Training in progress, step 22, checkpoint

checkpoint-22/1_Pooling/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-22/added_tokens.json

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

checkpoint-22/config.json

@@ -0,0 +1,35 @@
+{
+  "_name_or_path": "microsoft/deberta-v3-small",
+  "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": "4.42.3",
+  "type_vocab_size": 0,
+  "vocab_size": 128100
+}

checkpoint-22/config_sentence_transformers.json

@@ -0,0 +1,10 @@
+{
+  "__version__": {
+    "sentence_transformers": "3.0.1",
+    "transformers": "4.42.3",
+    "pytorch": "2.1.2"
+  },
+  "prompts": {},
+  "default_prompt_name": null,
+  "similarity_fn_name": null
+}

checkpoint-22/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-22/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8539aa1b6965657584b76742b43d76a69ffd50c3324dfb394b519c7fbf2e61aa
+size 1130520122

checkpoint-22/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdc35bb45cf1a7fc12552a11861fa204c62164025cca9f4bea942f1d5f7bccfd
+size 565251810

checkpoint-22/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ade7e6083110813e38f7c518497d010aeb9e329eb9d11adcd275aeac34305dd2
+size 14244

checkpoint-22/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e98539a41ac284b9d0f25a0c37e52f5f08c3158a42c16ac21682452a58936300
+size 1064

checkpoint-22/sentence_bert_config.json

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

checkpoint-22/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-22/spm.model

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

checkpoint-22/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-22/trainer_state.json

