AdinaM
/

hackathon_datacrowd

Mixture of Experts

Eval Results (legacy)

text-generation-inference

Model card Files Files and versions

AdinaM commited on Apr 19, 2024

Commit

d542c17

·

verified ·

1 Parent(s): c1ab4ad

Update config.json

Files changed (1) hide show

config.json +2 -2

config.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "architectures": [
-    "MixtralForCausalLM"
   ],
   "attention_dropout": 0.0,
   "bos_token_id": 1,
@@ -10,7 +10,7 @@
   "initializer_range": 0.02,
   "intermediate_size": 16384,
   "max_position_embeddings": 65536,
-  "model_type": "mixtral",
   "num_attention_heads": 48,
   "num_experts_per_tok": 2,
   "num_hidden_layers": 56,

 {
   "architectures": [
+    "Hackathon_DataCrowd"
   ],
   "attention_dropout": 0.0,
   "bos_token_id": 1,
   "initializer_range": 0.02,
   "intermediate_size": 16384,
   "max_position_embeddings": 65536,
+  "model_type": "hackathon_datacrowd",
   "num_attention_heads": 48,
   "num_experts_per_tok": 2,
   "num_hidden_layers": 56,