End of training

README.md

@@ -14,6 +14,8 @@ should probably proofread and complete it, then remove this comment. -->
 # cultureclip_sd_0226
 
 This model is a fine-tuned version of [openai/clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0008
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 10.0,
-    "eval_loss": 0.00021045817993581295,
-    "eval_runtime": 2.7761,
-    "eval_samples_per_second": 20.172,
-    "eval_steps_per_second": 2.521,
-    "total_flos": 9864573991549200.0,
-    "train_loss": 0.1267506488533907,
-    "train_runtime": 3359.5638,
-    "train_samples_per_second": 16.324,
-    "train_steps_per_second": 0.256
+    "epoch": 5.0,
+    "eval_loss": 0.0008266960503533483,
+    "eval_runtime": 2.8363,
+    "eval_samples_per_second": 19.744,
+    "eval_steps_per_second": 2.468,
+    "total_flos": 4932286995774600.0,
+    "train_loss": 0.030539887709103243,
+    "train_runtime": 1868.9702,
+    "train_samples_per_second": 14.671,
+    "train_steps_per_second": 0.918
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 10.0,
-    "eval_loss": 0.00021045817993581295,
-    "eval_runtime": 2.7761,
-    "eval_samples_per_second": 20.172,
-    "eval_steps_per_second": 2.521
+    "epoch": 5.0,
+    "eval_loss": 0.0008266960503533483,
+    "eval_runtime": 2.8363,
+    "eval_samples_per_second": 19.744,
+    "eval_steps_per_second": 2.468
 }

tokenizer.json

@@ -1,21 +1,7 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 77,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
-  "padding": {
-    "strategy": {
-      "Fixed": 77
-    },
-    "direction": "Right",
-    "pad_to_multiple_of": null,
-    "pad_id": 49407,
-    "pad_type_id": 0,
-    "pad_token": "<|endoftext|>"
-  },
+  "truncation": null,
+  "padding": null,
   "added_tokens": [
     {
       "id": 49406,

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 10.0,
-    "total_flos": 9864573991549200.0,
-    "train_loss": 0.1267506488533907,
-    "train_runtime": 3359.5638,
-    "train_samples_per_second": 16.324,
-    "train_steps_per_second": 0.256
+    "epoch": 5.0,
+    "total_flos": 4932286995774600.0,
+    "train_loss": 0.030539887709103243,
+    "train_runtime": 1868.9702,
+    "train_samples_per_second": 14.671,
+    "train_steps_per_second": 0.918
 }

trainer_state.json

@@ -1,34 +1,48 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 10.0,
-  "eval_steps": 500.0,
-  "global_step": 860,
+  "epoch": 5.0,
+  "eval_steps": 200.0,
+  "global_step": 1715,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 5.813953488372093,
-      "grad_norm": 5.5346174240112305,
-      "learning_rate": 3e-05,
-      "loss": 0.1406,
+      "epoch": 1.4577259475218658,
+      "grad_norm": 0.005862534046173096,
+      "learning_rate": 8.01980198019802e-07,
+      "loss": 0.0387,
       "step": 500
     },
     {
-      "epoch": 10.0,
-      "step": 860,
-      "total_flos": 9864573991549200.0,
-      "train_loss": 0.1267506488533907,
-      "train_runtime": 3359.5638,
-      "train_samples_per_second": 16.324,
-      "train_steps_per_second": 0.256
+      "epoch": 2.9154518950437316,
+      "grad_norm": 39.22230911254883,
+      "learning_rate": 4.7194719471947193e-07,
+      "loss": 0.0292,
+      "step": 1000
+    },
+    {
+      "epoch": 4.373177842565598,
+      "grad_norm": 29.80740737915039,
+      "learning_rate": 1.419141914191419e-07,
+      "loss": 0.0235,
+      "step": 1500
+    },
+    {
+      "epoch": 5.0,
+      "step": 1715,
+      "total_flos": 4932286995774600.0,
+      "train_loss": 0.030539887709103243,
+      "train_runtime": 1868.9702,
+      "train_samples_per_second": 14.671,
+      "train_steps_per_second": 0.918
     }
   ],
   "logging_steps": 500,
-  "max_steps": 860,
+  "max_steps": 1715,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -42,8 +56,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 9864573991549200.0,
-  "train_batch_size": 64,
+  "total_flos": 4932286995774600.0,
+  "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null
 }