@@ -0,0 +1,1804 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6395348837209303,
+  "eval_steps": 2,
+  "global_step": 22,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.029069767441860465,
+      "grad_norm": 13.73514461517334,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 6.7536,
+      "step": 1
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "grad_norm": 16.892961502075195,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 6.6203,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.109375,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.55078125,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9469717741012573,
+      "eval_VitaminC_cosine_ap": 0.5165514227338435,
+      "eval_VitaminC_cosine_f1": 0.6525198938992042,
+      "eval_VitaminC_cosine_f1_threshold": 0.4987494945526123,
+      "eval_VitaminC_cosine_precision": 0.484251968503937,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.55078125,
+      "eval_VitaminC_dot_accuracy_threshold": 417.45166015625,
+      "eval_VitaminC_dot_ap": 0.5125282416460941,
+      "eval_VitaminC_dot_f1": 0.6525198938992042,
+      "eval_VitaminC_dot_f1_threshold": 200.84530639648438,
+      "eval_VitaminC_dot_precision": 0.484251968503937,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.546875,
+      "eval_VitaminC_euclidean_accuracy_threshold": 6.841136932373047,
+      "eval_VitaminC_euclidean_ap": 0.512994913068431,
+      "eval_VitaminC_euclidean_f1": 0.6525198938992042,
+      "eval_VitaminC_euclidean_f1_threshold": 20.22179412841797,
+      "eval_VitaminC_euclidean_precision": 0.484251968503937,
+      "eval_VitaminC_euclidean_recall": 1.0,
+      "eval_VitaminC_manhattan_accuracy": 0.546875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 117.1143569946289,
+      "eval_VitaminC_manhattan_ap": 0.5177961641566705,
+      "eval_VitaminC_manhattan_f1": 0.6542553191489362,
+      "eval_VitaminC_manhattan_f1_threshold": 292.3347473144531,
+      "eval_VitaminC_manhattan_precision": 0.48616600790513836,
+      "eval_VitaminC_manhattan_recall": 1.0,
+      "eval_VitaminC_max_accuracy": 0.55078125,
+      "eval_VitaminC_max_accuracy_threshold": 417.45166015625,
+      "eval_VitaminC_max_ap": 0.5177961641566705,
+      "eval_VitaminC_max_f1": 0.6542553191489362,
+      "eval_VitaminC_max_f1_threshold": 292.3347473144531,
+      "eval_VitaminC_max_precision": 0.48616600790513836,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5177961641566705,
+      "eval_sts-test_pearson_cosine": 0.017178505918243114,
+      "eval_sts-test_pearson_dot": 0.16228524441844774,
+      "eval_sts-test_pearson_euclidean": 0.024391561236282323,
+      "eval_sts-test_pearson_manhattan": 0.059272519732015624,
+      "eval_sts-test_pearson_max": 0.16228524441844774,
+      "eval_sts-test_spearman_cosine": 0.07115563415775981,
+      "eval_sts-test_spearman_dot": 0.1967074668301984,
+      "eval_sts-test_spearman_euclidean": 0.05117306486959643,
+      "eval_sts-test_spearman_manhattan": 0.07641354546391785,
+      "eval_sts-test_spearman_max": 0.1967074668301984,
+      "eval_vitaminc-pairs_loss": 2.72658371925354,
+      "eval_vitaminc-pairs_runtime": 1.4719,
+      "eval_vitaminc-pairs_samples_per_second": 73.373,
+      "eval_vitaminc-pairs_steps_per_second": 1.359,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_negation-triplets_loss": 5.094177722930908,
+      "eval_negation-triplets_runtime": 0.3027,
+      "eval_negation-triplets_samples_per_second": 211.46,
+      "eval_negation-triplets_steps_per_second": 3.304,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_scitail-pairs-pos_loss": 1.9183871746063232,
+      "eval_scitail-pairs-pos_runtime": 0.3785,
+      "eval_scitail-pairs-pos_samples_per_second": 142.654,
+      "eval_scitail-pairs-pos_steps_per_second": 2.642,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_xsum-pairs_loss": 6.3277668952941895,
+      "eval_xsum-pairs_runtime": 3.4205,
+      "eval_xsum-pairs_samples_per_second": 37.422,
+      "eval_xsum-pairs_steps_per_second": 0.585,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_sciq_pairs_loss": 0.36539140343666077,
+      "eval_sciq_pairs_runtime": 3.4462,
+      "eval_sciq_pairs_samples_per_second": 37.143,
+      "eval_sciq_pairs_steps_per_second": 0.58,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_qasc_pairs_loss": 3.9689104557037354,
+      "eval_qasc_pairs_runtime": 0.6477,
+      "eval_qasc_pairs_samples_per_second": 197.636,
+      "eval_qasc_pairs_steps_per_second": 3.088,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_openbookqa_pairs_loss": 4.7439374923706055,
+      "eval_openbookqa_pairs_runtime": 0.5759,
+      "eval_openbookqa_pairs_samples_per_second": 222.248,
+      "eval_openbookqa_pairs_steps_per_second": 3.473,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_msmarco_pairs_loss": 10.51363754272461,
+      "eval_msmarco_pairs_runtime": 1.3036,
+      "eval_msmarco_pairs_samples_per_second": 98.189,
+      "eval_msmarco_pairs_steps_per_second": 1.534,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_nq_pairs_loss": 4.969011306762695,
+      "eval_nq_pairs_runtime": 2.5922,
+      "eval_nq_pairs_samples_per_second": 49.378,
+      "eval_nq_pairs_steps_per_second": 0.772,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_trivia_pairs_loss": 3.8609507083892822,
+      "eval_trivia_pairs_runtime": 4.3666,
+      "eval_trivia_pairs_samples_per_second": 29.314,
+      "eval_trivia_pairs_steps_per_second": 0.458,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_gooaq_pairs_loss": 8.041110038757324,
+      "eval_gooaq_pairs_runtime": 0.9119,
+      "eval_gooaq_pairs_samples_per_second": 140.361,
+      "eval_gooaq_pairs_steps_per_second": 2.193,
+      "step": 2
+    },
+    {
+      "epoch": 0.05813953488372093,
+      "eval_paws-pos_loss": 2.21897029876709,
+      "eval_paws-pos_runtime": 0.6905,
+      "eval_paws-pos_samples_per_second": 185.364,
+      "eval_paws-pos_steps_per_second": 2.896,
+      "step": 2
+    },
+    {
+      "epoch": 0.0872093023255814,
+      "grad_norm": 15.801941871643066,
+      "learning_rate": 1e-05,
+      "loss": 6.7963,
+      "step": 3
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "grad_norm": 11.94517993927002,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 6.4488,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.109375,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.55078125,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9437637329101562,
+      "eval_VitaminC_cosine_ap": 0.5161989667198088,
+      "eval_VitaminC_cosine_f1": 0.6525198938992042,
+      "eval_VitaminC_cosine_f1_threshold": 0.56722092628479,
+      "eval_VitaminC_cosine_precision": 0.484251968503937,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.546875,
+      "eval_VitaminC_dot_accuracy_threshold": 416.9976806640625,
+      "eval_VitaminC_dot_ap": 0.515337582265173,
+      "eval_VitaminC_dot_f1": 0.6507936507936508,
+      "eval_VitaminC_dot_f1_threshold": 223.1047821044922,
+      "eval_VitaminC_dot_precision": 0.4823529411764706,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.55078125,
+      "eval_VitaminC_euclidean_accuracy_threshold": 6.841610908508301,
+      "eval_VitaminC_euclidean_ap": 0.5167890569390883,
+      "eval_VitaminC_euclidean_f1": 0.6542553191489362,
+      "eval_VitaminC_euclidean_f1_threshold": 18.548978805541992,
+      "eval_VitaminC_euclidean_precision": 0.48616600790513836,
+      "eval_VitaminC_euclidean_recall": 1.0,
+      "eval_VitaminC_manhattan_accuracy": 0.54296875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 110.40855407714844,
+      "eval_VitaminC_manhattan_ap": 0.5219582103185639,
+      "eval_VitaminC_manhattan_f1": 0.6542553191489362,
+      "eval_VitaminC_manhattan_f1_threshold": 282.48602294921875,
+      "eval_VitaminC_manhattan_precision": 0.48616600790513836,
+      "eval_VitaminC_manhattan_recall": 1.0,
+      "eval_VitaminC_max_accuracy": 0.55078125,
+      "eval_VitaminC_max_accuracy_threshold": 416.9976806640625,
+      "eval_VitaminC_max_ap": 0.5219582103185639,
+      "eval_VitaminC_max_f1": 0.6542553191489362,
+      "eval_VitaminC_max_f1_threshold": 282.48602294921875,
+      "eval_VitaminC_max_precision": 0.48616600790513836,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5219582103185639,
+      "eval_sts-test_pearson_cosine": 0.025276809579094173,
+      "eval_sts-test_pearson_dot": 0.21470147681711343,
+      "eval_sts-test_pearson_euclidean": 0.027783806662194415,
+      "eval_sts-test_pearson_manhattan": 0.059736177006118926,
+      "eval_sts-test_pearson_max": 0.21470147681711343,
+      "eval_sts-test_spearman_cosine": 0.0776783582433888,
+      "eval_sts-test_spearman_dot": 0.2491347682204262,
+      "eval_sts-test_spearman_euclidean": 0.054640386155863654,
+      "eval_sts-test_spearman_manhattan": 0.07764379287610436,
+      "eval_sts-test_spearman_max": 0.2491347682204262,
+      "eval_vitaminc-pairs_loss": 2.7045021057128906,
+      "eval_vitaminc-pairs_runtime": 1.4394,
+      "eval_vitaminc-pairs_samples_per_second": 75.033,
+      "eval_vitaminc-pairs_steps_per_second": 1.389,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_negation-triplets_loss": 5.081584453582764,
+      "eval_negation-triplets_runtime": 0.2952,
+      "eval_negation-triplets_samples_per_second": 216.79,
