Upload 13 files

unsup-whitenedcse-bert-base-uncased/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "bert-base-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": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "multi_dropout": false,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "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-base-uncased/eval_results.txt

@@ -0,0 +1,12 @@
+epoch = 1.0
+eval_CR = 88.47
+eval_MPQA = 88.56
+eval_MR = 83.73
+eval_MRPC = 74.34
+eval_SST2 = 88.42
+eval_SUBJ = 94.47
+eval_TREC = 80.36
+eval_avg_sts = 0.7561232564078078
+eval_avg_transfer = 85.47857142857143
+eval_sickr_spearman = 0.7206682316875482
+eval_stsb_spearman = 0.7915782811280672

unsup-whitenedcse-bert-base-uncased/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b4fd94a826bbad7e019e3973538b5725a58ae316429e5d76e56385ef5cb9b4b
+size 875973285

unsup-whitenedcse-bert-base-uncased/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0185e4dbb506a8307a9da59d0bef90f31e3e929afc2d9ebd3daf1281fa14ac46
+size 438012465

unsup-whitenedcse-bert-base-uncased/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e44519aa377dfc697372b150b09cfa3048148d6bf5a0ea461a9012cf304d40e4
+size 623

unsup-whitenedcse-bert-base-uncased/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-base-uncased/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-base-uncased"}

unsup-whitenedcse-bert-base-uncased/train_results.txt

@@ -0,0 +1,3 @@
+epoch = 1.0
+train_runtime = 7194.6299
+train_samples_per_second = 1.086

