Training in progress epoch 0

README.md

@@ -14,9 +14,9 @@ probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Train Loss: 2.5466
-- Validation Loss: 2.5109
-- Epoch: 2
+- Train Loss: 3.1628
+- Validation Loss: 2.9340
+- Epoch: 0
 
 ## Model description
 
@@ -42,9 +42,7 @@ The following hyperparameters were used during training:
 
 | Train Loss | Validation Loss | Epoch |
 |:----------:|:---------------:|:-----:|
-| 3.2920     | 3.0536          | 0     |
-| 2.8585     | 2.6748          | 1     |
-| 2.5466     | 2.5109          | 2     |
+| 3.1628     | 2.9340          | 0     |
 
 
 ### Framework versions

config.json

@@ -9,67 +9,67 @@
   "dropout": 0.1,
   "hidden_dim": 3072,
   "id2label": {
-    "0": 0,
-    "1": 19,
-    "2": 20,
-    "3": 7,
-    "4": 8,
-    "5": 25,
-    "6": 26,
-    "7": 23,
-    "8": 24,
-    "9": 27,
-    "10": 21,
-    "11": 1,
-    "12": 2,
-    "13": 13,
-    "14": 14,
-    "15": 28,
-    "16": 11,
-    "17": 15,
-    "18": 16,
-    "19": 12,
-    "20": 3,
-    "21": 4,
-    "22": 5,
-    "23": 22,
-    "24": 9,
-    "25": 17,
-    "26": 18,
-    "27": 6,
-    "28": 10
+    "0": "O",
+    "1": "B-algorithm",
+    "2": "I-algorithm",
+    "3": "B-conference",
+    "4": "I-conference",
+    "5": "B-country",
+    "6": "I-country",
+    "7": "B-field",
+    "8": "I-field",
+    "9": "B-location",
+    "10": "I-location",
+    "11": "B-metrics",
+    "12": "I-metrics",
+    "13": "B-misc",
+    "14": "I-misc",
+    "15": "B-organisation",
+    "16": "I-organisation",
+    "17": "B-person",
+    "18": "I-person",
+    "19": "B-product",
+    "20": "I-product",
+    "21": "B-programlang",
+    "22": "I-programlang",
+    "23": "B-researcher",
+    "24": "I-researcher",
+    "25": "B-task",
+    "26": "I-task",
+    "27": "B-university",
+    "28": "I-university"
   },
   "initializer_range": 0.02,
   "label2id": {
-    "0": 0,
-    "1": 11,
-    "2": 12,
-    "3": 20,
-    "4": 21,
-    "5": 22,
-    "6": 27,
-    "7": 3,
-    "8": 4,
-    "9": 24,
-    "10": 28,
-    "11": 16,
-    "12": 19,
-    "13": 13,
-    "14": 14,
-    "15": 17,
-    "16": 18,
-    "17": 25,
-    "18": 26,
-    "19": 1,
-    "20": 2,
-    "21": 10,
-    "22": 23,
-    "23": 7,
-    "24": 8,
-    "25": 5,
-    "26": 6,
-    "27": 9,
-    "28": 15
+    "B-algorithm": 1,
+    "B-conference": 3,
+    "B-country": 5,
+    "B-field": 7,
+    "B-location": 9,
+    "B-metrics": 11,
+    "B-misc": 13,
+    "B-organisation": 15,
+    "B-person": 17,
+    "B-product": 19,
+    "B-programlang": 21,
+    "B-researcher": 23,
+    "B-task": 25,
+    "B-university": 27,
+    "I-algorithm": 2,
+    "I-conference": 4,
+    "I-country": 6,
+    "I-field": 8,
+    "I-location": 10,
+    "I-metrics": 12,
+    "I-misc": 14,
+    "I-organisation": 16,
+    "I-person": 18,
+    "I-product": 20,
+    "I-programlang": 22,
+    "I-researcher": 24,
+    "I-task": 26,
+    "I-university": 28,
+    "O": 0
   },
   "max_position_embeddings": 512,
   "model_type": "distilbert",

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12e9ab80e9ae58775098043b004549492022429f3d33146fcb6ced6159423ed0
+oid sha256:a16b0eac3d3c5f4ef0c67633f7f91a2f6a4c26b6490cdb5b562c7d8bb6c817e6
 size 265666912