+      "eval_negation-triplets_steps_per_second": 3.387,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_scitail-pairs-pos_loss": 1.9186776876449585,
+      "eval_scitail-pairs-pos_runtime": 0.3702,
+      "eval_scitail-pairs-pos_samples_per_second": 145.879,
+      "eval_scitail-pairs-pos_steps_per_second": 2.701,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_xsum-pairs_loss": 6.199001312255859,
+      "eval_xsum-pairs_runtime": 3.4227,
+      "eval_xsum-pairs_samples_per_second": 37.398,
+      "eval_xsum-pairs_steps_per_second": 0.584,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_sciq_pairs_loss": 0.3414052426815033,
+      "eval_sciq_pairs_runtime": 3.4403,
+      "eval_sciq_pairs_samples_per_second": 37.206,
+      "eval_sciq_pairs_steps_per_second": 0.581,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_qasc_pairs_loss": 3.662200927734375,
+      "eval_qasc_pairs_runtime": 0.6486,
+      "eval_qasc_pairs_samples_per_second": 197.349,
+      "eval_qasc_pairs_steps_per_second": 3.084,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_openbookqa_pairs_loss": 4.650759220123291,
+      "eval_openbookqa_pairs_runtime": 0.5722,
+      "eval_openbookqa_pairs_samples_per_second": 223.702,
+      "eval_openbookqa_pairs_steps_per_second": 3.495,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_msmarco_pairs_loss": 9.487885475158691,
+      "eval_msmarco_pairs_runtime": 1.2978,
+      "eval_msmarco_pairs_samples_per_second": 98.63,
+      "eval_msmarco_pairs_steps_per_second": 1.541,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_nq_pairs_loss": 4.4714131355285645,
+      "eval_nq_pairs_runtime": 2.569,
+      "eval_nq_pairs_samples_per_second": 49.824,
+      "eval_nq_pairs_steps_per_second": 0.779,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_trivia_pairs_loss": 3.5246145725250244,
+      "eval_trivia_pairs_runtime": 4.3557,
+      "eval_trivia_pairs_samples_per_second": 29.387,
+      "eval_trivia_pairs_steps_per_second": 0.459,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_gooaq_pairs_loss": 7.395092010498047,
+      "eval_gooaq_pairs_runtime": 0.9109,
+      "eval_gooaq_pairs_samples_per_second": 140.521,
+      "eval_gooaq_pairs_steps_per_second": 2.196,
+      "step": 4
+    },
+    {
+      "epoch": 0.11627906976744186,
+      "eval_paws-pos_loss": 2.233201742172241,
+      "eval_paws-pos_runtime": 0.6853,
+      "eval_paws-pos_samples_per_second": 186.773,
+      "eval_paws-pos_steps_per_second": 2.918,
+      "step": 4
+    },
+    {
+      "epoch": 0.14534883720930233,
+      "grad_norm": 6.336462497711182,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 6.5567,
+      "step": 5
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "grad_norm": 30.24117660522461,
+      "learning_rate": 2e-05,
+      "loss": 7.994,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.125,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.546875,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9505010843276978,
+      "eval_VitaminC_cosine_ap": 0.5220332354916479,
+      "eval_VitaminC_cosine_f1": 0.6542553191489362,
+      "eval_VitaminC_cosine_f1_threshold": 0.6742120385169983,
+      "eval_VitaminC_cosine_precision": 0.48616600790513836,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.55859375,
+      "eval_VitaminC_dot_accuracy_threshold": 410.96356201171875,
+      "eval_VitaminC_dot_ap": 0.5098545013010067,
+      "eval_VitaminC_dot_f1": 0.6507936507936508,
+      "eval_VitaminC_dot_f1_threshold": 259.4541015625,
+      "eval_VitaminC_dot_precision": 0.4823529411764706,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.55078125,
+      "eval_VitaminC_euclidean_accuracy_threshold": 6.85032844543457,
+      "eval_VitaminC_euclidean_ap": 0.5187640034678729,
+      "eval_VitaminC_euclidean_f1": 0.6522911051212938,
+      "eval_VitaminC_euclidean_f1_threshold": 15.907812118530273,
+      "eval_VitaminC_euclidean_precision": 0.4879032258064516,
+      "eval_VitaminC_euclidean_recall": 0.983739837398374,
+      "eval_VitaminC_manhattan_accuracy": 0.54296875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 128.04562377929688,
+      "eval_VitaminC_manhattan_ap": 0.5216728184773658,
+      "eval_VitaminC_manhattan_f1": 0.6542553191489362,
+      "eval_VitaminC_manhattan_f1_threshold": 266.50201416015625,
+      "eval_VitaminC_manhattan_precision": 0.48616600790513836,
+      "eval_VitaminC_manhattan_recall": 1.0,
+      "eval_VitaminC_max_accuracy": 0.55859375,
+      "eval_VitaminC_max_accuracy_threshold": 410.96356201171875,
+      "eval_VitaminC_max_ap": 0.5220332354916479,
+      "eval_VitaminC_max_f1": 0.6542553191489362,
+      "eval_VitaminC_max_f1_threshold": 266.50201416015625,
+      "eval_VitaminC_max_precision": 0.4879032258064516,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5220332354916479,
+      "eval_sts-test_pearson_cosine": 0.03816033089318139,
+      "eval_sts-test_pearson_dot": 0.2714501319216333,
+      "eval_sts-test_pearson_euclidean": 0.03534045207172642,
+      "eval_sts-test_pearson_manhattan": 0.061786120590017285,
+      "eval_sts-test_pearson_max": 0.2714501319216333,
+      "eval_sts-test_spearman_cosine": 0.09093735567962451,
+      "eval_sts-test_spearman_dot": 0.29608861437488704,
+      "eval_sts-test_spearman_euclidean": 0.06076512778358849,
+      "eval_sts-test_spearman_manhattan": 0.08153826351124775,
+      "eval_sts-test_spearman_max": 0.29608861437488704,
+      "eval_vitaminc-pairs_loss": 2.6921818256378174,
+      "eval_vitaminc-pairs_runtime": 1.4315,
+      "eval_vitaminc-pairs_samples_per_second": 75.443,
+      "eval_vitaminc-pairs_steps_per_second": 1.397,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_negation-triplets_loss": 5.068139553070068,
+      "eval_negation-triplets_runtime": 0.2981,
+      "eval_negation-triplets_samples_per_second": 214.716,
+      "eval_negation-triplets_steps_per_second": 3.355,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_scitail-pairs-pos_loss": 1.9319576025009155,
+      "eval_scitail-pairs-pos_runtime": 0.3653,
+      "eval_scitail-pairs-pos_samples_per_second": 147.818,
+      "eval_scitail-pairs-pos_steps_per_second": 2.737,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_xsum-pairs_loss": 6.079549789428711,
+      "eval_xsum-pairs_runtime": 3.4013,
+      "eval_xsum-pairs_samples_per_second": 37.633,
+      "eval_xsum-pairs_steps_per_second": 0.588,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_sciq_pairs_loss": 0.32075461745262146,
+      "eval_sciq_pairs_runtime": 3.442,
+      "eval_sciq_pairs_samples_per_second": 37.188,
+      "eval_sciq_pairs_steps_per_second": 0.581,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_qasc_pairs_loss": 3.36326265335083,
+      "eval_qasc_pairs_runtime": 0.6495,
+      "eval_qasc_pairs_samples_per_second": 197.087,
+      "eval_qasc_pairs_steps_per_second": 3.079,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_openbookqa_pairs_loss": 4.481137275695801,
+      "eval_openbookqa_pairs_runtime": 0.5729,
+      "eval_openbookqa_pairs_samples_per_second": 223.432,
+      "eval_openbookqa_pairs_steps_per_second": 3.491,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_msmarco_pairs_loss": 8.048794746398926,
+      "eval_msmarco_pairs_runtime": 1.2936,
+      "eval_msmarco_pairs_samples_per_second": 98.947,
+      "eval_msmarco_pairs_steps_per_second": 1.546,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_nq_pairs_loss": 3.7927472591400146,
+      "eval_nq_pairs_runtime": 2.5733,
+      "eval_nq_pairs_samples_per_second": 49.741,
+      "eval_nq_pairs_steps_per_second": 0.777,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_trivia_pairs_loss": 3.284496545791626,
+      "eval_trivia_pairs_runtime": 4.355,
+      "eval_trivia_pairs_samples_per_second": 29.391,
+      "eval_trivia_pairs_steps_per_second": 0.459,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_gooaq_pairs_loss": 6.677769184112549,
+      "eval_gooaq_pairs_runtime": 0.9178,
+      "eval_gooaq_pairs_samples_per_second": 139.457,
+      "eval_gooaq_pairs_steps_per_second": 2.179,
+      "step": 6
+    },
+    {
+      "epoch": 0.1744186046511628,
+      "eval_paws-pos_loss": 2.2626476287841797,
+      "eval_paws-pos_runtime": 0.6901,
+      "eval_paws-pos_samples_per_second": 185.475,
+      "eval_paws-pos_steps_per_second": 2.898,
+      "step": 6
+    },
+    {
+      "epoch": 0.20348837209302326,
+      "grad_norm": 11.503045082092285,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 7.1037,
+      "step": 7
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "grad_norm": 6.524661540985107,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 6.6239,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.140625,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.55078125,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9556466341018677,
+      "eval_VitaminC_cosine_ap": 0.5262234678756146,
+      "eval_VitaminC_cosine_f1": 0.6542553191489362,
+      "eval_VitaminC_cosine_f1_threshold": 0.7482036352157593,
+      "eval_VitaminC_cosine_precision": 0.48616600790513836,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.54296875,
+      "eval_VitaminC_dot_accuracy_threshold": 429.2105712890625,
+      "eval_VitaminC_dot_ap": 0.5105928204741528,
+      "eval_VitaminC_dot_f1": 0.6525198938992042,
+      "eval_VitaminC_dot_f1_threshold": 304.46807861328125,
