update the one trained with request

.gitattributes

@@ -350,3 +350,73 @@ modernbert-crux-researchy-flatten.b64_n512.1e-4/optimizer.pt filter=lfs diff=lfs
 .git/lfs/objects/3d/4a/3d4a3a4394bc050534ccc08756a61ea984b16584719563b3ae55a8471b6fa937 filter=lfs diff=lfs merge=lfs -text
 .git/lfs/objects/fd/82/fd82dce9a1914d1427efa2b58c72a13ee36999640c614adaa63b2f789fe49c72 filter=lfs diff=lfs merge=lfs -text
 .git/lfs/objects/92/0f/920f334aa8a2a64cf006d7a05ba27b3b53bb416a0f30df954d0ba598a097fa3b filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_zero.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_half.neg_zero.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_zero.neg_high.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_low.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_high.neg_quarter.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-2000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-3000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-1000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-4000/model.safetensors filter=lfs diff=lfs merge=lfs -text
+modernbert-crux-researchy-pos_low.neg_zero.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt filter=lfs diff=lfs merge=lfs -text

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca7454cf27ca9f5516aa6fff961b75093665ac6e1dc9b38172ef9865db0763e2
+size 298041696

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47e732067fbfe03b9b9d1db6e092429169ba6dee3aa99b55786815e28b2ed522
+size 596170443

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/scheduler.pt

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/trainer_state.json

@@ -0,0 +1,799 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7132667617689016,
+  "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": 0.5201586889046041,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.23997409265084552,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.34286887314902237,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6213496755517833,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4542903465876373,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6560051657630788,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3328835143785305,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.4570725655218753,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.832098944676056,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.25908236715595256,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7391066781087818,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.30431141846717047,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.4799335275762782,
+      "step": 0
+    },
+    {
+      "epoch": 0.007132667617689016,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8e-06,
+      "loss": 1.6901,
+      "step": 10
+    },
+    {
+      "epoch": 0.014265335235378032,
+      "grad_norm": 10.75,
+      "learning_rate": 3.8e-06,
+      "loss": 1.7675,
+      "step": 20
+    },
+    {
+      "epoch": 0.021398002853067047,
+      "grad_norm": 21.0,
+      "learning_rate": 5.8e-06,
+      "loss": 1.527,
+      "step": 30
+    },
+    {
+      "epoch": 0.028530670470756064,
+      "grad_norm": 19.0,
+      "learning_rate": 7.8e-06,
+      "loss": 1.7708,
+      "step": 40
+    },
+    {
+      "epoch": 0.03566333808844508,
+      "grad_norm": 17.125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.4093,
+      "step": 50
+    },
+    {
+      "epoch": 0.042796005706134094,
+      "grad_norm": 15.625,
+      "learning_rate": 1.18e-05,
+      "loss": 1.579,
+      "step": 60
+    },
+    {
+      "epoch": 0.04992867332382311,
+      "grad_norm": 14.0,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.4845,
+      "step": 70
+    },
+    {
+      "epoch": 0.05706134094151213,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.58e-05,
+      "loss": 1.598,
+      "step": 80
+    },
+    {
+      "epoch": 0.06419400855920114,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.78e-05,
+      "loss": 1.3226,
+      "step": 90
+    },
+    {
+      "epoch": 0.07132667617689016,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.4969,
+      "step": 100
+    },
+    {
+      "epoch": 0.07845934379457917,
+      "grad_norm": 13.125,
+      "learning_rate": 2.18e-05,
+      "loss": 1.3867,
+      "step": 110
+    },
+    {
+      "epoch": 0.08559201141226819,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.38e-05,
+      "loss": 1.1904,
+      "step": 120
+    },
+    {
+      "epoch": 0.09272467902995721,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.58e-05,
+      "loss": 1.1668,
+      "step": 130
+    },
+    {
+      "epoch": 0.09985734664764621,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.124,
+      "step": 140
+    },
+    {
+      "epoch": 0.10699001426533523,
+      "grad_norm": 9.75,
+      "learning_rate": 2.98e-05,
+      "loss": 1.1571,
+      "step": 150
+    },
+    {
+      "epoch": 0.11412268188302425,
+      "grad_norm": 12.25,
+      "learning_rate": 3.18e-05,
+      "loss": 1.2145,
+      "step": 160
+    },
+    {
+      "epoch": 0.12125534950071326,
+      "grad_norm": 10.625,
+      "learning_rate": 3.38e-05,
+      "loss": 1.155,
+      "step": 170
+    },
+    {
+      "epoch": 0.12838801711840228,
+      "grad_norm": 8.5,
+      "learning_rate": 3.58e-05,
+      "loss": 1.1513,
+      "step": 180
+    },
+    {
+      "epoch": 0.1355206847360913,
+      "grad_norm": 9.5,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.9803,
+      "step": 190
+    },
+    {
+      "epoch": 0.14265335235378032,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 1.0302,
+      "step": 200
+    },
+    {
+      "epoch": 0.14978601997146934,
+      "grad_norm": 9.25,
+      "learning_rate": 4.18e-05,
+      "loss": 0.9878,
+      "step": 210
+    },
+    {
+      "epoch": 0.15691868758915833,
+      "grad_norm": 10.25,
+      "learning_rate": 4.38e-05,
+      "loss": 0.9816,
+      "step": 220
+    },
+    {
+      "epoch": 0.16405135520684735,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.58e-05,
+      "loss": 1.0415,
+      "step": 230
+    },
+    {
+      "epoch": 0.17118402282453637,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.78e-05,
+      "loss": 1.1126,
+      "step": 240
+    },
+    {
+      "epoch": 0.1783166904422254,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 1.0127,
+      "step": 250
+    },
+    {
+      "epoch": 0.18544935805991442,
+      "grad_norm": 5.71875,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.9577,
+      "step": 260
+    },
+    {
+      "epoch": 0.19258202567760344,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.9518,
+      "step": 270
+    },
+    {
+      "epoch": 0.19971469329529243,
+      "grad_norm": 6.28125,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.7952,
+      "step": 280
+    },
+    {
+      "epoch": 0.20684736091298145,
+      "grad_norm": 6.875,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.9201,
+      "step": 290
+    },
+    {
+      "epoch": 0.21398002853067047,
+      "grad_norm": 6.09375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.951,
+      "step": 300
+    },
+    {
+      "epoch": 0.2211126961483595,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.18e-05,
+      "loss": 0.9199,
+      "step": 310
+    },
+    {
+      "epoch": 0.2282453637660485,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.38e-05,
+      "loss": 0.8773,
+      "step": 320
+    },
+    {
+      "epoch": 0.23537803138373753,
+      "grad_norm": 4.875,
+      "learning_rate": 6.58e-05,
+      "loss": 0.9386,
+      "step": 330
+    },
+    {
+      "epoch": 0.24251069900142652,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.9364,
+      "step": 340
+    },
+    {
+      "epoch": 0.24964336661911554,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.98e-05,
+      "loss": 0.9053,
+      "step": 350
+    },
+    {
+      "epoch": 0.25677603423680456,
+      "grad_norm": 5.8125,
+      "learning_rate": 7.18e-05,
+      "loss": 0.8493,
+      "step": 360
+    },
+    {
+      "epoch": 0.26390870185449355,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.38e-05,
+      "loss": 0.8218,
+      "step": 370
+    },
+    {
+      "epoch": 0.2710413694721826,
+      "grad_norm": 5.625,
+      "learning_rate": 7.58e-05,
+      "loss": 0.8975,
+      "step": 380
+    },
+    {
+      "epoch": 0.2781740370898716,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.8891,
+      "step": 390
+    },
+    {
+      "epoch": 0.28530670470756064,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.7831,
+      "step": 400
+    },
+    {
+      "epoch": 0.29243937232524964,
+      "grad_norm": 5.125,
+      "learning_rate": 8.18e-05,
+      "loss": 0.9017,
+      "step": 410
+    },
+    {
+      "epoch": 0.2995720399429387,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.38e-05,
+      "loss": 0.7839,
+      "step": 420
+    },
+    {
+      "epoch": 0.3067047075606277,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.886,
+      "step": 430
+    },
+    {
+      "epoch": 0.31383737517831667,
+      "grad_norm": 6.1875,
+      "learning_rate": 8.78e-05,
+      "loss": 0.863,
+      "step": 440
+    },
+    {
+      "epoch": 0.3209700427960057,
+      "grad_norm": 4.625,
+      "learning_rate": 8.98e-05,
+      "loss": 0.884,
+      "step": 450
+    },
+    {
+      "epoch": 0.3281027104136947,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.9209,
+      "step": 460
+    },
+    {
+      "epoch": 0.33523537803138376,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.8775,
+      "step": 470
+    },
+    {
+      "epoch": 0.34236804564907275,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.58e-05,
+      "loss": 0.7956,
+      "step": 480
+    },
+    {
+      "epoch": 0.34950071326676174,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.8632,
+      "step": 490
+    },
+    {
+      "epoch": 0.3566333808844508,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.98e-05,
+      "loss": 0.8308,
+      "step": 500
+    },
+    {
+      "epoch": 0.3637660485021398,
+      "grad_norm": 5.25,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.884,
+      "step": 510
+    },
+    {
+      "epoch": 0.37089871611982883,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.8663,
+      "step": 520
+    },
+    {
+      "epoch": 0.3780313837375178,
+      "grad_norm": 5.25,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.8722,
+      "step": 530
+    },
+    {
+      "epoch": 0.38516405135520687,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.8391,
+      "step": 540
+    },
+    {
+      "epoch": 0.39229671897289586,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.7823,
+      "step": 550
+    },
+    {
+      "epoch": 0.39942938659058486,
+      "grad_norm": 4.875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.8606,
+      "step": 560
+    },
+    {
+      "epoch": 0.4065620542082739,
+      "grad_norm": 5.625,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.8719,
+      "step": 570
+    },
+    {
+      "epoch": 0.4136947218259629,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.8619,
+      "step": 580
+    },
+    {
+      "epoch": 0.42082738944365194,
+      "grad_norm": 5.0,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.8439,
+      "step": 590
+    },
+    {
+      "epoch": 0.42796005706134094,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.886,
+      "step": 600
+    },
+    {
+      "epoch": 0.43509272467902993,
+      "grad_norm": 5.125,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.8346,
+      "step": 610
+    },
+    {
+      "epoch": 0.442225392296719,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.8145,
+      "step": 620
+    },
+    {
+      "epoch": 0.44935805991440797,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.8739,
+      "step": 630
+    },
+    {
+      "epoch": 0.456490727532097,
+      "grad_norm": 4.75,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.7826,
+      "step": 640
+    },
+    {
+      "epoch": 0.463623395149786,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.8541,
+      "step": 650
+    },
+    {
+      "epoch": 0.47075606276747506,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.8208,
+      "step": 660
+    },
+    {
+      "epoch": 0.47788873038516405,
+      "grad_norm": 5.375,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.7761,
+      "step": 670
+    },
+    {
+      "epoch": 0.48502139800285304,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.8346,
+      "step": 680
+    },
+    {
+      "epoch": 0.4921540656205421,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.9192,
+      "step": 690
+    },
+    {
+      "epoch": 0.4992867332382311,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.8037,
+      "step": 700
+    },
+    {
+      "epoch": 0.5064194008559201,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.7669,
+      "step": 710
+    },
+    {
+      "epoch": 0.5135520684736091,
+      "grad_norm": 6.875,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.7915,
+      "step": 720
+    },
+    {
+      "epoch": 0.5206847360912982,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.7863,
+      "step": 730
+    },
+    {
+      "epoch": 0.5278174037089871,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.7789,
+      "step": 740
+    },
+    {
+      "epoch": 0.5349500713266762,
+      "grad_norm": 4.0,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.7759,
+      "step": 750
+    },
+    {
+      "epoch": 0.5420827389443652,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.8504,
+      "step": 760
+    },
+    {
+      "epoch": 0.5492154065620543,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.809,
+      "step": 770
+    },
+    {
+      "epoch": 0.5563480741797432,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.7721,
+      "step": 780
+    },
+    {
+      "epoch": 0.5634807417974322,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.7319,
+      "step": 790
+    },
+    {
+      "epoch": 0.5706134094151213,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.7726,
+      "step": 800
+    },
+    {
+      "epoch": 0.5777460770328102,
+      "grad_norm": 4.625,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.8522,
+      "step": 810
+    },
+    {
+      "epoch": 0.5848787446504993,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.7761,
+      "step": 820
+    },
+    {
+      "epoch": 0.5920114122681883,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.8702,
+      "step": 830
+    },
+    {
+      "epoch": 0.5991440798858774,
+      "grad_norm": 5.25,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.8636,
+      "step": 840
+    },
+    {
+      "epoch": 0.6062767475035663,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.6974,
+      "step": 850
+    },
+    {
+      "epoch": 0.6134094151212554,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.8017,
+      "step": 860
+    },
+    {
+      "epoch": 0.6205420827389444,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.7799,
+      "step": 870
+    },
+    {
+      "epoch": 0.6276747503566333,
+      "grad_norm": 5.125,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.7808,
+      "step": 880
+    },
+    {
+      "epoch": 0.6348074179743224,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.8337,
+      "step": 890
+    },
+    {
+      "epoch": 0.6419400855920114,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.8043,
+      "step": 900
+    },
+    {
+      "epoch": 0.6490727532097005,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.8328,
+      "step": 910
+    },
+    {
+      "epoch": 0.6562054208273894,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.7745,
+      "step": 920
+    },
+    {
+      "epoch": 0.6633380884450785,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.745,
+      "step": 930
+    },
+    {
+      "epoch": 0.6704707560627675,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.82,
+      "step": 940
+    },
+    {
+      "epoch": 0.6776034236804565,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.8025,
+      "step": 950
+    },
+    {
+      "epoch": 0.6847360912981455,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.84,
+      "step": 960
+    },
+    {
+      "epoch": 0.6918687589158345,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.8552,
+      "step": 970
+    },
+    {
+      "epoch": 0.6990014265335235,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.8082,
+      "step": 980
+    },
+    {
+      "epoch": 0.7061340941512125,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.8265,
+      "step": 990
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "grad_norm": 4.875,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.8369,
+      "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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-1000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af7cf7c3c741add5e8996639d61d87fe74bb40ad0faa1b29901130f4b3a60fc3
+size 6353

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:096d65b355d6e0613adaabb2f411c1c64396ffbf43bdd18442af55813a089b23
+size 298041696

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:944f7393a46576d6228b50f137c797dcfc63520740b5d5942abe6fffc0c37b37
+size 596170443

