Improved supervisor with deepset/tinyroberta-6l-768d

config.json

@@ -1,7 +1,7 @@
 {
-  "_name_or_path": "./ernie-autosave/xlmroberta/1714666085842",
+  "_name_or_path": "deepset/tinyroberta-6l-768d",
   "architectures": [
-    "XLMRobertaForSequenceClassification"
+    "RobertaForSequenceClassification"
   ],
   "attention_probs_dropout_prob": 0.1,
   "bos_token_id": 0,
@@ -16,15 +16,15 @@
   "language": "english",
   "layer_norm_eps": 1e-05,
   "max_position_embeddings": 514,
-  "model_type": "xlm-roberta",
-  "name": "XLMRoberta",
+  "model_type": "roberta",
+  "name": "Roberta",
   "num_attention_heads": 12,
-  "num_hidden_layers": 12,
-  "output_past": true,
+  "num_hidden_layers": 6,
   "pad_token_id": 1,
   "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
   "transformers_version": "4.24.0",
   "type_vocab_size": 1,
   "use_cache": true,
-  "vocab_size": 250002
+  "vocab_size": 50265
 }

special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:324b63541b17025ff3d54e116087638b4d561e9527e96ca10f08926599fb9464
-size 1112470336
+oid sha256:8260bd1e75816fb830aa02bed77289950e69ae35eab0524d6f9a697c28912087
+size 328629336

tokenizer_config.json

@@ -0,0 +1,18 @@
+{
+  "add_prefix_space": false,
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "errors": "replace",
+  "mask_token": "<mask>",
+  "max_len": 512,
+  "model_max_length": 512,
+  "name_or_path": "deepset/tinyroberta-6l-768d",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "special_tokens_map_file": null,
+  "tokenizer_class": "RobertaTokenizer",
+  "trim_offsets": true,
+  "unk_token": "<unk>"
+}

training.json

@@ -0,0 +1,32 @@
+{
+    "epochs": 2,
+    "learning_rate": 1e-05,
+    "model_name": "deepset/tinyroberta-6l-768d",
+    "evaluation": [
+        0.0008178171037303006,
+        0.0007848538627519552,
+        0.9992092289536872,
+        0.0007827766833699012,
+        0.9991875151053535,
+        0.0007897196128086037,
+        0.999194377478458,
+        0.0007847069126737383,
+        0.9991912057171757,
+        0.0007855461690940979,
+        0.9991958935677381,
+        0.000785352314220833,
+        0.9991970388540179,
+        0.0007879251506279772,
+        0.9991914060750224,
+        0.0007884886635244337,
+        0.9992020577928902,
+        0.0007888689475750166,
+        0.9992294733418028,
+        0.0007840781050665324,
+        0.9992103814093027,
+        0.0007828270352403521,
+        0.9992048671582784,
+        0.0007829428568245644,
+        0.9992107893169417
+    ]
+}