Training in progress, epoch 1

added_tokens.json

@@ -1,50 +1,51 @@
 {
-  "</access_instructions>": 29037,
-  "</accommodation>": 28997,
-  "</activity>": 28999,
-  "</address>": 29023,
-  "</arrival location>": 29001,
-  "</attendee>": 29003,
-  "</cost>": 29005,
-  "</date>": 29007,
-  "</flight_type>": 29041,
-  "</itenerary>": 29009,
-  "</itinerary>": 29019,
-  "</link>": 29031,
-  "</location>": 29011,
-  "</occupancy>": 29035,
-  "</person>": 29043,
-  "</planner>": 29017,
-  "</probability>": 29039,
-  "</seat>": 29021,
-  "</sender>": 29013,
-  "</thing_needed>": 29025,
-  "</time>": 29015,
-  "</travel>": 29027,
-  "</trip>": 29033,
-  "</type>": 29029,
-  "<access_instructions>": 29036,
-  "<accommodation>": 28996,
-  "<activity>": 28998,
-  "<address>": 29022,
-  "<arrival location>": 29000,
-  "<attendee>": 29002,
-  "<cost>": 29004,
-  "<date>": 29006,
-  "<flight_type>": 29040,
-  "<itenerary>": 29008,
-  "<itinerary>": 29018,
-  "<link>": 29030,
-  "<location>": 29010,
-  "<occupancy>": 29034,
-  "<person>": 29042,
-  "<planner>": 29016,
-  "<probability>": 29038,
-  "<seat>": 29020,
-  "<sender>": 29012,
-  "<thing_needed>": 29024,
-  "<time>": 29014,
-  "<travel>": 29026,
-  "<trip>": 29032,
-  "<type>": 29028
+  "</access_instructions>": 128042,
+  "</accommodation>": 128002,
+  "</activity>": 128004,
+  "</address>": 128028,
+  "</arrival location>": 128006,
+  "</attendee>": 128008,
+  "</cost>": 128010,
+  "</date>": 128012,
+  "</flight_type>": 128046,
+  "</itenerary>": 128014,
+  "</itinerary>": 128024,
+  "</link>": 128036,
+  "</location>": 128016,
+  "</occupancy>": 128040,
+  "</person>": 128048,
+  "</planner>": 128022,
+  "</probability>": 128044,
+  "</seat>": 128026,
+  "</sender>": 128018,
+  "</thing_needed>": 128030,
+  "</time>": 128020,
+  "</travel>": 128032,
+  "</trip>": 128038,
+  "</type>": 128034,
+  "<access_instructions>": 128041,
+  "<accommodation>": 128001,
+  "<activity>": 128003,
+  "<address>": 128027,
+  "<arrival location>": 128005,
+  "<attendee>": 128007,
+  "<cost>": 128009,
+  "<date>": 128011,
+  "<flight_type>": 128045,
+  "<itenerary>": 128013,
+  "<itinerary>": 128023,
+  "<link>": 128035,
+  "<location>": 128015,
+  "<occupancy>": 128039,
+  "<person>": 128047,
+  "<planner>": 128021,
+  "<probability>": 128043,
+  "<seat>": 128025,
+  "<sender>": 128017,
+  "<thing_needed>": 128029,
+  "<time>": 128019,
+  "<travel>": 128031,
+  "<trip>": 128037,
+  "<type>": 128033,
+  "[MASK]": 128000
 }

