Training in progress, step 4000, checkpoint

checkpoint-4000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eec64f06ce034d2f53284ede69e0a18c462d9c6399c4a59f426816e19834006a
+oid sha256:72ab603b10249cbcbb1067d702621fd78353251374ff39cd2c1eb0e51b91ac09
 size 1711848436

checkpoint-4000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2b783ae0e8fa9b15c1f02c717ecfac7e4d692eeaaf55edd1365e61c3c9c4104
+oid sha256:73efabeb88f686fbf340f1f551427d8b3522727462a6248b0661a5d3bc8872b9
 size 3424043887

checkpoint-4000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:00bac7e79638e389e1d4407565bcf1f331ae392be1b0784fec2a32495e23ed5d
+oid sha256:617e91b64af20002dd5d45dd9ed192f9149ef56f7e3dde7bba59e871ba98d117
 size 623

checkpoint-4000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.6968969106674194,
+  "best_metric": 0.7458393573760986,
   "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l57-l/checkpoint-4000",
-  "epoch": 0.7369196757553427,
+  "epoch": 0.42662116040955633,
   "eval_steps": 500,
   "global_step": 4000,
   "is_hyper_param_search": false,
@@ -9,128 +9,128 @@
   "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
     },
     {
-      "epoch": 0.7369196757553427,
-      "grad_norm": 116.85651397705078,
-      "learning_rate": 3.771800540407762e-07,
-      "loss": 0.1367,
+      "epoch": 0.42662116040955633,
+      "grad_norm": 3.2739710807800293,
+      "learning_rate": 4.2889647326507393e-07,
+      "loss": 0.1113,
       "step": 4000
     },
     {
-      "epoch": 0.7369196757553427,
-      "eval_loss": 0.6968969106674194,
-      "eval_runtime": 77.1872,
-      "eval_samples_per_second": 15.637,
-      "eval_steps_per_second": 1.956,
+      "epoch": 0.42662116040955633,
+      "eval_loss": 0.7458393573760986,
+      "eval_runtime": 62.1803,
+      "eval_samples_per_second": 15.873,
+      "eval_steps_per_second": 1.994,
       "step": 4000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 16284,
+  "max_steps": 28128,
   "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:ea34fd4333686a44a60d290faa4ae8e00df783131842d9ae3ef1b9e874a39349
+oid sha256:bd16a1abd112ae60a2700f590eb3bd73e96acd9a0a37b30bbfa3f8ee2dc23bc1
 size 4847