End of training

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [OFA-Sys/chinese-clip-vit-base-patch16](https://huggingface.co/OFA-Sys/chinese-clip-vit-base-patch16) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 4.8862
+- Loss: 4.9099
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 60.0,
-    "eval_loss": 3.5839552879333496,
-    "eval_runtime": 31.5919,
-    "eval_samples_per_second": 307.516,
-    "eval_steps_per_second": 6.995,
+    "eval_loss": 4.9098591804504395,
+    "eval_runtime": 31.7328,
+    "eval_samples_per_second": 306.15,
+    "eval_steps_per_second": 6.964,
     "total_flos": 1.4015777943683174e+18,
-    "train_loss": 2.0418326404735745,
-    "train_runtime": 30807.4628,
-    "train_samples_per_second": 115.158,
-    "train_steps_per_second": 2.88
+    "train_loss": 2.04994026615675,
+    "train_runtime": 30910.218,
+    "train_samples_per_second": 114.776,
+    "train_steps_per_second": 2.871
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 60.0,
-    "eval_loss": 3.5839552879333496,
-    "eval_runtime": 31.5919,
-    "eval_samples_per_second": 307.516,
-    "eval_steps_per_second": 6.995
+    "eval_loss": 4.9098591804504395,
+    "eval_runtime": 31.7328,
+    "eval_samples_per_second": 306.15,
+    "eval_steps_per_second": 6.964
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 60.0,
     "total_flos": 1.4015777943683174e+18,
-    "train_loss": 2.0418326404735745,
-    "train_runtime": 30807.4628,
-    "train_samples_per_second": 115.158,
-    "train_steps_per_second": 2.88
+    "train_loss": 2.04994026615675,
+    "train_runtime": 30910.218,
+    "train_samples_per_second": 114.776,
+    "train_steps_per_second": 2.871
 }

