Upload folder using huggingface_hub
Browse files- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/config.json +63 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json +0 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json +23 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json +48 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/config.json +63 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/model.safetensors +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/optimizer.pt +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/rng_state.pth +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/scheduler.pt +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer.json +0 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer_config.json +23 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/trainer_state.json +48 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/training_args.bin +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/config.json +63 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/model.safetensors +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/optimizer.pt +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/rng_state.pth +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/scheduler.pt +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer.json +0 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer_config.json +23 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/trainer_state.json +41 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/training_args.bin +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/config.json +63 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/pytorch_model.bin +3 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/tokenizer.json +0 -0
- gorkem-pii-ready/SpanBased-CustomDeBERTa/tokenizer_config.json +23 -0
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/config.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ADDRESS",
|
| 15 |
+
"2": "CREDIT_CARD",
|
| 16 |
+
"3": "CRYPTO_ADDRESS",
|
| 17 |
+
"4": "EMAIL",
|
| 18 |
+
"5": "IBAN",
|
| 19 |
+
"6": "OFF_PLATFORM_ATTEMPT",
|
| 20 |
+
"7": "PHONE",
|
| 21 |
+
"8": "SOCIAL_MEDIA",
|
| 22 |
+
"9": "URL"
|
| 23 |
+
},
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 3072,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"ADDRESS": 1,
|
| 28 |
+
"CREDIT_CARD": 2,
|
| 29 |
+
"CRYPTO_ADDRESS": 3,
|
| 30 |
+
"EMAIL": 4,
|
| 31 |
+
"IBAN": 5,
|
| 32 |
+
"O": 0,
|
| 33 |
+
"OFF_PLATFORM_ATTEMPT": 6,
|
| 34 |
+
"PHONE": 7,
|
| 35 |
+
"SOCIAL_MEDIA": 8,
|
| 36 |
+
"URL": 9
|
| 37 |
+
},
|
| 38 |
+
"layer_norm_eps": 1e-07,
|
| 39 |
+
"legacy": true,
|
| 40 |
+
"max_position_embeddings": 512,
|
| 41 |
+
"max_relative_positions": -1,
|
| 42 |
+
"model_type": "deberta-v2",
|
| 43 |
+
"norm_rel_ebd": "layer_norm",
|
| 44 |
+
"num_attention_heads": 12,
|
| 45 |
+
"num_hidden_layers": 12,
|
| 46 |
+
"pad_token_id": 0,
|
| 47 |
+
"pooler_dropout": 0,
|
| 48 |
+
"pooler_hidden_act": "gelu",
|
| 49 |
+
"pooler_hidden_size": 768,
|
| 50 |
+
"pos_att_type": [
|
| 51 |
+
"p2c",
|
| 52 |
+
"c2p"
|
| 53 |
+
],
|
| 54 |
+
"position_biased_input": false,
|
| 55 |
+
"position_buckets": 256,
|
| 56 |
+
"relative_attention": true,
|
| 57 |
+
"share_att_key": true,
|
| 58 |
+
"tie_word_embeddings": true,
|
| 59 |
+
"transformers_version": "5.3.0",
|
| 60 |
+
"type_vocab_size": 0,
|
| 61 |
+
"use_cache": false,
|
| 62 |
+
"vocab_size": 128100
|
| 63 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b50cd5ab3c7657588b9a8806b8067cff40a5a88f08af036cb65a7c350c6e3910
|
| 3 |
+
size 741359744
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81efb579de78cb0c15910daed7b2813085eb412a6429ddf08ef0fd3f95424b29
|
| 3 |
+
size 1482843147
