Training in progress, step 856, checkpoint

checkpoint-856/1_WeightedPooling/config.json

@@ -0,0 +1,10 @@
+{
+  "word_embedding_dimension": 768,
+  "pooling_mode_cls_token": false,
+  "pooling_mode_mean_tokens": true,
+  "pooling_mode_max_tokens": false,
+  "pooling_mode_mean_sqrt_len_tokens": false,
+  "pooling_mode_weightedmean_tokens": false,
+  "pooling_mode_lasttoken": false,
+  "include_prompt": true
+}

checkpoint-856/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[MASK]": 128000
+}

checkpoint-856/config.json

@@ -0,0 +1,35 @@
+{
+  "_name_or_path": "bobox/DeBERTa-small-ST-v1-test-step3",
+  "architectures": [
+    "DebertaV2Model"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-07,
+  "max_position_embeddings": 512,
+  "max_relative_positions": -1,
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 6,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 768,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
+  "relative_attention": true,
+  "share_att_key": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.0",
+  "type_vocab_size": 0,
+  "vocab_size": 128100
+}

checkpoint-856/config_sentence_transformers.json

@@ -0,0 +1,10 @@
+{
+  "__version__": {
+    "sentence_transformers": "3.0.1",
+    "transformers": "4.44.0",
+    "pytorch": "2.4.0"
+  },
+  "prompts": {},
+  "default_prompt_name": null,
+  "similarity_fn_name": null
+}

checkpoint-856/modules.json

@@ -0,0 +1,14 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_WeightedPooling",
+    "type": "__main__"
+  }
+]

checkpoint-856/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3735b56d9678801dbb4fc3f9608ce67ae0f57b5607bc673d6d899259966cdc15
+size 9475334

checkpoint-856/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1886a286545822d70cd17732b6025a52902e7429a8ab7eec28bf9c1a4f143b6c
+size 565251810

checkpoint-856/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a49bc95ef60ca37dda4b75975617f0947174c06e9a80922a1dbab206db48276
+size 14244

checkpoint-856/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12a84c62d7e4d1b192d0b2286404c583ce0e76d01489f0466e9ccb4bbaa8d4f2
+size 1064

checkpoint-856/sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 512,
+  "do_lower_case": false
+}

checkpoint-856/special_tokens_map.json

@@ -0,0 +1,51 @@
+{
+  "bos_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-856/spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
+size 2464616

checkpoint-856/tokenizer_config.json

@@ -0,0 +1,65 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128000": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "[CLS]",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "max_length": 512,
+  "model_max_length": 512,
+  "pad_to_multiple_of": null,
+  "pad_token": "[PAD]",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "stride": 0,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}

