Training in progress, epoch 3, checkpoint

checkpoint-300/config.json

@@ -0,0 +1,32 @@
+{
+  "architectures": [
+    "DebertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-07,
+  "legacy": true,
+  "max_position_embeddings": 512,
+  "max_relative_positions": -1,
+  "model_type": "deberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 1024,
+  "pos_att_type": [
+    "c2p",
+    "p2c"
+  ],
+  "position_biased_input": false,
+  "relative_attention": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.50.3",
+  "type_vocab_size": 0,
+  "vocab_size": 50265
+}

checkpoint-300/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ae3c099612f4322af8e0822bc932a7fcbdf543f4c6a7b03e08be25e23424b6a
+size 1624907048

checkpoint-300/trainer_state.json

@@ -0,0 +1,172 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 50,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.202085256576538,
+      "learning_rate": 2.75e-05,
+      "loss": 0.7001,
+      "step": 50
+    },
+    {
+      "epoch": 0.5,
+      "eval_electronics_AUC": 0.8274661508704062,
+      "eval_electronics_accuracy": 0.5875,
+      "eval_electronics_f1": 0.0,
+      "eval_electronics_loss": 0.7029502987861633,
+      "eval_electronics_macro_f1": 0.3700787401574803,
+      "eval_electronics_micro_f1": 0.5875,
+      "eval_electronics_precision": 0.0,
+      "eval_electronics_recall": 0.0,
+      "eval_electronics_runtime": 1.1511,
+      "eval_electronics_samples_per_second": 347.485,
+      "eval_electronics_specificity": 1.0,
+      "eval_electronics_steps_per_second": 0.869,
+      "step": 50
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.9589738845825195,
+      "learning_rate": 2.5e-05,
+      "loss": 0.6259,
+      "step": 100
+    },
+    {
+      "epoch": 1.0,
+      "eval_electronics_AUC": 0.8812894906511928,
+      "eval_electronics_accuracy": 0.805,
+      "eval_electronics_f1": 0.7914438502673797,
+      "eval_electronics_loss": 0.44977375864982605,
+      "eval_electronics_macro_f1": 0.804172629359042,
+      "eval_electronics_micro_f1": 0.805,
+      "eval_electronics_precision": 0.7081339712918661,
+      "eval_electronics_recall": 0.896969696969697,
+      "eval_electronics_runtime": 1.1287,
+      "eval_electronics_samples_per_second": 354.379,
+      "eval_electronics_specificity": 0.7404255319148936,
+      "eval_electronics_steps_per_second": 0.886,
+      "step": 100
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.954824447631836,
+      "learning_rate": 2.25e-05,
+      "loss": 0.5285,
+      "step": 150
+    },
+    {
+      "epoch": 1.5,
+      "eval_electronics_AUC": 0.8546228239845262,
+      "eval_electronics_accuracy": 0.75,
+      "eval_electronics_f1": 0.647887323943662,
+      "eval_electronics_loss": 0.5688791275024414,
+      "eval_electronics_macro_f1": 0.7270444371656295,
+      "eval_electronics_micro_f1": 0.75,
+      "eval_electronics_precision": 0.773109243697479,
+      "eval_electronics_recall": 0.5575757575757576,
+      "eval_electronics_runtime": 1.1417,
+      "eval_electronics_samples_per_second": 350.348,
+      "eval_electronics_specificity": 0.8851063829787233,
+      "eval_electronics_steps_per_second": 0.876,
+      "step": 150
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.80739426612854,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.492,
+      "step": 200
+    },
+    {
+      "epoch": 2.0,
+      "eval_electronics_AUC": 0.8878401031592521,
+      "eval_electronics_accuracy": 0.79,
+      "eval_electronics_f1": 0.7741935483870968,
+      "eval_electronics_loss": 0.5164185762405396,
+      "eval_electronics_macro_f1": 0.7889659330720531,
+      "eval_electronics_micro_f1": 0.79,
+      "eval_electronics_precision": 0.6956521739130435,
+      "eval_electronics_recall": 0.8727272727272727,
+      "eval_electronics_runtime": 1.1374,
+      "eval_electronics_samples_per_second": 351.668,
+      "eval_electronics_specificity": 0.7319148936170212,
+      "eval_electronics_steps_per_second": 0.879,
+      "step": 200
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.75403594970703,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.4042,
+      "step": 250
+    },
+    {
+      "epoch": 2.5,
+      "eval_electronics_AUC": 0.8810831721470018,
+      "eval_electronics_accuracy": 0.7525,
+      "eval_electronics_f1": 0.7579462102689487,
+      "eval_electronics_loss": 0.6417726874351501,
+      "eval_electronics_macro_f1": 0.7523746396613286,
+      "eval_electronics_micro_f1": 0.7525,
+      "eval_electronics_precision": 0.6352459016393442,
+      "eval_electronics_recall": 0.9393939393939394,
+      "eval_electronics_runtime": 1.176,
+      "eval_electronics_samples_per_second": 340.145,
+      "eval_electronics_specificity": 0.6212765957446809,
+      "eval_electronics_steps_per_second": 0.85,
+      "step": 250
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 17.9029541015625,
+      "learning_rate": 1.5e-05,
+      "loss": 0.3676,
+      "step": 300
+    },
+    {
+      "epoch": 3.0,
+      "eval_electronics_AUC": 0.8895422308188267,
+      "eval_electronics_accuracy": 0.81,
+      "eval_electronics_f1": 0.7432432432432432,
+      "eval_electronics_loss": 0.5099908709526062,
+      "eval_electronics_macro_f1": 0.7962247962247961,
+      "eval_electronics_micro_f1": 0.81,
+      "eval_electronics_precision": 0.8396946564885496,
+      "eval_electronics_recall": 0.6666666666666666,
+      "eval_electronics_runtime": 1.1391,
+      "eval_electronics_samples_per_second": 351.16,
+      "eval_electronics_specificity": 0.9106382978723404,
+      "eval_electronics_steps_per_second": 0.878,
+      "step": 300
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 407456319744000.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-300/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47871193157119808f8d00b7d97c1b2cdfffd44e2a841575fc1c03ac79100709
+size 5432