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71cb42a9e4441862670a2ab5ca064012a2845f17a7eaf1f9dcd3ae09778900b8
|
| 3 |
+
size 14645
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ed3b7879c9d08310fa68bd1af5c1b9a3843a452ce133fa854cced73ec663d43
|
| 3 |
+
size 1465
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.3831258644536653,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.6915629322268326,
|
| 14 |
+
"grad_norm": 0.06808845698833466,
|
| 15 |
+
"learning_rate": 1.3098201936376211e-05,
|
| 16 |
+
"loss": 0.012068235397338867,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.3831258644536653,
|
| 21 |
+
"grad_norm": 0.031969357281923294,
|
| 22 |
+
"learning_rate": 6.182572614107885e-06,
|
| 23 |
+
"loss": 0.0036826863288879396,
|
| 24 |
+
"step": 1000
|
| 25 |
+
}
|
| 26 |
+
],
|
| 27 |
+
"logging_steps": 500,
|
| 28 |
+
"max_steps": 1446,
|
| 29 |
+
"num_input_tokens_seen": 0,
|
| 30 |
+
"num_train_epochs": 2,
|
| 31 |
+
"save_steps": 500,
|
| 32 |
+
"stateful_callbacks": {
|
| 33 |
+
"TrainerControl": {
|
| 34 |
+
"args": {
|
| 35 |
+
"should_epoch_stop": false,
|
| 36 |
+
"should_evaluate": false,
|
| 37 |
+
"should_log": false,
|
| 38 |
+
"should_save": true,
|
| 39 |
+
"should_training_stop": false
|
| 40 |
+
},
|
| 41 |
+
"attributes": {}
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"total_flos": 243931167235368.0,
|
| 45 |
+
"train_batch_size": 16,
|
| 46 |
+
"trial_name": null,
|
| 47 |
+
"trial_params": null
|
| 48 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e609457181ac50440df7d6728c3518d3b9316719fead1acbcffdb0f4941dd06
|
| 3 |
+
size 5201
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/config.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ADDRESS",
|
| 15 |
+
"2": "CREDIT_CARD",
|
| 16 |
+
"3": "CRYPTO_ADDRESS",
|
| 17 |
+
"4": "EMAIL",
|
| 18 |
+
"5": "IBAN",
|
| 19 |
+
"6": "OFF_PLATFORM_ATTEMPT",
|
| 20 |
+
"7": "PHONE",
|
| 21 |
+
"8": "SOCIAL_MEDIA",
|
| 22 |
+
"9": "URL"
|
| 23 |
+
},
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 3072,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"ADDRESS": 1,
|
| 28 |
+
"CREDIT_CARD": 2,
|
| 29 |
+
"CRYPTO_ADDRESS": 3,
|
| 30 |
+
"EMAIL": 4,
|
| 31 |
+
"IBAN": 5,
|
| 32 |
+
"O": 0,
|
| 33 |
+
"OFF_PLATFORM_ATTEMPT": 6,
|
| 34 |
+
"PHONE": 7,
|
| 35 |
+
"SOCIAL_MEDIA": 8,
|
| 36 |
+
"URL": 9
|
| 37 |
+
},
|
| 38 |
+
"layer_norm_eps": 1e-07,
|
| 39 |
+
"legacy": true,
|
| 40 |
+
"max_position_embeddings": 512,
|
| 41 |
+
"max_relative_positions": -1,
|
| 42 |
+
"model_type": "deberta-v2",
|
| 43 |
+
"norm_rel_ebd": "layer_norm",
|
| 44 |
+
"num_attention_heads": 12,
|
| 45 |
+
"num_hidden_layers": 12,
|
| 46 |
+
"pad_token_id": 0,
|
| 47 |
+
"pooler_dropout": 0,
|
| 48 |
+
"pooler_hidden_act": "gelu",
|
| 49 |
+
"pooler_hidden_size": 768,
|
| 50 |
+
"pos_att_type": [
|
| 51 |
+
"p2c",
|
| 52 |
+
"c2p"
|
| 53 |
+
],
|
| 54 |
+
"position_biased_input": false,
|
| 55 |
+
"position_buckets": 256,
|
| 56 |
+
"relative_attention": true,
|
| 57 |
+
"share_att_key": true,
|
| 58 |
+
"tie_word_embeddings": true,
|
| 59 |
+
"transformers_version": "5.3.0",
|
| 60 |
+
"type_vocab_size": 0,
|
| 61 |
+
"use_cache": false,
|
| 62 |
+
"vocab_size": 128100
|
| 63 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e0451ac8b4df0b6b6d533b31ec15c485262a83d16df5fbbf3ea11188baee6592
|
| 3 |
+
size 741359744
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f24e1abe2c8864c29da0680eee30c4c32cb4197b8014d068f5c3a96e0ec8bbab
|
| 3 |
+
size 1482843147
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35e65ddbf7e3c4fc0d4d3c23af53d056bf5459ff4970dcb0db5997c61b26527e
|
| 3 |
+
size 14645
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8742f08c44d981613576954e914b525feeeddfe29bf72a4b0ac0e765a99171c