+      "eval_VitaminC_dot_precision": 0.484251968503937,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.55078125,
+      "eval_VitaminC_euclidean_accuracy_threshold": 6.457396030426025,
+      "eval_VitaminC_euclidean_ap": 0.5262113804807371,
+      "eval_VitaminC_euclidean_f1": 0.6505376344086021,
+      "eval_VitaminC_euclidean_f1_threshold": 14.363262176513672,
+      "eval_VitaminC_euclidean_precision": 0.4859437751004016,
+      "eval_VitaminC_euclidean_recall": 0.983739837398374,
+      "eval_VitaminC_manhattan_accuracy": 0.546875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 144.2735137939453,
+      "eval_VitaminC_manhattan_ap": 0.5216238817948092,
+      "eval_VitaminC_manhattan_f1": 0.6542553191489362,
+      "eval_VitaminC_manhattan_f1_threshold": 248.78077697753906,
+      "eval_VitaminC_manhattan_precision": 0.48616600790513836,
+      "eval_VitaminC_manhattan_recall": 1.0,
+      "eval_VitaminC_max_accuracy": 0.55078125,
+      "eval_VitaminC_max_accuracy_threshold": 429.2105712890625,
+      "eval_VitaminC_max_ap": 0.5262234678756146,
+      "eval_VitaminC_max_f1": 0.6542553191489362,
+      "eval_VitaminC_max_f1_threshold": 304.46807861328125,
+      "eval_VitaminC_max_precision": 0.48616600790513836,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5262234678756146,
+      "eval_sts-test_pearson_cosine": 0.04822016371804072,
+      "eval_sts-test_pearson_dot": 0.2933829804301914,
+      "eval_sts-test_pearson_euclidean": 0.0440427995340741,
+      "eval_sts-test_pearson_manhattan": 0.06814345626491929,
+      "eval_sts-test_pearson_max": 0.2933829804301914,
+      "eval_sts-test_spearman_cosine": 0.10311449730221736,
+      "eval_sts-test_spearman_dot": 0.30840433621335606,
+      "eval_sts-test_spearman_euclidean": 0.06897119310369988,
+      "eval_sts-test_spearman_manhattan": 0.08808657402390177,
+      "eval_sts-test_spearman_max": 0.30840433621335606,
+      "eval_vitaminc-pairs_loss": 2.7006633281707764,
+      "eval_vitaminc-pairs_runtime": 1.435,
+      "eval_vitaminc-pairs_samples_per_second": 75.262,
+      "eval_vitaminc-pairs_steps_per_second": 1.394,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_negation-triplets_loss": 5.0855937004089355,
+      "eval_negation-triplets_runtime": 0.2946,
+      "eval_negation-triplets_samples_per_second": 217.208,
+      "eval_negation-triplets_steps_per_second": 3.394,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_scitail-pairs-pos_loss": 1.9701284170150757,
+      "eval_scitail-pairs-pos_runtime": 0.3657,
+      "eval_scitail-pairs-pos_samples_per_second": 147.656,
+      "eval_scitail-pairs-pos_steps_per_second": 2.734,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_xsum-pairs_loss": 6.050904273986816,
+      "eval_xsum-pairs_runtime": 3.4161,
+      "eval_xsum-pairs_samples_per_second": 37.47,
+      "eval_xsum-pairs_steps_per_second": 0.585,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_sciq_pairs_loss": 0.31550008058547974,
+      "eval_sciq_pairs_runtime": 3.4195,
+      "eval_sciq_pairs_samples_per_second": 37.432,
+      "eval_sciq_pairs_steps_per_second": 0.585,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_qasc_pairs_loss": 3.174623727798462,
+      "eval_qasc_pairs_runtime": 0.6393,
+      "eval_qasc_pairs_samples_per_second": 200.217,
+      "eval_qasc_pairs_steps_per_second": 3.128,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_openbookqa_pairs_loss": 4.325982570648193,
+      "eval_openbookqa_pairs_runtime": 0.5723,
+      "eval_openbookqa_pairs_samples_per_second": 223.64,
+      "eval_openbookqa_pairs_steps_per_second": 3.494,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_msmarco_pairs_loss": 6.989845275878906,
+      "eval_msmarco_pairs_runtime": 1.2986,
+      "eval_msmarco_pairs_samples_per_second": 98.569,
+      "eval_msmarco_pairs_steps_per_second": 1.54,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_nq_pairs_loss": 3.3526666164398193,
+      "eval_nq_pairs_runtime": 2.5766,
+      "eval_nq_pairs_samples_per_second": 49.677,
+      "eval_nq_pairs_steps_per_second": 0.776,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_trivia_pairs_loss": 3.2416629791259766,
+      "eval_trivia_pairs_runtime": 4.3655,
+      "eval_trivia_pairs_samples_per_second": 29.321,
+      "eval_trivia_pairs_steps_per_second": 0.458,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_gooaq_pairs_loss": 6.288425922393799,
+      "eval_gooaq_pairs_runtime": 0.9111,
+      "eval_gooaq_pairs_samples_per_second": 140.492,
+      "eval_gooaq_pairs_steps_per_second": 2.195,
+      "step": 8
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "eval_paws-pos_loss": 2.351145029067993,
+      "eval_paws-pos_runtime": 0.6907,
+      "eval_paws-pos_samples_per_second": 185.325,
+      "eval_paws-pos_steps_per_second": 2.896,
+      "step": 8
+    },
+    {
+      "epoch": 0.2616279069767442,
+      "grad_norm": 15.498976707458496,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 6.7359,
+      "step": 9
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "grad_norm": 17.9885311126709,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 7.0187,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.140625,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.55078125,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9581937789916992,
+      "eval_VitaminC_cosine_ap": 0.5266754197706615,
+      "eval_VitaminC_cosine_f1": 0.6507936507936508,
+      "eval_VitaminC_cosine_f1_threshold": 0.7856193780899048,
+      "eval_VitaminC_cosine_precision": 0.4823529411764706,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.54296875,
+      "eval_VitaminC_dot_accuracy_threshold": 461.7385559082031,
+      "eval_VitaminC_dot_ap": 0.5148613370532991,
+      "eval_VitaminC_dot_f1": 0.6542553191489362,
+      "eval_VitaminC_dot_f1_threshold": 349.2696838378906,
+      "eval_VitaminC_dot_precision": 0.48616600790513836,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.55078125,
+      "eval_VitaminC_euclidean_accuracy_threshold": 6.300814628601074,
+      "eval_VitaminC_euclidean_ap": 0.5255273000837065,
+      "eval_VitaminC_euclidean_f1": 0.6472148541114058,
+      "eval_VitaminC_euclidean_f1_threshold": 14.141785621643066,
+      "eval_VitaminC_euclidean_precision": 0.48031496062992124,
+      "eval_VitaminC_euclidean_recall": 0.991869918699187,
+      "eval_VitaminC_manhattan_accuracy": 0.546875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 111.04672241210938,
+      "eval_VitaminC_manhattan_ap": 0.5200203121459024,
+      "eval_VitaminC_manhattan_f1": 0.6542553191489362,
+      "eval_VitaminC_manhattan_f1_threshold": 232.8947296142578,
+      "eval_VitaminC_manhattan_precision": 0.48616600790513836,
+      "eval_VitaminC_manhattan_recall": 1.0,
+      "eval_VitaminC_max_accuracy": 0.55078125,
+      "eval_VitaminC_max_accuracy_threshold": 461.7385559082031,
+      "eval_VitaminC_max_ap": 0.5266754197706615,
+      "eval_VitaminC_max_f1": 0.6542553191489362,
+      "eval_VitaminC_max_f1_threshold": 349.2696838378906,
+      "eval_VitaminC_max_precision": 0.48616600790513836,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5266754197706615,
+      "eval_sts-test_pearson_cosine": 0.05719150926706661,
+      "eval_sts-test_pearson_dot": 0.2946717285740077,
+      "eval_sts-test_pearson_euclidean": 0.05619658108650874,
+      "eval_sts-test_pearson_manhattan": 0.07947531332122215,
+      "eval_sts-test_pearson_max": 0.2946717285740077,
+      "eval_sts-test_spearman_cosine": 0.11293393987132996,
+      "eval_sts-test_spearman_dot": 0.3047454601059247,
+      "eval_sts-test_spearman_euclidean": 0.07970793034154534,
+      "eval_sts-test_spearman_manhattan": 0.09745056138447042,
+      "eval_sts-test_spearman_max": 0.3047454601059247,
+      "eval_vitaminc-pairs_loss": 2.715181589126587,
+      "eval_vitaminc-pairs_runtime": 1.4378,
+      "eval_vitaminc-pairs_samples_per_second": 75.113,
+      "eval_vitaminc-pairs_steps_per_second": 1.391,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_negation-triplets_loss": 5.137111186981201,
+      "eval_negation-triplets_runtime": 0.296,
+      "eval_negation-triplets_samples_per_second": 216.18,
+      "eval_negation-triplets_steps_per_second": 3.378,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_scitail-pairs-pos_loss": 2.0431957244873047,
+      "eval_scitail-pairs-pos_runtime": 0.3627,
+      "eval_scitail-pairs-pos_samples_per_second": 148.876,
+      "eval_scitail-pairs-pos_steps_per_second": 2.757,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_xsum-pairs_loss": 5.958922863006592,
+      "eval_xsum-pairs_runtime": 3.4184,
+      "eval_xsum-pairs_samples_per_second": 37.445,
+      "eval_xsum-pairs_steps_per_second": 0.585,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_sciq_pairs_loss": 0.3122929334640503,
+      "eval_sciq_pairs_runtime": 3.4301,
+      "eval_sciq_pairs_samples_per_second": 37.316,
+      "eval_sciq_pairs_steps_per_second": 0.583,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_qasc_pairs_loss": 3.0287587642669678,
+      "eval_qasc_pairs_runtime": 0.6451,
+      "eval_qasc_pairs_samples_per_second": 198.424,
+      "eval_qasc_pairs_steps_per_second": 3.1,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_openbookqa_pairs_loss": 4.213818073272705,
+      "eval_openbookqa_pairs_runtime": 0.5726,
+      "eval_openbookqa_pairs_samples_per_second": 223.54,
