Training in progress, step 5000, checkpoint

checkpoint-5000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad6757f8a837ad3dfb09b6fc1dcb0b868b7af40b239f2b0dfc2692d66633eefb
+oid sha256:cdf67d3de471c326b240bdbd77dbc405e8c753c7d653f31582e509fdf6b55716
 size 1711848436

checkpoint-5000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3a1913b51c497214c016acbdc6ae948b5a70304cf0d9327cb56ace5b956ba6f0
+oid sha256:892c83acef92139cb512da738db48646f5d62f453aa9f1b50a370ae6ad100445
 size 3424043887

checkpoint-5000/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7035414576530457,
+  "best_metric": 0.7118513584136963,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-5000",
   "epoch": 0.5332764505119454,
   "eval_steps": 500,
@@ -10,152 +10,152 @@
   "log_history": [
     {
       "epoch": 0.05332764505119454,
-      "grad_norm": 414.2552795410156,
+      "grad_norm": 401.61187744140625,
       "learning_rate": 4.911120591581342e-07,
-      "loss": 0.3809,
+      "loss": 0.3808,
       "step": 500
     },
     {
       "epoch": 0.05332764505119454,
-      "eval_loss": 1.1109352111816406,
-      "eval_runtime": 63.6734,
-      "eval_samples_per_second": 15.501,
-      "eval_steps_per_second": 1.947,
+      "eval_loss": 1.1107642650604248,
+      "eval_runtime": 62.4169,
+      "eval_samples_per_second": 15.813,
+      "eval_steps_per_second": 1.987,
       "step": 500
     },
     {
       "epoch": 0.10665529010238908,
-      "grad_norm": 45.880027770996094,
+      "grad_norm": 42.57551193237305,
       "learning_rate": 4.822241183162685e-07,
-      "loss": 0.2649,
+      "loss": 0.2666,
       "step": 1000
     },
     {
       "epoch": 0.10665529010238908,
-      "eval_loss": 0.9588962197303772,
-      "eval_runtime": 62.6825,
-      "eval_samples_per_second": 15.746,
-      "eval_steps_per_second": 1.978,
+      "eval_loss": 0.9644754528999329,
+      "eval_runtime": 63.4304,
+      "eval_samples_per_second": 15.56,
+      "eval_steps_per_second": 1.955,
       "step": 1000
     },
     {
       "epoch": 0.1599829351535836,
-      "grad_norm": 432.9915466308594,
+      "grad_norm": 518.2286376953125,
       "learning_rate": 4.733361774744027e-07,
-      "loss": 0.2104,
+      "loss": 0.2085,
       "step": 1500
     },
     {
       "epoch": 0.1599829351535836,
-      "eval_loss": 0.90110844373703,
-      "eval_runtime": 62.6757,
-      "eval_samples_per_second": 15.748,
-      "eval_steps_per_second": 1.978,
+      "eval_loss": 0.9051361680030823,
+      "eval_runtime": 63.9009,
+      "eval_samples_per_second": 15.446,
+      "eval_steps_per_second": 1.941,
       "step": 1500
     },
     {
       "epoch": 0.21331058020477817,
-      "grad_norm": 0.014208819717168808,
+      "grad_norm": 0.0223398394882679,
       "learning_rate": 4.6444823663253695e-07,
-      "loss": 0.1849,
+      "loss": 0.1867,
       "step": 2000
     },
     {
       "epoch": 0.21331058020477817,
-      "eval_loss": 0.8570474982261658,
-      "eval_runtime": 63.7296,
-      "eval_samples_per_second": 15.487,
-      "eval_steps_per_second": 1.946,
+      "eval_loss": 0.8611643314361572,
+      "eval_runtime": 63.9649,
+      "eval_samples_per_second": 15.43,
+      "eval_steps_per_second": 1.939,
       "step": 2000
     },
     {
       "epoch": 0.2666382252559727,
-      "grad_norm": 174.73324584960938,
+      "grad_norm": 175.0380096435547,
       "learning_rate": 4.5556029579067116e-07,
-      "loss": 0.2056,
+      "loss": 0.2068,
       "step": 2500
     },
     {
       "epoch": 0.2666382252559727,
-      "eval_loss": 0.7895939350128174,
-      "eval_runtime": 63.8387,
