Training in progress, step 104, checkpoint

checkpoint-104/1_Pooling/config.json

@@ -0,0 +1,10 @@
+{
+  "word_embedding_dimension": 1536,
+  "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-104/added_tokens.json

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

checkpoint-104/config.json

@@ -0,0 +1,38 @@
+{
+  "_name_or_path": "microsoft/deberta-v2-xlarge",
+  "architectures": [
+    "DebertaV2Model"
+  ],
+  "attention_head_size": 64,
+  "attention_probs_dropout_prob": 0.1,
+  "conv_act": "gelu",
+  "conv_kernel_size": 3,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "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": 24,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 1536,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
+  "relative_attention": true,
+  "share_att_key": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.42.4",
+  "type_vocab_size": 0,
+  "vocab_size": 128100
+}

checkpoint-104/config_sentence_transformers.json

@@ -0,0 +1,10 @@
+{
+  "__version__": {
+    "sentence_transformers": "3.0.1",
+    "transformers": "4.42.4",
+    "pytorch": "2.4.0+cu121"
+  },
+  "prompts": {},
+  "default_prompt_name": null,
+  "similarity_fn_name": null
+}

checkpoint-104/modules.json

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

checkpoint-104/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03711586d86d98d07974115286b30db77f87b990c215cb55519faaf2e2349328
+size 7077084396

checkpoint-104/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:663b83f1d9a3333afdb6f468647c42e95752dee8bba6f61e56679fea719a9066
+size 3538506546

checkpoint-104/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8977fa1b1bcba243d602658153b863fe53a9a164e28641a100c49c959c6f6cbd
+size 14244

checkpoint-104/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43e3a487c8814200d9b028c0f0f955ce2a396f10a332ca3f721b9f7a9ede70e2
+size 1064

checkpoint-104/sentence_bert_config.json

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

checkpoint-104/special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-104/spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5598d5e96f339a8d980c15f9afd405a2e5e1be7db41de3ed13b0f03fac1e8c17
+size 2447305

checkpoint-104/tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "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]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}

