checkpoint step 1000

student/config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "albertklorer/ModernBERT-small",
+  "_name_or_path": "answerdotai/ModernBERT-large",
   "architectures": [
     "ModernBertForMaskedLM"
   ],
@@ -19,10 +19,10 @@
   "global_rope_theta": 160000.0,
   "gradient_checkpointing": false,
   "hidden_activation": "gelu",
-  "hidden_size": 512,
+  "hidden_size": 768,
   "initializer_cutoff_factor": 2.0,
   "initializer_range": 0.02,
-  "intermediate_size": 768,
+  "intermediate_size": 1152,
   "layer_norm_eps": 1e-05,
   "local_attention": 128,
   "local_rope_theta": 10000.0,
@@ -32,8 +32,8 @@
   "model_type": "modernbert",
   "norm_bias": false,
   "norm_eps": 1e-05,
-  "num_attention_heads": 8,
-  "num_hidden_layers": 16,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 11,
   "pad_token_id": 50283,
   "position_embedding_type": "absolute",
   "reference_compile": true,

student/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2e215854332031f7e7407aeb7cf8a278b5f5436914bb472403d0cf48fbee450b
-size 247090040
+oid sha256:323c81f8836cc5080fc51d39dbc062f93323d96411debe31e0a8c57f3effdf9a
+size 377966432

trainer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:52d420a9d1b16301e5abeb5e6ed693bd57a187c35a25c59a9717551ef26b0f0d
-size 357892833
+oid sha256:6e7f4c858751ded2d282cdb55ad5f1984fba0fdbab619ae73bc9089e762ba677
+size 724885121