Upload folder using huggingface_hub
Browse files- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/config.json +72 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/model.safetensors +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/optimizer.pt +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/rng_state.pth +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/scheduler.pt +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/tokenizer.json +0 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/tokenizer_config.json +15 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/trainer_state.json +56 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/training_args.bin +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/config.json +72 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/model.safetensors +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/optimizer.pt +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/rng_state.pth +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/scheduler.pt +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/tokenizer.json +0 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/tokenizer_config.json +15 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/trainer_state.json +85 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/training_args.bin +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/config.json +72 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/model.safetensors +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/optimizer.pt +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/rng_state.pth +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/scheduler.pt +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/tokenizer.json +0 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/tokenizer_config.json +15 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/trainer_state.json +107 -0
- financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/training_args.bin +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/config.json +72 -0
- financial-pii-ready-augmented/TokenBased-BERT/model.safetensors +3 -0
- financial-pii-ready-augmented/TokenBased-BERT/tokenizer.json +0 -0
- financial-pii-ready-augmented/TokenBased-BERT/tokenizer_config.json +15 -0
- financial-pii-ready-augmented/TokenBased-BERT/training_args.bin +3 -0
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d22d9f2aedd27b2f9b6bd2f6b47cc34e47b5806c3800a42246daf737f43b551
|
| 3 |
+
size 709127020
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93e180d6f1de491c6c071ee2cc4047e884ffe965433e27592ad070556776494e
|
| 3 |
+
size 1418377099
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f723918c1a31b1b5c3ddcc544f5440ab89339ab85bdc3c8f1287e36c8f7f8a9e
|
| 3 |
+
size 14645
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce6d3e9031acf7f981f453d679bbe342d984ff5aeb56532d27c6ea458de0d9d9
|
| 3 |
+
size 1465
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/trainer_state.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1266,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3949447077409163,
|
| 14 |
+
"grad_norm": 0.014986732043325901,
|
| 15 |
+
"learning_rate": 1.737230121116377e-05,
|
| 16 |
+
"loss": 0.058785953521728516,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7898894154818326,
|
| 21 |
+
"grad_norm": 0.03321343660354614,
|
| 22 |
+
"learning_rate": 1.4739336492890996e-05,
|
| 23 |
+
"loss": 0.0014266656637191772,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_loss": 0.000838952895719558,
|
| 29 |
+