+      "eval_openbookqa_pairs_steps_per_second": 3.493,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_msmarco_pairs_loss": 6.443023204803467,
+      "eval_msmarco_pairs_runtime": 1.2926,
+      "eval_msmarco_pairs_samples_per_second": 99.023,
+      "eval_msmarco_pairs_steps_per_second": 1.547,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_nq_pairs_loss": 3.1351583003997803,
+      "eval_nq_pairs_runtime": 2.579,
+      "eval_nq_pairs_samples_per_second": 49.632,
+      "eval_nq_pairs_steps_per_second": 0.776,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_trivia_pairs_loss": 3.116779088973999,
+      "eval_trivia_pairs_runtime": 4.3604,
+      "eval_trivia_pairs_samples_per_second": 29.355,
+      "eval_trivia_pairs_steps_per_second": 0.459,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_gooaq_pairs_loss": 6.086311340332031,
+      "eval_gooaq_pairs_runtime": 0.9098,
+      "eval_gooaq_pairs_samples_per_second": 140.686,
+      "eval_gooaq_pairs_steps_per_second": 2.198,
+      "step": 10
+    },
+    {
+      "epoch": 0.29069767441860467,
+      "eval_paws-pos_loss": 2.509471893310547,
+      "eval_paws-pos_runtime": 0.6868,
+      "eval_paws-pos_samples_per_second": 186.374,
+      "eval_paws-pos_steps_per_second": 2.912,
+      "step": 10
+    },
+    {
+      "epoch": 0.31976744186046513,
+      "grad_norm": 8.210014343261719,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 6.4394,
+      "step": 11
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "grad_norm": 4.276110649108887,
+      "learning_rate": 4e-05,
+      "loss": 6.2215,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.125,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.54296875,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.967127799987793,
+      "eval_VitaminC_cosine_ap": 0.5255548283042585,
+      "eval_VitaminC_cosine_f1": 0.648936170212766,
+      "eval_VitaminC_cosine_f1_threshold": 0.8157126903533936,
+      "eval_VitaminC_cosine_precision": 0.48221343873517786,
+      "eval_VitaminC_cosine_recall": 0.991869918699187,
+      "eval_VitaminC_dot_accuracy": 0.54296875,
+      "eval_VitaminC_dot_accuracy_threshold": 494.9104309082031,
+      "eval_VitaminC_dot_ap": 0.5147267697897451,
+      "eval_VitaminC_dot_f1": 0.6525198938992042,
+      "eval_VitaminC_dot_f1_threshold": 383.80987548828125,
+      "eval_VitaminC_dot_precision": 0.484251968503937,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.546875,
+      "eval_VitaminC_euclidean_accuracy_threshold": 5.804706573486328,
+      "eval_VitaminC_euclidean_ap": 0.5251569661235761,
+      "eval_VitaminC_euclidean_f1": 0.648936170212766,
+      "eval_VitaminC_euclidean_f1_threshold": 13.634963035583496,
+      "eval_VitaminC_euclidean_precision": 0.48221343873517786,
+      "eval_VitaminC_euclidean_recall": 0.991869918699187,
+      "eval_VitaminC_manhattan_accuracy": 0.55078125,
+      "eval_VitaminC_manhattan_accuracy_threshold": 115.37125396728516,
+      "eval_VitaminC_manhattan_ap": 0.5211715241173163,
+      "eval_VitaminC_manhattan_f1": 0.6525198938992042,
+      "eval_VitaminC_manhattan_f1_threshold": 222.44180297851562,
+      "eval_VitaminC_manhattan_precision": 0.484251968503937,
+      "eval_VitaminC_manhattan_recall": 1.0,
+      "eval_VitaminC_max_accuracy": 0.55078125,
+      "eval_VitaminC_max_accuracy_threshold": 494.9104309082031,
+      "eval_VitaminC_max_ap": 0.5255548283042585,
+      "eval_VitaminC_max_f1": 0.6525198938992042,
+      "eval_VitaminC_max_f1_threshold": 383.80987548828125,
+      "eval_VitaminC_max_precision": 0.484251968503937,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5255548283042585,
+      "eval_sts-test_pearson_cosine": 0.07362980702318303,
+      "eval_sts-test_pearson_dot": 0.2873931216538933,
+      "eval_sts-test_pearson_euclidean": 0.0777294747568082,
+      "eval_sts-test_pearson_manhattan": 0.09990133962066747,
+      "eval_sts-test_pearson_max": 0.2873931216538933,
+      "eval_sts-test_spearman_cosine": 0.1307277513542105,
+      "eval_sts-test_spearman_dot": 0.29102261593628775,
+      "eval_sts-test_spearman_euclidean": 0.10450013684970334,
+      "eval_sts-test_spearman_manhattan": 0.1162952592302797,
+      "eval_sts-test_spearman_max": 0.29102261593628775,
+      "eval_vitaminc-pairs_loss": 2.7209765911102295,
+      "eval_vitaminc-pairs_runtime": 1.4394,
+      "eval_vitaminc-pairs_samples_per_second": 75.03,
+      "eval_vitaminc-pairs_steps_per_second": 1.389,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_negation-triplets_loss": 5.181106090545654,
+      "eval_negation-triplets_runtime": 0.2954,
+      "eval_negation-triplets_samples_per_second": 216.692,
+      "eval_negation-triplets_steps_per_second": 3.386,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_scitail-pairs-pos_loss": 2.0938117504119873,
+      "eval_scitail-pairs-pos_runtime": 0.3664,
+      "eval_scitail-pairs-pos_samples_per_second": 147.368,
+      "eval_scitail-pairs-pos_steps_per_second": 2.729,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_xsum-pairs_loss": 5.798585891723633,
+      "eval_xsum-pairs_runtime": 3.4113,
+      "eval_xsum-pairs_samples_per_second": 37.522,
+      "eval_xsum-pairs_steps_per_second": 0.586,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_sciq_pairs_loss": 0.30859243869781494,
+      "eval_sciq_pairs_runtime": 3.4356,
+      "eval_sciq_pairs_samples_per_second": 37.256,
+      "eval_sciq_pairs_steps_per_second": 0.582,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_qasc_pairs_loss": 2.919435739517212,
+      "eval_qasc_pairs_runtime": 0.6393,
+      "eval_qasc_pairs_samples_per_second": 200.207,
+      "eval_qasc_pairs_steps_per_second": 3.128,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_openbookqa_pairs_loss": 4.164936542510986,
+      "eval_openbookqa_pairs_runtime": 0.5772,
+      "eval_openbookqa_pairs_samples_per_second": 221.764,
+      "eval_openbookqa_pairs_steps_per_second": 3.465,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_msmarco_pairs_loss": 6.209752082824707,
+      "eval_msmarco_pairs_runtime": 1.296,
+      "eval_msmarco_pairs_samples_per_second": 98.767,
+      "eval_msmarco_pairs_steps_per_second": 1.543,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_nq_pairs_loss": 3.008612632751465,
+      "eval_nq_pairs_runtime": 2.5746,
+      "eval_nq_pairs_samples_per_second": 49.716,
+      "eval_nq_pairs_steps_per_second": 0.777,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_trivia_pairs_loss": 2.97579288482666,
+      "eval_trivia_pairs_runtime": 4.3688,
+      "eval_trivia_pairs_samples_per_second": 29.299,
+      "eval_trivia_pairs_steps_per_second": 0.458,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_gooaq_pairs_loss": 5.966132640838623,
+      "eval_gooaq_pairs_runtime": 0.9077,
+      "eval_gooaq_pairs_samples_per_second": 141.017,
+      "eval_gooaq_pairs_steps_per_second": 2.203,
+      "step": 12
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "eval_paws-pos_loss": 2.638349771499634,
+      "eval_paws-pos_runtime": 0.6823,
+      "eval_paws-pos_samples_per_second": 187.606,
+      "eval_paws-pos_steps_per_second": 2.931,
+      "step": 12
+    },
+    {
+      "epoch": 0.37790697674418605,
+      "grad_norm": 3.267026662826538,
+      "learning_rate": 3.9847321628214e-05,
+      "loss": 6.2269,
+      "step": 13
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "grad_norm": 2.8631503582000732,
+      "learning_rate": 3.9392394604217464e-05,
+      "loss": 6.3233,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.125,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.546875,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9713834524154663,
+      "eval_VitaminC_cosine_ap": 0.5312073832689701,
+      "eval_VitaminC_cosine_f1": 0.648936170212766,
+      "eval_VitaminC_cosine_f1_threshold": 0.8415127992630005,
+      "eval_VitaminC_cosine_precision": 0.48221343873517786,
+      "eval_VitaminC_cosine_recall": 0.991869918699187,
+      "eval_VitaminC_dot_accuracy": 0.55078125,
+      "eval_VitaminC_dot_accuracy_threshold": 488.3980407714844,
+      "eval_VitaminC_dot_ap": 0.5169851993943358,
+      "eval_VitaminC_dot_f1": 0.6525198938992042,
+      "eval_VitaminC_dot_f1_threshold": 400.4466247558594,
+      "eval_VitaminC_dot_precision": 0.484251968503937,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.55078125,
+      "eval_VitaminC_euclidean_accuracy_threshold": 5.628746032714844,
+      "eval_VitaminC_euclidean_ap": 0.529795456897044,
+      "eval_VitaminC_euclidean_f1": 0.648936170212766,
+      "eval_VitaminC_euclidean_f1_threshold": 12.81901741027832,
+      "eval_VitaminC_euclidean_precision": 0.48221343873517786,
+      "eval_VitaminC_euclidean_recall": 0.991869918699187,
+      "eval_VitaminC_manhattan_accuracy": 0.546875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 113.0782470703125,
+      "eval_VitaminC_manhattan_ap": 0.5209454866207013,
+      "eval_VitaminC_manhattan_f1": 0.6522911051212938,
+      "eval_VitaminC_manhattan_f1_threshold": 189.79293823242188,
+      "eval_VitaminC_manhattan_precision": 0.4879032258064516,
+      "eval_VitaminC_manhattan_recall": 0.983739837398374,
+      "eval_VitaminC_max_accuracy": 0.55078125,
+      "eval_VitaminC_max_accuracy_threshold": 488.3980407714844,
+      "eval_VitaminC_max_ap": 0.5312073832689701,
