Upload add_sub_baseline_100K
Browse files
add_sub_baseline_100K/config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SorlModelWrapper"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"eos_token_id": null,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 512,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 2048,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention"
|
| 19 |
+
],
|
| 20 |
+
"max_position_embeddings": 128,
|
| 21 |
+
"max_window_layers": 28,
|
| 22 |
+
"model_type": "qwen3",
|
| 23 |
+
"num_attention_heads": 4,
|
| 24 |
+
"num_hidden_layers": 3,
|
| 25 |
+
"num_key_value_heads": 4,
|
| 26 |
+
"pad_token_id": null,
|
| 27 |
+
"rms_norm_eps": 1e-06,
|
| 28 |
+
"rope_parameters": {
|
| 29 |
+
"rope_theta": 10000.0,
|
| 30 |
+
"rope_type": "default"
|
| 31 |
+
},
|
| 32 |
+
"sliding_window": null,
|
| 33 |
+
"tie_word_embeddings": false,
|
| 34 |
+
"transformers_version": "5.5.0",
|
| 35 |
+
"use_cache": true,
|
| 36 |
+
"use_sliding_window": false,
|
| 37 |
+
"vocab_size": 151645
|
| 38 |
+
}
|
add_sub_baseline_100K/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"output_attentions": false,
|
| 4 |
+
"output_hidden_states": false,
|
| 5 |
+
"transformers_version": "5.5.0",
|
| 6 |
+
"use_cache": true
|
| 7 |
+
}
|
add_sub_baseline_100K/metrics.json
ADDED
|
@@ -0,0 +1,385 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"history": {
|
| 3 |
+
"step": [
|
| 4 |
+
50,
|
| 5 |
+
100,
|
| 6 |
+
150,
|
| 7 |
+
200,
|
| 8 |
+
250,
|
| 9 |
+
300,
|
| 10 |
+
350,
|
| 11 |
+
400,
|
| 12 |
+
450,
|
| 13 |
+
500,
|
| 14 |
+
550,
|
| 15 |
+
600,
|
| 16 |
+
650,
|
| 17 |
+
700,
|
| 18 |
+
750,
|
| 19 |
+
800,
|
| 20 |
+
850,
|
| 21 |
+
900,
|
| 22 |
+
950,
|
| 23 |
+
1000,
|
| 24 |
+
1050,
|
| 25 |
+
1100,
|
| 26 |
+
1150,
|
| 27 |
+
1200,
|
| 28 |
+
1250,
|
| 29 |
+
1300,
|
| 30 |
+
1350,
|
| 31 |
+
1400,
|
| 32 |
+
1450,
|
| 33 |
+
1500,
|
| 34 |
+
1550,
|
| 35 |
+
1600,
|
| 36 |
+
1650,
|
| 37 |
+
1700,
|
| 38 |
+
1750,
|
| 39 |
+
1800,
|
| 40 |
+
1850,
|
| 41 |
+
1900,
|
| 42 |
+
1950,
|
| 43 |
+
2000,
|
| 44 |
+
2050,
|
| 45 |
+
2100,
|
| 46 |
+
2150,
|
| 47 |
+
2200,
|
| 48 |
+
2250,
|
| 49 |
+
2300,
|
| 50 |
+
2350,
|
| 51 |
+
2400,
|
| 52 |
+
2450,
|
| 53 |
+
2500,
|
| 54 |
+
2550,
|
| 55 |
+
2600,
|
| 56 |
+
2650,
|
| 57 |
+
2700,
|
| 58 |
+
2750,
|
| 59 |
+
2800,
|
| 60 |
+
2850,
|
| 61 |
+
2900,
|
| 62 |
+
2950,
|
| 63 |
+
3000,
|
| 64 |
+
3050,
|
| 65 |
+
3100,
|
| 66 |
+
3150,
|
| 67 |
+
3200,
|
| 68 |
+
3250,
|
| 69 |
+
3300,
|
| 70 |
+
3350,
|
| 71 |
+
3400,
|
| 72 |
+
3450,
|
| 73 |
+
3500,
|
| 74 |
+
3550,
|
| 75 |
+
3600,
|
| 76 |
+
3650,
|
| 77 |
+
3700,
|
| 78 |
+
3750,
|
| 79 |
+
3800,
|
| 80 |
+
3850,
|
| 81 |
+
3900,
|
| 82 |
+
3950,
|
| 83 |
+
4000,
|
| 84 |
+
4050,
|
| 85 |
+
4100,
|