"eval_runtime": 16.245,
|
| 30 |
+
"eval_samples_per_second": 554.017,
|
| 31 |
+
"eval_steps_per_second": 69.252,
|
| 32 |
+
"step": 1266
|
| 33 |
+
}
|
| 34 |
+
],
|
| 35 |
+
"logging_steps": 500,
|
| 36 |
+
"max_steps": 3798,
|
| 37 |
+
"num_input_tokens_seen": 0,
|
| 38 |
+
"num_train_epochs": 3,
|
| 39 |
+
"save_steps": 500,
|
| 40 |
+
"stateful_callbacks": {
|
| 41 |
+
"TrainerControl": {
|
| 42 |
+
"args": {
|
| 43 |
+
"should_epoch_stop": false,
|
| 44 |
+
"should_evaluate": false,
|
| 45 |
+
"should_log": false,
|
| 46 |
+
"should_save": true,
|
| 47 |
+
"should_training_stop": false
|
| 48 |
+
},
|
| 49 |
+
"attributes": {}
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"total_flos": 2.1167907572736e+16,
|
| 53 |
+
"train_batch_size": 64,
|
| 54 |
+
"trial_name": null,
|
| 55 |
+
"trial_params": null
|
| 56 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-1266/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a5dcd6cd4928720e853b7aa9e45f03f28f0e7ec0feb786651bfbe24f71b4aa1
|
| 3 |
+
size 5201
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0d61f5284c270fe93de6bee8272cfb809fb7ff9f9a14925c5008fb560d9ea52
|
| 3 |
+
size 709127020
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8fc1a0308ec09cda791c69ad81a8cfd6674690d1c6f52836431a84acdad2b0a2
|
| 3 |
+
size 1418377099
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:891132493e9e6da9564eacd082621286d9a6baa5a3b2c5ef90a1f291ce8ac3b5
|
| 3 |
+
size 14645
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69efa8de68881242a3a1e1dafe3bccd449b7960d290a67b07e993ed60d667162
|
| 3 |
+
size 1465
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/trainer_state.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2532,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3949447077409163,
|
| 14 |
+
"grad_norm": 0.014986732043325901,
|
| 15 |
+
"learning_rate": 1.737230121116377e-05,
|
| 16 |
+
"loss": 0.058785953521728516,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7898894154818326,
|
| 21 |
+
"grad_norm": 0.03321343660354614,
|
| 22 |
+
"learning_rate": 1.4739336492890996e-05,
|
| 23 |
+
"loss": 0.0014266656637191772,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_loss": 0.000838952895719558,
|
| 29 |
+
"eval_runtime": 16.245,
|
| 30 |
+
"eval_samples_per_second": 554.017,
|
| 31 |
+
"eval_steps_per_second": 69.252,
|
| 32 |
+
"step": 1266
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.1848341232227488,
|
| 36 |
+
"grad_norm": 0.01166529580950737,
|
| 37 |
+
"learning_rate": 1.2106371774618222e-05,
|
| 38 |
+
"loss": 0.0008846146464347839,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.5797788309636651,
|
| 43 |
+
"grad_norm": 0.025141963735222816,
|
| 44 |
+
"learning_rate": 9.473407056345445e-06,
|
| 45 |
+
"loss": 0.0005799291133880615,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.9747235387045814,
|
| 50 |
+
"grad_norm": 0.0046472507528960705,
|
| 51 |
+
"learning_rate": 6.84044233807267e-06,
|
| 52 |
+
"loss": 0.00048410308361053467,
|
| 53 |
+
"step": 2500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 2.0,
|
| 57 |
+
"eval_loss": 0.00022992123558651656,
|
| 58 |
+
"eval_runtime": 16.2616,
|
| 59 |
+
"eval_samples_per_second": 553.451,
|
| 60 |
+
"eval_steps_per_second": 69.181,
|
| 61 |
+
"step": 2532
|
| 62 |
+
}
|
| 63 |
+
],
|
| 64 |
+
"logging_steps": 500,
|
| 65 |
+
"max_steps": 3798,
|
| 66 |
+
"num_input_tokens_seen": 0,
|
| 67 |
+
"num_train_epochs": 3,
|
| 68 |
+
"save_steps": 500,
|
| 69 |
+
"stateful_callbacks": {
|
| 70 |
+
"TrainerControl": {
|
| 71 |
+
"args": {
|
| 72 |
+
