Step 28592

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.9464938640594482,
+    "eval_loss": 0.30927810072898865,
+    "eval_runtime": 134.3304,
+    "eval_samples": 3009,
+    "eval_samples_per_second": 22.4,
+    "eval_steps_per_second": 1.407,
+    "train_loss": 0.1776402507509504,
+    "train_runtime": 7042.0183,
+    "train_samples": 21714,
+    "train_samples_per_second": 12.334,
+    "train_steps_per_second": 4.111
+}

config.json

@@ -1,7 +1,7 @@
 {
   "_name_or_path": "bert-base-chinese",
   "architectures": [
-    "BertForQuestionAnswering"
+    "BertForMultipleChoice"
   ],
   "attention_probs_dropout_prob": 0.1,
   "classifier_dropout": null,
@@ -24,7 +24,7 @@
   "pooler_type": "first_token_transform",
   "position_embedding_type": "absolute",
   "torch_dtype": "float32",
-  "transformers_version": "4.19.0.dev0",
+  "transformers_version": "4.18.0",
   "type_vocab_size": 2,
   "use_cache": true,
   "vocab_size": 21128

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "eval_accuracy": 0.9464938640594482,
+    "eval_loss": 0.30927810072898865,
+    "eval_runtime": 134.3304,
+    "eval_samples": 3009,
+    "eval_samples_per_second": 22.4,
+    "eval_steps_per_second": 1.407
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bdf94027cfa5a7831cc3c9964fd477de5599f3dca3c791da60d5997ca17d8030
-size 406787613
+oid sha256:c8f62f85c3b9b343d4a3af5fc2b862807c8bfc411d34a21b043bd6d0702d7fa8
+size 143720448

tokenizer.json

@@ -1,21 +1,7 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 384,
-    "strategy": "OnlySecond",
-    "stride": 128
-  },
-  "padding": {
-    "strategy": {
-      "Fixed": 384
-    },
-    "direction": "Right",
-    "pad_to_multiple_of": null,
-    "pad_id": 0,
-    "pad_type_id": 0,
-    "pad_token": "[PAD]"
-  },
+  "truncation": null,
+  "padding": null,
   "added_tokens": [
     {
       "id": 0,

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.1776402507509504,
+    "train_runtime": 7042.0183,
+    "train_samples": 21714,
+    "train_samples_per_second": 12.334,
+    "train_steps_per_second": 4.111
+}