checkpoint-856/trainer_state.json

@@ -0,0 +1,1344 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.001169590643275,
+  "eval_steps": 214,
+  "global_step": 856,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.012865497076023392,
+      "grad_norm": 0.07311347126960754,
+      "learning_rate": 3.5081502480510277e-07,
+      "loss": 0.1802,
+      "step": 11
+    },
+    {
+      "epoch": 0.025730994152046785,
+      "grad_norm": 0.3663940727710724,
+      "learning_rate": 7.016300496102055e-07,
+      "loss": 0.1573,
+      "step": 22
+    },
+    {
+      "epoch": 0.03859649122807018,
+      "grad_norm": 0.22339867055416107,
+      "learning_rate": 1.0524450744153084e-06,
+      "loss": 0.1184,
+      "step": 33
+    },
+    {
+      "epoch": 0.05146198830409357,
+      "grad_norm": 0.06488268822431564,
+      "learning_rate": 1.403260099220411e-06,
+      "loss": 0.1456,
+      "step": 44
+    },
+    {
+      "epoch": 0.06432748538011696,
+      "grad_norm": 0.7134692072868347,
+      "learning_rate": 1.754075124025514e-06,
+      "loss": 0.2102,
+      "step": 55
+    },
+    {
+      "epoch": 0.07719298245614035,
+      "grad_norm": 0.048255760222673416,
+      "learning_rate": 2.104890148830617e-06,
+      "loss": 0.151,
+      "step": 66
+    },
+    {
+      "epoch": 0.09005847953216374,
+      "grad_norm": 0.43056783080101013,
+      "learning_rate": 2.4557051736357195e-06,
+      "loss": 0.1356,
+      "step": 77
+    },
+    {
+      "epoch": 0.10292397660818714,
+      "grad_norm": 0.7663404941558838,
+      "learning_rate": 2.806520198440822e-06,
+      "loss": 0.2292,
+      "step": 88
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 0.48333388566970825,
+      "learning_rate": 3.1573352232459252e-06,
+      "loss": 0.126,
+      "step": 99
+    },
+    {
+      "epoch": 0.1286549707602339,
+      "grad_norm": 0.5180703997612,
+      "learning_rate": 3.508150248051028e-06,
+      "loss": 0.1942,
+      "step": 110
+    },
+    {
+      "epoch": 0.1415204678362573,
+      "grad_norm": 0.6447436809539795,
+      "learning_rate": 3.85896527285613e-06,
+      "loss": 0.2089,
+      "step": 121
+    },
+    {
+      "epoch": 0.1543859649122807,
+      "grad_norm": 0.5512965321540833,
+      "learning_rate": 4.209780297661234e-06,
+      "loss": 0.1225,
+      "step": 132
+    },
+    {
+      "epoch": 0.1672514619883041,
+      "grad_norm": 0.5181564688682556,
+      "learning_rate": 4.560595322466336e-06,
+      "loss": 0.2504,
+      "step": 143
+    },
+    {
+      "epoch": 0.18011695906432748,
+      "grad_norm": 1.549516201019287,
+      "learning_rate": 4.911410347271439e-06,
+      "loss": 0.1454,
+      "step": 154
+    },
+    {
+      "epoch": 0.19298245614035087,
+      "grad_norm": 0.06744886934757233,
+      "learning_rate": 5.262225372076542e-06,
+      "loss": 0.2052,
+      "step": 165
+    },
+    {
+      "epoch": 0.20584795321637428,
+      "grad_norm": 0.3412599265575409,
+      "learning_rate": 5.613040396881644e-06,
+      "loss": 0.1321,
+      "step": 176
+    },
+    {
+      "epoch": 0.21871345029239767,
+      "grad_norm": 0.4878319203853607,
+      "learning_rate": 5.963855421686748e-06,
+      "loss": 0.1975,
+      "step": 187
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 0.6517124176025391,
+      "learning_rate": 6.3146704464918505e-06,
+      "loss": 0.1615,
+      "step": 198
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 0.1128321960568428,
+      "learning_rate": 6.665485471296953e-06,
+      "loss": 0.2174,
+      "step": 209
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_Qnli-dev_cosine_accuracy": 0.697265625,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.6495540738105774,
+      "eval_Qnli-dev_cosine_ap": 0.7205205074888317,
+      "eval_Qnli-dev_cosine_f1": 0.6815286624203821,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.4835514724254608,
+      "eval_Qnli-dev_cosine_precision": 0.5459183673469388,
+      "eval_Qnli-dev_cosine_recall": 0.9067796610169492,
+      "eval_Qnli-dev_dot_accuracy": 0.666015625,
+      "eval_Qnli-dev_dot_accuracy_threshold": 307.405517578125,
+      "eval_Qnli-dev_dot_ap": 0.676904834512821,
+      "eval_Qnli-dev_dot_f1": 0.6718027734976887,
+      "eval_Qnli-dev_dot_f1_threshold": 182.19287109375,
+      "eval_Qnli-dev_dot_precision": 0.5278450363196125,
+      "eval_Qnli-dev_dot_recall": 0.923728813559322,
+      "eval_Qnli-dev_euclidean_accuracy": 0.70703125,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 17.179746627807617,
+      "eval_Qnli-dev_euclidean_ap": 0.7311696464051982,
+      "eval_Qnli-dev_euclidean_f1": 0.6860706860706861,
+      "eval_Qnli-dev_euclidean_f1_threshold": 17.50692367553711,
+      "eval_Qnli-dev_euclidean_precision": 0.673469387755102,
+      "eval_Qnli-dev_euclidean_recall": 0.6991525423728814,
+      "eval_Qnli-dev_manhattan_accuracy": 0.703125,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 357.37890625,
+      "eval_Qnli-dev_manhattan_ap": 0.7327744127614749,
+      "eval_Qnli-dev_manhattan_f1": 0.680161943319838,
+      "eval_Qnli-dev_manhattan_f1_threshold": 373.75103759765625,
+      "eval_Qnli-dev_manhattan_precision": 0.6511627906976745,
+      "eval_Qnli-dev_manhattan_recall": 0.711864406779661,
+      "eval_Qnli-dev_max_accuracy": 0.70703125,
+      "eval_Qnli-dev_max_accuracy_threshold": 357.37890625,
+      "eval_Qnli-dev_max_ap": 0.7327744127614749,
+      "eval_Qnli-dev_max_f1": 0.6860706860706861,
+      "eval_Qnli-dev_max_f1_threshold": 373.75103759765625,
+      "eval_Qnli-dev_max_precision": 0.673469387755102,
+      "eval_Qnli-dev_max_recall": 0.923728813559322,
+      "eval_allNLI-dev_cosine_accuracy": 0.740234375,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.7851241827011108,
+      "eval_allNLI-dev_cosine_ap": 0.6340322426598262,
+      "eval_allNLI-dev_cosine_f1": 0.6371681415929203,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.5923386812210083,
+      "eval_allNLI-dev_cosine_precision": 0.5161290322580645,
+      "eval_allNLI-dev_cosine_recall": 0.8323699421965318,
+      "eval_allNLI-dev_dot_accuracy": 0.724609375,
+      "eval_allNLI-dev_dot_accuracy_threshold": 294.0272521972656,
+      "eval_allNLI-dev_dot_ap": 0.5910094855503549,
+      "eval_allNLI-dev_dot_f1": 0.5960264900662251,
+      "eval_allNLI-dev_dot_f1_threshold": 227.6922149658203,
+      "eval_allNLI-dev_dot_precision": 0.48214285714285715,
+      "eval_allNLI-dev_dot_recall": 0.7803468208092486,
+      "eval_allNLI-dev_euclidean_accuracy": 0.748046875,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 13.517601013183594,
+      "eval_allNLI-dev_euclidean_ap": 0.6422252682731053,
+      "eval_allNLI-dev_euclidean_f1": 0.65,
+      "eval_allNLI-dev_euclidean_f1_threshold": 17.501617431640625,
