Upload 5 files
Browse files- config.json +31 -0
- model.safetensors +3 -0
- scales.json +278 -0
- tokenizer.json +0 -0
- train_stats.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "tiny-gpt",
|
| 3 |
+
"vocab_size": 1920,
|
| 4 |
+
"n_layers": 6,
|
| 5 |
+
"n_heads": 6,
|
| 6 |
+
"d_model": 240,
|
| 7 |
+
"d_ff": 960,
|
| 8 |
+
"n_ctx": 64,
|
| 9 |
+
"dropout": 0.0,
|
| 10 |
+
"tie_word_embeddings": true,
|
| 11 |
+
"special_tokens": {
|
| 12 |
+
"pad_token": "<PAD>",
|
| 13 |
+
"bos_token": "<BOS>",
|
| 14 |
+
"eos_token": "<EOS>",
|
| 15 |
+
"sep_token": "<SEP>",
|
| 16 |
+
"unk_token": "<UNK>",
|
| 17 |
+
"ids": {
|
| 18 |
+
"pad": 0,
|
| 19 |
+
"bos": 1,
|
| 20 |
+
"eos": 2,
|
| 21 |
+
"sep": 3,
|
| 22 |
+
"unk": 4
|
| 23 |
+
}
|
| 24 |
+
},
|
| 25 |
+
"quantization": {
|
| 26 |
+
"default": "int8",
|
| 27 |
+
"embedding_bits": 18,
|
| 28 |
+
"layernorm_bits": 24,
|
| 29 |
+
"format": "safetensors+scales"
|
| 30 |
+
}
|
| 31 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ac71d1ac78ca8be0cbbd6e2783d0a781173c16c38502b7be8c8d65dae12934c
|
| 3 |
+
size 6553752
|
scales.json
ADDED
|
@@ -0,0 +1,278 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"tok_emb.weight": {
|
| 3 |
+
"scale": 3.5722721575251916e-05,
|
| 4 |
+
"nbits": 18,
|
| 5 |
+
"dtype": "int32",
|
| 6 |
+
"emulated": true
|
| 7 |
+
},
|
| 8 |
+
"pos_emb.weight": {
|
| 9 |
+
"scale": 3.314935944296658e-05,
|
| 10 |
+
"nbits": 18,
|
| 11 |
+
"dtype": "int32",
|
| 12 |
+
"emulated": true
|
| 13 |
+
},
|
| 14 |
+
"blocks.0.norm1.weight": {
|
| 15 |
+
"scale": 1.3538360433878532e-07,
|
| 16 |
+
"nbits": 24,
|
| 17 |
+
"dtype": "int32",
|
| 18 |
+
"emulated": true
|
| 19 |
+
},
|
| 20 |
+
"blocks.0.attn.mask": {
|
| 21 |
+
"scale": 0.007874015826771653,
|
| 22 |
+
"nbits": 8,
|
| 23 |
+
"dtype": "int8",
|
| 24 |
+
"emulated": false
|
| 25 |
+
},
|
| 26 |
+
"blocks.0.attn.W_qkv.weight": {
|
| 27 |
+
"scale": 0.0013009633219017568,
|
| 28 |
+
"nbits": 8,
|
| 29 |
+
"dtype": "int8",
|
| 30 |
+
"emulated": false
|
| 31 |
+
},
|
| 32 |
+
"blocks.0.attn.W_o.weight": {
|
| 33 |
+
"scale": 0.0009806638006958249,
|
| 34 |
+
"nbits": 8,
|
| 35 |
+
"dtype": "int8",
|
| 36 |
+
"emulated": false
|
| 37 |
+
},
|
| 38 |
+
"blocks.0.norm2.weight": {
|
| 39 |
+
"scale": 1.3978162281289483e-07,
|
| 40 |
+
"nbits": 24,
|
| 41 |
+
"dtype": "int32",
|
| 42 |
+
"emulated": true
|
| 43 |
+
},
|
