Training in progress, step 4000, checkpoint

checkpoint-4000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:03f47678e62fd053b2491b45a7839eae2ac60f3cb639f72d05f41034907064b9
+oid sha256:4ea6bcfaa9efb649bf75add1834edb941c21ab025f126f1dd5b6800cc405e4e5
 size 1711848436

checkpoint-4000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f7d39ef6aff5c3dd50449ad89aca3752fbca89bbd47e12c135470fa11b24f054
+oid sha256:99e26c03042ed0a6b711ffb9eca502e23838330f954b8b5d1c19155f72340e24
 size 3424043887

checkpoint-4000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2804773ff24cfee4e1940fb5ed0e01151248a4999b512e43cad8a70eead751df
+oid sha256:ab15c562351807c0837f190f0e6bcaf245e60e971b09a9339abc67be0581f8b3
 size 623

checkpoint-4000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 2.0000314712524414,
-  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l55-l/checkpoint-4000",
-  "epoch": 0.7367839381101492,
+  "best_metric": 2.0315287113189697,
+  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l55-l/checkpoint-3500",
+  "epoch": 0.7369196757553427,
   "eval_steps": 500,
   "global_step": 4000,
   "is_hyper_param_search": false,
@@ -9,128 +9,128 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.09209799226376865,
-      "grad_norm": 0.009597355499863625,
-      "learning_rate": 4.846503346227052e-05,
-      "loss": 0.7336,
+      "epoch": 0.09211495946941783,
+      "grad_norm": 0.030529862269759178,
+      "learning_rate": 4.8464750675509704e-05,
+      "loss": 0.7076,
       "step": 500
     },
     {
-      "epoch": 0.09209799226376865,
-      "eval_loss": 2.0780556201934814,
-      "eval_runtime": 77.1404,
-      "eval_samples_per_second": 15.634,
-      "eval_steps_per_second": 1.957,
+      "epoch": 0.09211495946941783,
+      "eval_loss": 2.07871675491333,
+      "eval_runtime": 73.4415,
+      "eval_samples_per_second": 16.435,
+      "eval_steps_per_second": 2.056,
       "step": 500
     },
     {
-      "epoch": 0.1841959845275373,
-      "grad_norm": 0.10049372166395187,
-      "learning_rate": 4.693006692454105e-05,
-      "loss": 0.6953,
+      "epoch": 0.18422991893883567,
+      "grad_norm": 0.02040918357670307,
+      "learning_rate": 4.6929501351019406e-05,
+      "loss": 0.7045,
       "step": 1000
     },
     {
-      "epoch": 0.1841959845275373,
-      "eval_loss": 2.0782744884490967,
-      "eval_runtime": 77.8965,
-      "eval_samples_per_second": 15.482,
-      "eval_steps_per_second": 1.938,
+      "epoch": 0.18422991893883567,
+      "eval_loss": 2.0783257484436035,
+      "eval_runtime": 74.6654,
+      "eval_samples_per_second": 16.165,
+      "eval_steps_per_second": 2.022,
       "step": 1000
     },
     {
-      "epoch": 0.27629397679130596,
-      "grad_norm": 0.0041180383414030075,
-      "learning_rate": 4.539510038681157e-05,
-      "loss": 0.6953,
+      "epoch": 0.2763448784082535,
+      "grad_norm": 0.8010361194610596,
+      "learning_rate": 4.539425202652911e-05,
+      "loss": 0.7106,
       "step": 1500
     },
     {
-      "epoch": 0.27629397679130596,
-      "eval_loss": 2.077699899673462,
-      "eval_runtime": 78.9586,
-      "eval_samples_per_second": 15.274,
-      "eval_steps_per_second": 1.912,
+      "epoch": 0.2763448784082535,
+      "eval_loss": 2.0473792552948,
+      "eval_runtime": 75.4141,
+      "eval_samples_per_second": 16.005,
+      "eval_steps_per_second": 2.002,
       "step": 1500
     },
     {
-      "epoch": 0.3683919690550746,
-      "grad_norm": 0.29397013783454895,
-      "learning_rate": 4.386013384908209e-05,
-      "loss": 0.6932,
+      "epoch": 0.36845983787767134,
+      "grad_norm": 1.1971429586410522,
+      "learning_rate": 4.3859002702038817e-05,
+      "loss": 0.704,
