update the ones with request

.gitattributes

@@ -53,3 +53,43 @@ modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco/chec
 modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
 modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
 modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.msmarco.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/0c/e8/0ce892ac499c726f9f0ffc58f21eaed98059a99787e8a5e0b47b09519dd56c03 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/b5/a4/b5a4e3e60f62a3f1aa05b880c77e0b56f807efafa62f60a5f452019530af6711 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/2e/d8/2ed820e6dfb3db7d393c18aa5f5eb9c5baf64f9c9c906ace62bac5ba200d4cf2 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/0d/7d/0d7de69bdc8b2c1365cd9379c80c2e9660583d3b9d21c5da9f70bc756aef8c5c filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/c7/ca/c7ca94df487520ae2c411613298582fccb5e50edc91edd1e3311d560a4ff7427 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/40/49/4049b989fa1e41ca8b3f1d3acf616f06129dd1998d0d24bb71a4e9077014975b filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/31/5e/315edc508eedf0940ded0c431d5b741a533b41fd7759fa8996d6f2a53fa7a1f5 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/98/16/9816537f9d6d804d6866451e1583a3857ffd0af4e789dd38f7361227c01b1b62 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/eb/8c/eb8c18aefb03ba8afaf3ddafca1cd6ea74416c497632a03907346835d107adad filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/44/44/4444b588f984289c2834e5eed337f71344ab7374e6ca84c4ebf27f527fdfb31e filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/f8/78/f8788fb899610dac5c81332d3ad006033814e54fdc7526f95926d4a8228a5000 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/1e/e5/1ee565563332516e4024443fa4e532456a9f9be6bcbe4be7011aa8a9b89ed496 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/a5/4b/a54bb859b7620f51ba6e2f7feb41f03eed4f1e229e4d5a26ba32b89f7cfda494 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/8b/09/8b0972cf18a384a6114761e112ac68fee94239f09830af05f762267cbfd9ae4d filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/68/4d/684d731b0a9245de462cf0af539b773d34256b419759d530a7b160bea702561c filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/1b/a4/1ba44b14f4a98062eaecb16a59796513bafcff78abd007a3fba0d5e20e889ea8 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/0b/21/0b21d57153ab71ab9dadd49944db15e74447f04e9b0cb67c61655d0106e4a0d6 filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/6d/41/6d41fde859da34b9501b484af69475f50a37de92a28b6fa47a87c5caff3703ae filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/82/36/823620ff1dd0a6e380f4c8e0ec885b6b1aa2023dde15f5e41b13c2b6e43f80ab filter=lfs diff=lfs merge=lfs -text
+.git/lfs/objects/1f/8c/1f8cea23e59027c2c5b36168d940ec4d3844b1f29c814cb8d6202d62fa45f9c0 filter=lfs diff=lfs merge=lfs -text

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/config.json

@@ -0,0 +1,45 @@
+{
+  "architectures": [
+    "ModernBertModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "repad_logits_with_grad": false,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.53.0",
+  "vocab_size": 50368
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a268b8d3b223be4857defa18f90c46cf839f603e9f616465a0472f05acc3215a
+size 298041696

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5831f1d98a7fb576cf915ca7b4410402babd2e88a871c415aeb759500f4add58
+size 596170443

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3044a568f2c48f55afab6e63cf0c046d3586784e77e39b1af825abecfb2cdba8
+size 15365

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86accf27064cdd503053e90476a6bd10de333d4ff0594535ad55ea13a473c91d
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18ca8d714ef40be035404c1957b5a4dee84e1f43980408393f8aa710552ee6f6
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cfdebe99e40accc9c9d8f09c63136a14abda997d9b501969ec8e16e9d183179
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:053b181ac2db5ceae48310494587c492212456802136bc3852c62ccc7b9ff38c
+size 1465

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/trainer_state.json

@@ -0,0 +1,799 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7849293563579278,
+  "eval_steps": 1000.0,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval/nano_beir.arguana": 0.4400549254974839,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.22880053057901228,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.30026716538531234,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6419178788751609,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4370593595651809,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6723047477224562,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3437446614516137,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.436711543163819,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.8439997970269911,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.26304331393922176,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7599341408912704,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.29696490432942557,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.47206691403557904,
+      "step": 0
+    },
+    {
+      "epoch": 0.007849293563579277,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8e-06,
+      "loss": 0.7314,
+      "step": 10
+    },
+    {
+      "epoch": 0.015698587127158554,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.8e-06,
+      "loss": 0.8796,
+      "step": 20
+    },
+    {
+      "epoch": 0.023547880690737835,
+      "grad_norm": 7.3125,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7311,
+      "step": 30
+    },
+    {
+      "epoch": 0.03139717425431711,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.8e-06,
+      "loss": 0.8214,
+      "step": 40
+    },
+    {
+      "epoch": 0.03924646781789639,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.8489,
+      "step": 50
+    },
+    {
+      "epoch": 0.04709576138147567,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.7661,
+      "step": 60
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.7525,
+      "step": 70
+    },
+    {
+      "epoch": 0.06279434850863422,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.8653,
+      "step": 80
+    },
+    {
+      "epoch": 0.0706436420722135,
+      "grad_norm": 6.25,
+      "learning_rate": 1.78e-05,
+      "loss": 0.7607,
+      "step": 90
+    },
+    {
+      "epoch": 0.07849293563579278,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.7259,
+      "step": 100
+    },
+    {
+      "epoch": 0.08634222919937205,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.18e-05,
+      "loss": 0.7862,
+      "step": 110
+    },
+    {
+      "epoch": 0.09419152276295134,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.38e-05,
+      "loss": 0.8212,
+      "step": 120
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 6.0,
+      "learning_rate": 2.58e-05,
+      "loss": 0.6884,
+      "step": 130
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.712,
+      "step": 140
+    },
+    {
+      "epoch": 0.11773940345368916,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.98e-05,
+      "loss": 0.6231,
+      "step": 150
+    },
+    {
+      "epoch": 0.12558869701726844,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.18e-05,
+      "loss": 0.6733,
+      "step": 160
+    },
+    {
+      "epoch": 0.13343799058084774,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.38e-05,
+      "loss": 0.6973,
+      "step": 170
+    },
+    {
+      "epoch": 0.141287284144427,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.58e-05,
+      "loss": 0.5928,
+      "step": 180
+    },
+    {
+      "epoch": 0.14913657770800628,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.6189,
+      "step": 190
+    },
+    {
+      "epoch": 0.15698587127158556,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.6661,
+      "step": 200
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 4.875,
+      "learning_rate": 4.18e-05,
+      "loss": 0.634,
+      "step": 210
+    },
+    {
+      "epoch": 0.1726844583987441,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.38e-05,
+      "loss": 0.6963,
+      "step": 220
+    },
+    {
+      "epoch": 0.18053375196232338,
+      "grad_norm": 5.875,
+      "learning_rate": 4.58e-05,
+      "loss": 0.6744,
+      "step": 230
+    },
+    {
+      "epoch": 0.18838304552590268,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.78e-05,
+      "loss": 0.6635,
+      "step": 240
+    },
+    {
+      "epoch": 0.19623233908948196,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.6874,
+      "step": 250
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 5.28125,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.6516,
+      "step": 260
+    },
+    {
+      "epoch": 0.2119309262166405,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.7579,
+      "step": 270
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.6867,
+      "step": 280
+    },
+    {
+      "epoch": 0.22762951334379905,
+      "grad_norm": 6.59375,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.7038,
+      "step": 290
+    },
+    {
+      "epoch": 0.23547880690737832,
+      "grad_norm": 5.9375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.6681,
+      "step": 300
+    },
+    {
+      "epoch": 0.24332810047095763,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.18e-05,
+      "loss": 0.6133,
+      "step": 310
+    },
+    {
+      "epoch": 0.25117739403453687,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.38e-05,
+      "loss": 0.6203,
+      "step": 320
+    },
+    {
+      "epoch": 0.25902668759811615,
+      "grad_norm": 6.3125,
+      "learning_rate": 6.58e-05,
+      "loss": 0.7164,
+      "step": 330
+    },
+    {
+      "epoch": 0.2668759811616955,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.7157,
+      "step": 340
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.98e-05,
+      "loss": 0.7176,
+      "step": 350
+    },
+    {
+      "epoch": 0.282574568288854,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.18e-05,
+      "loss": 0.702,
+      "step": 360
+    },
+    {
+      "epoch": 0.2904238618524333,
+      "grad_norm": 5.25,
+      "learning_rate": 7.38e-05,
+      "loss": 0.5834,
+      "step": 370
+    },
+    {
+      "epoch": 0.29827315541601257,
+      "grad_norm": 5.46875,
+      "learning_rate": 7.58e-05,
+      "loss": 0.594,
+      "step": 380
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 6.625,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.7667,
+      "step": 390
+    },
+    {
+      "epoch": 0.3139717425431711,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.6477,
+      "step": 400
+    },
+    {
+      "epoch": 0.3218210361067504,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.18e-05,
+      "loss": 0.6104,
+      "step": 410
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.38e-05,
+      "loss": 0.6879,
+      "step": 420
+    },
+    {
+      "epoch": 0.33751962323390894,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.6514,
+      "step": 430
+    },
+    {
+      "epoch": 0.3453689167974882,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.78e-05,
+      "loss": 0.649,
+      "step": 440
+    },
+    {
+      "epoch": 0.3532182103610675,
+      "grad_norm": 4.75,
+      "learning_rate": 8.98e-05,
+      "loss": 0.6204,
+      "step": 450
+    },
+    {
+      "epoch": 0.36106750392464676,
+      "grad_norm": 5.125,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.6633,
+      "step": 460
+    },
+    {
+      "epoch": 0.36891679748822603,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.7058,
+      "step": 470
+    },
+    {
+      "epoch": 0.37676609105180536,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.58e-05,
+      "loss": 0.5653,
+      "step": 480
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.6143,
+      "step": 490
+    },
+    {
+      "epoch": 0.3924646781789639,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.98e-05,
+      "loss": 0.6218,
+      "step": 500
+    },
+    {
+      "epoch": 0.4003139717425432,
+      "grad_norm": 6.03125,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.6115,
+      "step": 510
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.6206,
+      "step": 520
+    },
+    {
+      "epoch": 0.41601255886970173,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.629,
+      "step": 530
+    },
+    {
+      "epoch": 0.423861852433281,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.6492,
+      "step": 540
+    },
+    {
+      "epoch": 0.4317111459968603,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.585,
+      "step": 550
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 7.1875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.558,
+      "step": 560
+    },
+    {
+      "epoch": 0.4474097331240188,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.6242,
+      "step": 570
+    },
+    {
+      "epoch": 0.4552590266875981,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.682,
+      "step": 580
+    },
+    {
+      "epoch": 0.4631083202511774,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.6194,
+      "step": 590
+    },
+    {
+      "epoch": 0.47095761381475665,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.6211,
+      "step": 600
+    },
+    {
+      "epoch": 0.478806907378336,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.617,
+      "step": 610
+    },
+    {
+      "epoch": 0.48665620094191525,
+      "grad_norm": 5.125,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.5901,
+      "step": 620
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.6067,
+      "step": 630
+    },
+    {
+      "epoch": 0.5023547880690737,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.5976,
+      "step": 640
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 5.75,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.6467,
+      "step": 650
+    },
+    {
+      "epoch": 0.5180533751962323,
+      "grad_norm": 7.25,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.5904,
+      "step": 660
+    },
+    {
+      "epoch": 0.5259026687598116,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.6814,
+      "step": 670
+    },
+    {
+      "epoch": 0.533751962323391,
+      "grad_norm": 6.125,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.6684,
+      "step": 680
+    },
+    {
+      "epoch": 0.5416012558869702,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.6206,
+      "step": 690
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 5.875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.6096,
+      "step": 700
+    },
+    {
+      "epoch": 0.5572998430141287,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.5468,
+      "step": 710
+    },
+    {
+      "epoch": 0.565149136577708,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.5883,
+      "step": 720
+    },
+    {
+      "epoch": 0.5729984301412873,
+      "grad_norm": 4.75,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.596,
+      "step": 730
+    },
+    {
+      "epoch": 0.5808477237048666,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.5178,
+      "step": 740
+    },
+    {
+      "epoch": 0.5886970172684458,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.6438,
+      "step": 750
+    },
+    {
+      "epoch": 0.5965463108320251,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.6975,
+      "step": 760
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 4.375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.5828,
+      "step": 770
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.5019,
+      "step": 780
+    },
+    {
+      "epoch": 0.6200941915227629,
+      "grad_norm": 6.75,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.6489,
+      "step": 790
+    },
+    {
+      "epoch": 0.6279434850863422,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.6364,
+      "step": 800
+    },
+    {
+      "epoch": 0.6357927786499215,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.5458,
+      "step": 810
+    },
+    {
+      "epoch": 0.6436420722135008,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.5859,
+      "step": 820
+    },
+    {
+      "epoch": 0.6514913657770801,
+      "grad_norm": 4.375,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.6369,
+      "step": 830
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.7126,
+      "step": 840
+    },
+    {
+      "epoch": 0.6671899529042387,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.5857,
+      "step": 850
+    },
+    {
+      "epoch": 0.6750392464678179,
+      "grad_norm": 5.0,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.645,
+      "step": 860
+    },
+    {
+      "epoch": 0.6828885400313972,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.6538,
+      "step": 870
+    },
+    {
+      "epoch": 0.6907378335949764,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.6045,
+      "step": 880
+    },
+    {
+      "epoch": 0.6985871271585558,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.6282,
+      "step": 890
+    },
+    {
+      "epoch": 0.706436420722135,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.5509,
+      "step": 900
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 7.125,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.6392,
+      "step": 910
+    },
+    {
+      "epoch": 0.7221350078492935,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.6152,
+      "step": 920
+    },
+    {
+      "epoch": 0.7299843014128728,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.7706,
+      "step": 930
+    },
+    {
+      "epoch": 0.7378335949764521,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.6656,
+      "step": 940
+    },
+    {
+      "epoch": 0.7456828885400314,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.6592,
+      "step": 950
+    },
+    {
+      "epoch": 0.7535321821036107,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.6141,
+      "step": 960
+    },
+    {
+      "epoch": 0.7613814756671899,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.6161,
+      "step": 970
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 4.25,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.7166,
+      "step": 980
+    },
+    {
+      "epoch": 0.7770800627943485,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.6028,
+      "step": 990
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.5262,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1000,
+  "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": 16,
+  "trial_name": null,
+  "trial_params": null
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-1000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9365177436ad70bb9d6000d96d3ccdc5c5ca38536defdcfec6b548401edbc07f
+size 6481

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/config.json

@@ -0,0 +1,45 @@
+{
+  "architectures": [
+    "ModernBertModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "repad_logits_with_grad": false,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.53.0",
+  "vocab_size": 50368
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf0a75f2dc3580598001eea2e77afba906bc5adc06f1a5696547fcb216fa27a6
+size 298041696

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f2241ba51f63d9e2dc7bf192d7e5d8fffa9c1d8416702001aa9bccb56377121
+size 596170443

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18bb6b6e7da6250d22c25b95c64287be4aea598d9d97ef67b0fd69b9be869ed7
+size 15365

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13db4dd10b22b5464f5ef0b539152b46390eae6a827b355108cb54b1af2eb830
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06e293f13fb6e873e8081636b6b07115410b397c55f1ed08b8772b059c8eb74e
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c595c077664f94b7f91088561951be20e39483e01865774a83bc46f3971a3152
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff65ca8b6e225d1d82cc63887ee464e8c896624115a4dc693eae1fff20815699
+size 1465