|
| 3 |
+
size 1465
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/trainer_state.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 1446,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.6915629322268326,
|
| 14 |
+
"grad_norm": 0.06808845698833466,
|
| 15 |
+
"learning_rate": 1.3098201936376211e-05,
|
| 16 |
+
"loss": 0.012068235397338867,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.3831258644536653,
|
| 21 |
+
"grad_norm": 0.031969357281923294,
|
| 22 |
+
"learning_rate": 6.182572614107885e-06,
|
| 23 |
+
"loss": 0.0036826863288879396,
|
| 24 |
+
"step": 1000
|
| 25 |
+
}
|
| 26 |
+
],
|
| 27 |
+
"logging_steps": 500,
|
| 28 |
+
"max_steps": 1446,
|
| 29 |
+
"num_input_tokens_seen": 0,
|
| 30 |
+
"num_train_epochs": 2,
|
| 31 |
+
"save_steps": 500,
|
| 32 |
+
"stateful_callbacks": {
|
| 33 |
+
"TrainerControl": {
|
| 34 |
+
"args": {
|
| 35 |
+
"should_epoch_stop": false,
|
| 36 |
+
"should_evaluate": false,
|
| 37 |
+
"should_log": false,
|
| 38 |
+
"should_save": true,
|
| 39 |
+
"should_training_stop": true
|
| 40 |
+
},
|
| 41 |
+
"attributes": {}
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"total_flos": 353148288762204.0,
|
| 45 |
+
"train_batch_size": 16,
|
| 46 |
+
"trial_name": null,
|
| 47 |
+
"trial_params": null
|
| 48 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e609457181ac50440df7d6728c3518d3b9316719fead1acbcffdb0f4941dd06
|
| 3 |
+
size 5201
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/config.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ADDRESS",
|
| 15 |
+
"2": "CREDIT_CARD",
|
| 16 |
+
"3": "CRYPTO_ADDRESS",
|
| 17 |
+
"4": "EMAIL",
|
| 18 |
+
"5": "IBAN",
|
| 19 |
+
"6": "OFF_PLATFORM_ATTEMPT",
|
| 20 |
+
"7": "PHONE",
|
| 21 |
+
"8": "SOCIAL_MEDIA",
|
| 22 |
+
"9": "URL"
|
| 23 |
+
},
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 3072,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"ADDRESS": 1,
|
| 28 |
+
"CREDIT_CARD": 2,
|
| 29 |
+
"CRYPTO_ADDRESS": 3,
|
| 30 |
+
"EMAIL": 4,
|
| 31 |
+
"IBAN": 5,
|
| 32 |
+
"O": 0,
|
| 33 |
+
"OFF_PLATFORM_ATTEMPT": 6,
|
| 34 |
+
"PHONE": 7,
|
| 35 |
+
"SOCIAL_MEDIA": 8,
|
| 36 |
+
"URL": 9
|
| 37 |
+
},
|
| 38 |
+
"layer_norm_eps": 1e-07,
|
| 39 |
+
"legacy": true,
|
| 40 |
+
"max_position_embeddings": 512,
|
| 41 |
+
"max_relative_positions": -1,
|
| 42 |
+
"model_type": "deberta-v2",
|
| 43 |
+
"norm_rel_ebd": "layer_norm",
|
| 44 |
+
"num_attention_heads": 12,
|
| 45 |
+
"num_hidden_layers": 12,
|
| 46 |
+
"pad_token_id": 0,
|
| 47 |
+
"pooler_dropout": 0,
|
| 48 |
+
"pooler_hidden_act": "gelu",
|
| 49 |
+
"pooler_hidden_size": 768,
|
| 50 |
+
"pos_att_type": [
|
| 51 |
+
"p2c",
|
| 52 |
+
"c2p"
|
| 53 |
+
],
|
| 54 |
+
"position_biased_input": false,
|
| 55 |
+
"position_buckets": 256,
|
| 56 |
+
"relative_attention": true,
|
| 57 |
+
"share_att_key": true,
|
| 58 |
+
"tie_word_embeddings": true,
|
| 59 |
+
"transformers_version": "5.3.0",
|
| 60 |
+
"type_vocab_size": 0,
|
| 61 |
+
"use_cache": false,
|
| 62 |
+
"vocab_size": 128100
|
| 63 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eaea070db4574cbd17227d4d3af5838d33529749916280fa241994ee375ce290
|
| 3 |
+
size 741359744
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09076ea10a3d61e7c2e1df3f6041fa3881cd48d41cc55b7dd24e43bbfab79c04
|
| 3 |
+
size 1482843147
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd28b96146b0c352b1c54b6d9faac47cc39eb31de8b2a770e08db7f2e62d4d5b
|
| 3 |
+
size 14645
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d1f43e93fb7b4c7684484d3ae04ff8f2520e0f67f3b6d0d2947cb202aa52ee2
|
| 3 |
+
size 1465
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/trainer_state.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6915629322268326,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.6915629322268326,