"should_epoch_stop": false,
|
| 73 |
+
"should_evaluate": false,
|
| 74 |
+
"should_log": false,
|
| 75 |
+
"should_save": true,
|
| 76 |
+
"should_training_stop": false
|
| 77 |
+
},
|
| 78 |
+
"attributes": {}
|
| 79 |
+
}
|
| 80 |
+
},
|
| 81 |
+
"total_flos": 4.2335815145472e+16,
|
| 82 |
+
"train_batch_size": 64,
|
| 83 |
+
"trial_name": null,
|
| 84 |
+
"trial_params": null
|
| 85 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-2532/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a5dcd6cd4928720e853b7aa9e45f03f28f0e7ec0feb786651bfbe24f71b4aa1
|
| 3 |
+
size 5201
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10d3e0748399553cd0a57f27853f816666656e57866d55ddc847b85e2072d6f9
|
| 3 |
+
size 709127020
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52ffe20cd5feda65016275b4988e648a43531525eddaea7ad4a1ba4fcbcee882
|
| 3 |
+
size 1418377099
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7929ac78388e0ec3d41cca0c87dd02a819e57ddb3890a9e554a75c577f089e97
|
| 3 |
+
size 14645
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3914220417dc768575571dac19f96808e2cdb6f306f00faab8e23bc6f26c4797
|
| 3 |
+
size 1465
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/trainer_state.json
ADDED
|
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3798,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3949447077409163,
|
| 14 |
+
"grad_norm": 0.014986732043325901,
|
| 15 |
+
"learning_rate": 1.737230121116377e-05,
|
| 16 |
+
"loss": 0.058785953521728516,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.7898894154818326,
|
| 21 |
+
"grad_norm": 0.03321343660354614,
|
| 22 |
+
"learning_rate": 1.4739336492890996e-05,
|
| 23 |
+
"loss": 0.0014266656637191772,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"eval_loss": 0.000838952895719558,
|
| 29 |
+
"eval_runtime": 16.245,
|
| 30 |
+
"eval_samples_per_second": 554.017,
|
| 31 |
+
"eval_steps_per_second": 69.252,
|
| 32 |
+
"step": 1266
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.1848341232227488,
|
| 36 |
+
"grad_norm": 0.01166529580950737,
|
| 37 |
+
"learning_rate": 1.2106371774618222e-05,
|
| 38 |
+
"loss": 0.0008846146464347839,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.5797788309636651,
|
| 43 |
+
"grad_norm": 0.025141963735222816,
|
| 44 |
+
"learning_rate": 9.473407056345445e-06,
|
| 45 |
+
"loss": 0.0005799291133880615,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.9747235387045814,
|
| 50 |
+
"grad_norm": 0.0046472507528960705,
|
| 51 |
+
"learning_rate": 6.84044233807267e-06,
|
| 52 |
+
"loss": 0.00048410308361053467,
|
| 53 |
+
"step": 2500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 2.0,
|
| 57 |
+
"eval_loss": 0.00022992123558651656,
|
| 58 |
+
"eval_runtime": 16.2616,
|
| 59 |
+
"eval_samples_per_second": 553.451,
|
| 60 |
+
"eval_steps_per_second": 69.181,
|
| 61 |
+
"step": 2532
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 2.3696682464454977,
|
| 65 |
+
"grad_norm": 0.02571079321205616,
|
| 66 |
+
"learning_rate": 4.207477619799895e-06,
|
| 67 |
+
"loss": 0.00036954617500305177,
|
| 68 |
+
"step": 3000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 2.764612954186414,
|
| 72 |
+
"grad_norm": 0.0029153793584555387,
|
| 73 |
+
"learning_rate": 1.5745129015271196e-06,
|
| 74 |
+
"loss": 0.00031371000409126284,
|
| 75 |
+
"step": 3500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 3.0,
|
| 79 |
+
"eval_loss": 0.00018665596144273877,
|
| 80 |
+
"eval_runtime": 16.2995,
|
| 81 |
+
"eval_samples_per_second": 552.165,
|
| 82 |
+
"eval_steps_per_second": 69.021,
|
| 83 |
+
