Training in progress, epoch 5, checkpoint

checkpoint-235/config.json

@@ -0,0 +1,43 @@
+{
+  "add_cross_attention": false,
+  "architectures": [
+    "ElectraForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": null,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "embedding_size": 128,
+  "eos_token_id": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 256,
+  "id2label": {
+    "0": "valid",
+    "1": "fallacy"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 1024,
+  "is_decoder": false,
+  "label2id": {
+    "fallacy": 1,
+    "valid": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "electra",
+  "num_attention_heads": 4,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.9.0",
+  "type_vocab_size": 2,
+  "use_cache": false,
+  "vocab_size": 30522
+}

checkpoint-235/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6f6a8505c36e37b1b5fcd38d4455a343639019d4e6f70f57493b8dc37fed351
+size 54221200

checkpoint-235/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5853af63ceaea0b2d7b3ed21127b9b6e8ae5ecbece3d5738b4661f18ada1d05
+size 108567563

checkpoint-235/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:678e4e519fa896a5f9c4841fa19a1e4dfd8dbe937f178c5f71171f52ac8d6a65
+size 14645

checkpoint-235/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be0e932021d58d3c46733ca20dc12b546e8bd64bc0a440d710adc37c34a403eb
+size 1383

checkpoint-235/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44fb94a88c15d58ffa1bf978fccff6134483f7613a240fb8f9d498a06f59dd4a
+size 1465

checkpoint-235/tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "backend": "tokenizers",
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

