End of training

README.md

@@ -15,6 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
 # clip-finetuned
 
 This model is a fine-tuned version of [openai/clip-vit-base-patch32](https://huggingface.co/openai/clip-vit-base-patch32) on the imagefolder dataset.
+It achieves the following results on the evaluation set:
+- Loss: 3.4175
 
 ## Model description
 

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 30.0,
-    "eval_loss": 5.64693546295166,
-    "eval_runtime": 1.9784,
-    "eval_samples_per_second": 21.229,
-    "eval_steps_per_second": 1.011,
-    "test_loss": 5.721177577972412,
-    "test_runtime": 0.9455,
-    "test_samples_per_second": 88.843,
-    "test_steps_per_second": 3.173,
-    "train_loss": 0.7521804347182766,
-    "train_runtime": 313.3926,
-    "train_samples_per_second": 32.451,
-    "train_steps_per_second": 1.053
+    "epoch": 10.0,
+    "eval_loss": 3.417513132095337,
+    "eval_runtime": 2.2084,
+    "eval_samples_per_second": 19.018,
+    "eval_steps_per_second": 0.906,
+    "test_loss": 3.6892688274383545,
+    "test_runtime": 0.9331,
+    "test_samples_per_second": 90.022,
+    "test_steps_per_second": 3.215,
+    "train_loss": 1.472021484375,
+    "train_runtime": 106.3445,
+    "train_samples_per_second": 31.878,
+    "train_steps_per_second": 1.034
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 30.0,
-    "eval_loss": 5.64693546295166,
-    "eval_runtime": 1.9784,
-    "eval_samples_per_second": 21.229,
-    "eval_steps_per_second": 1.011
+    "epoch": 10.0,
+    "eval_loss": 3.417513132095337,
+    "eval_runtime": 2.2084,
+    "eval_samples_per_second": 19.018,
+    "eval_steps_per_second": 0.906
 }

test_results.json

@@ -1,6 +1,6 @@
 {
-    "test_loss": 5.721177577972412,
-    "test_runtime": 0.9455,
-    "test_samples_per_second": 88.843,
-    "test_steps_per_second": 3.173
+    "test_loss": 3.6892688274383545,
+    "test_runtime": 0.9331,
+    "test_samples_per_second": 90.022,
+    "test_steps_per_second": 3.215
 }

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 30.0,
-    "train_loss": 0.7521804347182766,
-    "train_runtime": 313.3926,
-    "train_samples_per_second": 32.451,
-    "train_steps_per_second": 1.053
+    "epoch": 10.0,
+    "train_loss": 1.472021484375,
+    "train_runtime": 106.3445,
+    "train_samples_per_second": 31.878,
+    "train_steps_per_second": 1.034
 }

trainer_state.json

@@ -1,29 +1,29 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 30.0,
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 330,
+  "global_step": 110,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 30.0,
-      "step": 330,
-      "total_flos": 591558369314220.0,
-      "train_loss": 0.7521804347182766,
-      "train_runtime": 313.3926,
-      "train_samples_per_second": 32.451,
-      "train_steps_per_second": 1.053
+      "epoch": 10.0,
+      "step": 110,
+      "total_flos": 197186123104740.0,
+      "train_loss": 1.472021484375,
+      "train_runtime": 106.3445,
+      "train_samples_per_second": 31.878,
+      "train_steps_per_second": 1.034
     }
   ],
   "logging_steps": 500,
-  "max_steps": 330,
+  "max_steps": 110,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 30,
+  "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 591558369314220.0,
+  "total_flos": 197186123104740.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null