| 44 |
+
"blocks.0.ff.fc1.weight": {
|
| 45 |
+
"scale": 0.0012522846309088159,
|
| 46 |
+
"nbits": 8,
|
| 47 |
+
"dtype": "int8",
|
| 48 |
+
"emulated": false
|
| 49 |
+
},
|
| 50 |
+
"blocks.0.ff.fc2.weight": {
|
| 51 |
+
"scale": 0.0008591893586456118,
|
| 52 |
+
"nbits": 8,
|
| 53 |
+
"dtype": "int8",
|
| 54 |
+
"emulated": false
|
| 55 |
+
},
|
| 56 |
+
"blocks.1.norm1.weight": {
|
| 57 |
+
"scale": 1.3831941095351961e-07,
|
| 58 |
+
"nbits": 24,
|
| 59 |
+
"dtype": "int32",
|
| 60 |
+
"emulated": true
|
| 61 |
+
},
|
| 62 |
+
"blocks.1.attn.mask": {
|
| 63 |
+
"scale": 0.007874015826771653,
|
| 64 |
+
"nbits": 8,
|
| 65 |
+
"dtype": "int8",
|
| 66 |
+
"emulated": false
|
| 67 |
+
},
|
| 68 |
+
"blocks.1.attn.W_qkv.weight": {
|
| 69 |
+
"scale": 0.001282494329923795,
|
| 70 |
+
"nbits": 8,
|
| 71 |
+
"dtype": "int8",
|
| 72 |
+
"emulated": false
|
| 73 |
+
},
|
| 74 |
+
"blocks.1.attn.W_o.weight": {
|
| 75 |
+
"scale": 0.0011245226614926556,
|
| 76 |
+
"nbits": 8,
|
| 77 |
+
"dtype": "int8",
|
| 78 |
+
"emulated": false
|
| 79 |
+
},
|
| 80 |
+
"blocks.1.norm2.weight": {
|
| 81 |
+
"scale": 1.4735743036467565e-07,
|
| 82 |
+
"nbits": 24,
|
| 83 |
+
"dtype": "int32",
|
| 84 |
+
"emulated": true
|
| 85 |
+
},
|
| 86 |
+
"blocks.1.ff.fc1.weight": {
|
| 87 |
+
"scale": 0.001335447659535746,
|
| 88 |
+
"nbits": 8,
|
| 89 |
+
"dtype": "int8",
|
| 90 |
+
"emulated": false
|
| 91 |
+
},
|
| 92 |
+
"blocks.1.ff.fc2.weight": {
|
| 93 |
+
"scale": 0.0009407425338697058,
|
| 94 |
+
"nbits": 8,
|
| 95 |
+
"dtype": "int8",
|
| 96 |
+
"emulated": false
|
| 97 |
+
},
|
| 98 |
+
"blocks.2.norm1.weight": {
|
| 99 |
+
"scale": 1.4124889373715176e-07,
|
| 100 |
+
"nbits": 24,
|
| 101 |
+
"dtype": "int32",
|
| 102 |
+
"emulated": true
|
| 103 |
+
},
|
| 104 |
+
"blocks.2.attn.mask": {
|
| 105 |
+
"scale": 0.007874015826771653,
|
| 106 |
+
"nbits": 8,
|
| 107 |
+
"dtype": "int8",
|
| 108 |
+
"emulated": false
|
| 109 |
+
},
|
| 110 |
+
"blocks.2.attn.W_qkv.weight": {
|
| 111 |
+
"scale": 0.001387487442367734,
|
| 112 |
+
"nbits": 8,
|
| 113 |
+
"dtype": "int8",
|
| 114 |
+
"emulated": false
|
| 115 |
+
},
|
| 116 |
+
"blocks.2.attn.W_o.weight": {
|
| 117 |
+
"scale": 0.0012036952295510599,
|
| 118 |
+
"nbits": 8,
|
| 119 |
+
"dtype": "int8",
|
| 120 |
+
"emulated": false
|
| 121 |
+
},
|
| 122 |
+
"blocks.2.norm2.weight": {
|
| 123 |
+
"scale": 1.4793962073005056e-07,
