[Model] roberta-base-stsb

config.json

@@ -0,0 +1,35 @@
+{
+  "_name_or_path": "/mnt/lustre/zhangyunchen/transformers/roberta-base",
+  "architectures": [
+    "RobertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "finetuning_task": "stsb",
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "LABEL_0"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "LABEL_0": 0
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "problem_type": "regression",
+  "torch_dtype": "float32",
+  "transformers_version": "4.16.2",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50265
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50a8a493a7d39eb568f9b195a446a70c364e5af1a95ae6b1f642d6ff28808524
+size 997298211

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d53ad7964c0be3d02151eed8eee39491de64f6ba9f46a35685cda0c150160597
+size 498676627

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dba11b9dc46ca4213c43bd2be071fbcd910ded8a0373affa5542585650081e02
+size 14659

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d03ed95b8fbea10e20c6fbe6b4d37782180f1f48fec05fcebf936fd45bf5a1a6
+size 623

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "add_prefix_space": false, "errors": "replace", "sep_token": "</s>", "cls_token": "<s>", "pad_token": "<pad>", "mask_token": "<mask>", "trim_offsets": true, "special_tokens_map_file": null, "name_or_path": "/mnt/lustre/zhangyunchen/transformers/roberta-base", "tokenizer_class": "RobertaTokenizer"}

