update

.gitattributes

@@ -1,6 +1,7 @@
 *.7z filter=lfs diff=lfs merge=lfs -text
 *.arrow filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
 *.bz2 filter=lfs diff=lfs merge=lfs -text
 *.ckpt filter=lfs diff=lfs merge=lfs -text
 *.ftz filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,74 @@
+---
+license: mit
+tags:
+- generated_from_trainer
+metrics:
+- precision
+- recall
+- f1
+- accuracy
+model-index:
+- name: cv-ner
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# cv-ner
+
+This model is a fine-tuned version of [microsoft/mdeberta-v3-base](https://huggingface.co/microsoft/mdeberta-v3-base) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0956
+- Precision: 0.8906
+- Recall: 0.9325
+- F1: 0.9111
+- Accuracy: 0.9851
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-05
+- train_batch_size: 16
+- eval_batch_size: 4
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 10.0
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
+|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
+| No log        | 1.0   | 91   | 0.2049          | 0.6618    | 0.7362 | 0.6970 | 0.9534   |
+| 0.5036        | 2.0   | 182  | 0.1156          | 0.7873    | 0.8630 | 0.8234 | 0.9722   |
+| 0.1442        | 3.0   | 273  | 0.1078          | 0.8262    | 0.9039 | 0.8633 | 0.9771   |
+| 0.0757        | 4.0   | 364  | 0.1179          | 0.8652    | 0.9059 | 0.8851 | 0.9780   |
+| 0.0526        | 5.0   | 455  | 0.0907          | 0.888     | 0.9080 | 0.8979 | 0.9837   |
+| 0.0342        | 6.0   | 546  | 0.0972          | 0.8926    | 0.9346 | 0.9131 | 0.9832   |
+| 0.0245        | 7.0   | 637  | 0.1064          | 0.8937    | 0.9284 | 0.9107 | 0.9834   |
+| 0.0188        | 8.0   | 728  | 0.0965          | 0.8980    | 0.9366 | 0.9169 | 0.9850   |
+| 0.0159        | 9.0   | 819  | 0.0999          | 0.91      | 0.9305 | 0.9201 | 0.9846   |
+| 0.0141        | 10.0  | 910  | 0.0956          | 0.8906    | 0.9325 | 0.9111 | 0.9851   |
+
+
+### Framework versions
+
+- Transformers 4.24.0.dev0
+- Pytorch 1.12.1+cu113
+- Datasets 2.6.1
+- Tokenizers 0.13.1

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[MASK]": 250101
+}

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.985104873847401,
+    "eval_f1": 0.911088911088911,
+    "eval_loss": 0.09563781321048737,
+    "eval_precision": 0.890625,
+    "eval_recall": 0.9325153374233128,
+    "eval_runtime": 1.2647,
+    "eval_samples": 161,
+    "eval_samples_per_second": 127.3,
+    "eval_steps_per_second": 32.418,
+    "train_loss": 0.09724931471295409,
+    "train_runtime": 322.9196,
+    "train_samples": 1441,
+    "train_samples_per_second": 44.624,
+    "train_steps_per_second": 2.818
+}