|
| 14 |
+
"grad_norm": 0.06808845698833466,
|
| 15 |
+
"learning_rate": 1.3098201936376211e-05,
|
| 16 |
+
"loss": 0.012068235397338867,
|
| 17 |
+
"step": 500
|
| 18 |
+
}
|
| 19 |
+
],
|
| 20 |
+
"logging_steps": 500,
|
| 21 |
+
"max_steps": 1446,
|
| 22 |
+
"num_input_tokens_seen": 0,
|
| 23 |
+
"num_train_epochs": 2,
|
| 24 |
+
"save_steps": 500,
|
| 25 |
+
"stateful_callbacks": {
|
| 26 |
+
"TrainerControl": {
|
| 27 |
+
"args": {
|
| 28 |
+
"should_epoch_stop": false,
|
| 29 |
+
"should_evaluate": false,
|
| 30 |
+
"should_log": false,
|
| 31 |
+
"should_save": true,
|
| 32 |
+
"should_training_stop": false
|
| 33 |
+
},
|
| 34 |
+
"attributes": {}
|
| 35 |
+
}
|
| 36 |
+
},
|
| 37 |
+
"total_flos": 121295240486400.0,
|
| 38 |
+
"train_batch_size": 16,
|
| 39 |
+
"trial_name": null,
|
| 40 |
+
"trial_params": null
|
| 41 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e609457181ac50440df7d6728c3518d3b9316719fead1acbcffdb0f4941dd06
|
| 3 |
+
size 5201
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/config.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DebertaAdvancedSpanClassifier"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": null,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "O",
|
| 14 |
+
"1": "ADDRESS",
|
| 15 |
+
"2": "CREDIT_CARD",
|
| 16 |
+
"3": "CRYPTO_ADDRESS",
|
| 17 |
+
"4": "EMAIL",
|
| 18 |
+
"5": "IBAN",
|
| 19 |
+
"6": "OFF_PLATFORM_ATTEMPT",
|
| 20 |
+
"7": "PHONE",
|
| 21 |
+
"8": "SOCIAL_MEDIA",
|
| 22 |
+
"9": "URL"
|
| 23 |
+
},
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 3072,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"ADDRESS": 1,
|
| 28 |
+
"CREDIT_CARD": 2,
|
| 29 |
+
"CRYPTO_ADDRESS": 3,
|
| 30 |
+
"EMAIL": 4,
|
| 31 |
+
"IBAN": 5,
|
| 32 |
+
"O": 0,
|
| 33 |
+
"OFF_PLATFORM_ATTEMPT": 6,
|
| 34 |
+
"PHONE": 7,
|
| 35 |
+
"SOCIAL_MEDIA": 8,
|
| 36 |
+
"URL": 9
|
| 37 |
+
},
|
| 38 |
+
"layer_norm_eps": 1e-07,
|
| 39 |
+
"legacy": true,
|
| 40 |
+
"max_position_embeddings": 512,
|
| 41 |
+
"max_relative_positions": -1,
|
| 42 |
+
"model_type": "deberta-v2",
|
| 43 |
+
"norm_rel_ebd": "layer_norm",
|
| 44 |
+
"num_attention_heads": 12,
|
| 45 |
+
"num_hidden_layers": 12,
|
| 46 |
+
"pad_token_id": 0,
|
| 47 |
+
"pooler_dropout": 0,
|
| 48 |
+
"pooler_hidden_act": "gelu",
|
| 49 |
+
"pooler_hidden_size": 768,
|
| 50 |
+
"pos_att_type": [
|
| 51 |
+
"p2c",
|
| 52 |
+
"c2p"
|
| 53 |
+
],
|
| 54 |
+
"position_biased_input": false,
|
| 55 |
+
"position_buckets": 256,
|
| 56 |
+
"relative_attention": true,
|
| 57 |
+
"share_att_key": true,
|
| 58 |
+
"tie_word_embeddings": true,
|
| 59 |
+
"transformers_version": "5.3.0",
|
| 60 |
+
"type_vocab_size": 0,
|
| 61 |
+
"use_cache": false,
|
| 62 |
+
"vocab_size": 128100
|
| 63 |
+
}
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9d7fb44d2ab60a24cf4a3e1dce9b3baf23b551a5b131bda58d827ac494d5b1e
|
| 3 |
+
size 741421647
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
gorkem-pii-ready/SpanBased-CustomDeBERTa/tokenizer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": true,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "[CLS]",
|
| 5 |
+
"cls_token": "[CLS]",
|
| 6 |
+
"do_lower_case": false,
|
| 7 |
+
"eos_token": "[SEP]",
|
| 8 |
+
"extra_special_tokens": [
|
| 9 |
+
"[PAD]",
|
| 10 |
+
"[CLS]",
|
| 11 |
+
"[SEP]"
|
| 12 |
+
],
|
| 13 |
+
"is_local": false,
|
| 14 |
+
"mask_token": "[MASK]",
|
| 15 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 16 |
+
"pad_token": "[PAD]",
|
| 17 |
+
"sep_token": "[SEP]",
|
| 18 |
+
"split_by_punct": false,
|
| 19 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 20 |
+
"unk_id": 3,
|
| 21 |
+
"unk_token": "[UNK]",
|
| 22 |
+
"vocab_type": "spm"
|
| 23 |
+
}
|