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/trainer_state.json

@@ -0,0 +1,1564 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.5698587127158556,
+  "eval_steps": 1000.0,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval/nano_beir.arguana": 0.4400549254974839,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.22880053057901228,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.30026716538531234,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6419178788751609,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4370593595651809,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6723047477224562,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3437446614516137,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.436711543163819,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.8439997970269911,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.26304331393922176,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7599341408912704,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.29696490432942557,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.47206691403557904,
+      "step": 0
+    },
+    {
+      "epoch": 0.007849293563579277,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8e-06,
+      "loss": 0.7314,
+      "step": 10
+    },
+    {
+      "epoch": 0.015698587127158554,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.8e-06,
+      "loss": 0.8796,
+      "step": 20
+    },
+    {
+      "epoch": 0.023547880690737835,
+      "grad_norm": 7.3125,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7311,
+      "step": 30
+    },
+    {
+      "epoch": 0.03139717425431711,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.8e-06,
+      "loss": 0.8214,
+      "step": 40
+    },
+    {
+      "epoch": 0.03924646781789639,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.8489,
+      "step": 50
+    },
+    {
+      "epoch": 0.04709576138147567,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.7661,
+      "step": 60
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.7525,
+      "step": 70
+    },
+    {
+      "epoch": 0.06279434850863422,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.8653,
+      "step": 80
+    },
+    {
+      "epoch": 0.0706436420722135,
+      "grad_norm": 6.25,
+      "learning_rate": 1.78e-05,
+      "loss": 0.7607,
+      "step": 90
+    },
+    {
+      "epoch": 0.07849293563579278,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.7259,
+      "step": 100
+    },
+    {
+      "epoch": 0.08634222919937205,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.18e-05,
+      "loss": 0.7862,
+      "step": 110
+    },
+    {
+      "epoch": 0.09419152276295134,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.38e-05,
+      "loss": 0.8212,
+      "step": 120
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 6.0,
+      "learning_rate": 2.58e-05,
+      "loss": 0.6884,
+      "step": 130
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.712,
+      "step": 140
+    },
+    {
+      "epoch": 0.11773940345368916,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.98e-05,
+      "loss": 0.6231,
+      "step": 150
+    },
+    {
+      "epoch": 0.12558869701726844,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.18e-05,
+      "loss": 0.6733,
+      "step": 160
+    },
+    {
+      "epoch": 0.13343799058084774,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.38e-05,
+      "loss": 0.6973,
+      "step": 170
+    },
+    {
+      "epoch": 0.141287284144427,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.58e-05,
+      "loss": 0.5928,
+      "step": 180
+    },
+    {
+      "epoch": 0.14913657770800628,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.6189,
+      "step": 190
+    },
+    {
+      "epoch": 0.15698587127158556,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.6661,
+      "step": 200
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 4.875,
+      "learning_rate": 4.18e-05,
+      "loss": 0.634,
+      "step": 210
+    },
+    {
+      "epoch": 0.1726844583987441,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.38e-05,
+      "loss": 0.6963,
+      "step": 220
+    },
+    {
+      "epoch": 0.18053375196232338,
+      "grad_norm": 5.875,
+      "learning_rate": 4.58e-05,
+      "loss": 0.6744,
+      "step": 230
+    },
+    {
+      "epoch": 0.18838304552590268,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.78e-05,
+      "loss": 0.6635,
+      "step": 240
+    },
+    {
+      "epoch": 0.19623233908948196,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.6874,
+      "step": 250
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 5.28125,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.6516,
+      "step": 260
+    },
+    {
+      "epoch": 0.2119309262166405,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.7579,
+      "step": 270
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.6867,
+      "step": 280
+    },
+    {
+      "epoch": 0.22762951334379905,
+      "grad_norm": 6.59375,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.7038,
+      "step": 290
+    },
+    {
+      "epoch": 0.23547880690737832,
+      "grad_norm": 5.9375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.6681,
+      "step": 300
+    },
+    {
+      "epoch": 0.24332810047095763,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.18e-05,
+      "loss": 0.6133,
+      "step": 310
+    },
+    {
+      "epoch": 0.25117739403453687,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.38e-05,
+      "loss": 0.6203,
+      "step": 320
+    },
+    {
+      "epoch": 0.25902668759811615,
+      "grad_norm": 6.3125,
+      "learning_rate": 6.58e-05,
+      "loss": 0.7164,
+      "step": 330
+    },
+    {
+      "epoch": 0.2668759811616955,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.7157,
+      "step": 340
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.98e-05,
+      "loss": 0.7176,
+      "step": 350
+    },
+    {
+      "epoch": 0.282574568288854,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.18e-05,
+      "loss": 0.702,
+      "step": 360
+    },
+    {
+      "epoch": 0.2904238618524333,
+      "grad_norm": 5.25,
+      "learning_rate": 7.38e-05,
+      "loss": 0.5834,
+      "step": 370
+    },
+    {
+      "epoch": 0.29827315541601257,
+      "grad_norm": 5.46875,
+      "learning_rate": 7.58e-05,
+      "loss": 0.594,
+      "step": 380
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 6.625,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.7667,
+      "step": 390
+    },
+    {
+      "epoch": 0.3139717425431711,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.6477,
+      "step": 400
+    },
+    {
+      "epoch": 0.3218210361067504,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.18e-05,
+      "loss": 0.6104,
+      "step": 410
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.38e-05,
+      "loss": 0.6879,
+      "step": 420
+    },
+    {
+      "epoch": 0.33751962323390894,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.6514,
+      "step": 430
+    },
+    {
+      "epoch": 0.3453689167974882,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.78e-05,
+      "loss": 0.649,
+      "step": 440
+    },
+    {
+      "epoch": 0.3532182103610675,
+      "grad_norm": 4.75,
+      "learning_rate": 8.98e-05,
+      "loss": 0.6204,
+      "step": 450
+    },
+    {
+      "epoch": 0.36106750392464676,
+      "grad_norm": 5.125,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.6633,
+      "step": 460
+    },
+    {
+      "epoch": 0.36891679748822603,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.7058,
+      "step": 470
+    },
+    {
+      "epoch": 0.37676609105180536,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.58e-05,
+      "loss": 0.5653,
+      "step": 480
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.6143,
+      "step": 490
+    },
+    {
+      "epoch": 0.3924646781789639,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.98e-05,
+      "loss": 0.6218,
+      "step": 500
+    },
+    {
+      "epoch": 0.4003139717425432,
+      "grad_norm": 6.03125,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.6115,
+      "step": 510
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.6206,
+      "step": 520
+    },
+    {
+      "epoch": 0.41601255886970173,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.629,
+      "step": 530
+    },
+    {
+      "epoch": 0.423861852433281,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.6492,
+      "step": 540
+    },
+    {
+      "epoch": 0.4317111459968603,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.585,
+      "step": 550
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 7.1875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.558,
+      "step": 560
+    },
+    {
+      "epoch": 0.4474097331240188,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.6242,
+      "step": 570
+    },
+    {
+      "epoch": 0.4552590266875981,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.682,
+      "step": 580
+    },
+    {
+      "epoch": 0.4631083202511774,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.6194,
+      "step": 590
+    },
+    {
+      "epoch": 0.47095761381475665,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.6211,
+      "step": 600
+    },
+    {
+      "epoch": 0.478806907378336,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.617,
+      "step": 610
+    },
+    {
+      "epoch": 0.48665620094191525,
+      "grad_norm": 5.125,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.5901,
+      "step": 620
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.6067,
+      "step": 630
+    },
+    {
+      "epoch": 0.5023547880690737,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.5976,
+      "step": 640
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 5.75,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.6467,
+      "step": 650
+    },
+    {
+      "epoch": 0.5180533751962323,
+      "grad_norm": 7.25,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.5904,
+      "step": 660
+    },
+    {
+      "epoch": 0.5259026687598116,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.6814,
+      "step": 670
+    },
+    {
+      "epoch": 0.533751962323391,
+      "grad_norm": 6.125,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.6684,
+      "step": 680
+    },
+    {
+      "epoch": 0.5416012558869702,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.6206,
+      "step": 690
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 5.875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.6096,
+      "step": 700
+    },
+    {
+      "epoch": 0.5572998430141287,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.5468,
+      "step": 710
+    },
+    {
+      "epoch": 0.565149136577708,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.5883,
+      "step": 720
+    },
+    {
+      "epoch": 0.5729984301412873,
+      "grad_norm": 4.75,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.596,
+      "step": 730
+    },
+    {
+      "epoch": 0.5808477237048666,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.5178,
+      "step": 740
+    },
+    {
+      "epoch": 0.5886970172684458,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.6438,
+      "step": 750
+    },
+    {
+      "epoch": 0.5965463108320251,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.6975,
+      "step": 760
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 4.375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.5828,
+      "step": 770
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.5019,
+      "step": 780
+    },
+    {
+      "epoch": 0.6200941915227629,
+      "grad_norm": 6.75,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.6489,
+      "step": 790
+    },
+    {
+      "epoch": 0.6279434850863422,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.6364,
+      "step": 800
+    },
+    {
+      "epoch": 0.6357927786499215,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.5458,
+      "step": 810
+    },
+    {
+      "epoch": 0.6436420722135008,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.5859,
+      "step": 820
+    },
+    {
+      "epoch": 0.6514913657770801,
+      "grad_norm": 4.375,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.6369,
+      "step": 830
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.7126,
+      "step": 840
+    },
+    {
+      "epoch": 0.6671899529042387,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.5857,
+      "step": 850
+    },
+    {
+      "epoch": 0.6750392464678179,
+      "grad_norm": 5.0,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.645,
+      "step": 860
+    },
+    {
+      "epoch": 0.6828885400313972,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.6538,
+      "step": 870
+    },
+    {
+      "epoch": 0.6907378335949764,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.6045,
+      "step": 880
+    },
+    {
+      "epoch": 0.6985871271585558,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.6282,
+      "step": 890
+    },
+    {
+      "epoch": 0.706436420722135,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.5509,
+      "step": 900
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 7.125,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.6392,
+      "step": 910
+    },
+    {
+      "epoch": 0.7221350078492935,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.6152,
+      "step": 920
+    },
+    {
+      "epoch": 0.7299843014128728,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.7706,
+      "step": 930
+    },
+    {
+      "epoch": 0.7378335949764521,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.6656,
+      "step": 940
+    },
+    {
+      "epoch": 0.7456828885400314,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.6592,
+      "step": 950
+    },
+    {
+      "epoch": 0.7535321821036107,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.6141,
+      "step": 960
+    },
+    {
+      "epoch": 0.7613814756671899,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.6161,
+      "step": 970
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 4.25,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.7166,
+      "step": 980
+    },
+    {
+      "epoch": 0.7770800627943485,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.6028,
+      "step": 990
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.5262,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.arguana": 0.4712918167441636,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.climate_fever": 0.23854189010584295,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.dbpedia_entity": 0.33317100014903445,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.fever": 0.6283057972805869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.fiqa": 0.4448076388918781,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.hotpotqa": 0.6717658203813929,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.nfcorpus": 0.35468777200298285,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.nq": 0.41666515563087825,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.quora": 0.8413355846974869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.scidocs": 0.25553825348355963,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.scifact": 0.7484102747744907,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.webis_touche2020": 0.2895097071551271,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/avg": 0.47450255927478535,
+      "step": 1000
+    },
+    {
+      "epoch": 0.792778649921507,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.68762555129222e-05,
+      "loss": 0.5077,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8006279434850864,
+      "grad_norm": 5.0,
+      "learning_rate": 9.675366799960841e-05,
+      "loss": 0.7276,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8084772370486656,
+      "grad_norm": 6.59375,
+      "learning_rate": 9.662880177847594e-05,
+      "loss": 0.5999,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 3.875,
+      "learning_rate": 9.65016629353282e-05,
+      "loss": 0.6675,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 3.875,
+      "learning_rate": 9.637225766673307e-05,
+      "loss": 0.6682,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8320251177394035,
+      "grad_norm": 5.125,
+      "learning_rate": 9.624059227972075e-05,
+      "loss": 0.6381,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8398744113029827,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.610667319147647e-05,
+      "loss": 0.607,
+      "step": 1070
+    },
+    {
+      "epoch": 0.847723704866562,
+      "grad_norm": 4.75,
+      "learning_rate": 9.597050692902765e-05,
+      "loss": 0.5786,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8555729984301413,
+      "grad_norm": 6.125,
+      "learning_rate": 9.583210012892582e-05,
+      "loss": 0.5928,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8634222919937206,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.569145953692315e-05,
+      "loss": 0.595,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8712715855572999,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.55485920076437e-05,
+      "loss": 0.604,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.540350450424926e-05,
+      "loss": 0.6252,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8869701726844584,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.525620409810007e-05,
+      "loss": 0.5765,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8948194662480377,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.510669796841014e-05,
+      "loss": 0.5604,
+      "step": 1140
+    },
+    {
+      "epoch": 0.902668759811617,
+      "grad_norm": 6.34375,
+      "learning_rate": 9.495499340189728e-05,
+      "loss": 0.555,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9105180533751962,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.480109779242805e-05,
+      "loss": 0.6496,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9183673469387755,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.464501864065735e-05,
+      "loss": 0.584,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9262166405023547,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.448676355366282e-05,
+      "loss": 0.6644,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.432634024457414e-05,
+      "loss": 0.6077,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9419152276295133,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.416375653219709e-05,
+      "loss": 0.5531,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9497645211930926,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.399902034063242e-05,
+      "loss": 0.6593,
+      "step": 1210
+    },
+    {
+      "epoch": 0.957613814756672,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.383213969888972e-05,
+      "loss": 0.5477,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9654631083202512,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.366312274049602e-05,
+      "loss": 0.6071,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9733124018838305,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.34919777030994e-05,
+      "loss": 0.5946,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9811616954474097,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.331871292806759e-05,
+      "loss": 0.5837,
+      "step": 1250
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.314333686008125e-05,
+      "loss": 0.6056,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9968602825745683,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.296585804672252e-05,
+      "loss": 0.5941,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0047095761381475,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.278628513805838e-05,
+      "loss": 0.6723,
+      "step": 1280
+    },
+    {
+      "epoch": 1.012558869701727,
+      "grad_norm": 3.875,
+      "learning_rate": 9.260462688621905e-05,
+      "loss": 0.6221,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.242089214497145e-05,
+      "loss": 0.5627,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0282574568288854,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.223508986928766e-05,
+      "loss": 0.6045,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0361067503924646,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.204722911490846e-05,
+      "loss": 0.5723,
+      "step": 1320
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 5.625,
+      "learning_rate": 9.1857319037902e-05,
+      "loss": 0.5616,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0518053375196232,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.16653688942175e-05,
+      "loss": 0.5994,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0596546310832025,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.147138803923416e-05,
+      "loss": 0.6163,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0675039246467817,
+      "grad_norm": 8.0,
+      "learning_rate": 9.127538592730519e-05,
+      "loss": 0.6001,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0753532182103611,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.107737211129701e-05,
+      "loss": 0.55,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0832025117739403,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.087735624212365e-05,
+      "loss": 0.5109,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0910518053375196,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.06753480682764e-05,
+      "loss": 0.6688,
+      "step": 1390
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.047135743534868e-05,
+      "loss": 0.6504,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1067503924646782,
+      "grad_norm": 4.625,
+      "learning_rate": 9.02653942855561e-05,
+      "loss": 0.6315,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1145996860282574,
+      "grad_norm": 6.25,
+      "learning_rate": 9.005746865725207e-05,
+      "loss": 0.6297,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1224489795918366,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.984759068443831e-05,
+      "loss": 0.5962,
+      "step": 1430
+    },
+    {
+      "epoch": 1.130298273155416,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.963577059627118e-05,
+      "loss": 0.5681,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1381475667189953,
+      "grad_norm": 4.78125,
+      "learning_rate": 8.942201871656291e-05,
+      "loss": 0.6486,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1459968602825745,
+      "grad_norm": 5.40625,
+      "learning_rate": 8.920634546327858e-05,
+      "loss": 0.5462,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.898876134802826e-05,
+      "loss": 0.5653,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1616954474097332,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.87692769755548e-05,
+      "loss": 0.492,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1695447409733124,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.854790304321681e-05,
+      "loss": 0.5613,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1773940345368916,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.832465034046749e-05,
+      "loss": 0.5728,
+      "step": 1500
+    },
+    {
+      "epoch": 1.185243328100471,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.80995297483286e-05,
+      "loss": 0.5405,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1930926216640503,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.787255223886019e-05,
+      "loss": 0.5618,
+      "step": 1520
+    },
+    {
+      "epoch": 1.2009419152276295,
+      "grad_norm": 4.875,
+      "learning_rate": 8.764372887462586e-05,
+      "loss": 0.5731,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.741307080815357e-05,
+      "loss": 0.5745,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2166405023547882,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.718058928139205e-05,
+      "loss": 0.5266,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2244897959183674,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.694629562516294e-05,
+      "loss": 0.6069,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2323390894819466,
+      "grad_norm": 3.5,
+      "learning_rate": 8.671020125860851e-05,
+      "loss": 0.5433,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2401883830455258,
+      "grad_norm": 4.0,
+      "learning_rate": 8.647231768863512e-05,
+      "loss": 0.4842,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2480376766091053,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.623265650935234e-05,
+      "loss": 0.58,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2558869701726845,
+      "grad_norm": 4.125,
+      "learning_rate": 8.599122940150795e-05,
+      "loss": 0.5331,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 6.0,
+      "learning_rate": 8.574804813191858e-05,
+      "loss": 0.6408,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2715855572998431,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.550312455289625e-05,
+      "loss": 0.5044,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2794348508634223,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.525647060167063e-05,
+      "loss": 0.6098,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2872841444270016,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.500809829980734e-05,
+      "loss": 0.5911,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2951334379905808,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.4758019752622e-05,
+      "loss": 0.5876,
+      "step": 1650
+    },
+    {
+      "epoch": 1.30298273155416,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.450624714859016e-05,
+      "loss": 0.5618,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3108320251177394,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.425279275875336e-05,
+      "loss": 0.5629,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.399766893612096e-05,
+      "loss": 0.5672,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 5.5,
+      "learning_rate": 8.374088811506819e-05,
+      "loss": 0.6229,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3343799058084773,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.348246281072998e-05,
+      "loss": 0.5945,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3422291993720565,
+      "grad_norm": 4.125,
+      "learning_rate": 8.322240561839109e-05,
+      "loss": 0.5755,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3500784929356358,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.296072921287217e-05,
+      "loss": 0.5472,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3579277864992152,
+      "grad_norm": 6.34375,
+      "learning_rate": 8.269744634791208e-05,
+      "loss": 0.6287,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3657770800627944,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.243256985554621e-05,
+      "loss": 0.5684,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 5.875,
+      "learning_rate": 8.21661126454811e-05,
+      "loss": 0.5455,
+      "step": 1750
+    },
+    {
+      "epoch": 1.3814756671899528,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.189808770446528e-05,
+      "loss": 0.5403,
+      "step": 1760
+    },
+    {
+      "epoch": 1.389324960753532,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.5343,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3971742543171115,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.135738695798376e-05,
+      "loss": 0.5692,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4050235478806907,
+      "grad_norm": 5.125,
+      "learning_rate": 8.108473750550965e-05,
+      "loss": 0.5382,
+      "step": 1790
+    },
+    {
+      "epoch": 1.41287284144427,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.081057302678352e-05,
+      "loss": 0.5316,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4207221350078494,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.053490688419533e-05,
+      "loss": 0.5221,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 6.6875,
+      "learning_rate": 8.02577525133239e-05,
+      "loss": 0.6501,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4364207221350078,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.997912342228232e-05,
+      "loss": 0.5745,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4442700156985873,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.969903319105935e-05,
+      "loss": 0.5797,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4521193092621665,
+      "grad_norm": 6.21875,
+      "learning_rate": 7.941749547085777e-05,
+      "loss": 0.6899,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4599686028257457,
+      "grad_norm": 4.625,
+      "learning_rate": 7.913452398342881e-05,
+      "loss": 0.5891,
+      "step": 1860
+    },
+    {
+      "epoch": 1.467817896389325,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.885013252040359e-05,
+      "loss": 0.6118,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4756671899529041,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.856433494262077e-05,
+      "loss": 0.6102,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 5.0,
+      "learning_rate": 7.827714517945115e-05,
+      "loss": 0.5459,
+      "step": 1890
+    },
+    {
+      "epoch": 1.4913657770800628,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.798857722811856e-05,
+      "loss": 0.5717,
+      "step": 1900
+    },
+    {
+      "epoch": 1.499215070643642,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.769864515301787e-05,
+      "loss": 0.5279,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5070643642072215,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.740736308502938e-05,
+      "loss": 0.66,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5149136577708007,
+      "grad_norm": 6.5,
+      "learning_rate": 7.711474522083015e-05,
+      "loss": 0.546,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5227629513343799,
+      "grad_norm": 4.625,
+      "learning_rate": 7.682080582220206e-05,
+      "loss": 0.5207,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "grad_norm": 6.4375,
+      "learning_rate": 7.65255592153367e-05,
+      "loss": 0.6005,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.622901979013717e-05,
+      "loss": 0.5897,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5463108320251178,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.593120199951671e-05,
+      "loss": 0.5639,
+      "step": 1970
+    },
+    {
+      "epoch": 1.554160125588697,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.563212035869425e-05,
+      "loss": 0.5479,
+      "step": 1980
+    },
+    {
+      "epoch": 1.5620094191522762,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.533178944448705e-05,
+      "loss": 0.5539,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "grad_norm": 6.09375,
+      "learning_rate": 7.503022389460014e-05,
+      "loss": 0.625,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1000,
+  "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": 16,
+  "trial_name": null,
+  "trial_params": null
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-2000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9365177436ad70bb9d6000d96d3ccdc5c5ca38536defdcfec6b548401edbc07f
+size 6481