config.json

@@ -0,0 +1,86 @@
+{
+  "_name_or_path": "microsoft/mdeberta-v3-base",
+  "architectures": [
+    "DebertaV2ForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "finetuning_task": "ner",
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "B-ADDRESS",
+    "1": "B-COMPANY",
+    "2": "B-DOB",
+    "3": "B-EMAIL",
+    "4": "B-FULLNAME",
+    "5": "B-GENDER",
+    "6": "B-MAJOR",
+    "7": "B-PHONE_NUMBER",
+    "8": "B-POSITION",
+    "9": "B-SCHOOL",
+    "10": "B-TIME",
+    "11": "I-ADDRESS",
+    "12": "I-COMPANY",
+    "13": "I-DOB",
+    "14": "I-EMAIL",
+    "15": "I-FULLNAME",
+    "16": "I-GENDER",
+    "17": "I-MAJOR",
+    "18": "I-PHONE_NUMBER",
+    "19": "I-POSITION",
+    "20": "I-SCHOOL",
+    "21": "I-TIME",
+    "22": "O"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-ADDRESS": 0,
+    "B-COMPANY": 1,
+    "B-DOB": 2,
+    "B-EMAIL": 3,
+    "B-FULLNAME": 4,
+    "B-GENDER": 5,
+    "B-MAJOR": 6,
+    "B-PHONE_NUMBER": 7,
+    "B-POSITION": 8,
+    "B-SCHOOL": 9,
+    "B-TIME": 10,
+    "I-ADDRESS": 11,
+    "I-COMPANY": 12,
+    "I-DOB": 13,
+    "I-EMAIL": 14,
+    "I-FULLNAME": 15,
+    "I-GENDER": 16,
+    "I-MAJOR": 17,
+    "I-PHONE_NUMBER": 18,
+    "I-POSITION": 19,
+    "I-SCHOOL": 20,
+    "I-TIME": 21,
+    "O": 22
+  },
+  "layer_norm_eps": 1e-07,
+  "max_position_embeddings": 512,
+  "max_relative_positions": -1,
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 768,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
+  "relative_attention": true,
+  "share_att_key": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.24.0.dev0",
+  "type_vocab_size": 0,
+  "vocab_size": 251000
+}

eval_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 10.0,
+    "eval_accuracy": 0.985104873847401,
+    "eval_f1": 0.911088911088911,
+    "eval_loss": 0.09563781321048737,
+    "eval_precision": 0.890625,
+    "eval_recall": 0.9325153374233128,
+    "eval_runtime": 1.2647,
+    "eval_samples": 161,
+    "eval_samples_per_second": 127.3,
+    "eval_steps_per_second": 32.418
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5cc9b7eb748d73b8e5d1d1aa9ade4051451efbd33c0d050d45af8b8bd2553f8
+size 1113017327

special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13c8d666d62a7bc4ac8f040aab68e942c861f93303156cc28f5c7e885d86d6e3
+size 4305025

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbef9712c55ef75d0004007743c550a957b55a8f094bec9f147c42dc093ab471
+size 16331566

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "name_or_path": "microsoft/mdeberta-v3-base",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "special_tokens_map_file": null,
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 10.0,
+    "train_loss": 0.09724931471295409,
+    "train_runtime": 322.9196,
+    "train_samples": 1441,
+    "train_samples_per_second": 44.624,
+    "train_steps_per_second": 2.818
+}

