Training in progress, epoch 1

config.json

@@ -0,0 +1,82 @@
+{
+  "_name_or_path": "answerdotai/ModernBERT-base",
+  "architectures": [
+    "ModernBertForSequenceClassification"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "id2label": {
+    "0": "POLITICS",
+    "1": "WELLNESS",
+    "2": "ENTERTAINMENT",
+    "3": "TRAVEL",
+    "4": "STYLE & BEAUTY",
+    "5": "PARENTING",
+    "6": "HEALTHY LIVING",
+    "7": "QUEER VOICES",
+    "8": "FOOD & DRINK",
+    "9": "BUSINESS",
+    "10": "COMEDY",
+    "11": "SPORTS",
+    "12": "BLACK VOICES",
+    "13": "HOME & LIVING",
+    "14": "PARENTS"
+  },
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "label2id": {
+    "BLACK VOICES": 12,
+    "BUSINESS": 9,
+    "COMEDY": 10,
+    "ENTERTAINMENT": 2,
+    "FOOD & DRINK": 8,
+    "HEALTHY LIVING": 6,
+    "HOME & LIVING": 13,
+    "PARENTING": 5,
+    "PARENTS": 14,
+    "POLITICS": 0,
+    "QUEER VOICES": 7,
+    "SPORTS": 11,
+    "STYLE & BEAUTY": 4,
+    "TRAVEL": 3,
+    "WELLNESS": 1
+  },
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "reference_compile": false,
+  "repad_logits_with_grad": false,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.49.0.dev0",
+  "vocab_size": 50368
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30436f6e6464e8a76632e0cd14fa44972cb54fbb823ae7a4f24603129f9586f4
+size 598479780

runs/Jan19_12-41-07_06203a5cda3b/events.out.tfevents.1737290468.06203a5cda3b.18.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d86d4f2de380264fc9f3fe2961ff841c3f9f8d4cf963abf940484026cce6e2bb
+size 10163

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4283086b0dd82c6464d1f5236fa60bcdcdcf149d17a84cafca2b32d0d203fac
+size 5432