Step 29000

config.json

@@ -1,34 +1,37 @@
 {
-  "_name_or_path": "hfl/chinese-roberta-wwm-ext",
+  "_name_or_path": "checkpoint_qa",
+  "_num_labels": 2,
   "architectures": [
-    "BertForQuestionAnswering"
+    "AlbertForQuestionAnswering"
   ],
-  "attention_probs_dropout_prob": 0.1,
-  "bos_token_id": 0,
-  "classifier_dropout": null,
-  "directionality": "bidi",
-  "eos_token_id": 2,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "attention_probs_dropout_prob": 0,
+  "bos_token_id": 2,
+  "classifier_dropout_prob": 0.1,
+  "down_scale_factor": 1,
+  "embedding_size": 128,
+  "eos_token_id": 3,
+  "gap_size": 0,
+  "hidden_act": "relu",
+  "hidden_dropout_prob": 0,
+  "hidden_size": 1024,
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "inner_group_num": 1,
+  "intermediate_size": 4096,
   "layer_norm_eps": 1e-12,
+  "layers_to_keep": [],
   "max_position_embeddings": 512,
-  "model_type": "bert",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "model_type": "albert",
+  "net_structure_type": 0,
+  "num_attention_heads": 16,
+  "num_hidden_groups": 1,
+  "num_hidden_layers": 24,
+  "num_memory_blocks": 0,
   "output_past": true,
   "pad_token_id": 0,
-  "pooler_fc_size": 768,
-  "pooler_num_attention_heads": 12,
-  "pooler_num_fc_layers": 3,
-  "pooler_size_per_head": 128,
-  "pooler_type": "first_token_transform",
   "position_embedding_type": "absolute",
+  "tokenizer_class": "BertTokenizer",
   "torch_dtype": "float32",
-  "transformers_version": "4.19.0.dev0",
+  "transformers_version": "4.18.0",
   "type_vocab_size": 2,
-  "use_cache": true,
   "vocab_size": 21128
 }

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc9f5e76c94e45ab3f220067d7b9ff866d3770e8dbfb86e57f4b84dad114245c
-size 813550411
+oid sha256:cc0b6f9ede25848b4440e3a3b75ae481910266027a76a6f6d4ccfceaf2cbcbc4
+size 114425856

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:36a99e0482d2f085159df1862301a6b484be8ebfc234a8af773418b421692756
-size 406787530
+oid sha256:26aa43a71ef20ec5338c4fdde9142dfd63ea5b4748f0718dc9c1adadc2356b50
+size 62015575

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aace6fbf4a2c391628706d200a7bb32cc72926305b5643db22364e4c38c6ca12
-size 14649
+oid sha256:ff0b7d688f72cecc3d6db8a5558dfafab2669c75076ea7f5e8d0318cba901963
+size 14503

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f5737c5d76961855bdeb09713dea4f64813c7a4768dc6718d1076810a1b558e
+oid sha256:55379860d5993322de7e50bc90fdcc01f8bd3b79ab77cc4a24298562169a1da1
 size 623

tokenizer.json

@@ -2,13 +2,13 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 512,
+    "max_length": 384,
     "strategy": "OnlySecond",
     "stride": 128
   },
   "padding": {
     "strategy": {
-      "Fixed": 512
+      "Fixed": 384
     },
     "direction": "Right",
     "pad_to_multiple_of": null,
@@ -68,7 +68,7 @@
     "clean_text": true,
     "handle_chinese_chars": true,
     "strip_accents": null,
-    "lowercase": true
+    "lowercase": false
   },
   "pre_tokenizer": {
     "type": "BertPreTokenizer"

tokenizer_config.json

@@ -1 +1 @@
-{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": "/home/shung/.cache/huggingface/transformers/d521373fc7ac35f63d56cf303de74a202403dcf1aaa792cd01f653694be59563.dd8bd9bfd3664b530ea4e645105f557769387b3da9f79bdb55ed556bdd80611d", "name_or_path": "hfl/chinese-roberta-wwm-ext", "tokenizer_class": "BertTokenizer"}
+{"do_lower_case": false, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "do_basic_tokenize": true, "never_split": null, "special_tokens_map_file": "/root/.cache/huggingface/transformers/1f536158f09966b7ddca1f2c06264b78535edb420d115380141a23361d136d78.dd8bd9bfd3664b530ea4e645105f557769387b3da9f79bdb55ed556bdd80611d", "name_or_path": "checkpoint_qa", "tokenizer_class": "BertTokenizer"}

