Training in progress, epoch 1

config.json

@@ -0,0 +1,55 @@
+{
+  "_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": "small_llm",
+    "1": "large_llm"
+  },
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "label2id": {
+    "large_llm": "1",
+    "small_llm": "0"
+  },
+  "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": true,
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.48.0.dev0",
+  "vocab_size": 50368
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bb97b3a80466213cf0930f49cd9bc713a69ac1460722ed69afa91a24228308a
+size 598439784

runs/Jan05_06-27-14_df92ac5f6441/events.out.tfevents.1736058438.df92ac5f6441.4279.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e0f5a6c437568b8a955e30fb0da80a45fedc8c782c7254643da8c70bad997f9
+size 11520

runs/Jan05_06-29-21_df92ac5f6441/events.out.tfevents.1736058566.df92ac5f6441.4279.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bbe706c694c97cd07efd8cedfe9cbd462e42188854f5affedf9ce5e107af282
+size 5804

runs/Jan05_06-32-55_df92ac5f6441/events.out.tfevents.1736058779.df92ac5f6441.4279.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90bb0c6bad1f8533809ac2ef6ec62255c92a709a728dfa7cfb066a0078e2595d
+size 5804

runs/Jan05_06-33-27_df92ac5f6441/events.out.tfevents.1736058810.df92ac5f6441.4279.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73e5e435fb687162d54619cc9abc08ace1f8c9ef1314e78b42f1d130111d07e4
+size 5809

runs/Jan05_06-36-15_df92ac5f6441/events.out.tfevents.1736058978.df92ac5f6441.4279.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96726a89ec58990c90483085795f5834d820abb77df3af57773f693bc2ffe54e
+size 4184

runs/Jan05_06-37-41_df92ac5f6441/events.out.tfevents.1736059068.df92ac5f6441.7347.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9ea6242ce7bf5df701edf7be443c3783363999ded41023d14096be1c403b2dd
+size 5804

runs/Jan05_06-40-24_df92ac5f6441/events.out.tfevents.1736059228.df92ac5f6441.7347.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:403a7b733714112cf3427dbb1ae9e12bd861cedc16dfecb833049f42a55d62d9
+size 8016

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbeb7a73abcb55ee254d23c3b3c2d1fad18b76d265505d0a85db33645323249a
+size 5368