+      "eval_allNLI-dev_euclidean_precision": 0.5355805243445693,
+      "eval_allNLI-dev_euclidean_recall": 0.8265895953757225,
+      "eval_allNLI-dev_manhattan_accuracy": 0.74609375,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 298.3919677734375,
+      "eval_allNLI-dev_manhattan_ap": 0.6432220315530937,
+      "eval_allNLI-dev_manhattan_f1": 0.6618357487922705,
+      "eval_allNLI-dev_manhattan_f1_threshold": 358.1668701171875,
+      "eval_allNLI-dev_manhattan_precision": 0.5684647302904564,
+      "eval_allNLI-dev_manhattan_recall": 0.791907514450867,
+      "eval_allNLI-dev_max_accuracy": 0.748046875,
+      "eval_allNLI-dev_max_accuracy_threshold": 298.3919677734375,
+      "eval_allNLI-dev_max_ap": 0.6432220315530937,
+      "eval_allNLI-dev_max_f1": 0.6618357487922705,
+      "eval_allNLI-dev_max_f1_threshold": 358.1668701171875,
+      "eval_allNLI-dev_max_precision": 0.5684647302904564,
+      "eval_allNLI-dev_max_recall": 0.8323699421965318,
+      "eval_sequential_score": 0.7327744127614749,
+      "eval_sts-test_pearson_cosine": 0.8852985023833193,
+      "eval_sts-test_pearson_dot": 0.8698187375798812,
+      "eval_sts-test_pearson_euclidean": 0.9048697523934177,
+      "eval_sts-test_pearson_manhattan": 0.9054784772409876,
+      "eval_sts-test_pearson_max": 0.9054784772409876,
+      "eval_sts-test_spearman_cosine": 0.9079225861122531,
+      "eval_sts-test_spearman_dot": 0.8661416896957076,
+      "eval_sts-test_spearman_euclidean": 0.9007419777144345,
+      "eval_sts-test_spearman_manhattan": 0.9013345291665179,
+      "eval_sts-test_spearman_max": 0.9079225861122531,
+      "eval_vitaminc-pairs_loss": 2.7610974311828613,
+      "eval_vitaminc-pairs_runtime": 3.8706,
+      "eval_vitaminc-pairs_samples_per_second": 33.07,
+      "eval_vitaminc-pairs_steps_per_second": 0.258,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_negation-triplets_loss": 1.0606263875961304,
+      "eval_negation-triplets_runtime": 0.6073,
+      "eval_negation-triplets_samples_per_second": 210.757,
+      "eval_negation-triplets_steps_per_second": 1.647,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_scitail-pairs-pos_loss": 0.059617433696985245,
+      "eval_scitail-pairs-pos_runtime": 0.9151,
+      "eval_scitail-pairs-pos_samples_per_second": 139.868,
+      "eval_scitail-pairs-pos_steps_per_second": 1.093,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_scitail-pairs-qa_loss": 2.513320396246854e-07,
+      "eval_scitail-pairs-qa_runtime": 0.5906,
+      "eval_scitail-pairs-qa_samples_per_second": 216.739,
+      "eval_scitail-pairs-qa_steps_per_second": 1.693,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_xsum-pairs_loss": 0.004431054927408695,
+      "eval_xsum-pairs_runtime": 2.7873,
+      "eval_xsum-pairs_samples_per_second": 45.922,
+      "eval_xsum-pairs_steps_per_second": 0.359,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_sciq_pairs_loss": 0.013232991099357605,
+      "eval_sciq_pairs_runtime": 4.1862,
+      "eval_sciq_pairs_samples_per_second": 30.576,
+      "eval_sciq_pairs_steps_per_second": 0.239,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_qasc_pairs_loss": 0.049007147550582886,
+      "eval_qasc_pairs_runtime": 0.6288,
+      "eval_qasc_pairs_samples_per_second": 203.554,
+      "eval_qasc_pairs_steps_per_second": 1.59,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_openbookqa_pairs_loss": 1.3196656703948975,
+      "eval_openbookqa_pairs_runtime": 0.5848,
+      "eval_openbookqa_pairs_samples_per_second": 218.897,
+      "eval_openbookqa_pairs_steps_per_second": 1.71,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_msmarco_pairs_loss": 0.13587385416030884,
+      "eval_msmarco_pairs_runtime": 1.341,
+      "eval_msmarco_pairs_samples_per_second": 95.453,
+      "eval_msmarco_pairs_steps_per_second": 0.746,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_nq_pairs_loss": 0.04134061187505722,
+      "eval_nq_pairs_runtime": 2.9296,
+      "eval_nq_pairs_samples_per_second": 43.693,
+      "eval_nq_pairs_steps_per_second": 0.341,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_trivia_pairs_loss": 0.211629718542099,
+      "eval_trivia_pairs_runtime": 3.2033,
+      "eval_trivia_pairs_samples_per_second": 39.958,
+      "eval_trivia_pairs_steps_per_second": 0.312,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_gooaq_pairs_loss": 0.11869887262582779,
+      "eval_gooaq_pairs_runtime": 0.869,
+      "eval_gooaq_pairs_samples_per_second": 147.291,
+      "eval_gooaq_pairs_steps_per_second": 1.151,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_paws-pos_loss": 0.04590649530291557,
+      "eval_paws-pos_runtime": 0.6652,
+      "eval_paws-pos_samples_per_second": 192.433,
+      "eval_paws-pos_steps_per_second": 1.503,
+      "step": 214
+    },
+    {
+      "epoch": 0.25029239766081873,
+      "eval_global_dataset_loss": 0.21873071789741516,
+      "eval_global_dataset_runtime": 24.2472,
+      "eval_global_dataset_samples_per_second": 27.343,
+      "eval_global_dataset_steps_per_second": 0.247,
+      "step": 214
+    },
+    {
+      "epoch": 0.2573099415204678,
+      "grad_norm": 0.2186487466096878,
+      "learning_rate": 7.016300496102056e-06,
+      "loss": 0.1545,
+      "step": 220
+    },
+    {
+      "epoch": 0.27017543859649124,
+      "grad_norm": 0.16798946261405945,
+      "learning_rate": 7.3671155209071584e-06,
+      "loss": 0.1201,
+      "step": 231
+    },
+    {
+      "epoch": 0.2830409356725146,
+      "grad_norm": 0.5161852836608887,
+      "learning_rate": 7.71793054571226e-06,
+      "loss": 0.1389,
+      "step": 242
+    },
+    {
+      "epoch": 0.295906432748538,
+      "grad_norm": 0.6402292251586914,
+      "learning_rate": 8.068745570517365e-06,
+      "loss": 0.1657,
+      "step": 253
+    },
+    {
+      "epoch": 0.3087719298245614,
+      "grad_norm": 0.8064345717430115,
+      "learning_rate": 8.419560595322467e-06,
+      "loss": 0.237,
+      "step": 264
+    },
+    {
+      "epoch": 0.3216374269005848,
+      "grad_norm": 0.8077338337898254,
+      "learning_rate": 8.770375620127569e-06,
+      "loss": 0.1094,
+      "step": 275
+    },
+    {
+      "epoch": 0.3345029239766082,
+      "grad_norm": 0.23748837411403656,
+      "learning_rate": 9.121190644932673e-06,
+      "loss": 0.196,
+      "step": 286
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 0.26720044016838074,
+      "learning_rate": 9.472005669737774e-06,
+      "loss": 0.2164,
+      "step": 297
+    },
+    {
+      "epoch": 0.36023391812865496,
+      "grad_norm": 0.8082326650619507,
+      "learning_rate": 9.822820694542878e-06,
+      "loss": 0.1793,
+      "step": 308
+    },
+    {
+      "epoch": 0.3730994152046784,
+      "grad_norm": 0.8150445818901062,
+      "learning_rate": 1.017363571934798e-05,