trainer_state.json

@@ -1,346 +1,364 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 9.934971098265896,
-  "global_step": 27500,
+  "epoch": 9.122365523749608,
+  "global_step": 29000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.18,
-      "learning_rate": 2.945809248554913e-05,
-      "loss": 1.4801,
+      "epoch": 0.16,
+      "learning_rate": 2.764076753696131e-05,
+      "loss": 0.6436,
       "step": 500
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 2.8916184971098265e-05,
-      "loss": 0.9767,
+      "epoch": 0.31,
+      "learning_rate": 2.5281535073922616e-05,
+      "loss": 0.5523,
       "step": 1000
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 2.83742774566474e-05,
-      "loss": 0.8751,
+      "epoch": 0.47,
+      "learning_rate": 2.2922302610883924e-05,
+      "loss": 0.4916,
       "step": 1500
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 2.7832369942196533e-05,
-      "loss": 0.8366,
+      "epoch": 0.63,
+      "learning_rate": 2.0563070147845238e-05,
+      "loss": 0.4873,
       "step": 2000
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 2.7290462427745663e-05,
-      "loss": 0.7828,
+      "epoch": 0.79,
+      "learning_rate": 1.8203837684806546e-05,
+      "loss": 0.4573,
       "step": 2500
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 2.6748554913294797e-05,
-      "loss": 0.6836,
+      "epoch": 0.94,
+      "learning_rate": 1.5844605221767853e-05,
+      "loss": 0.442,
       "step": 3000
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 2.620664739884393e-05,
-      "loss": 0.5499,
+      "epoch": 1.1,
+      "learning_rate": 1.3485372758729161e-05,
+      "loss": 0.3298,
       "step": 3500
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 2.5664739884393065e-05,
-      "loss": 0.5279,
+      "epoch": 1.26,
+      "learning_rate": 1.1126140295690468e-05,
+      "loss": 0.2909,
       "step": 4000
     },
     {
-      "epoch": 1.63,
-      "learning_rate": 2.5122832369942196e-05,
-      "loss": 0.5512,
+      "epoch": 1.42,
+      "learning_rate": 8.766907832651778e-06,
+      "loss": 0.2898,
       "step": 4500
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 2.458092485549133e-05,
-      "loss": 0.5806,
+      "epoch": 1.57,
+      "learning_rate": 6.407675369613086e-06,
+      "loss": 0.2688,
       "step": 5000
     },
     {
-      "epoch": 1.99,
-      "learning_rate": 2.4039017341040463e-05,
-      "loss": 0.5244,
+      "epoch": 1.73,
+      "learning_rate": 4.048442906574395e-06,
+      "loss": 0.2815,
       "step": 5500
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 2.3497109826589597e-05,
-      "loss": 0.3951,
+      "epoch": 1.89,
+      "learning_rate": 1.6892104435357032e-06,
+      "loss": 0.2442,
       "step": 6000
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 2.2955202312138728e-05,
-      "loss": 0.3984,
+      "epoch": 2.04,
+      "learning_rate": 2.6932997798049702e-05,
+      "loss": 0.2957,
       "step": 6500
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 2.2413294797687862e-05,
-      "loss": 0.4088,
+      "epoch": 2.2,
+      "learning_rate": 2.669707455174583e-05,
+      "loss": 0.2735,
       "step": 7000
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 2.1871387283236992e-05,
-      "loss": 0.4008,
+      "epoch": 2.36,
+      "learning_rate": 2.6461151305441964e-05,
+      "loss": 0.307,
       "step": 7500
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 2.132947976878613e-05,
-      "loss": 0.4095,
+      "epoch": 2.52,
+      "learning_rate": 2.6225228059138093e-05,
+      "loss": 0.2875,
       "step": 8000
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 2.078757225433526e-05,
-      "loss": 0.347,
+      "epoch": 2.67,
+      "learning_rate": 2.5989304812834225e-05,
+      "loss": 0.3232,
       "step": 8500
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 2.0245664739884394e-05,
-      "loss": 0.3115,
+      "epoch": 2.83,
+      "learning_rate": 2.5753381566530358e-05,
+      "loss": 0.2979,
       "step": 9000
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 1.9703757225433524e-05,
-      "loss": 0.3224,
+      "epoch": 2.99,
+      "learning_rate": 2.5517458320226487e-05,
+      "loss": 0.312,
       "step": 9500
