Merge pull request #356 from exo-explore/nemotron

add nemotron-70b and nemotron-70b-bf16

exo/models.py

@@ -59,4 +59,11 @@ model_base_shards = {
   "qwen-2.5-math-72b": {
     "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Math-72B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=80),
   },
+  ### nemotron
+  "nemotron-70b": {
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/nvidia_Llama-3.1-Nemotron-70B-Instruct-HF_4bit", start_layer=0, end_layer=0, n_layers=80),
+  },
+  "nemotron-70b-bf16": {
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Llama-3.1-Nemotron-70B-Instruct-HF-bf16", start_layer=0, end_layer=0, n_layers=80),
+  },
 }