feat-add-configs

config.json

@@ -0,0 +1,40 @@
+{
+  "_name_or_path": "jinaai/jina-bert-flash-implementation",
+  "auto_map": {
+    "AutoConfig": "jinaai/jina-bert-flash-implementation--configuration_bert.JinaBertConfig",
+    "AutoModel": "jinaai/jina-bert-flash-implementation--modeling_bert.BertModel",
+    "AutoModelForPreTraining": "jinaai/jina-bert-flash-implementation--modeling_bert.BertForPreTraining",
+    "AutoModelForMaskedLM": "jinaai/jina-bert-flash-implementation--modeling_bert.BertForPreTraining"
+  },
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "dense_seq_output": false,
+  "emb_pooler": null,
+  "fused_bias_fc": false,
+  "fused_dropout_add_ln": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "last_layer_subset": false,
+  "layer_norm_eps": 1e-12,
+  "mlp_checkpoint_lvl": 0,
+  "mlp_type": "glu",
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "num_loras": 5,
+  "pad_token_id": 0,
+  "pad_vocab_size_multiple": 1,
+  "torch_dtype": "float16",
+  "transformers_version": "4.39.3",
+  "type_vocab_size": 2,
+  "use_flash_attn": true,
+  "use_qk_norm": false,
+  "vocab_size": 30528,
+  "window_size": [
+    -1,
+    -1
+  ]
+}