winglian
/

llama-3-8b-256k-PoSE

Text Generation

text-generation-inference

Inference Endpoints

Model card Files Files and versions Community

winglian commited on Apr 26

Commit

26ce3fa

•

1 Parent(s): d2591cd

Update config.json

Files changed (1) hide show

config.json +3 -6

config.json CHANGED Viewed

@@ -11,18 +11,15 @@
   "hidden_size": 4096,
   "initializer_range": 0.02,
   "intermediate_size": 14336,
-  "max_position_embeddings": 1048576,
   "model_type": "llama",
   "num_attention_heads": 32,
   "num_hidden_layers": 32,
   "num_key_value_heads": 8,
   "pretraining_tp": 1,
   "rms_norm_eps": 1e-05,
-  "rope_scaling": {
-    "factor": 16.0,
-    "type": "dynamic"
-  },
-  "rope_theta": 500000.0,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.40.0.dev0",

   "hidden_size": 4096,
   "initializer_range": 0.02,
   "intermediate_size": 14336,
+  "max_position_embeddings": 524288,
   "model_type": "llama",
   "num_attention_heads": 32,
   "num_hidden_layers": 32,
   "num_key_value_heads": 8,
   "pretraining_tp": 1,
   "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 2000000.0,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.40.0.dev0",