checkpoint-235/trainer_state.json

@@ -0,0 +1,250 @@
+{
+  "best_global_step": 235,
+  "best_metric": 0.65996599659966,
+  "best_model_checkpoint": "models/electra-small-touche-base-binary/trainer/checkpoint-235",
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 0.7140693068504333,
+      "learning_rate": 1.125e-05,
+      "loss": 0.69647216796875,
+      "step": 10
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 1.1431223154067993,
+      "learning_rate": 2.3749999999999998e-05,
+      "loss": 0.6931365966796875,
+      "step": 20
+    },
+    {
+      "epoch": 0.6382978723404256,
+      "grad_norm": 0.6381158828735352,
+      "learning_rate": 2.928909952606635e-05,
+      "loss": 0.69478759765625,
+      "step": 30
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 1.9318883419036865,
+      "learning_rate": 2.7867298578199053e-05,
+      "loss": 0.6951080322265625,
+      "step": 40
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.555,
+      "eval_fallacy_f1": 0.3597122302158273,
+      "eval_loss": 0.6898486614227295,
+      "eval_macro_f1": 0.5093580308167259,
+      "eval_runtime": 0.1669,
+      "eval_samples_per_second": 1198.414,
+      "eval_steps_per_second": 23.968,
+      "step": 47
+    },
+    {
+      "epoch": 1.0638297872340425,
+      "grad_norm": 0.6443229913711548,
+      "learning_rate": 2.6445497630331753e-05,
+      "loss": 0.6960784912109375,
+      "step": 50
+    },
+    {
+      "epoch": 1.2765957446808511,
+      "grad_norm": 1.3156065940856934,
+      "learning_rate": 2.5023696682464456e-05,
+      "loss": 0.6871307373046875,
+      "step": 60
+    },
+    {
+      "epoch": 1.4893617021276595,
+      "grad_norm": 0.9663533568382263,
+      "learning_rate": 2.360189573459716e-05,
+      "loss": 0.69168701171875,
+      "step": 70
+    },
+    {
+      "epoch": 1.702127659574468,
+      "grad_norm": 0.7499622702598572,
+      "learning_rate": 2.2180094786729858e-05,
+      "loss": 0.6835113525390625,
+      "step": 80
+    },
+    {
+      "epoch": 1.9148936170212765,
+      "grad_norm": 1.1466385126113892,
+      "learning_rate": 2.075829383886256e-05,
+      "loss": 0.6876190185546875,
+      "step": 90
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.64,
+      "eval_fallacy_f1": 0.64,
+      "eval_loss": 0.6786279082298279,
+      "eval_macro_f1": 0.64,
+      "eval_runtime": 0.2441,
+      "eval_samples_per_second": 819.224,
+      "eval_steps_per_second": 16.384,
+      "step": 94
+    },
+    {
+      "epoch": 2.127659574468085,
+      "grad_norm": 0.7131018042564392,
+      "learning_rate": 1.933649289099526e-05,
+      "loss": 0.6760498046875,
+      "step": 100
+    },
+    {
+      "epoch": 2.3404255319148937,
+      "grad_norm": 0.8485779762268066,
+      "learning_rate": 1.791469194312796e-05,
+      "loss": 0.67723388671875,
+      "step": 110
+    },
+    {
+      "epoch": 2.5531914893617023,
+      "grad_norm": 1.0046956539154053,
+      "learning_rate": 1.6492890995260666e-05,
+      "loss": 0.6680267333984375,
+      "step": 120
+    },
+    {
+      "epoch": 2.7659574468085104,
+      "grad_norm": 1.3051499128341675,
+      "learning_rate": 1.5071090047393365e-05,
+      "loss": 0.6558563232421875,
+      "step": 130
+    },
+    {
+      "epoch": 2.978723404255319,
+      "grad_norm": 1.3270968198776245,
+      "learning_rate": 1.3649289099526066e-05,
+      "loss": 0.6429000854492187,
+      "step": 140
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.63,
+      "eval_fallacy_f1": 0.5432098765432098,
+      "eval_loss": 0.6571618914604187,
+      "eval_macro_f1": 0.6161427533976553,
+      "eval_runtime": 0.1787,
+      "eval_samples_per_second": 1119.018,
+      "eval_steps_per_second": 22.38,
+      "step": 141
+    },
+    {
+      "epoch": 3.1914893617021276,
+      "grad_norm": 1.5919787883758545,
+      "learning_rate": 1.2227488151658767e-05,
+      "loss": 0.6370620727539062,
+      "step": 150
+    },
+    {
+      "epoch": 3.404255319148936,
+      "grad_norm": 1.217308759689331,
+      "learning_rate": 1.080568720379147e-05,
+      "loss": 0.6328628540039063,
+      "step": 160
+    },
+    {
+      "epoch": 3.617021276595745,
+      "grad_norm": 2.489976406097412,
+      "learning_rate": 9.383886255924171e-06,
+      "loss": 0.6132949829101563,
+      "step": 170
+    },
+    {
+      "epoch": 3.829787234042553,
+      "grad_norm": 1.557442307472229,
+      "learning_rate": 7.962085308056872e-06,
+      "loss": 0.6257461547851563,
+      "step": 180
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.645,
+      "eval_fallacy_f1": 0.6697674418604651,
+      "eval_loss": 0.6352868676185608,
+      "eval_macro_f1": 0.6429918290383407,
+      "eval_runtime": 0.1618,
+      "eval_samples_per_second": 1236.069,
+      "eval_steps_per_second": 24.721,
+      "step": 188
+    },
+    {
+      "epoch": 4.042553191489362,
+      "grad_norm": 2.8439695835113525,
+      "learning_rate": 6.5402843601895735e-06,
+      "loss": 0.60234375,
+      "step": 190
+    },
+    {
+      "epoch": 4.25531914893617,
+      "grad_norm": 1.99105703830719,
+      "learning_rate": 5.1184834123222755e-06,
+      "loss": 0.6104934692382813,
+      "step": 200
+    },
+    {
+      "epoch": 4.468085106382979,
+      "grad_norm": 1.5985376834869385,
+      "learning_rate": 3.696682464454976e-06,
+      "loss": 0.5836532592773438,
+      "step": 210
+    },
+    {
+      "epoch": 4.680851063829787,
+      "grad_norm": 1.6915240287780762,
+      "learning_rate": 2.274881516587678e-06,
+      "loss": 0.5967620849609375,
+      "step": 220
+    },
+    {
+      "epoch": 4.8936170212765955,
+      "grad_norm": 1.8389923572540283,
+      "learning_rate": 8.530805687203791e-07,
+      "loss": 0.5855438232421875,
+      "step": 230
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.66,
+      "eval_fallacy_f1": 0.6633663366336634,
+      "eval_loss": 0.6288647651672363,
+      "eval_macro_f1": 0.65996599659966,
+      "eval_runtime": 0.1577,
+      "eval_samples_per_second": 1268.019,
+      "eval_steps_per_second": 25.36,
+      "step": 235
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 40353289458552.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-235/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:feed2b83d33087270d16ed7f3e10fa99ab935ebc4c954f924db9b9235a327ec4
+size 5393