|
| 124 |
+
"nbits": 24,
|
| 125 |
+
"dtype": "int32",
|
| 126 |
+
"emulated": true
|
| 127 |
+
},
|
| 128 |
+
"blocks.2.ff.fc1.weight": {
|
| 129 |
+
"scale": 0.0015438962488531877,
|
| 130 |
+
"nbits": 8,
|
| 131 |
+
"dtype": "int8",
|
| 132 |
+
"emulated": false
|
| 133 |
+
},
|
| 134 |
+
"blocks.2.ff.fc2.weight": {
|
| 135 |
+
"scale": 0.0010037684617888083,
|
| 136 |
+
"nbits": 8,
|
| 137 |
+
"dtype": "int8",
|
| 138 |
+
"emulated": false
|
| 139 |
+
},
|
| 140 |
+
"blocks.3.norm1.weight": {
|
| 141 |
+
"scale": 1.3520645180278738e-07,
|
| 142 |
+
"nbits": 24,
|
| 143 |
+
"dtype": "int32",
|
| 144 |
+
"emulated": true
|
| 145 |
+
},
|
| 146 |
+
"blocks.3.attn.mask": {
|
| 147 |
+
"scale": 0.007874015826771653,
|
| 148 |
+
"nbits": 8,
|
| 149 |
+
"dtype": "int8",
|
| 150 |
+
"emulated": false
|
| 151 |
+
},
|
| 152 |
+
"blocks.3.attn.W_qkv.weight": {
|
| 153 |
+
"scale": 0.0011730166719523753,
|
| 154 |
+
"nbits": 8,
|
| 155 |
+
"dtype": "int8",
|
| 156 |
+
"emulated": false
|
| 157 |
+
},
|
| 158 |
+
"blocks.3.attn.W_o.weight": {
|
| 159 |
+
"scale": 0.0010373295140886681,
|
| 160 |
+
"nbits": 8,
|
| 161 |
+
"dtype": "int8",
|
| 162 |
+
"emulated": false
|
| 163 |
+
},
|
| 164 |
+
"blocks.3.norm2.weight": {
|
| 165 |
+
"scale": 1.3207615540723584e-07,
|
| 166 |
+
"nbits": 24,
|
| 167 |
+
"dtype": "int32",
|
| 168 |
+
"emulated": true
|
| 169 |
+
},
|
| 170 |
+
"blocks.3.ff.fc1.weight": {
|
| 171 |
+
"scale": 0.0011169617888365016,
|
| 172 |
+
"nbits": 8,
|
| 173 |
+
"dtype": "int8",
|
| 174 |
+
"emulated": false
|
| 175 |
+
},
|
| 176 |
+
"blocks.3.ff.fc2.weight": {
|
| 177 |
+
"scale": 0.0008502002038224287,
|
| 178 |
+
"nbits": 8,
|
| 179 |
+
"dtype": "int8",
|
| 180 |
+
"emulated": false
|
| 181 |
+
},
|
| 182 |
+
"blocks.4.norm1.weight": {
|
| 183 |
+
"scale": 1.3003884439983394e-07,
|
| 184 |
+
"nbits": 24,
|
| 185 |
+
"dtype": "int32",
|
| 186 |
+
"emulated": true
|
| 187 |
+
},
|
| 188 |
+
"blocks.4.attn.mask": {
|
| 189 |
+
"scale": 0.007874015826771653,
|
| 190 |
+
"nbits": 8,
|
| 191 |
+
"dtype": "int8",
|
| 192 |
+
"emulated": false
|
| 193 |
+
},
|
| 194 |
+
"blocks.4.attn.W_qkv.weight": {
|
| 195 |
+
"scale": 0.001148089610578582,
|
| 196 |
+
"nbits": 8,
|
| 197 |
+
"dtype": "int8",
|
| 198 |
+
"emulated": false
|
| 199 |
+
},
|
| 200 |
+
"blocks.4.attn.W_o.weight": {
|
| 201 |
+
"scale": 0.0010316128654612143,
|
| 202 |
