config.json

{
    "architectures": [
        "Cohere2ForCausalLM"
    ],
    "attention_bias": false,
    "attention_dropout": 0.0,
    "bos_token_id": 5,
    "cache_implementation": "hybrid",
    "eos_token_id": 255001,
    "head_dim": 128,
    "hidden_act": "silu",
    "hidden_size": 12288,
    "initializer_range": 0.02,
    "intermediate_size": 36864,
    "layer_norm_eps": 1e-05,
    "logit_scale": 0.25,
    "max_position_embeddings": 131072,
    "model_type": "cohere2",
    "num_attention_heads": 96,
    "num_hidden_layers": 64,
    "num_key_value_heads": 8,
    "order_of_interleaved_layers": "local_attn_first",
    "pad_token_id": 0,
    "position_embedding_type": "rope_gptj",
    "quantization": {
        "group_size": 64,
        "bits": 3
    },
    "quantization_config": {
        "group_size": 64,
        "bits": 3
    },
    "rope_scaling": null,
    "rope_theta": 50000,
    "rotary_pct": 1.0,
    "sliding_window": 4096,
    "sliding_window_pattern": 4,
    "torch_dtype": "bfloat16",
    "transformers_version": "4.48.0.dev0",
    "use_cache": true,
    "use_embedding_sharing": true,
    "use_gated_activation": true,
    "use_parallel_block": true,
    "use_parallel_embedding": true,
    "vocab_size": 256000
}