Upload folder using huggingface_hub
Browse files- financial-pii-ready/TokenBased-CRF/checkpoint-63/config.json +72 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/model.safetensors +3 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/optimizer.pt +3 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/rng_state.pth +3 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/scheduler.pt +3 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer.json +0 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer_config.json +15 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/trainer_state.json +33 -0
- financial-pii-ready/TokenBased-CRF/checkpoint-63/training_args.bin +3 -0
- financial-pii-ready/TokenBased-CRF/config.json +72 -0
- financial-pii-ready/TokenBased-CRF/pytorch_model.bin +3 -0
- financial-pii-ready/TokenBased-CRF/tokenizer.json +0 -0
- financial-pii-ready/TokenBased-CRF/tokenizer_config.json +15 -0
financial-pii-ready/TokenBased-CRF/checkpoint-63/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"TransformerCrfForTokenClassification"
|
| 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/TokenBased-CRF/checkpoint-63/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a368b03f677581bb8fd655dbdc802a4fbdb5e3edb0110378723bbafead161454
|
| 3 |
+
size 711492512
|
financial-pii-ready/TokenBased-CRF/checkpoint-63/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0d8931089cd75a43c0fb36b9673e981fe4bb116f8cad37679a9b3b181dac777
|
| 3 |
+
size 1418382155
|
financial-pii-ready/TokenBased-CRF/checkpoint-63/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:77e1bc0cc29528179cf7713b093fffbd0754c58c2875616aa74e1b6cd67bfe7d
|
| 3 |
+
size 14645
|
financial-pii-ready/TokenBased-CRF/checkpoint-63/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8baf2b5a46fac85f95c0a15caf2414a92c036d41b6ccf016dc0542ed26e499c2
|
| 3 |
+
size 1465
|
financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready/TokenBased-CRF/checkpoint-63/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/TokenBased-CRF/checkpoint-63/trainer_state.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 63,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [],
|
| 12 |
+
"logging_steps": 500,
|
| 13 |
+
"max_steps": 63,
|
| 14 |
+
"num_input_tokens_seen": 0,
|
| 15 |
+
"num_train_epochs": 1,
|
| 16 |
+
"save_steps": 500,
|
| 17 |
+
"stateful_callbacks": {
|
| 18 |
+
"TrainerControl": {
|
| 19 |
+
"args": {
|
| 20 |
+
"should_epoch_stop": false,
|
| 21 |
+
"should_evaluate": false,
|
| 22 |
+
"should_log": false,
|
| 23 |
+
"should_save": true,
|
| 24 |
+
"should_training_stop": true
|
| 25 |
+
},
|
| 26 |
+
"attributes": {}
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"total_flos": 131573741568000.0,
|
| 30 |
+
"train_batch_size": 8,
|
| 31 |
+
"trial_name": null,
|
| 32 |
+
"trial_params": null
|
| 33 |
+
}
|
financial-pii-ready/TokenBased-CRF/checkpoint-63/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a3e9ae25a3c620ff30c32b0562f13e8c74a0630c6e5f3c6ff751337dbed4ff7
|
| 3 |
+
size 5201
|
financial-pii-ready/TokenBased-CRF/config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_cross_attention": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"TransformerCrfForTokenClassification"
|
| 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/TokenBased-CRF/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9573b1a3c4c82ee47e0ea17df336a8a8274df97b6220f3aec5fd5c88440b2536
|
| 3 |
+
size 711554579
|
financial-pii-ready/TokenBased-CRF/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
financial-pii-ready/TokenBased-CRF/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 |
+
}
|