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/config.json

@@ -0,0 +1,45 @@
+{
+  "architectures": [
+    "ModernBertModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "repad_logits_with_grad": false,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.53.0",
+  "vocab_size": 50368
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bb963d2d2da684d38c649275acf790b53b2b9efeb1a54ccb045770f523ccc1d
+size 298041696

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:707d877637528a6ecd5001a42e1fe755b11b439f37b6d35579a2d7247f6dcfc2
+size 596170443

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac0e770c61e3acef7bd144efee65a450ed487b244ed293a8dd801394ffcae775
+size 15365

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fb125336725f7741cb4daa1e3d06e225bbacfde8d41c4dcabb6762c222e62c6
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:153c865f77c7129ba565bded50f334683d51c80f20e3cfec39e62f8737b86f0d
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d017ce00fcebac7edc058ddd138f194eb0340f2d8ad0879bdab08f922ed0846e
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01a8f419d963b829163d14296852a07ad341e5e63e49d44de7c0a10c9369d217
+size 1465

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/trainer_state.json

@@ -0,0 +1,2329 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.3547880690737832,
+  "eval_steps": 1000.0,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval/nano_beir.arguana": 0.4400549254974839,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.22880053057901228,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.30026716538531234,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6419178788751609,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4370593595651809,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6723047477224562,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3437446614516137,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.436711543163819,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.8439997970269911,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.26304331393922176,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7599341408912704,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.29696490432942557,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.47206691403557904,
+      "step": 0
+    },
+    {
+      "epoch": 0.007849293563579277,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8e-06,
+      "loss": 0.7314,
+      "step": 10
+    },
+    {
+      "epoch": 0.015698587127158554,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.8e-06,
+      "loss": 0.8796,
+      "step": 20
+    },
+    {
+      "epoch": 0.023547880690737835,
+      "grad_norm": 7.3125,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7311,
+      "step": 30
+    },
+    {
+      "epoch": 0.03139717425431711,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.8e-06,
+      "loss": 0.8214,
+      "step": 40
+    },
+    {
+      "epoch": 0.03924646781789639,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.8489,
+      "step": 50
+    },
+    {
+      "epoch": 0.04709576138147567,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.7661,
+      "step": 60
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.7525,
+      "step": 70
+    },
+    {
+      "epoch": 0.06279434850863422,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.8653,
+      "step": 80
+    },
+    {
+      "epoch": 0.0706436420722135,
+      "grad_norm": 6.25,
+      "learning_rate": 1.78e-05,
+      "loss": 0.7607,
+      "step": 90
+    },
+    {
+      "epoch": 0.07849293563579278,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.7259,
+      "step": 100
+    },
+    {
+      "epoch": 0.08634222919937205,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.18e-05,
+      "loss": 0.7862,
+      "step": 110
+    },
+    {
+      "epoch": 0.09419152276295134,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.38e-05,
+      "loss": 0.8212,
+      "step": 120
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 6.0,
+      "learning_rate": 2.58e-05,
+      "loss": 0.6884,
+      "step": 130
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.712,
+      "step": 140
+    },
+    {
+      "epoch": 0.11773940345368916,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.98e-05,
+      "loss": 0.6231,
+      "step": 150
+    },
+    {
+      "epoch": 0.12558869701726844,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.18e-05,
+      "loss": 0.6733,
+      "step": 160
+    },
+    {
+      "epoch": 0.13343799058084774,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.38e-05,
+      "loss": 0.6973,
+      "step": 170
+    },
+    {
+      "epoch": 0.141287284144427,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.58e-05,
+      "loss": 0.5928,
+      "step": 180
+    },
+    {
+      "epoch": 0.14913657770800628,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.6189,
+      "step": 190
+    },
+    {
+      "epoch": 0.15698587127158556,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.6661,
+      "step": 200
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 4.875,
+      "learning_rate": 4.18e-05,
+      "loss": 0.634,
+      "step": 210
+    },
+    {
+      "epoch": 0.1726844583987441,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.38e-05,
+      "loss": 0.6963,
+      "step": 220
+    },
+    {
+      "epoch": 0.18053375196232338,
+      "grad_norm": 5.875,
+      "learning_rate": 4.58e-05,
+      "loss": 0.6744,
+      "step": 230
+    },
+    {
+      "epoch": 0.18838304552590268,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.78e-05,
+      "loss": 0.6635,
+      "step": 240
+    },
+    {
+      "epoch": 0.19623233908948196,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.6874,
+      "step": 250
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 5.28125,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.6516,
+      "step": 260
+    },
+    {
+      "epoch": 0.2119309262166405,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.7579,
+      "step": 270
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.6867,
+      "step": 280
+    },
+    {
+      "epoch": 0.22762951334379905,
+      "grad_norm": 6.59375,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.7038,
+      "step": 290
+    },
+    {
+      "epoch": 0.23547880690737832,
+      "grad_norm": 5.9375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.6681,
+      "step": 300
+    },
+    {
+      "epoch": 0.24332810047095763,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.18e-05,
+      "loss": 0.6133,
+      "step": 310
+    },
+    {
+      "epoch": 0.25117739403453687,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.38e-05,
+      "loss": 0.6203,
+      "step": 320
+    },
+    {
+      "epoch": 0.25902668759811615,
+      "grad_norm": 6.3125,
+      "learning_rate": 6.58e-05,
+      "loss": 0.7164,
+      "step": 330
+    },
+    {
+      "epoch": 0.2668759811616955,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.7157,
+      "step": 340
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.98e-05,
+      "loss": 0.7176,
+      "step": 350
+    },
+    {
+      "epoch": 0.282574568288854,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.18e-05,
+      "loss": 0.702,
+      "step": 360
+    },
+    {
+      "epoch": 0.2904238618524333,
+      "grad_norm": 5.25,
+      "learning_rate": 7.38e-05,
+      "loss": 0.5834,
+      "step": 370
+    },
+    {
+      "epoch": 0.29827315541601257,
+      "grad_norm": 5.46875,
+      "learning_rate": 7.58e-05,
+      "loss": 0.594,
+      "step": 380
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 6.625,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.7667,
+      "step": 390
+    },
+    {
+      "epoch": 0.3139717425431711,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.6477,
+      "step": 400
+    },
+    {
+      "epoch": 0.3218210361067504,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.18e-05,
+      "loss": 0.6104,
+      "step": 410
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.38e-05,
+      "loss": 0.6879,
+      "step": 420
+    },
+    {
+      "epoch": 0.33751962323390894,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.6514,
+      "step": 430
+    },
+    {
+      "epoch": 0.3453689167974882,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.78e-05,
+      "loss": 0.649,
+      "step": 440
+    },
+    {
+      "epoch": 0.3532182103610675,
+      "grad_norm": 4.75,
+      "learning_rate": 8.98e-05,
+      "loss": 0.6204,
+      "step": 450
+    },
+    {
+      "epoch": 0.36106750392464676,
+      "grad_norm": 5.125,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.6633,
+      "step": 460
+    },
+    {
+      "epoch": 0.36891679748822603,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.7058,
+      "step": 470
+    },
+    {
+      "epoch": 0.37676609105180536,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.58e-05,
+      "loss": 0.5653,
+      "step": 480
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.6143,
+      "step": 490
+    },
+    {
+      "epoch": 0.3924646781789639,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.98e-05,
+      "loss": 0.6218,
+      "step": 500
+    },
+    {
+      "epoch": 0.4003139717425432,
+      "grad_norm": 6.03125,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.6115,
+      "step": 510
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.6206,
+      "step": 520
+    },
+    {
+      "epoch": 0.41601255886970173,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.629,
+      "step": 530
+    },
+    {
+      "epoch": 0.423861852433281,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.6492,
+      "step": 540
+    },
+    {
+      "epoch": 0.4317111459968603,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.585,
+      "step": 550
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 7.1875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.558,
+      "step": 560
+    },
+    {
+      "epoch": 0.4474097331240188,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.6242,
+      "step": 570
+    },
+    {
+      "epoch": 0.4552590266875981,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.682,
+      "step": 580
+    },
+    {
+      "epoch": 0.4631083202511774,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.6194,
+      "step": 590
+    },
+    {
+      "epoch": 0.47095761381475665,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.6211,
+      "step": 600
+    },
+    {
+      "epoch": 0.478806907378336,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.617,
+      "step": 610
+    },
+    {
+      "epoch": 0.48665620094191525,
+      "grad_norm": 5.125,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.5901,
+      "step": 620
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.6067,
+      "step": 630
+    },
+    {
+      "epoch": 0.5023547880690737,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.5976,
+      "step": 640
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 5.75,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.6467,
+      "step": 650
+    },
+    {
+      "epoch": 0.5180533751962323,
+      "grad_norm": 7.25,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.5904,
+      "step": 660
+    },
+    {
+      "epoch": 0.5259026687598116,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.6814,
+      "step": 670
+    },
+    {
+      "epoch": 0.533751962323391,
+      "grad_norm": 6.125,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.6684,
+      "step": 680
+    },
+    {
+      "epoch": 0.5416012558869702,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.6206,
+      "step": 690
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 5.875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.6096,
+      "step": 700
+    },
+    {
+      "epoch": 0.5572998430141287,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.5468,
+      "step": 710
+    },
+    {
+      "epoch": 0.565149136577708,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.5883,
+      "step": 720
+    },
+    {
+      "epoch": 0.5729984301412873,
+      "grad_norm": 4.75,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.596,
+      "step": 730
+    },
+    {
+      "epoch": 0.5808477237048666,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.5178,
+      "step": 740
+    },
+    {
+      "epoch": 0.5886970172684458,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.6438,
+      "step": 750
+    },
+    {
+      "epoch": 0.5965463108320251,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.6975,
+      "step": 760
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 4.375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.5828,
+      "step": 770
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.5019,
+      "step": 780
+    },
+    {
+      "epoch": 0.6200941915227629,
+      "grad_norm": 6.75,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.6489,
+      "step": 790
+    },
+    {
+      "epoch": 0.6279434850863422,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.6364,
+      "step": 800
+    },
+    {
+      "epoch": 0.6357927786499215,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.5458,
+      "step": 810
+    },
+    {
+      "epoch": 0.6436420722135008,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.5859,
+      "step": 820
+    },
+    {
+      "epoch": 0.6514913657770801,
+      "grad_norm": 4.375,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.6369,
+      "step": 830
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.7126,
+      "step": 840
+    },
+    {
+      "epoch": 0.6671899529042387,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.5857,
+      "step": 850
+    },
+    {
+      "epoch": 0.6750392464678179,
+      "grad_norm": 5.0,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.645,
+      "step": 860
+    },
+    {
+      "epoch": 0.6828885400313972,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.6538,
+      "step": 870
+    },
+    {
+      "epoch": 0.6907378335949764,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.6045,
+      "step": 880
+    },
+    {
+      "epoch": 0.6985871271585558,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.6282,
+      "step": 890
+    },
+    {
+      "epoch": 0.706436420722135,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.5509,
+      "step": 900
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 7.125,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.6392,
+      "step": 910
+    },
+    {
+      "epoch": 0.7221350078492935,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.6152,
+      "step": 920
+    },
+    {
+      "epoch": 0.7299843014128728,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.7706,
+      "step": 930
+    },
+    {
+      "epoch": 0.7378335949764521,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.6656,
+      "step": 940
+    },
+    {
+      "epoch": 0.7456828885400314,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.6592,
+      "step": 950
+    },
+    {
+      "epoch": 0.7535321821036107,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.6141,
+      "step": 960
+    },
+    {
+      "epoch": 0.7613814756671899,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.6161,
+      "step": 970
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 4.25,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.7166,
+      "step": 980
+    },
+    {
+      "epoch": 0.7770800627943485,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.6028,
+      "step": 990
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.5262,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.arguana": 0.4712918167441636,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.climate_fever": 0.23854189010584295,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.dbpedia_entity": 0.33317100014903445,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.fever": 0.6283057972805869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.fiqa": 0.4448076388918781,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.hotpotqa": 0.6717658203813929,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.nfcorpus": 0.35468777200298285,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.nq": 0.41666515563087825,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.quora": 0.8413355846974869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.scidocs": 0.25553825348355963,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.scifact": 0.7484102747744907,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.webis_touche2020": 0.2895097071551271,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/avg": 0.47450255927478535,
+      "step": 1000
+    },
+    {
+      "epoch": 0.792778649921507,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.68762555129222e-05,
+      "loss": 0.5077,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8006279434850864,
+      "grad_norm": 5.0,
+      "learning_rate": 9.675366799960841e-05,
+      "loss": 0.7276,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8084772370486656,
+      "grad_norm": 6.59375,
+      "learning_rate": 9.662880177847594e-05,
+      "loss": 0.5999,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 3.875,
+      "learning_rate": 9.65016629353282e-05,
+      "loss": 0.6675,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 3.875,
+      "learning_rate": 9.637225766673307e-05,
+      "loss": 0.6682,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8320251177394035,
+      "grad_norm": 5.125,
+      "learning_rate": 9.624059227972075e-05,
+      "loss": 0.6381,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8398744113029827,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.610667319147647e-05,
+      "loss": 0.607,
+      "step": 1070
+    },
+    {
+      "epoch": 0.847723704866562,
+      "grad_norm": 4.75,
+      "learning_rate": 9.597050692902765e-05,
+      "loss": 0.5786,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8555729984301413,
+      "grad_norm": 6.125,
+      "learning_rate": 9.583210012892582e-05,
+      "loss": 0.5928,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8634222919937206,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.569145953692315e-05,
+      "loss": 0.595,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8712715855572999,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.55485920076437e-05,
+      "loss": 0.604,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.540350450424926e-05,
+      "loss": 0.6252,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8869701726844584,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.525620409810007e-05,
+      "loss": 0.5765,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8948194662480377,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.510669796841014e-05,
+      "loss": 0.5604,
+      "step": 1140
+    },
+    {
+      "epoch": 0.902668759811617,
+      "grad_norm": 6.34375,
+      "learning_rate": 9.495499340189728e-05,
+      "loss": 0.555,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9105180533751962,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.480109779242805e-05,
+      "loss": 0.6496,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9183673469387755,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.464501864065735e-05,
+      "loss": 0.584,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9262166405023547,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.448676355366282e-05,
+      "loss": 0.6644,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.432634024457414e-05,
+      "loss": 0.6077,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9419152276295133,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.416375653219709e-05,
+      "loss": 0.5531,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9497645211930926,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.399902034063242e-05,
+      "loss": 0.6593,
+      "step": 1210
+    },
+    {
+      "epoch": 0.957613814756672,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.383213969888972e-05,
+      "loss": 0.5477,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9654631083202512,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.366312274049602e-05,
+      "loss": 0.6071,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9733124018838305,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.34919777030994e-05,
+      "loss": 0.5946,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9811616954474097,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.331871292806759e-05,
+      "loss": 0.5837,
+      "step": 1250
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.314333686008125e-05,
+      "loss": 0.6056,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9968602825745683,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.296585804672252e-05,
+      "loss": 0.5941,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0047095761381475,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.278628513805838e-05,
+      "loss": 0.6723,
+      "step": 1280
+    },
+    {
+      "epoch": 1.012558869701727,
+      "grad_norm": 3.875,
+      "learning_rate": 9.260462688621905e-05,
+      "loss": 0.6221,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.242089214497145e-05,
+      "loss": 0.5627,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0282574568288854,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.223508986928766e-05,
+      "loss": 0.6045,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0361067503924646,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.204722911490846e-05,
+      "loss": 0.5723,
+      "step": 1320
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 5.625,
+      "learning_rate": 9.1857319037902e-05,
+      "loss": 0.5616,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0518053375196232,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.16653688942175e-05,
+      "loss": 0.5994,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0596546310832025,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.147138803923416e-05,
+      "loss": 0.6163,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0675039246467817,
+      "grad_norm": 8.0,
+      "learning_rate": 9.127538592730519e-05,
+      "loss": 0.6001,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0753532182103611,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.107737211129701e-05,
+      "loss": 0.55,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0832025117739403,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.087735624212365e-05,
+      "loss": 0.5109,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0910518053375196,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.06753480682764e-05,
+      "loss": 0.6688,
+      "step": 1390
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.047135743534868e-05,
+      "loss": 0.6504,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1067503924646782,
+      "grad_norm": 4.625,
+      "learning_rate": 9.02653942855561e-05,
+      "loss": 0.6315,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1145996860282574,
+      "grad_norm": 6.25,
+      "learning_rate": 9.005746865725207e-05,
+      "loss": 0.6297,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1224489795918366,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.984759068443831e-05,
+      "loss": 0.5962,
+      "step": 1430
+    },
+    {
+      "epoch": 1.130298273155416,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.963577059627118e-05,
+      "loss": 0.5681,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1381475667189953,
+      "grad_norm": 4.78125,
+      "learning_rate": 8.942201871656291e-05,
+      "loss": 0.6486,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1459968602825745,
+      "grad_norm": 5.40625,
+      "learning_rate": 8.920634546327858e-05,
+      "loss": 0.5462,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.898876134802826e-05,
+      "loss": 0.5653,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1616954474097332,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.87692769755548e-05,
+      "loss": 0.492,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1695447409733124,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.854790304321681e-05,
+      "loss": 0.5613,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1773940345368916,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.832465034046749e-05,
+      "loss": 0.5728,
+      "step": 1500
+    },
+    {
+      "epoch": 1.185243328100471,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.80995297483286e-05,
+      "loss": 0.5405,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1930926216640503,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.787255223886019e-05,
+      "loss": 0.5618,
+      "step": 1520
+    },
+    {
+      "epoch": 1.2009419152276295,
+      "grad_norm": 4.875,
+      "learning_rate": 8.764372887462586e-05,