+
"nbits": 8,
|
| 203 |
+
"dtype": "int8",
|
| 204 |
+
"emulated": false
|
| 205 |
+
},
|
| 206 |
+
"blocks.4.norm2.weight": {
|
| 207 |
+
"scale": 1.3064186788070484e-07,
|
| 208 |
+
"nbits": 24,
|
| 209 |
+
"dtype": "int32",
|
| 210 |
+
"emulated": true
|
| 211 |
+
},
|
| 212 |
+
"blocks.4.ff.fc1.weight": {
|
| 213 |
+
"scale": 0.0011656039895843145,
|
| 214 |
+
"nbits": 8,
|
| 215 |
+
"dtype": "int8",
|
| 216 |
+
"emulated": false
|
| 217 |
+
},
|
| 218 |
+
"blocks.4.ff.fc2.weight": {
|
| 219 |
+
"scale": 0.0006925634595222173,
|
| 220 |
+
"nbits": 8,
|
| 221 |
+
"dtype": "int8",
|
| 222 |
+
"emulated": false
|
| 223 |
+
},
|
| 224 |
+
"blocks.5.norm1.weight": {
|
| 225 |
+
"scale": 1.302203312470367e-07,
|
| 226 |
+
"nbits": 24,
|
| 227 |
+
"dtype": "int32",
|
| 228 |
+
"emulated": true
|
| 229 |
+
},
|
| 230 |
+
"blocks.5.attn.mask": {
|
| 231 |
+
"scale": 0.007874015826771653,
|
| 232 |
+
"nbits": 8,
|
| 233 |
+
"dtype": "int8",
|
| 234 |
+
"emulated": false
|
| 235 |
+
},
|
| 236 |
+
"blocks.5.attn.W_qkv.weight": {
|
| 237 |
+
"scale": 0.0012216476088303093,
|
| 238 |
+
"nbits": 8,
|
| 239 |
+
"dtype": "int8",
|
| 240 |
+
"emulated": false
|
| 241 |
+
},
|
| 242 |
+
"blocks.5.attn.W_o.weight": {
|
| 243 |
+
"scale": 0.0009262990260556739,
|
| 244 |
+
"nbits": 8,
|
| 245 |
+
"dtype": "int8",
|
| 246 |
+
"emulated": false
|
| 247 |
+
},
|
| 248 |
+
"blocks.5.norm2.weight": {
|
| 249 |
+
"scale": 1.308989564840049e-07,
|
| 250 |
+
"nbits": 24,
|
| 251 |
+
"dtype": "int32",
|
| 252 |
+
"emulated": true
|
| 253 |
+
},
|
| 254 |
+
"blocks.5.ff.fc1.weight": {
|
| 255 |
+
"scale": 0.0009419608504622752,
|
| 256 |
+
"nbits": 8,
|
| 257 |
+
"dtype": "int8",
|
| 258 |
+
"emulated": false
|
| 259 |
+
},
|
| 260 |
+
"blocks.5.ff.fc2.weight": {
|
| 261 |
+
"scale": 0.0005326317604001864,
|
| 262 |
+
"nbits": 8,
|
| 263 |
+
"dtype": "int8",
|
| 264 |
+
"emulated": false
|
| 265 |
+
},
|
| 266 |
+
"norm_f.weight": {
|
| 267 |
+
"scale": 1.1641727667871719e-07,
|
| 268 |
+
"nbits": 24,
|
| 269 |
+
"dtype": "int32",
|
| 270 |
+
"emulated": true
|
| 271 |
+
},
|
| 272 |
+
"lm_head.weight": {
|
| 273 |
+
"scale": 0.036867817634565696,
|
| 274 |
+
"nbits": 8,
|
| 275 |
+
"dtype": "int8",
|
| 276 |
+
"emulated": false
|
| 277 |
+
}
|
| 278 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
train_stats.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|