+      "loss": 0.2878,
+      "step": 319
+    },
+    {
+      "epoch": 0.38596491228070173,
+      "grad_norm": 0.2097979038953781,
+      "learning_rate": 1.0524450744153083e-05,
+      "loss": 0.1189,
+      "step": 330
+    },
+    {
+      "epoch": 0.39883040935672515,
+      "grad_norm": 1.0323094129562378,
+      "learning_rate": 1.0875265768958187e-05,
+      "loss": 0.1475,
+      "step": 341
+    },
+    {
+      "epoch": 0.41169590643274856,
+      "grad_norm": 0.41917282342910767,
+      "learning_rate": 1.1226080793763289e-05,
+      "loss": 0.1019,
+      "step": 352
+    },
+    {
+      "epoch": 0.4245614035087719,
+      "grad_norm": 0.17783042788505554,
+      "learning_rate": 1.1576895818568394e-05,
+      "loss": 0.1587,
+      "step": 363
+    },
+    {
+      "epoch": 0.43742690058479533,
+      "grad_norm": 0.8943735957145691,
+      "learning_rate": 1.1927710843373496e-05,
+      "loss": 0.2483,
+      "step": 374
+    },
+    {
+      "epoch": 0.4502923976608187,
+      "grad_norm": 0.7853964567184448,
+      "learning_rate": 1.2278525868178599e-05,
+      "loss": 0.1427,
+      "step": 385
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 0.4753592908382416,
+      "learning_rate": 1.2629340892983701e-05,
+      "loss": 0.1199,
+      "step": 396
+    },
+    {
+      "epoch": 0.4760233918128655,
+      "grad_norm": 0.47377559542655945,
+      "learning_rate": 1.2980155917788803e-05,
+      "loss": 0.2037,
+      "step": 407
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 0.6249018311500549,
+      "learning_rate": 1.3330970942593906e-05,
+      "loss": 0.1317,
+      "step": 418
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_Qnli-dev_cosine_accuracy": 0.6953125,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.6539117097854614,
+      "eval_Qnli-dev_cosine_ap": 0.7174538904795291,
+      "eval_Qnli-dev_cosine_f1": 0.6773675762439807,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.4876832365989685,
+      "eval_Qnli-dev_cosine_precision": 0.5452196382428941,
+      "eval_Qnli-dev_cosine_recall": 0.8940677966101694,
+      "eval_Qnli-dev_dot_accuracy": 0.673828125,
+      "eval_Qnli-dev_dot_accuracy_threshold": 248.8419647216797,
+      "eval_Qnli-dev_dot_ap": 0.6692436401997881,
+      "eval_Qnli-dev_dot_f1": 0.6758832565284177,
+      "eval_Qnli-dev_dot_f1_threshold": 174.55992126464844,
+      "eval_Qnli-dev_dot_precision": 0.5301204819277109,
+      "eval_Qnli-dev_dot_recall": 0.9322033898305084,
+      "eval_Qnli-dev_euclidean_accuracy": 0.70703125,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 16.61339569091797,
+      "eval_Qnli-dev_euclidean_ap": 0.7278403442362325,
+      "eval_Qnli-dev_euclidean_f1": 0.6821052631578948,
+      "eval_Qnli-dev_euclidean_f1_threshold": 17.059036254882812,
+      "eval_Qnli-dev_euclidean_precision": 0.6778242677824268,
+      "eval_Qnli-dev_euclidean_recall": 0.6864406779661016,
+      "eval_Qnli-dev_manhattan_accuracy": 0.697265625,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 348.561279296875,
+      "eval_Qnli-dev_manhattan_ap": 0.7293997108333973,
+      "eval_Qnli-dev_manhattan_f1": 0.6797642436149313,
+      "eval_Qnli-dev_manhattan_f1_threshold": 371.4664306640625,
+      "eval_Qnli-dev_manhattan_precision": 0.6336996336996337,
+      "eval_Qnli-dev_manhattan_recall": 0.7330508474576272,
+      "eval_Qnli-dev_max_accuracy": 0.70703125,
+      "eval_Qnli-dev_max_accuracy_threshold": 348.561279296875,
+      "eval_Qnli-dev_max_ap": 0.7293997108333973,
+      "eval_Qnli-dev_max_f1": 0.6821052631578948,
+      "eval_Qnli-dev_max_f1_threshold": 371.4664306640625,
+      "eval_Qnli-dev_max_precision": 0.6778242677824268,
+      "eval_Qnli-dev_max_recall": 0.9322033898305084,
+      "eval_allNLI-dev_cosine_accuracy": 0.740234375,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.789513111114502,
+      "eval_allNLI-dev_cosine_ap": 0.6326730829325957,
+      "eval_allNLI-dev_cosine_f1": 0.6369710467706012,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.6001436710357666,
+      "eval_allNLI-dev_cosine_precision": 0.5181159420289855,
+      "eval_allNLI-dev_cosine_recall": 0.8265895953757225,
+      "eval_allNLI-dev_dot_accuracy": 0.728515625,
+      "eval_allNLI-dev_dot_accuracy_threshold": 284.5158996582031,
+      "eval_allNLI-dev_dot_ap": 0.587991265887336,
+      "eval_allNLI-dev_dot_f1": 0.5996204933586338,
+      "eval_allNLI-dev_dot_f1_threshold": 187.17926025390625,
+      "eval_allNLI-dev_dot_precision": 0.4463276836158192,
+      "eval_allNLI-dev_dot_recall": 0.9132947976878613,
+      "eval_allNLI-dev_euclidean_accuracy": 0.740234375,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 13.375718116760254,
+      "eval_allNLI-dev_euclidean_ap": 0.6391067622382595,
+      "eval_allNLI-dev_euclidean_f1": 0.6478260869565217,
+      "eval_allNLI-dev_euclidean_f1_threshold": 17.581111907958984,
+      "eval_allNLI-dev_euclidean_precision": 0.519163763066202,
+      "eval_allNLI-dev_euclidean_recall": 0.861271676300578,
+      "eval_allNLI-dev_manhattan_accuracy": 0.740234375,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 271.5396728515625,
+      "eval_allNLI-dev_manhattan_ap": 0.6397141445378819,
+      "eval_allNLI-dev_manhattan_f1": 0.6558139534883721,
+      "eval_allNLI-dev_manhattan_f1_threshold": 356.73675537109375,
+      "eval_allNLI-dev_manhattan_precision": 0.5486381322957199,
+      "eval_allNLI-dev_manhattan_recall": 0.815028901734104,
+      "eval_allNLI-dev_max_accuracy": 0.740234375,
+      "eval_allNLI-dev_max_accuracy_threshold": 284.5158996582031,
+      "eval_allNLI-dev_max_ap": 0.6397141445378819,
+      "eval_allNLI-dev_max_f1": 0.6558139534883721,
+      "eval_allNLI-dev_max_f1_threshold": 356.73675537109375,
+      "eval_allNLI-dev_max_precision": 0.5486381322957199,
+      "eval_allNLI-dev_max_recall": 0.9132947976878613,
+      "eval_sequential_score": 0.7293997108333973,
+      "eval_sts-test_pearson_cosine": 0.8826342923852873,
+      "eval_sts-test_pearson_dot": 0.8663977738237862,
+      "eval_sts-test_pearson_euclidean": 0.9007138184358896,
+      "eval_sts-test_pearson_manhattan": 0.9014088314200603,
+      "eval_sts-test_pearson_max": 0.9014088314200603,
+      "eval_sts-test_spearman_cosine": 0.9062742181501622,
+      "eval_sts-test_spearman_dot": 0.8628408814067393,
+      "eval_sts-test_spearman_euclidean": 0.897460680975433,
+      "eval_sts-test_spearman_manhattan": 0.8979340598184311,
+      "eval_sts-test_spearman_max": 0.9062742181501622,
+      "eval_vitaminc-pairs_loss": 2.7317824363708496,
+      "eval_vitaminc-pairs_runtime": 3.8325,
+      "eval_vitaminc-pairs_samples_per_second": 33.398,
+      "eval_vitaminc-pairs_steps_per_second": 0.261,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_negation-triplets_loss": 1.0631705522537231,