+      "loss": 0.5731,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.741307080815357e-05,
+      "loss": 0.5745,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2166405023547882,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.718058928139205e-05,
+      "loss": 0.5266,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2244897959183674,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.694629562516294e-05,
+      "loss": 0.6069,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2323390894819466,
+      "grad_norm": 3.5,
+      "learning_rate": 8.671020125860851e-05,
+      "loss": 0.5433,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2401883830455258,
+      "grad_norm": 4.0,
+      "learning_rate": 8.647231768863512e-05,
+      "loss": 0.4842,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2480376766091053,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.623265650935234e-05,
+      "loss": 0.58,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2558869701726845,
+      "grad_norm": 4.125,
+      "learning_rate": 8.599122940150795e-05,
+      "loss": 0.5331,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 6.0,
+      "learning_rate": 8.574804813191858e-05,
+      "loss": 0.6408,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2715855572998431,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.550312455289625e-05,
+      "loss": 0.5044,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2794348508634223,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.525647060167063e-05,
+      "loss": 0.6098,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2872841444270016,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.500809829980734e-05,
+      "loss": 0.5911,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2951334379905808,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.4758019752622e-05,
+      "loss": 0.5876,
+      "step": 1650
+    },
+    {
+      "epoch": 1.30298273155416,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.450624714859016e-05,
+      "loss": 0.5618,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3108320251177394,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.425279275875336e-05,
+      "loss": 0.5629,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.399766893612096e-05,
+      "loss": 0.5672,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 5.5,
+      "learning_rate": 8.374088811506819e-05,
+      "loss": 0.6229,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3343799058084773,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.348246281072998e-05,
+      "loss": 0.5945,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3422291993720565,
+      "grad_norm": 4.125,
+      "learning_rate": 8.322240561839109e-05,
+      "loss": 0.5755,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3500784929356358,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.296072921287217e-05,
+      "loss": 0.5472,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3579277864992152,
+      "grad_norm": 6.34375,
+      "learning_rate": 8.269744634791208e-05,
+      "loss": 0.6287,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3657770800627944,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.243256985554621e-05,
+      "loss": 0.5684,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 5.875,
+      "learning_rate": 8.21661126454811e-05,
+      "loss": 0.5455,
+      "step": 1750
+    },
+    {
+      "epoch": 1.3814756671899528,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.189808770446528e-05,
+      "loss": 0.5403,
+      "step": 1760
+    },
+    {
+      "epoch": 1.389324960753532,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.5343,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3971742543171115,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.135738695798376e-05,
+      "loss": 0.5692,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4050235478806907,
+      "grad_norm": 5.125,
+      "learning_rate": 8.108473750550965e-05,
+      "loss": 0.5382,
+      "step": 1790
+    },
+    {
+      "epoch": 1.41287284144427,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.081057302678352e-05,
+      "loss": 0.5316,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4207221350078494,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.053490688419533e-05,
+      "loss": 0.5221,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 6.6875,
+      "learning_rate": 8.02577525133239e-05,
+      "loss": 0.6501,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4364207221350078,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.997912342228232e-05,
+      "loss": 0.5745,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4442700156985873,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.969903319105935e-05,
+      "loss": 0.5797,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4521193092621665,
+      "grad_norm": 6.21875,
+      "learning_rate": 7.941749547085777e-05,
+      "loss": 0.6899,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4599686028257457,
+      "grad_norm": 4.625,
+      "learning_rate": 7.913452398342881e-05,
+      "loss": 0.5891,
+      "step": 1860
+    },
+    {
+      "epoch": 1.467817896389325,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.885013252040359e-05,
+      "loss": 0.6118,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4756671899529041,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.856433494262077e-05,
+      "loss": 0.6102,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 5.0,
+      "learning_rate": 7.827714517945115e-05,
+      "loss": 0.5459,
+      "step": 1890
+    },
+    {
+      "epoch": 1.4913657770800628,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.798857722811856e-05,
+      "loss": 0.5717,
+      "step": 1900
+    },
+    {
+      "epoch": 1.499215070643642,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.769864515301787e-05,
+      "loss": 0.5279,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5070643642072215,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.740736308502938e-05,
+      "loss": 0.66,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5149136577708007,
+      "grad_norm": 6.5,
+      "learning_rate": 7.711474522083015e-05,
+      "loss": 0.546,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5227629513343799,
+      "grad_norm": 4.625,
+      "learning_rate": 7.682080582220206e-05,
+      "loss": 0.5207,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "grad_norm": 6.4375,
+      "learning_rate": 7.65255592153367e-05,
+      "loss": 0.6005,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.622901979013717e-05,
+      "loss": 0.5897,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5463108320251178,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.593120199951671e-05,
+      "loss": 0.5639,
+      "step": 1970
+    },
+    {
+      "epoch": 1.554160125588697,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.563212035869425e-05,
+      "loss": 0.5479,
+      "step": 1980
+    },
+    {
+      "epoch": 1.5620094191522762,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.533178944448705e-05,
+      "loss": 0.5539,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "grad_norm": 6.09375,
+      "learning_rate": 7.503022389460014e-05,
+      "loss": 0.625,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.arguana": 0.4666002720912347,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.climate_fever": 0.22462604477746737,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.dbpedia_entity": 0.3175867497382319,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.fever": 0.6140103747506924,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.fiqa": 0.43978199849769206,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.hotpotqa": 0.6788521964877974,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.nfcorpus": 0.3485625465712446,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.nq": 0.42598841107613383,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.quora": 0.833039807676506,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.scidocs": 0.24848467409467515,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.scifact": 0.7682290881646785,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.webis_touche2020": 0.2836464597787185,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/avg": 0.47078405197542267,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5777080062794349,
+      "grad_norm": 5.15625,
+      "learning_rate": 7.472743840691299e-05,
+      "loss": 0.4956,
+      "step": 2010
+    },
+    {
+      "epoch": 1.585557299843014,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.44234477387631e-05,
+      "loss": 0.4504,
+      "step": 2020
+    },
+    {
+      "epoch": 1.5934065934065935,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.411826670622676e-05,
+      "loss": 0.5722,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6012558869701727,
+      "grad_norm": 4.875,
+      "learning_rate": 7.381191018339696e-05,
+      "loss": 0.5777,
+      "step": 2040
+    },
+    {
+      "epoch": 1.609105180533752,
+      "grad_norm": 5.34375,
+      "learning_rate": 7.350439310165842e-05,
+      "loss": 0.5645,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6169544740973314,
+      "grad_norm": 5.125,
+      "learning_rate": 7.319573044895986e-05,
+      "loss": 0.5454,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6248037676609104,
+      "grad_norm": 5.4375,
+      "learning_rate": 7.28859372690835e-05,
+      "loss": 0.5681,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6326530612244898,
+      "grad_norm": 4.96875,
+      "learning_rate": 7.257502866091192e-05,
+      "loss": 0.5132,
+      "step": 2080
+    },
+    {
+      "epoch": 1.640502354788069,
+      "grad_norm": 5.125,
+      "learning_rate": 7.226301977769199e-05,
+      "loss": 0.5463,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.194992582629654e-05,
+      "loss": 0.6572,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6562009419152277,
+      "grad_norm": 4.625,
+      "learning_rate": 7.163576206648304e-05,
+      "loss": 0.5608,
+      "step": 2110
+    },
+    {
+      "epoch": 1.664050235478807,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.132054381014995e-05,
+      "loss": 0.5352,
+      "step": 2120
+    },
+    {
+      "epoch": 1.6718995290423861,
+      "grad_norm": 6.25,
+      "learning_rate": 7.100428642059033e-05,
+      "loss": 0.6148,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6797488226059656,
+      "grad_norm": 4.75,
+      "learning_rate": 7.068700531174319e-05,
+      "loss": 0.6296,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6875981161695446,
+      "grad_norm": 5.125,
+      "learning_rate": 7.036871594744217e-05,
+      "loss": 0.5741,
+      "step": 2150
+    },
+    {
+      "epoch": 1.695447409733124,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.004943384066187e-05,
+      "loss": 0.5457,
+      "step": 2160
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.97291745527617e-05,
+      "loss": 0.6556,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7111459968602825,
+      "grad_norm": 5.875,
+      "learning_rate": 6.940795369272754e-05,
+      "loss": 0.6284,
+      "step": 2180
+    },
+    {
+      "epoch": 1.718995290423862,
+      "grad_norm": 4.40625,
+      "learning_rate": 6.908578691641092e-05,
+      "loss": 0.5214,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7268445839874411,
+      "grad_norm": 5.96875,
+      "learning_rate": 6.876268992576604e-05,
+      "loss": 0.5484,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7346938775510203,
+      "grad_norm": 5.75,
+      "learning_rate": 6.843867846808438e-05,
+      "loss": 0.5804,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7425431711145998,
+      "grad_norm": 5.59375,
+      "learning_rate": 6.811376833522729e-05,
+      "loss": 0.5653,
+      "step": 2220
+    },
+    {
+      "epoch": 1.750392464678179,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.778797536285624e-05,
+      "loss": 0.5328,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7582417582417582,
+      "grad_norm": 4.0,
+      "learning_rate": 6.746131542966112e-05,
+      "loss": 0.5909,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7660910518053377,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.713380445658618e-05,
+      "loss": 0.5715,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7739403453689166,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.680545840605423e-05,
+      "loss": 0.4487,
+      "step": 2260
+    },
+    {
+      "epoch": 1.781789638932496,
+      "grad_norm": 4.9375,
+      "learning_rate": 6.647629328118851e-05,
+      "loss": 0.5661,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7896389324960753,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.614632512503288e-05,
+      "loss": 0.5829,
+      "step": 2280
+    },
+    {
+      "epoch": 1.7974882260596545,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.58155700197697e-05,
+      "loss": 0.5994,
+      "step": 2290
+    },
+    {
+      "epoch": 1.805337519623234,
+      "grad_norm": 4.71875,
+      "learning_rate": 6.548404408593621e-05,
+      "loss": 0.579,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8131868131868132,
+      "grad_norm": 4.25,
+      "learning_rate": 6.515176348163871e-05,
+      "loss": 0.488,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8210361067503924,
+      "grad_norm": 4.25,
+      "learning_rate": 6.481874440176506e-05,
+      "loss": 0.5083,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8288854003139718,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.448500307719538e-05,
+      "loss": 0.6211,
+      "step": 2330
+    },
+    {
+      "epoch": 1.836734693877551,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.415055577401102e-05,
+      "loss": 0.5724,
+      "step": 2340
+    },
+    {
+      "epoch": 1.8445839874411303,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.381541879270169e-05,
+      "loss": 0.5523,
+      "step": 2350
+    },
+    {
+      "epoch": 1.8524332810047097,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.347960846737106e-05,
+      "loss": 0.5615,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8602825745682887,
+      "grad_norm": 4.375,
+      "learning_rate": 6.31431411649406e-05,
+      "loss": 0.4643,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8681318681318682,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.280603328435198e-05,
+      "loss": 0.5643,
+      "step": 2380
+    },
+    {
+      "epoch": 1.8759811616954474,
+      "grad_norm": 5.375,
+      "learning_rate": 6.24683012557677e-05,
+      "loss": 0.5446,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8838304552590266,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.212996153977037e-05,
+      "loss": 0.5759,
+      "step": 2400
+    },
+    {
+      "epoch": 1.891679748822606,
+      "grad_norm": 5.46875,
+      "learning_rate": 6.179103062656042e-05,
+      "loss": 0.5231,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8995290423861853,
+      "grad_norm": 5.5625,
+      "learning_rate": 6.145152503515239e-05,
+      "loss": 0.5679,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9073783359497645,
+      "grad_norm": 4.78125,
+      "learning_rate": 6.111146131256983e-05,
+      "loss": 0.5482,
+      "step": 2430
+    },
+    {
+      "epoch": 1.915227629513344,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.077085603303882e-05,
+      "loss": 0.5287,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 5.8125,
+      "learning_rate": 6.042972579718019e-05,
+      "loss": 0.6531,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9309262166405023,
+      "grad_norm": 4.75,
+      "learning_rate": 6.008808723120035e-05,
+      "loss": 0.5463,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9387755102040818,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.974595698608103e-05,
+      "loss": 0.5132,
+      "step": 2470
+    },
+    {
+      "epoch": 1.9466248037676608,
+      "grad_norm": 4.6875,
+      "learning_rate": 5.9403351736767696e-05,
+      "loss": 0.6158,
+      "step": 2480
+    },
+    {
+      "epoch": 1.9544740973312402,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.906028818135687e-05,
+      "loss": 0.5588,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9623233908948194,
+      "grad_norm": 5.125,
+      "learning_rate": 5.8716783040282244e-05,
+      "loss": 0.5456,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9701726844583987,
+      "grad_norm": 4.65625,
+      "learning_rate": 5.837285305549978e-05,
+      "loss": 0.6038,
+      "step": 2510
+    },
+    {
+      "epoch": 1.978021978021978,
+      "grad_norm": 5.25,
+      "learning_rate": 5.8028514989671724e-05,
+      "loss": 0.5504,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9858712715855573,
+      "grad_norm": 5.0,
+      "learning_rate": 5.7683785625349616e-05,
+      "loss": 0.5705,
+      "step": 2530
+    },
+    {
+      "epoch": 1.9937205651491365,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.733868176415633e-05,
+      "loss": 0.5866,
+      "step": 2540
+    },
+    {
+      "epoch": 2.001569858712716,
+      "grad_norm": 5.125,
+      "learning_rate": 5.699322022596722e-05,
+      "loss": 0.6031,
+      "step": 2550
+    },
+    {
+      "epoch": 2.009419152276295,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.664741784809022e-05,
+      "loss": 0.5695,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0172684458398744,
+      "grad_norm": 5.65625,
+      "learning_rate": 5.630129148444543e-05,
+      "loss": 0.5491,
+      "step": 2570
+    },
+    {
+      "epoch": 2.025117739403454,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.595485800474349e-05,
+      "loss": 0.5192,
+      "step": 2580
+    },
+    {
+      "epoch": 2.032967032967033,
+      "grad_norm": 6.25,
+      "learning_rate": 5.560813429366345e-05,
+      "loss": 0.5584,
+      "step": 2590
+    },
+    {
+      "epoch": 2.0408163265306123,
+      "grad_norm": 4.78125,
+      "learning_rate": 5.5261137250029835e-05,
+      "loss": 0.4623,
+      "step": 2600
+    },
+    {
+      "epoch": 2.0486656200941917,
+      "grad_norm": 4.53125,
+      "learning_rate": 5.4913883785988993e-05,
+      "loss": 0.5125,
+      "step": 2610
+    },
+    {
+      "epoch": 2.0565149136577707,
+      "grad_norm": 4.5,
+      "learning_rate": 5.456639082618489e-05,
+      "loss": 0.4937,
+      "step": 2620
+    },
+    {
+      "epoch": 2.06436420722135,
+      "grad_norm": 4.65625,
+      "learning_rate": 5.421867530693414e-05,
+      "loss": 0.5231,
+      "step": 2630
+    },
+    {
+      "epoch": 2.072213500784929,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.3870754175400595e-05,
+      "loss": 0.422,
+      "step": 2640
+    },
+    {
+      "epoch": 2.0800627943485086,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.352264438876935e-05,
+      "loss": 0.5353,
+      "step": 2650
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.3174362913420306e-05,
+      "loss": 0.4964,
+      "step": 2660
+    },
+    {
+      "epoch": 2.095761381475667,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.2825926724101236e-05,
+      "loss": 0.5701,
+      "step": 2670
+    },
+    {
+      "epoch": 2.1036106750392465,
+      "grad_norm": 5.5,
+      "learning_rate": 5.247735280310041e-05,
+      "loss": 0.5823,
+      "step": 2680
+    },
+    {
+      "epoch": 2.111459968602826,
+      "grad_norm": 4.125,
+      "learning_rate": 5.212865813941899e-05,
+      "loss": 0.4692,
+      "step": 2690
+    },
+    {
+      "epoch": 2.119309262166405,
+      "grad_norm": 6.9375,
+      "learning_rate": 5.1779859727942924e-05,
+      "loss": 0.5522,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1271585557299844,
+      "grad_norm": 5.78125,
+      "learning_rate": 5.143097456861474e-05,
+      "loss": 0.5413,
+      "step": 2710
+    },
+    {
+      "epoch": 2.1350078492935634,
+      "grad_norm": 4.625,
+      "learning_rate": 5.1082019665604895e-05,
+      "loss": 0.4929,
+      "step": 2720
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.073301202648304e-05,
+      "loss": 0.5183,
+      "step": 2730
+    },
+    {
+      "epoch": 2.1507064364207222,
+      "grad_norm": 6.0,
+      "learning_rate": 5.0383968661389146e-05,
+      "loss": 0.5698,
+      "step": 2740
+    },
+    {
+      "epoch": 2.1585557299843012,
+      "grad_norm": 5.0,
+      "learning_rate": 5.003490658220438e-05,
+      "loss": 0.4615,
+      "step": 2750
+    },
+    {
+      "epoch": 2.1664050235478807,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.968584280172206e-05,
+      "loss": 0.5037,
+      "step": 2760
+    },
+    {
+      "epoch": 2.17425431711146,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.933679433281836e-05,
+      "loss": 0.5184,
+      "step": 2770
+    },
+    {
+      "epoch": 2.182103610675039,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.8987778187623245e-05,
+      "loss": 0.5307,
+      "step": 2780
+    },
+    {
+      "epoch": 2.1899529042386185,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.863881137669123e-05,
+      "loss": 0.4549,
+      "step": 2790
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.8289910908172376e-05,
+      "loss": 0.5787,
+      "step": 2800
+    },
+    {
+      "epoch": 2.205651491365777,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.794109378698327e-05,
+      "loss": 0.6173,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2135007849293564,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7592377013978306e-05,
+      "loss": 0.5365,
+      "step": 2820
+    },
+    {
+      "epoch": 2.221350078492936,
+      "grad_norm": 5.125,
+      "learning_rate": 4.724377758512103e-05,
+      "loss": 0.4651,
+      "step": 2830
+    },
+    {
+      "epoch": 2.229199372056515,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.68953124906558e-05,
+      "loss": 0.5659,
+      "step": 2840
+    },
+    {
+      "epoch": 2.2370486656200943,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.654699871427971e-05,
+      "loss": 0.6009,
+      "step": 2850
+    },
+    {
+      "epoch": 2.2448979591836733,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.619885323231484e-05,
+      "loss": 0.5581,
+      "step": 2860
+    },
+    {
+      "epoch": 2.2527472527472527,
+      "grad_norm": 5.25,
+      "learning_rate": 4.58508930128808e-05,
+      "loss": 0.5357,
+      "step": 2870
+    },
+    {
+      "epoch": 2.260596546310832,
+      "grad_norm": 5.625,
+      "learning_rate": 4.550313501506781e-05,
+      "loss": 0.5235,
+      "step": 2880
+    },
+    {
+      "epoch": 2.268445839874411,
+      "grad_norm": 4.875,
+      "learning_rate": 4.515559618811005e-05,
+      "loss": 0.442,
+      "step": 2890
+    },
+    {
+      "epoch": 2.2762951334379906,
+      "grad_norm": 5.375,
+      "learning_rate": 4.4808293470559643e-05,
+      "loss": 0.4889,
+      "step": 2900
+    },
+    {
+      "epoch": 2.2841444270015696,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.446124378946107e-05,
+      "loss": 0.6102,
+      "step": 2910
+    },
+    {
+      "epoch": 2.291993720565149,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.411446405952618e-05,
+      "loss": 0.5462,
+      "step": 2920
+    },
+    {
+      "epoch": 2.2998430141287285,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.376797118230979e-05,
+      "loss": 0.4926,
+      "step": 2930
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 5.5,
+      "learning_rate": 4.342178204538588e-05,
+      "loss": 0.5301,
+      "step": 2940
+    },
+    {
+      "epoch": 2.315541601255887,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.307591352152459e-05,
+      "loss": 0.5053,
+      "step": 2950
+    },
+    {
+      "epoch": 2.3233908948194664,
+      "grad_norm": 5.0,
+      "learning_rate": 4.273038246786986e-05,
+      "loss": 0.5717,
+      "step": 2960
+    },
+    {
+      "epoch": 2.3312401883830454,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.238520572511773e-05,
+      "loss": 0.5054,
+      "step": 2970
+    },
+    {
+      "epoch": 2.339089481946625,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.2040400116695666e-05,
+      "loss": 0.4606,
+      "step": 2980
+    },
+    {
+      "epoch": 2.3469387755102042,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.169598244794261e-05,
+      "loss": 0.4655,
+      "step": 2990
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.135196950528982e-05,
+      "loss": 0.4932,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1000,
+  "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": 16,
+  "trial_name": null,
+  "trial_params": null
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-3000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9365177436ad70bb9d6000d96d3ccdc5c5ca38536defdcfec6b548401edbc07f
+size 6481

