Training in progress, epoch 1

added_tokens.json

@@ -1,7 +1,7 @@
 {
-  "[CLS]": 101,
-  "[MASK]": 103,
+  "[CLS]": 2,
+  "[MASK]": 4,
   "[PAD]": 0,
-  "[SEP]": 102,
-  "[UNK]": 100
+  "[SEP]": 3,
+  "[UNK]": 1
 }

config.json

@@ -1,14 +1,15 @@
 {
-  "_name_or_path": "ai-forever/ruBert-base",
+  "_name_or_path": "ai-forever/ruElectra-medium",
   "architectures": [
-    "BertForSequenceClassification"
+    "ElectraForSequenceClassification"
   ],
   "attention_probs_dropout_prob": 0.1,
   "classifier_dropout": null,
-  "directionality": "bidi",
+  "embedding_size": 576,
+  "generator_size": "0.25",
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "hidden_size": 576,
   "id2label": {
     "0": "appreciation",
     "1": "positive answer",
@@ -22,7 +23,7 @@
     "9": "yes/no question"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "intermediate_size": 2304,
   "label2id": {
     "LABEL_0": 0,
     "LABEL_1": 1,
@@ -37,20 +38,19 @@
   },
   "layer_norm_eps": 1e-12,
   "max_position_embeddings": 512,
-  "model_type": "bert",
-  "num_attention_heads": 12,
+  "model_type": "electra",
+  "num_attention_heads": 9,
   "num_hidden_layers": 12,
   "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",
   "problem_type": "single_label_classification",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
   "torch_dtype": "float32",
   "transformers_version": "4.34.0",
   "type_vocab_size": 2,
   "use_cache": true,
-  "vocab_size": 120138
+  "vocab_size": 64000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3aa034ec4c9a7b6fd647fe7efcb9a27480c75619de55d2916d157fccc2dd9e72
-size 713328878
+oid sha256:a66a682a6dc05daafb5eda4cefb1ca0f1e29cc0af27787e880f5fff6d47ab79b
+size 341529134

tokenizer_config.json

@@ -8,7 +8,7 @@
       "single_word": false,
       "special": true
     },
-    "100": {
+    "1": {
       "content": "[UNK]",
       "lstrip": false,
       "normalized": false,
@@ -16,7 +16,7 @@
       "single_word": false,
       "special": true
     },
-    "101": {
+    "2": {
       "content": "[CLS]",
       "lstrip": false,
       "normalized": false,
@@ -24,7 +24,7 @@
       "single_word": false,
       "special": true
     },
-    "102": {
+    "3": {
       "content": "[SEP]",
       "lstrip": false,
       "normalized": false,
@@ -32,7 +32,7 @@
       "single_word": false,
       "special": true
     },
-    "103": {
+    "4": {
       "content": "[MASK]",
       "lstrip": false,
       "normalized": false,
@@ -53,6 +53,6 @@
   "sep_token": "[SEP]",
   "strip_accents": null,
   "tokenize_chinese_chars": true,
-  "tokenizer_class": "BertTokenizer",
+  "tokenizer_class": "ElectraTokenizer",
   "unk_token": "[UNK]"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:afa5405c13038c5067a7d94b550b40924e22c83677db865f976cefc3cc21b8a4
+oid sha256:e3678b6ce7243be45f7d9665be5c222c4910edf038ab5706b549a2eaf1e1e065
 size 4536