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/scheduler.pt

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/trainer_state.json

@@ -0,0 +1,1564 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.4265335235378032,
+  "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": 0.5201586889046041,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.23997409265084552,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.34286887314902237,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6213496755517833,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4542903465876373,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6560051657630788,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3328835143785305,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.4570725655218753,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.832098944676056,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.25908236715595256,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7391066781087818,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.30431141846717047,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.4799335275762782,
+      "step": 0
+    },
+    {
+      "epoch": 0.007132667617689016,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8e-06,
+      "loss": 1.6901,
+      "step": 10
+    },
+    {
+      "epoch": 0.014265335235378032,
+      "grad_norm": 10.75,
+      "learning_rate": 3.8e-06,
+      "loss": 1.7675,
+      "step": 20
+    },
+    {
+      "epoch": 0.021398002853067047,
+      "grad_norm": 21.0,
+      "learning_rate": 5.8e-06,
+      "loss": 1.527,
+      "step": 30
+    },
+    {
+      "epoch": 0.028530670470756064,
+      "grad_norm": 19.0,
+      "learning_rate": 7.8e-06,
+      "loss": 1.7708,
+      "step": 40
+    },
+    {
+      "epoch": 0.03566333808844508,
+      "grad_norm": 17.125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.4093,
+      "step": 50
+    },
+    {
+      "epoch": 0.042796005706134094,
+      "grad_norm": 15.625,
+      "learning_rate": 1.18e-05,
+      "loss": 1.579,
+      "step": 60
+    },
+    {
+      "epoch": 0.04992867332382311,
+      "grad_norm": 14.0,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.4845,
+      "step": 70
+    },
+    {
+      "epoch": 0.05706134094151213,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.58e-05,
+      "loss": 1.598,
+      "step": 80
+    },
+    {
+      "epoch": 0.06419400855920114,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.78e-05,
+      "loss": 1.3226,
+      "step": 90
+    },
+    {
+      "epoch": 0.07132667617689016,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.4969,
+      "step": 100
+    },
+    {
+      "epoch": 0.07845934379457917,
+      "grad_norm": 13.125,
+      "learning_rate": 2.18e-05,
+      "loss": 1.3867,
+      "step": 110
+    },
+    {
+      "epoch": 0.08559201141226819,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.38e-05,
+      "loss": 1.1904,
+      "step": 120
+    },
+    {
+      "epoch": 0.09272467902995721,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.58e-05,
+      "loss": 1.1668,
+      "step": 130
+    },
+    {
+      "epoch": 0.09985734664764621,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.124,
+      "step": 140
+    },
+    {
+      "epoch": 0.10699001426533523,
+      "grad_norm": 9.75,
+      "learning_rate": 2.98e-05,
+      "loss": 1.1571,
+      "step": 150
+    },
+    {
+      "epoch": 0.11412268188302425,
+      "grad_norm": 12.25,
+      "learning_rate": 3.18e-05,
+      "loss": 1.2145,
+      "step": 160
+    },
+    {
+      "epoch": 0.12125534950071326,
+      "grad_norm": 10.625,
+      "learning_rate": 3.38e-05,
+      "loss": 1.155,
+      "step": 170
+    },
+    {
+      "epoch": 0.12838801711840228,
+      "grad_norm": 8.5,
+      "learning_rate": 3.58e-05,
+      "loss": 1.1513,
+      "step": 180
+    },
+    {
+      "epoch": 0.1355206847360913,
+      "grad_norm": 9.5,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.9803,
+      "step": 190
+    },
+    {
+      "epoch": 0.14265335235378032,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 1.0302,
+      "step": 200
+    },
+    {
+      "epoch": 0.14978601997146934,
+      "grad_norm": 9.25,
+      "learning_rate": 4.18e-05,
+      "loss": 0.9878,
+      "step": 210
+    },
+    {
+      "epoch": 0.15691868758915833,
+      "grad_norm": 10.25,
+      "learning_rate": 4.38e-05,
+      "loss": 0.9816,
+      "step": 220
+    },
+    {
+      "epoch": 0.16405135520684735,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.58e-05,
+      "loss": 1.0415,
+      "step": 230
+    },
+    {
+      "epoch": 0.17118402282453637,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.78e-05,
+      "loss": 1.1126,
+      "step": 240
+    },
+    {
+      "epoch": 0.1783166904422254,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 1.0127,
+      "step": 250
+    },
+    {
+      "epoch": 0.18544935805991442,
+      "grad_norm": 5.71875,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.9577,
+      "step": 260
+    },
+    {
+      "epoch": 0.19258202567760344,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.9518,
+      "step": 270
+    },
+    {
+      "epoch": 0.19971469329529243,
+      "grad_norm": 6.28125,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.7952,
+      "step": 280
+    },
+    {
+      "epoch": 0.20684736091298145,
+      "grad_norm": 6.875,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.9201,
+      "step": 290
+    },
+    {
+      "epoch": 0.21398002853067047,
+      "grad_norm": 6.09375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.951,
+      "step": 300
+    },
+    {
+      "epoch": 0.2211126961483595,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.18e-05,
+      "loss": 0.9199,
+      "step": 310
+    },
+    {
+      "epoch": 0.2282453637660485,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.38e-05,
+      "loss": 0.8773,
+      "step": 320
+    },
+    {
+      "epoch": 0.23537803138373753,
+      "grad_norm": 4.875,
+      "learning_rate": 6.58e-05,
+      "loss": 0.9386,
+      "step": 330
+    },
+    {
+      "epoch": 0.24251069900142652,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.9364,
+      "step": 340
+    },
+    {
+      "epoch": 0.24964336661911554,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.98e-05,
+      "loss": 0.9053,
+      "step": 350
+    },
+    {
+      "epoch": 0.25677603423680456,
+      "grad_norm": 5.8125,
+      "learning_rate": 7.18e-05,
+      "loss": 0.8493,
+      "step": 360
+    },
+    {
+      "epoch": 0.26390870185449355,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.38e-05,
+      "loss": 0.8218,
+      "step": 370
+    },
+    {
+      "epoch": 0.2710413694721826,
+      "grad_norm": 5.625,
+      "learning_rate": 7.58e-05,
+      "loss": 0.8975,
+      "step": 380
+    },
+    {
+      "epoch": 0.2781740370898716,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.8891,
+      "step": 390
+    },
+    {
+      "epoch": 0.28530670470756064,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.7831,
+      "step": 400
+    },
+    {
+      "epoch": 0.29243937232524964,
+      "grad_norm": 5.125,
+      "learning_rate": 8.18e-05,
+      "loss": 0.9017,
+      "step": 410
+    },
+    {
+      "epoch": 0.2995720399429387,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.38e-05,
+      "loss": 0.7839,
+      "step": 420
+    },
+    {
+      "epoch": 0.3067047075606277,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.886,
+      "step": 430
+    },
+    {
+      "epoch": 0.31383737517831667,
+      "grad_norm": 6.1875,
+      "learning_rate": 8.78e-05,
+      "loss": 0.863,
+      "step": 440
+    },
+    {
+      "epoch": 0.3209700427960057,
+      "grad_norm": 4.625,
+      "learning_rate": 8.98e-05,
+      "loss": 0.884,
+      "step": 450
+    },
+    {
+      "epoch": 0.3281027104136947,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.9209,
+      "step": 460
+    },
+    {
+      "epoch": 0.33523537803138376,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.8775,
+      "step": 470
+    },
+    {
+      "epoch": 0.34236804564907275,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.58e-05,
+      "loss": 0.7956,
+      "step": 480
+    },
+    {
+      "epoch": 0.34950071326676174,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.8632,
+      "step": 490
+    },
+    {
+      "epoch": 0.3566333808844508,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.98e-05,
+      "loss": 0.8308,
+      "step": 500
+    },
+    {
+      "epoch": 0.3637660485021398,
+      "grad_norm": 5.25,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.884,
+      "step": 510
+    },
+    {
+      "epoch": 0.37089871611982883,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.8663,
+      "step": 520
+    },
+    {
+      "epoch": 0.3780313837375178,
+      "grad_norm": 5.25,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.8722,
+      "step": 530
+    },
+    {
+      "epoch": 0.38516405135520687,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.8391,
+      "step": 540
+    },
+    {
+      "epoch": 0.39229671897289586,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.7823,
+      "step": 550
+    },
+    {
+      "epoch": 0.39942938659058486,
+      "grad_norm": 4.875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.8606,
+      "step": 560
+    },
+    {
+      "epoch": 0.4065620542082739,
+      "grad_norm": 5.625,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.8719,
+      "step": 570
+    },
+    {
+      "epoch": 0.4136947218259629,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.8619,
+      "step": 580
+    },
+    {
+      "epoch": 0.42082738944365194,
+      "grad_norm": 5.0,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.8439,
+      "step": 590
+    },
+    {
+      "epoch": 0.42796005706134094,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.886,
+      "step": 600
+    },
+    {
+      "epoch": 0.43509272467902993,
+      "grad_norm": 5.125,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.8346,
+      "step": 610
+    },
+    {
+      "epoch": 0.442225392296719,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.8145,
+      "step": 620
+    },
+    {
+      "epoch": 0.44935805991440797,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.8739,
+      "step": 630
+    },
+    {
+      "epoch": 0.456490727532097,
+      "grad_norm": 4.75,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.7826,
+      "step": 640
+    },
+    {
+      "epoch": 0.463623395149786,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.8541,
+      "step": 650
+    },
+    {
+      "epoch": 0.47075606276747506,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.8208,
+      "step": 660
+    },
+    {
+      "epoch": 0.47788873038516405,
+      "grad_norm": 5.375,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.7761,
+      "step": 670
+    },
+    {
+      "epoch": 0.48502139800285304,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.8346,
+      "step": 680
+    },
+    {
+      "epoch": 0.4921540656205421,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.9192,
+      "step": 690
+    },
+    {
+      "epoch": 0.4992867332382311,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.8037,
+      "step": 700
+    },
+    {
+      "epoch": 0.5064194008559201,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.7669,
+      "step": 710
+    },
+    {
+      "epoch": 0.5135520684736091,
+      "grad_norm": 6.875,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.7915,
+      "step": 720
+    },
+    {
+      "epoch": 0.5206847360912982,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.7863,
+      "step": 730
+    },
+    {
+      "epoch": 0.5278174037089871,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.7789,
+      "step": 740
+    },
+    {
+      "epoch": 0.5349500713266762,
+      "grad_norm": 4.0,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.7759,
+      "step": 750
+    },
+    {
+      "epoch": 0.5420827389443652,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.8504,
+      "step": 760
+    },
+    {
+      "epoch": 0.5492154065620543,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.809,
+      "step": 770
+    },
+    {
+      "epoch": 0.5563480741797432,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.7721,
+      "step": 780
+    },
+    {
+      "epoch": 0.5634807417974322,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.7319,
+      "step": 790
+    },
+    {
+      "epoch": 0.5706134094151213,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.7726,
+      "step": 800
+    },
+    {
+      "epoch": 0.5777460770328102,
+      "grad_norm": 4.625,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.8522,
+      "step": 810
+    },
+    {
+      "epoch": 0.5848787446504993,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.7761,
+      "step": 820
+    },
+    {
+      "epoch": 0.5920114122681883,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.8702,
+      "step": 830
+    },
+    {
+      "epoch": 0.5991440798858774,
+      "grad_norm": 5.25,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.8636,
+      "step": 840
+    },
+    {
+      "epoch": 0.6062767475035663,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.6974,
+      "step": 850
+    },
+    {
+      "epoch": 0.6134094151212554,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.8017,
+      "step": 860
+    },
+    {
+      "epoch": 0.6205420827389444,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.7799,
+      "step": 870
+    },
+    {
+      "epoch": 0.6276747503566333,
+      "grad_norm": 5.125,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.7808,
+      "step": 880
+    },
+    {
+      "epoch": 0.6348074179743224,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.8337,
+      "step": 890
+    },
+    {
+      "epoch": 0.6419400855920114,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.8043,
+      "step": 900
+    },
+    {
+      "epoch": 0.6490727532097005,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.8328,
+      "step": 910
+    },
+    {
+      "epoch": 0.6562054208273894,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.7745,
+      "step": 920
+    },
+    {
+      "epoch": 0.6633380884450785,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.745,
+      "step": 930
+    },
+    {
+      "epoch": 0.6704707560627675,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.82,
+      "step": 940
+    },
+    {
+      "epoch": 0.6776034236804565,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.8025,
+      "step": 950
+    },
+    {
+      "epoch": 0.6847360912981455,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.84,
+      "step": 960
+    },
+    {
+      "epoch": 0.6918687589158345,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.8552,
+      "step": 970
+    },
+    {
+      "epoch": 0.6990014265335235,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.8082,
+      "step": 980
+    },
+    {
+      "epoch": 0.7061340941512125,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.8265,
+      "step": 990
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "grad_norm": 4.875,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.8369,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.arguana": 0.47952715537701884,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.climate_fever": 0.22702207934204544,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.dbpedia_entity": 0.3287396364958694,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.fever": 0.5532628324042129,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.fiqa": 0.4160855038689836,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.hotpotqa": 0.6599716759215029,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.nfcorpus": 0.3449425020579428,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.nq": 0.39316532973343116,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.quora": 0.8295275476072758,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.scidocs": 0.2526832556333601,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.scifact": 0.7219633063518482,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.webis_touche2020": 0.2787587910246796,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/avg": 0.4571374679848476,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7203994293865906,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.68762555129222e-05,
+      "loss": 0.8872,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7275320970042796,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.675366799960841e-05,
+      "loss": 0.8334,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7346647646219686,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.662880177847594e-05,
+      "loss": 0.8014,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7417974322396577,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.65016629353282e-05,
+      "loss": 0.848,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7489300998573466,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.637225766673307e-05,
+      "loss": 0.7203,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7560627674750356,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.624059227972075e-05,
+      "loss": 0.8405,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7631954350927247,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.610667319147647e-05,
+      "loss": 0.8418,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7703281027104137,
+      "grad_norm": 5.75,
+      "learning_rate": 9.597050692902765e-05,
+      "loss": 0.8853,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7774607703281027,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.583210012892582e-05,
+      "loss": 0.7733,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7845934379457917,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.569145953692315e-05,
+      "loss": 0.7698,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7917261055634808,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.55485920076437e-05,
+      "loss": 0.7759,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7988587731811697,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.540350450424926e-05,
+      "loss": 0.8025,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8059914407988588,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.525620409810007e-05,
+      "loss": 0.8364,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8131241084165478,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.510669796841014e-05,
+      "loss": 0.7523,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8202567760342369,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.495499340189728e-05,
+      "loss": 0.7333,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8273894436519258,
+      "grad_norm": 4.875,
+      "learning_rate": 9.480109779242805e-05,
+      "loss": 0.7325,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8345221112696148,
+      "grad_norm": 5.0,
+      "learning_rate": 9.464501864065735e-05,
+      "loss": 0.7996,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8416547788873039,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.448676355366282e-05,
+      "loss": 0.7946,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8487874465049928,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.432634024457414e-05,
+      "loss": 0.9133,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8559201141226819,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.416375653219709e-05,
+      "loss": 0.7218,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8630527817403709,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.399902034063242e-05,
+      "loss": 0.8453,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8701854493580599,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.383213969888972e-05,
+      "loss": 0.7261,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8773181169757489,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.366312274049602e-05,
+      "loss": 0.6459,
+      "step": 1230
+    },
+    {
+      "epoch": 0.884450784593438,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.34919777030994e-05,
+      "loss": 0.7202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.891583452211127,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.331871292806759e-05,
+      "loss": 0.7874,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8987161198288159,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.314333686008125e-05,
+      "loss": 0.8524,
+      "step": 1260
+    },
+    {
+      "epoch": 0.905848787446505,
+      "grad_norm": 5.75,
+      "learning_rate": 9.296585804672252e-05,
+      "loss": 0.7854,
+      "step": 1270
+    },
+    {
+      "epoch": 0.912981455064194,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.278628513805838e-05,
+      "loss": 0.8386,
+      "step": 1280
+    },
+    {
+      "epoch": 0.920114122681883,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.260462688621905e-05,
+      "loss": 0.8279,
+      "step": 1290
+    },
+    {
+      "epoch": 0.927246790299572,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.242089214497145e-05,
+      "loss": 0.8345,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9343794579172611,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.223508986928766e-05,
+      "loss": 0.7596,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9415121255349501,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.204722911490846e-05,
+      "loss": 0.6704,
+      "step": 1320
+    },
+    {
+      "epoch": 0.948644793152639,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.1857319037902e-05,
+      "loss": 0.7792,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9557774607703281,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.16653688942175e-05,
+      "loss": 0.813,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9629101283880172,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.147138803923416e-05,
+      "loss": 0.834,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9700427960057061,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.127538592730519e-05,
+      "loss": 0.8299,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9771754636233951,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.107737211129701e-05,
+      "loss": 0.7407,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9843081312410842,
+      "grad_norm": 5.0,
+      "learning_rate": 9.087735624212365e-05,
+      "loss": 0.8821,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9914407988587732,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.06753480682764e-05,
+      "loss": 0.8595,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9985734664764622,
+      "grad_norm": 4.5,
+      "learning_rate": 9.047135743534868e-05,
+      "loss": 0.869,
+      "step": 1400
+    },
+    {
+      "epoch": 1.005706134094151,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.02653942855561e-05,
+      "loss": 0.8397,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0128388017118402,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.005746865725207e-05,
+      "loss": 0.8047,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0199714693295292,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.984759068443831e-05,
+      "loss": 0.72,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0271041369472182,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.963577059627118e-05,
+      "loss": 0.7724,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0342368045649073,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.942201871656291e-05,
+      "loss": 0.8014,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0413694721825963,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.920634546327858e-05,
+      "loss": 0.8532,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0485021398002854,
+      "grad_norm": 3.75,
+      "learning_rate": 8.898876134802826e-05,
+      "loss": 0.7285,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0556348074179742,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.87692769755548e-05,
+      "loss": 0.8574,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0627674750356633,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.854790304321681e-05,
+      "loss": 0.76,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0699001426533523,
+      "grad_norm": 4.75,
+      "learning_rate": 8.832465034046749e-05,
+      "loss": 0.8441,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0770328102710414,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.80995297483286e-05,
+      "loss": 0.764,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0841654778887304,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.787255223886019e-05,
+      "loss": 0.7432,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0912981455064195,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.764372887462586e-05,
+      "loss": 0.7894,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0984308131241085,
+      "grad_norm": 4.5,
+      "learning_rate": 8.741307080815357e-05,
+      "loss": 0.761,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1055634807417973,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.718058928139205e-05,
+      "loss": 0.9432,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1126961483594864,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.694629562516294e-05,
+      "loss": 0.7863,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1198288159771754,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.671020125860851e-05,
+      "loss": 0.7732,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1269614835948645,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.647231768863512e-05,
+      "loss": 0.7651,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1340941512125535,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.623265650935234e-05,
+      "loss": 0.8021,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1412268188302426,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.599122940150795e-05,
+      "loss": 0.8561,
+      "step": 1600
+    },
+    {
+      "epoch": 1.1483594864479316,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.574804813191858e-05,
+      "loss": 0.7551,
+      "step": 1610
+    },
+    {
+      "epoch": 1.1554921540656204,
+      "grad_norm": 5.375,
+      "learning_rate": 8.550312455289625e-05,
+      "loss": 0.8045,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1626248216833095,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.525647060167063e-05,
+      "loss": 0.7145,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1697574893009985,
+      "grad_norm": 4.125,
+      "learning_rate": 8.500809829980734e-05,
+      "loss": 0.8007,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1768901569186876,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.4758019752622e-05,
+      "loss": 0.8009,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1840228245363766,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.450624714859016e-05,
+      "loss": 0.833,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1911554921540657,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.425279275875336e-05,
+      "loss": 0.8149,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1982881597717547,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.399766893612096e-05,
+      "loss": 0.808,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2054208273894436,
+      "grad_norm": 4.625,
+      "learning_rate": 8.374088811506819e-05,
+      "loss": 0.7549,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2125534950071326,
+      "grad_norm": 5.71875,
+      "learning_rate": 8.348246281072998e-05,
+      "loss": 0.7755,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2196861626248217,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.322240561839109e-05,
+      "loss": 0.8038,
+      "step": 1710
+    },
+    {
+      "epoch": 1.2268188302425107,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.296072921287217e-05,
+      "loss": 0.7207,
+      "step": 1720
+    },
+    {
+      "epoch": 1.2339514978601998,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.269744634791208e-05,
+      "loss": 0.7994,
+      "step": 1730
+    },
+    {
+      "epoch": 1.2410841654778888,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.243256985554621e-05,
+      "loss": 0.7989,
+      "step": 1740
+    },
+    {
+      "epoch": 1.2482168330955776,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.21661126454811e-05,
+      "loss": 0.6516,
+      "step": 1750
+    },
+    {
+      "epoch": 1.2553495007132667,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.189808770446528e-05,
+      "loss": 0.7919,
+      "step": 1760
+    },
+    {
+      "epoch": 1.2624821683309557,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.9372,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2696148359486448,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.135738695798376e-05,
+      "loss": 0.7683,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2767475035663338,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.108473750550965e-05,
+      "loss": 0.737,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2838801711840229,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.081057302678352e-05,
+      "loss": 0.7507,
+      "step": 1800
+    },
+    {
+      "epoch": 1.291012838801712,
+      "grad_norm": 4.875,
+      "learning_rate": 8.053490688419533e-05,
+      "loss": 0.7016,
+      "step": 1810
+    },
+    {
+      "epoch": 1.298145506419401,
+      "grad_norm": 4.5,
+      "learning_rate": 8.02577525133239e-05,
+      "loss": 0.7151,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3052781740370898,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.997912342228232e-05,
+      "loss": 0.8559,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3124108416547788,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.969903319105935e-05,
+      "loss": 0.7118,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3195435092724679,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.941749547085777e-05,
+      "loss": 0.7855,
+      "step": 1850
+    },
+    {
+      "epoch": 1.326676176890157,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.913452398342881e-05,
+      "loss": 0.7438,
+      "step": 1860
+    },
+    {
+      "epoch": 1.333808844507846,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.885013252040359e-05,
+      "loss": 0.7018,
+      "step": 1870
+    },
+    {
+      "epoch": 1.340941512125535,
+      "grad_norm": 6.9375,
+      "learning_rate": 7.856433494262077e-05,
+      "loss": 0.7083,
+      "step": 1880
+    },
+    {
+      "epoch": 1.3480741797432239,
+      "grad_norm": 4.875,
+      "learning_rate": 7.827714517945115e-05,
+      "loss": 0.8265,
+      "step": 1890
+    },
+    {
+      "epoch": 1.355206847360913,
+      "grad_norm": 5.5625,
+      "learning_rate": 7.798857722811856e-05,
+      "loss": 0.8735,
+      "step": 1900
+    },
+    {
+      "epoch": 1.362339514978602,
+      "grad_norm": 5.71875,
+      "learning_rate": 7.769864515301787e-05,
+      "loss": 0.7256,
+      "step": 1910
+    },
+    {
+      "epoch": 1.369472182596291,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.740736308502938e-05,
+      "loss": 0.7472,
+      "step": 1920
+    },
+    {
+      "epoch": 1.37660485021398,
+      "grad_norm": 5.875,
+      "learning_rate": 7.711474522083015e-05,
+      "loss": 0.7601,
+      "step": 1930
+    },
+    {
+      "epoch": 1.383737517831669,
+      "grad_norm": 5.75,
+      "learning_rate": 7.682080582220206e-05,
+      "loss": 0.8857,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3908701854493581,
+      "grad_norm": 6.40625,
+      "learning_rate": 7.65255592153367e-05,
+      "loss": 0.7296,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3980028530670472,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.622901979013717e-05,
+      "loss": 0.7331,
+      "step": 1960
+    },
+    {
+      "epoch": 1.405135520684736,
+      "grad_norm": 5.09375,
+      "learning_rate": 7.593120199951671e-05,
+      "loss": 0.8462,
+      "step": 1970
+    },
+    {
+      "epoch": 1.412268188302425,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.563212035869425e-05,
+      "loss": 0.7307,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4194008559201141,
+      "grad_norm": 4.90625,
+      "learning_rate": 7.533178944448705e-05,
+      "loss": 0.7711,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.503022389460014e-05,
+      "loss": 0.7485,
+      "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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-2000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af7cf7c3c741add5e8996639d61d87fe74bb40ad0faa1b29901130f4b3a60fc3
+size 6353