| 86 |
+
4150,
|
| 87 |
+
4200,
|
| 88 |
+
4250,
|
| 89 |
+
4300,
|
| 90 |
+
4350,
|
| 91 |
+
4400,
|
| 92 |
+
4450,
|
| 93 |
+
4500,
|
| 94 |
+
4550,
|
| 95 |
+
4600,
|
| 96 |
+
4650
|
| 97 |
+
],
|
| 98 |
+
"loss": [
|
| 99 |
+
10.446642875671387,
|
| 100 |
+
8.204933166503906,
|
| 101 |
+
7.287497043609619,
|
| 102 |
+
6.563387393951416,
|
| 103 |
+
5.592346668243408,
|
| 104 |
+
4.6155476570129395,
|
| 105 |
+
3.5229132175445557,
|
| 106 |
+
2.671051025390625,
|
| 107 |
+
2.1285922527313232,
|
| 108 |
+
1.856718897819519,
|
| 109 |
+
1.7758562564849854,
|
| 110 |
+
1.636636734008789,
|
| 111 |
+
1.4528383016586304,
|
| 112 |
+
0.7716706395149231,
|
| 113 |
+
0.47104448080062866,
|
| 114 |
+
0.30569639801979065,
|
| 115 |
+
0.2391534149646759,
|
| 116 |
+
0.17922067642211914,
|
| 117 |
+
0.15349432826042175,
|
| 118 |
+
0.1651923954486847,
|
| 119 |
+
0.17999541759490967,
|
| 120 |
+
0.19358131289482117,
|
| 121 |
+
0.16539938747882843,
|
| 122 |
+
0.1219974234700203,
|
| 123 |
+
0.08426837623119354,
|
| 124 |
+
0.07237092405557632,
|
| 125 |
+
0.048519425094127655,
|
| 126 |
+
0.10355325788259506,
|
| 127 |
+
0.051684457808732986,
|
| 128 |
+
0.09696381539106369,
|
| 129 |
+
0.04080639407038689,
|
| 130 |
+
0.036543793976306915,
|
| 131 |
+
0.02337537333369255,
|
| 132 |
+
0.023294461891055107,
|
| 133 |
+
0.027975469827651978,
|
| 134 |
+
0.03485460206866264,
|
| 135 |
+
0.01895260438323021,
|
| 136 |
+
0.014638577587902546,
|
| 137 |
+
0.020639639347791672,
|
| 138 |
+
0.010763270780444145,
|
| 139 |
+
0.007144362665712833,
|
| 140 |
+
0.024989277124404907,
|
| 141 |
+
0.01930844970047474,
|
| 142 |
+
0.023015184327960014,
|
| 143 |
+
0.0063614509999752045,
|
| 144 |
+
0.0028356898110359907,
|
| 145 |
+
0.002119298791512847,
|
| 146 |
+
0.020854128524661064,
|
| 147 |
+
0.0018869990017265081,
|
| 148 |
+
0.00558065390214324,
|
| 149 |
+
0.006174446549266577,
|
| 150 |
+
0.006082491483539343,
|
| 151 |
+
0.0037679767701774836,
|
| 152 |
+
0.007742547430098057,
|
| 153 |
+
0.0011493159690871835,
|
| 154 |
+
0.006048537325114012,
|
| 155 |
+
0.00800343882292509,
|
| 156 |
+
0.0007379369926638901,
|
| 157 |
+
0.0028449923265725374,
|
| 158 |
+
0.0014030687743797898,
|
| 159 |
+
0.0009590271511115134,
|
| 160 |
+
0.0025089881382882595,
|
| 161 |
+
0.0003829688939731568,
|
| 162 |
+
0.0003303142439108342,
|
| 163 |
+
0.0003975001454818994,
|
| 164 |
+
0.0034398504067212343,
|
| 165 |
+
0.00010646107693901286,
|
| 166 |
+
0.00044194384827278554,
|
| 167 |
+
0.00020321234478615224,
|
| 168 |
+
0.00010634677892085165,
|
| 169 |
+
0.00026234029792249203,
|
| 170 |
+
8.515844820067286e-05,
|
| 171 |
+
6.779689283575863e-05,
|
| 172 |
+
0.0002607057977002114,
|
| 173 |
+
0.00036781796370632946,
|
| 174 |
+
6.288080476224422e-05,
|
| 175 |
+