config.json

@@ -1,13 +1,12 @@
 {
-  "_name_or_path": "distilbert-base-uncased",
-  "activation": "gelu",
+  "_name_or_path": "microsoft/deberta-v3-small",
   "architectures": [
-    "DistilBertForSequenceClassification"
+    "DebertaV2ForSequenceClassification"
   ],
-  "attention_dropout": 0.1,
-  "dim": 768,
-  "dropout": 0.1,
-  "hidden_dim": 3072,
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
   "id2label": {
     "0": "accommodation cost",
     "1": "accommodation planner",
@@ -72,6 +71,7 @@
     "60": "itinerary link"
   },
   "initializer_range": 0.02,
+  "intermediate_size": 3072,
   "label2id": {
     "accommodation access_instructions": 41,
     "accommodation address": 36,
@@ -135,17 +135,28 @@
     "trip travel": 29,
     "type probability": 30
   },
+  "layer_norm_eps": 1e-07,
   "max_position_embeddings": 512,
-  "model_type": "distilbert",
-  "n_heads": 12,
-  "n_layers": 6,
+  "max_relative_positions": -1,
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 6,
   "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 768,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
   "problem_type": "multi_label_classification",
-  "qa_dropout": 0.1,
-  "seq_classif_dropout": 0.2,
-  "sinusoidal_pos_embds": false,
-  "tie_weights_": true,
+  "relative_attention": true,
+  "share_att_key": true,
   "torch_dtype": "float32",
   "transformers_version": "4.35.2",
-  "vocab_size": 29044
+  "type_vocab_size": 0,
+  "vocab_size": 128049
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0be413c88627e99625665c44c86daf911488d040fbec6e5d73abc0a6bbe4524d
-size 263473636
+oid sha256:2f412e319c853bd843e88723ce5e86a8882f4c04c19ac683dc128d4ef39720dc
+size 567623380

runs/Feb10_23-15-08_eacf5aebda76/events.out.tfevents.1707607012.eacf5aebda76.364.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75d0b6e81514c60d1dd1ec3bbe964d2b67d891fafd556c23149571fe382a61bc
+size 23926

runs/Feb10_23-38-37_eacf5aebda76/events.out.tfevents.1707608319.eacf5aebda76.364.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e7008d1e290a59c48de727786d80773200e977cfc20439d73142f4c75dea47f
+size 8305

special_tokens_map.json

@@ -1,7 +1,15 @@
 {
+  "bos_token": "[CLS]",
   "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
   "mask_token": "[MASK]",
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
-  "unk_token": "[UNK]"
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
 }

spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
+size 2464616

tokenizer_config.json

@@ -8,31 +8,31 @@
       "single_word": false,
       "special": true
     },
-    "100": {
-      "content": "[UNK]",
+    "1": {
+      "content": "[CLS]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
     },
-    "101": {
-      "content": "[CLS]",
+    "2": {
+      "content": "[SEP]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
     },
-    "102": {
-      "content": "[SEP]",
+    "3": {
+      "content": "[UNK]",
       "lstrip": false,
-      "normalized": false,
+      "normalized": true,
       "rstrip": false,
       "single_word": false,
       "special": true
     },
-    "103": {
+    "128000": {
       "content": "[MASK]",
       "lstrip": false,
       "normalized": false,
@@ -40,7 +40,7 @@
       "single_word": false,
       "special": true
     },
-    "28996": {
+    "128001": {
       "content": "<accommodation>",
       "lstrip": false,
       "normalized": true,
@@ -48,7 +48,7 @@
       "single_word": false,
       "special": false
     },
-    "28997": {
+    "128002": {
       "content": "</accommodation>",
       "lstrip": false,
       "normalized": true,
@@ -56,7 +56,7 @@
       "single_word": false,
       "special": false
     },
-    "28998": {
+    "128003": {
       "content": "<activity>",
       "lstrip": false,
       "normalized": true,
@@ -64,7 +64,7 @@
       "single_word": false,
       "special": false
     },
-    "28999": {
+    "128004": {
       "content": "</activity>",
       "lstrip": false,
       "normalized": true,
@@ -72,7 +72,7 @@
       "single_word": false,
       "special": false
     },
-    "29000": {
+    "128005": {
       "content": "<arrival location>",
       "lstrip": false,
       "normalized": true,
@@ -80,7 +80,7 @@
       "single_word": false,
       "special": false
     },
-    "29001": {
+    "128006": {
       "content": "</arrival location>",
       "lstrip": false,
       "normalized": true,
@@ -88,7 +88,7 @@
       "single_word": false,
       "special": false
     },
-    "29002": {
+    "128007": {
       "content": "<attendee>",
       "lstrip": false,
       "normalized": true,
@@ -96,7 +96,7 @@
       "single_word": false,
       "special": false
     },
-    "29003": {
+    "128008": {
       "content": "</attendee>",
       "lstrip": false,
       "normalized": true,
@@ -104,7 +104,7 @@
       "single_word": false,
       "special": false
     },
-    "29004": {
+    "128009": {
       "content": "<cost>",
       "lstrip": false,
       "normalized": true,
@@ -112,7 +112,7 @@
       "single_word": false,
       "special": false
     },
-    "29005": {
+    "128010": {
       "content": "</cost>",
       "lstrip": false,
       "normalized": true,
@@ -120,7 +120,7 @@
       "single_word": false,
       "special": false
     },
-    "29006": {
+    "128011": {
       "content": "<date>",
       "lstrip": false,
       "normalized": true,
@@ -128,7 +128,7 @@
       "single_word": false,
       "special": false
     },
-    "29007": {
+    "128012": {
       "content": "</date>",
       "lstrip": false,
       "normalized": true,
@@ -136,7 +136,7 @@
       "single_word": false,
       "special": false
     },
-    "29008": {
+    "128013": {
       "content": "<itenerary>",
       "lstrip": false,
       "normalized": true,
@@ -144,7 +144,7 @@
       "single_word": false,
       "special": false
     },
-    "29009": {
+    "128014": {
       "content": "</itenerary>",
       "lstrip": false,
       "normalized": true,
@@ -152,7 +152,7 @@
       "single_word": false,
       "special": false
     },
-    "29010": {
+    "128015": {
       "content": "<location>",
       "lstrip": false,
       "normalized": true,
@@ -160,7 +160,7 @@
       "single_word": false,
       "special": false
     },
-    "29011": {
+    "128016": {
       "content": "</location>",
       "lstrip": false,
       "normalized": true,
@@ -168,7 +168,7 @@
       "single_word": false,
       "special": false
     },
-    "29012": {
+    "128017": {
       "content": "<sender>",
       "lstrip": false,
       "normalized": true,
@@ -176,7 +176,7 @@
       "single_word": false,
       "special": false
     },
-    "29013": {
+    "128018": {
       "content": "</sender>",
       "lstrip": false,
       "normalized": true,
@@ -184,7 +184,7 @@
       "single_word": false,
       "special": false
     },
-    "29014": {
+    "128019": {
       "content": "<time>",
       "lstrip": false,
       "normalized": true,
@@ -192,7 +192,7 @@
       "single_word": false,
       "special": false
     },
-    "29015": {
+    "128020": {
       "content": "</time>",
       "lstrip": false,
       "normalized": true,
@@ -200,7 +200,7 @@
       "single_word": false,
       "special": false
     },
-    "29016": {
+    "128021": {
       "content": "<planner>",
       "lstrip": false,
       "normalized": true,
@@ -208,7 +208,7 @@
       "single_word": false,
       "special": false
     },
-    "29017": {
+    "128022": {
       "content": "</planner>",
       "lstrip": false,
       "normalized": true,
@@ -216,7 +216,7 @@
       "single_word": false,
       "special": false
     },
-    "29018": {
+    "128023": {
       "content": "<itinerary>",
       "lstrip": false,
       "normalized": true,
@@ -224,7 +224,7 @@
       "single_word": false,
       "special": false
     },
-    "29019": {
+    "128024": {
       "content": "</itinerary>",