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73bd96a7a178cbb1b6042cb786b17db797070f08fb5cdd2b8c9292b9919e8ea6
+size 298041696

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65d16bc47175a6c00670f9468b6365667a23c8eaf3d7b5528554ef44bf7728ee
+size 596170443

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/scheduler.pt

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/trainer_state.json

@@ -0,0 +1,2329 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.1398002853067046,
+  "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": 0.5201586889046041,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.23997409265084552,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.34286887314902237,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6213496755517833,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4542903465876373,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6560051657630788,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3328835143785305,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.4570725655218753,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.832098944676056,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.25908236715595256,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7391066781087818,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.30431141846717047,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.4799335275762782,
+      "step": 0
+    },
+    {
+      "epoch": 0.007132667617689016,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8e-06,
+      "loss": 1.6901,
+      "step": 10
+    },
+    {
+      "epoch": 0.014265335235378032,
+      "grad_norm": 10.75,
+      "learning_rate": 3.8e-06,
+      "loss": 1.7675,
+      "step": 20
+    },
+    {
+      "epoch": 0.021398002853067047,
+      "grad_norm": 21.0,
+      "learning_rate": 5.8e-06,
+      "loss": 1.527,
+      "step": 30
+    },
+    {
+      "epoch": 0.028530670470756064,
+      "grad_norm": 19.0,
+      "learning_rate": 7.8e-06,
+      "loss": 1.7708,
+      "step": 40
+    },
+    {
+      "epoch": 0.03566333808844508,
+      "grad_norm": 17.125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.4093,
+      "step": 50
+    },
+    {
+      "epoch": 0.042796005706134094,
+      "grad_norm": 15.625,
+      "learning_rate": 1.18e-05,
+      "loss": 1.579,
+      "step": 60
+    },
+    {
+      "epoch": 0.04992867332382311,
+      "grad_norm": 14.0,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.4845,
+      "step": 70
+    },
+    {
+      "epoch": 0.05706134094151213,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.58e-05,
+      "loss": 1.598,
+      "step": 80
+    },
+    {
+      "epoch": 0.06419400855920114,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.78e-05,
+      "loss": 1.3226,
+      "step": 90
+    },
+    {
+      "epoch": 0.07132667617689016,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.4969,
+      "step": 100
+    },
+    {
+      "epoch": 0.07845934379457917,
+      "grad_norm": 13.125,
+      "learning_rate": 2.18e-05,
+      "loss": 1.3867,
+      "step": 110
+    },
+    {
+      "epoch": 0.08559201141226819,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.38e-05,
+      "loss": 1.1904,
+      "step": 120
+    },
+    {
+      "epoch": 0.09272467902995721,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.58e-05,
+      "loss": 1.1668,
+      "step": 130
+    },
+    {
+      "epoch": 0.09985734664764621,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.124,
+      "step": 140
+    },
+    {
+      "epoch": 0.10699001426533523,
+      "grad_norm": 9.75,
+      "learning_rate": 2.98e-05,
+      "loss": 1.1571,
+      "step": 150
+    },
+    {
+      "epoch": 0.11412268188302425,
+      "grad_norm": 12.25,
+      "learning_rate": 3.18e-05,
+      "loss": 1.2145,
+      "step": 160
+    },
+    {
+      "epoch": 0.12125534950071326,
+      "grad_norm": 10.625,
+      "learning_rate": 3.38e-05,
+      "loss": 1.155,
+      "step": 170
+    },
+    {
+      "epoch": 0.12838801711840228,
+      "grad_norm": 8.5,
+      "learning_rate": 3.58e-05,
+      "loss": 1.1513,
+      "step": 180
+    },
+    {
+      "epoch": 0.1355206847360913,
+      "grad_norm": 9.5,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.9803,
+      "step": 190
+    },
+    {
+      "epoch": 0.14265335235378032,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 1.0302,
+      "step": 200
+    },
+    {
+      "epoch": 0.14978601997146934,
+      "grad_norm": 9.25,
+      "learning_rate": 4.18e-05,
+      "loss": 0.9878,
+      "step": 210
+    },
+    {
+      "epoch": 0.15691868758915833,
+      "grad_norm": 10.25,
+      "learning_rate": 4.38e-05,
+      "loss": 0.9816,
+      "step": 220
+    },
+    {
+      "epoch": 0.16405135520684735,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.58e-05,
+      "loss": 1.0415,
+      "step": 230
+    },
+    {
+      "epoch": 0.17118402282453637,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.78e-05,
+      "loss": 1.1126,
+      "step": 240
+    },
+    {
+      "epoch": 0.1783166904422254,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 1.0127,
+      "step": 250
+    },
+    {
+      "epoch": 0.18544935805991442,
+      "grad_norm": 5.71875,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.9577,
+      "step": 260
+    },
+    {
+      "epoch": 0.19258202567760344,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.9518,
+      "step": 270
+    },
+    {
+      "epoch": 0.19971469329529243,
+      "grad_norm": 6.28125,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.7952,
+      "step": 280
+    },
+    {
+      "epoch": 0.20684736091298145,
+      "grad_norm": 6.875,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.9201,
+      "step": 290
+    },
+    {
+      "epoch": 0.21398002853067047,
+      "grad_norm": 6.09375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.951,
+      "step": 300
+    },
+    {
+      "epoch": 0.2211126961483595,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.18e-05,
+      "loss": 0.9199,
+      "step": 310
+    },
+    {
+      "epoch": 0.2282453637660485,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.38e-05,
+      "loss": 0.8773,
+      "step": 320
+    },
+    {
+      "epoch": 0.23537803138373753,
+      "grad_norm": 4.875,
+      "learning_rate": 6.58e-05,
+      "loss": 0.9386,
+      "step": 330
+    },
+    {
+      "epoch": 0.24251069900142652,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.9364,
+      "step": 340
+    },
+    {
+      "epoch": 0.24964336661911554,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.98e-05,
+      "loss": 0.9053,
+      "step": 350
+    },
+    {
+      "epoch": 0.25677603423680456,
+      "grad_norm": 5.8125,
+      "learning_rate": 7.18e-05,
+      "loss": 0.8493,
+      "step": 360
+    },
+    {
+      "epoch": 0.26390870185449355,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.38e-05,
+      "loss": 0.8218,
+      "step": 370
+    },
+    {
+      "epoch": 0.2710413694721826,
+      "grad_norm": 5.625,
+      "learning_rate": 7.58e-05,
+      "loss": 0.8975,
+      "step": 380
+    },
+    {
+      "epoch": 0.2781740370898716,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.8891,
+      "step": 390
+    },
+    {
+      "epoch": 0.28530670470756064,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.7831,
+      "step": 400
+    },
+    {
+      "epoch": 0.29243937232524964,
+      "grad_norm": 5.125,
+      "learning_rate": 8.18e-05,
+      "loss": 0.9017,
+      "step": 410
+    },
+    {
+      "epoch": 0.2995720399429387,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.38e-05,
+      "loss": 0.7839,
+      "step": 420
+    },
+    {
+      "epoch": 0.3067047075606277,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.886,
+      "step": 430
+    },
+    {
+      "epoch": 0.31383737517831667,
+      "grad_norm": 6.1875,
+      "learning_rate": 8.78e-05,
+      "loss": 0.863,
+      "step": 440
+    },
+    {
+      "epoch": 0.3209700427960057,
+      "grad_norm": 4.625,
+      "learning_rate": 8.98e-05,
+      "loss": 0.884,
+      "step": 450
+    },
+    {
+      "epoch": 0.3281027104136947,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.9209,
+      "step": 460
+    },
+    {
+      "epoch": 0.33523537803138376,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.8775,
+      "step": 470
+    },
+    {
+      "epoch": 0.34236804564907275,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.58e-05,
+      "loss": 0.7956,
+      "step": 480
+    },
+    {
+      "epoch": 0.34950071326676174,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.8632,
+      "step": 490
+    },
+    {
+      "epoch": 0.3566333808844508,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.98e-05,
+      "loss": 0.8308,
+      "step": 500
+    },
+    {
+      "epoch": 0.3637660485021398,
+      "grad_norm": 5.25,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.884,
+      "step": 510
+    },
+    {
+      "epoch": 0.37089871611982883,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.8663,
+      "step": 520
+    },
+    {
+      "epoch": 0.3780313837375178,
+      "grad_norm": 5.25,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.8722,
+      "step": 530
+    },
+    {
+      "epoch": 0.38516405135520687,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.8391,
+      "step": 540
+    },
+    {
+      "epoch": 0.39229671897289586,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.7823,
+      "step": 550
+    },
+    {
+      "epoch": 0.39942938659058486,
+      "grad_norm": 4.875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.8606,
+      "step": 560
+    },
+    {
+      "epoch": 0.4065620542082739,
+      "grad_norm": 5.625,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.8719,
+      "step": 570
+    },
+    {
+      "epoch": 0.4136947218259629,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.8619,
+      "step": 580
+    },
+    {
+      "epoch": 0.42082738944365194,
+      "grad_norm": 5.0,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.8439,
+      "step": 590
+    },
+    {
+      "epoch": 0.42796005706134094,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.886,
+      "step": 600
+    },
+    {
+      "epoch": 0.43509272467902993,
+      "grad_norm": 5.125,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.8346,
+      "step": 610
+    },
+    {
+      "epoch": 0.442225392296719,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.8145,
+      "step": 620
+    },
+    {
+      "epoch": 0.44935805991440797,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.8739,
+      "step": 630
+    },
+    {
+      "epoch": 0.456490727532097,
+      "grad_norm": 4.75,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.7826,
+      "step": 640
+    },
+    {
+      "epoch": 0.463623395149786,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.8541,
+      "step": 650
+    },
+    {
+      "epoch": 0.47075606276747506,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.8208,
+      "step": 660
+    },
+    {
+      "epoch": 0.47788873038516405,
+      "grad_norm": 5.375,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.7761,
+      "step": 670
+    },
+    {
+      "epoch": 0.48502139800285304,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.8346,
+      "step": 680
+    },
+    {
+      "epoch": 0.4921540656205421,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.9192,
+      "step": 690
+    },
+    {
+      "epoch": 0.4992867332382311,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.8037,
+      "step": 700
+    },
+    {
+      "epoch": 0.5064194008559201,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.7669,
+      "step": 710
+    },
+    {
+      "epoch": 0.5135520684736091,
+      "grad_norm": 6.875,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.7915,
+      "step": 720
+    },
+    {
+      "epoch": 0.5206847360912982,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.7863,
+      "step": 730
+    },
+    {
+      "epoch": 0.5278174037089871,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.7789,
+      "step": 740
+    },
+    {
+      "epoch": 0.5349500713266762,
+      "grad_norm": 4.0,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.7759,
+      "step": 750
+    },
+    {
+      "epoch": 0.5420827389443652,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.8504,
+      "step": 760
+    },
+    {
+      "epoch": 0.5492154065620543,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.809,
+      "step": 770
+    },
+    {
+      "epoch": 0.5563480741797432,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.7721,
+      "step": 780
+    },
+    {
+      "epoch": 0.5634807417974322,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.7319,
+      "step": 790
+    },
+    {
+      "epoch": 0.5706134094151213,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.7726,
+      "step": 800
+    },
+    {
+      "epoch": 0.5777460770328102,
+      "grad_norm": 4.625,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.8522,
+      "step": 810
+    },
+    {
+      "epoch": 0.5848787446504993,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.7761,
+      "step": 820
+    },
+    {
+      "epoch": 0.5920114122681883,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.8702,
+      "step": 830
+    },
+    {
+      "epoch": 0.5991440798858774,
+      "grad_norm": 5.25,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.8636,
+      "step": 840
+    },
+    {
+      "epoch": 0.6062767475035663,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.6974,
+      "step": 850
+    },
+    {
+      "epoch": 0.6134094151212554,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.8017,
+      "step": 860
+    },
+    {
+      "epoch": 0.6205420827389444,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.7799,
+      "step": 870
+    },
+    {
+      "epoch": 0.6276747503566333,
+      "grad_norm": 5.125,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.7808,
+      "step": 880
+    },
+    {
+      "epoch": 0.6348074179743224,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.8337,
+      "step": 890
+    },
+    {
+      "epoch": 0.6419400855920114,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.8043,
+      "step": 900
+    },
+    {
+      "epoch": 0.6490727532097005,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.8328,
+      "step": 910
+    },
+    {
+      "epoch": 0.6562054208273894,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.7745,
+      "step": 920
+    },
+    {
+      "epoch": 0.6633380884450785,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.745,
+      "step": 930
+    },
+    {
+      "epoch": 0.6704707560627675,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.82,
+      "step": 940
+    },
+    {
+      "epoch": 0.6776034236804565,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.8025,
+      "step": 950
+    },
+    {
+      "epoch": 0.6847360912981455,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.84,
+      "step": 960
+    },
+    {
+      "epoch": 0.6918687589158345,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.8552,
+      "step": 970
+    },
+    {
+      "epoch": 0.6990014265335235,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.8082,
+      "step": 980
+    },
+    {
+      "epoch": 0.7061340941512125,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.8265,
+      "step": 990
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "grad_norm": 4.875,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.8369,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.arguana": 0.47952715537701884,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.climate_fever": 0.22702207934204544,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.dbpedia_entity": 0.3287396364958694,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.fever": 0.5532628324042129,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.fiqa": 0.4160855038689836,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.hotpotqa": 0.6599716759215029,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.nfcorpus": 0.3449425020579428,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.nq": 0.39316532973343116,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.quora": 0.8295275476072758,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.scidocs": 0.2526832556333601,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.scifact": 0.7219633063518482,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.webis_touche2020": 0.2787587910246796,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/avg": 0.4571374679848476,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7203994293865906,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.68762555129222e-05,
+      "loss": 0.8872,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7275320970042796,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.675366799960841e-05,
+      "loss": 0.8334,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7346647646219686,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.662880177847594e-05,
+      "loss": 0.8014,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7417974322396577,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.65016629353282e-05,
+      "loss": 0.848,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7489300998573466,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.637225766673307e-05,
+      "loss": 0.7203,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7560627674750356,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.624059227972075e-05,
+      "loss": 0.8405,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7631954350927247,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.610667319147647e-05,
+      "loss": 0.8418,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7703281027104137,
+      "grad_norm": 5.75,
+      "learning_rate": 9.597050692902765e-05,
+      "loss": 0.8853,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7774607703281027,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.583210012892582e-05,
+      "loss": 0.7733,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7845934379457917,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.569145953692315e-05,
+      "loss": 0.7698,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7917261055634808,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.55485920076437e-05,
+      "loss": 0.7759,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7988587731811697,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.540350450424926e-05,
+      "loss": 0.8025,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8059914407988588,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.525620409810007e-05,
+      "loss": 0.8364,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8131241084165478,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.510669796841014e-05,
+      "loss": 0.7523,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8202567760342369,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.495499340189728e-05,
+      "loss": 0.7333,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8273894436519258,
+      "grad_norm": 4.875,
+      "learning_rate": 9.480109779242805e-05,
+      "loss": 0.7325,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8345221112696148,
+      "grad_norm": 5.0,
+      "learning_rate": 9.464501864065735e-05,
+      "loss": 0.7996,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8416547788873039,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.448676355366282e-05,
+      "loss": 0.7946,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8487874465049928,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.432634024457414e-05,
+      "loss": 0.9133,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8559201141226819,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.416375653219709e-05,
+      "loss": 0.7218,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8630527817403709,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.399902034063242e-05,
+      "loss": 0.8453,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8701854493580599,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.383213969888972e-05,
+      "loss": 0.7261,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8773181169757489,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.366312274049602e-05,
+      "loss": 0.6459,
+      "step": 1230
+    },
+    {
+      "epoch": 0.884450784593438,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.34919777030994e-05,
+      "loss": 0.7202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.891583452211127,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.331871292806759e-05,
+      "loss": 0.7874,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8987161198288159,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.314333686008125e-05,
+      "loss": 0.8524,
+      "step": 1260
+    },
+    {
+      "epoch": 0.905848787446505,
+      "grad_norm": 5.75,
+      "learning_rate": 9.296585804672252e-05,
+      "loss": 0.7854,
+      "step": 1270
+    },
+    {
+      "epoch": 0.912981455064194,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.278628513805838e-05,
+      "loss": 0.8386,
+      "step": 1280
+    },
+    {
+      "epoch": 0.920114122681883,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.260462688621905e-05,
+      "loss": 0.8279,
+      "step": 1290
+    },
+    {
+      "epoch": 0.927246790299572,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.242089214497145e-05,
+      "loss": 0.8345,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9343794579172611,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.223508986928766e-05,
+      "loss": 0.7596,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9415121255349501,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.204722911490846e-05,
+      "loss": 0.6704,
+      "step": 1320
+    },
+    {
+      "epoch": 0.948644793152639,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.1857319037902e-05,
+      "loss": 0.7792,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9557774607703281,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.16653688942175e-05,
+      "loss": 0.813,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9629101283880172,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.147138803923416e-05,
+      "loss": 0.834,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9700427960057061,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.127538592730519e-05,
+      "loss": 0.8299,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9771754636233951,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.107737211129701e-05,
+      "loss": 0.7407,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9843081312410842,
+      "grad_norm": 5.0,
+      "learning_rate": 9.087735624212365e-05,
+      "loss": 0.8821,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9914407988587732,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.06753480682764e-05,
+      "loss": 0.8595,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9985734664764622,
+      "grad_norm": 4.5,
+      "learning_rate": 9.047135743534868e-05,
+      "loss": 0.869,
+      "step": 1400
+    },
+    {
+      "epoch": 1.005706134094151,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.02653942855561e-05,
+      "loss": 0.8397,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0128388017118402,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.005746865725207e-05,
+      "loss": 0.8047,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0199714693295292,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.984759068443831e-05,
+      "loss": 0.72,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0271041369472182,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.963577059627118e-05,
+      "loss": 0.7724,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0342368045649073,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.942201871656291e-05,
+      "loss": 0.8014,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0413694721825963,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.920634546327858e-05,
+      "loss": 0.8532,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0485021398002854,
+      "grad_norm": 3.75,
+      "learning_rate": 8.898876134802826e-05,
+      "loss": 0.7285,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0556348074179742,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.87692769755548e-05,
+      "loss": 0.8574,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0627674750356633,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.854790304321681e-05,
+      "loss": 0.76,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0699001426533523,
+      "grad_norm": 4.75,
+      "learning_rate": 8.832465034046749e-05,
+      "loss": 0.8441,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0770328102710414,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.80995297483286e-05,
+      "loss": 0.764,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0841654778887304,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.787255223886019e-05,
+      "loss": 0.7432,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0912981455064195,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.764372887462586e-05,
+      "loss": 0.7894,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0984308131241085,
+      "grad_norm": 4.5,
+      "learning_rate": 8.741307080815357e-05,
