updating the repo with the fine-tuned model

README.md

@@ -0,0 +1,73 @@
+---
+license: apache-2.0
+tags:
+- generated_from_trainer
+metrics:
+- precision
+- recall
+- f1
+- accuracy
+model-index:
+- name: atco2_test_set_1h
+  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. -->
+
+# atco2_test_set_1h
+
+This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.4282
+- Precision: 0.6195
+- Recall: 0.7071
+- F1: 0.6604
+- Accuracy: 0.8182
+
+## 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: 32
+- eval_batch_size: 16
+- seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 64
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 500
+- training_steps: 3000
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
+|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
+| No log        | 125.0 | 500  | 0.8692          | 0.6396    | 0.7172 | 0.6762 | 0.8307   |
+| 0.2158        | 250.0 | 1000 | 1.0074          | 0.5702    | 0.6970 | 0.6273 | 0.8245   |
+| 0.2158        | 375.0 | 1500 | 1.3560          | 0.6577    | 0.7374 | 0.6952 | 0.8119   |
+| 0.0184        | 500.0 | 2000 | 1.3393          | 0.6182    | 0.6869 | 0.6507 | 0.8056   |
+| 0.0184        | 625.0 | 2500 | 1.3528          | 0.6087    | 0.7071 | 0.6542 | 0.8213   |
+| 0.0175        | 750.0 | 3000 | 1.4282          | 0.6195    | 0.7071 | 0.6604 | 0.8182   |
+
+
+### Framework versions
+
+- Transformers 4.24.0
+- Pytorch 1.13.0+cu117
+- Datasets 2.7.0
+- Tokenizers 0.13.2

all_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 750.0,
+    "train_loss": 0.0839117234547933,
+    "train_runtime": 334.6639,
+    "train_samples_per_second": 573.71,
+    "train_steps_per_second": 8.964
+}

classification_report

@@ -0,0 +1,34 @@
+	************* Report B/I tags*************	
+              precision    recall  f1-score   support
+
+         B-O       0.71      0.61      0.66      3106
+  B-callsign       0.85      0.89      0.87      2951
+   B-command       0.69      0.73      0.71      2357
+     B-value       0.58      0.55      0.56      3055
+         I-O       0.73      0.53      0.61      5403
+  I-callsign       0.92      0.92      0.92      8397
+   I-command       0.62      0.71      0.66      2795
+     I-value       0.73      0.85      0.78      7817
+
+    accuracy                           0.76     35881
+   macro avg       0.73      0.72      0.72     35881
+weighted avg       0.76      0.76      0.75     35881
+
+ 	************ Report with merged classes ***********	
+              precision    recall  f1-score   support
+
+           O       0.80      0.63      0.70      8509
+    callsign       0.93      0.95      0.94     11348
+     command       0.70      0.78      0.74      5152
+       value       0.77      0.85      0.81     10872
+
+    accuracy                           0.82     35881
+   macro avg       0.80      0.80      0.80     35881
+weighted avg       0.82      0.82      0.81     35881
+
+
+ JACCARD ERROR RATE (JER): [51.10824742 22.8462217  44.672      60.66494966 55.62913907 14.40397351
+ 50.74626866 35.74402169]
+ JER - WEIGHTED          : 37.89041510368749 
+
+

config.json

@@ -0,0 +1,48 @@
+{
+  "_name_or_path": "experiments/results/ner/baseline/bert-base-uncased/1234/atco2_test_set_1h//",
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "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": "B-O",
+    "1": "O",
+    "2": "I-O",
+    "3": "B-value",
+    "4": "I-value",
+    "5": "B-callsign",
+    "6": "B-command",
+    "7": "I-callsign",
+    "8": "I-command"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "B-O": 0,
+    "B-callsign": 5,
+    "B-command": 6,
+    "B-value": 3,
+    "I-O": 2,
+    "I-callsign": 7,
+    "I-command": 8,
+    "I-value": 4,
+    "O": 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",
+  "torch_dtype": "float32",
+  "transformers_version": "4.24.0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d324a065c486a170865e72c584a817114112ad8c3fe38a4bd31527f72f62887
+size 435663597