+      "eval_negation-triplets_runtime": 0.6008,
+      "eval_negation-triplets_samples_per_second": 213.032,
+      "eval_negation-triplets_steps_per_second": 1.664,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_scitail-pairs-pos_loss": 0.06051468849182129,
+      "eval_scitail-pairs-pos_runtime": 0.8927,
+      "eval_scitail-pairs-pos_samples_per_second": 143.381,
+      "eval_scitail-pairs-pos_steps_per_second": 1.12,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_scitail-pairs-qa_loss": 3.731894366865163e-07,
+      "eval_scitail-pairs-qa_runtime": 0.5829,
+      "eval_scitail-pairs-qa_samples_per_second": 219.584,
+      "eval_scitail-pairs-qa_steps_per_second": 1.716,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_xsum-pairs_loss": 0.004530374892055988,
+      "eval_xsum-pairs_runtime": 2.7882,
+      "eval_xsum-pairs_samples_per_second": 45.908,
+      "eval_xsum-pairs_steps_per_second": 0.359,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_sciq_pairs_loss": 0.0028694826178252697,
+      "eval_sciq_pairs_runtime": 4.1341,
+      "eval_sciq_pairs_samples_per_second": 30.962,
+      "eval_sciq_pairs_steps_per_second": 0.242,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_qasc_pairs_loss": 0.0601225420832634,
+      "eval_qasc_pairs_runtime": 0.6017,
+      "eval_qasc_pairs_samples_per_second": 212.728,
+      "eval_qasc_pairs_steps_per_second": 1.662,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_openbookqa_pairs_loss": 1.2716314792633057,
+      "eval_openbookqa_pairs_runtime": 0.5746,
+      "eval_openbookqa_pairs_samples_per_second": 222.757,
+      "eval_openbookqa_pairs_steps_per_second": 1.74,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_msmarco_pairs_loss": 0.16218337416648865,
+      "eval_msmarco_pairs_runtime": 1.3355,
+      "eval_msmarco_pairs_samples_per_second": 95.842,
+      "eval_msmarco_pairs_steps_per_second": 0.749,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_nq_pairs_loss": 0.039985254406929016,
+      "eval_nq_pairs_runtime": 2.9238,
+      "eval_nq_pairs_samples_per_second": 43.779,
+      "eval_nq_pairs_steps_per_second": 0.342,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_trivia_pairs_loss": 0.20936143398284912,
+      "eval_trivia_pairs_runtime": 3.1822,
+      "eval_trivia_pairs_samples_per_second": 40.223,
+      "eval_trivia_pairs_steps_per_second": 0.314,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_gooaq_pairs_loss": 0.12705038487911224,
+      "eval_gooaq_pairs_runtime": 0.8602,
+      "eval_gooaq_pairs_samples_per_second": 148.795,
+      "eval_gooaq_pairs_steps_per_second": 1.162,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_paws-pos_loss": 0.046078916639089584,
+      "eval_paws-pos_runtime": 0.6578,
+      "eval_paws-pos_samples_per_second": 194.575,
+      "eval_paws-pos_steps_per_second": 1.52,
+      "step": 428
+    },
+    {
+      "epoch": 0.5005847953216375,
+      "eval_global_dataset_loss": 0.21221815049648285,
+      "eval_global_dataset_runtime": 24.2782,
+      "eval_global_dataset_samples_per_second": 27.308,
+      "eval_global_dataset_steps_per_second": 0.247,
+      "step": 428
+    },
+    {
+      "epoch": 0.5017543859649123,
+      "grad_norm": 0.6901716589927673,
+      "learning_rate": 1.3681785967399008e-05,
+      "loss": 0.1293,
+      "step": 429
+    },
+    {
+      "epoch": 0.5146198830409356,
+      "grad_norm": 1.1465657949447632,
+      "learning_rate": 1.4032600992204112e-05,
+      "loss": 0.1902,
+      "step": 440
+    },
+    {
+      "epoch": 0.5274853801169591,
+      "grad_norm": 1.0795351266860962,
+      "learning_rate": 1.4383416017009213e-05,
+      "loss": 0.1429,
+      "step": 451
+    },
+    {
+      "epoch": 0.5403508771929825,
+      "grad_norm": 0.9797111749649048,
+      "learning_rate": 1.4734231041814317e-05,
+      "loss": 0.2446,
+      "step": 462
+    },
+    {
+      "epoch": 0.5532163742690058,
+      "grad_norm": 0.21088336408138275,
+      "learning_rate": 1.5085046066619419e-05,
+      "loss": 0.1623,
+      "step": 473
+    },
+    {
+      "epoch": 0.5660818713450292,
+      "grad_norm": 0.4258130192756653,
+      "learning_rate": 1.543586109142452e-05,
+      "loss": 0.0707,
+      "step": 484
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.8029544949531555,
+      "learning_rate": 1.5786676116229624e-05,
+      "loss": 0.1557,
+      "step": 495
+    },
+    {
+      "epoch": 0.591812865497076,
+      "grad_norm": 2.734902858734131,
+      "learning_rate": 1.613749114103473e-05,
+      "loss": 0.2016,
+      "step": 506
+    },
+    {
+      "epoch": 0.6046783625730994,
+      "grad_norm": 0.4968857169151306,
+      "learning_rate": 1.648830616583983e-05,
+      "loss": 0.1018,
+      "step": 517
+    },
+    {
+      "epoch": 0.6175438596491228,
+      "grad_norm": 1.5250153541564941,
+      "learning_rate": 1.6839121190644935e-05,
+      "loss": 0.1821,
+      "step": 528
+    },
+    {
+      "epoch": 0.6304093567251462,
+      "grad_norm": 0.7500655055046082,
+      "learning_rate": 1.7189936215450038e-05,
+      "loss": 0.1437,
+      "step": 539
+    },
+    {
+      "epoch": 0.6432748538011696,
+      "grad_norm": 0.19580113887786865,
+      "learning_rate": 1.7540751240255138e-05,
+      "loss": 0.1112,
+      "step": 550
+    },
+    {
+      "epoch": 0.656140350877193,
+      "grad_norm": 0.1376626342535019,
+      "learning_rate": 1.789156626506024e-05,
+      "loss": 0.12,
+      "step": 561
+    },
+    {
+      "epoch": 0.6690058479532164,
+      "grad_norm": 0.49516016244888306,
+      "learning_rate": 1.8242381289865345e-05,
+      "loss": 0.0933,
+      "step": 572
+    },
+    {
+      "epoch": 0.6818713450292397,
+      "grad_norm": 0.20120573043823242,
+      "learning_rate": 1.859319631467045e-05,
+      "loss": 0.0939,
+      "step": 583
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 0.3648889362812042,
+      "learning_rate": 1.894401133947555e-05,
+      "loss": 0.2064,
+      "step": 594
+    },
+    {
+      "epoch": 0.7076023391812866,
+      "grad_norm": 0.07958167791366577,
+      "learning_rate": 1.9294826364280652e-05,
+      "loss": 0.131,
+      "step": 605
+    },
+    {
+      "epoch": 0.7204678362573099,
+      "grad_norm": 0.7020072340965271,
+      "learning_rate": 1.9645641389085756e-05,
+      "loss": 0.161,
+      "step": 616
+    },
+    {
+      "epoch": 0.7333333333333333,
+      "grad_norm": 1.3057806491851807,
+      "learning_rate": 1.999645641389086e-05,
+      "loss": 0.213,
+      "step": 627
+    },
+    {
+      "epoch": 0.7461988304093568,
+      "grad_norm": 0.7668664455413818,
+      "learning_rate": 2.034727143869596e-05,
+      "loss": 0.1853,
+      "step": 638