8.045908907661214e-05,
|
| 176 |
+
0.00011293154238956049,
|
| 177 |
+
0.0005836685886606574,
|
| 178 |
+
5.503458669409156e-05,
|
| 179 |
+
8.547406469006091e-05,
|
| 180 |
+
7.086647383403033e-05,
|
| 181 |
+
4.5273562136571854e-05,
|
| 182 |
+
4.361348692327738e-05,
|
| 183 |
+
3.827679392998107e-05,
|
| 184 |
+
4.5121512812329456e-05,
|
| 185 |
+
5.49159522051923e-05,
|
| 186 |
+
7.655313675059006e-05,
|
| 187 |
+
4.943844396620989e-05,
|
| 188 |
+
3.6672678106697276e-05,
|
| 189 |
+
4.1609022446209565e-05,
|
| 190 |
+
7.665202429052442e-05,
|
| 191 |
+
3.566138548194431e-05
|
| 192 |
+
],
|
| 193 |
+
"base_loss": [
|
| 194 |
+
10.446642875671387,
|
| 195 |
+
8.204933166503906,
|
| 196 |
+
7.287497043609619,
|
| 197 |
+
6.563387393951416,
|
| 198 |
+
5.592346668243408,
|
| 199 |
+
4.6155476570129395,
|
| 200 |
+
3.5229132175445557,
|
| 201 |
+
2.671051025390625,
|
| 202 |
+
2.1285922527313232,
|
| 203 |
+
1.856718897819519,
|
| 204 |
+
1.7758562564849854,
|
| 205 |
+
1.636636734008789,
|
| 206 |
+
1.4528383016586304,
|
| 207 |
+
0.7716706395149231,
|
| 208 |
+
0.47104448080062866,
|
| 209 |
+
0.30569639801979065,
|
| 210 |
+
0.2391534149646759,
|
| 211 |
+
0.17922067642211914,
|
| 212 |
+
0.15349432826042175,
|
| 213 |
+
0.1651923954486847,
|
| 214 |
+
0.17999541759490967,
|
| 215 |
+
0.19358131289482117,
|
| 216 |
+
0.16539938747882843,
|
| 217 |
+
0.1219974234700203,
|
| 218 |
+
0.08426837623119354,
|
| 219 |
+
0.07237092405557632,
|
| 220 |
+
0.048519425094127655,
|
| 221 |
+
0.10355325788259506,
|
| 222 |
+
0.051684457808732986,
|
| 223 |
+
0.09696381539106369,
|
| 224 |
+
0.04080639407038689,
|
| 225 |
+
0.036543793976306915,
|
| 226 |
+
0.02337537333369255,
|
| 227 |
+
0.023294461891055107,
|
| 228 |
+
0.027975469827651978,
|
| 229 |
+
0.03485460206866264,
|
| 230 |
+
0.01895260438323021,
|
| 231 |
+
0.014638577587902546,
|
| 232 |
+
0.020639639347791672,
|
| 233 |
+
0.010763270780444145,
|
| 234 |
+
0.007144362665712833,
|
| 235 |
+
0.024989277124404907,
|
| 236 |
+
0.01930844970047474,
|
| 237 |
+
0.023015184327960014,
|
| 238 |
+
0.0063614509999752045,
|
| 239 |
+
0.0028356898110359907,
|
| 240 |
+
0.002119298791512847,
|
| 241 |
+
0.020854128524661064,
|
| 242 |
+
0.0018869990017265081,
|
| 243 |
+
0.00558065390214324,
|
| 244 |
+
0.006174446549266577,
|
| 245 |
+
0.006082491483539343,
|
| 246 |
+
0.0037679767701774836,
|
| 247 |
+
0.007742547430098057,
|
| 248 |
+
0.0011493159690871835,
|
| 249 |
+
0.006048537325114012,
|
| 250 |
+
0.00800343882292509,
|
| 251 |
+
0.0007379369926638901,
|
| 252 |
+
0.0028449923265725374,
|
| 253 |
+
0.0014030687743797898,
|
| 254 |
+
0.0009590271511115134,
|
| 255 |
+
0.0025089881382882595,
|
| 256 |
+
0.0003829688939731568,
|
| 257 |
+
0.0003303142439108342,
|
| 258 |
+
0.0003975001454818994,
|