trainer_state.json

@@ -0,0 +1,199 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "global_step": 910,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9533894011551323,
+      "eval_f1": 0.6969990319457888,
+      "eval_loss": 0.20486080646514893,
+      "eval_precision": 0.6617647058823529,
+      "eval_recall": 0.7361963190184049,
+      "eval_runtime": 1.2373,
+      "eval_samples_per_second": 130.123,
+      "eval_steps_per_second": 33.137,
+      "step": 91
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.4505494505494504e-05,
+      "loss": 0.5036,
+      "step": 100
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9722362954706657,
+      "eval_f1": 0.8234146341463414,
+      "eval_loss": 0.11564752459526062,
+      "eval_precision": 0.7873134328358209,
+      "eval_recall": 0.8629856850715747,
+      "eval_runtime": 1.2367,
+      "eval_samples_per_second": 130.181,
+      "eval_steps_per_second": 33.152,
+      "step": 182
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.901098901098901e-05,
+      "loss": 0.1442,
+      "step": 200
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9771000101327388,
+      "eval_f1": 0.8632812499999999,
+      "eval_loss": 0.10779214650392532,
+      "eval_precision": 0.8261682242990654,
+      "eval_recall": 0.9038854805725971,
+      "eval_runtime": 1.2508,
+      "eval_samples_per_second": 128.715,
+      "eval_steps_per_second": 32.778,
+      "step": 273
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.3516483516483513e-05,
+      "loss": 0.0757,
+      "step": 300
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9780119566318776,
+      "eval_f1": 0.8851148851148851,
+      "eval_loss": 0.1178947389125824,
+      "eval_precision": 0.865234375,
+      "eval_recall": 0.9059304703476483,
+      "eval_runtime": 1.2453,
+      "eval_samples_per_second": 129.288,
+      "eval_steps_per_second": 32.924,
+      "step": 364
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.8021978021978025e-05,
+      "loss": 0.0526,
+      "step": 400
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9836862904042962,
+      "eval_f1": 0.897876643073812,
+      "eval_loss": 0.0906740128993988,
+      "eval_precision": 0.888,
+      "eval_recall": 0.9079754601226994,
+      "eval_runtime": 1.2584,
+      "eval_samples_per_second": 127.94,
+      "eval_steps_per_second": 32.581,
+      "step": 455
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.252747252747253e-05,
+      "loss": 0.0342,
+      "step": 500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9831796534603303,
+      "eval_f1": 0.9130869130869131,
+      "eval_loss": 0.0972292423248291,
+      "eval_precision": 0.892578125,
+      "eval_recall": 0.934560327198364,
+      "eval_runtime": 1.2748,
+      "eval_samples_per_second": 126.296,
+      "eval_steps_per_second": 32.162,
+      "step": 546
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 1.7032967032967035e-05,
+      "loss": 0.0245,
+      "step": 600
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9833823082379167,
+      "eval_f1": 0.9107321965897693,
+      "eval_loss": 0.10643576085567474,
+      "eval_precision": 0.8937007874015748,
+      "eval_recall": 0.9284253578732107,
+      "eval_runtime": 1.2478,
+      "eval_samples_per_second": 129.031,
+      "eval_steps_per_second": 32.859,
+      "step": 637
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 0.0188,
+      "step": 700
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9850035464586078,
+      "eval_f1": 0.9169169169169169,
+      "eval_loss": 0.09653466939926147,
+      "eval_precision": 0.8980392156862745,
+      "eval_recall": 0.9366053169734151,
+      "eval_runtime": 1.2451,
+      "eval_samples_per_second": 129.306,
+      "eval_steps_per_second": 32.929,
+      "step": 728
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 6.043956043956044e-06,
+      "loss": 0.0159,
+      "step": 800
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.984598236903435,
+      "eval_f1": 0.9201213346814965,
+      "eval_loss": 0.09986742585897446,
+      "eval_precision": 0.91,
+      "eval_recall": 0.9304703476482618,
+      "eval_runtime": 1.2461,
+      "eval_samples_per_second": 129.208,
+      "eval_steps_per_second": 32.904,
+      "step": 819
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5.494505494505495e-07,
+      "loss": 0.0141,
+      "step": 900
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.985104873847401,
+      "eval_f1": 0.911088911088911,
+      "eval_loss": 0.09563781321048737,
+      "eval_precision": 0.890625,
+      "eval_recall": 0.9325153374233128,
+      "eval_runtime": 1.2493,
+      "eval_samples_per_second": 128.877,
+      "eval_steps_per_second": 32.819,
+      "step": 910
+    },
+    {
+      "epoch": 10.0,
+      "step": 910,
+      "total_flos": 1883034567828480.0,
+      "train_loss": 0.09724931471295409,
+      "train_runtime": 322.9196,
+      "train_samples_per_second": 44.624,
+      "train_steps_per_second": 2.818
+    }
+  ],
+  "max_steps": 910,
+  "num_train_epochs": 10,
+  "total_flos": 1883034567828480.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:872b0c9a30857c46bfc90e2d51e67a262aea692254c5bde161bebb6cafe01c96
+size 3375