+      "loss": 0.761,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1055634807417973,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.718058928139205e-05,
+      "loss": 0.9432,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1126961483594864,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.694629562516294e-05,
+      "loss": 0.7863,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1198288159771754,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.671020125860851e-05,
+      "loss": 0.7732,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1269614835948645,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.647231768863512e-05,
+      "loss": 0.7651,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1340941512125535,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.623265650935234e-05,
+      "loss": 0.8021,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1412268188302426,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.599122940150795e-05,
+      "loss": 0.8561,
+      "step": 1600
+    },
+    {
+      "epoch": 1.1483594864479316,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.574804813191858e-05,
+      "loss": 0.7551,
+      "step": 1610
+    },
+    {
+      "epoch": 1.1554921540656204,
+      "grad_norm": 5.375,
+      "learning_rate": 8.550312455289625e-05,
+      "loss": 0.8045,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1626248216833095,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.525647060167063e-05,
+      "loss": 0.7145,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1697574893009985,
+      "grad_norm": 4.125,
+      "learning_rate": 8.500809829980734e-05,
+      "loss": 0.8007,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1768901569186876,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.4758019752622e-05,
+      "loss": 0.8009,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1840228245363766,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.450624714859016e-05,
+      "loss": 0.833,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1911554921540657,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.425279275875336e-05,
+      "loss": 0.8149,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1982881597717547,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.399766893612096e-05,
+      "loss": 0.808,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2054208273894436,
+      "grad_norm": 4.625,
+      "learning_rate": 8.374088811506819e-05,
+      "loss": 0.7549,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2125534950071326,
+      "grad_norm": 5.71875,
+      "learning_rate": 8.348246281072998e-05,
+      "loss": 0.7755,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2196861626248217,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.322240561839109e-05,
+      "loss": 0.8038,
+      "step": 1710
+    },
+    {
+      "epoch": 1.2268188302425107,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.296072921287217e-05,
+      "loss": 0.7207,
+      "step": 1720
+    },
+    {
+      "epoch": 1.2339514978601998,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.269744634791208e-05,
+      "loss": 0.7994,
+      "step": 1730
+    },
+    {
+      "epoch": 1.2410841654778888,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.243256985554621e-05,
+      "loss": 0.7989,
+      "step": 1740
+    },
+    {
+      "epoch": 1.2482168330955776,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.21661126454811e-05,
+      "loss": 0.6516,
+      "step": 1750
+    },
+    {
+      "epoch": 1.2553495007132667,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.189808770446528e-05,
+      "loss": 0.7919,
+      "step": 1760
+    },
+    {
+      "epoch": 1.2624821683309557,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.9372,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2696148359486448,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.135738695798376e-05,
+      "loss": 0.7683,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2767475035663338,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.108473750550965e-05,
+      "loss": 0.737,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2838801711840229,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.081057302678352e-05,
+      "loss": 0.7507,
+      "step": 1800
+    },
+    {
+      "epoch": 1.291012838801712,
+      "grad_norm": 4.875,
+      "learning_rate": 8.053490688419533e-05,
+      "loss": 0.7016,
+      "step": 1810
+    },
+    {
+      "epoch": 1.298145506419401,
+      "grad_norm": 4.5,
+      "learning_rate": 8.02577525133239e-05,
+      "loss": 0.7151,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3052781740370898,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.997912342228232e-05,
+      "loss": 0.8559,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3124108416547788,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.969903319105935e-05,
+      "loss": 0.7118,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3195435092724679,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.941749547085777e-05,
+      "loss": 0.7855,
+      "step": 1850
+    },
+    {
+      "epoch": 1.326676176890157,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.913452398342881e-05,
+      "loss": 0.7438,
+      "step": 1860
+    },
+    {
+      "epoch": 1.333808844507846,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.885013252040359e-05,
+      "loss": 0.7018,
+      "step": 1870
+    },
+    {
+      "epoch": 1.340941512125535,
+      "grad_norm": 6.9375,
+      "learning_rate": 7.856433494262077e-05,
+      "loss": 0.7083,
+      "step": 1880
+    },
+    {
+      "epoch": 1.3480741797432239,
+      "grad_norm": 4.875,
+      "learning_rate": 7.827714517945115e-05,
+      "loss": 0.8265,
+      "step": 1890
+    },
+    {
+      "epoch": 1.355206847360913,
+      "grad_norm": 5.5625,
+      "learning_rate": 7.798857722811856e-05,
+      "loss": 0.8735,
+      "step": 1900
+    },
+    {
+      "epoch": 1.362339514978602,
+      "grad_norm": 5.71875,
+      "learning_rate": 7.769864515301787e-05,
+      "loss": 0.7256,
+      "step": 1910
+    },
+    {
+      "epoch": 1.369472182596291,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.740736308502938e-05,
+      "loss": 0.7472,
+      "step": 1920
+    },
+    {
+      "epoch": 1.37660485021398,
+      "grad_norm": 5.875,
+      "learning_rate": 7.711474522083015e-05,
+      "loss": 0.7601,
+      "step": 1930
+    },
+    {
+      "epoch": 1.383737517831669,
+      "grad_norm": 5.75,
+      "learning_rate": 7.682080582220206e-05,
+      "loss": 0.8857,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3908701854493581,
+      "grad_norm": 6.40625,
+      "learning_rate": 7.65255592153367e-05,
+      "loss": 0.7296,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3980028530670472,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.622901979013717e-05,
+      "loss": 0.7331,
+      "step": 1960
+    },
+    {
+      "epoch": 1.405135520684736,
+      "grad_norm": 5.09375,
+      "learning_rate": 7.593120199951671e-05,
+      "loss": 0.8462,
+      "step": 1970
+    },
+    {
+      "epoch": 1.412268188302425,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.563212035869425e-05,
+      "loss": 0.7307,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4194008559201141,
+      "grad_norm": 4.90625,
+      "learning_rate": 7.533178944448705e-05,
+      "loss": 0.7711,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.503022389460014e-05,
+      "loss": 0.7485,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.arguana": 0.47494654864753366,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.climate_fever": 0.2496059194905162,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.dbpedia_entity": 0.3371368383795953,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.fever": 0.6065422117591313,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.fiqa": 0.4135337243548038,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.hotpotqa": 0.6804056480114672,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.nfcorpus": 0.35238501477624823,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.nq": 0.39770361480597743,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.quora": 0.8217479889085385,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.scidocs": 0.2592573629546752,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.scifact": 0.7479618950231317,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.webis_touche2020": 0.2719065772753103,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/avg": 0.467761112032244,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4336661911554922,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.472743840691299e-05,
+      "loss": 0.8505,
+      "step": 2010
+    },
+    {
+      "epoch": 1.440798858773181,
+      "grad_norm": 6.15625,
+      "learning_rate": 7.44234477387631e-05,
+      "loss": 0.7449,
+      "step": 2020
+    },
+    {
+      "epoch": 1.44793152639087,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.411826670622676e-05,
+      "loss": 0.6961,
+      "step": 2030
+    },
+    {
+      "epoch": 1.4550641940085591,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.381191018339696e-05,
+      "loss": 0.8105,
+      "step": 2040
+    },
+    {
+      "epoch": 1.4621968616262482,
+      "grad_norm": 5.59375,
+      "learning_rate": 7.350439310165842e-05,
+      "loss": 0.8111,
+      "step": 2050
+    },
+    {
+      "epoch": 1.4693295292439372,
+      "grad_norm": 4.625,
+      "learning_rate": 7.319573044895986e-05,
+      "loss": 0.7972,
+      "step": 2060
+    },
+    {
+      "epoch": 1.4764621968616263,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.28859372690835e-05,
+      "loss": 0.7745,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4835948644793153,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.257502866091192e-05,
+      "loss": 0.7904,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4907275320970044,
+      "grad_norm": 4.875,
+      "learning_rate": 7.226301977769199e-05,
+      "loss": 0.7724,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4978601997146934,
+      "grad_norm": 5.59375,
+      "learning_rate": 7.194992582629654e-05,
+      "loss": 0.7693,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5049928673323825,
+      "grad_norm": 4.75,
+      "learning_rate": 7.163576206648304e-05,
+      "loss": 0.6929,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5121255349500713,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.132054381014995e-05,
+      "loss": 0.8008,
+      "step": 2120
+    },
+    {
+      "epoch": 1.5192582025677603,
+      "grad_norm": 6.28125,
+      "learning_rate": 7.100428642059033e-05,
+      "loss": 0.8399,
+      "step": 2130
+    },
+    {
+      "epoch": 1.5263908701854494,
+      "grad_norm": 6.3125,
+      "learning_rate": 7.068700531174319e-05,
+      "loss": 0.7399,
+      "step": 2140
+    },
+    {
+      "epoch": 1.5335235378031382,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.036871594744217e-05,
+      "loss": 0.7796,
+      "step": 2150
+    },
+    {
+      "epoch": 1.5406562054208273,
+      "grad_norm": 5.34375,
+      "learning_rate": 7.004943384066187e-05,
+      "loss": 0.8685,
+      "step": 2160
+    },
+    {
+      "epoch": 1.5477888730385163,
+      "grad_norm": 4.8125,
+      "learning_rate": 6.97291745527617e-05,
+      "loss": 0.7448,
+      "step": 2170
+    },
+    {
+      "epoch": 1.5549215406562054,
+      "grad_norm": 6.09375,
+      "learning_rate": 6.940795369272754e-05,
+      "loss": 0.8527,
+      "step": 2180
+    },
+    {
+      "epoch": 1.5620542082738944,
+      "grad_norm": 5.15625,
+      "learning_rate": 6.908578691641092e-05,
+      "loss": 0.8704,
+      "step": 2190
+    },
+    {
+      "epoch": 1.5691868758915835,
+      "grad_norm": 4.3125,
+      "learning_rate": 6.876268992576604e-05,
+      "loss": 0.7033,
+      "step": 2200
+    },
+    {
+      "epoch": 1.5763195435092725,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.843867846808438e-05,
+      "loss": 0.7664,
+      "step": 2210
+    },
+    {
+      "epoch": 1.5834522111269616,
+      "grad_norm": 4.5625,
+      "learning_rate": 6.811376833522729e-05,
+      "loss": 0.7902,
+      "step": 2220
+    },
+    {
+      "epoch": 1.5905848787446506,
+      "grad_norm": 4.6875,
+      "learning_rate": 6.778797536285624e-05,
+      "loss": 0.7585,
+      "step": 2230
+    },
+    {
+      "epoch": 1.5977175463623396,
+      "grad_norm": 5.375,
+      "learning_rate": 6.746131542966112e-05,
+      "loss": 0.8124,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6048502139800287,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.713380445658618e-05,
+      "loss": 0.7936,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6119828815977175,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.680545840605423e-05,
+      "loss": 0.6785,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6191155492154066,
+      "grad_norm": 5.875,
+      "learning_rate": 6.647629328118851e-05,
+      "loss": 0.8321,
+      "step": 2270
+    },
+    {
+      "epoch": 1.6262482168330956,
+      "grad_norm": 5.125,
+      "learning_rate": 6.614632512503288e-05,
+      "loss": 0.8184,
+      "step": 2280
+    },
+    {
+      "epoch": 1.6333808844507844,
+      "grad_norm": 4.5,
+      "learning_rate": 6.58155700197697e-05,
+      "loss": 0.7187,
+      "step": 2290
+    },
+    {
+      "epoch": 1.6405135520684735,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.548404408593621e-05,
+      "loss": 0.6873,
+      "step": 2300
+    },
+    {
+      "epoch": 1.6476462196861625,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.515176348163871e-05,
+      "loss": 0.7537,
+      "step": 2310
+    },
+    {
+      "epoch": 1.6547788873038516,
+      "grad_norm": 4.71875,
+      "learning_rate": 6.481874440176506e-05,
+      "loss": 0.8058,
+      "step": 2320
+    },
+    {
+      "epoch": 1.6619115549215406,
+      "grad_norm": 5.375,
+      "learning_rate": 6.448500307719538e-05,
+      "loss": 0.8117,
+      "step": 2330
+    },
+    {
+      "epoch": 1.6690442225392297,
+      "grad_norm": 5.15625,
+      "learning_rate": 6.415055577401102e-05,
+      "loss": 0.7938,
+      "step": 2340
+    },
+    {
+      "epoch": 1.6761768901569187,
+      "grad_norm": 6.15625,
+      "learning_rate": 6.381541879270169e-05,
+      "loss": 0.7898,
+      "step": 2350
+    },
+    {
+      "epoch": 1.6833095577746078,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.347960846737106e-05,
+      "loss": 0.7896,
+      "step": 2360
+    },
+    {
+      "epoch": 1.6904422253922968,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.31431411649406e-05,
+      "loss": 0.6994,
+      "step": 2370
+    },
+    {
+      "epoch": 1.6975748930099859,
+      "grad_norm": 5.71875,
+      "learning_rate": 6.280603328435198e-05,
+      "loss": 0.7416,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7047075606276747,
+      "grad_norm": 4.8125,
+      "learning_rate": 6.24683012557677e-05,
+      "loss": 0.7143,
+      "step": 2390
+    },
+    {
+      "epoch": 1.7118402282453637,
+      "grad_norm": 5.28125,
+      "learning_rate": 6.212996153977037e-05,
+      "loss": 0.6699,
+      "step": 2400
+    },
+    {
+      "epoch": 1.7189728958630528,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.179103062656042e-05,
+      "loss": 0.7084,
+      "step": 2410
+    },
+    {
+      "epoch": 1.7261055634807418,
+      "grad_norm": 5.125,
+      "learning_rate": 6.145152503515239e-05,
+      "loss": 0.7638,
+      "step": 2420
+    },
+    {
+      "epoch": 1.7332382310984307,
+      "grad_norm": 5.0,
+      "learning_rate": 6.111146131256983e-05,
+      "loss": 0.7875,
+      "step": 2430
+    },
+    {
+      "epoch": 1.7403708987161197,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.077085603303882e-05,
+      "loss": 0.8864,
+      "step": 2440
+    },
+    {
+      "epoch": 1.7475035663338088,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.042972579718019e-05,
+      "loss": 0.7354,
+      "step": 2450
+    },
+    {
+      "epoch": 1.7546362339514978,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.008808723120035e-05,
+      "loss": 0.7581,
+      "step": 2460
+    },
+    {
+      "epoch": 1.7617689015691869,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.974595698608103e-05,
+      "loss": 0.738,
+      "step": 2470
+    },
+    {
+      "epoch": 1.768901569186876,
+      "grad_norm": 4.53125,
+      "learning_rate": 5.9403351736767696e-05,
+      "loss": 0.6686,
+      "step": 2480
+    },
+    {
+      "epoch": 1.776034236804565,
+      "grad_norm": 5.125,
+      "learning_rate": 5.906028818135687e-05,
+      "loss": 0.6354,
+      "step": 2490
+    },
+    {
+      "epoch": 1.783166904422254,
+      "grad_norm": 5.125,
+      "learning_rate": 5.8716783040282244e-05,
+      "loss": 0.7869,
+      "step": 2500
+    },
+    {
+      "epoch": 1.790299572039943,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.837285305549978e-05,
+      "loss": 0.7463,
+      "step": 2510
+    },
+    {
+      "epoch": 1.797432239657632,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.8028514989671724e-05,
+      "loss": 0.7524,
+      "step": 2520
+    },
+    {
+      "epoch": 1.804564907275321,
+      "grad_norm": 6.15625,
+      "learning_rate": 5.7683785625349616e-05,
+      "loss": 0.7558,
+      "step": 2530
+    },
+    {
+      "epoch": 1.81169757489301,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.733868176415633e-05,
+      "loss": 0.8002,
+      "step": 2540
+    },
+    {
+      "epoch": 1.818830242510699,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.699322022596722e-05,
+      "loss": 0.7829,
+      "step": 2550
+    },
+    {
+      "epoch": 1.825962910128388,
+      "grad_norm": 4.375,
+      "learning_rate": 5.664741784809022e-05,
+      "loss": 0.7097,
+      "step": 2560
+    },
+    {
+      "epoch": 1.833095577746077,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.630129148444543e-05,
+      "loss": 0.6793,
+      "step": 2570
+    },
+    {
+      "epoch": 1.840228245363766,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.595485800474349e-05,
+      "loss": 0.7081,
+      "step": 2580
+    },
+    {
+      "epoch": 1.847360912981455,
+      "grad_norm": 4.25,
+      "learning_rate": 5.560813429366345e-05,
+      "loss": 0.911,
+      "step": 2590
+    },
+    {
+      "epoch": 1.854493580599144,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.5261137250029835e-05,
+      "loss": 0.8128,
+      "step": 2600
+    },
+    {
+      "epoch": 1.861626248216833,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.4913883785988993e-05,
+      "loss": 0.7784,
+      "step": 2610
+    },
+    {
+      "epoch": 1.8687589158345221,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.456639082618489e-05,
+      "loss": 0.6765,
+      "step": 2620
+    },
+    {
+      "epoch": 1.8758915834522112,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.421867530693414e-05,
+      "loss": 0.7464,
+      "step": 2630
+    },
+    {
+      "epoch": 1.8830242510699002,
+      "grad_norm": 4.4375,
+      "learning_rate": 5.3870754175400595e-05,
+      "loss": 0.7452,
+      "step": 2640
+    },
+    {
+      "epoch": 1.8901569186875893,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.352264438876935e-05,
+      "loss": 0.7475,
+      "step": 2650
+    },
+    {
+      "epoch": 1.8972895863052783,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.3174362913420306e-05,
+      "loss": 0.761,
+      "step": 2660
+    },
+    {
+      "epoch": 1.9044222539229672,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.2825926724101236e-05,
+      "loss": 0.7183,
+      "step": 2670
+    },
+    {
+      "epoch": 1.9115549215406562,
+      "grad_norm": 4.875,
+      "learning_rate": 5.247735280310041e-05,
+      "loss": 0.756,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9186875891583453,
+      "grad_norm": 4.8125,
+      "learning_rate": 5.212865813941899e-05,
+      "loss": 0.6997,
+      "step": 2690
+    },
+    {
+      "epoch": 1.925820256776034,
+      "grad_norm": 4.75,
+      "learning_rate": 5.1779859727942924e-05,
+      "loss": 0.7978,
+      "step": 2700
+    },
+    {
+      "epoch": 1.9329529243937231,
+      "grad_norm": 4.5,
+      "learning_rate": 5.143097456861474e-05,
+      "loss": 0.7099,
+      "step": 2710
+    },
+    {
+      "epoch": 1.9400855920114122,
+      "grad_norm": 4.8125,
+      "learning_rate": 5.1082019665604895e-05,
+      "loss": 0.7029,
+      "step": 2720
+    },
+    {
+      "epoch": 1.9472182596291012,
+      "grad_norm": 5.875,
+      "learning_rate": 5.073301202648304e-05,
+      "loss": 0.7328,
+      "step": 2730
+    },
+    {
+      "epoch": 1.9543509272467903,
+      "grad_norm": 3.875,
+      "learning_rate": 5.0383968661389146e-05,
+      "loss": 0.7982,
+      "step": 2740
+    },
+    {
+      "epoch": 1.9614835948644793,
+      "grad_norm": 4.71875,
+      "learning_rate": 5.003490658220438e-05,
+      "loss": 0.7126,
+      "step": 2750
+    },
+    {
+      "epoch": 1.9686162624821684,
+      "grad_norm": 4.125,
+      "learning_rate": 4.968584280172206e-05,
+      "loss": 0.6598,
+      "step": 2760
+    },
+    {
+      "epoch": 1.9757489300998574,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.933679433281836e-05,
+      "loss": 0.7039,
+      "step": 2770
+    },
+    {
+      "epoch": 1.9828815977175465,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.8987778187623245e-05,
+      "loss": 0.6858,
+      "step": 2780
+    },
+    {
+      "epoch": 1.9900142653352355,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.863881137669123e-05,
+      "loss": 0.7924,
+      "step": 2790
+    },
+    {
+      "epoch": 1.9971469329529246,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8289910908172376e-05,
+      "loss": 0.7606,
+      "step": 2800
+    },
+    {
+      "epoch": 2.0042796005706136,
+      "grad_norm": 5.0,
+      "learning_rate": 4.794109378698327e-05,
+      "loss": 0.7584,
+      "step": 2810
+    },
+    {
+      "epoch": 2.011412268188302,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.7592377013978306e-05,
+      "loss": 0.6966,
+      "step": 2820
+    },
+    {
+      "epoch": 2.0185449358059913,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.724377758512103e-05,
+      "loss": 0.8236,
+      "step": 2830
+    },
+    {
+      "epoch": 2.0256776034236803,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.68953124906558e-05,
+      "loss": 0.7729,
+      "step": 2840
+    },
+    {
+      "epoch": 2.0328102710413694,
+      "grad_norm": 4.875,
+      "learning_rate": 4.654699871427971e-05,
+      "loss": 0.6784,
+      "step": 2850
+    },
+    {
+      "epoch": 2.0399429386590584,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.619885323231484e-05,
+      "loss": 0.8262,
+      "step": 2860
+    },
+    {
+      "epoch": 2.0470756062767475,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.58508930128808e-05,
+      "loss": 0.7451,
+      "step": 2870
+    },
+    {
+      "epoch": 2.0542082738944365,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.550313501506781e-05,
+      "loss": 0.7622,
+      "step": 2880
+    },
+    {
+      "epoch": 2.0613409415121255,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.515559618811005e-05,
+      "loss": 0.6851,
+      "step": 2890
+    },
+    {
+      "epoch": 2.0684736091298146,
+      "grad_norm": 4.375,
+      "learning_rate": 4.4808293470559643e-05,
+      "loss": 0.7706,
+      "step": 2900
+    },
+    {
+      "epoch": 2.0756062767475036,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.446124378946107e-05,
+      "loss": 0.7432,
+      "step": 2910
+    },
+    {
+      "epoch": 2.0827389443651927,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.411446405952618e-05,
+      "loss": 0.7158,
+      "step": 2920
+    },
+    {
+      "epoch": 2.0898716119828817,
+      "grad_norm": 4.375,
+      "learning_rate": 4.376797118230979e-05,
+      "loss": 0.7035,
+      "step": 2930
+    },
+    {
+      "epoch": 2.097004279600571,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.342178204538588e-05,
+      "loss": 0.7611,
+      "step": 2940
+    },
+    {
+      "epoch": 2.10413694721826,
+      "grad_norm": 4.625,
+      "learning_rate": 4.307591352152459e-05,
+      "loss": 0.6861,
+      "step": 2950
+    },
+    {
+      "epoch": 2.1112696148359484,
+      "grad_norm": 5.875,
+      "learning_rate": 4.273038246786986e-05,
+      "loss": 0.8394,
+      "step": 2960
+    },
+    {
+      "epoch": 2.1184022824536375,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.238520572511773e-05,
+      "loss": 0.6647,
+      "step": 2970
+    },
+    {
+      "epoch": 2.1255349500713265,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.2040400116695666e-05,
+      "loss": 0.7305,
+      "step": 2980
+    },
+    {
+      "epoch": 2.1326676176890156,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.169598244794261e-05,
+      "loss": 0.7693,
+      "step": 2990
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.135196950528982e-05,
+      "loss": 0.7392,
+      "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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-3000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af7cf7c3c741add5e8996639d61d87fe74bb40ad0faa1b29901130f4b3a60fc3
+size 6353

