Training in progress, epoch 3, checkpoint

checkpoint-300/config.json

@@ -0,0 +1,33 @@
+{
+  "_name_or_path": "microsoft/deberta-large",
+  "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": "4.48.3",
+  "type_vocab_size": 0,
+  "vocab_size": 50265
+}

checkpoint-300/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bad678d67e21a5f8ff39b1f5d9235030da01967807bd9ef3f3d811a014d2541
+size 1624907048

checkpoint-300/trainer_state.json

@@ -0,0 +1,171 @@
+{
+  "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": 9.394545555114746,
+      "learning_rate": 2.75e-05,
+      "loss": 0.52,
+      "step": 50
+    },
+    {
+      "epoch": 0.5,
+      "eval_pet_AUC": 0.9145265189137036,
+      "eval_pet_accuracy": 0.845,
+      "eval_pet_f1": 0.7703703703703704,
+      "eval_pet_loss": 0.3643406927585602,
+      "eval_pet_macro_f1": 0.8266946191474493,
+      "eval_pet_micro_f1": 0.845,
+      "eval_pet_precision": 0.7482014388489209,
+      "eval_pet_recall": 0.7938931297709924,
+      "eval_pet_runtime": 0.9635,
+      "eval_pet_samples_per_second": 415.172,
+      "eval_pet_specificity": 0.8698884758364313,
+      "eval_pet_steps_per_second": 1.038,
+      "step": 50
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.517026901245117,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4033,
+      "step": 100
+    },
+    {
+      "epoch": 1.0,
+      "eval_pet_AUC": 0.9307017792786401,
+      "eval_pet_accuracy": 0.85,
+      "eval_pet_f1": 0.7744360902255639,
+      "eval_pet_loss": 0.32512709498405457,
+      "eval_pet_macro_f1": 0.831038269831883,
+      "eval_pet_micro_f1": 0.85,
+      "eval_pet_precision": 0.762962962962963,
+      "eval_pet_recall": 0.7862595419847328,
+      "eval_pet_runtime": 0.9568,
+      "eval_pet_samples_per_second": 418.039,
+      "eval_pet_specificity": 0.8810408921933085,
+      "eval_pet_steps_per_second": 1.045,
+      "step": 100
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.107645034790039,
+      "learning_rate": 2.25e-05,
+      "loss": 0.2658,
+      "step": 150
+    },
+    {
+      "epoch": 1.5,
+      "eval_pet_AUC": 0.945259513607083,
+      "eval_pet_accuracy": 0.855,
+      "eval_pet_f1": 0.7943262411347518,
+      "eval_pet_loss": 0.375436931848526,
+      "eval_pet_macro_f1": 0.8411785645828199,
+      "eval_pet_micro_f1": 0.855,
+      "eval_pet_precision": 0.7417218543046358,
+      "eval_pet_recall": 0.8549618320610687,
+      "eval_pet_runtime": 0.9663,
+      "eval_pet_samples_per_second": 413.932,
+      "eval_pet_specificity": 0.8550185873605948,
+      "eval_pet_steps_per_second": 1.035,
+      "step": 150
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.23839282989502,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.2444,
+      "step": 200
+    },
+    {
+      "epoch": 2.0,
+      "eval_pet_AUC": 0.9240614092340873,
+      "eval_pet_accuracy": 0.845,
+      "eval_pet_f1": 0.7785714285714286,
+      "eval_pet_loss": 0.38033241033554077,
+      "eval_pet_macro_f1": 0.8296703296703296,
+      "eval_pet_micro_f1": 0.845,
+      "eval_pet_precision": 0.7315436241610739,
+      "eval_pet_recall": 0.8320610687022901,
+      "eval_pet_runtime": 0.9687,
+      "eval_pet_samples_per_second": 412.922,
+      "eval_pet_specificity": 0.8513011152416357,
+      "eval_pet_steps_per_second": 1.032,
+      "step": 200
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.208889961242676,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.1705,
+      "step": 250
+    },
+    {
+      "epoch": 2.5,
+      "eval_pet_AUC": 0.9323760606146599,
+      "eval_pet_accuracy": 0.8625,
+      "eval_pet_f1": 0.7639484978540773,
+      "eval_pet_loss": 0.6355132460594177,
+      "eval_pet_macro_f1": 0.8334733670928234,
+      "eval_pet_micro_f1": 0.8625,
+      "eval_pet_precision": 0.8725490196078431,
+      "eval_pet_recall": 0.6793893129770993,
+      "eval_pet_runtime": 0.9639,
+      "eval_pet_samples_per_second": 415.002,
+      "eval_pet_specificity": 0.9516728624535316,
+      "eval_pet_steps_per_second": 1.038,
+      "step": 250
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 9.768027305603027,
+      "learning_rate": 1.5e-05,
+      "loss": 0.1972,
+      "step": 300
+    },
+    {
+      "epoch": 3.0,
+      "eval_pet_AUC": 0.9315531087715316,
+      "eval_pet_accuracy": 0.8525,
+      "eval_pet_f1": 0.7773584905660378,
+      "eval_pet_loss": 0.43593117594718933,
+      "eval_pet_macro_f1": 0.833539058367131,
+      "eval_pet_micro_f1": 0.8525,
+      "eval_pet_precision": 0.7686567164179104,
+      "eval_pet_recall": 0.7862595419847328,
+      "eval_pet_runtime": 0.9547,
+      "eval_pet_samples_per_second": 418.986,
+      "eval_pet_specificity": 0.8847583643122676,
+      "eval_pet_steps_per_second": 1.047,
+      "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 sha256:782038bc3d4de4077e2809a63f4570c5ba00c054cf84c7f0402dfaeb26a11ccd
+size 5432