Feat/add nemotron nano v3 tests (#33345)
This commit is contained in:
@@ -355,5 +355,22 @@
|
||||
"is_deepseek_mla": true,
|
||||
"is_multimodal_model": false,
|
||||
"dtype": "torch.float32"
|
||||
},
|
||||
"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16": {
|
||||
"architectures": [
|
||||
"NemotronHForCausalLM"
|
||||
],
|
||||
"model_type": "nemotron_h",
|
||||
"text_model_type": "nemotron_h",
|
||||
"hidden_size": 2688,
|
||||
"total_num_hidden_layers": 52,
|
||||
"total_num_attention_heads": 32,
|
||||
"head_size": 128,
|
||||
"vocab_size": 131072,
|
||||
"total_num_kv_heads": 2,
|
||||
"num_experts": 128,
|
||||
"is_deepseek_mla": false,
|
||||
"is_multimodal_model": false,
|
||||
"dtype": "torch.bfloat16"
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user