Upload NemotronForCausalLM

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/ckpts/nemotron/nemotron3-8b-hf",
+  "_name_or_path": "/ckpts/nemotron/nemotron3-8b-hf-new",
   "architectures": [
     "NemotronForCausalLM"
   ],
@@ -7,6 +7,7 @@
   "attention_dropout": 0.0,
   "bos_token_id": 2,
   "eos_token_id": 3,
+  "head_dim": 128,
   "hidden_act": "relu2",
   "hidden_size": 4096,
   "initializer_range": 0.01,
@@ -18,13 +19,11 @@
   "num_attention_heads": 32,
   "num_hidden_layers": 32,
   "num_key_value_heads": 32,
-  "rope_percent": 0.5,
-  "rope_percentage": 0.5,
-  "rope_scaling": null,
+  "partial_rotary_factor": 0.5,
   "rope_theta": 10000,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.42.0.dev0",
+  "transformers_version": "4.44.0",
   "use_cache": true,
   "vocab_size": 256000
 }

generation_config.json

@@ -2,5 +2,5 @@
   "_from_model_config": true,
   "bos_token_id": 2,
   "eos_token_id": 3,
-  "transformers_version": "4.42.0.dev0"
+  "transformers_version": "4.44.0"
 }