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/config.json

@@ -0,0 +1,45 @@
+{
+  "architectures": [
+    "ModernBertModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "repad_logits_with_grad": false,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.53.0",
+  "vocab_size": 50368
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1dd3dcdc59285c708a09271b560d3abb25a372173e901184f4c94aa67938defe
+size 298041696

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:421aa92cd21b5f582224f33575aab4f4b0cb5e42700e3b3279f086383cb122a0
+size 596170443

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24c12cceaffe8acec04189ba71dbb81f100dd5cd03b86390253117b3fab2911d
+size 15365

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69adbd9997461473344beb9c44d2e496e24fbc4d6fe69245ab0bd127882efd96
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:baaf7026d9e588ca2ba5b4de8768a379982b7530f39e0b88fa44af9bef8e8bc9
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ec81cbdf599c90b49fb13998f6e46e2492b55345216231b6ce078f88cf04eae
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9218a8f161ba45a13a24f24cd020f32b0abec4bcc53a2764e48b480568dd2d13
+size 1465

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/trainer_state.json

@@ -0,0 +1,3094 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.1397174254317113,
+  "eval_steps": 1000.0,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval/nano_beir.arguana": 0.4400549254974839,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.22880053057901228,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.30026716538531234,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6419178788751609,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4370593595651809,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6723047477224562,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3437446614516137,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.436711543163819,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.8439997970269911,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.26304331393922176,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7599341408912704,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.29696490432942557,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.47206691403557904,
+      "step": 0
+    },
+    {
+      "epoch": 0.007849293563579277,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8e-06,
+      "loss": 0.7314,
+      "step": 10
+    },
+    {
+      "epoch": 0.015698587127158554,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.8e-06,
+      "loss": 0.8796,
+      "step": 20
+    },
+    {
+      "epoch": 0.023547880690737835,
+      "grad_norm": 7.3125,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7311,
+      "step": 30
+    },
+    {
+      "epoch": 0.03139717425431711,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.8e-06,
+      "loss": 0.8214,
+      "step": 40
+    },
+    {
+      "epoch": 0.03924646781789639,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.8489,
+      "step": 50
+    },
+    {
+      "epoch": 0.04709576138147567,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.7661,
+      "step": 60
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.7525,
+      "step": 70
+    },
+    {
+      "epoch": 0.06279434850863422,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.8653,
+      "step": 80
+    },
+    {
+      "epoch": 0.0706436420722135,
+      "grad_norm": 6.25,
+      "learning_rate": 1.78e-05,
+      "loss": 0.7607,
+      "step": 90
+    },
+    {
+      "epoch": 0.07849293563579278,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.7259,
+      "step": 100
+    },
+    {
+      "epoch": 0.08634222919937205,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.18e-05,
+      "loss": 0.7862,
+      "step": 110
+    },
+    {
+      "epoch": 0.09419152276295134,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.38e-05,
+      "loss": 0.8212,
+      "step": 120
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 6.0,
+      "learning_rate": 2.58e-05,
+      "loss": 0.6884,
+      "step": 130
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.712,
+      "step": 140
+    },
+    {
+      "epoch": 0.11773940345368916,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.98e-05,
+      "loss": 0.6231,
+      "step": 150
+    },
+    {
+      "epoch": 0.12558869701726844,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.18e-05,
+      "loss": 0.6733,
+      "step": 160
+    },
+    {
+      "epoch": 0.13343799058084774,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.38e-05,
+      "loss": 0.6973,
+      "step": 170
+    },
+    {
+      "epoch": 0.141287284144427,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.58e-05,
+      "loss": 0.5928,
+      "step": 180
+    },
+    {
+      "epoch": 0.14913657770800628,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.6189,
+      "step": 190
+    },
+    {
+      "epoch": 0.15698587127158556,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.6661,
+      "step": 200
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 4.875,
+      "learning_rate": 4.18e-05,
+      "loss": 0.634,
+      "step": 210
+    },
+    {
+      "epoch": 0.1726844583987441,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.38e-05,
+      "loss": 0.6963,
+      "step": 220
+    },
+    {
+      "epoch": 0.18053375196232338,
+      "grad_norm": 5.875,
+      "learning_rate": 4.58e-05,
+      "loss": 0.6744,
+      "step": 230
+    },
+    {
+      "epoch": 0.18838304552590268,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.78e-05,
+      "loss": 0.6635,
+      "step": 240
+    },
+    {
+      "epoch": 0.19623233908948196,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.6874,
+      "step": 250
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 5.28125,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.6516,
+      "step": 260
+    },
+    {
+      "epoch": 0.2119309262166405,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.7579,
+      "step": 270
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.6867,
+      "step": 280
+    },
+    {
+      "epoch": 0.22762951334379905,
+      "grad_norm": 6.59375,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.7038,
+      "step": 290
+    },
+    {
+      "epoch": 0.23547880690737832,
+      "grad_norm": 5.9375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.6681,
+      "step": 300
+    },
+    {
+      "epoch": 0.24332810047095763,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.18e-05,
+      "loss": 0.6133,
+      "step": 310
+    },
+    {
+      "epoch": 0.25117739403453687,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.38e-05,
+      "loss": 0.6203,
+      "step": 320
+    },
+    {
+      "epoch": 0.25902668759811615,
+      "grad_norm": 6.3125,
+      "learning_rate": 6.58e-05,
+      "loss": 0.7164,
+      "step": 330
+    },
+    {
+      "epoch": 0.2668759811616955,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.7157,
+      "step": 340
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.98e-05,
+      "loss": 0.7176,
+      "step": 350
+    },
+    {
+      "epoch": 0.282574568288854,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.18e-05,
+      "loss": 0.702,
+      "step": 360
+    },
+    {
+      "epoch": 0.2904238618524333,
+      "grad_norm": 5.25,
+      "learning_rate": 7.38e-05,
+      "loss": 0.5834,
+      "step": 370
+    },
+    {
+      "epoch": 0.29827315541601257,
+      "grad_norm": 5.46875,
+      "learning_rate": 7.58e-05,
+      "loss": 0.594,
+      "step": 380
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 6.625,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.7667,
+      "step": 390
+    },
+    {
+      "epoch": 0.3139717425431711,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.6477,
+      "step": 400
+    },
+    {
+      "epoch": 0.3218210361067504,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.18e-05,
+      "loss": 0.6104,
+      "step": 410
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.38e-05,
+      "loss": 0.6879,
+      "step": 420
+    },
+    {
+      "epoch": 0.33751962323390894,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.6514,
+      "step": 430
+    },
+    {
+      "epoch": 0.3453689167974882,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.78e-05,
+      "loss": 0.649,
+      "step": 440
+    },
+    {
+      "epoch": 0.3532182103610675,
+      "grad_norm": 4.75,
+      "learning_rate": 8.98e-05,
+      "loss": 0.6204,
+      "step": 450
+    },
+    {
+      "epoch": 0.36106750392464676,
+      "grad_norm": 5.125,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.6633,
+      "step": 460
+    },
+    {
+      "epoch": 0.36891679748822603,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.7058,
+      "step": 470
+    },
+    {
+      "epoch": 0.37676609105180536,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.58e-05,
+      "loss": 0.5653,
+      "step": 480
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.6143,
+      "step": 490
+    },
+    {
+      "epoch": 0.3924646781789639,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.98e-05,
+      "loss": 0.6218,
+      "step": 500
+    },
+    {
+      "epoch": 0.4003139717425432,
+      "grad_norm": 6.03125,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.6115,
+      "step": 510
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.6206,
+      "step": 520
+    },
+    {
+      "epoch": 0.41601255886970173,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.629,
+      "step": 530
+    },
+    {
+      "epoch": 0.423861852433281,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.6492,
+      "step": 540
+    },
+    {
+      "epoch": 0.4317111459968603,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.585,
+      "step": 550
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 7.1875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.558,
+      "step": 560
+    },
+    {
+      "epoch": 0.4474097331240188,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.6242,
+      "step": 570
+    },
+    {
+      "epoch": 0.4552590266875981,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.682,
+      "step": 580
+    },
+    {
+      "epoch": 0.4631083202511774,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.6194,
+      "step": 590
+    },
+    {
+      "epoch": 0.47095761381475665,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.6211,
+      "step": 600
+    },
+    {
+      "epoch": 0.478806907378336,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.617,
+      "step": 610
+    },
+    {
+      "epoch": 0.48665620094191525,
+      "grad_norm": 5.125,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.5901,
+      "step": 620
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.6067,
+      "step": 630
+    },
+    {
+      "epoch": 0.5023547880690737,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.5976,
+      "step": 640
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 5.75,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.6467,
+      "step": 650
+    },
+    {
+      "epoch": 0.5180533751962323,
+      "grad_norm": 7.25,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.5904,
+      "step": 660
+    },
+    {
+      "epoch": 0.5259026687598116,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.6814,
+      "step": 670
+    },
+    {
+      "epoch": 0.533751962323391,
+      "grad_norm": 6.125,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.6684,
+      "step": 680
+    },
+    {
+      "epoch": 0.5416012558869702,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.6206,
+      "step": 690
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 5.875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.6096,
+      "step": 700
+    },
+    {
+      "epoch": 0.5572998430141287,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.5468,
+      "step": 710
+    },
+    {
+      "epoch": 0.565149136577708,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.5883,
+      "step": 720
+    },
+    {
+      "epoch": 0.5729984301412873,
+      "grad_norm": 4.75,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.596,
+      "step": 730
+    },
+    {
+      "epoch": 0.5808477237048666,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.5178,
+      "step": 740
+    },
+    {
+      "epoch": 0.5886970172684458,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.6438,
+      "step": 750
+    },
+    {
+      "epoch": 0.5965463108320251,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.6975,
+      "step": 760
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 4.375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.5828,
+      "step": 770
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.5019,
+      "step": 780
+    },
+    {
+      "epoch": 0.6200941915227629,
+      "grad_norm": 6.75,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.6489,
+      "step": 790
+    },
+    {
+      "epoch": 0.6279434850863422,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.6364,
+      "step": 800
+    },
+    {
+      "epoch": 0.6357927786499215,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.5458,
+      "step": 810
+    },
+    {
+      "epoch": 0.6436420722135008,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.5859,
+      "step": 820
+    },
+    {
+      "epoch": 0.6514913657770801,
+      "grad_norm": 4.375,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.6369,
+      "step": 830
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.7126,
+      "step": 840
+    },
+    {
+      "epoch": 0.6671899529042387,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.5857,
+      "step": 850
+    },
+    {
+      "epoch": 0.6750392464678179,
+      "grad_norm": 5.0,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.645,
+      "step": 860
+    },
+    {
+      "epoch": 0.6828885400313972,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.6538,
+      "step": 870
+    },
+    {
+      "epoch": 0.6907378335949764,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.6045,
+      "step": 880
+    },
+    {
+      "epoch": 0.6985871271585558,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.6282,
+      "step": 890
+    },
+    {
+      "epoch": 0.706436420722135,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.5509,
+      "step": 900
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 7.125,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.6392,
+      "step": 910
+    },
+    {
+      "epoch": 0.7221350078492935,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.6152,
+      "step": 920
+    },
+    {
+      "epoch": 0.7299843014128728,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.7706,
+      "step": 930
+    },
+    {
+      "epoch": 0.7378335949764521,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.6656,
+      "step": 940
+    },
+    {
+      "epoch": 0.7456828885400314,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.6592,
+      "step": 950
+    },
+    {
+      "epoch": 0.7535321821036107,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.6141,
+      "step": 960
+    },
+    {
+      "epoch": 0.7613814756671899,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.6161,
+      "step": 970
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 4.25,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.7166,
+      "step": 980
+    },
+    {
+      "epoch": 0.7770800627943485,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.6028,
+      "step": 990
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.5262,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.arguana": 0.4712918167441636,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.climate_fever": 0.23854189010584295,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.dbpedia_entity": 0.33317100014903445,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.fever": 0.6283057972805869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.fiqa": 0.4448076388918781,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.hotpotqa": 0.6717658203813929,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.nfcorpus": 0.35468777200298285,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.nq": 0.41666515563087825,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.quora": 0.8413355846974869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.scidocs": 0.25553825348355963,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.scifact": 0.7484102747744907,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/nano_beir.webis_touche2020": 0.2895097071551271,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "eval/avg": 0.47450255927478535,
+      "step": 1000
+    },
+    {
+      "epoch": 0.792778649921507,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.68762555129222e-05,
+      "loss": 0.5077,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8006279434850864,
+      "grad_norm": 5.0,
+      "learning_rate": 9.675366799960841e-05,
+      "loss": 0.7276,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8084772370486656,
+      "grad_norm": 6.59375,
+      "learning_rate": 9.662880177847594e-05,
+      "loss": 0.5999,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 3.875,