checkpoint-104/trainer_state.json

@@ -0,0 +1,1043 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.20051413881748073,
+  "eval_steps": 26,
+  "global_step": 104,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005784061696658098,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 7.8503,
+      "step": 3
+    },
+    {
+      "epoch": 0.011568123393316195,
+      "grad_norm": 127.71849060058594,
+      "learning_rate": 5.791505791505792e-08,
+      "loss": 8.4022,
+      "step": 6
+    },
+    {
+      "epoch": 0.017352185089974295,
+      "grad_norm": 244.984619140625,
+      "learning_rate": 1.7374517374517373e-07,
+      "loss": 11.1776,
+      "step": 9
+    },
+    {
+      "epoch": 0.02313624678663239,
+      "grad_norm": 425.6427917480469,
+      "learning_rate": 3.4749034749034746e-07,
+      "loss": 9.7845,
+      "step": 12
+    },
+    {
+      "epoch": 0.02892030848329049,
+      "grad_norm": 238.8634490966797,
+      "learning_rate": 5.212355212355212e-07,
+      "loss": 8.9224,
+      "step": 15
+    },
+    {
+      "epoch": 0.03470437017994859,
+      "grad_norm": 173.4210968017578,
+      "learning_rate": 6.949806949806949e-07,
+      "loss": 11.1202,
+      "step": 18
+    },
+    {
+      "epoch": 0.04048843187660668,
+      "grad_norm": 77.89308166503906,
+      "learning_rate": 8.687258687258688e-07,
+      "loss": 7.413,
+      "step": 21
+    },
+    {
+      "epoch": 0.04627249357326478,
+      "grad_norm": Infinity,
+      "learning_rate": 9.845559845559844e-07,
+      "loss": 7.7803,
+      "step": 24
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_Qnli-dev_cosine_accuracy": 0.609375,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.9602330923080444,
+      "eval_Qnli-dev_cosine_ap": 0.587737015984062,
+      "eval_Qnli-dev_cosine_f1": 0.6358695652173914,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.7741116285324097,
+      "eval_Qnli-dev_cosine_precision": 0.468,
+      "eval_Qnli-dev_cosine_recall": 0.9915254237288136,
+      "eval_Qnli-dev_dot_accuracy": 0.56640625,
+      "eval_Qnli-dev_dot_accuracy_threshold": 750.123779296875,
+      "eval_Qnli-dev_dot_ap": 0.5198069471819926,
+      "eval_Qnli-dev_dot_f1": 0.6318607764390897,
+      "eval_Qnli-dev_dot_f1_threshold": 387.37371826171875,
+      "eval_Qnli-dev_dot_precision": 0.461839530332681,
+      "eval_Qnli-dev_dot_recall": 1.0,
+      "eval_Qnli-dev_euclidean_accuracy": 0.6171875,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 7.253280162811279,
+      "eval_Qnli-dev_euclidean_ap": 0.5993562103857255,
+      "eval_Qnli-dev_euclidean_f1": 0.6376021798365122,
+      "eval_Qnli-dev_euclidean_f1_threshold": 17.42431640625,
+      "eval_Qnli-dev_euclidean_precision": 0.46987951807228917,
+      "eval_Qnli-dev_euclidean_recall": 0.9915254237288136,
+      "eval_Qnli-dev_manhattan_accuracy": 0.625,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 212.0604705810547,
+      "eval_Qnli-dev_manhattan_ap": 0.609409682371795,
+      "eval_Qnli-dev_manhattan_f1": 0.6358695652173914,
+      "eval_Qnli-dev_manhattan_f1_threshold": 461.66192626953125,
+      "eval_Qnli-dev_manhattan_precision": 0.468,
+      "eval_Qnli-dev_manhattan_recall": 0.9915254237288136,
+      "eval_Qnli-dev_max_accuracy": 0.625,
+      "eval_Qnli-dev_max_accuracy_threshold": 750.123779296875,
+      "eval_Qnli-dev_max_ap": 0.609409682371795,
+      "eval_Qnli-dev_max_f1": 0.6376021798365122,
+      "eval_Qnli-dev_max_f1_threshold": 461.66192626953125,
+      "eval_Qnli-dev_max_precision": 0.46987951807228917,
+      "eval_Qnli-dev_max_recall": 1.0,
+      "eval_allNLI-dev_cosine_accuracy": 0.666015625,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9931652545928955,
+      "eval_allNLI-dev_cosine_ap": 0.3330277035108763,
+      "eval_allNLI-dev_cosine_f1": 0.5088235294117648,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.40660202503204346,
+      "eval_allNLI-dev_cosine_precision": 0.34122287968441817,
+      "eval_allNLI-dev_cosine_recall": 1.0,
+      "eval_allNLI-dev_dot_accuracy": 0.662109375,
+      "eval_allNLI-dev_dot_accuracy_threshold": 810.7718505859375,
+      "eval_allNLI-dev_dot_ap": 0.2980492563498839,
+      "eval_allNLI-dev_dot_f1": 0.5088235294117648,
+      "eval_allNLI-dev_dot_f1_threshold": 294.67071533203125,
+      "eval_allNLI-dev_dot_precision": 0.34122287968441817,
+      "eval_allNLI-dev_dot_recall": 1.0,
+      "eval_allNLI-dev_euclidean_accuracy": 0.666015625,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 3.3851852416992188,
+      "eval_allNLI-dev_euclidean_ap": 0.33790858355605125,
+      "eval_allNLI-dev_euclidean_f1": 0.5095729013254787,
+      "eval_allNLI-dev_euclidean_f1_threshold": 28.78042984008789,
+      "eval_allNLI-dev_euclidean_precision": 0.34189723320158105,
+      "eval_allNLI-dev_euclidean_recall": 1.0,
+      "eval_allNLI-dev_manhattan_accuracy": 0.666015625,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 98.2369613647461,
+      "eval_allNLI-dev_manhattan_ap": 0.3452833632811151,
+      "eval_allNLI-dev_manhattan_f1": 0.5091463414634146,
+      "eval_allNLI-dev_manhattan_f1_threshold": 626.5145874023438,
+      "eval_allNLI-dev_manhattan_precision": 0.34575569358178054,
+      "eval_allNLI-dev_manhattan_recall": 0.9653179190751445,
+      "eval_allNLI-dev_max_accuracy": 0.666015625,
+      "eval_allNLI-dev_max_accuracy_threshold": 810.7718505859375,
+      "eval_allNLI-dev_max_ap": 0.3452833632811151,
+      "eval_allNLI-dev_max_f1": 0.5095729013254787,
+      "eval_allNLI-dev_max_f1_threshold": 626.5145874023438,
+      "eval_allNLI-dev_max_precision": 0.34575569358178054,
+      "eval_allNLI-dev_max_recall": 1.0,