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "name_or_path": "experiments/results/ner/baseline/bert-base-uncased/1234/atco2_test_set_1h//",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 750.0,
+    "train_loss": 0.0839117234547933,
+    "train_runtime": 334.6639,
+    "train_samples_per_second": 573.71,
+    "train_steps_per_second": 8.964
+}

trainer_state.json

@@ -0,0 +1,115 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 750.0,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 125.0,
+      "eval_accuracy": 0.8307210031347962,
+      "eval_f1": 0.6761904761904761,
+      "eval_loss": 0.869185745716095,
+      "eval_precision": 0.6396396396396397,
+      "eval_recall": 0.7171717171717171,
+      "eval_runtime": 1.1326,
+      "eval_samples_per_second": 24.722,
+      "eval_steps_per_second": 1.766,
+      "step": 500
+    },
+    {
+      "epoch": 250.0,
+      "learning_rate": 4e-05,
+      "loss": 0.2158,
+      "step": 1000
+    },
+    {
+      "epoch": 250.0,
+      "eval_accuracy": 0.8244514106583072,
+      "eval_f1": 0.6272727272727273,
+      "eval_loss": 1.0074251890182495,
+      "eval_precision": 0.5702479338842975,
+      "eval_recall": 0.696969696969697,
+      "eval_runtime": 0.5366,
+      "eval_samples_per_second": 52.181,
+      "eval_steps_per_second": 3.727,
+      "step": 1000
+    },
+    {
+      "epoch": 375.0,
+      "eval_accuracy": 0.8119122257053292,
+      "eval_f1": 0.6952380952380952,
+      "eval_loss": 1.3559716939926147,
+      "eval_precision": 0.6576576576576577,
+      "eval_recall": 0.7373737373737373,
+      "eval_runtime": 0.5707,
+      "eval_samples_per_second": 49.06,
+      "eval_steps_per_second": 3.504,
+      "step": 1500
+    },
+    {
+      "epoch": 500.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 2000
+    },
+    {
+      "epoch": 500.0,
+      "eval_accuracy": 0.8056426332288401,
+      "eval_f1": 0.6507177033492823,
+      "eval_loss": 1.339294195175171,
+      "eval_precision": 0.6181818181818182,
+      "eval_recall": 0.6868686868686869,
+      "eval_runtime": 0.5948,
+      "eval_samples_per_second": 47.076,
+      "eval_steps_per_second": 3.363,
+      "step": 2000
+    },
+    {
+      "epoch": 625.0,
+      "eval_accuracy": 0.8213166144200627,
+      "eval_f1": 0.6542056074766355,
+      "eval_loss": 1.3527517318725586,
+      "eval_precision": 0.6086956521739131,
+      "eval_recall": 0.7070707070707071,
+      "eval_runtime": 0.5373,
+      "eval_samples_per_second": 52.109,
+      "eval_steps_per_second": 3.722,
+      "step": 2500
+    },
+    {
+      "epoch": 750.0,
+      "learning_rate": 0.0,
+      "loss": 0.0175,
+      "step": 3000
+    },
+    {
+      "epoch": 750.0,
+      "eval_accuracy": 0.8181818181818182,
+      "eval_f1": 0.660377358490566,
+      "eval_loss": 1.4282338619232178,
+      "eval_precision": 0.6194690265486725,
+      "eval_recall": 0.7070707070707071,
+      "eval_runtime": 0.5599,
+      "eval_samples_per_second": 50.01,
+      "eval_steps_per_second": 3.572,
+      "step": 3000
+    },
+    {
+      "epoch": 750.0,
+      "step": 3000,
+      "total_flos": 3253656728250000.0,
+      "train_loss": 0.0839117234547933,
+      "train_runtime": 334.6639,
+      "train_samples_per_second": 573.71,
+      "train_steps_per_second": 8.964
+    }
+  ],
+  "max_steps": 3000,
+  "num_train_epochs": 750,
+  "total_flos": 3253656728250000.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0608652fbed1fef6adc00256e46dbf72205c353bc0368a023a085bff3e9d020c
+size 3451