best models for abstract classification

config.json

@@ -0,0 +1,35 @@
+{
+  "_name_or_path": "/projects/SolarWindowsADSP/shu/models/revision/bert/batterybert-cased/",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "battery",
+    "1": "non-battery"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "battery": 0,
+    "non-battery": 1
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.17.0.dev0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 28996
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53d839ce28a09c97761c439c2404396e5ac1c3836bebccb8aed2ed145efc3979
+size 866620625

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a88afc77160e1cc1a49c77dbeb667509cf858d5ab469d0e725e052ab946df1a1
+size 433336585

rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:669f513d2f3d83b1cfc47acb74a028b2fff6a683892aaf557414b3620bd5759f
+size 14657

rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b6717eb06ff20db6f9eac1cded2bb929982b4950f278a2fced1857225fb05fa
+size 14657

rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c258e9591dbfbc83f78322f1b8658b9d088739eb22afd4748b52f7c624013ad3
+size 14657

rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90c0f37f676709fd76ef32d973467797029597ad1ea93ea57976ca633752305d
+size 14657

rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c60e238ce3b778719c51f8e86cb8be7d0e3bc4219bab295cd3ff9591aae767c1
+size 14657

rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68ce3bdf8f5fbc09a548a7f66846141fbd610c1f17bd123f4e61edeec44c17b0
+size 14657

rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49156254176c1ba3764b6c8d5e2b3f3ad1ce32da1bb3e17674d926ba123c5366
+size 14657

rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f8c908c8a57bf3b8d7c9f727b95764749b2f8090a0f812d8357e2494da134b9
+size 14657

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb2d6ea316aafd776d25c89bd0c36e3f70f448d4042f9bbb2dee342f9cb4d1aa
+size 623

special_tokens_map.json

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

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": false, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "/projects/SolarWindowsADSP/shu/models/revision/bert/batterybert-cased/", "tokenizer_class": "BertTokenizer"}

trainer_state.json

@@ -0,0 +1,247 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 11.0,
+  "global_step": 11231,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9347045380346067e-05,
+      "loss": 0.1628,
+      "step": 500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.8694090760692135e-05,
+      "loss": 0.1424,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9604628682136536,
+      "eval_loss": 0.13337615132331848,
+      "eval_runtime": 3.606,
+      "eval_samples_per_second": 2588.168,
+      "eval_steps_per_second": 80.976,
+      "step": 1021
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.80411361410382e-05,
+      "loss": 0.1054,
+      "step": 1500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7388181521384265e-05,
+      "loss": 0.1097,
+      "step": 2000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9652844667434692,
+      "eval_loss": 0.11623150110244751,
+      "eval_runtime": 3.5866,
+      "eval_samples_per_second": 2602.181,
+      "eval_steps_per_second": 81.414,
+      "step": 2042
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.673522690173033e-05,
+      "loss": 0.0804,
+      "step": 2500
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.6082272282076395e-05,
+      "loss": 0.0755,
+      "step": 3000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9673202633857727,
+      "eval_loss": 0.1282455027103424,
+      "eval_runtime": 3.6184,
+      "eval_samples_per_second": 2579.339,
+      "eval_steps_per_second": 80.699,
+      "step": 3063
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.5429317662422463e-05,
+      "loss": 0.0591,
+      "step": 3500
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.4776363042768528e-05,
+      "loss": 0.0494,
+      "step": 4000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9676417112350464,
+      "eval_loss": 0.14640793204307556,
+      "eval_runtime": 3.593,
+      "eval_samples_per_second": 2597.544,
+      "eval_steps_per_second": 81.269,
+      "step": 4084
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.4123408423114595e-05,
+      "loss": 0.0328,
+      "step": 4500
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.3470453803460662e-05,
+      "loss": 0.0353,
+      "step": 5000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9633558392524719,
+      "eval_loss": 0.19047445058822632,
+      "eval_runtime": 3.6111,
+      "eval_samples_per_second": 2584.563,
+      "eval_steps_per_second": 80.863,
+      "step": 5105
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1.2817499183806728e-05,
+      "loss": 0.0209,
+      "step": 5500
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 1.2164544564152792e-05,
+      "loss": 0.0182,
+      "step": 6000
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.968820333480835,
+      "eval_loss": 0.20608273148536682,
+      "eval_runtime": 3.6122,
+      "eval_samples_per_second": 2583.731,
+      "eval_steps_per_second": 80.837,
+      "step": 6126
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.1511589944498858e-05,
+      "loss": 0.0144,
+      "step": 6500
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.0858635324844923e-05,
+      "loss": 0.0126,
+      "step": 7000
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9689274430274963,
+      "eval_loss": 0.24051280319690704,
+      "eval_runtime": 3.603,
+      "eval_samples_per_second": 2590.356,
+      "eval_steps_per_second": 81.044,
+      "step": 7147
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 1.020568070519099e-05,
+      "loss": 0.0086,
+      "step": 7500
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 9.552726085537057e-06,
+      "loss": 0.0087,
+      "step": 8000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9695703387260437,
+      "eval_loss": 0.259203165769577,
+      "eval_runtime": 3.6023,
+      "eval_samples_per_second": 2590.86,
+      "eval_steps_per_second": 81.06,
+      "step": 8168
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 8.899771465883122e-06,
+      "loss": 0.0056,
+      "step": 8500
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 8.246816846229188e-06,
+      "loss": 0.0068,
+      "step": 9000
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9706417918205261,
+      "eval_loss": 0.2531001567840576,
+      "eval_runtime": 3.6041,
+      "eval_samples_per_second": 2589.528,
+      "eval_steps_per_second": 81.018,
+      "step": 9189
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 7.593862226575253e-06,
+      "loss": 0.0041,
+      "step": 9500
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 6.94090760692132e-06,
+      "loss": 0.0034,
+      "step": 10000
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9699989557266235,
+      "eval_loss": 0.281872034072876,
+      "eval_runtime": 3.597,
+      "eval_samples_per_second": 2594.647,
+      "eval_steps_per_second": 81.178,
+      "step": 10210
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 6.287952987267385e-06,
+      "loss": 0.0034,
+      "step": 10500
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 5.634998367613451e-06,
+      "loss": 0.0032,
+      "step": 11000
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.9728918671607971,
+      "eval_loss": 0.27420875430107117,
+      "eval_runtime": 3.6954,
+      "eval_samples_per_second": 2525.56,
+      "eval_steps_per_second": 79.017,
+      "step": 11231
+    }
+  ],
+  "max_steps": 15315,
+  "num_train_epochs": 15,
+  "total_flos": 2.363421365580595e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb5615137be998cf6339f47a4a8f352fd31a3bc9c39c2ea617f40308858790df
+size 3183