+      "eval_sequential_score": 0.609409682371795,
+      "eval_sts-test_pearson_cosine": 0.24255956652335486,
+      "eval_sts-test_pearson_dot": 0.18022025813169615,
+      "eval_sts-test_pearson_euclidean": 0.28657386934462287,
+      "eval_sts-test_pearson_manhattan": 0.32587411772454034,
+      "eval_sts-test_pearson_max": 0.32587411772454034,
+      "eval_sts-test_spearman_cosine": 0.32225123770630715,
+      "eval_sts-test_spearman_dot": 0.17184726031108927,
+      "eval_sts-test_spearman_euclidean": 0.3279675400801376,
+      "eval_sts-test_spearman_manhattan": 0.3635547276663291,
+      "eval_sts-test_spearman_max": 0.3635547276663291,
+      "eval_vitaminc-pairs_loss": 4.122440338134766,
+      "eval_vitaminc-pairs_runtime": 4.7337,
+      "eval_vitaminc-pairs_samples_per_second": 27.04,
+      "eval_vitaminc-pairs_steps_per_second": 0.211,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_negation-triplets_loss": 3.917590379714966,
+      "eval_negation-triplets_runtime": 3.3679,
+      "eval_negation-triplets_samples_per_second": 38.006,
+      "eval_negation-triplets_steps_per_second": 0.297,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_scitail-pairs-pos_loss": 2.5062460899353027,
+      "eval_scitail-pairs-pos_runtime": 2.7274,
+      "eval_scitail-pairs-pos_samples_per_second": 46.931,
+      "eval_scitail-pairs-pos_steps_per_second": 0.367,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_scitail-pairs-qa_loss": 2.978612184524536,
+      "eval_scitail-pairs-qa_runtime": 2.2452,
+      "eval_scitail-pairs-qa_samples_per_second": 57.011,
+      "eval_scitail-pairs-qa_steps_per_second": 0.445,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_xsum-pairs_loss": 5.394700050354004,
+      "eval_xsum-pairs_runtime": 3.0886,
+      "eval_xsum-pairs_samples_per_second": 41.442,
+      "eval_xsum-pairs_steps_per_second": 0.324,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_sciq_pairs_loss": 0.8977569937705994,
+      "eval_sciq_pairs_runtime": 4.7516,
+      "eval_sciq_pairs_samples_per_second": 26.938,
+      "eval_sciq_pairs_steps_per_second": 0.21,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_qasc_pairs_loss": 6.6808977127075195,
+      "eval_qasc_pairs_runtime": 2.1087,
+      "eval_qasc_pairs_samples_per_second": 60.701,
+      "eval_qasc_pairs_steps_per_second": 0.474,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_openbookqa_pairs_loss": 6.402599334716797,
+      "eval_openbookqa_pairs_runtime": 2.2222,
+      "eval_openbookqa_pairs_samples_per_second": 57.601,
+      "eval_openbookqa_pairs_steps_per_second": 0.45,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_msmarco_pairs_loss": 15.782674789428711,
+      "eval_msmarco_pairs_runtime": 2.307,
+      "eval_msmarco_pairs_samples_per_second": 55.483,
+      "eval_msmarco_pairs_steps_per_second": 0.433,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_nq_pairs_loss": 10.447425842285156,
+      "eval_nq_pairs_runtime": 3.6239,
+      "eval_nq_pairs_samples_per_second": 35.321,
+      "eval_nq_pairs_steps_per_second": 0.276,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_trivia_pairs_loss": 6.107333183288574,
+      "eval_trivia_pairs_runtime": 3.1299,
+      "eval_trivia_pairs_samples_per_second": 40.895,
+      "eval_trivia_pairs_steps_per_second": 0.319,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_gooaq_pairs_loss": 8.095273971557617,
+      "eval_gooaq_pairs_runtime": 2.1098,
+      "eval_gooaq_pairs_samples_per_second": 60.669,
+      "eval_gooaq_pairs_steps_per_second": 0.474,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_paws-pos_loss": 1.6077361106872559,
+      "eval_paws-pos_runtime": 2.2412,
+      "eval_paws-pos_samples_per_second": 57.112,
+      "eval_paws-pos_steps_per_second": 0.446,
+      "step": 26
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "eval_global_dataset_loss": 6.348417282104492,
+      "eval_global_dataset_runtime": 8.3683,
+      "eval_global_dataset_samples_per_second": 38.837,
+      "eval_global_dataset_steps_per_second": 0.358,
+      "step": 26
+    },
+    {
+      "epoch": 0.05205655526992288,
+      "grad_norm": 172.1571807861328,
+      "learning_rate": 1.1583011583011583e-06,
+      "loss": 7.9729,
+      "step": 27
+    },
+    {
+      "epoch": 0.05784061696658098,
+      "grad_norm": 106.302001953125,
+      "learning_rate": 1.3320463320463321e-06,
+      "loss": 6.0587,
+      "step": 30
+    },
+    {
+      "epoch": 0.06362467866323908,
+      "grad_norm": 72.31057739257812,
+      "learning_rate": 1.5057915057915057e-06,
+      "loss": 5.6742,
+      "step": 33
+    },
+    {
+      "epoch": 0.06940874035989718,
+      "grad_norm": 258.2577209472656,
+      "learning_rate": 1.6795366795366796e-06,
+      "loss": 6.5406,
+      "step": 36
+    },
+    {
+      "epoch": 0.07519280205655526,
+      "grad_norm": 156.0019073486328,
+      "learning_rate": 1.7953667953667953e-06,
+      "loss": 5.4429,
+      "step": 39
+    },
+    {
+      "epoch": 0.08097686375321336,
+      "grad_norm": 227.6062469482422,
+      "learning_rate": 1.969111969111969e-06,
+      "loss": 6.7855,
+      "step": 42
+    },
+    {
+      "epoch": 0.08676092544987146,
+      "grad_norm": 221.12347412109375,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 5.3403,
+      "step": 45
+    },
+    {
+      "epoch": 0.09254498714652956,
+      "grad_norm": 363.6319274902344,
+      "learning_rate": 2.3166023166023166e-06,
+      "loss": 4.2282,
+      "step": 48
+    },
+    {
+      "epoch": 0.09832904884318766,
+      "grad_norm": 119.22310638427734,
+      "learning_rate": 2.4903474903474904e-06,