+      "learning_rate": 9.65016629353282e-05,
+      "loss": 0.6675,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 3.875,
+      "learning_rate": 9.637225766673307e-05,
+      "loss": 0.6682,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8320251177394035,
+      "grad_norm": 5.125,
+      "learning_rate": 9.624059227972075e-05,
+      "loss": 0.6381,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8398744113029827,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.610667319147647e-05,
+      "loss": 0.607,
+      "step": 1070
+    },
+    {
+      "epoch": 0.847723704866562,
+      "grad_norm": 4.75,
+      "learning_rate": 9.597050692902765e-05,
+      "loss": 0.5786,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8555729984301413,
+      "grad_norm": 6.125,
+      "learning_rate": 9.583210012892582e-05,
+      "loss": 0.5928,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8634222919937206,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.569145953692315e-05,
+      "loss": 0.595,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8712715855572999,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.55485920076437e-05,
+      "loss": 0.604,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.540350450424926e-05,
+      "loss": 0.6252,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8869701726844584,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.525620409810007e-05,
+      "loss": 0.5765,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8948194662480377,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.510669796841014e-05,
+      "loss": 0.5604,
+      "step": 1140
+    },
+    {
+      "epoch": 0.902668759811617,
+      "grad_norm": 6.34375,
+      "learning_rate": 9.495499340189728e-05,
+      "loss": 0.555,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9105180533751962,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.480109779242805e-05,
+      "loss": 0.6496,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9183673469387755,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.464501864065735e-05,
+      "loss": 0.584,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9262166405023547,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.448676355366282e-05,
+      "loss": 0.6644,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.432634024457414e-05,
+      "loss": 0.6077,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9419152276295133,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.416375653219709e-05,
+      "loss": 0.5531,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9497645211930926,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.399902034063242e-05,
+      "loss": 0.6593,
+      "step": 1210
+    },
+    {
+      "epoch": 0.957613814756672,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.383213969888972e-05,
+      "loss": 0.5477,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9654631083202512,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.366312274049602e-05,
+      "loss": 0.6071,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9733124018838305,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.34919777030994e-05,
+      "loss": 0.5946,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9811616954474097,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.331871292806759e-05,
+      "loss": 0.5837,
+      "step": 1250
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.314333686008125e-05,
+      "loss": 0.6056,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9968602825745683,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.296585804672252e-05,
+      "loss": 0.5941,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0047095761381475,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.278628513805838e-05,
+      "loss": 0.6723,
+      "step": 1280
+    },
+    {
+      "epoch": 1.012558869701727,
+      "grad_norm": 3.875,
+      "learning_rate": 9.260462688621905e-05,
+      "loss": 0.6221,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.242089214497145e-05,
+      "loss": 0.5627,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0282574568288854,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.223508986928766e-05,
+      "loss": 0.6045,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0361067503924646,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.204722911490846e-05,
+      "loss": 0.5723,
+      "step": 1320
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 5.625,
+      "learning_rate": 9.1857319037902e-05,
+      "loss": 0.5616,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0518053375196232,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.16653688942175e-05,
+      "loss": 0.5994,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0596546310832025,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.147138803923416e-05,
+      "loss": 0.6163,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0675039246467817,
+      "grad_norm": 8.0,
+      "learning_rate": 9.127538592730519e-05,
+      "loss": 0.6001,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0753532182103611,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.107737211129701e-05,
+      "loss": 0.55,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0832025117739403,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.087735624212365e-05,
+      "loss": 0.5109,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0910518053375196,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.06753480682764e-05,
+      "loss": 0.6688,
+      "step": 1390
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.047135743534868e-05,
+      "loss": 0.6504,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1067503924646782,
+      "grad_norm": 4.625,
+      "learning_rate": 9.02653942855561e-05,
+      "loss": 0.6315,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1145996860282574,
+      "grad_norm": 6.25,
+      "learning_rate": 9.005746865725207e-05,
+      "loss": 0.6297,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1224489795918366,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.984759068443831e-05,
+      "loss": 0.5962,
+      "step": 1430
+    },
+    {
+      "epoch": 1.130298273155416,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.963577059627118e-05,
+      "loss": 0.5681,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1381475667189953,
+      "grad_norm": 4.78125,
+      "learning_rate": 8.942201871656291e-05,
+      "loss": 0.6486,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1459968602825745,
+      "grad_norm": 5.40625,
+      "learning_rate": 8.920634546327858e-05,
+      "loss": 0.5462,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.898876134802826e-05,
+      "loss": 0.5653,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1616954474097332,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.87692769755548e-05,
+      "loss": 0.492,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1695447409733124,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.854790304321681e-05,
+      "loss": 0.5613,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1773940345368916,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.832465034046749e-05,
+      "loss": 0.5728,
+      "step": 1500
+    },
+    {
+      "epoch": 1.185243328100471,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.80995297483286e-05,
+      "loss": 0.5405,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1930926216640503,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.787255223886019e-05,
+      "loss": 0.5618,
+      "step": 1520
+    },
+    {
+      "epoch": 1.2009419152276295,
+      "grad_norm": 4.875,
+      "learning_rate": 8.764372887462586e-05,
+      "loss": 0.5731,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.741307080815357e-05,
+      "loss": 0.5745,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2166405023547882,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.718058928139205e-05,
+      "loss": 0.5266,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2244897959183674,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.694629562516294e-05,
+      "loss": 0.6069,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2323390894819466,
+      "grad_norm": 3.5,
+      "learning_rate": 8.671020125860851e-05,
+      "loss": 0.5433,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2401883830455258,
+      "grad_norm": 4.0,
+      "learning_rate": 8.647231768863512e-05,
+      "loss": 0.4842,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2480376766091053,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.623265650935234e-05,
+      "loss": 0.58,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2558869701726845,
+      "grad_norm": 4.125,
+      "learning_rate": 8.599122940150795e-05,
+      "loss": 0.5331,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 6.0,
+      "learning_rate": 8.574804813191858e-05,
+      "loss": 0.6408,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2715855572998431,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.550312455289625e-05,
+      "loss": 0.5044,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2794348508634223,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.525647060167063e-05,
+      "loss": 0.6098,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2872841444270016,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.500809829980734e-05,
+      "loss": 0.5911,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2951334379905808,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.4758019752622e-05,
+      "loss": 0.5876,
+      "step": 1650
+    },
+    {
+      "epoch": 1.30298273155416,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.450624714859016e-05,
+      "loss": 0.5618,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3108320251177394,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.425279275875336e-05,
+      "loss": 0.5629,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.399766893612096e-05,
+      "loss": 0.5672,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 5.5,
+      "learning_rate": 8.374088811506819e-05,
+      "loss": 0.6229,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3343799058084773,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.348246281072998e-05,
+      "loss": 0.5945,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3422291993720565,
+      "grad_norm": 4.125,
+      "learning_rate": 8.322240561839109e-05,
+      "loss": 0.5755,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3500784929356358,
+      "grad_norm": 5.90625,
+      "learning_rate": 8.296072921287217e-05,
+      "loss": 0.5472,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3579277864992152,
+      "grad_norm": 6.34375,
+      "learning_rate": 8.269744634791208e-05,
+      "loss": 0.6287,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3657770800627944,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.243256985554621e-05,
+      "loss": 0.5684,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 5.875,
+      "learning_rate": 8.21661126454811e-05,
+      "loss": 0.5455,
+      "step": 1750
+    },
+    {
+      "epoch": 1.3814756671899528,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.189808770446528e-05,
+      "loss": 0.5403,
+      "step": 1760
+    },
+    {
+      "epoch": 1.389324960753532,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.5343,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3971742543171115,
+      "grad_norm": 4.53125,
+      "learning_rate": 8.135738695798376e-05,
+      "loss": 0.5692,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4050235478806907,
+      "grad_norm": 5.125,
+      "learning_rate": 8.108473750550965e-05,
+      "loss": 0.5382,
+      "step": 1790
+    },
+    {
+      "epoch": 1.41287284144427,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.081057302678352e-05,
+      "loss": 0.5316,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4207221350078494,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.053490688419533e-05,
+      "loss": 0.5221,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 6.6875,
+      "learning_rate": 8.02577525133239e-05,
+      "loss": 0.6501,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4364207221350078,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.997912342228232e-05,
+      "loss": 0.5745,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4442700156985873,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.969903319105935e-05,
+      "loss": 0.5797,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4521193092621665,
+      "grad_norm": 6.21875,
+      "learning_rate": 7.941749547085777e-05,
+      "loss": 0.6899,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4599686028257457,
+      "grad_norm": 4.625,
+      "learning_rate": 7.913452398342881e-05,
+      "loss": 0.5891,
+      "step": 1860
+    },
+    {
+      "epoch": 1.467817896389325,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.885013252040359e-05,
+      "loss": 0.6118,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4756671899529041,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.856433494262077e-05,
+      "loss": 0.6102,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 5.0,
+      "learning_rate": 7.827714517945115e-05,
+      "loss": 0.5459,
+      "step": 1890
+    },
+    {
+      "epoch": 1.4913657770800628,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.798857722811856e-05,
+      "loss": 0.5717,
+      "step": 1900
+    },
+    {
+      "epoch": 1.499215070643642,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.769864515301787e-05,
+      "loss": 0.5279,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5070643642072215,
+      "grad_norm": 4.6875,
+      "learning_rate": 7.740736308502938e-05,
+      "loss": 0.66,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5149136577708007,
+      "grad_norm": 6.5,
+      "learning_rate": 7.711474522083015e-05,
+      "loss": 0.546,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5227629513343799,
+      "grad_norm": 4.625,
+      "learning_rate": 7.682080582220206e-05,
+      "loss": 0.5207,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "grad_norm": 6.4375,
+      "learning_rate": 7.65255592153367e-05,
+      "loss": 0.6005,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.622901979013717e-05,
+      "loss": 0.5897,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5463108320251178,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.593120199951671e-05,
+      "loss": 0.5639,
+      "step": 1970
+    },
+    {
+      "epoch": 1.554160125588697,
+      "grad_norm": 4.59375,
+      "learning_rate": 7.563212035869425e-05,
+      "loss": 0.5479,
+      "step": 1980
+    },
+    {
+      "epoch": 1.5620094191522762,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.533178944448705e-05,
+      "loss": 0.5539,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "grad_norm": 6.09375,
+      "learning_rate": 7.503022389460014e-05,
+      "loss": 0.625,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.arguana": 0.4666002720912347,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.climate_fever": 0.22462604477746737,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.dbpedia_entity": 0.3175867497382319,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.fever": 0.6140103747506924,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.fiqa": 0.43978199849769206,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.hotpotqa": 0.6788521964877974,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.nfcorpus": 0.3485625465712446,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.nq": 0.42598841107613383,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.quora": 0.833039807676506,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.scidocs": 0.24848467409467515,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.scifact": 0.7682290881646785,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/nano_beir.webis_touche2020": 0.2836464597787185,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5698587127158556,
+      "eval/avg": 0.47078405197542267,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5777080062794349,
+      "grad_norm": 5.15625,
+      "learning_rate": 7.472743840691299e-05,
+      "loss": 0.4956,
+      "step": 2010
+    },
+    {
+      "epoch": 1.585557299843014,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.44234477387631e-05,
+      "loss": 0.4504,
+      "step": 2020
+    },
+    {
+      "epoch": 1.5934065934065935,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.411826670622676e-05,
+      "loss": 0.5722,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6012558869701727,
+      "grad_norm": 4.875,
+      "learning_rate": 7.381191018339696e-05,
+      "loss": 0.5777,
+      "step": 2040
+    },
+    {
+      "epoch": 1.609105180533752,
+      "grad_norm": 5.34375,
+      "learning_rate": 7.350439310165842e-05,
+      "loss": 0.5645,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6169544740973314,
+      "grad_norm": 5.125,
+      "learning_rate": 7.319573044895986e-05,
+      "loss": 0.5454,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6248037676609104,
+      "grad_norm": 5.4375,
