Training in progress, step 6500, checkpoint

checkpoint-6500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4d6f457609c93ca7de3b2cd62cb48b504d81a31b9b8cd9b8e274b1031fbcbbd7
+oid sha256:1a56cc3805b01f02c61cf46bf24bbce7a1a90c177e79b491706af2c11c42fa3e
 size 1711848436

checkpoint-6500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2de85f7f6aad835d12f6ec19c41a4dca30cfb202fe9797b57a9e1f1d3dbc7c99
+oid sha256:2c6ef7d95228db335d8919dff2bca57d03703c8abe5c30c9a03377945295b4f1
 size 3424043887

checkpoint-6500/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.6261336207389832,
+  "best_metric": 0.628384530544281,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l57-l/checkpoint-6500",
   "epoch": 1.1974944731024317,
   "eval_steps": 500,
@@ -10,197 +10,197 @@
   "log_history": [
     {
       "epoch": 0.09211495946941783,
-      "grad_norm": 3.083087682723999,
+      "grad_norm": 2.6780805587768555,
       "learning_rate": 4.84647506755097e-07,
-      "loss": 0.2708,
+      "loss": 0.2705,
       "step": 500
     },
     {
       "epoch": 0.09211495946941783,
-      "eval_loss": 1.0631108283996582,
-      "eval_runtime": 74.1964,
-      "eval_samples_per_second": 16.268,
-      "eval_steps_per_second": 2.035,
+      "eval_loss": 1.0680681467056274,
+      "eval_runtime": 74.5723,
+      "eval_samples_per_second": 16.186,
+      "eval_steps_per_second": 2.025,
       "step": 500
     },
     {
       "epoch": 0.18422991893883567,
-      "grad_norm": 7.726487159729004,
+      "grad_norm": 8.301130294799805,
       "learning_rate": 4.69295013510194e-07,
-      "loss": 0.251,
+      "loss": 0.2545,
       "step": 1000
     },
     {
       "epoch": 0.18422991893883567,
-      "eval_loss": 0.942773699760437,
-      "eval_runtime": 74.6047,
-      "eval_samples_per_second": 16.179,
-      "eval_steps_per_second": 2.024,
+      "eval_loss": 0.9443553686141968,
+      "eval_runtime": 74.9643,
+      "eval_samples_per_second": 16.101,
+      "eval_steps_per_second": 2.014,
       "step": 1000
     },
     {
       "epoch": 0.2763448784082535,
-      "grad_norm": 0.02718021161854267,
+      "grad_norm": 0.018898434937000275,
       "learning_rate": 4.5394252026529107e-07,
-      "loss": 0.235,
+      "loss": 0.234,
       "step": 1500
     },
     {
       "epoch": 0.2763448784082535,
-      "eval_loss": 0.8716810345649719,
-      "eval_runtime": 75.03,
-      "eval_samples_per_second": 16.087,
-      "eval_steps_per_second": 2.013,
+      "eval_loss": 0.876915693283081,
+      "eval_runtime": 75.6401,
+      "eval_samples_per_second": 15.957,
+      "eval_steps_per_second": 1.996,
       "step": 1500
     },
     {
       "epoch": 0.36845983787767134,
-      "grad_norm": 0.0001215290030813776,
+      "grad_norm": 0.00010984255641233176,
       "learning_rate": 4.385900270203881e-07,
-      "loss": 0.1529,
+      "loss": 0.1539,
       "step": 2000
     },
     {
       "epoch": 0.36845983787767134,
-      "eval_loss": 0.8318113088607788,
-      "eval_runtime": 76.6917,
-      "eval_samples_per_second": 15.738,
-      "eval_steps_per_second": 1.969,
+      "eval_loss": 0.8415330648422241,
+      "eval_runtime": 76.9959,
+      "eval_samples_per_second": 15.676,
+      "eval_steps_per_second": 1.961,
       "step": 2000
     },
     {
       "epoch": 0.46057479734708917,
-      "grad_norm": 419.545166015625,
+      "grad_norm": 395.890625,
       "learning_rate": 4.232375337754851e-07,
-      "loss": 0.1781,
+      "loss": 0.1766,
       "step": 2500
     },
     {
       "epoch": 0.46057479734708917,
-      "eval_loss": 0.7549223303794861,
-      "eval_runtime": 75.2036,
-      "eval_samples_per_second": 16.05,
-      "eval_steps_per_second": 2.008,
+      "eval_loss": 0.7660450339317322,
+      "eval_runtime": 76.733,
+      "eval_samples_per_second": 15.73,
+      "eval_steps_per_second": 1.968,
       "step": 2500
     },
     {
       "epoch": 0.552689756816507,
-      "grad_norm": 368.7484436035156,
+      "grad_norm": 363.6924743652344,
       "learning_rate": 4.0788504053058217e-07,
-      "loss": 0.1681,
+      "loss": 0.1679,
       "step": 3000
     },
     {
       "epoch": 0.552689756816507,
-      "eval_loss": 0.7217888236045837,
-      "eval_runtime": 76.0606,
-      "eval_samples_per_second": 15.869,
-      "eval_steps_per_second": 1.985,
+      "eval_loss": 0.7269155383110046,
+      "eval_runtime": 76.7928,
+      "eval_samples_per_second": 15.718,
+      "eval_steps_per_second": 1.966,
       "step": 3000
     },
     {
       "epoch": 0.6448047162859248,
