Update config.json

config.json

@@ -1,56 +1,63 @@
 {
+  "arch": {
     "architectures": [
-        "ScriptableCrammedBERT"
+      "ScriptableCrammedBERT"
     ],
-    "num_transformer_layers": 16,
+    "attention": {
+      "causal_attention": false,
+      "dropout_prob": 0.1,
+      "num_attention_heads": 12,
+      "qkv_bias": false,
+      "rotary_embedding": false,
+      "seq_op_in_fp32": false,
+      "sequence_op": "torch-softmax",
+      "skip_output_projection": false,
+      "type": "self-attention"
+    },
+    "classification_head": {
+      "classifier_dropout": 0.1,
+      "head_dim": 1024,
+      "include_ff_layer": true,
+      "nonlin": "Tanh",
+      "pooler": "zero_index"
+    },
+    "decoder_bias": false,
+    "embedding": {
+      "dropout_prob": 0.1,
+      "embedding_dim": 768,
+      "max_seq_length": 128,
+      "normalization": true,
+      "pad_token_id": 0,
+      "pos_embedding": "scaled-sinusoidal",
+      "stable_low_precision": false,
+      "vocab_size": 32768
+    },
+    "ffn_layer_frequency": 1,
+    "final_norm": true,
+    "hidden_dropout_prob": 0.1,
     "hidden_size": 768,
+    "init": {
+      "std": 0.02,
+      "type": "normal"
+    },
     "intermed_size": 3072,
-    "hidden_dropout_prob": 0.1,
+    "loss": "cross-entropy",
+    "nonlin": "GELUglu",
     "norm": "LayerNorm",
     "norm_eps": 1e-12,
     "norm_scheme": "pre",
-    "nonlin": "GELUglu",
-    "tie_weights": true,
-    "decoder_bias": false,
-    "sparse_prediction": 0.25,
-    "loss": "cross-entropy",
+    "num_labels": null,
+    "num_transformer_layers": 16,
     "objective_layout": "MLM",
-    "embedding": {
-        "vocab_size": null,
-        "pos_embedding": "scaled-sinusoidal",
-        "dropout_prob": 0.1,
-        "pad_token_id": 0,
-        "max_seq_length": 128,
-        "embedding_dim": 768,
-        "normalization": true,
-        "stable_low_precision": false
-    },
-    "attention": {
-        "type": "self-attention",
-        "causal_attention": false,
-        "num_attention_heads": 12,
-        "dropout_prob": 0.1,
-        "skip_output_projection": false,
-        "qkv_bias": false,
-        "rotary_embedding": false,
-        "seq_op_in_fp32": false,
-        "sequence_op": "torch-softmax"
-    },
-    "init": {
-        "type": "normal",
-        "std": 0.02
-    },
-    "ffn_layer_frequency": 1,
     "skip_head_transform": true,
-    "use_bias": false,
-    "final_norm": true,
-    "num_labels": 0,
-    "classification_head": {
-        "pooler": "avg",
-        "include_ff_layer": true,
-        "head_dim": 1024,
-        "nonlin": "Tanh",
-        "classifier_dropout": 0.1
-    },
-    "attn_implementation": null
+    "sparse_prediction": 0.25,
+    "tie_weights": true,
+    "use_bias": false
+  },
+  "architectures": [
+    "ScriptableLMForPreTraining"
+  ],
+  "model_type": "crammedBERT",
+  "torch_dtype": "float32",
+  "transformers_version": "4.29.2"
 }