+      "learning_rate": 7.28859372690835e-05,
+      "loss": 0.5681,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6326530612244898,
+      "grad_norm": 4.96875,
+      "learning_rate": 7.257502866091192e-05,
+      "loss": 0.5132,
+      "step": 2080
+    },
+    {
+      "epoch": 1.640502354788069,
+      "grad_norm": 5.125,
+      "learning_rate": 7.226301977769199e-05,
+      "loss": 0.5463,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.194992582629654e-05,
+      "loss": 0.6572,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6562009419152277,
+      "grad_norm": 4.625,
+      "learning_rate": 7.163576206648304e-05,
+      "loss": 0.5608,
+      "step": 2110
+    },
+    {
+      "epoch": 1.664050235478807,
+      "grad_norm": 4.8125,
+      "learning_rate": 7.132054381014995e-05,
+      "loss": 0.5352,
+      "step": 2120
+    },
+    {
+      "epoch": 1.6718995290423861,
+      "grad_norm": 6.25,
+      "learning_rate": 7.100428642059033e-05,
+      "loss": 0.6148,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6797488226059656,
+      "grad_norm": 4.75,
+      "learning_rate": 7.068700531174319e-05,
+      "loss": 0.6296,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6875981161695446,
+      "grad_norm": 5.125,
+      "learning_rate": 7.036871594744217e-05,
+      "loss": 0.5741,
+      "step": 2150
+    },
+    {
+      "epoch": 1.695447409733124,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.004943384066187e-05,
+      "loss": 0.5457,
+      "step": 2160
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.97291745527617e-05,
+      "loss": 0.6556,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7111459968602825,
+      "grad_norm": 5.875,
+      "learning_rate": 6.940795369272754e-05,
+      "loss": 0.6284,
+      "step": 2180
+    },
+    {
+      "epoch": 1.718995290423862,
+      "grad_norm": 4.40625,
+      "learning_rate": 6.908578691641092e-05,
+      "loss": 0.5214,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7268445839874411,
+      "grad_norm": 5.96875,
+      "learning_rate": 6.876268992576604e-05,
+      "loss": 0.5484,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7346938775510203,
+      "grad_norm": 5.75,
+      "learning_rate": 6.843867846808438e-05,
+      "loss": 0.5804,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7425431711145998,
+      "grad_norm": 5.59375,
+      "learning_rate": 6.811376833522729e-05,
+      "loss": 0.5653,
+      "step": 2220
+    },
+    {
+      "epoch": 1.750392464678179,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.778797536285624e-05,
+      "loss": 0.5328,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7582417582417582,
+      "grad_norm": 4.0,
+      "learning_rate": 6.746131542966112e-05,
+      "loss": 0.5909,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7660910518053377,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.713380445658618e-05,
+      "loss": 0.5715,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7739403453689166,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.680545840605423e-05,
+      "loss": 0.4487,
+      "step": 2260
+    },
+    {
+      "epoch": 1.781789638932496,
+      "grad_norm": 4.9375,
+      "learning_rate": 6.647629328118851e-05,
+      "loss": 0.5661,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7896389324960753,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.614632512503288e-05,
+      "loss": 0.5829,
+      "step": 2280
+    },
+    {
+      "epoch": 1.7974882260596545,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.58155700197697e-05,
+      "loss": 0.5994,
+      "step": 2290
+    },
+    {
+      "epoch": 1.805337519623234,
+      "grad_norm": 4.71875,
+      "learning_rate": 6.548404408593621e-05,
+      "loss": 0.579,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8131868131868132,
+      "grad_norm": 4.25,
+      "learning_rate": 6.515176348163871e-05,
+      "loss": 0.488,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8210361067503924,
+      "grad_norm": 4.25,
+      "learning_rate": 6.481874440176506e-05,
+      "loss": 0.5083,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8288854003139718,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.448500307719538e-05,
+      "loss": 0.6211,
+      "step": 2330
+    },
+    {
+      "epoch": 1.836734693877551,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.415055577401102e-05,
+      "loss": 0.5724,
+      "step": 2340
+    },
+    {
+      "epoch": 1.8445839874411303,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.381541879270169e-05,
+      "loss": 0.5523,
+      "step": 2350
+    },
+    {
+      "epoch": 1.8524332810047097,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.347960846737106e-05,
+      "loss": 0.5615,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8602825745682887,
+      "grad_norm": 4.375,
+      "learning_rate": 6.31431411649406e-05,
+      "loss": 0.4643,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8681318681318682,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.280603328435198e-05,
+      "loss": 0.5643,
+      "step": 2380
+    },
+    {
+      "epoch": 1.8759811616954474,
+      "grad_norm": 5.375,
+      "learning_rate": 6.24683012557677e-05,
+      "loss": 0.5446,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8838304552590266,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.212996153977037e-05,
+      "loss": 0.5759,
+      "step": 2400
+    },
+    {
+      "epoch": 1.891679748822606,
+      "grad_norm": 5.46875,
+      "learning_rate": 6.179103062656042e-05,
+      "loss": 0.5231,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8995290423861853,
+      "grad_norm": 5.5625,
+      "learning_rate": 6.145152503515239e-05,
+      "loss": 0.5679,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9073783359497645,
+      "grad_norm": 4.78125,
+      "learning_rate": 6.111146131256983e-05,
+      "loss": 0.5482,
+      "step": 2430
+    },
+    {
+      "epoch": 1.915227629513344,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.077085603303882e-05,
+      "loss": 0.5287,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 5.8125,
+      "learning_rate": 6.042972579718019e-05,
+      "loss": 0.6531,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9309262166405023,
+      "grad_norm": 4.75,
+      "learning_rate": 6.008808723120035e-05,
+      "loss": 0.5463,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9387755102040818,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.974595698608103e-05,
+      "loss": 0.5132,
+      "step": 2470
+    },
+    {
+      "epoch": 1.9466248037676608,
+      "grad_norm": 4.6875,
+      "learning_rate": 5.9403351736767696e-05,
+      "loss": 0.6158,
+      "step": 2480
+    },
+    {
+      "epoch": 1.9544740973312402,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.906028818135687e-05,
+      "loss": 0.5588,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9623233908948194,
+      "grad_norm": 5.125,
+      "learning_rate": 5.8716783040282244e-05,
+      "loss": 0.5456,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9701726844583987,
+      "grad_norm": 4.65625,
+      "learning_rate": 5.837285305549978e-05,
+      "loss": 0.6038,
+      "step": 2510
+    },
+    {
+      "epoch": 1.978021978021978,
+      "grad_norm": 5.25,
+      "learning_rate": 5.8028514989671724e-05,
+      "loss": 0.5504,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9858712715855573,
+      "grad_norm": 5.0,
+      "learning_rate": 5.7683785625349616e-05,
+      "loss": 0.5705,
+      "step": 2530
+    },
+    {
+      "epoch": 1.9937205651491365,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.733868176415633e-05,
+      "loss": 0.5866,
+      "step": 2540
+    },
+    {
+      "epoch": 2.001569858712716,
+      "grad_norm": 5.125,
+      "learning_rate": 5.699322022596722e-05,
+      "loss": 0.6031,
+      "step": 2550
+    },
+    {
+      "epoch": 2.009419152276295,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.664741784809022e-05,
+      "loss": 0.5695,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0172684458398744,
+      "grad_norm": 5.65625,
+      "learning_rate": 5.630129148444543e-05,
+      "loss": 0.5491,
+      "step": 2570
+    },
+    {
+      "epoch": 2.025117739403454,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.595485800474349e-05,
+      "loss": 0.5192,
+      "step": 2580
+    },
+    {
+      "epoch": 2.032967032967033,
+      "grad_norm": 6.25,
+      "learning_rate": 5.560813429366345e-05,
+      "loss": 0.5584,
+      "step": 2590
+    },
+    {
+      "epoch": 2.0408163265306123,
+      "grad_norm": 4.78125,
+      "learning_rate": 5.5261137250029835e-05,
+      "loss": 0.4623,
+      "step": 2600
+    },
+    {
+      "epoch": 2.0486656200941917,
+      "grad_norm": 4.53125,
+      "learning_rate": 5.4913883785988993e-05,
+      "loss": 0.5125,
+      "step": 2610
+    },
+    {
+      "epoch": 2.0565149136577707,
+      "grad_norm": 4.5,
+      "learning_rate": 5.456639082618489e-05,
+      "loss": 0.4937,
+      "step": 2620
+    },
+    {
+      "epoch": 2.06436420722135,
+      "grad_norm": 4.65625,
+      "learning_rate": 5.421867530693414e-05,
+      "loss": 0.5231,
+      "step": 2630
+    },
+    {
+      "epoch": 2.072213500784929,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.3870754175400595e-05,
+      "loss": 0.422,
+      "step": 2640
+    },
+    {
+      "epoch": 2.0800627943485086,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.352264438876935e-05,
+      "loss": 0.5353,
+      "step": 2650
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.3174362913420306e-05,
+      "loss": 0.4964,
+      "step": 2660
+    },
+    {
+      "epoch": 2.095761381475667,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.2825926724101236e-05,
+      "loss": 0.5701,
+      "step": 2670
+    },
+    {
+      "epoch": 2.1036106750392465,
+      "grad_norm": 5.5,
+      "learning_rate": 5.247735280310041e-05,
+      "loss": 0.5823,
+      "step": 2680
+    },
+    {
+      "epoch": 2.111459968602826,
+      "grad_norm": 4.125,
+      "learning_rate": 5.212865813941899e-05,
+      "loss": 0.4692,
+      "step": 2690
+    },
+    {
+      "epoch": 2.119309262166405,
+      "grad_norm": 6.9375,
+      "learning_rate": 5.1779859727942924e-05,
+      "loss": 0.5522,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1271585557299844,
+      "grad_norm": 5.78125,
+      "learning_rate": 5.143097456861474e-05,
+      "loss": 0.5413,
+      "step": 2710
+    },
+    {
+      "epoch": 2.1350078492935634,
+      "grad_norm": 4.625,
+      "learning_rate": 5.1082019665604895e-05,
+      "loss": 0.4929,
+      "step": 2720
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.073301202648304e-05,
+      "loss": 0.5183,
+      "step": 2730
+    },
+    {
+      "epoch": 2.1507064364207222,
+      "grad_norm": 6.0,
+      "learning_rate": 5.0383968661389146e-05,
+      "loss": 0.5698,
+      "step": 2740
+    },
+    {
+      "epoch": 2.1585557299843012,
+      "grad_norm": 5.0,
+      "learning_rate": 5.003490658220438e-05,
+      "loss": 0.4615,
+      "step": 2750
+    },
+    {
+      "epoch": 2.1664050235478807,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.968584280172206e-05,
+      "loss": 0.5037,
+      "step": 2760
+    },
+    {
+      "epoch": 2.17425431711146,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.933679433281836e-05,
+      "loss": 0.5184,
+      "step": 2770
+    },
+    {
+      "epoch": 2.182103610675039,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.8987778187623245e-05,
+      "loss": 0.5307,
+      "step": 2780
+    },
+    {
+      "epoch": 2.1899529042386185,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.863881137669123e-05,
+      "loss": 0.4549,
+      "step": 2790
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.8289910908172376e-05,
+      "loss": 0.5787,
+      "step": 2800
+    },
+    {
+      "epoch": 2.205651491365777,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.794109378698327e-05,
+      "loss": 0.6173,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2135007849293564,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7592377013978306e-05,
+      "loss": 0.5365,
+      "step": 2820
+    },
+    {
+      "epoch": 2.221350078492936,
+      "grad_norm": 5.125,
+      "learning_rate": 4.724377758512103e-05,
+      "loss": 0.4651,
+      "step": 2830
+    },
+    {
+      "epoch": 2.229199372056515,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.68953124906558e-05,
+      "loss": 0.5659,
+      "step": 2840
+    },
+    {
+      "epoch": 2.2370486656200943,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.654699871427971e-05,
+      "loss": 0.6009,
+      "step": 2850
+    },
+    {
+      "epoch": 2.2448979591836733,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.619885323231484e-05,
+      "loss": 0.5581,
+      "step": 2860
+    },
+    {
+      "epoch": 2.2527472527472527,
+      "grad_norm": 5.25,
+      "learning_rate": 4.58508930128808e-05,
+      "loss": 0.5357,
+      "step": 2870
+    },
+    {
+      "epoch": 2.260596546310832,
+      "grad_norm": 5.625,
+      "learning_rate": 4.550313501506781e-05,
+      "loss": 0.5235,
+      "step": 2880
+    },
+    {
+      "epoch": 2.268445839874411,
+      "grad_norm": 4.875,
+      "learning_rate": 4.515559618811005e-05,
+      "loss": 0.442,
+      "step": 2890
+    },
+    {
+      "epoch": 2.2762951334379906,
+      "grad_norm": 5.375,
+      "learning_rate": 4.4808293470559643e-05,
+      "loss": 0.4889,
+      "step": 2900
+    },
+    {
+      "epoch": 2.2841444270015696,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.446124378946107e-05,
+      "loss": 0.6102,
+      "step": 2910
+    },
+    {
+      "epoch": 2.291993720565149,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.411446405952618e-05,
+      "loss": 0.5462,
+      "step": 2920
+    },
+    {
+      "epoch": 2.2998430141287285,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.376797118230979e-05,
+      "loss": 0.4926,
+      "step": 2930
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 5.5,
+      "learning_rate": 4.342178204538588e-05,
+      "loss": 0.5301,
+      "step": 2940
+    },
+    {
+      "epoch": 2.315541601255887,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.307591352152459e-05,
+      "loss": 0.5053,
+      "step": 2950
+    },
+    {
+      "epoch": 2.3233908948194664,
+      "grad_norm": 5.0,
+      "learning_rate": 4.273038246786986e-05,
+      "loss": 0.5717,
+      "step": 2960
+    },
+    {
+      "epoch": 2.3312401883830454,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.238520572511773e-05,
+      "loss": 0.5054,
+      "step": 2970
+    },
+    {
+      "epoch": 2.339089481946625,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.2040400116695666e-05,
+      "loss": 0.4606,
+      "step": 2980
+    },
+    {
+      "epoch": 2.3469387755102042,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.169598244794261e-05,
+      "loss": 0.4655,
+      "step": 2990
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.135196950528982e-05,
+      "loss": 0.4932,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.arguana": 0.4735606937464546,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.climate_fever": 0.23710577392038215,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.dbpedia_entity": 0.32334770186384726,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.fever": 0.6341234371233305,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.fiqa": 0.44598842357063445,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.hotpotqa": 0.6704826316013306,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.nfcorpus": 0.35000164335019307,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.nq": 0.40244906153552973,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.quora": 0.822792613681568,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.scidocs": 0.24785649115277483,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.scifact": 0.7593341539154295,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/nano_beir.webis_touche2020": 0.28737508267615175,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3547880690737832,
+      "eval/avg": 0.47120147567813553,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3626373626373627,
+      "grad_norm": 4.625,
+      "learning_rate": 4.100837805544279e-05,
+      "loss": 0.5136,
+      "step": 3010
+    },
+    {
+      "epoch": 2.370486656200942,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.0665224844564053e-05,
+      "loss": 0.5163,
+      "step": 3020
+    },
+    {
+      "epoch": 2.378335949764521,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.032252659745699e-05,
+      "loss": 0.4874,
+      "step": 3030
+    },
+    {
+      "epoch": 2.3861852433281006,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.9980300016750694e-05,
+      "loss": 0.5288,
+      "step": 3040
+    },
+    {
+      "epoch": 2.39403453689168,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.963856178208588e-05,
+      "loss": 0.5765,
+      "step": 3050
+    },
+    {
+      "epoch": 2.401883830455259,
+      "grad_norm": 5.75,
+      "learning_rate": 3.9297328549302e-05,
+      "loss": 0.5288,
