config.json

{
    "architectures": [
        "NemotronForCausalLM"
    ],
    "bos_token_id": 2,
    "eos_token_id": 3,
    "hidden_act": "relu2",
    "hidden_size": 3072,
    "initializer_range": 0.0134,
    "intermediate_size": 9216,
    "kv_channels": 128,
    "max_position_embeddings": 4096,
    "model_type": "nemotron",
    "norm_eps": 1e-05,
    "num_attention_heads": 24,
    "num_hidden_layers": 32,
    "num_key_value_heads": 8,
    "partial_rotary_factor": 0.5,
    "quantization": {
        "group_size": 64,
        "bits": 4,
        "mode": "affine"
    },
    "quantization_config": {
        "group_size": 64,
        "bits": 4,
        "mode": "affine"
    },
    "rope_theta": 10000,
    "tie_word_embeddings": false,
    "torch_dtype": "bfloat16",
    "transformers_version": "4.32.0.dev0",
    "use_cache": true,
    "vocab_size": 256000
}