End of training

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 4.0,
-    "eval_loss": 0.49561700224876404,
-    "eval_runtime": 72.0108,
-    "eval_samples_per_second": 16.761,
-    "eval_steps_per_second": 2.097,
-    "total_flos": 7811792839259640.0,
-    "train_loss": 0.07694113771750283,
-    "train_runtime": 16520.0957,
-    "train_samples_per_second": 2.629,
-    "train_steps_per_second": 1.314
+    "eval_loss": 0.17657655477523804,
+    "eval_runtime": 7.3821,
+    "eval_samples_per_second": 16.391,
+    "eval_steps_per_second": 2.167,
+    "total_flos": 8593187978196360.0,
+    "train_loss": 0.0035514953327706023,
+    "train_runtime": 1405.6232,
+    "train_samples_per_second": 33.984,
+    "train_steps_per_second": 16.992
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 4.0,
-    "eval_loss": 0.49561700224876404,
-    "eval_runtime": 72.0108,
-    "eval_samples_per_second": 16.761,
-    "eval_steps_per_second": 2.097
+    "eval_loss": 0.17657655477523804,
+    "eval_runtime": 7.3821,
+    "eval_samples_per_second": 16.391,
+    "eval_steps_per_second": 2.167
 }

runs/Aug23_10-36-57_winter-park/events.out.tfevents.1724432742.winter-park.1153123.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c02e9bee007be85dbdcda75ef4b7c638613cf554c9da4539f759c6239bc99ce
+size 316

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 4.0,
-    "total_flos": 7811792839259640.0,
-    "train_loss": 0.07694113771750283,
-    "train_runtime": 16520.0957,
-    "train_samples_per_second": 2.629,
-    "train_steps_per_second": 1.314
+    "total_flos": 8593187978196360.0,
+    "train_loss": 0.0035514953327706023,
+    "train_runtime": 1405.6232,
+    "train_samples_per_second": 33.984,
+    "train_steps_per_second": 16.992
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.49561700224876404,
-  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-21000",
+  "best_metric": 0.17657655477523804,
+  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-23500",
   "epoch": 4.0,
   "eval_steps": 500,
-  "global_step": 21712,
+  "global_step": 23884,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -653,18 +653,78 @@
       "eval_steps_per_second": 2.025,
       "step": 21500
     },
+    {
+      "epoch": 3.6844749623178696,
+      "grad_norm": 0.022144177928566933,
+      "learning_rate": 3.944062971026629e-08,
+      "loss": 0.0432,
+      "step": 22000
+    },
+    {
+      "epoch": 3.6844749623178696,
+      "eval_loss": 0.17835669219493866,
+      "eval_runtime": 7.4507,
+      "eval_samples_per_second": 16.24,
+      "eval_steps_per_second": 2.147,
+      "step": 22000
+    },
+    {
+      "epoch": 3.7682130296432756,
+      "grad_norm": 2.4324751279891643e-07,
+      "learning_rate": 2.897337129459052e-08,
+      "loss": 0.0461,
+      "step": 22500
+    },
+    {
+      "epoch": 3.7682130296432756,
+      "eval_loss": 0.17669138312339783,
+      "eval_runtime": 7.5623,
+      "eval_samples_per_second": 16.0,
+      "eval_steps_per_second": 2.116,
+      "step": 22500
+    },
+    {
+      "epoch": 3.851951096968682,
+      "grad_norm": 4.172260048562748e-07,
+      "learning_rate": 1.8506112878914754e-08,
+      "loss": 0.0513,
+      "step": 23000
+    },
+    {
+      "epoch": 3.851951096968682,
+      "eval_loss": 0.1774093359708786,
+      "eval_runtime": 7.8193,
+      "eval_samples_per_second": 15.475,
+      "eval_steps_per_second": 2.046,
+      "step": 23000
+    },
+    {
+      "epoch": 3.935689164294088,
+      "grad_norm": 0.5274502038955688,
+      "learning_rate": 8.038854463238988e-09,
+      "loss": 0.0326,
+      "step": 23500
+    },
+    {
+      "epoch": 3.935689164294088,
+      "eval_loss": 0.17657655477523804,
+      "eval_runtime": 7.7889,
+      "eval_samples_per_second": 15.535,
+      "eval_steps_per_second": 2.054,
+      "step": 23500
+    },
     {
       "epoch": 4.0,
-      "step": 21712,
-      "total_flos": 7811792839259640.0,
-      "train_loss": 0.07694113771750283,
-      "train_runtime": 16520.0957,
-      "train_samples_per_second": 2.629,
-      "train_steps_per_second": 1.314
+      "step": 23884,
+      "total_flos": 8593187978196360.0,
+      "train_loss": 0.0035514953327706023,
+      "train_runtime": 1405.6232,
+      "train_samples_per_second": 33.984,
+      "train_steps_per_second": 16.992
     }
   ],
   "logging_steps": 500,
-  "max_steps": 21712,
+  "max_steps": 23884,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,
@@ -680,7 +740,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 7811792839259640.0,
+  "total_flos": 8593187978196360.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null