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86aa8a0a858b4170a0a0701c1871f2d27d9c905bd960a613015f68b2f98fbb58
+size 298041696

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcd48107e84feb43b1ddcf1ad31643963a6e81e3b99f3a74da919de2742d9444
+size 596170443

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/rng_state_0.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/rng_state_1.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/rng_state_2.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/rng_state_3.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/scheduler.pt

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/trainer_state.json

@@ -0,0 +1,3094 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.8530670470756063,
+  "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": 0.5201586889046041,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.climate_fever": 0.23997409265084552,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.dbpedia_entity": 0.34286887314902237,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fever": 0.6213496755517833,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.fiqa": 0.4542903465876373,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.hotpotqa": 0.6560051657630788,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nfcorpus": 0.3328835143785305,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.nq": 0.4570725655218753,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.quora": 0.832098944676056,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scidocs": 0.25908236715595256,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.scifact": 0.7391066781087818,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/nano_beir.webis_touche2020": 0.30431141846717047,
+      "step": 0
+    },
+    {
+      "epoch": 0,
+      "eval/avg": 0.4799335275762782,
+      "step": 0
+    },
+    {
+      "epoch": 0.007132667617689016,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8e-06,
+      "loss": 1.6901,
+      "step": 10
+    },
+    {
+      "epoch": 0.014265335235378032,
+      "grad_norm": 10.75,
+      "learning_rate": 3.8e-06,
+      "loss": 1.7675,
+      "step": 20
+    },
+    {
+      "epoch": 0.021398002853067047,
+      "grad_norm": 21.0,
+      "learning_rate": 5.8e-06,
+      "loss": 1.527,
+      "step": 30
+    },
+    {
+      "epoch": 0.028530670470756064,
+      "grad_norm": 19.0,
+      "learning_rate": 7.8e-06,
+      "loss": 1.7708,
+      "step": 40
+    },
+    {
+      "epoch": 0.03566333808844508,
+      "grad_norm": 17.125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.4093,
+      "step": 50
+    },
+    {
+      "epoch": 0.042796005706134094,
+      "grad_norm": 15.625,
+      "learning_rate": 1.18e-05,
+      "loss": 1.579,
+      "step": 60
+    },
+    {
+      "epoch": 0.04992867332382311,
+      "grad_norm": 14.0,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.4845,
+      "step": 70
+    },
+    {
+      "epoch": 0.05706134094151213,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.58e-05,
+      "loss": 1.598,
+      "step": 80
+    },
+    {
+      "epoch": 0.06419400855920114,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.78e-05,
+      "loss": 1.3226,
+      "step": 90
+    },
+    {
+      "epoch": 0.07132667617689016,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.4969,
+      "step": 100
+    },
+    {
+      "epoch": 0.07845934379457917,
+      "grad_norm": 13.125,
+      "learning_rate": 2.18e-05,
+      "loss": 1.3867,
+      "step": 110
+    },
+    {
+      "epoch": 0.08559201141226819,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.38e-05,
+      "loss": 1.1904,
+      "step": 120
+    },
+    {
+      "epoch": 0.09272467902995721,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.58e-05,
+      "loss": 1.1668,
+      "step": 130
+    },
+    {
+      "epoch": 0.09985734664764621,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.124,
+      "step": 140
+    },
+    {
+      "epoch": 0.10699001426533523,
+      "grad_norm": 9.75,
+      "learning_rate": 2.98e-05,
+      "loss": 1.1571,
+      "step": 150
+    },
+    {
+      "epoch": 0.11412268188302425,
+      "grad_norm": 12.25,
+      "learning_rate": 3.18e-05,
+      "loss": 1.2145,
+      "step": 160
+    },
+    {
+      "epoch": 0.12125534950071326,
+      "grad_norm": 10.625,
+      "learning_rate": 3.38e-05,
+      "loss": 1.155,
+      "step": 170
+    },
+    {
+      "epoch": 0.12838801711840228,
+      "grad_norm": 8.5,
+      "learning_rate": 3.58e-05,
+      "loss": 1.1513,
+      "step": 180
+    },
+    {
+      "epoch": 0.1355206847360913,
+      "grad_norm": 9.5,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.9803,
+      "step": 190
+    },
+    {
+      "epoch": 0.14265335235378032,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 1.0302,
+      "step": 200
+    },
+    {
+      "epoch": 0.14978601997146934,
+      "grad_norm": 9.25,
+      "learning_rate": 4.18e-05,
+      "loss": 0.9878,
+      "step": 210
+    },
+    {
+      "epoch": 0.15691868758915833,
+      "grad_norm": 10.25,
+      "learning_rate": 4.38e-05,
+      "loss": 0.9816,
+      "step": 220
+    },
+    {
+      "epoch": 0.16405135520684735,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.58e-05,
+      "loss": 1.0415,
+      "step": 230
+    },
+    {
+      "epoch": 0.17118402282453637,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.78e-05,
+      "loss": 1.1126,
+      "step": 240
+    },
+    {
+      "epoch": 0.1783166904422254,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 1.0127,
+      "step": 250
+    },
+    {
+      "epoch": 0.18544935805991442,
+      "grad_norm": 5.71875,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.9577,
+      "step": 260
+    },
+    {
+      "epoch": 0.19258202567760344,
+      "grad_norm": 6.21875,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.9518,
+      "step": 270
+    },
+    {
+      "epoch": 0.19971469329529243,
+      "grad_norm": 6.28125,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.7952,
+      "step": 280
+    },
+    {
+      "epoch": 0.20684736091298145,
+      "grad_norm": 6.875,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.9201,
+      "step": 290
+    },
+    {
+      "epoch": 0.21398002853067047,
+      "grad_norm": 6.09375,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.951,
+      "step": 300
+    },
+    {
+      "epoch": 0.2211126961483595,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.18e-05,
+      "loss": 0.9199,
+      "step": 310
+    },
+    {
+      "epoch": 0.2282453637660485,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.38e-05,
+      "loss": 0.8773,
+      "step": 320
+    },
+    {
+      "epoch": 0.23537803138373753,
+      "grad_norm": 4.875,
+      "learning_rate": 6.58e-05,
+      "loss": 0.9386,
+      "step": 330
+    },
+    {
+      "epoch": 0.24251069900142652,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.9364,
+      "step": 340
+    },
+    {
+      "epoch": 0.24964336661911554,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.98e-05,
+      "loss": 0.9053,
+      "step": 350
+    },
+    {
+      "epoch": 0.25677603423680456,
+      "grad_norm": 5.8125,
+      "learning_rate": 7.18e-05,
+      "loss": 0.8493,
+      "step": 360
+    },
+    {
+      "epoch": 0.26390870185449355,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.38e-05,
+      "loss": 0.8218,
+      "step": 370
+    },
+    {
+      "epoch": 0.2710413694721826,
+      "grad_norm": 5.625,
+      "learning_rate": 7.58e-05,
+      "loss": 0.8975,
+      "step": 380
+    },
+    {
+      "epoch": 0.2781740370898716,
+      "grad_norm": 5.1875,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.8891,
+      "step": 390
+    },
+    {
+      "epoch": 0.28530670470756064,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.98e-05,
+      "loss": 0.7831,
+      "step": 400
+    },
+    {
+      "epoch": 0.29243937232524964,
+      "grad_norm": 5.125,
+      "learning_rate": 8.18e-05,
+      "loss": 0.9017,
+      "step": 410
+    },
+    {
+      "epoch": 0.2995720399429387,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.38e-05,
+      "loss": 0.7839,
+      "step": 420
+    },
+    {
+      "epoch": 0.3067047075606277,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.58e-05,
+      "loss": 0.886,
+      "step": 430
+    },
+    {
+      "epoch": 0.31383737517831667,
+      "grad_norm": 6.1875,
+      "learning_rate": 8.78e-05,
+      "loss": 0.863,
+      "step": 440
+    },
+    {
+      "epoch": 0.3209700427960057,
+      "grad_norm": 4.625,
+      "learning_rate": 8.98e-05,
+      "loss": 0.884,
+      "step": 450
+    },
+    {
+      "epoch": 0.3281027104136947,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.9209,
+      "step": 460
+    },
+    {
+      "epoch": 0.33523537803138376,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.38e-05,
+      "loss": 0.8775,
+      "step": 470
+    },
+    {
+      "epoch": 0.34236804564907275,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.58e-05,
+      "loss": 0.7956,
+      "step": 480
+    },
+    {
+      "epoch": 0.34950071326676174,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.78e-05,
+      "loss": 0.8632,
+      "step": 490
+    },
+    {
+      "epoch": 0.3566333808844508,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.98e-05,
+      "loss": 0.8308,
+      "step": 500
+    },
+    {
+      "epoch": 0.3637660485021398,
+      "grad_norm": 5.25,
+      "learning_rate": 9.999901304280685e-05,
+      "loss": 0.884,
+      "step": 510
+    },
+    {
+      "epoch": 0.37089871611982883,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999560138895238e-05,
+      "loss": 0.8663,
+      "step": 520
+    },
+    {
+      "epoch": 0.3780313837375178,
+      "grad_norm": 5.25,
+      "learning_rate": 9.99897530200195e-05,
+      "loss": 0.8722,
+      "step": 530
+    },
+    {
+      "epoch": 0.38516405135520687,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.998146822104943e-05,
+      "loss": 0.8391,
+      "step": 540
+    },
+    {
+      "epoch": 0.39229671897289586,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.997074739583162e-05,
+      "loss": 0.7823,
+      "step": 550
+    },
+    {
+      "epoch": 0.39942938659058486,
+      "grad_norm": 4.875,
+      "learning_rate": 9.995759106688393e-05,
+      "loss": 0.8606,
+      "step": 560
+    },
+    {
+      "epoch": 0.4065620542082739,
+      "grad_norm": 5.625,
+      "learning_rate": 9.99419998754273e-05,
+      "loss": 0.8719,
+      "step": 570
+    },
+    {
+      "epoch": 0.4136947218259629,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.992397458135438e-05,
+      "loss": 0.8619,
+      "step": 580
+    },
+    {
+      "epoch": 0.42082738944365194,
+      "grad_norm": 5.0,
+      "learning_rate": 9.990351606319261e-05,
+      "loss": 0.8439,
+      "step": 590
+    },
+    {
+      "epoch": 0.42796005706134094,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.988062531806126e-05,
+      "loss": 0.886,
+      "step": 600
+    },
+    {
+      "epoch": 0.43509272467902993,
+      "grad_norm": 5.125,
+      "learning_rate": 9.985530346162299e-05,
+      "loss": 0.8346,
+      "step": 610
+    },
+    {
+      "epoch": 0.442225392296719,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.982755172802933e-05,
+      "loss": 0.8145,
+      "step": 620
+    },
+    {
+      "epoch": 0.44935805991440797,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.979737146986064e-05,
+      "loss": 0.8739,
+      "step": 630
+    },
+    {
+      "epoch": 0.456490727532097,
+      "grad_norm": 4.75,
+      "learning_rate": 9.976476415806013e-05,
+      "loss": 0.7826,
+      "step": 640
+    },
+    {
+      "epoch": 0.463623395149786,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.972973138186217e-05,
+      "loss": 0.8541,
+      "step": 650
+    },
+    {
+      "epoch": 0.47075606276747506,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.969227484871484e-05,
+      "loss": 0.8208,
+      "step": 660
+    },
+    {
+      "epoch": 0.47788873038516405,
+      "grad_norm": 5.375,
+      "learning_rate": 9.965239638419672e-05,
+      "loss": 0.7761,
+      "step": 670
+    },
+    {
+      "epoch": 0.48502139800285304,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.961009793192792e-05,
+      "loss": 0.8346,
+      "step": 680
+    },
+    {
+      "epoch": 0.4921540656205421,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.956538155347534e-05,
+      "loss": 0.9192,
+      "step": 690
+    },
+    {
+      "epoch": 0.4992867332382311,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.951824942825215e-05,
+      "loss": 0.8037,
+      "step": 700
+    },
+    {
+      "epoch": 0.5064194008559201,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.946870385341167e-05,
+      "loss": 0.7669,
+      "step": 710
+    },
+    {
+      "epoch": 0.5135520684736091,
+      "grad_norm": 6.875,
+      "learning_rate": 9.941674724373531e-05,
+      "loss": 0.7915,
+      "step": 720
+    },
+    {
+      "epoch": 0.5206847360912982,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.936238213151491e-05,
+      "loss": 0.7863,
+      "step": 730
+    },
+    {
+      "epoch": 0.5278174037089871,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.930561116642935e-05,
+      "loss": 0.7789,
+      "step": 740
+    },
+    {
+      "epoch": 0.5349500713266762,
+      "grad_norm": 4.0,
+      "learning_rate": 9.924643711541539e-05,
+      "loss": 0.7759,
+      "step": 750
+    },
+    {
+      "epoch": 0.5420827389443652,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.918486286253279e-05,
+      "loss": 0.8504,
+      "step": 760
+    },
+    {
+      "epoch": 0.5492154065620543,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.912089140882376e-05,
+      "loss": 0.809,
+      "step": 770
+    },
+    {
+      "epoch": 0.5563480741797432,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.90545258721667e-05,
+      "loss": 0.7721,
+      "step": 780
+    },
+    {
+      "epoch": 0.5634807417974322,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.898576948712426e-05,
+      "loss": 0.7319,
+      "step": 790
+    },
+    {
+      "epoch": 0.5706134094151213,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.891462560478562e-05,
+      "loss": 0.7726,
+      "step": 800
+    },
+    {
+      "epoch": 0.5777460770328102,
+      "grad_norm": 4.625,
+      "learning_rate": 9.884109769260325e-05,
+      "loss": 0.8522,
+      "step": 810
+    },
+    {
+      "epoch": 0.5848787446504993,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.876518933422386e-05,
+      "loss": 0.7761,
+      "step": 820
+    },
+    {
+      "epoch": 0.5920114122681883,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.868690422931372e-05,
+      "loss": 0.8702,
+      "step": 830
+    },
+    {
+      "epoch": 0.5991440798858774,
+      "grad_norm": 5.25,
+      "learning_rate": 9.860624619337844e-05,
+      "loss": 0.8636,
+      "step": 840
+    },
+    {
+      "epoch": 0.6062767475035663,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.852321915757687e-05,
+      "loss": 0.6974,
+      "step": 850
+    },
+    {
+      "epoch": 0.6134094151212554,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.843782716852963e-05,
+      "loss": 0.8017,
+      "step": 860
+    },
+    {
+      "epoch": 0.6205420827389444,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.835007438812177e-05,
+      "loss": 0.7799,
+      "step": 870
+    },
+    {
+      "epoch": 0.6276747503566333,
+      "grad_norm": 5.125,
+      "learning_rate": 9.82599650933e-05,
+      "loss": 0.7808,
+      "step": 880
+    },
+    {
+      "epoch": 0.6348074179743224,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.816750367586425e-05,
+      "loss": 0.8337,
+      "step": 890
+    },
+    {
+      "epoch": 0.6419400855920114,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.807269464225355e-05,
+      "loss": 0.8043,
+      "step": 900
+    },
+    {
+      "epoch": 0.6490727532097005,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.797554261332645e-05,
+      "loss": 0.8328,
+      "step": 910
+    },
+    {
+      "epoch": 0.6562054208273894,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.787605232413574e-05,
+      "loss": 0.7745,
+      "step": 920
+    },
+    {
+      "epoch": 0.6633380884450785,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.777422862369783e-05,
+      "loss": 0.745,
+      "step": 930
+    },
+    {
+      "epoch": 0.6704707560627675,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.767007647475617e-05,
+      "loss": 0.82,
+      "step": 940
+    },
+    {
+      "epoch": 0.6776034236804565,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.756360095353956e-05,
+      "loss": 0.8025,
+      "step": 950
+    },
+    {
+      "epoch": 0.6847360912981455,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.745480724951473e-05,
+      "loss": 0.84,
+      "step": 960
+    },
+    {
+      "epoch": 0.6918687589158345,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.73437006651333e-05,
+      "loss": 0.8552,
+      "step": 970
+    },
+    {
+      "epoch": 0.6990014265335235,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.723028661557345e-05,
+      "loss": 0.8082,
+      "step": 980
+    },
+    {
+      "epoch": 0.7061340941512125,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.711457062847595e-05,
+      "loss": 0.8265,
+      "step": 990
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "grad_norm": 4.875,
+      "learning_rate": 9.699655834367478e-05,