+      "loss": 4.7411,
+      "step": 51
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_Qnli-dev_cosine_accuracy": 0.58984375,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.9407284259796143,
+      "eval_Qnli-dev_cosine_ap": 0.5957496162735323,
+      "eval_Qnli-dev_cosine_f1": 0.6314363143631436,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.7475543022155762,
+      "eval_Qnli-dev_cosine_precision": 0.4641434262948207,
+      "eval_Qnli-dev_cosine_recall": 0.9872881355932204,
+      "eval_Qnli-dev_dot_accuracy": 0.591796875,
+      "eval_Qnli-dev_dot_accuracy_threshold": 787.171630859375,
+      "eval_Qnli-dev_dot_ap": 0.556540262374452,
+      "eval_Qnli-dev_dot_f1": 0.6328767123287672,
+      "eval_Qnli-dev_dot_f1_threshold": 576.0565795898438,
+      "eval_Qnli-dev_dot_precision": 0.4676113360323887,
+      "eval_Qnli-dev_dot_recall": 0.9788135593220338,
+      "eval_Qnli-dev_euclidean_accuracy": 0.587890625,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 9.92269515991211,
+      "eval_Qnli-dev_euclidean_ap": 0.5913529071853763,
+      "eval_Qnli-dev_euclidean_f1": 0.6322930800542741,
+      "eval_Qnli-dev_euclidean_f1_threshold": 19.546676635742188,
+      "eval_Qnli-dev_euclidean_precision": 0.46506986027944114,
+      "eval_Qnli-dev_euclidean_recall": 0.9872881355932204,
+      "eval_Qnli-dev_manhattan_accuracy": 0.591796875,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 314.9248352050781,
+      "eval_Qnli-dev_manhattan_ap": 0.5944730102616272,
+      "eval_Qnli-dev_manhattan_f1": 0.632952691680261,
+      "eval_Qnli-dev_manhattan_f1_threshold": 424.9676818847656,
+      "eval_Qnli-dev_manhattan_precision": 0.5145888594164456,
+      "eval_Qnli-dev_manhattan_recall": 0.8220338983050848,
+      "eval_Qnli-dev_max_accuracy": 0.591796875,
+      "eval_Qnli-dev_max_accuracy_threshold": 787.171630859375,
+      "eval_Qnli-dev_max_ap": 0.5957496162735323,
+      "eval_Qnli-dev_max_f1": 0.632952691680261,
+      "eval_Qnli-dev_max_f1_threshold": 576.0565795898438,
+      "eval_Qnli-dev_max_precision": 0.5145888594164456,
+      "eval_Qnli-dev_max_recall": 0.9872881355932204,
+      "eval_allNLI-dev_cosine_accuracy": 0.666015625,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9878466129302979,
+      "eval_allNLI-dev_cosine_ap": 0.3898337405414008,
+      "eval_allNLI-dev_cosine_f1": 0.5065885797950219,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.6534674167633057,
+      "eval_allNLI-dev_cosine_precision": 0.3392156862745098,
+      "eval_allNLI-dev_cosine_recall": 1.0,
+      "eval_allNLI-dev_dot_accuracy": 0.666015625,
+      "eval_allNLI-dev_dot_accuracy_threshold": 925.5989990234375,
+      "eval_allNLI-dev_dot_ap": 0.3391870470132937,
+      "eval_allNLI-dev_dot_f1": 0.5046153846153847,
+      "eval_allNLI-dev_dot_f1_threshold": 608.7012939453125,
+      "eval_allNLI-dev_dot_precision": 0.3438155136268344,
+      "eval_allNLI-dev_dot_recall": 0.9479768786127167,
+      "eval_allNLI-dev_euclidean_accuracy": 0.666015625,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 4.708977699279785,
+      "eval_allNLI-dev_euclidean_ap": 0.39556533875401867,
+      "eval_allNLI-dev_euclidean_f1": 0.5080763582966226,
+      "eval_allNLI-dev_euclidean_f1_threshold": 23.160057067871094,
+      "eval_allNLI-dev_euclidean_precision": 0.3405511811023622,
+      "eval_allNLI-dev_euclidean_recall": 1.0,
+      "eval_allNLI-dev_manhattan_accuracy": 0.666015625,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 142.7260284423828,
+      "eval_allNLI-dev_manhattan_ap": 0.4050661514301679,
+      "eval_allNLI-dev_manhattan_f1": 0.5118483412322274,
+      "eval_allNLI-dev_manhattan_f1_threshold": 504.7154846191406,
+      "eval_allNLI-dev_manhattan_precision": 0.3521739130434783,
+      "eval_allNLI-dev_manhattan_recall": 0.9364161849710982,
+      "eval_allNLI-dev_max_accuracy": 0.666015625,
+      "eval_allNLI-dev_max_accuracy_threshold": 925.5989990234375,
+      "eval_allNLI-dev_max_ap": 0.4050661514301679,
+      "eval_allNLI-dev_max_f1": 0.5118483412322274,
+      "eval_allNLI-dev_max_f1_threshold": 608.7012939453125,
+      "eval_allNLI-dev_max_precision": 0.3521739130434783,
+      "eval_allNLI-dev_max_recall": 1.0,
+      "eval_sequential_score": 0.5957496162735323,
+      "eval_sts-test_pearson_cosine": 0.5067228548513025,
+      "eval_sts-test_pearson_dot": 0.28867418819353924,
+      "eval_sts-test_pearson_euclidean": 0.5315956249457565,
+      "eval_sts-test_pearson_manhattan": 0.5620907203006935,
+      "eval_sts-test_pearson_max": 0.5620907203006935,
+      "eval_sts-test_spearman_cosine": 0.5257409961657202,
+      "eval_sts-test_spearman_dot": 0.280153994622986,
+      "eval_sts-test_spearman_euclidean": 0.5306475244646732,
+      "eval_sts-test_spearman_manhattan": 0.5605804796765843,
+      "eval_sts-test_spearman_max": 0.5605804796765843,
+      "eval_vitaminc-pairs_loss": 3.929797887802124,
+      "eval_vitaminc-pairs_runtime": 4.6018,
+      "eval_vitaminc-pairs_samples_per_second": 27.815,
+      "eval_vitaminc-pairs_steps_per_second": 0.217,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_negation-triplets_loss": 3.267047166824341,
+      "eval_negation-triplets_runtime": 3.2813,
+      "eval_negation-triplets_samples_per_second": 39.009,
+      "eval_negation-triplets_steps_per_second": 0.305,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_scitail-pairs-pos_loss": 1.134178876876831,
+      "eval_scitail-pairs-pos_runtime": 2.6261,
+      "eval_scitail-pairs-pos_samples_per_second": 48.742,
+      "eval_scitail-pairs-pos_steps_per_second": 0.381,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_scitail-pairs-qa_loss": 2.0085601806640625,