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_Qnli-dev_cosine_accuracy": 0.693359375,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.6550551652908325,
+      "eval_Qnli-dev_cosine_ap": 0.7155098805507838,
+      "eval_Qnli-dev_cosine_f1": 0.6794871794871794,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.4866919219493866,
+      "eval_Qnli-dev_cosine_precision": 0.5463917525773195,
+      "eval_Qnli-dev_cosine_recall": 0.8983050847457628,
+      "eval_Qnli-dev_dot_accuracy": 0.67578125,
+      "eval_Qnli-dev_dot_accuracy_threshold": 258.037841796875,
+      "eval_Qnli-dev_dot_ap": 0.672701841309456,
+      "eval_Qnli-dev_dot_f1": 0.6731078904991948,
+      "eval_Qnli-dev_dot_f1_threshold": 197.6945343017578,
+      "eval_Qnli-dev_dot_precision": 0.5428571428571428,
+      "eval_Qnli-dev_dot_recall": 0.885593220338983,
+      "eval_Qnli-dev_euclidean_accuracy": 0.697265625,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 16.745851516723633,
+      "eval_Qnli-dev_euclidean_ap": 0.7236455475550527,
+      "eval_Qnli-dev_euclidean_f1": 0.6750524109014674,
+      "eval_Qnli-dev_euclidean_f1_threshold": 17.19139862060547,
+      "eval_Qnli-dev_euclidean_precision": 0.6680497925311203,
+      "eval_Qnli-dev_euclidean_recall": 0.6822033898305084,
+      "eval_Qnli-dev_manhattan_accuracy": 0.697265625,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 362.6970520019531,
+      "eval_Qnli-dev_manhattan_ap": 0.7257321921740054,
+      "eval_Qnli-dev_manhattan_f1": 0.6774847870182557,
+      "eval_Qnli-dev_manhattan_f1_threshold": 369.7593994140625,
+      "eval_Qnli-dev_manhattan_precision": 0.6498054474708171,
+      "eval_Qnli-dev_manhattan_recall": 0.7076271186440678,
+      "eval_Qnli-dev_max_accuracy": 0.697265625,
+      "eval_Qnli-dev_max_accuracy_threshold": 362.6970520019531,
+      "eval_Qnli-dev_max_ap": 0.7257321921740054,
+      "eval_Qnli-dev_max_f1": 0.6794871794871794,
+      "eval_Qnli-dev_max_f1_threshold": 369.7593994140625,
+      "eval_Qnli-dev_max_precision": 0.6680497925311203,
+      "eval_Qnli-dev_max_recall": 0.8983050847457628,
+      "eval_allNLI-dev_cosine_accuracy": 0.73828125,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.7900433540344238,
+      "eval_allNLI-dev_cosine_ap": 0.6359348988299178,
+      "eval_allNLI-dev_cosine_f1": 0.639269406392694,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.6079586744308472,
+      "eval_allNLI-dev_cosine_precision": 0.5283018867924528,
+      "eval_allNLI-dev_cosine_recall": 0.8092485549132948,
+      "eval_allNLI-dev_dot_accuracy": 0.7265625,
+      "eval_allNLI-dev_dot_accuracy_threshold": 282.57489013671875,
+      "eval_allNLI-dev_dot_ap": 0.589207838274076,
+      "eval_allNLI-dev_dot_f1": 0.6059405940594059,
+      "eval_allNLI-dev_dot_f1_threshold": 203.0603485107422,
+      "eval_allNLI-dev_dot_precision": 0.4608433734939759,
+      "eval_allNLI-dev_dot_recall": 0.884393063583815,
+      "eval_allNLI-dev_euclidean_accuracy": 0.744140625,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 13.129840850830078,
+      "eval_allNLI-dev_euclidean_ap": 0.6431587171633473,
+      "eval_allNLI-dev_euclidean_f1": 0.6506550218340611,
+      "eval_allNLI-dev_euclidean_f1_threshold": 17.757301330566406,
+      "eval_allNLI-dev_euclidean_precision": 0.5228070175438596,
+      "eval_allNLI-dev_euclidean_recall": 0.861271676300578,
+      "eval_allNLI-dev_manhattan_accuracy": 0.74609375,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 276.65826416015625,
+      "eval_allNLI-dev_manhattan_ap": 0.6424686977323816,
+      "eval_allNLI-dev_manhattan_f1": 0.6589861751152074,
+      "eval_allNLI-dev_manhattan_f1_threshold": 362.68414306640625,
+      "eval_allNLI-dev_manhattan_precision": 0.5478927203065134,
+      "eval_allNLI-dev_manhattan_recall": 0.8265895953757225,
+      "eval_allNLI-dev_max_accuracy": 0.74609375,
+      "eval_allNLI-dev_max_accuracy_threshold": 282.57489013671875,
+      "eval_allNLI-dev_max_ap": 0.6431587171633473,
+      "eval_allNLI-dev_max_f1": 0.6589861751152074,
+      "eval_allNLI-dev_max_f1_threshold": 362.68414306640625,
+      "eval_allNLI-dev_max_precision": 0.5478927203065134,
+      "eval_allNLI-dev_max_recall": 0.884393063583815,
+      "eval_sequential_score": 0.7257321921740054,
+      "eval_sts-test_pearson_cosine": 0.8848372605057793,
+      "eval_sts-test_pearson_dot": 0.8709019996065221,
+      "eval_sts-test_pearson_euclidean": 0.9038234321498232,
+      "eval_sts-test_pearson_manhattan": 0.9044737426180963,
+      "eval_sts-test_pearson_max": 0.9044737426180963,
+      "eval_sts-test_spearman_cosine": 0.9073185783372533,
+      "eval_sts-test_spearman_dot": 0.8668432552205406,
+      "eval_sts-test_spearman_euclidean": 0.899716547310912,
+      "eval_sts-test_spearman_manhattan": 0.9006361409812439,
+      "eval_sts-test_spearman_max": 0.9073185783372533,
+      "eval_vitaminc-pairs_loss": 2.7074272632598877,
+      "eval_vitaminc-pairs_runtime": 3.7948,
+      "eval_vitaminc-pairs_samples_per_second": 33.73,
+      "eval_vitaminc-pairs_steps_per_second": 0.264,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_negation-triplets_loss": 1.0683540105819702,
+      "eval_negation-triplets_runtime": 0.5897,
+      "eval_negation-triplets_samples_per_second": 217.072,
+      "eval_negation-triplets_steps_per_second": 1.696,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_scitail-pairs-pos_loss": 0.05928462743759155,
+      "eval_scitail-pairs-pos_runtime": 0.8353,
+      "eval_scitail-pairs-pos_samples_per_second": 153.247,
+      "eval_scitail-pairs-pos_steps_per_second": 1.197,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_scitail-pairs-qa_loss": 2.8886071845590777e-07,
+      "eval_scitail-pairs-qa_runtime": 0.58,
+      "eval_scitail-pairs-qa_samples_per_second": 220.677,
+      "eval_scitail-pairs-qa_steps_per_second": 1.724,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_xsum-pairs_loss": 0.005155234597623348,
+      "eval_xsum-pairs_runtime": 2.8136,
+      "eval_xsum-pairs_samples_per_second": 45.494,
+      "eval_xsum-pairs_steps_per_second": 0.355,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_sciq_pairs_loss": 0.0030471039935946465,
+      "eval_sciq_pairs_runtime": 4.0962,
+      "eval_sciq_pairs_samples_per_second": 31.248,
+      "eval_sciq_pairs_steps_per_second": 0.244,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_qasc_pairs_loss": 0.05629788711667061,
+      "eval_qasc_pairs_runtime": 0.5944,
+      "eval_qasc_pairs_samples_per_second": 215.353,
+      "eval_qasc_pairs_steps_per_second": 1.682,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_openbookqa_pairs_loss": 1.2897449731826782,
+      "eval_openbookqa_pairs_runtime": 0.5597,
+      "eval_openbookqa_pairs_samples_per_second": 228.687,