       "lstrip": false,
       "normalized": true,
@@ -232,7 +232,7 @@
       "single_word": false,
       "special": false
     },
-    "29020": {
+    "128025": {
       "content": "<seat>",
       "lstrip": false,
       "normalized": true,
@@ -240,7 +240,7 @@
       "single_word": false,
       "special": false
     },
-    "29021": {
+    "128026": {
       "content": "</seat>",
       "lstrip": false,
       "normalized": true,
@@ -248,7 +248,7 @@
       "single_word": false,
       "special": false
     },
-    "29022": {
+    "128027": {
       "content": "<address>",
       "lstrip": false,
       "normalized": true,
@@ -256,7 +256,7 @@
       "single_word": false,
       "special": false
     },
-    "29023": {
+    "128028": {
       "content": "</address>",
       "lstrip": false,
       "normalized": true,
@@ -264,7 +264,7 @@
       "single_word": false,
       "special": false
     },
-    "29024": {
+    "128029": {
       "content": "<thing_needed>",
       "lstrip": false,
       "normalized": true,
@@ -272,7 +272,7 @@
       "single_word": false,
       "special": false
     },
-    "29025": {
+    "128030": {
       "content": "</thing_needed>",
       "lstrip": false,
       "normalized": true,
@@ -280,7 +280,7 @@
       "single_word": false,
       "special": false
     },
-    "29026": {
+    "128031": {
       "content": "<travel>",
       "lstrip": false,
       "normalized": true,
@@ -288,7 +288,7 @@
       "single_word": false,
       "special": false
     },
-    "29027": {
+    "128032": {
       "content": "</travel>",
       "lstrip": false,
       "normalized": true,
@@ -296,7 +296,7 @@
       "single_word": false,
       "special": false
     },
-    "29028": {
+    "128033": {
       "content": "<type>",
       "lstrip": false,
       "normalized": true,
@@ -304,7 +304,7 @@
       "single_word": false,
       "special": false
     },
-    "29029": {
+    "128034": {
       "content": "</type>",
       "lstrip": false,
       "normalized": true,
@@ -312,7 +312,7 @@
       "single_word": false,
       "special": false
     },
-    "29030": {
+    "128035": {
       "content": "<link>",
       "lstrip": false,
       "normalized": true,
@@ -320,7 +320,7 @@
       "single_word": false,
       "special": false
     },
-    "29031": {
+    "128036": {
       "content": "</link>",
       "lstrip": false,
       "normalized": true,
@@ -328,7 +328,7 @@
       "single_word": false,
       "special": false
     },
-    "29032": {
+    "128037": {
       "content": "<trip>",
       "lstrip": false,
       "normalized": true,
@@ -336,7 +336,7 @@
       "single_word": false,
       "special": false
     },
-    "29033": {
+    "128038": {
       "content": "</trip>",
       "lstrip": false,
       "normalized": true,
@@ -344,7 +344,7 @@
       "single_word": false,
       "special": false
     },
-    "29034": {
+    "128039": {
       "content": "<occupancy>",
       "lstrip": false,
       "normalized": true,
@@ -352,7 +352,7 @@
       "single_word": false,
       "special": false
     },
-    "29035": {
+    "128040": {
       "content": "</occupancy>",
       "lstrip": false,
       "normalized": true,
@@ -360,7 +360,7 @@
       "single_word": false,
       "special": false
     },
-    "29036": {
+    "128041": {
       "content": "<access_instructions>",
       "lstrip": false,
       "normalized": true,
@@ -368,7 +368,7 @@
       "single_word": false,
       "special": false
     },
-    "29037": {
+    "128042": {
       "content": "</access_instructions>",
       "lstrip": false,
       "normalized": true,
@@ -376,7 +376,7 @@
       "single_word": false,
       "special": false
     },
-    "29038": {
+    "128043": {
       "content": "<probability>",
       "lstrip": false,
       "normalized": true,
@@ -384,7 +384,7 @@
       "single_word": false,
       "special": false
     },
-    "29039": {
+    "128044": {
       "content": "</probability>",
       "lstrip": false,
       "normalized": true,
@@ -392,7 +392,7 @@
       "single_word": false,
       "special": false
     },
-    "29040": {
+    "128045": {
       "content": "<flight_type>",
       "lstrip": false,
       "normalized": true,
@@ -400,7 +400,7 @@
       "single_word": false,
       "special": false
     },
-    "29041": {
+    "128046": {
       "content": "</flight_type>",
       "lstrip": false,
       "normalized": true,
@@ -408,7 +408,7 @@
       "single_word": false,
       "special": false
     },
-    "29042": {
+    "128047": {
       "content": "<person>",
       "lstrip": false,
       "normalized": true,
@@ -416,7 +416,7 @@
       "single_word": false,
       "special": false
     },
-    "29043": {
+    "128048": {
       "content": "</person>",
       "lstrip": false,
       "normalized": true,
@@ -425,15 +425,18 @@
       "special": false
     }
   },
+  "bos_token": "[CLS]",
   "clean_up_tokenization_spaces": true,
   "cls_token": "[CLS]",
   "do_lower_case": false,
+  "eos_token": "[SEP]",
   "mask_token": "[MASK]",
-  "model_max_length": 512,
+  "model_max_length": 1000000000000000019884624838656,
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
-  "strip_accents": null,
-  "tokenize_chinese_chars": true,
-  "tokenizer_class": "DistilBertTokenizer",
-  "unk_token": "[UNK]"
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:912d19e81081dbdc2e3ebc85394bd6b36f8df061d34a6dd6bc8f46787df0bde8
+oid sha256:f5430ad25b5443de74a5526a3b54250516022a4d768261559b70158ecc6c4209
 size 4600