Training in progress, epoch 7, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:622bd7a4380e21525cb42dca1bf92fe97484dd66422519f9b0981f649a7d1d76
+oid sha256:91eb96d3922be157532116be345e306074157b7a7a4a7335c0238c11a13104fb
 size 711443456

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0bc28c79f43492462a8e2f42da31e595112127cce6a03da5c2e702d15e1ed78d
+oid sha256:1914f1fa557a8fa584d003645654c4d6913c76b6a764548c9acfbff31bb2d3d0
 size 1423007994

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b38a2c951656be0a70f6e69a88eae78f0b9438cb8d5b065b93eccc853ae555d
+oid sha256:c46246640d1543a5d18433713b555d6e3842e91778e76a1852735bc0ed173fe2
 size 15984

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:46675ed5a6b24b22c0bc3653a3cff8c05654a637f35a241f134284b79d176577
+oid sha256:f5ed2125154aeb523313b13badf0ffb177c19ce0c77a62f1b348e5f1cb4c3b09
 size 15984

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e90a9c419257426cf85763420b7dbc2cf1c88ef69e3e43f3f41c305218425f7
+oid sha256:6d5d3bc9f42a0ece76532c1de7bc707e6df336aca81bca6f2066c30c05126a2f
 size 15984

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4cb8903b61a89e1d9a23300ebd09815a631f69a4be52ab57b2fde4f1c2a83641
+oid sha256:ecbdd4a506b7c616a788f2418e754c6a53c2bcc7a63cff96be4697ca6bc0706a
 size 15984

last-checkpoint/rng_state_4.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:72583e173a04498b93ecdfe16f02d7ea59d0ffbaf05725099e03a133a9df2495
+oid sha256:f2100657e88a6589af92eae97d1bbf8683394c289c642981d0c2937312744e3b
 size 15984

last-checkpoint/rng_state_5.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12374950fb047bf2c810a7f5336254aff2f096c940cf390221a067240e146773
+oid sha256:7a771fb324acc9bce75a763b43a2389416b47091b1973e7a8b137672d65216c7
 size 15984

last-checkpoint/rng_state_6.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7786ec19b24090398a272236946ef1a91af4ed2b9c3b02f64db208a504cc2fb7
+oid sha256:8efe5d627912fd093b28bfa11c19eed1c8663d5c4427be33733700365128d5f9
 size 15984

last-checkpoint/rng_state_7.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0fca6ef03b1b3ff8c55b8c6be1793f194760dcf2e05a9004b0687c57f6621896
+oid sha256:162222fe11954a835fdb6646ebeec9ede29106cf49f7358b96266185f60e2fe4
 size 15984

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fe885b47262b2377e6c40e91e10de848879660bf5ce05715ebb7bb06b6a672f
+oid sha256:df2fb6fb1693a9f822931c64464c1e0bfe205ebfb9fe72b6c70c024b845afee2
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.7946750167423844,
-  "best_model_checkpoint": "checkpoint/cross_encoder_20250522_full_data/checkpoint-13902",
-  "epoch": 6.0,
+  "best_metric": 0.7979517910648003,
+  "best_model_checkpoint": "checkpoint/cross_encoder_20250522_full_data/checkpoint-16219",
+  "epoch": 7.0,
   "eval_steps": 500,
-  "global_step": 13902,
+  "global_step": 16219,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -563,6 +563,101 @@
       "eval_samples_per_second": 808.07,
       "eval_steps_per_second": 6.313,
       "step": 13902
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 1.9424314498901367,
+      "learning_rate": 4.0591573080173995e-06,
+      "loss": 0.3873,
+      "step": 14000
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 2.831543445587158,
+      "learning_rate": 3.911643687576664e-06,
+      "loss": 0.3807,
+      "step": 14200
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 2.7500486373901367,
+      "learning_rate": 3.7651180761199505e-06,
+      "loss": 0.3801,
+      "step": 14400
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 4.015429496765137,
+      "learning_rate": 3.619713489479354e-06,
+      "loss": 0.3824,
+      "step": 14600
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 2.4831278324127197,
+      "learning_rate": 3.4755619258215407e-06,
+      "loss": 0.3808,
+      "step": 14800
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 7.442523002624512,
+      "learning_rate": 3.3327942458199193e-06,
+      "loss": 0.385,
+      "step": 15000
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 4.242701530456543,
+      "learning_rate": 3.1915400538594333e-06,
+      "loss": 0.3832,
+      "step": 15200
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 5.000258922576904,
+      "learning_rate": 3.0519275803818014e-06,
+      "loss": 0.3805,
+      "step": 15400
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 2.0560362339019775,
+      "learning_rate": 2.914083565478024e-06,
+      "loss": 0.381,
+      "step": 15600
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 1.8777313232421875,
+      "learning_rate": 2.7781331438338317e-06,
+      "loss": 0.3831,
+      "step": 15800
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 1.0822844505310059,
+      "learning_rate": 2.6441997311325196e-06,
+      "loss": 0.3797,
+      "step": 16000
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 3.2328083515167236,
+      "learning_rate": 2.5124049120182916e-06,
+      "loss": 0.383,
+      "step": 16200
+    },
+    {
+      "epoch": 7.0,
+      "eval_f2": 0.7979517910648003,
+      "eval_loss": 0.4335618019104004,
+      "eval_precision": 0.5632793509486378,
+      "eval_recall": 0.8907248402555911,
+      "eval_runtime": 148.5629,
+      "eval_samples_per_second": 809.031,
+      "eval_steps_per_second": 6.321,
+      "step": 16219
     }
   ],
   "logging_steps": 200,
@@ -570,7 +665,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 4.681945532459909e+17,
+  "total_flos": 5.462269789659464e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null