Upload 13 files

unsup-whitenedcse-bert-large/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "bert-large-uncased",
+  "architectures": [
+    "BertForCL"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "attention_probs_dropout_prob_noise": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_dropout_prob_noise": 0.1,
+  "hidden_dropout_prob_noise2": 0.2,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "multi_dropout": false,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.2.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

unsup-whitenedcse-bert-large/eval_results.txt

@@ -0,0 +1,12 @@
+epoch = 3.0
+eval_CR = 86.8
+eval_MPQA = 88.93
+eval_MR = 81.34
+eval_MRPC = 75.32
+eval_SST2 = 86.35
+eval_SUBJ = 94.95
+eval_TREC = 80.06
+eval_avg_sts = 0.7633403555204041
+eval_avg_transfer = 84.82142857142857
+eval_sickr_spearman = 0.72891405751156
+eval_stsb_spearman = 0.7977666535292482

unsup-whitenedcse-bert-large/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01dd4bb57212223c323c088fb260c22c243108bc79c1b6be17b1538bd082ccf9
+size 2681362661

unsup-whitenedcse-bert-large/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b261a34a2f332726a7867ac78e20eeb149d03902ebd72b2e56b476c305e7788a
+size 1340737771

unsup-whitenedcse-bert-large/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce4f5d162c374e981a14c1f0f3013fc440027580b3ef5476b48ad3425d21b24f
+size 623

unsup-whitenedcse-bert-large/special_tokens_map.json

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

unsup-whitenedcse-bert-large/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, "model_max_length": 512, "name_or_path": "bert-large-uncased"}

unsup-whitenedcse-bert-large/train_results.txt

@@ -0,0 +1,3 @@
+epoch = 3.0
+train_runtime = 6475.209
+train_samples_per_second = 1.81

unsup-whitenedcse-bert-large/trainer_state.json

@@ -0,0 +1,139 @@
+{
+  "best_metric": 0.815129401249849,
+  "best_model_checkpoint": "result/my-unsup-simcse-bert-base-uncased",
+  "epoch": 0.2399846409829771,
+  "global_step": 1875,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "eval_avg_sts": 0.7118506537188065,
+      "eval_sickr_spearman": 0.7191480759127982,
+      "eval_stsb_spearman": 0.704553231524815,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "eval_avg_sts": 0.7490824841945876,
+      "eval_sickr_spearman": 0.7415587518643453,
+      "eval_stsb_spearman": 0.7566062165248297,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "eval_avg_sts": 0.761657495121701,
+      "eval_sickr_spearman": 0.7473965027359571,
+      "eval_stsb_spearman": 0.7759184875074447,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.8080122872136184e-05,
+      "loss": 0.0028,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_avg_sts": 0.7557478973832961,
+      "eval_sickr_spearman": 0.739644291594853,
+      "eval_stsb_spearman": 0.7718515031717392,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "eval_avg_sts": 0.7582679338253449,
+      "eval_sickr_spearman": 0.745471220921864,
+      "eval_stsb_spearman": 0.7710646467288258,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "eval_avg_sts": 0.7584985773227231,
+      "eval_sickr_spearman": 0.7513467923228714,
+      "eval_stsb_spearman": 0.7656503623225748,
+      "step": 750
+    },
+    {
+      "epoch": 0.11,
+      "eval_avg_sts": 0.7382867459735918,
+      "eval_sickr_spearman": 0.6915170632864946,
+      "eval_stsb_spearman": 0.785056428660689,
+      "step": 875
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.6160245744272367e-05,
+      "loss": 0.001,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "eval_avg_sts": 0.7696010959205108,
+      "eval_sickr_spearman": 0.7403704289249832,
+      "eval_stsb_spearman": 0.7988317629160384,
+      "step": 1000
+    },
+    {
+      "epoch": 0.14,
+      "eval_avg_sts": 0.7688633369693862,
+      "eval_sickr_spearman": 0.7358842236914375,
+      "eval_stsb_spearman": 0.8018424502473349,
+      "step": 1125
+    },
+    {
+      "epoch": 0.16,
+      "eval_avg_sts": 0.7735971189091344,
+      "eval_sickr_spearman": 0.7425422731621982,
+      "eval_stsb_spearman": 0.8046519646560706,
+      "step": 1250
+    },
+    {
+      "epoch": 0.18,
+      "eval_avg_sts": 0.7754623637694573,
+      "eval_sickr_spearman": 0.7437673388828963,
+      "eval_stsb_spearman": 0.8071573886560185,
+      "step": 1375
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.424036861640855e-05,
+      "loss": 0.0002,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "eval_avg_sts": 0.7741734659454558,
+      "eval_sickr_spearman": 0.7443858813788181,
+      "eval_stsb_spearman": 0.8039610505120934,
+      "step": 1500
+    },
+    {
+      "epoch": 0.21,
+      "eval_avg_sts": 0.779812815773511,
+      "eval_sickr_spearman": 0.7505964761581395,
+      "eval_stsb_spearman": 0.8090291553888825,
+      "step": 1625
+    },
+    {
+      "epoch": 0.22,
+      "eval_avg_sts": 0.7806364317746171,
+      "eval_sickr_spearman": 0.7509446342591667,
+      "eval_stsb_spearman": 0.8103282292900675,
+      "step": 1750
+    },
+    {
+      "epoch": 0.24,
+      "eval_avg_sts": 0.785057765369954,
+      "eval_sickr_spearman": 0.7549861294900588,
+      "eval_stsb_spearman": 0.815129401249849,
+      "step": 1875
+    }
+  ],
+  "max_steps": 7813,
+  "num_train_epochs": 1,
+  "total_flos": 0,
+  "trial_name": null,
+  "trial_params": null
+}

unsup-whitenedcse-bert-large/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21ec10519f14953ba791c9d015a02d6df160857ff68907c6bdbf92483ddde31e
+size 2095