vetta-granite-2b-v3 / config.json
asifdotpy's picture
Upload merged Vetta Granite fine-tuned model v3 (16-bit)
ea5c81d verified
{
"architectures": [
"GraniteForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.1,
"attention_multiplier": 0.015625,
"bos_token_id": 0,
"dtype": "float16",
"embedding_multiplier": 12.0,
"eos_token_id": 0,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 8192,
"logits_scaling": 8.0,
"max_position_embeddings": 4096,
"mlp_bias": false,
"model_type": "granite",
"num_attention_heads": 32,
"num_hidden_layers": 40,
"num_key_value_heads": 8,
"pad_token_id": 0,
"residual_multiplier": 0.22,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 10000,
"tie_word_embeddings": true,
"transformers_version": "4.57.2",
"unsloth_version": "2025.11.4",
"use_cache": true,
"vocab_size": 49155
}