+      "eval_scitail-pairs-qa_runtime": 2.2161,
+      "eval_scitail-pairs-qa_samples_per_second": 57.759,
+      "eval_scitail-pairs-qa_steps_per_second": 0.451,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_xsum-pairs_loss": 4.462350845336914,
+      "eval_xsum-pairs_runtime": 3.1126,
+      "eval_xsum-pairs_samples_per_second": 41.124,
+      "eval_xsum-pairs_steps_per_second": 0.321,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_sciq_pairs_loss": 0.71834796667099,
+      "eval_sciq_pairs_runtime": 4.7505,
+      "eval_sciq_pairs_samples_per_second": 26.944,
+      "eval_sciq_pairs_steps_per_second": 0.211,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_qasc_pairs_loss": 4.889453887939453,
+      "eval_qasc_pairs_runtime": 2.1101,
+      "eval_qasc_pairs_samples_per_second": 60.662,
+      "eval_qasc_pairs_steps_per_second": 0.474,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_openbookqa_pairs_loss": 3.4273369312286377,
+      "eval_openbookqa_pairs_runtime": 2.2311,
+      "eval_openbookqa_pairs_samples_per_second": 57.371,
+      "eval_openbookqa_pairs_steps_per_second": 0.448,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_msmarco_pairs_loss": 7.532540321350098,
+      "eval_msmarco_pairs_runtime": 2.2959,
+      "eval_msmarco_pairs_samples_per_second": 55.751,
+      "eval_msmarco_pairs_steps_per_second": 0.436,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_nq_pairs_loss": 6.29541015625,
+      "eval_nq_pairs_runtime": 3.661,
+      "eval_nq_pairs_samples_per_second": 34.963,
+      "eval_nq_pairs_steps_per_second": 0.273,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_trivia_pairs_loss": 5.789977073669434,
+      "eval_trivia_pairs_runtime": 3.0826,
+      "eval_trivia_pairs_samples_per_second": 41.523,
+      "eval_trivia_pairs_steps_per_second": 0.324,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_gooaq_pairs_loss": 5.785163879394531,
+      "eval_gooaq_pairs_runtime": 2.1382,
+      "eval_gooaq_pairs_samples_per_second": 59.863,
+      "eval_gooaq_pairs_steps_per_second": 0.468,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_paws-pos_loss": 0.47830095887184143,
+      "eval_paws-pos_runtime": 2.3113,
+      "eval_paws-pos_samples_per_second": 55.379,
+      "eval_paws-pos_steps_per_second": 0.433,
+      "step": 52
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "eval_global_dataset_loss": 3.8913991451263428,
+      "eval_global_dataset_runtime": 8.3623,
+      "eval_global_dataset_samples_per_second": 38.865,
+      "eval_global_dataset_steps_per_second": 0.359,
+      "step": 52
+    },
+    {
+      "epoch": 0.10411311053984576,
+      "grad_norm": 55.83478546142578,
+      "learning_rate": 2.6640926640926642e-06,
+      "loss": 3.9082,
+      "step": 54
+    },
+    {
+      "epoch": 0.10989717223650386,
+      "grad_norm": 57.4188117980957,
+      "learning_rate": 2.7799227799227797e-06,
+      "loss": 4.3922,
+      "step": 57
+    },
+    {
+      "epoch": 0.11568123393316196,
+      "grad_norm": 116.89701843261719,
+      "learning_rate": 2.9536679536679535e-06,
+      "loss": 3.2655,
+      "step": 60
+    },
+    {
+      "epoch": 0.12146529562982006,
+      "grad_norm": 53.49580001831055,
+      "learning_rate": 3.1274131274131274e-06,
+      "loss": 3.1043,
+      "step": 63
+    },
+    {
+      "epoch": 0.12724935732647816,
+      "grad_norm": 51.044410705566406,
+      "learning_rate": 3.3011583011583012e-06,
+      "loss": 2.2074,
+      "step": 66
+    },
+    {
+      "epoch": 0.13303341902313626,
+      "grad_norm": 139.9288787841797,
+      "learning_rate": 3.474903474903475e-06,
+      "loss": 1.4414,
+      "step": 69
+    },
+    {
+      "epoch": 0.13881748071979436,
+      "grad_norm": 37.221168518066406,
+      "learning_rate": 3.5907335907335905e-06,
+      "loss": 1.5937,
+      "step": 72
+    },
+    {
+      "epoch": 0.14460154241645246,
+      "grad_norm": 25.855661392211914,
+      "learning_rate": 3.7644787644787644e-06,
+      "loss": 1.0306,
+      "step": 75
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "grad_norm": 29.79730796813965,
+      "learning_rate": 3.938223938223938e-06,
+      "loss": 1.0784,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_Qnli-dev_cosine_accuracy": 0.638671875,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.856202244758606,
+      "eval_Qnli-dev_cosine_ap": 0.6292803059732273,
+      "eval_Qnli-dev_cosine_f1": 0.6408839779005525,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.6206527352333069,
+      "eval_Qnli-dev_cosine_precision": 0.47540983606557374,
+      "eval_Qnli-dev_cosine_recall": 0.9830508474576272,
+      "eval_Qnli-dev_dot_accuracy": 0.626953125,
+      "eval_Qnli-dev_dot_accuracy_threshold": 887.830810546875,
+      "eval_Qnli-dev_dot_ap": 0.592075047542057,
+      "eval_Qnli-dev_dot_f1": 0.640990371389271,
+      "eval_Qnli-dev_dot_f1_threshold": 601.355224609375,
+      "eval_Qnli-dev_dot_precision": 0.4745417515274949,
+      "eval_Qnli-dev_dot_recall": 0.9872881355932204,
+      "eval_Qnli-dev_euclidean_accuracy": 0.638671875,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 17.094371795654297,
+      "eval_Qnli-dev_euclidean_ap": 0.6289192660680781,
+      "eval_Qnli-dev_euclidean_f1": 0.6423357664233577,
+      "eval_Qnli-dev_euclidean_f1_threshold": 24.894683837890625,
+      "eval_Qnli-dev_euclidean_precision": 0.48997772828507796,
+      "eval_Qnli-dev_euclidean_recall": 0.9322033898305084,
+      "eval_Qnli-dev_manhattan_accuracy": 0.640625,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 526.0111083984375,
