Training in progress, step 3500, checkpoint

checkpoint-3500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17791a516b4f73114fc2dfc5f9400ee1ab4d9561daec3596c7ca868a2624770b
+oid sha256:16848ce2260c74b866b85895c079b644c666ea2cb641eb77b69f0c4ff2745139
 size 1711848436

checkpoint-3500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e615bc9c471c9fc666f5af20a2b401877e2b84e50a4302f736360fd69631ca7
+oid sha256:6a50e3dc976b10827e928a557aaf6aba76c4ff82da45ad78fbba79f8dabec0aa
 size 3424043887

checkpoint-3500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9233cb0bba958a77fdb1c6b7d9bee0340b217e653dfb7039b5c9f521e3361168
+oid sha256:bac4a5dffca0cdb0a61239173d0604da73e5e26d6cec5c59a3769fc12563057a
 size 623

checkpoint-3500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.7097663283348083,
+  "best_metric": 0.7610095739364624,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l57-l/checkpoint-3500",
-  "epoch": 0.6448047162859248,
+  "epoch": 0.37329351535836175,
   "eval_steps": 500,
   "global_step": 3500,
   "is_hyper_param_search": false,
@@ -9,113 +9,113 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.09211495946941783,
-      "grad_norm": 2.6780805587768555,
-      "learning_rate": 4.84647506755097e-07,
-      "loss": 0.2705,
+      "epoch": 0.05332764505119454,
+      "grad_norm": 414.62841796875,
+      "learning_rate": 4.911120591581342e-07,
+      "loss": 0.3812,
       "step": 500
     },
     {
-      "epoch": 0.09211495946941783,
-      "eval_loss": 1.0680681467056274,
-      "eval_runtime": 74.5723,
-      "eval_samples_per_second": 16.186,
-      "eval_steps_per_second": 2.025,
+      "epoch": 0.05332764505119454,
+      "eval_loss": 1.1163015365600586,
+      "eval_runtime": 61.8962,
+      "eval_samples_per_second": 15.946,
+      "eval_steps_per_second": 2.003,
       "step": 500
     },
     {
-      "epoch": 0.18422991893883567,
-      "grad_norm": 8.301130294799805,
-      "learning_rate": 4.69295013510194e-07,
-      "loss": 0.2545,
+      "epoch": 0.10665529010238908,
+      "grad_norm": 34.08418655395508,
+      "learning_rate": 4.822241183162685e-07,
+      "loss": 0.2683,
       "step": 1000
     },
     {
-      "epoch": 0.18422991893883567,
-      "eval_loss": 0.9443553686141968,
-      "eval_runtime": 74.9643,
-      "eval_samples_per_second": 16.101,
-      "eval_steps_per_second": 2.014,
+      "epoch": 0.10665529010238908,
+      "eval_loss": 0.9684447646141052,
+      "eval_runtime": 61.821,
+      "eval_samples_per_second": 15.965,
+      "eval_steps_per_second": 2.006,
       "step": 1000
     },
     {
-      "epoch": 0.2763448784082535,
-      "grad_norm": 0.018898434937000275,
-      "learning_rate": 4.5394252026529107e-07,
-      "loss": 0.234,
+      "epoch": 0.1599829351535836,
+      "grad_norm": 350.4574890136719,
+      "learning_rate": 4.733361774744027e-07,
+      "loss": 0.2119,
       "step": 1500
     },
     {
-      "epoch": 0.2763448784082535,
-      "eval_loss": 0.876915693283081,
-      "eval_runtime": 75.6401,
-      "eval_samples_per_second": 15.957,
-      "eval_steps_per_second": 1.996,
+      "epoch": 0.1599829351535836,
+      "eval_loss": 0.9099885821342468,
+      "eval_runtime": 61.9606,
+      "eval_samples_per_second": 15.929,
+      "eval_steps_per_second": 2.001,
       "step": 1500