+      "eval_openbookqa_pairs_steps_per_second": 1.787,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_msmarco_pairs_loss": 0.14785167574882507,
+      "eval_msmarco_pairs_runtime": 1.3218,
+      "eval_msmarco_pairs_samples_per_second": 96.836,
+      "eval_msmarco_pairs_steps_per_second": 0.757,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_nq_pairs_loss": 0.04229257255792618,
+      "eval_nq_pairs_runtime": 2.9207,
+      "eval_nq_pairs_samples_per_second": 43.825,
+      "eval_nq_pairs_steps_per_second": 0.342,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_trivia_pairs_loss": 0.2112799882888794,
+      "eval_trivia_pairs_runtime": 3.1802,
+      "eval_trivia_pairs_samples_per_second": 40.249,
+      "eval_trivia_pairs_steps_per_second": 0.314,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_gooaq_pairs_loss": 0.11469888687133789,
+      "eval_gooaq_pairs_runtime": 0.8601,
+      "eval_gooaq_pairs_samples_per_second": 148.825,
+      "eval_gooaq_pairs_steps_per_second": 1.163,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_paws-pos_loss": 0.0460214801132679,
+      "eval_paws-pos_runtime": 0.6612,
+      "eval_paws-pos_samples_per_second": 193.581,
+      "eval_paws-pos_steps_per_second": 1.512,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508771929824561,
+      "eval_global_dataset_loss": 0.20304369926452637,
+      "eval_global_dataset_runtime": 24.1397,
+      "eval_global_dataset_samples_per_second": 27.465,
+      "eval_global_dataset_steps_per_second": 0.249,
+      "step": 642
+    },
+    {
+      "epoch": 0.7590643274853801,
+      "grad_norm": 0.5546406507492065,
+      "learning_rate": 2.0698086463501066e-05,
+      "loss": 0.1919,
+      "step": 649
+    },
+    {
+      "epoch": 0.7719298245614035,
+      "grad_norm": 0.16179773211479187,
+      "learning_rate": 2.1048901488306166e-05,
+      "loss": 0.1395,
+      "step": 660
+    },
+    {
+      "epoch": 0.7847953216374269,
+      "grad_norm": 0.4947220981121063,
+      "learning_rate": 2.139971651311127e-05,
+      "loss": 0.2047,
+      "step": 671
+    },
+    {
+      "epoch": 0.7976608187134503,
+      "grad_norm": 0.4697045683860779,
+      "learning_rate": 2.1750531537916374e-05,
+      "loss": 0.1421,
+      "step": 682
+    },
+    {
+      "epoch": 0.8105263157894737,
+      "grad_norm": 0.7013148665428162,
+      "learning_rate": 2.2101346562721477e-05,
+      "loss": 0.1227,
+      "step": 693
+    },
+    {
+      "epoch": 0.8233918128654971,
+      "grad_norm": 1.1045671701431274,
+      "learning_rate": 2.2452161587526577e-05,
+      "loss": 0.1235,
+      "step": 704
+    },
+    {
+      "epoch": 0.8362573099415205,
+      "grad_norm": 1.5613625049591064,
+      "learning_rate": 2.280297661233168e-05,
+      "loss": 0.2004,
+      "step": 715
+    },
+    {
+      "epoch": 0.8491228070175438,
+      "grad_norm": 0.23709431290626526,
+      "learning_rate": 2.3153791637136788e-05,
+      "loss": 0.1568,
+      "step": 726
+    },
+    {
+      "epoch": 0.8619883040935673,
+      "grad_norm": 1.1035298109054565,
+      "learning_rate": 2.3504606661941888e-05,
+      "loss": 0.1598,
+      "step": 737
+    },
+    {
+      "epoch": 0.8748538011695907,
+      "grad_norm": 0.7472051978111267,
+      "learning_rate": 2.385542168674699e-05,
+      "loss": 0.1328,
+      "step": 748
+    },
+    {
+      "epoch": 0.887719298245614,
+      "grad_norm": 0.57217937707901,
+      "learning_rate": 2.420623671155209e-05,
+      "loss": 0.0999,
+      "step": 759
+    },
+    {
+      "epoch": 0.9005847953216374,
+      "grad_norm": 0.3447597920894623,
+      "learning_rate": 2.4557051736357198e-05,
+      "loss": 0.1058,
+      "step": 770
+    },
+    {
+      "epoch": 0.9134502923976608,
+      "grad_norm": 1.0243078470230103,
+      "learning_rate": 2.4907866761162295e-05,
+      "loss": 0.1673,
+      "step": 781
+    },
+    {
+      "epoch": 0.9263157894736842,
+      "grad_norm": 1.0685806274414062,
+      "learning_rate": 2.5258681785967402e-05,
+      "loss": 0.1905,
+      "step": 792
+    },
+    {
+      "epoch": 0.9391812865497076,
+      "grad_norm": 1.1175950765609741,
+      "learning_rate": 2.5609496810772502e-05,
+      "loss": 0.1463,
+      "step": 803
+    },
+    {
+      "epoch": 0.952046783625731,
+      "grad_norm": 0.13670021295547485,
+      "learning_rate": 2.5960311835577605e-05,
+      "loss": 0.1294,
+      "step": 814
+    },
+    {
+      "epoch": 0.9649122807017544,
+      "grad_norm": 0.6462442278862,
+      "learning_rate": 2.6311126860382706e-05,
+      "loss": 0.1312,
+      "step": 825
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 0.1836312711238861,
+      "learning_rate": 2.6661941885187812e-05,
+      "loss": 0.1308,
+      "step": 836
+    },
+    {
+      "epoch": 0.9906432748538012,
+      "grad_norm": 4.2791407395270653e-07,
+      "learning_rate": 2.7012756909992913e-05,
+      "loss": 0.1076,
+      "step": 847
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_Qnli-dev_cosine_accuracy": 0.69140625,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.657210111618042,
+      "eval_Qnli-dev_cosine_ap": 0.714041144664227,
+      "eval_Qnli-dev_cosine_f1": 0.6783439490445861,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.48659271001815796,
+      "eval_Qnli-dev_cosine_precision": 0.5433673469387755,
+      "eval_Qnli-dev_cosine_recall": 0.902542372881356,
+      "eval_Qnli-dev_dot_accuracy": 0.669921875,
+      "eval_Qnli-dev_dot_accuracy_threshold": 256.53155517578125,
+      "eval_Qnli-dev_dot_ap": 0.6648303273464243,
+      "eval_Qnli-dev_dot_f1": 0.678125,
+      "eval_Qnli-dev_dot_f1_threshold": 178.63389587402344,
+      "eval_Qnli-dev_dot_precision": 0.5371287128712872,
+      "eval_Qnli-dev_dot_recall": 0.9194915254237288,
+      "eval_Qnli-dev_euclidean_accuracy": 0.6953125,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 16.298263549804688,
+      "eval_Qnli-dev_euclidean_ap": 0.7198419743004867,
+      "eval_Qnli-dev_euclidean_f1": 0.6771929824561403,
+      "eval_Qnli-dev_euclidean_f1_threshold": 18.734228134155273,
+      "eval_Qnli-dev_euclidean_precision": 0.5778443113772455,
+      "eval_Qnli-dev_euclidean_recall": 0.8177966101694916,
+      "eval_Qnli-dev_manhattan_accuracy": 0.697265625,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 350.3825988769531,
+      "eval_Qnli-dev_manhattan_ap": 0.7225583351616828,
+      "eval_Qnli-dev_manhattan_f1": 0.6733466933867736,
+      "eval_Qnli-dev_manhattan_f1_threshold": 364.5922546386719,
+      "eval_Qnli-dev_manhattan_precision": 0.6387832699619772,
+      "eval_Qnli-dev_manhattan_recall": 0.711864406779661,
+      "eval_Qnli-dev_max_accuracy": 0.697265625,
+      "eval_Qnli-dev_max_accuracy_threshold": 350.3825988769531,
+      "eval_Qnli-dev_max_ap": 0.7225583351616828,