trainer_state.json

@@ -0,0 +1,288 @@
+{
+  "best_metric": 0.9103662307301734,
+  "best_model_checkpoint": "./fp32_3e_5/models/stsb-roberta-base/checkpoint-1600",
+  "epoch": 8.88888888888889,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 3.7756,
+      "step": 100
+    },
+    {
+      "epoch": 0.56,
+      "eval_combined_score": 0.8535264998873677,
+      "eval_loss": 0.6678845286369324,
+      "eval_pearson": 0.852268663307789,
+      "eval_runtime": 7.0932,
+      "eval_samples_per_second": 211.47,
+      "eval_spearmanr": 0.8547843364669463,
+      "eval_steps_per_second": 6.626,
+      "step": 100
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.8368794326241135e-05,
+      "loss": 0.6723,
+      "step": 200
+    },
+    {
+      "epoch": 1.11,
+      "eval_combined_score": 0.8880544841766463,
+      "eval_loss": 0.5207411050796509,
+      "eval_pearson": 0.888720755423408,
+      "eval_runtime": 6.6011,
+      "eval_samples_per_second": 227.235,
+      "eval_spearmanr": 0.8873882129298847,
+      "eval_steps_per_second": 7.12,
+      "step": 200
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.6595744680851064e-05,
+      "loss": 0.5148,
+      "step": 300
+    },
+    {
+      "epoch": 1.67,
+      "eval_combined_score": 0.8963770244622205,
+      "eval_loss": 0.4456341564655304,
+      "eval_pearson": 0.8974699069640754,
+      "eval_runtime": 4.0841,
+      "eval_samples_per_second": 367.278,
+      "eval_spearmanr": 0.8952841419603657,
+      "eval_steps_per_second": 11.508,
+      "step": 300
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4822695035460992e-05,
+      "loss": 0.3898,
+      "step": 400
+    },
+    {
+      "epoch": 2.22,
+      "eval_combined_score": 0.900132405880169,
+      "eval_loss": 0.44346898794174194,
+      "eval_pearson": 0.9017121586504461,
+      "eval_runtime": 3.8243,
+      "eval_samples_per_second": 392.231,
+      "eval_spearmanr": 0.898552653109892,
+      "eval_steps_per_second": 12.29,
+      "step": 400
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.3049645390070924e-05,
+      "loss": 0.3167,
+      "step": 500
+    },
+    {
+      "epoch": 2.78,
+      "eval_combined_score": 0.9007290034038564,
+      "eval_loss": 0.48722508549690247,
+      "eval_pearson": 0.9027466027895519,
+      "eval_runtime": 3.7737,
+      "eval_samples_per_second": 397.491,
+      "eval_spearmanr": 0.898711404018161,
+      "eval_steps_per_second": 12.455,
+      "step": 500
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 0.253,
+      "step": 600
+    },
+    {
+      "epoch": 3.33,
+      "eval_combined_score": 0.9052751739076982,
+      "eval_loss": 0.4403521716594696,
+      "eval_pearson": 0.907271329521877,
+      "eval_runtime": 5.4187,
+      "eval_samples_per_second": 276.821,
+      "eval_spearmanr": 0.9032790182935194,
+      "eval_steps_per_second": 8.674,
+      "step": 600
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.950354609929078e-05,
+      "loss": 0.2197,
+      "step": 700
+    },
+    {
+      "epoch": 3.89,
+      "eval_combined_score": 0.9059989545154765,
+      "eval_loss": 0.40818288922309875,
+      "eval_pearson": 0.9078220385222855,
+      "eval_runtime": 3.7738,
+      "eval_samples_per_second": 397.48,
+      "eval_spearmanr": 0.9041758705086674,
+      "eval_steps_per_second": 12.454,
+      "step": 700
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.773049645390071e-05,
+      "loss": 0.1904,
+      "step": 800
+    },
+    {
+      "epoch": 4.44,
+      "eval_combined_score": 0.9059072653893803,
+      "eval_loss": 0.41569381952285767,
+      "eval_pearson": 0.9063948213345204,
+      "eval_runtime": 4.1987,
+      "eval_samples_per_second": 357.256,
+      "eval_spearmanr": 0.9054197094442402,
+      "eval_steps_per_second": 11.194,
+      "step": 800
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.5957446808510637e-05,
+      "loss": 0.1708,
+      "step": 900
+    },
+    {
+      "epoch": 5.0,
+      "eval_combined_score": 0.906494497393801,
+      "eval_loss": 0.46862491965293884,
+      "eval_pearson": 0.9083108072067478,
+      "eval_runtime": 4.0762,
+      "eval_samples_per_second": 367.992,
+      "eval_spearmanr": 0.904678187580854,
+      "eval_steps_per_second": 11.53,
+      "step": 900
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 1.4184397163120568e-05,
+      "loss": 0.1363,
+      "step": 1000
+    },
+    {
+      "epoch": 5.56,
+      "eval_combined_score": 0.9055447457890553,
+      "eval_loss": 0.4190002977848053,
+      "eval_pearson": 0.9069385161043947,
+      "eval_runtime": 3.7515,
+      "eval_samples_per_second": 399.845,
+      "eval_spearmanr": 0.9041509754737158,
+      "eval_steps_per_second": 12.528,
+      "step": 1000
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.2411347517730496e-05,
+      "loss": 0.1291,
+      "step": 1100
+    },
+    {
+      "epoch": 6.11,
+      "eval_combined_score": 0.9081783832953019,
+      "eval_loss": 0.4186760187149048,
+      "eval_pearson": 0.9095907875352112,
+      "eval_runtime": 4.6081,
+      "eval_samples_per_second": 325.513,
+      "eval_spearmanr": 0.9067659790553926,
+      "eval_steps_per_second": 10.199,
+      "step": 1100
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 0.1102,
+      "step": 1200
+    },
+    {
+      "epoch": 6.67,
+      "eval_combined_score": 0.907499747827053,
+      "eval_loss": 0.4207873046398163,
+      "eval_pearson": 0.9091866720599195,
+      "eval_runtime": 4.3818,
+      "eval_samples_per_second": 342.325,
+      "eval_spearmanr": 0.9058128235941866,
+      "eval_steps_per_second": 10.726,
+      "step": 1200
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 8.865248226950355e-06,
+      "loss": 0.0984,
+      "step": 1300
+    },
+    {
+      "epoch": 7.22,
+      "eval_combined_score": 0.906452303345736,
+      "eval_loss": 0.4338637590408325,
+      "eval_pearson": 0.907816988223447,
+      "eval_runtime": 3.7556,
+      "eval_samples_per_second": 399.402,
+      "eval_spearmanr": 0.905087618468025,
+      "eval_steps_per_second": 12.515,
+      "step": 1300
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 7.092198581560284e-06,
+      "loss": 0.092,
+      "step": 1400
+    },
+    {
+      "epoch": 7.78,
+      "eval_combined_score": 0.9068694360771773,
+      "eval_loss": 0.418102502822876,
+      "eval_pearson": 0.9085830258812985,
+      "eval_runtime": 4.0479,
+      "eval_samples_per_second": 370.565,
+      "eval_spearmanr": 0.9051558462730561,
+      "eval_steps_per_second": 11.611,
+      "step": 1400
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 0.0834,
+      "step": 1500
+    },
+    {
+      "epoch": 8.33,
+      "eval_combined_score": 0.9068391544864225,
+      "eval_loss": 0.42239323258399963,
+      "eval_pearson": 0.9082395585902145,
+      "eval_runtime": 4.2304,
+      "eval_samples_per_second": 354.575,
+      "eval_spearmanr": 0.9054387503826306,
+      "eval_steps_per_second": 11.11,
+      "step": 1500
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 3.546099290780142e-06,
+      "loss": 0.0798,
+      "step": 1600
+    },
+    {
+      "epoch": 8.89,
+      "eval_combined_score": 0.9087789493455298,
+      "eval_loss": 0.3950077295303345,
+      "eval_pearson": 0.9103662307301734,
+      "eval_runtime": 3.8353,
+      "eval_samples_per_second": 391.105,
+      "eval_spearmanr": 0.9071916679608862,
+      "eval_steps_per_second": 12.255,
+      "step": 1600
+    }
+  ],
+  "max_steps": 1800,
+  "num_train_epochs": 10,
+  "total_flos": 3362002879051776.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97367c6cd3e15f25c0f4cd81a95b79ed29b32469d7c3c3f416ee6cfd5ef048f0
+size 3055