Upload config.json
Browse files
onnx/directml/phi-3-small-128k-instruct-int4/config.json
ADDED
|
@@ -0,0 +1,187 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Phi-3-small-128k-instruct",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Phi3SmallForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout_prob": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "configuration_phi3_small.Phi3SmallConfig",
|
| 9 |
+
"AutoModelForCausalLM": "modeling_phi3_small.Phi3SmallForCausalLM",
|
| 10 |
+
"AutoTokenizer": "tokenization_phi3_small.Phi3SmallTokenizer"
|
| 11 |
+
},
|
| 12 |
+
"blocksparse_block_size": 64,
|
| 13 |
+
"blocksparse_homo_head_pattern": false,
|
| 14 |
+
"blocksparse_num_local_blocks": 16,
|
| 15 |
+
"blocksparse_triton_kernel_block_size": 64,
|
| 16 |
+
"blocksparse_vert_stride": 8,
|
| 17 |
+
"bos_token_id": 100257,
|
| 18 |
+
"dense_attention_every_n_layers": 2,
|
| 19 |
+
"embedding_dropout_prob": 0.1,
|
| 20 |
+
"eos_token_id": 100257,
|
| 21 |
+
"ff_dim_multiplier": null,
|
| 22 |
+
"ff_intermediate_size": 14336,
|
| 23 |
+
"ffn_dropout_prob": 0.1,
|
| 24 |
+
"gegelu_limit": 20.0,
|
| 25 |
+
"gegelu_pad_to_256": true,
|
| 26 |
+
"hidden_act": "gegelu",
|
| 27 |
+
"hidden_size": 4096,
|
| 28 |
+
"initializer_range": 0.02,
|
| 29 |
+
"layer_norm_epsilon": 1e-05,
|
| 30 |
+
"max_position_embeddings": 131072,
|
| 31 |
+
"model_type": "phi3small",
|
| 32 |
+
"mup_attn_multiplier": 1.0,
|
| 33 |
+
"mup_embedding_multiplier": 10.0,
|
| 34 |
+
"mup_use_scaling": true,
|
| 35 |
+
"mup_width_multiplier": 8.0,
|
| 36 |
+
"num_attention_heads": 32,
|
| 37 |
+
"num_hidden_layers": 32,
|
| 38 |
+
"num_key_value_heads": 8,
|
| 39 |
+
"original_max_position_embeddings": 8192,
|
| 40 |
+
"pad_sequence_to_multiple_of_64": true,
|
| 41 |
+
"reorder_and_upcast_attn": false,
|
| 42 |
+
"rope_embedding_base": 1000000,
|
| 43 |
+
"rope_position_scale": 1.0,
|
| 44 |
+
"rope_scaling": {
|
| 45 |
+
"long_factor": [
|
| 46 |
+
1.0,
|
| 47 |
+
1.01,
|
| 48 |
+
1.01,
|
| 49 |
+
1.02,
|
| 50 |
+
1.04,
|
| 51 |
+
1.04,
|
| 52 |
+
1.04,
|
| 53 |
+
1.05,
|
| 54 |
+
1.05,
|
| 55 |
+
1.06,
|
| 56 |
+
1.07,
|
| 57 |
+
1.08,
|
| 58 |
+
1.08,
|
| 59 |
+
1.08,
|
| 60 |
+
1.08,
|
| 61 |
+
1.08,
|
| 62 |
+
1.08,
|
| 63 |
+
1.08,
|
| 64 |
+
1.09,
|
| 65 |
+
1.09,
|
| 66 |
+
1.2,
|
| 67 |
+
2.31,
|
| 68 |
+
3.76,
|
| 69 |
+
9.38,
|
| 70 |
+
10.1,
|
| 71 |
+
10.8,
|
| 72 |
+
18.1,
|
| 73 |
+
25.2,
|
| 74 |
+
25.3,
|
| 75 |
+
26.1,
|
| 76 |
+
26.6,
|
| 77 |
+
30.2,
|
| 78 |
+
33.0,
|
| 79 |
+
41.5,
|
| 80 |
+
44.4,
|
| 81 |
+
44.8,
|
| 82 |
+
50.2,
|
| 83 |
+
51.9,
|
| 84 |
+
59.3,
|
| 85 |
+
62.7,
|
| 86 |
+
66.1,
|
| 87 |
+
66.3,
|
| 88 |
+
85.8,
|
| 89 |
+
