KafkaLM-15B / config.json
doubledsbv's picture
Upload MistralForCausalLM
ca524d5 verified
{
"architectures": [
"MistralForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 1,
"eos_token_id": 2,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 25600,
"max_position_embeddings": 32768,
"model_type": "mistral",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pad_token_id": 11,
"rms_norm_eps": 1e-05,
"rope_theta": 100000000.0,
"sliding_window": null,
"sparsity": [
{
"layer_idx": 0,
"layer_type": "self_attn",
"param_name": "model.layers.0.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 1,
"layer_type": "self_attn",
"param_name": "model.layers.1.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 1,
"layer_type": "self_attn",
"param_name": "model.layers.1.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 2,
"layer_type": "self_attn",
"param_name": "model.layers.2.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 2,
"layer_type": "self_attn",
"param_name": "model.layers.2.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 3,
"layer_type": "self_attn",
"param_name": "model.layers.3.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 4,
"layer_type": "self_attn",
"param_name": "model.layers.4.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 4,
"layer_type": "self_attn",
"param_name": "model.layers.4.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 5,
"layer_type": "self_attn",
"param_name": "model.layers.5.self_attn.k_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 5,
"layer_type": "self_attn",
"param_name": "model.layers.5.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 7,
"layer_type": "self_attn",
"param_name": "model.layers.7.self_attn.k_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 7,
"layer_type": "self_attn",
"param_name": "model.layers.7.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 8,
"layer_type": "self_attn",
"param_name": "model.layers.8.self_attn.k_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 9,
"layer_type": "self_attn",
"param_name": "model.layers.9.self_attn.k_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 12,
"layer_type": "self_attn",
"param_name": "model.layers.12.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 15,
"layer_type": "self_attn",
"param_name": "model.layers.15.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 15,
"layer_type": "self_attn",
"param_name": "model.layers.15.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 16,
"layer_type": "self_attn",
"param_name": "model.layers.16.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 17,
"layer_type": "self_attn",
"param_name": "model.layers.17.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 18,
"layer_type": "self_attn",
"param_name": "model.layers.18.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 19,
"layer_type": "self_attn",
"param_name": "model.layers.19.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 19,
"layer_type": "self_attn",
"param_name": "model.layers.19.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 20,
"layer_type": "self_attn",
"param_name": "model.layers.20.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 20,
"layer_type": "self_attn",
"param_name": "model.layers.20.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 24,
"layer_type": "self_attn",
"param_name": "model.layers.24.self_attn.k_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 24,
"layer_type": "self_attn",
"param_name": "model.layers.24.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 24,
"layer_type": "self_attn",
"param_name": "model.layers.24.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 25,
"layer_type": "self_attn",
"param_name": "model.layers.25.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 25,
"layer_type": "self_attn",
"param_name": "model.layers.25.self_attn.k_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 25,
"layer_type": "self_attn",
"param_name": "model.layers.25.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 25,
"layer_type": "self_attn",
"param_name": "model.layers.25.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 26,
"layer_type": "self_attn",
"param_name": "model.layers.26.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 26,
"layer_type": "self_attn",
"param_name": "model.layers.26.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 27,
"layer_type": "self_attn",
"param_name": "model.layers.27.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 27,
"layer_type": "mlp",
"param_name": "model.layers.27.mlp.down_proj.weight",
"sparsity": 50.0,
"total": 131072000,
"zeroed": 65536000
},
{
"layer_idx": 28,
"layer_type": "self_attn",
"param_name": "model.layers.28.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 30,
"layer_type": "self_attn",
"param_name": "model.layers.30.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 31,
"layer_type": "self_attn",
"param_name": "model.layers.31.self_attn.q_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
},
{
"layer_idx": 31,
"layer_type": "self_attn",
"param_name": "model.layers.31.self_attn.v_proj.weight",
"sparsity": 50.0,
"total": 5242880,
"zeroed": 2621440
},
{
"layer_idx": 31,
"layer_type": "self_attn",
"param_name": "model.layers.31.self_attn.o_proj.weight",
"sparsity": 50.0,
"total": 20971520,
"zeroed": 10485760
}
],
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.51.1",
"unsloth_fixed": true,
"use_cache": true,
"vocab_size": 131072
}