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: 3.4139
 
 ## Model description
 

all_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 10.0,
-    "eval_loss": 3.675844192504883,
-    "eval_runtime": 1.6492,
-    "eval_samples_per_second": 25.467,
-    "eval_steps_per_second": 0.606,
-    "test_loss": 4.198638439178467,
-    "test_runtime": 3.9304,
-    "test_samples_per_second": 21.372,
-    "test_steps_per_second": 0.509,
-    "train_loss": 2.1348843892415363,
-    "train_runtime": 100.2817,
-    "train_samples_per_second": 33.805,
-    "train_steps_per_second": 0.598
+    "eval_loss": 3.4139339923858643,
+    "eval_runtime": 1.4962,
+    "eval_samples_per_second": 28.071,
+    "eval_steps_per_second": 1.337,
+    "test_loss": 3.7118871212005615,
+    "test_runtime": 0.9098,
+    "test_samples_per_second": 92.324,
+    "test_steps_per_second": 3.297,
+    "train_loss": 1.4720843228426848,
+    "train_runtime": 106.1671,
+    "train_samples_per_second": 31.931,
+    "train_steps_per_second": 1.036
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 10.0,
-    "eval_loss": 3.675844192504883,
-    "eval_runtime": 1.6492,
-    "eval_samples_per_second": 25.467,
-    "eval_steps_per_second": 0.606
+    "eval_loss": 3.4139339923858643,
+    "eval_runtime": 1.4962,
+    "eval_samples_per_second": 28.071,
+    "eval_steps_per_second": 1.337
 }

test_results.json

@@ -1,6 +1,6 @@
 {
-    "test_loss": 4.198638439178467,
-    "test_runtime": 3.9304,
-    "test_samples_per_second": 21.372,
-    "test_steps_per_second": 0.509
+    "test_loss": 3.7118871212005615,
+    "test_runtime": 0.9098,
+    "test_samples_per_second": 92.324,
+    "test_steps_per_second": 3.297
 }

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 10.0,
-    "train_loss": 2.1348843892415363,
-    "train_runtime": 100.2817,
-    "train_samples_per_second": 33.805,
-    "train_steps_per_second": 0.598
+    "train_loss": 1.4720843228426848,
+    "train_runtime": 106.1671,
+    "train_samples_per_second": 31.931,
+    "train_steps_per_second": 1.036
 }

trainer_state.json

@@ -3,28 +3,28 @@
   "best_model_checkpoint": null,
   "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 60,
+  "global_step": 110,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 10.0,
-      "step": 60,
+      "step": 110,
       "total_flos": 197186123104740.0,
-      "train_loss": 2.1348843892415363,
-      "train_runtime": 100.2817,
-      "train_samples_per_second": 33.805,
-      "train_steps_per_second": 0.598
+      "train_loss": 1.4720843228426848,
+      "train_runtime": 106.1671,
+      "train_samples_per_second": 31.931,
+      "train_steps_per_second": 1.036
     }
   ],
   "logging_steps": 500,
-  "max_steps": 60,
+  "max_steps": 110,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
   "total_flos": 197186123104740.0,
-  "train_batch_size": 64,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null
 }