+      "step": 3060
+    },
+    {
+      "epoch": 2.4097331240188384,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.895661694962542e-05,
+      "loss": 0.6263,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4175824175824174,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.8616443588858805e-05,
+      "loss": 0.5104,
+      "step": 3080
+    },
+    {
+      "epoch": 2.425431711145997,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.827682504657187e-05,
+      "loss": 0.5545,
+      "step": 3090
+    },
+    {
+      "epoch": 2.4332810047095763,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.7937777875293244e-05,
+      "loss": 0.4756,
+      "step": 3100
+    },
+    {
+      "epoch": 2.4411302982731553,
+      "grad_norm": 4.5625,
+      "learning_rate": 3.759931859970374e-05,
+      "loss": 0.568,
+      "step": 3110
+    },
+    {
+      "epoch": 2.4489795918367347,
+      "grad_norm": 5.1875,
+      "learning_rate": 3.72614637158309e-05,
+      "loss": 0.4515,
+      "step": 3120
+    },
+    {
+      "epoch": 2.4568288854003137,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.692422969024516e-05,
+      "loss": 0.4946,
+      "step": 3130
+    },
+    {
+      "epoch": 2.464678178963893,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6587632959257165e-05,
+      "loss": 0.553,
+      "step": 3140
+    },
+    {
+      "epoch": 2.4725274725274726,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.625168992811671e-05,
+      "loss": 0.4509,
+      "step": 3150
+    },
+    {
+      "epoch": 2.4803767660910516,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.591641697021317e-05,
+      "loss": 0.4393,
+      "step": 3160
+    },
+    {
+      "epoch": 2.488226059654631,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.558183042627755e-05,
+      "loss": 0.5044,
+      "step": 3170
+    },
+    {
+      "epoch": 2.4960753532182105,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.524794660358593e-05,
+      "loss": 0.4536,
+      "step": 3180
+    },
+    {
+      "epoch": 2.5039246467817895,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.491478177516484e-05,
+      "loss": 0.5021,
+      "step": 3190
+    },
+    {
+      "epoch": 2.511773940345369,
+      "grad_norm": 4.0,
+      "learning_rate": 3.4582352178997935e-05,
+      "loss": 0.5415,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5196232339089484,
+      "grad_norm": 6.125,
+      "learning_rate": 3.425067401723477e-05,
+      "loss": 0.542,
+      "step": 3210
+    },
+    {
+      "epoch": 2.5274725274725274,
+      "grad_norm": 5.3125,
+      "learning_rate": 3.3919763455401015e-05,
+      "loss": 0.5314,
+      "step": 3220
+    },
+    {
+      "epoch": 2.535321821036107,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.358963662161062e-05,
+      "loss": 0.4573,
+      "step": 3230
+    },
+    {
+      "epoch": 2.5431711145996863,
+      "grad_norm": 4.9375,
+      "learning_rate": 3.326030960577972e-05,
+      "loss": 0.5061,
+      "step": 3240
+    },
+    {
+      "epoch": 2.5510204081632653,
+      "grad_norm": 3.25,
+      "learning_rate": 3.293179845884245e-05,
+      "loss": 0.4469,
+      "step": 3250
+    },
+    {
+      "epoch": 2.5588697017268447,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.2604119191968654e-05,
+      "loss": 0.4709,
+      "step": 3260
+    },
+    {
+      "epoch": 2.566718995290424,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.2277287775783525e-05,
+      "loss": 0.5695,
+      "step": 3270
+    },
+    {
+      "epoch": 2.574568288854003,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.195132013958918e-05,
+      "loss": 0.5323,
+      "step": 3280
+    },
+    {
+      "epoch": 2.5824175824175826,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.162623217058834e-05,
+      "loss": 0.5312,
+      "step": 3290
+    },
+    {
+      "epoch": 2.5902668759811616,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.130203971310999e-05,
+      "loss": 0.5346,
+      "step": 3300
+    },
+    {
+      "epoch": 2.598116169544741,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.097875856783713e-05,
+      "loss": 0.4832,
+      "step": 3310
+    },
+    {
+      "epoch": 2.60596546310832,
+      "grad_norm": 4.6875,
+      "learning_rate": 3.065640449103669e-05,
+      "loss": 0.5371,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6138147566718994,
+      "grad_norm": 4.84375,
+      "learning_rate": 3.033499319379163e-05,
+      "loss": 0.4707,
+      "step": 3330
+    },
+    {
+      "epoch": 2.621664050235479,
+      "grad_norm": 5.125,
+      "learning_rate": 3.001454034123512e-05,
+      "loss": 0.5356,
+      "step": 3340
+    },
+    {
+      "epoch": 2.629513343799058,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.9695061551787108e-05,
+      "loss": 0.5273,
+      "step": 3350
+    },
+    {
+      "epoch": 2.6373626373626373,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.9376572396393048e-05,
+      "loss": 0.5746,
+      "step": 3360
+    },
+    {
+      "epoch": 2.6452119309262168,
+      "grad_norm": 5.34375,
+      "learning_rate": 2.9059088397765088e-05,
+      "loss": 0.5145,
+      "step": 3370
+    },
+    {
+      "epoch": 2.6530612244897958,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.874262502962537e-05,
+      "loss": 0.4906,
+      "step": 3380
+    },
+    {
+      "epoch": 2.660910518053375,
+      "grad_norm": 4.625,
+      "learning_rate": 2.8427197715952047e-05,
+      "loss": 0.4909,
+      "step": 3390
+    },
+    {
+      "epoch": 2.6687598116169546,
+      "grad_norm": 5.28125,
+      "learning_rate": 2.811282183022736e-05,
+      "loss": 0.535,
+      "step": 3400
+    },
+    {
+      "epoch": 2.6766091051805336,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.7799512694688468e-05,
+      "loss": 0.524,
+      "step": 3410
+    },
+    {
+      "epoch": 2.684458398744113,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.7487285579580637e-05,
+      "loss": 0.4084,
+      "step": 3420
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 5.0,
+      "learning_rate": 2.7176155702412937e-05,
+      "loss": 0.5571,
+      "step": 3430
+    },
+    {
+      "epoch": 2.7001569858712715,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.686613822721666e-05,
+      "loss": 0.4846,
+      "step": 3440
+    },
+    {
+      "epoch": 2.708006279434851,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.6557248263806174e-05,
+      "loss": 0.5363,
+      "step": 3450
+    },
+    {
+      "epoch": 2.7158555729984304,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.6249500867042522e-05,
+      "loss": 0.5401,
+      "step": 3460
+    },
+    {
+      "epoch": 2.7237048665620094,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.5942911036099655e-05,
+      "loss": 0.5538,
+      "step": 3470
+    },
+    {
+      "epoch": 2.731554160125589,
+      "grad_norm": 4.75,
+      "learning_rate": 2.5637493713733374e-05,
+      "loss": 0.5288,
+      "step": 3480
+    },
+    {
+      "epoch": 2.739403453689168,
+      "grad_norm": 5.0625,
+      "learning_rate": 2.533326378555314e-05,
+      "loss": 0.5771,
+      "step": 3490
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.5030236079296444e-05,
+      "loss": 0.478,
+      "step": 3500
+    },
+    {
+      "epoch": 2.7551020408163263,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.4728425364106135e-05,
+      "loss": 0.4903,
+      "step": 3510
+    },
+    {
+      "epoch": 2.7629513343799057,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.442784634981071e-05,
+      "loss": 0.4784,
+      "step": 3520
+    },
+    {
+      "epoch": 2.770800627943485,
+      "grad_norm": 4.84375,
+      "learning_rate": 2.412851368620726e-05,
+      "loss": 0.5052,
+      "step": 3530
+    },
+    {
+      "epoch": 2.778649921507064,
+      "grad_norm": 5.0,
+      "learning_rate": 2.383044196234753e-05,
+      "loss": 0.5316,
+      "step": 3540
+    },
+    {
+      "epoch": 2.7864992150706436,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.3533645705826806e-05,
+      "loss": 0.4896,
+      "step": 3550
+    },
+    {
+      "epoch": 2.794348508634223,
+      "grad_norm": 5.53125,
+      "learning_rate": 2.323813938207593e-05,
+      "loss": 0.4697,
+      "step": 3560
+    },
+    {
+      "epoch": 2.802197802197802,
+      "grad_norm": 5.125,
+      "learning_rate": 2.294393739365621e-05,
+      "loss": 0.5038,
+      "step": 3570
+    },
+    {
+      "epoch": 2.8100470957613815,
+      "grad_norm": 5.21875,
+      "learning_rate": 2.265105407955752e-05,
+      "loss": 0.5708,
+      "step": 3580
+    },
+    {
+      "epoch": 2.817896389324961,
+      "grad_norm": 5.34375,
+      "learning_rate": 2.235950371449938e-05,
+      "loss": 0.5062,
+      "step": 3590
+    },
+    {
+      "epoch": 2.82574568288854,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.2069300508235275e-05,
+      "loss": 0.5216,
+      "step": 3600
+    },
+    {
+      "epoch": 2.8335949764521193,
+      "grad_norm": 4.25,
+      "learning_rate": 2.1780458604860056e-05,
+      "loss": 0.4725,
+      "step": 3610
+    },
+    {
+      "epoch": 2.8414442700156988,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.14929920821206e-05,
+      "loss": 0.5209,
+      "step": 3620
+    },
+    {
+      "epoch": 2.8492935635792778,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.1206914950729672e-05,
+      "loss": 0.4435,
+      "step": 3630
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 5.09375,
+      "learning_rate": 2.0922241153683066e-05,
+      "loss": 0.5311,
+      "step": 3640
+    },
+    {
+      "epoch": 2.8649921507064366,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.0638984565580023e-05,
+      "loss": 0.5394,
+      "step": 3650
+    },
+    {
+      "epoch": 2.8728414442700156,
+      "grad_norm": 5.09375,
+      "learning_rate": 2.035715899194704e-05,
+      "loss": 0.496,
+      "step": 3660
+    },
+    {
+      "epoch": 2.880690737833595,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.007677816856498e-05,
+      "loss": 0.5404,
+      "step": 3670
+    },
+    {
+      "epoch": 2.8885400313971745,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.9797855760799612e-05,
+      "loss": 0.4908,
+      "step": 3680
+    },
+    {
+      "epoch": 2.8963893249607535,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.9520405362935594e-05,
+      "loss": 0.4916,
+      "step": 3690
+    },
+    {
+      "epoch": 2.904238618524333,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.9244440497513893e-05,
+      "loss": 0.4959,
+      "step": 3700
+    },
+    {
+      "epoch": 2.912087912087912,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.896997461467272e-05,
+      "loss": 0.5591,
+      "step": 3710
+    },
+    {
+      "epoch": 2.9199372056514914,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.869702109149199e-05,
+      "loss": 0.4443,
+      "step": 3720
+    },
+    {
+      "epoch": 2.9277864992150704,
+      "grad_norm": 5.9375,
+      "learning_rate": 1.842559323134136e-05,
+      "loss": 0.5501,
+      "step": 3730
+    },
+    {
+      "epoch": 2.93563579277865,
+      "grad_norm": 5.625,
+      "learning_rate": 1.8155704263231776e-05,
+      "loss": 0.5145,
+      "step": 3740
+    },
+    {
+      "epoch": 2.9434850863422293,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.788736734117078e-05,
+      "loss": 0.6076,
+      "step": 3750
+    },
+    {
+      "epoch": 2.9513343799058083,
+      "grad_norm": 4.78125,
+      "learning_rate": 1.762059554352143e-05,
+      "loss": 0.4832,
+      "step": 3760
+    },
+    {
+      "epoch": 2.9591836734693877,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.7355401872364758e-05,
+      "loss": 0.5021,
+      "step": 3770
+    },
+    {
+      "epoch": 2.967032967032967,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.709179925286617e-05,
+      "loss": 0.616,
+      "step": 3780
+    },
+    {
+      "epoch": 2.974882260596546,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.6829800532645447e-05,
+      "loss": 0.4569,
+      "step": 3790
+    },
+    {
+      "epoch": 2.9827315541601256,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.6569418481150595e-05,
+      "loss": 0.464,
+      "step": 3800
+    },
+    {
+      "epoch": 2.990580847723705,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.6310665789035467e-05,
+      "loss": 0.5054,
+      "step": 3810
+    },
+    {
+      "epoch": 2.998430141287284,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.605355506754121e-05,
+      "loss": 0.4882,
+      "step": 3820
+    },
+    {
+      "epoch": 3.0062794348508635,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.5798098847881666e-05,
+      "loss": 0.4937,
+      "step": 3830
+    },
+    {
+      "epoch": 3.014128728414443,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.554430958063259e-05,
+      "loss": 0.4828,
+      "step": 3840
+    },
+    {
+      "epoch": 3.021978021978022,
+      "grad_norm": 5.65625,
+      "learning_rate": 1.529219963512481e-05,
+      "loss": 0.4813,
+      "step": 3850
+    },
+    {
+      "epoch": 3.0298273155416013,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.5041781298841423e-05,
+      "loss": 0.4902,
+      "step": 3860
+    },
+    {
+      "epoch": 3.0376766091051803,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.4793066776818842e-05,
+      "loss": 0.5018,
+      "step": 3870
+    },
+    {
+      "epoch": 3.0455259026687598,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.4546068191051987e-05,
+      "loss": 0.5047,
+      "step": 3880
+    },
+    {
+      "epoch": 3.053375196232339,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.4300797579903474e-05,
+      "loss": 0.5615,
+      "step": 3890
+    },
+    {
+      "epoch": 3.061224489795918,
+      "grad_norm": 4.125,
+      "learning_rate": 1.4057266897516841e-05,
+      "loss": 0.4684,
+      "step": 3900
+    },
+    {
+      "epoch": 3.0690737833594977,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.3815488013233984e-05,
+      "loss": 0.5193,
+      "step": 3910
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.3575472711016635e-05,
+      "loss": 0.4829,
+      "step": 3920
+    },
+    {
+      "epoch": 3.084772370486656,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.3337232688872009e-05,
+      "loss": 0.5745,
+      "step": 3930
+    },
+    {
+      "epoch": 3.0926216640502355,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3100779558282672e-05,
+      "loss": 0.4937,
+      "step": 3940
+    },
+    {
+      "epoch": 3.100470957613815,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.2866124843640615e-05,
+      "loss": 0.5324,
+      "step": 3950
+    },
+    {
+      "epoch": 3.108320251177394,
+      "grad_norm": 5.9375,
+      "learning_rate": 1.2633279981685608e-05,
+      "loss": 0.5407,
+      "step": 3960
+    },
+    {
+      "epoch": 3.1161695447409734,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.240225632094773e-05,
+      "loss": 0.574,
+      "step": 3970
+    },
+    {
+      "epoch": 3.1240188383045524,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.217306512119425e-05,
+      "loss": 0.5532,
+      "step": 3980
+    },
+    {
+      "epoch": 3.131868131868132,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.194571755288092e-05,
+      "loss": 0.4584,
+      "step": 3990
+    },
+    {
+      "epoch": 3.1397174254317113,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.1720224696607474e-05,
+      "loss": 0.5262,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1000,
+  "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": 16,
+  "trial_name": null,
+  "trial_params": null
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/checkpoint-4000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9365177436ad70bb9d6000d96d3ccdc5c5ca38536defdcfec6b548401edbc07f
+size 6481

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/config.json

@@ -0,0 +1,45 @@
+{
+  "architectures": [
+    "ModernBertModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "repad_logits_with_grad": false,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.53.0",
+  "vocab_size": 50368
+}

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:461d9cd6f69078fd798ee95a1eb88776ce5adc5748c48ce1477e6be2d2377af0
+size 298041696

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab613bcea54280917982d51c432e28d3ebba6d4e6765fa600dfbc4db6e1f64d5
+size 596170443

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24c12cceaffe8acec04189ba71dbb81f100dd5cd03b86390253117b3fab2911d
+size 15365

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69adbd9997461473344beb9c44d2e496e24fbc4d6fe69245ab0bd127882efd96
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:baaf7026d9e588ca2ba5b4de8768a379982b7530f39e0b88fa44af9bef8e8bc9
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ec81cbdf599c90b49fb13998f6e46e2492b55345216231b6ce078f88cf04eae
+size 15429

modernbert-two-stage-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.crux-researchy.request/runs/Jan01_23-09-47_nid005111/events.out.tfevents.1767301973.nid005111.95408.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a1266b963e5671de48195e4a46d80d1eb6a4537b58b81ada0e638328d7bf13b
+size 119511