-      "eval_samples_per_second": 15.461,
-      "eval_steps_per_second": 1.942,
+      "eval_loss": 0.7947371006011963,
+      "eval_runtime": 64.0455,
+      "eval_samples_per_second": 15.411,
+      "eval_steps_per_second": 1.936,
       "step": 2500
     },
     {
       "epoch": 0.3199658703071672,
-      "grad_norm": 9.852073823424234e-09,
+      "grad_norm": 9.192651084788395e-09,
       "learning_rate": 4.4667235494880547e-07,
-      "loss": 0.159,
+      "loss": 0.163,
       "step": 3000
     },
     {
       "epoch": 0.3199658703071672,
-      "eval_loss": 0.7647623419761658,
-      "eval_runtime": 63.9155,
-      "eval_samples_per_second": 15.442,
-      "eval_steps_per_second": 1.94,
+      "eval_loss": 0.7677457332611084,
+      "eval_runtime": 64.1593,
+      "eval_samples_per_second": 15.384,
+      "eval_steps_per_second": 1.933,
       "step": 3000
     },
     {
       "epoch": 0.37329351535836175,
-      "grad_norm": 0.0025215104687958956,
+      "grad_norm": 0.002516463166102767,
       "learning_rate": 4.377844141069397e-07,
-      "loss": 0.1696,
+      "loss": 0.1729,
       "step": 3500
     },
     {
       "epoch": 0.37329351535836175,
-      "eval_loss": 0.7638036012649536,
-      "eval_runtime": 63.8003,
-      "eval_samples_per_second": 15.47,
-      "eval_steps_per_second": 1.944,
+      "eval_loss": 0.7635349035263062,
+      "eval_runtime": 62.8257,
+      "eval_samples_per_second": 15.71,
+      "eval_steps_per_second": 1.974,
       "step": 3500
     },
     {
       "epoch": 0.42662116040955633,
-      "grad_norm": 2.4611830711364746,
+      "grad_norm": 7.581352710723877,
       "learning_rate": 4.2889647326507393e-07,
-      "loss": 0.1125,
+      "loss": 0.1149,
       "step": 4000
     },
     {
       "epoch": 0.42662116040955633,
-      "eval_loss": 0.7486374378204346,
-      "eval_runtime": 63.9847,
-      "eval_samples_per_second": 15.426,
-      "eval_steps_per_second": 1.938,
+      "eval_loss": 0.7458549737930298,
+      "eval_runtime": 64.2184,
+      "eval_samples_per_second": 15.369,
+      "eval_steps_per_second": 1.931,
       "step": 4000
     },
     {
       "epoch": 0.47994880546075086,
-      "grad_norm": 5.230295658111572,
+      "grad_norm": 1.3885610103607178,
       "learning_rate": 4.2000853242320814e-07,
-      "loss": 0.1322,
+      "loss": 0.1264,
       "step": 4500
     },
     {
       "epoch": 0.47994880546075086,
-      "eval_loss": 0.7251659035682678,
-      "eval_runtime": 64.4782,
-      "eval_samples_per_second": 15.308,
-      "eval_steps_per_second": 1.923,
+      "eval_loss": 0.7185410261154175,
+      "eval_runtime": 62.6998,
+      "eval_samples_per_second": 15.742,
+      "eval_steps_per_second": 1.978,
       "step": 4500
     },
     {
       "epoch": 0.5332764505119454,
-      "grad_norm": 3.821002974291332e-05,
+      "grad_norm": 4.22987541242037e-05,
       "learning_rate": 4.1112059158134245e-07,
-      "loss": 0.1633,
+      "loss": 0.1638,
       "step": 5000
     },
     {
       "epoch": 0.5332764505119454,
-      "eval_loss": 0.7035414576530457,
-      "eval_runtime": 62.6186,
-      "eval_samples_per_second": 15.762,
-      "eval_steps_per_second": 1.98,
+      "eval_loss": 0.7118513584136963,
+      "eval_runtime": 63.9674,
+      "eval_samples_per_second": 15.43,
+      "eval_steps_per_second": 1.938,
       "step": 5000
     }
   ],

checkpoint-5000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:057f6bbf1ff3e34abda0863071bbf5fdcc7edb1102f04dba05c7a15087bc3396
+oid sha256:54c5f1c69a0ead76c168a8c74b80758f8f8eb77228d72de2bfc13a6b8cdc784b
 size 4847