+      "eval_Qnli-dev_manhattan_ap": 0.6256143574081243,
+      "eval_Qnli-dev_manhattan_f1": 0.6417704011065007,
+      "eval_Qnli-dev_manhattan_f1_threshold": 853.242431640625,
+      "eval_Qnli-dev_manhattan_precision": 0.47638603696098564,
+      "eval_Qnli-dev_manhattan_recall": 0.9830508474576272,
+      "eval_Qnli-dev_max_accuracy": 0.640625,
+      "eval_Qnli-dev_max_accuracy_threshold": 887.830810546875,
+      "eval_Qnli-dev_max_ap": 0.6292803059732273,
+      "eval_Qnli-dev_max_f1": 0.6423357664233577,
+      "eval_Qnli-dev_max_f1_threshold": 853.242431640625,
+      "eval_Qnli-dev_max_precision": 0.48997772828507796,
+      "eval_Qnli-dev_max_recall": 0.9872881355932204,
+      "eval_allNLI-dev_cosine_accuracy": 0.6796875,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9426460862159729,
+      "eval_allNLI-dev_cosine_ap": 0.473508692504868,
+      "eval_allNLI-dev_cosine_f1": 0.5328947368421053,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.6884603500366211,
+      "eval_allNLI-dev_cosine_precision": 0.3724137931034483,
+      "eval_allNLI-dev_cosine_recall": 0.9364161849710982,
+      "eval_allNLI-dev_dot_accuracy": 0.673828125,
+      "eval_allNLI-dev_dot_accuracy_threshold": 980.9593505859375,
+      "eval_allNLI-dev_dot_ap": 0.44188020452374555,
+      "eval_allNLI-dev_dot_f1": 0.53156146179402,
+      "eval_allNLI-dev_dot_f1_threshold": 683.7774658203125,
+      "eval_allNLI-dev_dot_precision": 0.372960372960373,
+      "eval_allNLI-dev_dot_recall": 0.9248554913294798,
+      "eval_allNLI-dev_euclidean_accuracy": 0.6796875,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 10.647405624389648,
+      "eval_allNLI-dev_euclidean_ap": 0.47440611686548606,
+      "eval_allNLI-dev_euclidean_f1": 0.5292207792207793,
+      "eval_allNLI-dev_euclidean_f1_threshold": 25.08370590209961,
+      "eval_allNLI-dev_euclidean_precision": 0.36794582392776526,
+      "eval_allNLI-dev_euclidean_recall": 0.9421965317919075,
+      "eval_allNLI-dev_manhattan_accuracy": 0.6796875,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 318.408203125,
+      "eval_allNLI-dev_manhattan_ap": 0.4747137405292771,
+      "eval_allNLI-dev_manhattan_f1": 0.5305343511450381,
+      "eval_allNLI-dev_manhattan_f1_threshold": 697.8466796875,
+      "eval_allNLI-dev_manhattan_precision": 0.396011396011396,
+      "eval_allNLI-dev_manhattan_recall": 0.8034682080924855,
+      "eval_allNLI-dev_max_accuracy": 0.6796875,
+      "eval_allNLI-dev_max_accuracy_threshold": 980.9593505859375,
+      "eval_allNLI-dev_max_ap": 0.4747137405292771,
+      "eval_allNLI-dev_max_f1": 0.5328947368421053,
+      "eval_allNLI-dev_max_f1_threshold": 697.8466796875,
+      "eval_allNLI-dev_max_precision": 0.396011396011396,
+      "eval_allNLI-dev_max_recall": 0.9421965317919075,
+      "eval_sequential_score": 0.6292803059732273,
+      "eval_sts-test_pearson_cosine": 0.8216305666941104,
+      "eval_sts-test_pearson_dot": 0.7893058557928705,
+      "eval_sts-test_pearson_euclidean": 0.841701309677291,
+      "eval_sts-test_pearson_manhattan": 0.8438559225680149,
+      "eval_sts-test_pearson_max": 0.8438559225680149,
+      "eval_sts-test_spearman_cosine": 0.8291168678373173,
+      "eval_sts-test_spearman_dot": 0.7782024179758329,
+      "eval_sts-test_spearman_euclidean": 0.829029110613729,
+      "eval_sts-test_spearman_manhattan": 0.8308224279114665,
+      "eval_sts-test_spearman_max": 0.8308224279114665,
+      "eval_vitaminc-pairs_loss": 3.763850212097168,
+      "eval_vitaminc-pairs_runtime": 4.6142,
+      "eval_vitaminc-pairs_samples_per_second": 27.741,
+      "eval_vitaminc-pairs_steps_per_second": 0.217,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_negation-triplets_loss": 1.4772971868515015,
+      "eval_negation-triplets_runtime": 3.4145,
+      "eval_negation-triplets_samples_per_second": 37.488,
+      "eval_negation-triplets_steps_per_second": 0.293,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_scitail-pairs-pos_loss": 0.11086193472146988,
+      "eval_scitail-pairs-pos_runtime": 2.8091,
+      "eval_scitail-pairs-pos_samples_per_second": 45.565,
+      "eval_scitail-pairs-pos_steps_per_second": 0.356,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_scitail-pairs-qa_loss": 0.12822097539901733,
+      "eval_scitail-pairs-qa_runtime": 2.312,
+      "eval_scitail-pairs-qa_samples_per_second": 55.364,
+      "eval_scitail-pairs-qa_steps_per_second": 0.433,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_xsum-pairs_loss": 0.43598121404647827,
+      "eval_xsum-pairs_runtime": 3.1471,
+      "eval_xsum-pairs_samples_per_second": 40.673,
+      "eval_xsum-pairs_steps_per_second": 0.318,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_sciq_pairs_loss": 0.20338551700115204,
+      "eval_sciq_pairs_runtime": 4.8592,
+      "eval_sciq_pairs_samples_per_second": 26.342,
+      "eval_sciq_pairs_steps_per_second": 0.206,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_qasc_pairs_loss": 1.3059492111206055,
+      "eval_qasc_pairs_runtime": 2.1191,
+      "eval_qasc_pairs_samples_per_second": 60.403,
+      "eval_qasc_pairs_steps_per_second": 0.472,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_openbookqa_pairs_loss": 1.1711227893829346,
+      "eval_openbookqa_pairs_runtime": 2.2105,
+      "eval_openbookqa_pairs_samples_per_second": 57.907,
+      "eval_openbookqa_pairs_steps_per_second": 0.452,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_msmarco_pairs_loss": 1.3856205940246582,
+      "eval_msmarco_pairs_runtime": 2.3065,
+      "eval_msmarco_pairs_samples_per_second": 55.494,
