End of training

README.md

@@ -1,4 +1,6 @@
 ---
+license: gpl-3.0
+base_model: ckiplab/bert-base-chinese
 tags:
 - generated_from_trainer
 model-index:
@@ -11,7 +13,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # clip-roberta-finetuned
 
-This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
+This model is a fine-tuned version of [ckiplab/bert-base-chinese](https://huggingface.co/ckiplab/bert-base-chinese) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0543
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 10.0,
-    "eval_loss": 2.635524034500122,
-    "eval_runtime": 3.7818,
-    "eval_samples_per_second": 331.589,
-    "eval_steps_per_second": 5.289,
-    "total_flos": 1658811247948800.0,
-    "train_loss": 3.298740234375,
-    "train_runtime": 93.9998,
-    "train_samples_per_second": 133.405,
-    "train_steps_per_second": 2.128
+    "epoch": 100.0,
+    "eval_loss": 0.054326437413692474,
+    "eval_runtime": 3.844,
+    "eval_samples_per_second": 326.223,
+    "eval_steps_per_second": 3.382,
+    "total_flos": 1.6588112479488e+16,
+    "train_loss": 0.4321923828125,
+    "train_runtime": 849.2414,
+    "train_samples_per_second": 147.661,
+    "train_steps_per_second": 1.884
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 10.0,
-    "eval_loss": 2.635524034500122,
-    "eval_runtime": 3.7818,
-    "eval_samples_per_second": 331.589,
-    "eval_steps_per_second": 5.289
+    "epoch": 100.0,
+    "eval_loss": 0.054326437413692474,
+    "eval_runtime": 3.844,
+    "eval_samples_per_second": 326.223,
+    "eval_steps_per_second": 3.382
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 10.0,
-    "total_flos": 1658811247948800.0,
-    "train_loss": 3.298740234375,
-    "train_runtime": 93.9998,
-    "train_samples_per_second": 133.405,
-    "train_steps_per_second": 2.128
+    "epoch": 100.0,
+    "total_flos": 1.6588112479488e+16,
+    "train_loss": 0.4321923828125,
+    "train_runtime": 849.2414,
+    "train_samples_per_second": 147.661,
+    "train_steps_per_second": 1.884
 }

trainer_state.json

@@ -1,27 +1,48 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 10.0,
+  "epoch": 100.0,
   "eval_steps": 500,
-  "global_step": 200,
+  "global_step": 1600,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 10.0,
-      "step": 200,
-      "total_flos": 1658811247948800.0,
-      "train_loss": 3.298740234375,
-      "train_runtime": 93.9998,
-      "train_samples_per_second": 133.405,
-      "train_steps_per_second": 2.128
+      "epoch": 31.25,
+      "grad_norm": 1.9496452808380127,
+      "learning_rate": 3.4375e-05,
+      "loss": 1.2881,
+      "step": 500
+    },
+    {
+      "epoch": 62.5,
+      "grad_norm": 0.6210575699806213,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.0667,
+      "step": 1000
+    },
+    {
+      "epoch": 93.75,
+      "grad_norm": 0.8036883473396301,
+      "learning_rate": 3.125e-06,
+      "loss": 0.0245,
+      "step": 1500
+    },
+    {
+      "epoch": 100.0,
+      "step": 1600,
+      "total_flos": 1.6588112479488e+16,
+      "train_loss": 0.4321923828125,
+      "train_runtime": 849.2414,
+      "train_samples_per_second": 147.661,
+      "train_steps_per_second": 1.884
     }
   ],
   "logging_steps": 500,
-  "max_steps": 200,
+  "max_steps": 1600,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 100,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -35,8 +56,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 1658811247948800.0,
-  "train_batch_size": 64,
+  "total_flos": 1.6588112479488e+16,
+  "train_batch_size": 80,
   "trial_name": null,
   "trial_params": null
 }