tolga-ozturk
/

mGPT-nsp

next-sentence-prediction

text-generation-inference

Inference Endpoints

Model card Files Files and versions Community

tolga-ozturk commited on Jul 24, 2023

Commit

787baf9

·

1 Parent(s): 1669290

Update config.json

Files changed (1) hide show

config.json +3 -2

config.json CHANGED Viewed

@@ -2,7 +2,7 @@
   "_name_or_path": "THUMT/mGPT",
   "activation_function": "gelu_new",
   "architectures": [
-    "GPT2LMHeadModel"
   ],
   "attn_pdrop": 0.1,
   "bos_token_id": 1,
@@ -16,7 +16,7 @@
   "n_embd": 1024,
   "n_head": 16,
   "n_inner": 4096,
-  "n_layer": 24,
   "n_positions": 1024,
   "reorder_and_upcast_attn": false,
   "resid_pdrop": 0.1,
@@ -27,6 +27,7 @@
   "summary_proj_to_labels": true,
   "summary_type": "cls_index",
   "summary_use_proj": true,
   "transformers_version": "4.18.0",
   "use_cache": true,
   "vocab_size": 250100

   "_name_or_path": "THUMT/mGPT",
   "activation_function": "gelu_new",
   "architectures": [
+    "ModelNSP"
   ],
   "attn_pdrop": 0.1,
   "bos_token_id": 1,
   "n_embd": 1024,
   "n_head": 16,
   "n_inner": 4096,
+  "n_layer": 27,
   "n_positions": 1024,
   "reorder_and_upcast_attn": false,
   "resid_pdrop": 0.1,
   "summary_proj_to_labels": true,
   "summary_type": "cls_index",
   "summary_use_proj": true,
+  "torch_dtype": "float32",
   "transformers_version": "4.18.0",
   "use_cache": true,
   "vocab_size": 250100