+      "eval_msmarco_pairs_steps_per_second": 0.434,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_nq_pairs_loss": 1.2713654041290283,
+      "eval_nq_pairs_runtime": 3.6729,
+      "eval_nq_pairs_samples_per_second": 34.85,
+      "eval_nq_pairs_steps_per_second": 0.272,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_trivia_pairs_loss": 1.082784652709961,
+      "eval_trivia_pairs_runtime": 3.145,
+      "eval_trivia_pairs_samples_per_second": 40.7,
+      "eval_trivia_pairs_steps_per_second": 0.318,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_gooaq_pairs_loss": 0.703183650970459,
+      "eval_gooaq_pairs_runtime": 2.1975,
+      "eval_gooaq_pairs_samples_per_second": 58.248,
+      "eval_gooaq_pairs_steps_per_second": 0.455,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_paws-pos_loss": 0.04489962384104729,
+      "eval_paws-pos_runtime": 2.3636,
+      "eval_paws-pos_samples_per_second": 54.155,
+      "eval_paws-pos_steps_per_second": 0.423,
+      "step": 78
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "eval_global_dataset_loss": 0.6713590621948242,
+      "eval_global_dataset_runtime": 8.4648,
+      "eval_global_dataset_samples_per_second": 38.394,
+      "eval_global_dataset_steps_per_second": 0.354,
+      "step": 78
+    },
+    {
+      "epoch": 0.15616966580976863,
+      "grad_norm": 22.163389205932617,
+      "learning_rate": 4.1119691119691125e-06,
+      "loss": 0.9674,
+      "step": 81
+    },
+    {
+      "epoch": 0.16195372750642673,
+      "grad_norm": 25.121131896972656,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.9335,
+      "step": 84
+    },
+    {
+      "epoch": 0.16773778920308482,
+      "grad_norm": 44.308528900146484,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.8806,
+      "step": 87
+    },
+    {
+      "epoch": 0.17352185089974292,
+      "grad_norm": 9.972362518310547,
+      "learning_rate": 4.633204633204633e-06,
+      "loss": 0.631,
+      "step": 90
+    },
+    {
+      "epoch": 0.17930591259640102,
+      "grad_norm": 13.609597206115723,
+      "learning_rate": 4.806949806949808e-06,
+      "loss": 0.3384,
+      "step": 93
+    },
+    {
+      "epoch": 0.18508997429305912,
+      "grad_norm": 6.892751693725586,
+      "learning_rate": 4.980694980694981e-06,
+      "loss": 0.404,
+      "step": 96
+    },
+    {
+      "epoch": 0.19087403598971722,
+      "grad_norm": 7.414526462554932,
+      "learning_rate": 5.154440154440155e-06,
+      "loss": 0.6488,
+      "step": 99
+    },
+    {
+      "epoch": 0.19665809768637532,
+      "grad_norm": 23.75099754333496,
+      "learning_rate": 5.3281853281853285e-06,
+      "loss": 0.4728,
+      "step": 102
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_Qnli-dev_cosine_accuracy": 0.693359375,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.8276629447937012,
+      "eval_Qnli-dev_cosine_ap": 0.7127503402574797,
+      "eval_Qnli-dev_cosine_f1": 0.6730038022813688,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.7761298418045044,
+      "eval_Qnli-dev_cosine_precision": 0.6103448275862069,
+      "eval_Qnli-dev_cosine_recall": 0.75,
+      "eval_Qnli-dev_dot_accuracy": 0.681640625,
+      "eval_Qnli-dev_dot_accuracy_threshold": 857.5540771484375,
+      "eval_Qnli-dev_dot_ap": 0.6617256136301952,
+      "eval_Qnli-dev_dot_f1": 0.6678141135972461,
+      "eval_Qnli-dev_dot_f1_threshold": 756.1309204101562,
+      "eval_Qnli-dev_dot_precision": 0.5623188405797102,
+      "eval_Qnli-dev_dot_recall": 0.8220338983050848,
+      "eval_Qnli-dev_euclidean_accuracy": 0.69140625,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 18.45541763305664,
+      "eval_Qnli-dev_euclidean_ap": 0.7154480594709788,
+      "eval_Qnli-dev_euclidean_f1": 0.6725043782837128,
+      "eval_Qnli-dev_euclidean_f1_threshold": 22.723068237304688,
+      "eval_Qnli-dev_euclidean_precision": 0.573134328358209,
+      "eval_Qnli-dev_euclidean_recall": 0.8135593220338984,
+      "eval_Qnli-dev_manhattan_accuracy": 0.69140625,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 585.3442993164062,
+      "eval_Qnli-dev_manhattan_ap": 0.713289230749026,
+      "eval_Qnli-dev_manhattan_f1": 0.6733333333333333,
+      "eval_Qnli-dev_manhattan_f1_threshold": 733.2077026367188,
+      "eval_Qnli-dev_manhattan_precision": 0.554945054945055,
+      "eval_Qnli-dev_manhattan_recall": 0.8559322033898306,
+      "eval_Qnli-dev_max_accuracy": 0.693359375,
+      "eval_Qnli-dev_max_accuracy_threshold": 857.5540771484375,
+      "eval_Qnli-dev_max_ap": 0.7154480594709788,
+      "eval_Qnli-dev_max_f1": 0.6733333333333333,
+      "eval_Qnli-dev_max_f1_threshold": 756.1309204101562,
+      "eval_Qnli-dev_max_precision": 0.6103448275862069,
+      "eval_Qnli-dev_max_recall": 0.8559322033898306,
+      "eval_allNLI-dev_cosine_accuracy": 0.693359375,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.8820043802261353,
+      "eval_allNLI-dev_cosine_ap": 0.5380653098565689,
+      "eval_allNLI-dev_cosine_f1": 0.5689655172413792,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.7308224439620972,
+      "eval_allNLI-dev_cosine_precision": 0.4536082474226804,
+      "eval_allNLI-dev_cosine_recall": 0.7630057803468208,
+      "eval_allNLI-dev_dot_accuracy": 0.6875,
+      "eval_allNLI-dev_dot_accuracy_threshold": 904.0134887695312,
+      "eval_allNLI-dev_dot_ap": 0.5175334662627952,
+      "eval_allNLI-dev_dot_f1": 0.5675675675675675,
+      "eval_allNLI-dev_dot_f1_threshold": 762.652587890625,
+      "eval_allNLI-dev_dot_precision": 0.46494464944649444,
+      "eval_allNLI-dev_dot_recall": 0.7283236994219653,
+      "eval_allNLI-dev_euclidean_accuracy": 0.6953125,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 15.50227165222168,