-      "grad_norm": 20.03949737548828,
+      "grad_norm": 28.914026260375977,
       "learning_rate": 3.925325472856792e-07,
-      "loss": 0.1064,
+      "loss": 0.1104,
       "step": 3500
     },
     {
       "epoch": 0.6448047162859248,
-      "eval_loss": 0.7048410177230835,
-      "eval_runtime": 76.0578,
-      "eval_samples_per_second": 15.87,
-      "eval_steps_per_second": 1.985,
+      "eval_loss": 0.7097663283348083,
+      "eval_runtime": 76.6897,
+      "eval_samples_per_second": 15.739,
+      "eval_steps_per_second": 1.969,
       "step": 3500
     },
     {
       "epoch": 0.7369196757553427,
-      "grad_norm": 363.6124267578125,
+      "grad_norm": 116.85651397705078,
       "learning_rate": 3.771800540407762e-07,
-      "loss": 0.1357,
+      "loss": 0.1367,
       "step": 4000
     },
     {
       "epoch": 0.7369196757553427,
-      "eval_loss": 0.6961681246757507,
-      "eval_runtime": 76.3574,
-      "eval_samples_per_second": 15.807,
-      "eval_steps_per_second": 1.978,
+      "eval_loss": 0.6968969106674194,
+      "eval_runtime": 77.1872,
+      "eval_samples_per_second": 15.637,
+      "eval_steps_per_second": 1.956,
       "step": 4000
     },
     {
       "epoch": 0.8290346352247605,
-      "grad_norm": 494.03570556640625,
+      "grad_norm": 391.0343322753906,
       "learning_rate": 3.618275607958732e-07,
-      "loss": 0.1098,
+      "loss": 0.1129,
       "step": 4500
     },
     {
       "epoch": 0.8290346352247605,
-      "eval_loss": 0.6778352856636047,
-      "eval_runtime": 76.4844,
-      "eval_samples_per_second": 15.781,
-      "eval_steps_per_second": 1.974,
+      "eval_loss": 0.6776800751686096,
+      "eval_runtime": 75.0351,
+      "eval_samples_per_second": 16.086,
+      "eval_steps_per_second": 2.012,
       "step": 4500
     },
     {
       "epoch": 0.9211495946941783,
-      "grad_norm": 4.2969353671651334e-05,
+      "grad_norm": 2.0514346033451147e-05,
       "learning_rate": 3.464750675509703e-07,
-      "loss": 0.1142,
+      "loss": 0.1125,
       "step": 5000
     },
     {
       "epoch": 0.9211495946941783,
-      "eval_loss": 0.6656659245491028,
-      "eval_runtime": 76.6475,
-      "eval_samples_per_second": 15.747,
-      "eval_steps_per_second": 1.97,
+      "eval_loss": 0.6658498048782349,
+      "eval_runtime": 77.111,
+      "eval_samples_per_second": 15.653,
+      "eval_steps_per_second": 1.958,
       "step": 5000
     },
     {
       "epoch": 1.013264554163596,
-      "grad_norm": 0.0010161151876673102,
+      "grad_norm": 0.000568103976547718,
       "learning_rate": 3.311225743060673e-07,
-      "loss": 0.1113,
+      "loss": 0.1071,
       "step": 5500
     },
     {
       "epoch": 1.013264554163596,
-      "eval_loss": 0.6430702209472656,
-      "eval_runtime": 76.4915,
-      "eval_samples_per_second": 15.78,
-      "eval_steps_per_second": 1.974,
+      "eval_loss": 0.6465296149253845,
+      "eval_runtime": 75.2951,
+      "eval_samples_per_second": 16.03,
+      "eval_steps_per_second": 2.005,
       "step": 5500
     },
     {
       "epoch": 1.105379513633014,
-      "grad_norm": 0.05515381693840027,
+      "grad_norm": 0.02698369137942791,
       "learning_rate": 3.157700810611643e-07,
-      "loss": 0.0572,
+      "loss": 0.0553,
       "step": 6000
     },
     {
       "epoch": 1.105379513633014,
-      "eval_loss": 0.636746883392334,
-      "eval_runtime": 76.0169,
-      "eval_samples_per_second": 15.878,
-      "eval_steps_per_second": 1.986,
+      "eval_loss": 0.6357071995735168,
+      "eval_runtime": 76.6726,
+      "eval_samples_per_second": 15.742,
+      "eval_steps_per_second": 1.969,
       "step": 6000
     },
     {
       "epoch": 1.1974944731024317,
-      "grad_norm": 0.0038496837951242924,
+      "grad_norm": 0.0005627564387395978,
       "learning_rate": 3.0041758781626137e-07,
-      "loss": 0.0746,
+      "loss": 0.0729,
       "step": 6500
     },
     {
       "epoch": 1.1974944731024317,
-      "eval_loss": 0.6261336207389832,
-      "eval_runtime": 76.1916,
-      "eval_samples_per_second": 15.842,
-      "eval_steps_per_second": 1.982,
+      "eval_loss": 0.628384530544281,
+      "eval_runtime": 75.6565,
+      "eval_samples_per_second": 15.954,
+      "eval_steps_per_second": 1.996,
       "step": 6500
     }
   ],

checkpoint-6500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5c9589d95f7287a546fe28eed1162f0176f1b95b18274e91ad5fc36f7db12aeb
+oid sha256:ea34fd4333686a44a60d290faa4ae8e00df783131842d9ae3ef1b9e874a39349
 size 4847