+      "loss": 0.8369,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.arguana": 0.47952715537701884,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.climate_fever": 0.22702207934204544,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.dbpedia_entity": 0.3287396364958694,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.fever": 0.5532628324042129,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.fiqa": 0.4160855038689836,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.hotpotqa": 0.6599716759215029,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.nfcorpus": 0.3449425020579428,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.nq": 0.39316532973343116,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.quora": 0.8295275476072758,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.scidocs": 0.2526832556333601,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.scifact": 0.7219633063518482,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/nano_beir.webis_touche2020": 0.2787587910246796,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7132667617689016,
+      "eval/avg": 0.4571374679848476,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7203994293865906,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.68762555129222e-05,
+      "loss": 0.8872,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7275320970042796,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.675366799960841e-05,
+      "loss": 0.8334,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7346647646219686,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.662880177847594e-05,
+      "loss": 0.8014,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7417974322396577,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.65016629353282e-05,
+      "loss": 0.848,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7489300998573466,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.637225766673307e-05,
+      "loss": 0.7203,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7560627674750356,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.624059227972075e-05,
+      "loss": 0.8405,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7631954350927247,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.610667319147647e-05,
+      "loss": 0.8418,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7703281027104137,
+      "grad_norm": 5.75,
+      "learning_rate": 9.597050692902765e-05,
+      "loss": 0.8853,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7774607703281027,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.583210012892582e-05,
+      "loss": 0.7733,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7845934379457917,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.569145953692315e-05,
+      "loss": 0.7698,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7917261055634808,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.55485920076437e-05,
+      "loss": 0.7759,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7988587731811697,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.540350450424926e-05,
+      "loss": 0.8025,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8059914407988588,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.525620409810007e-05,
+      "loss": 0.8364,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8131241084165478,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.510669796841014e-05,
+      "loss": 0.7523,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8202567760342369,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.495499340189728e-05,
+      "loss": 0.7333,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8273894436519258,
+      "grad_norm": 4.875,
+      "learning_rate": 9.480109779242805e-05,
+      "loss": 0.7325,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8345221112696148,
+      "grad_norm": 5.0,
+      "learning_rate": 9.464501864065735e-05,
+      "loss": 0.7996,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8416547788873039,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.448676355366282e-05,
+      "loss": 0.7946,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8487874465049928,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.432634024457414e-05,
+      "loss": 0.9133,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8559201141226819,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.416375653219709e-05,
+      "loss": 0.7218,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8630527817403709,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.399902034063242e-05,
+      "loss": 0.8453,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8701854493580599,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.383213969888972e-05,
+      "loss": 0.7261,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8773181169757489,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.366312274049602e-05,
+      "loss": 0.6459,
+      "step": 1230
+    },
+    {
+      "epoch": 0.884450784593438,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.34919777030994e-05,
+      "loss": 0.7202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.891583452211127,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.331871292806759e-05,
+      "loss": 0.7874,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8987161198288159,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.314333686008125e-05,
+      "loss": 0.8524,
+      "step": 1260
+    },
+    {
+      "epoch": 0.905848787446505,
+      "grad_norm": 5.75,
+      "learning_rate": 9.296585804672252e-05,
+      "loss": 0.7854,
+      "step": 1270
+    },
+    {
+      "epoch": 0.912981455064194,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.278628513805838e-05,
+      "loss": 0.8386,
+      "step": 1280
+    },
+    {
+      "epoch": 0.920114122681883,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.260462688621905e-05,
+      "loss": 0.8279,
+      "step": 1290
+    },
+    {
+      "epoch": 0.927246790299572,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.242089214497145e-05,
+      "loss": 0.8345,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9343794579172611,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.223508986928766e-05,
+      "loss": 0.7596,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9415121255349501,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.204722911490846e-05,
+      "loss": 0.6704,
+      "step": 1320
+    },
+    {
+      "epoch": 0.948644793152639,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.1857319037902e-05,
+      "loss": 0.7792,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9557774607703281,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.16653688942175e-05,
+      "loss": 0.813,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9629101283880172,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.147138803923416e-05,
+      "loss": 0.834,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9700427960057061,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.127538592730519e-05,
+      "loss": 0.8299,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9771754636233951,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.107737211129701e-05,
+      "loss": 0.7407,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9843081312410842,
+      "grad_norm": 5.0,
+      "learning_rate": 9.087735624212365e-05,
+      "loss": 0.8821,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9914407988587732,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.06753480682764e-05,
+      "loss": 0.8595,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9985734664764622,
+      "grad_norm": 4.5,
+      "learning_rate": 9.047135743534868e-05,
+      "loss": 0.869,
+      "step": 1400
+    },
+    {
+      "epoch": 1.005706134094151,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.02653942855561e-05,
+      "loss": 0.8397,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0128388017118402,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.005746865725207e-05,
+      "loss": 0.8047,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0199714693295292,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.984759068443831e-05,
+      "loss": 0.72,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0271041369472182,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.963577059627118e-05,
+      "loss": 0.7724,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0342368045649073,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.942201871656291e-05,
+      "loss": 0.8014,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0413694721825963,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.920634546327858e-05,
+      "loss": 0.8532,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0485021398002854,
+      "grad_norm": 3.75,
+      "learning_rate": 8.898876134802826e-05,
+      "loss": 0.7285,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0556348074179742,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.87692769755548e-05,
+      "loss": 0.8574,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0627674750356633,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.854790304321681e-05,
+      "loss": 0.76,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0699001426533523,
+      "grad_norm": 4.75,
+      "learning_rate": 8.832465034046749e-05,
+      "loss": 0.8441,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0770328102710414,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.80995297483286e-05,
+      "loss": 0.764,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0841654778887304,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.787255223886019e-05,
+      "loss": 0.7432,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0912981455064195,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.764372887462586e-05,
+      "loss": 0.7894,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0984308131241085,
+      "grad_norm": 4.5,
+      "learning_rate": 8.741307080815357e-05,
+      "loss": 0.761,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1055634807417973,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.718058928139205e-05,
+      "loss": 0.9432,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1126961483594864,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.694629562516294e-05,
+      "loss": 0.7863,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1198288159771754,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.671020125860851e-05,
+      "loss": 0.7732,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1269614835948645,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.647231768863512e-05,
+      "loss": 0.7651,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1340941512125535,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.623265650935234e-05,
+      "loss": 0.8021,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1412268188302426,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.599122940150795e-05,
+      "loss": 0.8561,
+      "step": 1600
+    },
+    {
+      "epoch": 1.1483594864479316,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.574804813191858e-05,
+      "loss": 0.7551,
+      "step": 1610
+    },
+    {
+      "epoch": 1.1554921540656204,
+      "grad_norm": 5.375,
+      "learning_rate": 8.550312455289625e-05,
+      "loss": 0.8045,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1626248216833095,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.525647060167063e-05,
+      "loss": 0.7145,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1697574893009985,
+      "grad_norm": 4.125,
+      "learning_rate": 8.500809829980734e-05,
+      "loss": 0.8007,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1768901569186876,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.4758019752622e-05,
+      "loss": 0.8009,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1840228245363766,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.450624714859016e-05,
+      "loss": 0.833,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1911554921540657,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.425279275875336e-05,
+      "loss": 0.8149,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1982881597717547,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.399766893612096e-05,
+      "loss": 0.808,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2054208273894436,
+      "grad_norm": 4.625,
+      "learning_rate": 8.374088811506819e-05,
+      "loss": 0.7549,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2125534950071326,
+      "grad_norm": 5.71875,
+      "learning_rate": 8.348246281072998e-05,
+      "loss": 0.7755,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2196861626248217,
+      "grad_norm": 4.5625,
+      "learning_rate": 8.322240561839109e-05,
+      "loss": 0.8038,
+      "step": 1710
+    },
+    {
+      "epoch": 1.2268188302425107,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.296072921287217e-05,
+      "loss": 0.7207,
+      "step": 1720
+    },
+    {
+      "epoch": 1.2339514978601998,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.269744634791208e-05,
+      "loss": 0.7994,
+      "step": 1730
+    },
+    {
+      "epoch": 1.2410841654778888,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.243256985554621e-05,
+      "loss": 0.7989,
+      "step": 1740
+    },
+    {
+      "epoch": 1.2482168330955776,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.21661126454811e-05,
+      "loss": 0.6516,
+      "step": 1750
+    },
+    {
+      "epoch": 1.2553495007132667,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.189808770446528e-05,
+      "loss": 0.7919,
+      "step": 1760
+    },
+    {
+      "epoch": 1.2624821683309557,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.9372,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2696148359486448,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.135738695798376e-05,
+      "loss": 0.7683,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2767475035663338,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.108473750550965e-05,
+      "loss": 0.737,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2838801711840229,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.081057302678352e-05,
+      "loss": 0.7507,
+      "step": 1800
+    },
+    {
+      "epoch": 1.291012838801712,
+      "grad_norm": 4.875,
+      "learning_rate": 8.053490688419533e-05,
+      "loss": 0.7016,
+      "step": 1810
+    },
+    {
+      "epoch": 1.298145506419401,
+      "grad_norm": 4.5,
+      "learning_rate": 8.02577525133239e-05,
+      "loss": 0.7151,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3052781740370898,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.997912342228232e-05,
+      "loss": 0.8559,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3124108416547788,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.969903319105935e-05,
+      "loss": 0.7118,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3195435092724679,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.941749547085777e-05,
+      "loss": 0.7855,
+      "step": 1850
+    },
+    {
+      "epoch": 1.326676176890157,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.913452398342881e-05,
+      "loss": 0.7438,
+      "step": 1860
+    },
+    {
+      "epoch": 1.333808844507846,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.885013252040359e-05,
+      "loss": 0.7018,
+      "step": 1870
+    },
+    {
+      "epoch": 1.340941512125535,
+      "grad_norm": 6.9375,
+      "learning_rate": 7.856433494262077e-05,
+      "loss": 0.7083,
+      "step": 1880
+    },
+    {
+      "epoch": 1.3480741797432239,
+      "grad_norm": 4.875,
+      "learning_rate": 7.827714517945115e-05,
+      "loss": 0.8265,
+      "step": 1890
+    },
+    {
+      "epoch": 1.355206847360913,
+      "grad_norm": 5.5625,
+      "learning_rate": 7.798857722811856e-05,
+      "loss": 0.8735,
+      "step": 1900
+    },
+    {
+      "epoch": 1.362339514978602,
+      "grad_norm": 5.71875,
+      "learning_rate": 7.769864515301787e-05,
+      "loss": 0.7256,
+      "step": 1910
+    },
+    {
+      "epoch": 1.369472182596291,
+      "grad_norm": 4.53125,
+      "learning_rate": 7.740736308502938e-05,
+      "loss": 0.7472,
+      "step": 1920
+    },
+    {
+      "epoch": 1.37660485021398,
+      "grad_norm": 5.875,
+      "learning_rate": 7.711474522083015e-05,
+      "loss": 0.7601,
+      "step": 1930
+    },
+    {
+      "epoch": 1.383737517831669,
+      "grad_norm": 5.75,
+      "learning_rate": 7.682080582220206e-05,
+      "loss": 0.8857,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3908701854493581,
+      "grad_norm": 6.40625,
+      "learning_rate": 7.65255592153367e-05,
+      "loss": 0.7296,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3980028530670472,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.622901979013717e-05,
+      "loss": 0.7331,
+      "step": 1960
+    },
+    {
+      "epoch": 1.405135520684736,
+      "grad_norm": 5.09375,
+      "learning_rate": 7.593120199951671e-05,
+      "loss": 0.8462,
+      "step": 1970
+    },
+    {
+      "epoch": 1.412268188302425,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.563212035869425e-05,
+      "loss": 0.7307,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4194008559201141,
+      "grad_norm": 4.90625,
+      "learning_rate": 7.533178944448705e-05,
+      "loss": 0.7711,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.503022389460014e-05,
+      "loss": 0.7485,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.arguana": 0.47494654864753366,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.climate_fever": 0.2496059194905162,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.dbpedia_entity": 0.3371368383795953,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.fever": 0.6065422117591313,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.fiqa": 0.4135337243548038,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.hotpotqa": 0.6804056480114672,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.nfcorpus": 0.35238501477624823,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.nq": 0.39770361480597743,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.quora": 0.8217479889085385,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.scidocs": 0.2592573629546752,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.scifact": 0.7479618950231317,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/nano_beir.webis_touche2020": 0.2719065772753103,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4265335235378032,
