592
+    },
+    {
+      "epoch": 2.67,
+      "eval_ANS_f1": 0.25107308099671105,
+      "eval_ANS_number": 33325,
+      "eval_ANS_precision": 0.3296896349795042,
+      "eval_ANS_recall": 0.20273068267066766,
+      "eval_loss": 0.16195876896381378,
+      "eval_overall_accuracy": 0.9512107322907231,
+      "eval_overall_f1": 0.25107308099671105,
+      "eval_overall_precision": 0.3296896349795042,
+      "eval_overall_recall": 0.20273068267066766,
+      "eval_runtime": 275.1376,
+      "eval_samples_per_second": 63.674,
+      "step": 14592
+    }
+  ],
+  "max_steps": 16419,
+  "num_train_epochs": 3,
+  "total_flos": 3.702162559715073e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7b5aabe905b80a3f338326b74bc91382252767f6d9066d870b24d084942b025
+size 2415