+      "eval_VitaminC_max_f1": 0.6525198938992042,
+      "eval_VitaminC_max_f1_threshold": 400.4466247558594,
+      "eval_VitaminC_max_precision": 0.4879032258064516,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5312073832689701,
+      "eval_sts-test_pearson_cosine": 0.10291373182144412,
+      "eval_sts-test_pearson_dot": 0.2825251334692641,
+      "eval_sts-test_pearson_euclidean": 0.10855326758086989,
+      "eval_sts-test_pearson_manhattan": 0.1290688905692468,
+      "eval_sts-test_pearson_max": 0.2825251334692641,
+      "eval_sts-test_spearman_cosine": 0.1576804098079836,
+      "eval_sts-test_spearman_dot": 0.29155896943554727,
+      "eval_sts-test_spearman_euclidean": 0.13546971780560696,
+      "eval_sts-test_spearman_manhattan": 0.14696980995695116,
+      "eval_sts-test_spearman_max": 0.29155896943554727,
+      "eval_vitaminc-pairs_loss": 2.7145626544952393,
+      "eval_vitaminc-pairs_runtime": 1.435,
+      "eval_vitaminc-pairs_samples_per_second": 75.259,
+      "eval_vitaminc-pairs_steps_per_second": 1.394,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_negation-triplets_loss": 5.162342548370361,
+      "eval_negation-triplets_runtime": 0.2955,
+      "eval_negation-triplets_samples_per_second": 216.609,
+      "eval_negation-triplets_steps_per_second": 3.385,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_scitail-pairs-pos_loss": 2.047813892364502,
+      "eval_scitail-pairs-pos_runtime": 0.3677,
+      "eval_scitail-pairs-pos_samples_per_second": 146.843,
+      "eval_scitail-pairs-pos_steps_per_second": 2.719,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_xsum-pairs_loss": 5.738613605499268,
+      "eval_xsum-pairs_runtime": 3.4232,
+      "eval_xsum-pairs_samples_per_second": 37.392,
+      "eval_xsum-pairs_steps_per_second": 0.584,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_sciq_pairs_loss": 0.30535876750946045,
+      "eval_sciq_pairs_runtime": 3.4492,
+      "eval_sciq_pairs_samples_per_second": 37.11,
+      "eval_sciq_pairs_steps_per_second": 0.58,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_qasc_pairs_loss": 2.853950262069702,
+      "eval_qasc_pairs_runtime": 0.6383,
+      "eval_qasc_pairs_samples_per_second": 200.542,
+      "eval_qasc_pairs_steps_per_second": 3.133,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_openbookqa_pairs_loss": 4.114726543426514,
+      "eval_openbookqa_pairs_runtime": 0.5752,
+      "eval_openbookqa_pairs_samples_per_second": 222.522,
+      "eval_openbookqa_pairs_steps_per_second": 3.477,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_msmarco_pairs_loss": 6.093601226806641,
+      "eval_msmarco_pairs_runtime": 1.2998,
+      "eval_msmarco_pairs_samples_per_second": 98.476,
+      "eval_msmarco_pairs_steps_per_second": 1.539,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_nq_pairs_loss": 2.938187599182129,
+      "eval_nq_pairs_runtime": 2.5746,
+      "eval_nq_pairs_samples_per_second": 49.716,
+      "eval_nq_pairs_steps_per_second": 0.777,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_trivia_pairs_loss": 2.9295105934143066,
+      "eval_trivia_pairs_runtime": 4.3725,
+      "eval_trivia_pairs_samples_per_second": 29.274,
+      "eval_trivia_pairs_steps_per_second": 0.457,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_gooaq_pairs_loss": 5.916049003601074,
+      "eval_gooaq_pairs_runtime": 0.9073,
+      "eval_gooaq_pairs_samples_per_second": 141.085,
+      "eval_gooaq_pairs_steps_per_second": 2.204,
+      "step": 14
+    },
+    {
+      "epoch": 0.4069767441860465,
+      "eval_paws-pos_loss": 2.6019551753997803,
+      "eval_paws-pos_runtime": 0.6822,
+      "eval_paws-pos_samples_per_second": 187.639,
+      "eval_paws-pos_steps_per_second": 2.932,
+      "step": 14
+    },
+    {
+      "epoch": 0.436046511627907,
+      "grad_norm": 3.9955942630767822,
+      "learning_rate": 3.864447993031778e-05,
+      "loss": 6.2096,
+      "step": 15
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "grad_norm": 4.366024494171143,
+      "learning_rate": 3.7618802992467724e-05,
+      "loss": 6.0716,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.09375,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.5546875,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9691735506057739,
+      "eval_VitaminC_cosine_ap": 0.5349126870584762,
+      "eval_VitaminC_cosine_f1": 0.6472148541114058,
+      "eval_VitaminC_cosine_f1_threshold": 0.8376737236976624,
+      "eval_VitaminC_cosine_precision": 0.48031496062992124,
+      "eval_VitaminC_cosine_recall": 0.991869918699187,
+      "eval_VitaminC_dot_accuracy": 0.55078125,
+      "eval_VitaminC_dot_accuracy_threshold": 498.48065185546875,
+      "eval_VitaminC_dot_ap": 0.5211586848370118,
+      "eval_VitaminC_dot_f1": 0.6524064171122995,
+      "eval_VitaminC_dot_f1_threshold": 403.0154113769531,
+      "eval_VitaminC_dot_precision": 0.4860557768924303,
+      "eval_VitaminC_dot_recall": 0.991869918699187,
+      "eval_VitaminC_euclidean_accuracy": 0.55078125,
+      "eval_VitaminC_euclidean_accuracy_threshold": 5.804729461669922,
+      "eval_VitaminC_euclidean_ap": 0.5339291578094064,
+      "eval_VitaminC_euclidean_f1": 0.648936170212766,
+      "eval_VitaminC_euclidean_f1_threshold": 12.604965209960938,
+      "eval_VitaminC_euclidean_precision": 0.48221343873517786,
+      "eval_VitaminC_euclidean_recall": 0.991869918699187,
+      "eval_VitaminC_manhattan_accuracy": 0.5390625,
+      "eval_VitaminC_manhattan_accuracy_threshold": 100.77384948730469,
+      "eval_VitaminC_manhattan_ap": 0.5227366051874334,
+      "eval_VitaminC_manhattan_f1": 0.6522911051212938,
+      "eval_VitaminC_manhattan_f1_threshold": 195.17767333984375,
+      "eval_VitaminC_manhattan_precision": 0.4879032258064516,
+      "eval_VitaminC_manhattan_recall": 0.983739837398374,
+      "eval_VitaminC_max_accuracy": 0.5546875,
+      "eval_VitaminC_max_accuracy_threshold": 498.48065185546875,
+      "eval_VitaminC_max_ap": 0.5349126870584762,
+      "eval_VitaminC_max_f1": 0.6524064171122995,
+      "eval_VitaminC_max_f1_threshold": 403.0154113769531,
+      "eval_VitaminC_max_precision": 0.4879032258064516,
+      "eval_VitaminC_max_recall": 0.991869918699187,
+      "eval_sequential_score": 0.5349126870584762,
+      "eval_sts-test_pearson_cosine": 0.14325660008800628,
+      "eval_sts-test_pearson_dot": 0.2835408583342159,
+      "eval_sts-test_pearson_euclidean": 0.1460831550527121,
+      "eval_sts-test_pearson_manhattan": 0.1637804658889356,
+      "eval_sts-test_pearson_max": 0.2835408583342159,
+      "eval_sts-test_spearman_cosine": 0.19517789263204333,
+      "eval_sts-test_spearman_dot": 0.2999712694790054,
+      "eval_sts-test_spearman_euclidean": 0.1699467212439841,
+      "eval_sts-test_spearman_manhattan": 0.18602551069074194,
+      "eval_sts-test_spearman_max": 0.2999712694790054,
+      "eval_vitaminc-pairs_loss": 2.6983861923217773,
+      "eval_vitaminc-pairs_runtime": 1.4329,
+      "eval_vitaminc-pairs_samples_per_second": 75.372,
+      "eval_vitaminc-pairs_steps_per_second": 1.396,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_negation-triplets_loss": 5.085224628448486,
+      "eval_negation-triplets_runtime": 0.2968,
+      "eval_negation-triplets_samples_per_second": 215.65,
+      "eval_negation-triplets_steps_per_second": 3.37,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_scitail-pairs-pos_loss": 1.9003973007202148,
+      "eval_scitail-pairs-pos_runtime": 0.3648,
+      "eval_scitail-pairs-pos_samples_per_second": 148.014,
+      "eval_scitail-pairs-pos_steps_per_second": 2.741,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_xsum-pairs_loss": 5.669831275939941,
+      "eval_xsum-pairs_runtime": 3.42,
+      "eval_xsum-pairs_samples_per_second": 37.427,
+      "eval_xsum-pairs_steps_per_second": 0.585,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_sciq_pairs_loss": 0.29814526438713074,
+      "eval_sciq_pairs_runtime": 3.4152,
+      "eval_sciq_pairs_samples_per_second": 37.48,
+      "eval_sciq_pairs_steps_per_second": 0.586,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_qasc_pairs_loss": 2.7809503078460693,
+      "eval_qasc_pairs_runtime": 0.6393,
+      "eval_qasc_pairs_samples_per_second": 200.214,
+      "eval_qasc_pairs_steps_per_second": 3.128,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_openbookqa_pairs_loss": 4.030131816864014,
+      "eval_openbookqa_pairs_runtime": 0.5753,
+      "eval_openbookqa_pairs_samples_per_second": 222.483,
+      "eval_openbookqa_pairs_steps_per_second": 3.476,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_msmarco_pairs_loss": 5.903888702392578,
+      "eval_msmarco_pairs_runtime": 1.2958,
+      "eval_msmarco_pairs_samples_per_second": 98.78,
+      "eval_msmarco_pairs_steps_per_second": 1.543,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_nq_pairs_loss": 2.8589022159576416,
+      "eval_nq_pairs_runtime": 2.5698,
+      "eval_nq_pairs_samples_per_second": 49.809,
+      "eval_nq_pairs_steps_per_second": 0.778,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_trivia_pairs_loss": 2.8898680210113525,
+      "eval_trivia_pairs_runtime": 4.362,
+      "eval_trivia_pairs_samples_per_second": 29.344,
+      "eval_trivia_pairs_steps_per_second": 0.459,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_gooaq_pairs_loss": 5.800903797149658,
+      "eval_gooaq_pairs_runtime": 0.9084,