+      "eval_allNLI-dev_euclidean_ap": 0.5410854453600531,
+      "eval_allNLI-dev_euclidean_f1": 0.5645514223194749,
+      "eval_allNLI-dev_euclidean_f1_threshold": 23.451454162597656,
+      "eval_allNLI-dev_euclidean_precision": 0.45422535211267606,
+      "eval_allNLI-dev_euclidean_recall": 0.7456647398843931,
+      "eval_allNLI-dev_manhattan_accuracy": 0.6953125,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 518.75146484375,
+      "eval_allNLI-dev_manhattan_ap": 0.5430149701961973,
+      "eval_allNLI-dev_manhattan_f1": 0.5639913232104121,
+      "eval_allNLI-dev_manhattan_f1_threshold": 733.8897705078125,
+      "eval_allNLI-dev_manhattan_precision": 0.4513888888888889,
+      "eval_allNLI-dev_manhattan_recall": 0.7514450867052023,
+      "eval_allNLI-dev_max_accuracy": 0.6953125,
+      "eval_allNLI-dev_max_accuracy_threshold": 904.0134887695312,
+      "eval_allNLI-dev_max_ap": 0.5430149701961973,
+      "eval_allNLI-dev_max_f1": 0.5689655172413792,
+      "eval_allNLI-dev_max_f1_threshold": 762.652587890625,
+      "eval_allNLI-dev_max_precision": 0.46494464944649444,
+      "eval_allNLI-dev_max_recall": 0.7630057803468208,
+      "eval_sequential_score": 0.7154480594709788,
+      "eval_sts-test_pearson_cosine": 0.8841456597076958,
+      "eval_sts-test_pearson_dot": 0.8597263949983345,
+      "eval_sts-test_pearson_euclidean": 0.8977156894521804,
+      "eval_sts-test_pearson_manhattan": 0.8973099980705126,
+      "eval_sts-test_pearson_max": 0.8977156894521804,
+      "eval_sts-test_spearman_cosine": 0.8862653474156692,
+      "eval_sts-test_spearman_dot": 0.8451249310521576,
+      "eval_sts-test_spearman_euclidean": 0.8866065310325173,
+      "eval_sts-test_spearman_manhattan": 0.8860841943095571,
+      "eval_sts-test_spearman_max": 0.8866065310325173,
+      "eval_vitaminc-pairs_loss": 3.314167022705078,
+      "eval_vitaminc-pairs_runtime": 4.4886,
+      "eval_vitaminc-pairs_samples_per_second": 28.517,
+      "eval_vitaminc-pairs_steps_per_second": 0.223,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_negation-triplets_loss": 1.3400739431381226,
+      "eval_negation-triplets_runtime": 3.3068,
+      "eval_negation-triplets_samples_per_second": 38.708,
+      "eval_negation-triplets_steps_per_second": 0.302,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_scitail-pairs-pos_loss": 0.07263079285621643,
+      "eval_scitail-pairs-pos_runtime": 2.6213,
+      "eval_scitail-pairs-pos_samples_per_second": 48.83,
+      "eval_scitail-pairs-pos_steps_per_second": 0.381,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_scitail-pairs-qa_loss": 0.048478204756975174,
+      "eval_scitail-pairs-qa_runtime": 2.202,
+      "eval_scitail-pairs-qa_samples_per_second": 58.13,
+      "eval_scitail-pairs-qa_steps_per_second": 0.454,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_xsum-pairs_loss": 0.14091075956821442,
+      "eval_xsum-pairs_runtime": 3.0891,
+      "eval_xsum-pairs_samples_per_second": 41.436,
+      "eval_xsum-pairs_steps_per_second": 0.324,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_sciq_pairs_loss": 0.1505264937877655,
+      "eval_sciq_pairs_runtime": 4.762,
+      "eval_sciq_pairs_samples_per_second": 26.879,
+      "eval_sciq_pairs_steps_per_second": 0.21,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_qasc_pairs_loss": 0.4244489073753357,
+      "eval_qasc_pairs_runtime": 2.0783,
+      "eval_qasc_pairs_samples_per_second": 61.589,
+      "eval_qasc_pairs_steps_per_second": 0.481,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_openbookqa_pairs_loss": 0.5923758745193481,
+      "eval_openbookqa_pairs_runtime": 2.1938,
+      "eval_openbookqa_pairs_samples_per_second": 58.345,
+      "eval_openbookqa_pairs_steps_per_second": 0.456,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_msmarco_pairs_loss": 0.5882836580276489,
+      "eval_msmarco_pairs_runtime": 2.2483,
+      "eval_msmarco_pairs_samples_per_second": 56.933,
+      "eval_msmarco_pairs_steps_per_second": 0.445,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_nq_pairs_loss": 0.5424538254737854,
+      "eval_nq_pairs_runtime": 3.6424,
+      "eval_nq_pairs_samples_per_second": 35.142,
+      "eval_nq_pairs_steps_per_second": 0.275,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_trivia_pairs_loss": 0.6437886953353882,
+      "eval_trivia_pairs_runtime": 3.0809,
+      "eval_trivia_pairs_samples_per_second": 41.546,
+      "eval_trivia_pairs_steps_per_second": 0.325,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_gooaq_pairs_loss": 0.22378715872764587,
+      "eval_gooaq_pairs_runtime": 2.101,
+      "eval_gooaq_pairs_samples_per_second": 60.924,
+      "eval_gooaq_pairs_steps_per_second": 0.476,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_paws-pos_loss": 0.03581737354397774,
+      "eval_paws-pos_runtime": 2.2447,
+      "eval_paws-pos_samples_per_second": 57.024,
+      "eval_paws-pos_steps_per_second": 0.445,
+      "step": 104
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "eval_global_dataset_loss": 0.3989756405353546,
+      "eval_global_dataset_runtime": 8.402,
+      "eval_global_dataset_samples_per_second": 38.681,
+      "eval_global_dataset_steps_per_second": 0.357,
+      "step": 104
+    }
+  ],
+  "logging_steps": 3,
+  "max_steps": 1036,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 52,
+  "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": 64,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-104/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73b62473318b3b28c705f94b87acfb0646732bae15a12e95a938b998dc89c2e7
+size 5560