| 259 |
+
0.0034398504067212343,
|
| 260 |
+
0.00010646107693901286,
|
| 261 |
+
0.00044194384827278554,
|
| 262 |
+
0.00020321234478615224,
|
| 263 |
+
0.00010634677892085165,
|
| 264 |
+
0.00026234029792249203,
|
| 265 |
+
8.515844820067286e-05,
|
| 266 |
+
6.779689283575863e-05,
|
| 267 |
+
0.0002607057977002114,
|
| 268 |
+
0.00036781796370632946,
|
| 269 |
+
6.288080476224422e-05,
|
| 270 |
+
8.045908907661214e-05,
|
| 271 |
+
0.00011293154238956049,
|
| 272 |
+
0.0005836685886606574,
|
| 273 |
+
5.503458669409156e-05,
|
| 274 |
+
8.547406469006091e-05,
|
| 275 |
+
7.086647383403033e-05,
|
| 276 |
+
4.5273562136571854e-05,
|
| 277 |
+
4.361348692327738e-05,
|
| 278 |
+
3.827679392998107e-05,
|
| 279 |
+
4.5121512812329456e-05,
|
| 280 |
+
5.49159522051923e-05,
|
| 281 |
+
7.655313675059006e-05,
|
| 282 |
+
4.943844396620989e-05,
|
| 283 |
+
3.6672678106697276e-05,
|
| 284 |
+
4.1609022446209565e-05,
|
| 285 |
+
7.665202429052442e-05,
|
| 286 |
+
3.566138548194431e-05
|
| 287 |
+
],
|
| 288 |
+
"lr": [
|
| 289 |
+
4.183564567769478e-06,
|
| 290 |
+
8.452508004268944e-06,
|
| 291 |
+
1.2721451440768411e-05,
|
| 292 |
+
1.6990394877267876e-05,
|
| 293 |
+
2.1259338313767342e-05,
|
| 294 |
+
2.5528281750266808e-05,
|
| 295 |
+
2.979722518676628e-05,
|
| 296 |
+
3.406616862326574e-05,
|
| 297 |
+
3.8335112059765206e-05,
|
| 298 |
+
4.260405549626468e-05,
|
| 299 |
+
4.6872998932764145e-05,
|
| 300 |
+
5.1141942369263614e-05,
|
| 301 |
+
5.541088580576308e-05,
|
| 302 |
+
5.9679829242262546e-05,
|
| 303 |
+
6.394877267876202e-05,
|
| 304 |
+
6.821771611526148e-05,
|
| 305 |
+
7.248665955176095e-05,
|
| 306 |
+
7.675560298826042e-05,
|
| 307 |
+
7.999798087633018e-05,
|
| 308 |
+
7.994611226578842e-05,
|
| 309 |
+
7.982423928087593e-05,
|
| 310 |
+
7.963257550037617e-05,
|
| 311 |
+
7.937145680936735e-05,
|
| 312 |
+
7.904134081059389e-05,
|
| 313 |
+
7.864280602253109e-05,
|
| 314 |
+
7.817655086554887e-05,
|
| 315 |
+
7.764339243795064e-05,
|
| 316 |
+
7.704426508403295e-05,
|
| 317 |
+
7.638021875667483e-05,
|
| 318 |
+
7.565241717732657e-05,
|
| 319 |
+
7.486213579662251e-05,
|
| 320 |
+
7.401075955919167e-05,
|
| 321 |
+
7.309978047658348e-05,
|
| 322 |
+
7.21307950125619e-05,
|
| 323 |
+
7.11055012853501e-05,
|
| 324 |
+
7.002569609172871e-05,
|
| 325 |
+
6.889327175820302e-05,
|
| 326 |
+
6.771021282475686e-05,
|
| 327 |
+
6.647859256700532e-05,
|
| 328 |
+
6.520056936284096e-05,
|
| 329 |
+
6.387838290994056e-05,
|
| 330 |
+
6.251435030076157e-05,
|
| 331 |
+
6.111086196190657e-05,
|
| 332 |
+
5.967037746497144e-05,
|
| 333 |
+
5.819542121621955e-05,
|
| 334 |
+
5.668857803263459e-05,
|
| 335 |
+
5.515248861210576e-05,
|
| 336 |
+
5.358984490568317e-05,
|
| 337 |
+
5.200338540001364e-05,
|
| 338 |
+
5.0395890318224263e-05,
|
| 339 |
+
