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 sha256:4e0499c421257ff151b632893a1766632a6bc44f26db045d4ae62e5b31fc4143
+size 598439784

runs/Dec29_16-19-07_srv04-ubuntu/events.out.tfevents.1735485567.srv04-ubuntu.27614.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d3f1ab2a0e91253ddb65a7127e12752e262d0fd2e34390da12cb7287a6b98dc
+size 5809

runs/Dec29_16-31-49_srv04-ubuntu/events.out.tfevents.1735486325.srv04-ubuntu.7073.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:773f88908c46c21bded5f158c038d479751979d137386d2b3ceeb001f2022cac
+size 5809

runs/Dec30_12-16-38_srv04-ubuntu/events.out.tfevents.1735557437.srv04-ubuntu.10315.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dabbac2067d8aef952a13b474dc503d376f665a6c7a96a531ac7e3f2e7ba6a09
+size 5809

runs/Dec30_12-33-38_srv04-ubuntu/events.out.tfevents.1735558418.srv04-ubuntu.19779.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa2a9703f3fe01b1593da29945c444932bfcb75aa316210095156d1f18a895af
+size 5809

runs/Dec30_13-32-43_srv04-ubuntu/events.out.tfevents.1735561964.srv04-ubuntu.41375.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:560d84414555472735e071a5138cf49d21b89a30db13e6f36564a75ad825cd91
+size 5807

runs/Dec30_13-37-30_srv04-ubuntu/events.out.tfevents.1735562250.srv04-ubuntu.43412.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24a5c74fb6de9a045bb52b57422b4cb50ed017f1dfff4388ecd63351445be8ab
+size 6964

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b5827d144e6bf8d30702d43ddc00cdf10166764fe607301c548a94259947318
+size 5368