Training in progress, epoch 8

checkpoint-1890/config.json

@@ -0,0 +1,41 @@
+{
+  "_name_or_path": "ynie/electra-large-discriminator-snli_mnli_fever_anli_R1_R2_R3-nli",
+  "architectures": [
+    "ElectraForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "embedding_size": 1024,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "entailment",
+    "1": "neutral",
+    "2": "contradiction"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "contradiction": 2,
+    "entailment": 0,
+    "neutral": 1
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "electra",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "multi_label_classification",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.17.0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

checkpoint-1890/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13d2188589d5e440f89aac549891c91190c44add3ba29e702a5208bd3c8db9af
+size 2681386461

checkpoint-1890/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e742a3a54a5b90c9d5d5c36842221c5911704b73b0f3c88fb76e04cd4f6ae38
+size 1340745837

checkpoint-1890/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7eb5cbf4378adb59bda0a91794c54519073600b5a4bed55a5d93a952cb160812
+size 14503

checkpoint-1890/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d8b557902c70dd900c18634e2e38ce5ed11a803cb71ada011bcb53cb0a166fe
+size 559

checkpoint-1890/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff5b8bb070e67ec665ad97fd5084cc54f04a84f1488b08d9e54380c950cace67
+size 623

checkpoint-1890/special_tokens_map.json

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

checkpoint-1890/tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "normalization": true, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "google/electra-large-discriminator", "tokenizer_class": "ElectraTokenizer"}

checkpoint-1890/trainer_state.json

@@ -0,0 +1,160 @@
+{
+  "best_metric": 0.3247086703777313,
+  "best_model_checkpoint": "outputs/electra-nli-efl/checkpoint-1050",
+  "epoch": 8.995555555555555,
+  "global_step": 1890,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.499999999999999e-07,
+      "loss": 0.4175,
+      "step": 210
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7316666666666667,
+      "eval_f1": 0.7305281480923569,
+      "eval_loss": 0.4019607603549957,
+      "eval_runtime": 11.5986,
+      "eval_samples_per_second": 258.651,
+      "eval_steps_per_second": 32.331,
+      "step": 210
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9e-07,
+      "loss": 0.3061,
+      "step": 420
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.768,
+      "eval_f1": 0.7674936528439713,
+      "eval_loss": 0.3519645631313324,
+      "eval_runtime": 11.6033,
+      "eval_samples_per_second": 258.548,
+      "eval_steps_per_second": 32.319,
+      "step": 420
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.499999999999999e-07,
+      "loss": 0.2588,
+      "step": 630
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.79,
+      "eval_f1": 0.7887799932409598,
+      "eval_loss": 0.32531991600990295,
+      "eval_runtime": 11.5951,
+      "eval_samples_per_second": 258.729,
+      "eval_steps_per_second": 32.341,
+      "step": 630
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 8e-07,
+      "loss": 0.234,
+      "step": 840
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.788,
+      "eval_f1": 0.7877041067016072,
+      "eval_loss": 0.33730262517929077,
+      "eval_runtime": 11.6935,
+      "eval_samples_per_second": 256.553,
+      "eval_steps_per_second": 32.069,
+      "step": 840
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.5e-07,
+      "loss": 0.2116,
+      "step": 1050
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.804,
+      "eval_f1": 0.8032759681298146,
+      "eval_loss": 0.3247086703777313,
+      "eval_runtime": 11.5938,
+      "eval_samples_per_second": 258.758,
+      "eval_steps_per_second": 32.345,
+      "step": 1050
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 7e-07,
+      "loss": 0.1974,
+      "step": 1260
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.793,
+      "eval_f1": 0.7928414316247181,
+      "eval_loss": 0.34000951051712036,
+      "eval_runtime": 11.6038,
+      "eval_samples_per_second": 258.537,
+      "eval_steps_per_second": 32.317,
+      "step": 1260
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6.5e-07,
+      "loss": 0.1807,
+      "step": 1470
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7973333333333333,
+      "eval_f1": 0.7969281824938732,
+      "eval_loss": 0.35112178325653076,
+      "eval_runtime": 11.589,
+      "eval_samples_per_second": 258.866,
+      "eval_steps_per_second": 32.358,
+      "step": 1470
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 6e-07,
+      "loss": 0.1715,
+      "step": 1680
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7993333333333333,
+      "eval_f1": 0.79885693716362,
+      "eval_loss": 0.34963640570640564,
+      "eval_runtime": 11.6019,
+      "eval_samples_per_second": 258.579,
+      "eval_steps_per_second": 32.322,
+      "step": 1680
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.5e-07,
+      "loss": 0.1577,
+      "step": 1890
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.8043333333333333,
+      "eval_f1": 0.803166306363764,
+      "eval_loss": 0.3507426679134369,
+      "eval_runtime": 11.6051,
+      "eval_samples_per_second": 258.507,
+      "eval_steps_per_second": 32.313,
+      "step": 1890
+    }
+  ],
+  "max_steps": 4200,
+  "num_train_epochs": 20,
+  "total_flos": 3.742159735087181e+16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1890/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c104103a95a8e00f67dc451eba8bec78c0cbf934c84f4059200155cf9a45b752
+size 3119

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d40fe5153ca834ed2067708d539c5e680410874311663d2da4391439b5f653f3
+oid sha256:1e742a3a54a5b90c9d5d5c36842221c5911704b73b0f3c88fb76e04cd4f6ae38
 size 1340745837