89.3,
|
| 90 |
+
90.0,
|
| 91 |
+
99.9,
|
| 92 |
+
107.0,
|
| 93 |
+
110.0,
|
| 94 |
+
111.0,
|
| 95 |
+
117.0,
|
| 96 |
+
118.0,
|
| 97 |
+
121.0,
|
| 98 |
+
122.0,
|
| 99 |
+
127.0,
|
| 100 |
+
127.0,
|
| 101 |
+
128.0,
|
| 102 |
+
128.0,
|
| 103 |
+
128.0,
|
| 104 |
+
128.0,
|
| 105 |
+
128.0,
|
| 106 |
+
128.0,
|
| 107 |
+
129.0,
|
| 108 |
+
129.0,
|
| 109 |
+
129.0
|
| 110 |
+
],
|
| 111 |
+
"long_mscale": 1.1902380714238083,
|
| 112 |
+
"original_max_position_embeddings": 8192,
|
| 113 |
+
"short_factor": [
|
| 114 |
+
1.02,
|
| 115 |
+
1.02,
|
| 116 |
+
1.05,
|
| 117 |
+
1.05,
|
| 118 |
+
1.06,
|
| 119 |
+
1.08,
|
| 120 |
+
1.08,
|
| 121 |
+
1.08,
|
| 122 |
+
1.08,
|
| 123 |
+
1.12,
|
| 124 |
+
1.1800000000000002,
|
| 125 |
+
1.1900000000000002,
|
| 126 |
+
1.1900000000000002,
|
| 127 |
+
1.2100000000000002,
|
| 128 |
+
1.2300000000000002,
|
| 129 |
+
1.2400000000000002,
|
| 130 |
+
1.2400000000000002,
|
| 131 |
+
1.2500000000000002,
|
| 132 |
+
1.3000000000000003,
|
| 133 |
+
1.3100000000000003,
|
| 134 |
+
1.4600000000000004,
|
| 135 |
+
1.5100000000000005,
|
| 136 |
+
1.7000000000000006,
|
| 137 |
+
1.9300000000000008,
|
| 138 |
+
2.080000000000001,
|
| 139 |
+
2.4399999999999933,
|
| 140 |
+
3.2199999999999767,
|
| 141 |
+
3.4499999999999718,
|
| 142 |
+
3.579999999999969,
|
| 143 |
+
4.669999999999946,
|
| 144 |
+
4.779999999999943,
|
| 145 |
+
5.999999999999917,
|
| 146 |
+
6.009999999999917,
|
| 147 |
+
6.4199999999999084,
|
| 148 |
+
6.619999999999904,
|
| 149 |
+
7.189999999999892,
|
| 150 |
+
7.3099999999998895,
|
| 151 |
+
7.339999999999889,
|
| 152 |
+
7.479999999999886,
|
| 153 |
+
9.749999999999837,
|
| 154 |
+
10.919999999999812,
|
| 155 |
+
11.219999999999805,
|
| 156 |
+
11.749999999999794,
|
| 157 |
+
11.979999999999789,
|
| 158 |
+
13.239999999999762,
|
| 159 |
+
13.579999999999755,
|
| 160 |
+
13.669999999999753,
|
| 161 |
+
13.82999999999975,
|
| 162 |
+
14.009999999999746,
|
| 163 |
+
14.679999999999731,
|
| 164 |
+
14.889999999999727,
|
| 165 |
+
15.769999999999708,
|
| 166 |
+
15.769999999999708,
|
| 167 |
+
15.819999999999707,
|
| 168 |
+
15.839999999999707,
|
| 169 |
+
15.919999999999705,
|
| 170 |
+
16.029999999999703,
|
| 171 |
+
16.12999999999972,
|
| 172 |
+
16.44999999999977,
|
| 173 |
+
16.44999999999977,
|
| 174 |
+
16.77999999999982,
|
| 175 |
+
16.83999999999983,
|
| 176 |
+
16.83999999999983,
|
| 177 |
+
16.889999999999837
|
| 178 |
+
],
|
| 179 |
+
"short_mscale": 1.0,
|
| 180 |
+
"type": "su"
|
| 181 |
+
},
|
| 182 |
+
"torch_dtype": "bfloat16",
|
| 183 |
+
"transformers_version": "4.38.1",
|
| 184 |
+
"use_cache": true,
|
| 185 |
+
"attention_bias": false,
|
| 186 |
+
"vocab_size": 100352
|
| 187 |
+
}
|