unsup-whitenedcse-bert-base-uncased/trainer_state.json

@@ -0,0 +1,241 @@
+{
+  "best_metric": 0.8064888341848981,
+  "best_model_checkpoint": "result/my-unsup-simcse-bert-base-uncased",
+  "epoch": 0.8638341438443818,
+  "global_step": 3375,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03,
+      "eval_avg_sts": 0.7035617984039024,
+      "eval_sickr_spearman": 0.6848745407030954,
+      "eval_stsb_spearman": 0.7222490561047095,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "eval_avg_sts": 0.7266642057578656,
+      "eval_sickr_spearman": 0.70274752572599,
+      "eval_stsb_spearman": 0.7505808857897414,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "eval_avg_sts": 0.7461885709126848,
+      "eval_sickr_spearman": 0.7217250002652101,
+      "eval_stsb_spearman": 0.7706521415601595,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.6160737138469415e-05,
+      "loss": 0.0013,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "eval_avg_sts": 0.7463389976451559,
+      "eval_sickr_spearman": 0.7199349731494692,
+      "eval_stsb_spearman": 0.7727430221408426,
+      "step": 500
+    },
+    {
+      "epoch": 0.16,
+      "eval_avg_sts": 0.743938555014451,
+      "eval_sickr_spearman": 0.7180073918502566,
+      "eval_stsb_spearman": 0.7698697181786454,
+      "step": 625
+    },
+    {
+      "epoch": 0.19,
+      "eval_avg_sts": 0.7550432481097468,
+      "eval_sickr_spearman": 0.7234769647664954,
+      "eval_stsb_spearman": 0.7866095314529982,
+      "step": 750
+    },
+    {
+      "epoch": 0.22,
+      "eval_avg_sts": 0.7565707728294095,
+      "eval_sickr_spearman": 0.7258336460737318,
+      "eval_stsb_spearman": 0.7873078995850874,
+      "step": 875
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.232147427693883e-05,
+      "loss": 0.0004,
+      "step": 1000
+    },
+    {
+      "epoch": 0.26,
+      "eval_avg_sts": 0.7492950228664624,
+      "eval_sickr_spearman": 0.7233381976577096,
+      "eval_stsb_spearman": 0.7752518480752153,
+      "step": 1000
+    },
+    {
+      "epoch": 0.29,
+      "eval_avg_sts": 0.7522516406505854,
+      "eval_sickr_spearman": 0.7287197031901267,
+      "eval_stsb_spearman": 0.775783578111044,
+      "step": 1125
+    },
+    {
+      "epoch": 0.32,
+      "eval_avg_sts": 0.7552964382615647,
+      "eval_sickr_spearman": 0.729468085275092,
+      "eval_stsb_spearman": 0.7811247912480374,
+      "step": 1250
+    },
+    {
+      "epoch": 0.35,
+      "eval_avg_sts": 0.7568556891093262,
+      "eval_sickr_spearman": 0.7272718476322126,
+      "eval_stsb_spearman": 0.7864395305864398,
+      "step": 1375
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8482211415408245e-05,
+      "loss": 0.0004,
+      "step": 1500
+    },
+    {
+      "epoch": 0.38,
+      "eval_avg_sts": 0.7657081366190621,
+      "eval_sickr_spearman": 0.7298737076524129,
+      "eval_stsb_spearman": 0.8015425655857115,
+      "step": 1500
+    },
+    {
+      "epoch": 0.42,
+      "eval_avg_sts": 0.7645138872097019,
+      "eval_sickr_spearman": 0.7293550763436135,
+      "eval_stsb_spearman": 0.7996726980757902,
+      "step": 1625
+    },
+    {
+      "epoch": 0.45,
+      "eval_avg_sts": 0.7604696897809164,
+      "eval_sickr_spearman": 0.7309107216292885,
+      "eval_stsb_spearman": 0.7900286579325445,
+      "step": 1750
+    },
+    {
+      "epoch": 0.48,
+      "eval_avg_sts": 0.7665958188961237,
+      "eval_sickr_spearman": 0.7308310370977922,
+      "eval_stsb_spearman": 0.8023606006944553,
+      "step": 1875
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.4642948553877656e-05,
+      "loss": 0.0004,
+      "step": 2000
+    },
+    {
+      "epoch": 0.51,
+      "eval_avg_sts": 0.7650091226879439,
+      "eval_sickr_spearman": 0.729455872126414,
+      "eval_stsb_spearman": 0.8005623732494735,
+      "step": 2000
+    },
+    {
+      "epoch": 0.54,
+      "eval_avg_sts": 0.7652971356117957,
+      "eval_sickr_spearman": 0.7297830140299209,
+      "eval_stsb_spearman": 0.8008112571936706,
+      "step": 2125
+    },
+    {
+      "epoch": 0.58,
+      "eval_avg_sts": 0.7629965694314595,
+      "eval_sickr_spearman": 0.7272991579004704,
+      "eval_stsb_spearman": 0.7986939809624487,
+      "step": 2250
+    },
+    {
+      "epoch": 0.61,
+      "eval_avg_sts": 0.7639744709473972,
+      "eval_sickr_spearman": 0.7219387730993263,
+      "eval_stsb_spearman": 0.8060101687954682,
+      "step": 2375
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.080368569234707e-05,
+      "loss": 0.0003,
+      "step": 2500
+    },
+    {
+      "epoch": 0.64,
+      "eval_avg_sts": 0.7649654793413658,
+      "eval_sickr_spearman": 0.7252475539569496,
+      "eval_stsb_spearman": 0.8046834047257821,
+      "step": 2500
+    },
+    {
+      "epoch": 0.67,
+      "eval_avg_sts": 0.7645559824476773,
+      "eval_sickr_spearman": 0.7272004062812213,
+      "eval_stsb_spearman": 0.8019115586141332,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7,
+      "eval_avg_sts": 0.76482808857313,
+      "eval_sickr_spearman": 0.7270758876759156,
+      "eval_stsb_spearman": 0.8025802894703444,
+      "step": 2750
+    },
+    {
+      "epoch": 0.74,
+      "eval_avg_sts": 0.7659614109106493,
+      "eval_sickr_spearman": 0.7289854911304676,
+      "eval_stsb_spearman": 0.802937330690831,
+      "step": 2875
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 6.964422830816484e-06,
+      "loss": 0.0005,
+      "step": 3000
+    },
+    {
+      "epoch": 0.77,
+      "eval_avg_sts": 0.7669678121477455,
+      "eval_sickr_spearman": 0.7313520084790959,
+      "eval_stsb_spearman": 0.802583615816395,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8,
+      "eval_avg_sts": 0.767953736021239,
+      "eval_sickr_spearman": 0.7305751992550179,
+      "eval_stsb_spearman": 0.8053322727874599,
+      "step": 3125
+    },
+    {
+      "epoch": 0.83,
+      "eval_avg_sts": 0.7669771718476196,
+      "eval_sickr_spearman": 0.7286785670473275,
+      "eval_stsb_spearman": 0.8052757766479117,
+      "step": 3250
+    },
+    {
+      "epoch": 0.86,
+      "eval_avg_sts": 0.7669738896262592,
+      "eval_sickr_spearman": 0.7274589450676203,
+      "eval_stsb_spearman": 0.8064888341848981,
+      "step": 3375
+    }
+  ],
+  "max_steps": 3907,
+  "num_train_epochs": 1,
+  "total_flos": 0,
+  "trial_name": null,
+  "trial_params": null
+}

unsup-whitenedcse-bert-base-uncased/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc017acc014637a03799adfe67fafbf28a757d517eb124cfdc7199e10ef49e09
+size 2095