trainer_state.json

@@ -10,162 +10,162 @@
   "log_history": [
     {
       "epoch": 6.0,
-      "grad_norm": 2.476245641708374,
-      "learning_rate": 9.0009015100293e-06,
-      "loss": 2.4036,
+      "grad_norm": 2.8650519847869873,
+      "learning_rate": 9.001014198782962e-06,
+      "loss": 2.4122,
       "step": 8874
     },
     {
       "epoch": 6.0,
-      "eval_loss": 3.011286497116089,
-      "eval_runtime": 32.2427,
-      "eval_samples_per_second": 301.308,
-      "eval_steps_per_second": 6.854,
+      "eval_loss": 3.9393680095672607,
+      "eval_runtime": 31.1663,
+      "eval_samples_per_second": 311.715,
+      "eval_steps_per_second": 7.091,
       "step": 8874
     },
     {
       "epoch": 12.0,
-      "grad_norm": 10.134785652160645,
+      "grad_norm": 11.994268417358398,
       "learning_rate": 8.001239576290288e-06,
-      "loss": 2.1954,
+      "loss": 2.2167,
       "step": 17748
     },
     {
       "epoch": 12.0,
-      "eval_loss": 3.1597039699554443,
-      "eval_runtime": 31.8518,
-      "eval_samples_per_second": 305.006,
-      "eval_steps_per_second": 6.938,
+      "eval_loss": 4.1646857261657715,
+      "eval_runtime": 31.5415,
+      "eval_samples_per_second": 308.007,
+      "eval_steps_per_second": 7.007,
       "step": 17748
     },
     {
       "epoch": 18.0,
-      "grad_norm": 14.89818000793457,
+      "grad_norm": 20.35555076599121,
       "learning_rate": 7.001690331304937e-06,
-      "loss": 2.0709,
+      "loss": 2.0965,
       "step": 26622
     },
     {
       "epoch": 18.0,
-      "eval_loss": 3.206042528152466,
-      "eval_runtime": 33.1331,
-      "eval_samples_per_second": 293.211,
-      "eval_steps_per_second": 6.67,
+      "eval_loss": 4.430016994476318,
+      "eval_runtime": 31.5349,
+      "eval_samples_per_second": 308.071,
+      "eval_steps_per_second": 7.008,
       "step": 26622
     },
     {
       "epoch": 24.0,
-      "grad_norm": 20.7427921295166,
+      "grad_norm": 3.9490978717803955,
       "learning_rate": 6.002141086319586e-06,
-      "loss": 2.0132,
+      "loss": 2.0238,
       "step": 35496
     },
     {
       "epoch": 24.0,
-      "eval_loss": 3.307133913040161,
-      "eval_runtime": 31.6331,
-      "eval_samples_per_second": 307.115,
-      "eval_steps_per_second": 6.986,
+      "eval_loss": 4.574044227600098,
+      "eval_runtime": 31.6749,
+      "eval_samples_per_second": 306.709,
+      "eval_steps_per_second": 6.977,
       "step": 35496
     },
     {
       "epoch": 30.0,
-      "grad_norm": 10.81413745880127,
+      "grad_norm": 9.400227546691895,
       "learning_rate": 5.0025918413342355e-06,
-      "loss": 1.9783,
+      "loss": 1.9938,
       "step": 44370
     },
     {
       "epoch": 30.0,
-      "eval_loss": 3.3543457984924316,
-      "eval_runtime": 31.6425,
-      "eval_samples_per_second": 307.024,
-      "eval_steps_per_second": 6.984,
+      "eval_loss": 4.62649393081665,
+      "eval_runtime": 31.7226,
+      "eval_samples_per_second": 306.249,
+      "eval_steps_per_second": 6.967,
       "step": 44370
     },
     {
       "epoch": 36.0,
-      "grad_norm": 13.179546356201172,
-      "learning_rate": 4.003155285102547e-06,
-      "loss": 1.9672,
+      "grad_norm": 5.468742370605469,
+      "learning_rate": 4.003042596348885e-06,
+      "loss": 1.973,
       "step": 53244
     },
     {
       "epoch": 36.0,
-      "eval_loss": 3.45923113822937,
-      "eval_runtime": 31.6946,
-      "eval_samples_per_second": 306.519,
-      "eval_steps_per_second": 6.973,
+      "eval_loss": 4.671382427215576,
+      "eval_runtime": 31.6041,
+      "eval_samples_per_second": 307.396,
+      "eval_steps_per_second": 6.993,
       "step": 53244
     },
     {
       "epoch": 42.0,
-      "grad_norm": 35.83116149902344,
+      "grad_norm": 0.6607534289360046,
       "learning_rate": 3.003606040117197e-06,
-      "loss": 1.9536,
+      "loss": 1.9583,
       "step": 62118
     },
     {
       "epoch": 42.0,
-      "eval_loss": 3.473649740219116,
-      "eval_runtime": 31.7269,
-      "eval_samples_per_second": 306.207,
-      "eval_steps_per_second": 6.966,
+      "eval_loss": 4.793060302734375,
+      "eval_runtime": 31.7099,
+      "eval_samples_per_second": 306.372,
+      "eval_steps_per_second": 6.969,
       "step": 62118
     },
     {
       "epoch": 48.0,
-      "grad_norm": 2.7362611293792725,
-      "learning_rate": 2.0041694838855083e-06,
-      "loss": 1.9473,
+      "grad_norm": 1.4975688457489014,
+      "learning_rate": 2.004056795131846e-06,
+      "loss": 1.9466,
       "step": 70992
     },
     {
       "epoch": 48.0,
-      "eval_loss": 3.531722068786621,
-      "eval_runtime": 31.6356,
-      "eval_samples_per_second": 307.091,
-      "eval_steps_per_second": 6.986,
+      "eval_loss": 4.7913103103637695,
+      "eval_runtime": 33.0493,
+      "eval_samples_per_second": 293.955,
+      "eval_steps_per_second": 6.687,
       "step": 70992
     },
     {
       "epoch": 54.0,
-      "grad_norm": 3.277851104736328,
-      "learning_rate": 1.0046202389001578e-06,
-      "loss": 1.9479,
+      "grad_norm": 2.9576919078826904,
+      "learning_rate": 1.0045075501464953e-06,
+      "loss": 1.9415,
       "step": 79866
     },
     {
       "epoch": 54.0,
-      "eval_loss": 3.5619990825653076,
-      "eval_runtime": 31.6245,
-      "eval_samples_per_second": 307.199,
-      "eval_steps_per_second": 6.988,
+      "eval_loss": 4.844839572906494,
+      "eval_runtime": 31.7934,
+      "eval_samples_per_second": 305.566,
+      "eval_steps_per_second": 6.951,
       "step": 79866
     },
     {
       "epoch": 60.0,
-      "grad_norm": 2.524662733078003,
+      "grad_norm": 0.5801959037780762,
       "learning_rate": 5.070993914807302e-09,
-      "loss": 1.9411,
+      "loss": 1.9369,
       "step": 88740
     },
     {
       "epoch": 60.0,
-      "eval_loss": 3.5839552879333496,
-      "eval_runtime": 31.4889,
-      "eval_samples_per_second": 308.522,
-      "eval_steps_per_second": 7.018,
+      "eval_loss": 4.886190891265869,
+      "eval_runtime": 31.6305,
+      "eval_samples_per_second": 307.141,
+      "eval_steps_per_second": 6.987,
       "step": 88740
     },
     {
       "epoch": 60.0,
       "step": 88740,
       "total_flos": 1.4015777943683174e+18,
-      "train_loss": 2.0418326404735745,
-      "train_runtime": 30807.4628,
-      "train_samples_per_second": 115.158,
-      "train_steps_per_second": 2.88
+      "train_loss": 2.04994026615675,
+      "train_runtime": 30910.218,
+      "train_samples_per_second": 114.776,
+      "train_steps_per_second": 2.871
     }
   ],
   "logging_steps": 8874,