4.8770176747664026e-05,
|
| 340 |
+
4.712909370304193e-05,
|
| 341 |
+
4.547551713361344e-05,
|
| 342 |
+
4.3812344883164867e-05,
|
| 343 |
+
4.214249161162829e-05,
|
| 344 |
+
4.046888368722666e-05,
|
| 345 |
+
3.879445405810047e-05,
|
| 346 |
+
3.712213711240329e-05,
|
| 347 |
+
3.5454863535873805e-05,
|
| 348 |
+
3.3795555175896186e-05,
|
| 349 |
+
3.214711992104925e-05,
|
| 350 |
+
3.051244660511836e-05,
|
| 351 |
+
2.8894399944499974e-05,
|
| 352 |
+
2.729581551787133e-05,
|
| 353 |
+
2.5719494796923085e-05,
|
| 354 |
+
2.4168200236863477e-05,
|
| 355 |
+
2.264465043529768e-05,
|
| 356 |
+
2.1151515367966308e-05,
|
| 357 |
+
1.969141170969228e-05,
|
| 358 |
+
1.826689824873599e-05,
|
| 359 |
+
1.6880471402594932e-05,
|
| 360 |
+
1.553456084310621e-05,
|
| 361 |
+
1.4231525238518917e-05,
|
| 362 |
+
1.2973648119998136e-05,
|
| 363 |
+
1.1763133879804411e-05,
|
| 364 |
+
1.0602103908161819e-05,
|
| 365 |
+
9.492592875584728e-06,
|
| 366 |
+
8.436545167178147e-06,
|
| 367 |
+
7.43581147516069e-06,
|
| 368 |
+
6.492145555581349e-06,
|
| 369 |
+
5.607201154914275e-06,
|
| 370 |
+
4.78252911191718e-06,
|
| 371 |
+
4.019574639832518e-06,
|
| 372 |
+
3.3196747936942342e-06,
|
| 373 |
+
2.6840561271785694e-06,
|
| 374 |
+
2.113832543105203e-06,
|
| 375 |
+
1.6100033413556416e-06,
|
| 376 |
+
1.1734514676299224e-06,
|
| 377 |
+
8.049419661105129e-07,
|
| 378 |
+
5.051206387450869e-07,
|
| 379 |
+
2.7451291349783527e-07,
|
| 380 |
+
1.1352292355256922e-07,
|
| 381 |
+
2.2432799081339997e-08
|
| 382 |
+
]
|
| 383 |
+
},
|
| 384 |
+
"final_accuracy": 1.0
|
| 385 |
+
}
|
add_sub_baseline_100K/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1fa25258e7da94543218ca081ac521cb27604e416139b5a6a65144bb8ba97da
|
| 3 |
+
size 671794850
|
add_sub_baseline_100K/train_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"mode": "baseline",
|
| 3 |
+
"ops": "add_sub",
|
| 4 |
+
"n_digits": 6,
|
| 5 |
+
"n_layer": 3,
|
| 6 |
+
"n_head": 4,
|
| 7 |
+
"n_embd": 512,
|
| 8 |
+
"abs_vocab": 0,
|
| 9 |
+
"K": 4,
|
| 10 |
+
"batch_size": 64,
|
| 11 |
+
"num_epochs": 3,
|
| 12 |
+
"dataset_size": 100000,
|
| 13 |
+
"lr": 8e-05,
|
| 14 |
+
"output_dir": "ckpt/r/add_sub_baseline_100K",
|
| 15 |
+
"device": "cuda",
|
| 16 |
+
"push_to_hub": true,
|
| 17 |
+
"no_wandb": false,
|
| 18 |
+
"n_params": 167871744,
|
| 19 |
+
"run_name": "add_sub_baseline_100K",
|
| 20 |
+
"git_commit": "9e4530548a98f8c7f5c14930ac4aec4886bb4b1b",
|
| 21 |
+
"timestamp": "2026-04-07T05:30:02.482767",
|
| 22 |
+
"tokenizer": "Qwen/Qwen3-0.6B",
|
| 23 |
+
"dataset_repo": "thoughtworks/arithmetic-sorl-data",
|
| 24 |
+
"dataset_config": "add_sub_6digit",
|
| 25 |
+
"model_repo": "thoughtworks/arithmetic-sorl",
|
| 26 |
+
"trainer_version": "sft",
|
| 27 |
+
"final_accuracy": 1.0
|
| 28 |
+
}
|