+      "eval_Qnli-dev_max_f1": 0.6783439490445861,
+      "eval_Qnli-dev_max_f1_threshold": 364.5922546386719,
+      "eval_Qnli-dev_max_precision": 0.6387832699619772,
+      "eval_Qnli-dev_max_recall": 0.9194915254237288,
+      "eval_allNLI-dev_cosine_accuracy": 0.73828125,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.7839766144752502,
+      "eval_allNLI-dev_cosine_ap": 0.6377040389504978,
+      "eval_allNLI-dev_cosine_f1": 0.6371681415929203,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.5921227931976318,
+      "eval_allNLI-dev_cosine_precision": 0.5161290322580645,
+      "eval_allNLI-dev_cosine_recall": 0.8323699421965318,
+      "eval_allNLI-dev_dot_accuracy": 0.72265625,
+      "eval_allNLI-dev_dot_accuracy_threshold": 287.0504150390625,
+      "eval_allNLI-dev_dot_ap": 0.5976338107788919,
+      "eval_allNLI-dev_dot_f1": 0.615,
+      "eval_allNLI-dev_dot_f1_threshold": 242.96902465820312,
+      "eval_allNLI-dev_dot_precision": 0.5418502202643172,
+      "eval_allNLI-dev_dot_recall": 0.7109826589595376,
+      "eval_allNLI-dev_euclidean_accuracy": 0.740234375,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 13.528583526611328,
+      "eval_allNLI-dev_euclidean_ap": 0.6364974679625237,
+      "eval_allNLI-dev_euclidean_f1": 0.6419753086419753,
+      "eval_allNLI-dev_euclidean_f1_threshold": 16.527751922607422,
+      "eval_allNLI-dev_euclidean_precision": 0.5603448275862069,
+      "eval_allNLI-dev_euclidean_recall": 0.7514450867052023,
+      "eval_allNLI-dev_manhattan_accuracy": 0.73828125,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 287.6244812011719,
+      "eval_allNLI-dev_manhattan_ap": 0.6359733229236403,
+      "eval_allNLI-dev_manhattan_f1": 0.6461538461538462,
+      "eval_allNLI-dev_manhattan_f1_threshold": 341.6584167480469,
+      "eval_allNLI-dev_manhattan_precision": 0.5806451612903226,
+      "eval_allNLI-dev_manhattan_recall": 0.7283236994219653,
+      "eval_allNLI-dev_max_accuracy": 0.740234375,
+      "eval_allNLI-dev_max_accuracy_threshold": 287.6244812011719,
+      "eval_allNLI-dev_max_ap": 0.6377040389504978,
+      "eval_allNLI-dev_max_f1": 0.6461538461538462,
+      "eval_allNLI-dev_max_f1_threshold": 341.6584167480469,
+      "eval_allNLI-dev_max_precision": 0.5806451612903226,
+      "eval_allNLI-dev_max_recall": 0.8323699421965318,
+      "eval_sequential_score": 0.7225583351616828,
+      "eval_sts-test_pearson_cosine": 0.8846526004146165,
+      "eval_sts-test_pearson_dot": 0.8694917588757636,
+      "eval_sts-test_pearson_euclidean": 0.903648547377558,
+      "eval_sts-test_pearson_manhattan": 0.9046900352481024,
+      "eval_sts-test_pearson_max": 0.9046900352481024,
+      "eval_sts-test_spearman_cosine": 0.9069678627017289,
+      "eval_sts-test_spearman_dot": 0.8661496554202353,
+      "eval_sts-test_spearman_euclidean": 0.8991392112877015,
+      "eval_sts-test_spearman_manhattan": 0.9004067012640901,
+      "eval_sts-test_spearman_max": 0.9069678627017289,
+      "eval_vitaminc-pairs_loss": 2.658703327178955,
+      "eval_vitaminc-pairs_runtime": 3.7894,
+      "eval_vitaminc-pairs_samples_per_second": 33.778,
+      "eval_vitaminc-pairs_steps_per_second": 0.264,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_negation-triplets_loss": 1.048051357269287,
+      "eval_negation-triplets_runtime": 0.5948,
+      "eval_negation-triplets_samples_per_second": 215.195,
+      "eval_negation-triplets_steps_per_second": 1.681,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_scitail-pairs-pos_loss": 0.05958803743124008,
+      "eval_scitail-pairs-pos_runtime": 0.8408,
+      "eval_scitail-pairs-pos_samples_per_second": 152.231,
+      "eval_scitail-pairs-pos_steps_per_second": 1.189,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_scitail-pairs-qa_loss": 4.770551527144562e-07,
+      "eval_scitail-pairs-qa_runtime": 0.5629,
+      "eval_scitail-pairs-qa_samples_per_second": 227.394,
+      "eval_scitail-pairs-qa_steps_per_second": 1.777,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_xsum-pairs_loss": 0.005623215343803167,
+      "eval_xsum-pairs_runtime": 2.7656,
+      "eval_xsum-pairs_samples_per_second": 46.283,
+      "eval_xsum-pairs_steps_per_second": 0.362,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_sciq_pairs_loss": 0.0031388087663799524,
+      "eval_sciq_pairs_runtime": 4.099,
+      "eval_sciq_pairs_samples_per_second": 31.227,
+      "eval_sciq_pairs_steps_per_second": 0.244,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_qasc_pairs_loss": 0.05423235148191452,
+      "eval_qasc_pairs_runtime": 0.596,
+      "eval_qasc_pairs_samples_per_second": 214.767,
+      "eval_qasc_pairs_steps_per_second": 1.678,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_openbookqa_pairs_loss": 1.2992339134216309,
+      "eval_openbookqa_pairs_runtime": 0.5529,
+      "eval_openbookqa_pairs_samples_per_second": 231.507,
+      "eval_openbookqa_pairs_steps_per_second": 1.809,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_msmarco_pairs_loss": 0.15928690135478973,
+      "eval_msmarco_pairs_runtime": 1.3201,
+      "eval_msmarco_pairs_samples_per_second": 96.963,
+      "eval_msmarco_pairs_steps_per_second": 0.758,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_nq_pairs_loss": 0.04052837938070297,
+      "eval_nq_pairs_runtime": 2.9252,
+      "eval_nq_pairs_samples_per_second": 43.758,
+      "eval_nq_pairs_steps_per_second": 0.342,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_trivia_pairs_loss": 0.21009953320026398,
+      "eval_trivia_pairs_runtime": 3.1824,
+      "eval_trivia_pairs_samples_per_second": 40.221,
+      "eval_trivia_pairs_steps_per_second": 0.314,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_gooaq_pairs_loss": 0.11401581764221191,
+      "eval_gooaq_pairs_runtime": 0.8553,
+      "eval_gooaq_pairs_samples_per_second": 149.66,
+      "eval_gooaq_pairs_steps_per_second": 1.169,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_paws-pos_loss": 0.04620524123311043,
+      "eval_paws-pos_runtime": 0.6561,
+      "eval_paws-pos_samples_per_second": 195.099,
+      "eval_paws-pos_steps_per_second": 1.524,
+      "step": 856
+    },
+    {
+      "epoch": 1.001169590643275,
+      "eval_global_dataset_loss": 0.20265640318393707,
+      "eval_global_dataset_runtime": 24.1741,
+      "eval_global_dataset_samples_per_second": 27.426,
+      "eval_global_dataset_steps_per_second": 0.248,
+      "step": 856
+    }
+  ],
+  "logging_steps": 11,
+  "max_steps": 4275,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 214,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 96,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-856/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aed6626d5e4516eecaff94019f86ab02eb588df865930f776737c69908518829
+size 5624