     },
     {
-      "epoch": 0.36845983787767134,
-      "grad_norm": 0.00010984255641233176,
-      "learning_rate": 4.385900270203881e-07,
-      "loss": 0.1539,
+      "epoch": 0.21331058020477817,
+      "grad_norm": 0.01787523180246353,
+      "learning_rate": 4.6444823663253695e-07,
+      "loss": 0.1889,
       "step": 2000
     },
     {
-      "epoch": 0.36845983787767134,
-      "eval_loss": 0.8415330648422241,
-      "eval_runtime": 76.9959,
-      "eval_samples_per_second": 15.676,
-      "eval_steps_per_second": 1.961,
+      "epoch": 0.21331058020477817,
+      "eval_loss": 0.8620074987411499,
+      "eval_runtime": 63.2824,
+      "eval_samples_per_second": 15.597,
+      "eval_steps_per_second": 1.959,
       "step": 2000
     },
     {
-      "epoch": 0.46057479734708917,
-      "grad_norm": 395.890625,
-      "learning_rate": 4.232375337754851e-07,
-      "loss": 0.1766,
+      "epoch": 0.2666382252559727,
+      "grad_norm": 118.66907501220703,
+      "learning_rate": 4.5556029579067116e-07,
+      "loss": 0.2071,
       "step": 2500
     },
     {
-      "epoch": 0.46057479734708917,
-      "eval_loss": 0.7660450339317322,
-      "eval_runtime": 76.733,
-      "eval_samples_per_second": 15.73,
-      "eval_steps_per_second": 1.968,
+      "epoch": 0.2666382252559727,
+      "eval_loss": 0.791786253452301,
+      "eval_runtime": 63.7932,
+      "eval_samples_per_second": 15.472,
+      "eval_steps_per_second": 1.944,
       "step": 2500
     },
     {
-      "epoch": 0.552689756816507,
-      "grad_norm": 363.6924743652344,
-      "learning_rate": 4.0788504053058217e-07,
-      "loss": 0.1679,
+      "epoch": 0.3199658703071672,
+      "grad_norm": 7.21087323185543e-09,
+      "learning_rate": 4.4667235494880547e-07,
+      "loss": 0.1588,
       "step": 3000
     },
     {
-      "epoch": 0.552689756816507,
-      "eval_loss": 0.7269155383110046,
-      "eval_runtime": 76.7928,
-      "eval_samples_per_second": 15.718,
-      "eval_steps_per_second": 1.966,
+      "epoch": 0.3199658703071672,
+      "eval_loss": 0.7657254338264465,
+      "eval_runtime": 62.4738,
+      "eval_samples_per_second": 15.799,
+      "eval_steps_per_second": 1.985,
       "step": 3000
     },
     {
-      "epoch": 0.6448047162859248,
-      "grad_norm": 28.914026260375977,
-      "learning_rate": 3.925325472856792e-07,
-      "loss": 0.1104,
+      "epoch": 0.37329351535836175,
+      "grad_norm": 0.0014836661284789443,
+      "learning_rate": 4.377844141069397e-07,
+      "loss": 0.1718,
       "step": 3500
     },
     {
-      "epoch": 0.6448047162859248,
-      "eval_loss": 0.7097663283348083,
-      "eval_runtime": 76.6897,
-      "eval_samples_per_second": 15.739,
-      "eval_steps_per_second": 1.969,
+      "epoch": 0.37329351535836175,
+      "eval_loss": 0.7610095739364624,
+      "eval_runtime": 63.9118,
+      "eval_samples_per_second": 15.443,
+      "eval_steps_per_second": 1.94,
       "step": 3500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 16284,
+  "max_steps": 28128,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,

checkpoint-3500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea34fd4333686a44a60d290faa4ae8e00df783131842d9ae3ef1b9e874a39349
+oid sha256:bd16a1abd112ae60a2700f590eb3bd73e96acd9a0a37b30bbfa3f8ee2dc23bc1
 size 4847