+      "eval_gooaq_pairs_samples_per_second": 140.913,
+      "eval_gooaq_pairs_steps_per_second": 2.202,
+      "step": 16
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "eval_paws-pos_loss": 2.4279778003692627,
+      "eval_paws-pos_runtime": 0.6819,
+      "eval_paws-pos_samples_per_second": 187.706,
+      "eval_paws-pos_steps_per_second": 2.933,
+      "step": 16
+    },
+    {
+      "epoch": 0.4941860465116279,
+      "grad_norm": 5.666141033172607,
+      "learning_rate": 3.633624361531388e-05,
+      "loss": 5.9185,
+      "step": 17
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "grad_norm": 7.217273235321045,
+      "learning_rate": 3.482291100917928e-05,
+      "loss": 5.7074,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.078125,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.55078125,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9663318395614624,
+      "eval_VitaminC_cosine_ap": 0.5331049287016129,
+      "eval_VitaminC_cosine_f1": 0.648936170212766,
+      "eval_VitaminC_cosine_f1_threshold": 0.8216172456741333,
+      "eval_VitaminC_cosine_precision": 0.48221343873517786,
+      "eval_VitaminC_cosine_recall": 0.991869918699187,
+      "eval_VitaminC_dot_accuracy": 0.5546875,
+      "eval_VitaminC_dot_accuracy_threshold": 468.02716064453125,
+      "eval_VitaminC_dot_ap": 0.5202948869489447,
+      "eval_VitaminC_dot_f1": 0.6507936507936508,
+      "eval_VitaminC_dot_f1_threshold": 358.44915771484375,
+      "eval_VitaminC_dot_precision": 0.4823529411764706,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.546875,
+      "eval_VitaminC_euclidean_accuracy_threshold": 5.734438896179199,
+      "eval_VitaminC_euclidean_ap": 0.5282344411865632,
+      "eval_VitaminC_euclidean_f1": 0.6507936507936508,
+      "eval_VitaminC_euclidean_f1_threshold": 13.58138656616211,
+      "eval_VitaminC_euclidean_precision": 0.4823529411764706,
+      "eval_VitaminC_euclidean_recall": 1.0,
+      "eval_VitaminC_manhattan_accuracy": 0.54296875,
+      "eval_VitaminC_manhattan_accuracy_threshold": 112.2923355102539,
+      "eval_VitaminC_manhattan_ap": 0.5237821619629712,
+      "eval_VitaminC_manhattan_f1": 0.6540540540540541,
+      "eval_VitaminC_manhattan_f1_threshold": 210.52694702148438,
+      "eval_VitaminC_manhattan_precision": 0.4898785425101215,
+      "eval_VitaminC_manhattan_recall": 0.983739837398374,
+      "eval_VitaminC_max_accuracy": 0.5546875,
+      "eval_VitaminC_max_accuracy_threshold": 468.02716064453125,
+      "eval_VitaminC_max_ap": 0.5331049287016129,
+      "eval_VitaminC_max_f1": 0.6540540540540541,
+      "eval_VitaminC_max_f1_threshold": 358.44915771484375,
+      "eval_VitaminC_max_precision": 0.4898785425101215,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5331049287016129,
+      "eval_sts-test_pearson_cosine": 0.1771905788413257,
+      "eval_sts-test_pearson_dot": 0.27769356372160947,
+      "eval_sts-test_pearson_euclidean": 0.17465663546843996,
+      "eval_sts-test_pearson_manhattan": 0.18333784202455092,
+      "eval_sts-test_pearson_max": 0.27769356372160947,
+      "eval_sts-test_spearman_cosine": 0.2225047095682771,
+      "eval_sts-test_spearman_dot": 0.30350268106992373,
+      "eval_sts-test_spearman_euclidean": 0.19950882182232507,
+      "eval_sts-test_spearman_manhattan": 0.20700380700256382,
+      "eval_sts-test_spearman_max": 0.30350268106992373,
+      "eval_vitaminc-pairs_loss": 2.6848371028900146,
+      "eval_vitaminc-pairs_runtime": 1.4384,
+      "eval_vitaminc-pairs_samples_per_second": 75.083,
+      "eval_vitaminc-pairs_steps_per_second": 1.39,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_negation-triplets_loss": 4.926700592041016,
+      "eval_negation-triplets_runtime": 0.3007,
+      "eval_negation-triplets_samples_per_second": 212.811,
+      "eval_negation-triplets_steps_per_second": 3.325,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_scitail-pairs-pos_loss": 1.5993835926055908,
+      "eval_scitail-pairs-pos_runtime": 0.3639,
+      "eval_scitail-pairs-pos_samples_per_second": 148.377,
+      "eval_scitail-pairs-pos_steps_per_second": 2.748,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_xsum-pairs_loss": 5.547531604766846,
+      "eval_xsum-pairs_runtime": 3.4245,
+      "eval_xsum-pairs_samples_per_second": 37.378,
+      "eval_xsum-pairs_steps_per_second": 0.584,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_sciq_pairs_loss": 0.284468412399292,
+      "eval_sciq_pairs_runtime": 3.4609,
+      "eval_sciq_pairs_samples_per_second": 36.985,
+      "eval_sciq_pairs_steps_per_second": 0.578,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_qasc_pairs_loss": 2.660043478012085,
+      "eval_qasc_pairs_runtime": 0.6389,
+      "eval_qasc_pairs_samples_per_second": 200.332,
+      "eval_qasc_pairs_steps_per_second": 3.13,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_openbookqa_pairs_loss": 3.8950119018554688,
+      "eval_openbookqa_pairs_runtime": 0.5754,
+      "eval_openbookqa_pairs_samples_per_second": 222.436,
+      "eval_openbookqa_pairs_steps_per_second": 3.476,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_msmarco_pairs_loss": 5.680139541625977,
+      "eval_msmarco_pairs_runtime": 1.2956,
+      "eval_msmarco_pairs_samples_per_second": 98.794,
+      "eval_msmarco_pairs_steps_per_second": 1.544,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_nq_pairs_loss": 2.7670226097106934,
+      "eval_nq_pairs_runtime": 2.5834,
+      "eval_nq_pairs_samples_per_second": 49.547,
+      "eval_nq_pairs_steps_per_second": 0.774,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_trivia_pairs_loss": 2.8672029972076416,
+      "eval_trivia_pairs_runtime": 4.3936,
+      "eval_trivia_pairs_samples_per_second": 29.134,
+      "eval_trivia_pairs_steps_per_second": 0.455,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_gooaq_pairs_loss": 5.63346529006958,
+      "eval_gooaq_pairs_runtime": 0.9135,
+      "eval_gooaq_pairs_samples_per_second": 140.115,
+      "eval_gooaq_pairs_steps_per_second": 2.189,
+      "step": 18
+    },
+    {
+      "epoch": 0.5232558139534884,
+      "eval_paws-pos_loss": 2.020176887512207,
+      "eval_paws-pos_runtime": 0.7044,
+      "eval_paws-pos_samples_per_second": 181.707,
+      "eval_paws-pos_steps_per_second": 2.839,
+      "step": 18
+    },
+    {
+      "epoch": 0.5523255813953488,
+      "grad_norm": 6.092338562011719,
+      "learning_rate": 3.310961226183396e-05,
+      "loss": 5.7699,
+      "step": 19
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "grad_norm": 11.575896263122559,
+      "learning_rate": 3.12312251950283e-05,
+      "loss": 4.982,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.0625,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.54296875,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.9424006938934326,
+      "eval_VitaminC_cosine_ap": 0.5256196316622103,
+      "eval_VitaminC_cosine_f1": 0.6525198938992042,
+      "eval_VitaminC_cosine_f1_threshold": 0.7586345076560974,
+      "eval_VitaminC_cosine_precision": 0.484251968503937,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.5546875,
+      "eval_VitaminC_dot_accuracy_threshold": 454.70672607421875,
+      "eval_VitaminC_dot_ap": 0.5181548712256367,
+      "eval_VitaminC_dot_f1": 0.648936170212766,
+      "eval_VitaminC_dot_f1_threshold": 305.1645812988281,
+      "eval_VitaminC_dot_precision": 0.48221343873517786,
+      "eval_VitaminC_dot_recall": 0.991869918699187,
+      "eval_VitaminC_euclidean_accuracy": 0.5390625,
+      "eval_VitaminC_euclidean_accuracy_threshold": 7.773402214050293,
+      "eval_VitaminC_euclidean_ap": 0.5234752447708622,
+      "eval_VitaminC_euclidean_f1": 0.6525198938992042,
+      "eval_VitaminC_euclidean_f1_threshold": 14.282221794128418,
+      "eval_VitaminC_euclidean_precision": 0.484251968503937,
+      "eval_VitaminC_euclidean_recall": 1.0,
+      "eval_VitaminC_manhattan_accuracy": 0.5390625,
+      "eval_VitaminC_manhattan_accuracy_threshold": 139.0895538330078,
+      "eval_VitaminC_manhattan_ap": 0.5206276713713462,
+      "eval_VitaminC_manhattan_f1": 0.6576086956521738,
+      "eval_VitaminC_manhattan_f1_threshold": 228.93572998046875,
+      "eval_VitaminC_manhattan_precision": 0.49387755102040815,
+      "eval_VitaminC_manhattan_recall": 0.983739837398374,
+      "eval_VitaminC_max_accuracy": 0.5546875,
+      "eval_VitaminC_max_accuracy_threshold": 454.70672607421875,
+      "eval_VitaminC_max_ap": 0.5256196316622103,
+      "eval_VitaminC_max_f1": 0.6576086956521738,
+      "eval_VitaminC_max_f1_threshold": 305.1645812988281,
+      "eval_VitaminC_max_precision": 0.49387755102040815,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5256196316622103,
+      "eval_sts-test_pearson_cosine": 0.19452438299121508,
+      "eval_sts-test_pearson_dot": 0.2630327257559114,
+      "eval_sts-test_pearson_euclidean": 0.1906533741460466,
+      "eval_sts-test_pearson_manhattan": 0.18925349181409473,
+      "eval_sts-test_pearson_max": 0.2630327257559114,
+      "eval_sts-test_spearman_cosine": 0.23210464950295256,
+      "eval_sts-test_spearman_dot": 0.2940857028429396,
+      "eval_sts-test_spearman_euclidean": 0.20718891698862332,
+      "eval_sts-test_spearman_manhattan": 0.209132034772656,
+      "eval_sts-test_spearman_max": 0.2940857028429396,
+      "eval_vitaminc-pairs_loss": 2.680999755859375,