+      "eval/avg": 0.467761112032244,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4336661911554922,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.472743840691299e-05,
+      "loss": 0.8505,
+      "step": 2010
+    },
+    {
+      "epoch": 1.440798858773181,
+      "grad_norm": 6.15625,
+      "learning_rate": 7.44234477387631e-05,
+      "loss": 0.7449,
+      "step": 2020
+    },
+    {
+      "epoch": 1.44793152639087,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.411826670622676e-05,
+      "loss": 0.6961,
+      "step": 2030
+    },
+    {
+      "epoch": 1.4550641940085591,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.381191018339696e-05,
+      "loss": 0.8105,
+      "step": 2040
+    },
+    {
+      "epoch": 1.4621968616262482,
+      "grad_norm": 5.59375,
+      "learning_rate": 7.350439310165842e-05,
+      "loss": 0.8111,
+      "step": 2050
+    },
+    {
+      "epoch": 1.4693295292439372,
+      "grad_norm": 4.625,
+      "learning_rate": 7.319573044895986e-05,
+      "loss": 0.7972,
+      "step": 2060
+    },
+    {
+      "epoch": 1.4764621968616263,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.28859372690835e-05,
+      "loss": 0.7745,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4835948644793153,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.257502866091192e-05,
+      "loss": 0.7904,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4907275320970044,
+      "grad_norm": 4.875,
+      "learning_rate": 7.226301977769199e-05,
+      "loss": 0.7724,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4978601997146934,
+      "grad_norm": 5.59375,
+      "learning_rate": 7.194992582629654e-05,
+      "loss": 0.7693,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5049928673323825,
+      "grad_norm": 4.75,
+      "learning_rate": 7.163576206648304e-05,
+      "loss": 0.6929,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5121255349500713,
+      "grad_norm": 5.78125,
+      "learning_rate": 7.132054381014995e-05,
+      "loss": 0.8008,
+      "step": 2120
+    },
+    {
+      "epoch": 1.5192582025677603,
+      "grad_norm": 6.28125,
+      "learning_rate": 7.100428642059033e-05,
+      "loss": 0.8399,
+      "step": 2130
+    },
+    {
+      "epoch": 1.5263908701854494,
+      "grad_norm": 6.3125,
+      "learning_rate": 7.068700531174319e-05,
+      "loss": 0.7399,
+      "step": 2140
+    },
+    {
+      "epoch": 1.5335235378031382,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.036871594744217e-05,
+      "loss": 0.7796,
+      "step": 2150
+    },
+    {
+      "epoch": 1.5406562054208273,
+      "grad_norm": 5.34375,
+      "learning_rate": 7.004943384066187e-05,
+      "loss": 0.8685,
+      "step": 2160
+    },
+    {
+      "epoch": 1.5477888730385163,
+      "grad_norm": 4.8125,
+      "learning_rate": 6.97291745527617e-05,
+      "loss": 0.7448,
+      "step": 2170
+    },
+    {
+      "epoch": 1.5549215406562054,
+      "grad_norm": 6.09375,
+      "learning_rate": 6.940795369272754e-05,
+      "loss": 0.8527,
+      "step": 2180
+    },
+    {
+      "epoch": 1.5620542082738944,
+      "grad_norm": 5.15625,
+      "learning_rate": 6.908578691641092e-05,
+      "loss": 0.8704,
+      "step": 2190
+    },
+    {
+      "epoch": 1.5691868758915835,
+      "grad_norm": 4.3125,
+      "learning_rate": 6.876268992576604e-05,
+      "loss": 0.7033,
+      "step": 2200
+    },
+    {
+      "epoch": 1.5763195435092725,
+      "grad_norm": 5.6875,
+      "learning_rate": 6.843867846808438e-05,
+      "loss": 0.7664,
+      "step": 2210
+    },
+    {
+      "epoch": 1.5834522111269616,
+      "grad_norm": 4.5625,
+      "learning_rate": 6.811376833522729e-05,
+      "loss": 0.7902,
+      "step": 2220
+    },
+    {
+      "epoch": 1.5905848787446506,
+      "grad_norm": 4.6875,
+      "learning_rate": 6.778797536285624e-05,
+      "loss": 0.7585,
+      "step": 2230
+    },
+    {
+      "epoch": 1.5977175463623396,
+      "grad_norm": 5.375,
+      "learning_rate": 6.746131542966112e-05,
+      "loss": 0.8124,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6048502139800287,
+      "grad_norm": 4.59375,
+      "learning_rate": 6.713380445658618e-05,
+      "loss": 0.7936,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6119828815977175,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.680545840605423e-05,
+      "loss": 0.6785,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6191155492154066,
+      "grad_norm": 5.875,
+      "learning_rate": 6.647629328118851e-05,
+      "loss": 0.8321,
+      "step": 2270
+    },
+    {
+      "epoch": 1.6262482168330956,
+      "grad_norm": 5.125,
+      "learning_rate": 6.614632512503288e-05,
+      "loss": 0.8184,
+      "step": 2280
+    },
+    {
+      "epoch": 1.6333808844507844,
+      "grad_norm": 4.5,
+      "learning_rate": 6.58155700197697e-05,
+      "loss": 0.7187,
+      "step": 2290
+    },
+    {
+      "epoch": 1.6405135520684735,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.548404408593621e-05,
+      "loss": 0.6873,
+      "step": 2300
+    },
+    {
+      "epoch": 1.6476462196861625,
+      "grad_norm": 4.96875,
+      "learning_rate": 6.515176348163871e-05,
+      "loss": 0.7537,
+      "step": 2310
+    },
+    {
+      "epoch": 1.6547788873038516,
+      "grad_norm": 4.71875,
+      "learning_rate": 6.481874440176506e-05,
+      "loss": 0.8058,
+      "step": 2320
+    },
+    {
+      "epoch": 1.6619115549215406,
+      "grad_norm": 5.375,
+      "learning_rate": 6.448500307719538e-05,
+      "loss": 0.8117,
+      "step": 2330
+    },
+    {
+      "epoch": 1.6690442225392297,
+      "grad_norm": 5.15625,
+      "learning_rate": 6.415055577401102e-05,
+      "loss": 0.7938,
+      "step": 2340
+    },
+    {
+      "epoch": 1.6761768901569187,
+      "grad_norm": 6.15625,
+      "learning_rate": 6.381541879270169e-05,
+      "loss": 0.7898,
+      "step": 2350
+    },
+    {
+      "epoch": 1.6833095577746078,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.347960846737106e-05,
+      "loss": 0.7896,
+      "step": 2360
+    },
+    {
+      "epoch": 1.6904422253922968,
+      "grad_norm": 4.28125,
+      "learning_rate": 6.31431411649406e-05,
+      "loss": 0.6994,
+      "step": 2370
+    },
+    {
+      "epoch": 1.6975748930099859,
+      "grad_norm": 5.71875,
+      "learning_rate": 6.280603328435198e-05,
+      "loss": 0.7416,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7047075606276747,
+      "grad_norm": 4.8125,
+      "learning_rate": 6.24683012557677e-05,
+      "loss": 0.7143,
+      "step": 2390
+    },
+    {
+      "epoch": 1.7118402282453637,
+      "grad_norm": 5.28125,
+      "learning_rate": 6.212996153977037e-05,
+      "loss": 0.6699,
+      "step": 2400
+    },
+    {
+      "epoch": 1.7189728958630528,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.179103062656042e-05,
+      "loss": 0.7084,
+      "step": 2410
+    },
+    {
+      "epoch": 1.7261055634807418,
+      "grad_norm": 5.125,
+      "learning_rate": 6.145152503515239e-05,
+      "loss": 0.7638,
+      "step": 2420
+    },
+    {
+      "epoch": 1.7332382310984307,
+      "grad_norm": 5.0,
+      "learning_rate": 6.111146131256983e-05,
+      "loss": 0.7875,
+      "step": 2430
+    },
+    {
+      "epoch": 1.7403708987161197,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.077085603303882e-05,
+      "loss": 0.8864,
+      "step": 2440
+    },
+    {
+      "epoch": 1.7475035663338088,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.042972579718019e-05,
+      "loss": 0.7354,
+      "step": 2450
+    },
+    {
+      "epoch": 1.7546362339514978,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.008808723120035e-05,
+      "loss": 0.7581,
+      "step": 2460
+    },
+    {
+      "epoch": 1.7617689015691869,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.974595698608103e-05,
+      "loss": 0.738,
+      "step": 2470
+    },
+    {
+      "epoch": 1.768901569186876,
+      "grad_norm": 4.53125,
+      "learning_rate": 5.9403351736767696e-05,
+      "loss": 0.6686,
+      "step": 2480
+    },
+    {
+      "epoch": 1.776034236804565,
+      "grad_norm": 5.125,
+      "learning_rate": 5.906028818135687e-05,
+      "loss": 0.6354,
+      "step": 2490
+    },
+    {
+      "epoch": 1.783166904422254,
+      "grad_norm": 5.125,
+      "learning_rate": 5.8716783040282244e-05,
+      "loss": 0.7869,
+      "step": 2500
+    },
+    {
+      "epoch": 1.790299572039943,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.837285305549978e-05,
+      "loss": 0.7463,
+      "step": 2510
+    },
+    {
+      "epoch": 1.797432239657632,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.8028514989671724e-05,
+      "loss": 0.7524,
+      "step": 2520
+    },
+    {
+      "epoch": 1.804564907275321,
+      "grad_norm": 6.15625,
+      "learning_rate": 5.7683785625349616e-05,
+      "loss": 0.7558,
+      "step": 2530
+    },
+    {
+      "epoch": 1.81169757489301,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.733868176415633e-05,
+      "loss": 0.8002,
+      "step": 2540
+    },
+    {
+      "epoch": 1.818830242510699,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.699322022596722e-05,
+      "loss": 0.7829,
+      "step": 2550
+    },
+    {
+      "epoch": 1.825962910128388,
+      "grad_norm": 4.375,
+      "learning_rate": 5.664741784809022e-05,
+      "loss": 0.7097,
+      "step": 2560
+    },
+    {
+      "epoch": 1.833095577746077,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.630129148444543e-05,
+      "loss": 0.6793,
+      "step": 2570
+    },
+    {
+      "epoch": 1.840228245363766,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.595485800474349e-05,
+      "loss": 0.7081,
+      "step": 2580
+    },
+    {
+      "epoch": 1.847360912981455,
+      "grad_norm": 4.25,
+      "learning_rate": 5.560813429366345e-05,
+      "loss": 0.911,
+      "step": 2590
+    },
+    {
+      "epoch": 1.854493580599144,
+      "grad_norm": 5.5625,
+      "learning_rate": 5.5261137250029835e-05,
+      "loss": 0.8128,
+      "step": 2600
+    },
+    {
+      "epoch": 1.861626248216833,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.4913883785988993e-05,
+      "loss": 0.7784,
+      "step": 2610
+    },
+    {
+      "epoch": 1.8687589158345221,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.456639082618489e-05,
+      "loss": 0.6765,
+      "step": 2620
+    },
+    {
+      "epoch": 1.8758915834522112,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.421867530693414e-05,
+      "loss": 0.7464,
+      "step": 2630
+    },
+    {
+      "epoch": 1.8830242510699002,
+      "grad_norm": 4.4375,
+      "learning_rate": 5.3870754175400595e-05,
+      "loss": 0.7452,
+      "step": 2640
+    },
+    {
+      "epoch": 1.8901569186875893,
+      "grad_norm": 4.40625,
+      "learning_rate": 5.352264438876935e-05,
+      "loss": 0.7475,
+      "step": 2650
+    },
+    {
+      "epoch": 1.8972895863052783,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.3174362913420306e-05,
+      "loss": 0.761,
+      "step": 2660
+    },
+    {
+      "epoch": 1.9044222539229672,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.2825926724101236e-05,
+      "loss": 0.7183,
+      "step": 2670
+    },
+    {
+      "epoch": 1.9115549215406562,
+      "grad_norm": 4.875,
+      "learning_rate": 5.247735280310041e-05,
+      "loss": 0.756,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9186875891583453,
+      "grad_norm": 4.8125,
+      "learning_rate": 5.212865813941899e-05,
+      "loss": 0.6997,
+      "step": 2690
+    },
+    {
+      "epoch": 1.925820256776034,
+      "grad_norm": 4.75,
+      "learning_rate": 5.1779859727942924e-05,
+      "loss": 0.7978,
+      "step": 2700
+    },
+    {
+      "epoch": 1.9329529243937231,
+      "grad_norm": 4.5,
+      "learning_rate": 5.143097456861474e-05,
+      "loss": 0.7099,
+      "step": 2710
+    },
+    {
+      "epoch": 1.9400855920114122,
+      "grad_norm": 4.8125,
+      "learning_rate": 5.1082019665604895e-05,
+      "loss": 0.7029,
+      "step": 2720
+    },
+    {
+      "epoch": 1.9472182596291012,
+      "grad_norm": 5.875,
+      "learning_rate": 5.073301202648304e-05,
+      "loss": 0.7328,
+      "step": 2730
+    },
+    {
+      "epoch": 1.9543509272467903,
+      "grad_norm": 3.875,
+      "learning_rate": 5.0383968661389146e-05,
+      "loss": 0.7982,
+      "step": 2740
+    },
+    {
+      "epoch": 1.9614835948644793,
+      "grad_norm": 4.71875,
+      "learning_rate": 5.003490658220438e-05,
+      "loss": 0.7126,
+      "step": 2750
+    },
+    {
+      "epoch": 1.9686162624821684,
+      "grad_norm": 4.125,
+      "learning_rate": 4.968584280172206e-05,
+      "loss": 0.6598,
+      "step": 2760
+    },
+    {
+      "epoch": 1.9757489300998574,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.933679433281836e-05,
+      "loss": 0.7039,
+      "step": 2770
+    },
+    {
+      "epoch": 1.9828815977175465,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.8987778187623245e-05,
+      "loss": 0.6858,
+      "step": 2780
+    },
+    {
+      "epoch": 1.9900142653352355,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.863881137669123e-05,
+      "loss": 0.7924,
+      "step": 2790
+    },
+    {
+      "epoch": 1.9971469329529246,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8289910908172376e-05,
+      "loss": 0.7606,
+      "step": 2800
+    },
+    {
+      "epoch": 2.0042796005706136,
+      "grad_norm": 5.0,
+      "learning_rate": 4.794109378698327e-05,
+      "loss": 0.7584,
+      "step": 2810
+    },
+    {
+      "epoch": 2.011412268188302,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.7592377013978306e-05,
+      "loss": 0.6966,
+      "step": 2820
+    },
+    {
+      "epoch": 2.0185449358059913,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.724377758512103e-05,
+      "loss": 0.8236,
+      "step": 2830
+    },
+    {
+      "epoch": 2.0256776034236803,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.68953124906558e-05,
+      "loss": 0.7729,
+      "step": 2840
+    },
+    {
+      "epoch": 2.0328102710413694,
+      "grad_norm": 4.875,
+      "learning_rate": 4.654699871427971e-05,
+      "loss": 0.6784,
+      "step": 2850
+    },
+    {
+      "epoch": 2.0399429386590584,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.619885323231484e-05,
+      "loss": 0.8262,
+      "step": 2860
+    },
+    {
+      "epoch": 2.0470756062767475,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.58508930128808e-05,
+      "loss": 0.7451,
+      "step": 2870
+    },
+    {
+      "epoch": 2.0542082738944365,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.550313501506781e-05,
+      "loss": 0.7622,
+      "step": 2880
+    },
+    {
+      "epoch": 2.0613409415121255,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.515559618811005e-05,
+      "loss": 0.6851,
+      "step": 2890
+    },
+    {
+      "epoch": 2.0684736091298146,
+      "grad_norm": 4.375,
+      "learning_rate": 4.4808293470559643e-05,
+      "loss": 0.7706,
+      "step": 2900
+    },
+    {
+      "epoch": 2.0756062767475036,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.446124378946107e-05,
+      "loss": 0.7432,
+      "step": 2910
+    },
+    {
+      "epoch": 2.0827389443651927,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.411446405952618e-05,
+      "loss": 0.7158,
+      "step": 2920
+    },
+    {
+      "epoch": 2.0898716119828817,
+      "grad_norm": 4.375,
+      "learning_rate": 4.376797118230979e-05,
+      "loss": 0.7035,
+      "step": 2930
+    },
+    {
+      "epoch": 2.097004279600571,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.342178204538588e-05,
+      "loss": 0.7611,
+      "step": 2940
+    },
+    {
+      "epoch": 2.10413694721826,
+      "grad_norm": 4.625,
+      "learning_rate": 4.307591352152459e-05,
+      "loss": 0.6861,
+      "step": 2950
+    },
+    {
+      "epoch": 2.1112696148359484,
+      "grad_norm": 5.875,
+      "learning_rate": 4.273038246786986e-05,
+      "loss": 0.8394,
+      "step": 2960
+    },
+    {
+      "epoch": 2.1184022824536375,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.238520572511773e-05,
+      "loss": 0.6647,
+      "step": 2970
+    },
+    {
+      "epoch": 2.1255349500713265,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.2040400116695666e-05,
+      "loss": 0.7305,
+      "step": 2980
+    },
+    {
+      "epoch": 2.1326676176890156,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.169598244794261e-05,
+      "loss": 0.7693,
+      "step": 2990
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.135196950528982e-05,
+      "loss": 0.7392,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.arguana": 0.46876474079443375,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.climate_fever": 0.24845534369031178,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.dbpedia_entity": 0.3518699079588141,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.fever": 0.5976831251710185,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.fiqa": 0.40792762165216695,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.hotpotqa": 0.6722148076317351,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.nfcorpus": 0.34539657615301445,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.nq": 0.38890377559243106,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.quora": 0.826494799130688,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.scidocs": 0.2518276219672329,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.scifact": 0.7707161514794307,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/nano_beir.webis_touche2020": 0.2854043806246762,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1398002853067046,