       "step": 2000
     },
     {
-      "epoch": 0.3683919690550746,
-      "eval_loss": 2.0721774101257324,
-      "eval_runtime": 77.349,
-      "eval_samples_per_second": 15.592,
-      "eval_steps_per_second": 1.952,
+      "epoch": 0.36845983787767134,
+      "eval_loss": 2.281594753265381,
+      "eval_runtime": 76.3969,
+      "eval_samples_per_second": 15.799,
+      "eval_steps_per_second": 1.977,
       "step": 2000
     },
     {
-      "epoch": 0.46048996131884323,
-      "grad_norm": 0.9136261343955994,
-      "learning_rate": 4.232516731135262e-05,
-      "loss": 0.6907,
+      "epoch": 0.46057479734708917,
+      "grad_norm": 1.5381091833114624,
+      "learning_rate": 4.232375337754851e-05,
+      "loss": 0.6996,
       "step": 2500
     },
     {
-      "epoch": 0.46048996131884323,
-      "eval_loss": 2.079211711883545,
-      "eval_runtime": 79.3204,
-      "eval_samples_per_second": 15.204,
-      "eval_steps_per_second": 1.904,
+      "epoch": 0.46057479734708917,
+      "eval_loss": 2.045022964477539,
+      "eval_runtime": 76.3329,
+      "eval_samples_per_second": 15.812,
+      "eval_steps_per_second": 1.978,
       "step": 2500
     },
     {
-      "epoch": 0.5525879535826119,
-      "grad_norm": 4.484213829040527,
-      "learning_rate": 4.079020077362314e-05,
-      "loss": 0.7043,
+      "epoch": 0.552689756816507,
+      "grad_norm": 5.1257524490356445,
+      "learning_rate": 4.078850405305822e-05,
+      "loss": 0.6739,
       "step": 3000
     },
     {
-      "epoch": 0.5525879535826119,
-      "eval_loss": 2.0204577445983887,
-      "eval_runtime": 78.612,
-      "eval_samples_per_second": 15.341,
-      "eval_steps_per_second": 1.921,
+      "epoch": 0.552689756816507,
+      "eval_loss": 2.0969045162200928,
+      "eval_runtime": 76.2547,
+      "eval_samples_per_second": 15.829,
+      "eval_steps_per_second": 1.98,
       "step": 3000
     },
     {
-      "epoch": 0.6446859458463805,
-      "grad_norm": 1.1021015644073486,
-      "learning_rate": 3.9255234235893664e-05,
-      "loss": 0.7078,
+      "epoch": 0.6448047162859248,
+      "grad_norm": 6.156252861022949,
+      "learning_rate": 3.925325472856792e-05,
+      "loss": 0.6687,
       "step": 3500
     },
     {
-      "epoch": 0.6446859458463805,
-      "eval_loss": 2.1256344318389893,
-      "eval_runtime": 78.63,
-      "eval_samples_per_second": 15.338,
-      "eval_steps_per_second": 1.92,
+      "epoch": 0.6448047162859248,
+      "eval_loss": 2.0315287113189697,
+      "eval_runtime": 76.4108,
+      "eval_samples_per_second": 15.796,
+      "eval_steps_per_second": 1.976,
       "step": 3500
     },
     {
-      "epoch": 0.7367839381101492,
-      "grad_norm": 16.864044189453125,
-      "learning_rate": 3.772026769816418e-05,
-      "loss": 0.7034,
+      "epoch": 0.7369196757553427,
+      "grad_norm": 0.053296659141778946,
+      "learning_rate": 3.7718005404077624e-05,
+      "loss": 0.8387,
       "step": 4000
     },
     {
-      "epoch": 0.7367839381101492,
-      "eval_loss": 2.0000314712524414,
-      "eval_runtime": 78.6543,
-      "eval_samples_per_second": 15.333,
-      "eval_steps_per_second": 1.92,
+      "epoch": 0.7369196757553427,
+      "eval_loss": 2.3810112476348877,
+      "eval_runtime": 74.9737,
+      "eval_samples_per_second": 16.099,
+      "eval_steps_per_second": 2.014,
       "step": 4000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 16287,
+  "max_steps": 16284,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,

checkpoint-4000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d9885fd590ddfbffc4f8db1738c1993002dfa8c1603f7992a1cda9b93bbe7d34
+oid sha256:59e6c277b84a863e8b2dbf87019462e313e3d40a89afb31c644b26985cc0095f
 size 4847