trainer_state.json

@@ -1,130 +1,367 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.988361119849009,
-  "global_step": 9500,
+  "epoch": 4.0,
+  "global_step": 28952,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.16,
-      "learning_rate": 2.8427178357974207e-05,
-      "loss": 1.375,
+      "epoch": 0.07,
+      "learning_rate": 4.913650179607627e-05,
+      "loss": 0.3611,
       "step": 500
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 2.6854356715948413e-05,
-      "loss": 0.8081,
+      "epoch": 0.14,
+      "learning_rate": 4.827300359215253e-05,
+      "loss": 0.2469,
       "step": 1000
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 2.5281535073922616e-05,
-      "loss": 0.6701,
+      "epoch": 0.21,
+      "learning_rate": 4.740950538822879e-05,
+      "loss": 0.307,
       "step": 1500
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 2.3708713431896822e-05,
-      "loss": 0.6539,
+      "epoch": 0.28,
+      "learning_rate": 4.6546007184305056e-05,
+      "loss": 0.3492,
       "step": 2000
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 2.213589178987103e-05,
-      "loss": 0.6176,
+      "epoch": 0.35,
+      "learning_rate": 4.568250898038132e-05,
+      "loss": 0.2257,
       "step": 2500
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 2.0563070147845238e-05,
-      "loss": 0.5754,
+      "epoch": 0.41,
+      "learning_rate": 4.4819010776457586e-05,
+      "loss": 0.293,
       "step": 3000
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 1.899024850581944e-05,
-      "loss": 0.4299,
+      "epoch": 0.48,
+      "learning_rate": 4.395551257253385e-05,
+      "loss": 0.3023,
       "step": 3500
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 1.7417426863793647e-05,
-      "loss": 0.3507,
+      "epoch": 0.55,
+      "learning_rate": 4.3092014368610116e-05,
+      "loss": 0.2834,
       "step": 4000
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 1.5844605221767853e-05,
-      "loss": 0.3356,
+      "epoch": 0.62,
+      "learning_rate": 4.2228516164686374e-05,
+      "loss": 0.3275,
       "step": 4500
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 1.4271783579742058e-05,
-      "loss": 0.3311,
+      "epoch": 0.69,
+      "learning_rate": 4.136501796076264e-05,
+      "loss": 0.2694,
       "step": 5000
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 1.2698961937716264e-05,
-      "loss": 0.3037,
+      "epoch": 0.76,
+      "learning_rate": 4.0501519756838904e-05,
+      "loss": 0.2631,
       "step": 5500
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 1.1126140295690468e-05,
-      "loss": 0.2988,
+      "epoch": 0.83,
+      "learning_rate": 3.963802155291517e-05,
+      "loss": 0.2773,
       "step": 6000
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 9.553318653664675e-06,
-      "loss": 0.247,
+      "epoch": 0.9,
+      "learning_rate": 3.8774523348991434e-05,
+      "loss": 0.3141,
       "step": 6500
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 7.980497011638881e-06,
-      "loss": 0.1414,
+      "epoch": 0.97,
+      "learning_rate": 3.79110251450677e-05,
+      "loss": 0.2893,
       "step": 7000
     },
     {
-      "epoch": 2.36,
-      "learning_rate": 6.407675369613086e-06,
-      "loss": 0.1402,
+      "epoch": 1.04,
+      "learning_rate": 3.7047526941143964e-05,
+      "loss": 0.2038,
       "step": 7500
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 4.834853727587292e-06,
-      "loss": 0.1443,
+      "epoch": 1.11,
+      "learning_rate": 3.618402873722023e-05,
+      "loss": 0.2157,
       "step": 8000
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 3.2620320855614974e-06,
-      "loss": 0.1387,
+      "epoch": 1.17,
+      "learning_rate": 3.5320530533296494e-05,
+      "loss": 0.2771,
       "step": 8500
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 1.6892104435357032e-06,
-      "loss": 0.1413,
+      "epoch": 1.24,
+      "learning_rate": 3.445703232937276e-05,
+      "loss": 0.2511,
       "step": 9000
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.1638880150990878e-07,
-      "loss": 0.1322,
+      "epoch": 1.31,
+      "learning_rate": 3.3593534125449024e-05,
+      "loss": 0.2678,
       "step": 9500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.273003592152529e-05,
+      "loss": 0.1795,
+      "step": 10000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.186653771760155e-05,
+      "loss": 0.2357,
+      "step": 10500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.100303951367781e-05,
+      "loss": 0.2187,
+      "step": 11000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.0139541309754078e-05,
+      "loss": 0.2315,
+      "step": 11500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.9276043105830343e-05,
+      "loss": 0.2478,
+      "step": 12000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.8412544901906608e-05,
+      "loss": 0.2444,
+      "step": 12500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.7549046697982873e-05,
+      "loss": 0.3056,
+      "step": 13000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.668554849405913e-05,
+      "loss": 0.2881,
+      "step": 13500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.5822050290135396e-05,
+      "loss": 0.3426,
+      "step": 14000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.495855208621166e-05,
+      "loss": 0.1894,
+      "step": 14500
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.4095053882287926e-05,
+      "loss": 0.1381,
+      "step": 15000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3231555678364188e-05,
+      "loss": 0.1359,
+      "step": 15500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.2368057474440453e-05,
+      "loss": 0.1179,
+      "step": 16000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.1504559270516718e-05,
+      "loss": 0.1048,
+      "step": 16500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.0641061066592983e-05,
+      "loss": 0.141,
+      "step": 17000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9777562862669248e-05,
+      "loss": 0.1726,
+      "step": 17500
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8914064658745513e-05,
+      "loss": 0.1307,
+      "step": 18000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.8050566454821775e-05,
+      "loss": 0.1875,
+      "step": 18500
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.718706825089804e-05,
+      "loss": 0.0803,
+      "step": 19000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.6323570046974305e-05,
+      "loss": 0.1028,
+      "step": 19500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5460071843050566e-05,
+      "loss": 0.1484,
+      "step": 20000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4596573639126831e-05,
+      "loss": 0.1042,
+      "step": 20500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.3733075435203096e-05,
+      "loss": 0.1325,
+      "step": 21000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.286957723127936e-05,
+      "loss": 0.1005,
+      "step": 21500
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.2006079027355625e-05,
+      "loss": 0.0656,
+      "step": 22000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.1142580823431888e-05,
+      "loss": 0.0533,
+      "step": 22500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0279082619508151e-05,
+      "loss": 0.0527,
+      "step": 23000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.415584415584416e-06,
+      "loss": 0.0241,
+      "step": 23500
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 8.55208621166068e-06,
+      "loss": 0.0463,
+      "step": 24000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 7.688588007736943e-06,
+      "loss": 0.0746,
+      "step": 24500
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 6.825089803813209e-06,
+      "loss": 0.0723,
+      "step": 25000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.961591599889472e-06,
+      "loss": 0.044,
+      "step": 25500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.098093395965737e-06,
+      "loss": 0.0451,
+      "step": 26000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.2345951920420006e-06,
+      "loss": 0.0695,
+      "step": 26500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.371096988118265e-06,
+      "loss": 0.0228,
+      "step": 27000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 2.507598784194529e-06,
+      "loss": 0.0257,
+      "step": 27500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.644100580270793e-06,
+      "loss": 0.0158,
+      "step": 28000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.806023763470573e-07,
+      "loss": 0.0432,
+      "step": 28500
+    },
+    {
+      "epoch": 4.0,
+      "step": 28952,
+      "total_flos": 9.075809388129869e+16,
+      "train_loss": 0.1776402507509504,
+      "train_runtime": 7042.0183,
+      "train_samples_per_second": 12.334,
+      "train_steps_per_second": 4.111
     }
   ],
-  "max_steps": 9537,
-  "num_train_epochs": 3,
-  "total_flos": 2.233891297525248e+16,
+  "max_steps": 28952,
+  "num_train_epochs": 4,
+  "total_flos": 9.075809388129869e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:380a2bd8ea374526b1001ab1982b8d9e44d3c139b79e06b2f8a1f4d10919bedc
-size 3119
+oid sha256:cc04f9294df23795653d167d6ac3debb3207bdd8002b4db84878c01887fcd2ef
+size 3055