+      "eval/avg": 0.46797157098716285,
+      "step": 3000
+    },
+    {
+      "epoch": 2.1469329529243937,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.100837805544279e-05,
+      "loss": 0.7768,
+      "step": 3010
+    },
+    {
+      "epoch": 2.1540656205420827,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.0665224844564053e-05,
+      "loss": 0.6445,
+      "step": 3020
+    },
+    {
+      "epoch": 2.1611982881597718,
+      "grad_norm": 4.5,
+      "learning_rate": 4.032252659745699e-05,
+      "loss": 0.7335,
+      "step": 3030
+    },
+    {
+      "epoch": 2.168330955777461,
+      "grad_norm": 4.59375,
+      "learning_rate": 3.9980300016750694e-05,
+      "loss": 0.742,
+      "step": 3040
+    },
+    {
+      "epoch": 2.17546362339515,
+      "grad_norm": 4.34375,
+      "learning_rate": 3.963856178208588e-05,
+      "loss": 0.7422,
+      "step": 3050
+    },
+    {
+      "epoch": 2.182596291012839,
+      "grad_norm": 5.125,
+      "learning_rate": 3.9297328549302e-05,
+      "loss": 0.7333,
+      "step": 3060
+    },
+    {
+      "epoch": 2.189728958630528,
+      "grad_norm": 4.5625,
+      "learning_rate": 3.895661694962542e-05,
+      "loss": 0.8071,
+      "step": 3070
+    },
+    {
+      "epoch": 2.196861626248217,
+      "grad_norm": 5.5,
+      "learning_rate": 3.8616443588858805e-05,
+      "loss": 0.7712,
+      "step": 3080
+    },
+    {
+      "epoch": 2.2039942938659056,
+      "grad_norm": 4.53125,
+      "learning_rate": 3.827682504657187e-05,
+      "loss": 0.6875,
+      "step": 3090
+    },
+    {
+      "epoch": 2.2111269614835947,
+      "grad_norm": 5.0,
+      "learning_rate": 3.7937777875293244e-05,
+      "loss": 0.7984,
+      "step": 3100
+    },
+    {
+      "epoch": 2.2182596291012837,
+      "grad_norm": 4.78125,
+      "learning_rate": 3.759931859970374e-05,
+      "loss": 0.7392,
+      "step": 3110
+    },
+    {
+      "epoch": 2.2253922967189728,
+      "grad_norm": 4.375,
+      "learning_rate": 3.72614637158309e-05,
+      "loss": 0.796,
+      "step": 3120
+    },
+    {
+      "epoch": 2.232524964336662,
+      "grad_norm": 4.75,
+      "learning_rate": 3.692422969024516e-05,
+      "loss": 0.7527,
+      "step": 3130
+    },
+    {
+      "epoch": 2.239657631954351,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.6587632959257165e-05,
+      "loss": 0.7767,
+      "step": 3140
+    },
+    {
+      "epoch": 2.24679029957204,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.625168992811671e-05,
+      "loss": 0.7154,
+      "step": 3150
+    },
+    {
+      "epoch": 2.253922967189729,
+      "grad_norm": 5.5,
+      "learning_rate": 3.591641697021317e-05,
+      "loss": 0.7039,
+      "step": 3160
+    },
+    {
+      "epoch": 2.261055634807418,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.558183042627755e-05,
+      "loss": 0.7844,
+      "step": 3170
+    },
+    {
+      "epoch": 2.268188302425107,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.524794660358593e-05,
+      "loss": 0.7178,
+      "step": 3180
+    },
+    {
+      "epoch": 2.275320970042796,
+      "grad_norm": 5.03125,
+      "learning_rate": 3.491478177516484e-05,
+      "loss": 0.8085,
+      "step": 3190
+    },
+    {
+      "epoch": 2.282453637660485,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.4582352178997935e-05,
+      "loss": 0.7032,
+      "step": 3200
+    },
+    {
+      "epoch": 2.289586305278174,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.425067401723477e-05,
+      "loss": 0.7552,
+      "step": 3210
+    },
+    {
+      "epoch": 2.2967189728958632,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.3919763455401015e-05,
+      "loss": 0.7586,
+      "step": 3220
+    },
+    {
+      "epoch": 2.3038516405135523,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.358963662161062e-05,
+      "loss": 0.8767,
+      "step": 3230
+    },
+    {
+      "epoch": 2.310984308131241,
+      "grad_norm": 5.625,
+      "learning_rate": 3.326030960577972e-05,
+      "loss": 0.7965,
+      "step": 3240
+    },
+    {
+      "epoch": 2.31811697574893,
+      "grad_norm": 5.03125,
+      "learning_rate": 3.293179845884245e-05,
+      "loss": 0.725,
+      "step": 3250
+    },
+    {
+      "epoch": 2.325249643366619,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.2604119191968654e-05,
+      "loss": 0.6246,
+      "step": 3260
+    },
+    {
+      "epoch": 2.332382310984308,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.2277287775783525e-05,
+      "loss": 0.7659,
+      "step": 3270
+    },
+    {
+      "epoch": 2.339514978601997,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.195132013958918e-05,
+      "loss": 0.7054,
+      "step": 3280
+    },
+    {
+      "epoch": 2.346647646219686,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.162623217058834e-05,
+      "loss": 0.7813,
+      "step": 3290
+    },
+    {
+      "epoch": 2.353780313837375,
+      "grad_norm": 5.25,
+      "learning_rate": 3.130203971310999e-05,
+      "loss": 0.7004,
+      "step": 3300
+    },
+    {
+      "epoch": 2.3609129814550642,
+      "grad_norm": 4.59375,
+      "learning_rate": 3.097875856783713e-05,
+      "loss": 0.7509,
+      "step": 3310
+    },
+    {
+      "epoch": 2.3680456490727533,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.065640449103669e-05,
+      "loss": 0.7423,
+      "step": 3320
+    },
+    {
+      "epoch": 2.3751783166904423,
+      "grad_norm": 4.5,
+      "learning_rate": 3.033499319379163e-05,
+      "loss": 0.7703,
+      "step": 3330
+    },
+    {
+      "epoch": 2.3823109843081314,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.001454034123512e-05,
+      "loss": 0.7913,
+      "step": 3340
+    },
+    {
+      "epoch": 2.3894436519258204,
+      "grad_norm": 5.46875,
+      "learning_rate": 2.9695061551787108e-05,
+      "loss": 0.7646,
+      "step": 3350
+    },
+    {
+      "epoch": 2.3965763195435095,
+      "grad_norm": 4.78125,
+      "learning_rate": 2.9376572396393048e-05,
+      "loss": 0.751,
+      "step": 3360
+    },
+    {
+      "epoch": 2.403708987161198,
+      "grad_norm": 4.84375,
+      "learning_rate": 2.9059088397765088e-05,
+      "loss": 0.7969,
+      "step": 3370
+    },
+    {
+      "epoch": 2.410841654778887,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.874262502962537e-05,
+      "loss": 0.7995,
+      "step": 3380
+    },
+    {
+      "epoch": 2.417974322396576,
+      "grad_norm": 9.625,
+      "learning_rate": 2.8427197715952047e-05,
+      "loss": 0.7095,
+      "step": 3390
+    },
+    {
+      "epoch": 2.425106990014265,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.811282183022736e-05,
+      "loss": 0.8267,
+      "step": 3400
+    },
+    {
+      "epoch": 2.4322396576319543,
+      "grad_norm": 4.5625,
+      "learning_rate": 2.7799512694688468e-05,
+      "loss": 0.7129,
+      "step": 3410
+    },
+    {
+      "epoch": 2.4393723252496433,
+      "grad_norm": 6.5,
+      "learning_rate": 2.7487285579580637e-05,
+      "loss": 0.8285,
+      "step": 3420
+    },
+    {
+      "epoch": 2.4465049928673324,
+      "grad_norm": 5.09375,
+      "learning_rate": 2.7176155702412937e-05,
+      "loss": 0.7694,
+      "step": 3430
+    },
+    {
+      "epoch": 2.4536376604850214,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.686613822721666e-05,
+      "loss": 0.7047,
+      "step": 3440
+    },
+    {
+      "epoch": 2.4607703281027105,
+      "grad_norm": 5.1875,
+      "learning_rate": 2.6557248263806174e-05,
+      "loss": 0.7772,
+      "step": 3450
+    },
+    {
+      "epoch": 2.4679029957203995,
+      "grad_norm": 4.6875,
+      "learning_rate": 2.6249500867042522e-05,
+      "loss": 0.8148,
+      "step": 3460
+    },
+    {
+      "epoch": 2.4750356633380886,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.5942911036099655e-05,
+      "loss": 0.689,
+      "step": 3470
+    },
+    {
+      "epoch": 2.4821683309557776,
+      "grad_norm": 5.21875,
+      "learning_rate": 2.5637493713733374e-05,
+      "loss": 0.7762,
+      "step": 3480
+    },
+    {
+      "epoch": 2.4893009985734667,
+      "grad_norm": 5.59375,
+      "learning_rate": 2.533326378555314e-05,
+      "loss": 0.7899,
+      "step": 3490
+    },
+    {
+      "epoch": 2.4964336661911553,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.5030236079296444e-05,
+      "loss": 0.747,
+      "step": 3500
+    },
+    {
+      "epoch": 2.5035663338088447,
+      "grad_norm": 4.125,
+      "learning_rate": 2.4728425364106135e-05,
+      "loss": 0.7218,
+      "step": 3510
+    },
+    {
+      "epoch": 2.5106990014265333,
+      "grad_norm": 4.96875,
+      "learning_rate": 2.442784634981071e-05,
+      "loss": 0.7172,
+      "step": 3520
+    },
+    {
+      "epoch": 2.5178316690442224,
+      "grad_norm": 4.40625,
+      "learning_rate": 2.412851368620726e-05,
+      "loss": 0.6116,
+      "step": 3530
+    },
+    {
+      "epoch": 2.5249643366619114,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.383044196234753e-05,
+      "loss": 0.7398,
+      "step": 3540
+    },
+    {
+      "epoch": 2.5320970042796005,
+      "grad_norm": 5.09375,
+      "learning_rate": 2.3533645705826806e-05,
+      "loss": 0.8014,
+      "step": 3550
+    },
+    {
+      "epoch": 2.5392296718972895,
+      "grad_norm": 5.1875,
+      "learning_rate": 2.323813938207593e-05,
+      "loss": 0.775,
+      "step": 3560
+    },
+    {
+      "epoch": 2.5463623395149786,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.294393739365621e-05,
+      "loss": 0.8234,
+      "step": 3570
+    },
+    {
+      "epoch": 2.5534950071326676,
+      "grad_norm": 4.9375,
+      "learning_rate": 2.265105407955752e-05,
+      "loss": 0.7388,
+      "step": 3580
+    },
+    {
+      "epoch": 2.5606276747503567,
+      "grad_norm": 5.125,
+      "learning_rate": 2.235950371449938e-05,
+      "loss": 0.6721,
+      "step": 3590
+    },
+    {
+      "epoch": 2.5677603423680457,
+      "grad_norm": 5.3125,
+      "learning_rate": 2.2069300508235275e-05,
+      "loss": 0.7251,
+      "step": 3600
+    },
+    {
+      "epoch": 2.574893009985735,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.1780458604860056e-05,
+      "loss": 0.7803,
+      "step": 3610
+    },
+    {
+      "epoch": 2.582025677603424,
+      "grad_norm": 4.78125,
+      "learning_rate": 2.14929920821206e-05,
+      "loss": 0.7447,
+      "step": 3620
+    },
+    {
+      "epoch": 2.5891583452211124,
+      "grad_norm": 5.15625,
+      "learning_rate": 2.1206914950729672e-05,
+      "loss": 0.7869,
+      "step": 3630
+    },
+    {
+      "epoch": 2.596291012838802,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.0922241153683066e-05,
+      "loss": 0.7073,
+      "step": 3640
+    },
+    {
+      "epoch": 2.6034236804564905,
+      "grad_norm": 5.125,
+      "learning_rate": 2.0638984565580023e-05,
+      "loss": 0.7132,
+      "step": 3650
+    },
+    {
+      "epoch": 2.6105563480741796,
+      "grad_norm": 4.875,
+      "learning_rate": 2.035715899194704e-05,
+      "loss": 0.7917,
+      "step": 3660
+    },
+    {
+      "epoch": 2.6176890156918686,
+      "grad_norm": 4.84375,
+      "learning_rate": 2.007677816856498e-05,
+      "loss": 0.7058,
+      "step": 3670
+    },
+    {
+      "epoch": 2.6248216833095577,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.9797855760799612e-05,
+      "loss": 0.7053,
+      "step": 3680
+    },
+    {
+      "epoch": 2.6319543509272467,
+      "grad_norm": 5.65625,
+      "learning_rate": 1.9520405362935594e-05,
+      "loss": 0.7725,
+      "step": 3690
+    },
+    {
+      "epoch": 2.6390870185449358,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.9244440497513893e-05,
+      "loss": 0.712,
+      "step": 3700
+    },
+    {
+      "epoch": 2.646219686162625,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.896997461467272e-05,
+      "loss": 0.7954,
+      "step": 3710
+    },
+    {
+      "epoch": 2.653352353780314,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.869702109149199e-05,
+      "loss": 0.727,
+      "step": 3720
+    },
+    {
+      "epoch": 2.660485021398003,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.842559323134136e-05,
+      "loss": 0.6694,
+      "step": 3730
+    },
+    {
+      "epoch": 2.667617689015692,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.8155704263231776e-05,
+      "loss": 0.7812,
+      "step": 3740
+    },
+    {
+      "epoch": 2.674750356633381,
+      "grad_norm": 5.25,
+      "learning_rate": 1.788736734117078e-05,
+      "loss": 0.756,
+      "step": 3750
+    },
+    {
+      "epoch": 2.68188302425107,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.762059554352143e-05,
+      "loss": 0.6954,
+      "step": 3760
+    },
+    {
+      "epoch": 2.689015691868759,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.7355401872364758e-05,
+      "loss": 0.7037,
+      "step": 3770
+    },
+    {
+      "epoch": 2.6961483594864477,
+      "grad_norm": 5.65625,
+      "learning_rate": 1.709179925286617e-05,
+      "loss": 0.7861,
+      "step": 3780
+    },
+    {
+      "epoch": 2.703281027104137,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.6829800532645447e-05,
+      "loss": 0.7455,
+      "step": 3790
+    },
+    {
+      "epoch": 2.710413694721826,
+      "grad_norm": 5.375,
+      "learning_rate": 1.6569418481150595e-05,
+      "loss": 0.6718,
+      "step": 3800
+    },
+    {
+      "epoch": 2.717546362339515,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.6310665789035467e-05,
+      "loss": 0.7419,
+      "step": 3810
+    },
+    {
+      "epoch": 2.724679029957204,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.605355506754121e-05,
+      "loss": 0.7336,
+      "step": 3820
+    },
+    {
+      "epoch": 2.731811697574893,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5798098847881666e-05,
+      "loss": 0.6514,
+      "step": 3830
+    },
+    {
+      "epoch": 2.738944365192582,
+      "grad_norm": 4.25,
+      "learning_rate": 1.554430958063259e-05,
+      "loss": 0.6845,
+      "step": 3840
+    },
+    {
+      "epoch": 2.746077032810271,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.529219963512481e-05,
+      "loss": 0.7397,
+      "step": 3850
+    },
+    {
+      "epoch": 2.75320970042796,
+      "grad_norm": 3.875,
+      "learning_rate": 1.5041781298841423e-05,
+      "loss": 0.7563,
+      "step": 3860
+    },
+    {
+      "epoch": 2.760342368045649,
+      "grad_norm": 4.875,
+      "learning_rate": 1.4793066776818842e-05,
+      "loss": 0.7316,
+      "step": 3870
+    },
+    {
+      "epoch": 2.767475035663338,
+      "grad_norm": 4.125,
+      "learning_rate": 1.4546068191051987e-05,
+      "loss": 0.7016,
+      "step": 3880
+    },
+    {
+      "epoch": 2.7746077032810272,
+      "grad_norm": 5.4375,
+      "learning_rate": 1.4300797579903474e-05,
+      "loss": 0.7698,
+      "step": 3890
+    },
+    {
+      "epoch": 2.7817403708987163,
+      "grad_norm": 5.625,
+      "learning_rate": 1.4057266897516841e-05,
+      "loss": 0.8646,
+      "step": 3900
+    },
+    {
+      "epoch": 2.788873038516405,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.3815488013233984e-05,
+      "loss": 0.7541,
+      "step": 3910
+    },
+    {
+      "epoch": 2.7960057061340944,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.3575472711016635e-05,
+      "loss": 0.8134,
+      "step": 3920
+    },
+    {
+      "epoch": 2.803138373751783,
+      "grad_norm": 5.0,
+      "learning_rate": 1.3337232688872009e-05,
+      "loss": 0.7447,
+      "step": 3930
+    },
+    {
+      "epoch": 2.810271041369472,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.3100779558282672e-05,
+      "loss": 0.6445,
+      "step": 3940
+    },
+    {
+      "epoch": 2.817403708987161,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.2866124843640615e-05,
+      "loss": 0.7147,
+      "step": 3950
+    },
+    {
+      "epoch": 2.82453637660485,
+      "grad_norm": 5.0,
+      "learning_rate": 1.2633279981685608e-05,
+      "loss": 0.7744,
+      "step": 3960
+    },
+    {
+      "epoch": 2.831669044222539,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.240225632094773e-05,
+      "loss": 0.7242,
+      "step": 3970
+    },
+    {
+      "epoch": 2.8388017118402282,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.217306512119425e-05,
+      "loss": 0.7314,
+      "step": 3980
+    },
+    {
+      "epoch": 2.8459343794579173,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.194571755288092e-05,
+      "loss": 0.7505,
+      "step": 3990
+    },
+    {
+      "epoch": 2.8530670470756063,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.1720224696607474e-05,
+      "loss": 0.7146,
+      "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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/checkpoint-4000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af7cf7c3c741add5e8996639d61d87fe74bb40ad0faa1b29901130f4b3a60fc3
+size 6353

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.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-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b8cee98ee2a3b201eac58f67d136f3636e25ea77f2d6456aed0e76ba9806ba7
+size 298041696

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:466c5a0c45557ab3167c47df6fc8d83c9a350c6d83e31f3f63b06650c8ebae6e
+size 596170443

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/rng_state_0.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/rng_state_1.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/rng_state_2.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/rng_state_3.pth

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

modernbert-crux-researchy-pos_20.neg_51.b64_n512.1e-4.request/runs/Jan01_22-56-07_nid005051/events.out.tfevents.1767301064.nid005051.19581.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f851e7139ea8140e18315ef864967a2ff4a57ec78684c6323cb8173fc14f535
+size 8790