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 1.9161849710982662e-05,
-      "loss": 0.3421,
+      "epoch": 3.15,
+      "learning_rate": 2.5281535073922616e-05,
+      "loss": 0.2297,
       "step": 10000
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 1.8619942196531792e-05,
-      "loss": 0.3212,
+      "epoch": 3.3,
+      "learning_rate": 2.504561182761875e-05,
+      "loss": 0.2251,
       "step": 10500
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 1.8078034682080926e-05,
-      "loss": 0.3241,
+      "epoch": 3.46,
+      "learning_rate": 2.480968858131488e-05,
+      "loss": 0.2427,
       "step": 11000
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 1.7536127167630057e-05,
-      "loss": 0.2641,
+      "epoch": 3.62,
+      "learning_rate": 2.457376533501101e-05,
+      "loss": 0.2549,
       "step": 11500
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 1.6994219653179194e-05,
-      "loss": 0.2823,
+      "epoch": 3.77,
+      "learning_rate": 2.4337842088707142e-05,
+      "loss": 0.2566,
       "step": 12000
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 1.6452312138728324e-05,
-      "loss": 0.278,
+      "epoch": 3.93,
+      "learning_rate": 2.410191884240327e-05,
+      "loss": 0.2538,
       "step": 12500
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 1.591040462427746e-05,
-      "loss": 0.2774,
+      "epoch": 4.09,
+      "learning_rate": 2.3865995596099404e-05,
+      "loss": 0.1936,
       "step": 13000
     },
     {
-      "epoch": 4.88,
-      "learning_rate": 1.536849710982659e-05,
-      "loss": 0.2984,
+      "epoch": 4.25,
+      "learning_rate": 2.3630072349795533e-05,
+      "loss": 0.1742,
       "step": 13500
     },
     {
-      "epoch": 5.06,
-      "learning_rate": 1.4826589595375723e-05,
-      "loss": 0.276,
+      "epoch": 4.4,
+      "learning_rate": 2.3394149103491665e-05,
+      "loss": 0.1871,
       "step": 14000
     },
     {
-      "epoch": 5.24,
-      "learning_rate": 1.4284682080924855e-05,
-      "loss": 0.2485,
+      "epoch": 4.56,
+      "learning_rate": 2.3158225857187798e-05,
+      "loss": 0.1868,
       "step": 14500
     },
     {
-      "epoch": 5.42,
-      "learning_rate": 1.3742774566473989e-05,
-      "loss": 0.2513,
+      "epoch": 4.72,
+      "learning_rate": 2.2922302610883924e-05,
+      "loss": 0.1962,
       "step": 15000
     },
     {
-      "epoch": 5.6,
-      "learning_rate": 1.3200867052023121e-05,
-      "loss": 0.2741,
+      "epoch": 4.88,
+      "learning_rate": 2.2686379364580056e-05,
+      "loss": 0.2056,
       "step": 15500
     },
     {
-      "epoch": 5.78,
-      "learning_rate": 1.2658959537572255e-05,
-      "loss": 0.2581,
+      "epoch": 5.03,
+      "learning_rate": 2.245045611827619e-05,
+      "loss": 0.1918,
       "step": 16000
     },
     {
-      "epoch": 5.96,
-      "learning_rate": 1.2117052023121387e-05,
-      "loss": 0.2682,
+      "epoch": 5.19,
+      "learning_rate": 2.221453287197232e-05,
+      "loss": 0.145,
       "step": 16500
     },
     {
-      "epoch": 6.14,
-      "learning_rate": 1.1575144508670521e-05,
-      "loss": 0.2436,
+      "epoch": 5.35,
+      "learning_rate": 2.197860962566845e-05,
+      "loss": 0.14,
       "step": 17000
     },
     {
-      "epoch": 6.32,
-      "learning_rate": 1.1033236994219653e-05,
-      "loss": 0.2364,
+      "epoch": 5.5,
+      "learning_rate": 2.174268637936458e-05,
+      "loss": 0.1442,
       "step": 17500
     },
     {
-      "epoch": 6.5,
-      "learning_rate": 1.0491329479768787e-05,
-      "loss": 0.2336,
+      "epoch": 5.66,
+      "learning_rate": 2.150676313306071e-05,
+      "loss": 0.1617,
       "step": 18000
     },
     {
-      "epoch": 6.68,
-      "learning_rate": 9.94942196531792e-06,
-      "loss": 0.2354,
+      "epoch": 5.82,
+      "learning_rate": 2.127083988675684e-05,
+      "loss": 0.1525,
       "step": 18500
     },
     {
-      "epoch": 6.86,
-      "learning_rate": 9.407514450867053e-06,
-      "loss": 0.2486,
+      "epoch": 5.98,
+      "learning_rate": 2.1034916640452973e-05,
+      "loss": 0.159,
       "step": 19000
     },
     {
-      "epoch": 7.04,
-      "learning_rate": 8.865606936416185e-06,
-      "loss": 0.2461,
+      "epoch": 6.13,
+      "learning_rate": 2.0798993394149106e-05,
+      "loss": 0.1022,