+      "eval_vitaminc-pairs_runtime": 1.4364,
+      "eval_vitaminc-pairs_samples_per_second": 75.185,
+      "eval_vitaminc-pairs_steps_per_second": 1.392,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_negation-triplets_loss": 4.7232513427734375,
+      "eval_negation-triplets_runtime": 0.2954,
+      "eval_negation-triplets_samples_per_second": 216.661,
+      "eval_negation-triplets_steps_per_second": 3.385,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_scitail-pairs-pos_loss": 1.1812373399734497,
+      "eval_scitail-pairs-pos_runtime": 0.3904,
+      "eval_scitail-pairs-pos_samples_per_second": 138.319,
+      "eval_scitail-pairs-pos_steps_per_second": 2.561,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_xsum-pairs_loss": 5.3592987060546875,
+      "eval_xsum-pairs_runtime": 3.4181,
+      "eval_xsum-pairs_samples_per_second": 37.447,
+      "eval_xsum-pairs_steps_per_second": 0.585,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_sciq_pairs_loss": 0.2662390470504761,
+      "eval_sciq_pairs_runtime": 3.4207,
+      "eval_sciq_pairs_samples_per_second": 37.419,
+      "eval_sciq_pairs_steps_per_second": 0.585,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_qasc_pairs_loss": 2.4869425296783447,
+      "eval_qasc_pairs_runtime": 0.6388,
+      "eval_qasc_pairs_samples_per_second": 200.382,
+      "eval_qasc_pairs_steps_per_second": 3.131,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_openbookqa_pairs_loss": 3.7582223415374756,
+      "eval_openbookqa_pairs_runtime": 0.5726,
+      "eval_openbookqa_pairs_samples_per_second": 223.534,
+      "eval_openbookqa_pairs_steps_per_second": 3.493,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_msmarco_pairs_loss": 5.391519546508789,
+      "eval_msmarco_pairs_runtime": 1.2972,
+      "eval_msmarco_pairs_samples_per_second": 98.675,
+      "eval_msmarco_pairs_steps_per_second": 1.542,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_nq_pairs_loss": 2.6773831844329834,
+      "eval_nq_pairs_runtime": 2.5735,
+      "eval_nq_pairs_samples_per_second": 49.738,
+      "eval_nq_pairs_steps_per_second": 0.777,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_trivia_pairs_loss": 2.9395601749420166,
+      "eval_trivia_pairs_runtime": 4.3602,
+      "eval_trivia_pairs_samples_per_second": 29.356,
+      "eval_trivia_pairs_steps_per_second": 0.459,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_gooaq_pairs_loss": 5.372612476348877,
+      "eval_gooaq_pairs_runtime": 0.9156,
+      "eval_gooaq_pairs_samples_per_second": 139.801,
+      "eval_gooaq_pairs_steps_per_second": 2.184,
+      "step": 20
+    },
+    {
+      "epoch": 0.5813953488372093,
+      "eval_paws-pos_loss": 1.354379653930664,
+      "eval_paws-pos_runtime": 0.6816,
+      "eval_paws-pos_samples_per_second": 187.78,
+      "eval_paws-pos_steps_per_second": 2.934,
+      "step": 20
+    },
+    {
+      "epoch": 0.6104651162790697,
+      "grad_norm": 14.814264297485352,
+      "learning_rate": 2.9225988352621447e-05,
+      "loss": 4.4632,
+      "step": 21
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "grad_norm": 9.865734100341797,
+      "learning_rate": 2.713472257409928e-05,
+      "loss": 5.8961,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_NLI-v2_cosine_accuracy": 1.0,
+      "eval_NLI-v2_dot_accuracy": 0.03125,
+      "eval_NLI-v2_euclidean_accuracy": 1.0,
+      "eval_NLI-v2_manhattan_accuracy": 1.0,
+      "eval_NLI-v2_max_accuracy": 1.0,
+      "eval_VitaminC_cosine_accuracy": 0.55078125,
+      "eval_VitaminC_cosine_accuracy_threshold": 0.8842157125473022,
+      "eval_VitaminC_cosine_ap": 0.5210602959566645,
+      "eval_VitaminC_cosine_f1": 0.6542553191489362,
+      "eval_VitaminC_cosine_f1_threshold": 0.6854724884033203,
+      "eval_VitaminC_cosine_precision": 0.48616600790513836,
+      "eval_VitaminC_cosine_recall": 1.0,
+      "eval_VitaminC_dot_accuracy": 0.5390625,
+      "eval_VitaminC_dot_accuracy_threshold": 443.89678955078125,
+      "eval_VitaminC_dot_ap": 0.5051685354280885,
+      "eval_VitaminC_dot_f1": 0.6507936507936508,
+      "eval_VitaminC_dot_f1_threshold": 240.06381225585938,
+      "eval_VitaminC_dot_precision": 0.4823529411764706,
+      "eval_VitaminC_dot_recall": 1.0,
+      "eval_VitaminC_euclidean_accuracy": 0.5546875,
+      "eval_VitaminC_euclidean_accuracy_threshold": 10.32968521118164,
+      "eval_VitaminC_euclidean_ap": 0.5269597637184864,
+      "eval_VitaminC_euclidean_f1": 0.6559999999999999,
+      "eval_VitaminC_euclidean_f1_threshold": 15.269627571105957,
+      "eval_VitaminC_euclidean_precision": 0.4880952380952381,
+      "eval_VitaminC_euclidean_recall": 1.0,
+      "eval_VitaminC_manhattan_accuracy": 0.55078125,
+      "eval_VitaminC_manhattan_accuracy_threshold": 182.53717041015625,
+      "eval_VitaminC_manhattan_ap": 0.5192375962842349,
+      "eval_VitaminC_manhattan_f1": 0.6541554959785523,
+      "eval_VitaminC_manhattan_f1_threshold": 270.3173828125,
+      "eval_VitaminC_manhattan_precision": 0.488,
+      "eval_VitaminC_manhattan_recall": 0.991869918699187,
+      "eval_VitaminC_max_accuracy": 0.5546875,
+      "eval_VitaminC_max_accuracy_threshold": 443.89678955078125,
+      "eval_VitaminC_max_ap": 0.5269597637184864,
+      "eval_VitaminC_max_f1": 0.6559999999999999,
+      "eval_VitaminC_max_f1_threshold": 270.3173828125,
+      "eval_VitaminC_max_precision": 0.4880952380952381,
+      "eval_VitaminC_max_recall": 1.0,
+      "eval_sequential_score": 0.5269597637184864,
+      "eval_sts-test_pearson_cosine": 0.19359932329736743,
+      "eval_sts-test_pearson_dot": 0.24827045025174932,
+      "eval_sts-test_pearson_euclidean": 0.19635314257478567,
+      "eval_sts-test_pearson_manhattan": 0.19317635420262674,
+      "eval_sts-test_pearson_max": 0.24827045025174932,
+      "eval_sts-test_spearman_cosine": 0.23349279537520143,
+      "eval_sts-test_spearman_dot": 0.2861062808906728,
+      "eval_sts-test_spearman_euclidean": 0.2124858553518036,
+      "eval_sts-test_spearman_manhattan": 0.21175257633545957,
+      "eval_sts-test_spearman_max": 0.2861062808906728,
+      "eval_vitaminc-pairs_loss": 2.6802568435668945,
+      "eval_vitaminc-pairs_runtime": 1.4504,
+      "eval_vitaminc-pairs_samples_per_second": 74.465,
+      "eval_vitaminc-pairs_steps_per_second": 1.379,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_negation-triplets_loss": 4.573638916015625,
+      "eval_negation-triplets_runtime": 0.2964,
+      "eval_negation-triplets_samples_per_second": 215.93,
+      "eval_negation-triplets_steps_per_second": 3.374,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_scitail-pairs-pos_loss": 0.8745603561401367,
+      "eval_scitail-pairs-pos_runtime": 0.3676,
+      "eval_scitail-pairs-pos_samples_per_second": 146.892,
+      "eval_scitail-pairs-pos_steps_per_second": 2.72,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_xsum-pairs_loss": 5.152139663696289,
+      "eval_xsum-pairs_runtime": 3.4135,
+      "eval_xsum-pairs_samples_per_second": 37.498,
+      "eval_xsum-pairs_steps_per_second": 0.586,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_sciq_pairs_loss": 0.25412440299987793,
+      "eval_sciq_pairs_runtime": 3.4101,
+      "eval_sciq_pairs_samples_per_second": 37.536,
+      "eval_sciq_pairs_steps_per_second": 0.586,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_qasc_pairs_loss": 2.3517777919769287,
+      "eval_qasc_pairs_runtime": 0.6408,
+      "eval_qasc_pairs_samples_per_second": 199.764,
+      "eval_qasc_pairs_steps_per_second": 3.121,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_openbookqa_pairs_loss": 3.83288311958313,
+      "eval_openbookqa_pairs_runtime": 0.5773,
+      "eval_openbookqa_pairs_samples_per_second": 221.737,
+      "eval_openbookqa_pairs_steps_per_second": 3.465,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_msmarco_pairs_loss": 5.144412994384766,
+      "eval_msmarco_pairs_runtime": 1.2921,
+      "eval_msmarco_pairs_samples_per_second": 99.065,
+      "eval_msmarco_pairs_steps_per_second": 1.548,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_nq_pairs_loss": 2.5934412479400635,
+      "eval_nq_pairs_runtime": 2.5707,
+      "eval_nq_pairs_samples_per_second": 49.791,
+      "eval_nq_pairs_steps_per_second": 0.778,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_trivia_pairs_loss": 3.1416494846343994,
+      "eval_trivia_pairs_runtime": 4.3639,
+      "eval_trivia_pairs_samples_per_second": 29.332,
+      "eval_trivia_pairs_steps_per_second": 0.458,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_gooaq_pairs_loss": 5.084894180297852,
+      "eval_gooaq_pairs_runtime": 0.9118,
+      "eval_gooaq_pairs_samples_per_second": 140.378,
+      "eval_gooaq_pairs_steps_per_second": 2.193,
+      "step": 22
+    },
+    {
+      "epoch": 0.6395348837209303,
+      "eval_paws-pos_loss": 0.8249754309654236,
+      "eval_paws-pos_runtime": 0.6994,
+      "eval_paws-pos_samples_per_second": 183.015,
+      "eval_paws-pos_steps_per_second": 2.86,
+      "step": 22
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 34,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2,
+  "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": 224,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-22/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edceb7011ecd574f3e3fa5380c854a14e3b5904426b5550bdb62acd6cda68870
+size 5688