"step": 3798
|
| 84 |
+
}
|
| 85 |
+
],
|
| 86 |
+
"logging_steps": 500,
|
| 87 |
+
"max_steps": 3798,
|
| 88 |
+
"num_input_tokens_seen": 0,
|
| 89 |
+
"num_train_epochs": 3,
|
| 90 |
+
"save_steps": 500,
|
| 91 |
+
"stateful_callbacks": {
|
| 92 |
+
"TrainerControl": {
|
| 93 |
+
"args": {
|
| 94 |
+
"should_epoch_stop": false,
|
| 95 |
+
"should_evaluate": false,
|
| 96 |
+
"should_log": false,
|
| 97 |
+
"should_save": true,
|
| 98 |
+
"should_training_stop": true
|
| 99 |
+
},
|
| 100 |
+
"attributes": {}
|
| 101 |
+
}
|
| 102 |
+
},
|
| 103 |
+
"total_flos": 6.3503722718208e+16,
|
| 104 |
+
"train_batch_size": 64,
|
| 105 |
+
"trial_name": null,
|
| 106 |
+
"trial_params": null
|
| 107 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/checkpoint-3798/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a5dcd6cd4928720e853b7aa9e45f03f28f0e7ec0feb786651bfbe24f71b4aa1
|
| 3 |
+
size 5201
|
financial-pii-ready-augmented/TokenBased-BERT/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForTokenClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"directionality": "bidi",
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_act": "gelu",
|
| 13 |
+
"hidden_dropout_prob": 0.1,
|
| 14 |
+
"hidden_size": 768,
|
| 15 |
+
"id2label": {
|
| 16 |
+
"0": "O",
|
| 17 |
+
"1": "B-ADDRESS",
|
| 18 |
+
"2": "I-ADDRESS",
|
| 19 |
+
"3": "B-COMPANY",
|
| 20 |
+
"4": "I-COMPANY",
|
| 21 |
+
"5": "B-CREDIT_CARD",
|
| 22 |
+
"6": "I-CREDIT_CARD",
|
| 23 |
+
"7": "B-EMAIL",
|
| 24 |
+
"8": "I-EMAIL",
|
| 25 |
+
"9": "B-NAME",
|
| 26 |
+
"10": "I-NAME",
|
| 27 |
+
"11": "B-PHONE",
|
| 28 |
+
"12": "I-PHONE",
|
| 29 |
+
"13": "B-SSN",
|
| 30 |
+
"14": "I-SSN",
|
| 31 |
+
"15": "B-URL",
|
| 32 |
+
"16": "I-URL"
|
| 33 |
+
},
|
| 34 |
+
"initializer_range": 0.02,
|
| 35 |
+
"intermediate_size": 3072,
|
| 36 |
+
"is_decoder": false,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"B-ADDRESS": 1,
|
| 39 |
+
"B-COMPANY": 3,
|
| 40 |
+
"B-CREDIT_CARD": 5,
|
| 41 |
+
"B-EMAIL": 7,
|
| 42 |
+
"B-NAME": 9,
|
| 43 |
+
"B-PHONE": 11,
|
| 44 |
+
"B-SSN": 13,
|
| 45 |
+
"B-URL": 15,
|
| 46 |
+
"I-ADDRESS": 2,
|
| 47 |
+
"I-COMPANY": 4,
|
| 48 |
+
"I-CREDIT_CARD": 6,
|
| 49 |
+
"I-EMAIL": 8,
|
| 50 |
+
"I-NAME": 10,
|
| 51 |
+
"I-PHONE": 12,
|
| 52 |
+
"I-SSN": 14,
|
| 53 |
+
"I-URL": 16,
|
| 54 |
+
"O": 0
|
| 55 |
+
},
|
| 56 |
+
"layer_norm_eps": 1e-12,
|
| 57 |
+
"max_position_embeddings": 512,
|
| 58 |
+
"model_type": "bert",
|
| 59 |
+
"num_attention_heads": 12,
|
| 60 |
+
"num_hidden_layers": 12,
|
| 61 |
+
"pad_token_id": 0,
|
| 62 |
+
"pooler_fc_size": 768,
|
| 63 |
+
"pooler_num_attention_heads": 12,
|
| 64 |
+
"pooler_num_fc_layers": 3,
|
| 65 |
+
"pooler_size_per_head": 128,
|
| 66 |
+
"pooler_type": "first_token_transform",
|
| 67 |
+
"tie_word_embeddings": true,
|
| 68 |
+
"transformers_version": "5.3.0",
|
| 69 |
+
"type_vocab_size": 2,
|
| 70 |
+
"use_cache": false,
|
| 71 |
+
"vocab_size": 119547
|
| 72 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10d3e0748399553cd0a57f27853f816666656e57866d55ddc847b85e2072d6f9
|
| 3 |
+
size 709127020
|
financial-pii-ready-augmented/TokenBased-BERT/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready-augmented/TokenBased-BERT/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"cls_token": "[CLS]",
|
| 5 |
+
"do_lower_case": false,
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
financial-pii-ready-augmented/TokenBased-BERT/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a5dcd6cd4928720e853b7aa9e45f03f28f0e7ec0feb786651bfbe24f71b4aa1
|
| 3 |
+
size 5201
|