       "step": 19500
     },
     {
-      "epoch": 7.23,
-      "learning_rate": 8.323699421965318e-06,
-      "loss": 0.2161,
+      "epoch": 6.29,
+      "learning_rate": 2.0563070147845238e-05,
+      "loss": 0.1082,
       "step": 20000
     },
     {
-      "epoch": 7.41,
-      "learning_rate": 7.781791907514452e-06,
-      "loss": 0.2437,
+      "epoch": 6.45,
+      "learning_rate": 2.0327146901541364e-05,
+      "loss": 0.1143,
       "step": 20500
     },
     {
-      "epoch": 7.59,
-      "learning_rate": 7.239884393063584e-06,
-      "loss": 0.2121,
+      "epoch": 6.61,
+      "learning_rate": 2.0091223655237496e-05,
+      "loss": 0.1268,
       "step": 21000
     },
     {
-      "epoch": 7.77,
-      "learning_rate": 6.697976878612717e-06,
-      "loss": 0.2227,
+      "epoch": 6.76,
+      "learning_rate": 1.985530040893363e-05,
+      "loss": 0.1178,
       "step": 21500
     },
     {
-      "epoch": 7.95,
-      "learning_rate": 6.15606936416185e-06,
-      "loss": 0.2341,
+      "epoch": 6.92,
+      "learning_rate": 1.9619377162629758e-05,
+      "loss": 0.1102,
       "step": 22000
     },
     {
-      "epoch": 8.13,
-      "learning_rate": 5.614161849710983e-06,
-      "loss": 0.2182,
+      "epoch": 7.08,
+      "learning_rate": 1.938345391632589e-05,
+      "loss": 0.0948,
       "step": 22500
     },
     {
-      "epoch": 8.31,
-      "learning_rate": 5.072254335260116e-06,
-      "loss": 0.2148,
+      "epoch": 7.23,
+      "learning_rate": 1.914753067002202e-05,
+      "loss": 0.0709,
       "step": 23000
     },
     {
-      "epoch": 8.49,
-      "learning_rate": 4.530346820809248e-06,
-      "loss": 0.2247,
+      "epoch": 7.39,
+      "learning_rate": 1.891160742371815e-05,
+      "loss": 0.0904,
       "step": 23500
     },
     {
-      "epoch": 8.67,
-      "learning_rate": 3.988439306358381e-06,
-      "loss": 0.2268,
+      "epoch": 7.55,
+      "learning_rate": 1.867568417741428e-05,
+      "loss": 0.114,
       "step": 24000
     },
     {
-      "epoch": 8.85,
-      "learning_rate": 3.4465317919075147e-06,
-      "loss": 0.2187,
+      "epoch": 7.71,
+      "learning_rate": 1.8439760931110413e-05,
+      "loss": 0.1083,
       "step": 24500
     },
     {
-      "epoch": 9.03,
-      "learning_rate": 2.9046242774566473e-06,
-      "loss": 0.2188,
+      "epoch": 7.86,
+      "learning_rate": 1.8203837684806546e-05,
+      "loss": 0.0931,
       "step": 25000
     },
     {
-      "epoch": 9.21,
-      "learning_rate": 2.3627167630057803e-06,
-      "loss": 0.2196,
+      "epoch": 8.02,
+      "learning_rate": 1.796791443850267e-05,
+      "loss": 0.0845,
       "step": 25500
     },
     {
-      "epoch": 9.39,
-      "learning_rate": 1.8208092485549132e-06,
-      "loss": 0.2114,
+      "epoch": 8.18,
+      "learning_rate": 1.7731991192198804e-05,
+      "loss": 0.0557,
       "step": 26000
     },
     {
-      "epoch": 9.57,
-      "learning_rate": 1.2789017341040462e-06,
-      "loss": 0.2159,
+      "epoch": 8.34,
+      "learning_rate": 1.7496067945894936e-05,
+      "loss": 0.0824,
       "step": 26500
     },
     {
-      "epoch": 9.75,
-      "learning_rate": 7.369942196531793e-07,
-      "loss": 0.2153,
+      "epoch": 8.49,
+      "learning_rate": 1.726014469959107e-05,
+      "loss": 0.064,
       "step": 27000
     },
     {
-      "epoch": 9.93,
-      "learning_rate": 1.9508670520231215e-07,
-      "loss": 0.2102,
+      "epoch": 8.65,
+      "learning_rate": 1.7024221453287198e-05,
+      "loss": 0.076,
       "step": 27500
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 1.678829820698333e-05,
+      "loss": 0.0699,
+      "step": 28000
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 1.655237496067946e-05,
+      "loss": 0.0771,
+      "step": 28500
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 1.631645171437559e-05,
+      "loss": 0.0455,
+      "step": 29000
     }
   ],
-  "max_steps": 27680,
-  "num_train_epochs": 10,
-  "total_flos": 7.184484974834688e+16,
+  "max_steps": 63580,
+  "num_train_epochs": 20,
+  "total_flos": 1.020599417048832e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc79ec42c6d6f21322d15ba12920b9b996dc67964be1f40bbfdade533067a0d6
-size 3119
+oid sha256:9fb7de8132dd45150ceed627536a3d3a28474d6c4c4c99909b69d24204dcc308
+size 3055