Upload folder using huggingface_hub
Browse files- config.json +64 -0
- experiment_cfg/metadata.json +335 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +0 -0
- trainer_state.json +1434 -0
config.json
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"action_dim": 32,
|
| 3 |
+
"action_head_cfg": {
|
| 4 |
+
"action_dim": 32,
|
| 5 |
+
"action_horizon": 16,
|
| 6 |
+
"add_pos_embed": true,
|
| 7 |
+
"backbone_embedding_dim": 2048,
|
| 8 |
+
"diffusion_model_cfg": {
|
| 9 |
+
"attention_head_dim": 48,
|
| 10 |
+
"cross_attention_dim": 2048,
|
| 11 |
+
"dropout": 0.2,
|
| 12 |
+
"final_dropout": true,
|
| 13 |
+
"interleave_self_attention": true,
|
| 14 |
+
"norm_type": "ada_norm",
|
| 15 |
+
"num_attention_heads": 32,
|
| 16 |
+
"num_layers": 16,
|
| 17 |
+
"output_dim": 1024,
|
| 18 |
+
"positional_embeddings": null
|
| 19 |
+
},
|
| 20 |
+
"hidden_size": 1024,
|
| 21 |
+
"input_embedding_dim": 1536,
|
| 22 |
+
"max_action_dim": 32,
|
| 23 |
+
"max_state_dim": 64,
|
| 24 |
+
"model_dtype": "float32",
|
| 25 |
+
"noise_beta_alpha": 1.5,
|
| 26 |
+
"noise_beta_beta": 1.0,
|
| 27 |
+
"noise_s": 0.999,
|
| 28 |
+
"num_inference_timesteps": 4,
|
| 29 |
+
"num_target_vision_tokens": 32,
|
| 30 |
+
"num_timestep_buckets": 1000,
|
| 31 |
+
"tune_diffusion_model": true,
|
| 32 |
+
"tune_projector": true,
|
| 33 |
+
"use_vlln": true,
|
| 34 |
+
"vl_self_attention_cfg": {
|
| 35 |
+
"attention_head_dim": 64,
|
| 36 |
+
"dropout": 0.2,
|
| 37 |
+
"final_dropout": true,
|
| 38 |
+
"num_attention_heads": 32,
|
| 39 |
+
"num_layers": 4,
|
| 40 |
+
"positional_embeddings": null
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"action_horizon": 16,
|
| 44 |
+
"architectures": [
|
| 45 |
+
"GR00T_N1_5"
|
| 46 |
+
],
|
| 47 |
+
"attn_implementation": null,
|
| 48 |
+
"backbone_cfg": {
|
| 49 |
+
"eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops",
|
| 50 |
+
"load_bf16": false,
|
| 51 |
+
"project_to_dim": null,
|
| 52 |
+
"reproject_vision": false,
|
| 53 |
+
"select_layer": 12,
|
| 54 |
+
"tune_llm": false,
|
| 55 |
+
"tune_visual": true,
|
| 56 |
+
"use_flash_attention": true
|
| 57 |
+
},
|
| 58 |
+
"compute_dtype": "bfloat16",
|
| 59 |
+
"hidden_size": 2048,
|
| 60 |
+
"model_dtype": "float32",
|
| 61 |
+
"model_type": "gr00t_n1_5",
|
| 62 |
+
"torch_dtype": "bfloat16",
|
| 63 |
+
"transformers_version": "4.51.3"
|
| 64 |
+
}
|
experiment_cfg/metadata.json
ADDED
|
@@ -0,0 +1,335 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"new_embodiment": {
|
| 3 |
+
"statistics": {
|
| 4 |
+
"state": {
|
| 5 |
+
"left_arm": {
|
| 6 |
+
"max": [
|
| 7 |
+
2.1083366870880127,
|
| 8 |
+
1.6273977756500244,
|
| 9 |
+
1.728448748588562,
|
| 10 |
+
-0.25446105003356934,
|
| 11 |
+
1.7066854238510132,
|
| 12 |
+
1.1433908939361572,
|
| 13 |
+
0.6067746877670288,
|
| 14 |
+
1.0975148677825928
|
| 15 |
+
],
|
| 16 |
+
"min": [
|
| 17 |
+
-1.1136820316314697,
|
| 18 |
+
0.0,
|
| 19 |
+
-1.3611562252044678,
|
| 20 |
+
-2.3707194328308105,
|
| 21 |
+
-1.5844343900680542,
|
| 22 |
+
-1.0783765316009521,
|
| 23 |
+
-1.5404508113861084,
|
| 24 |
+
-0.0014851351734250784
|
| 25 |
+
],
|
| 26 |
+
"mean": [
|
| 27 |
+
0.5819931626319885,
|
| 28 |
+
0.35832464694976807,
|
| 29 |
+
-0.0866265743970871,
|
| 30 |
+
-1.876600980758667,
|
| 31 |
+
0.27218031883239746,
|
| 32 |
+
-0.2956414520740509,
|
| 33 |
+
-0.10110995918512344,
|
| 34 |
+
0.31076350808143616
|
| 35 |
+
],
|
| 36 |
+
"std": [
|
| 37 |
+
0.5412445068359375,
|
| 38 |
+
0.3857501149177551,
|
| 39 |
+
0.38675743341445923,
|
| 40 |
+
0.26987379789352417,
|
| 41 |
+
0.42374488711357117,
|
| 42 |
+
0.4309515357017517,
|
| 43 |
+
0.20699480175971985,
|
| 44 |
+
0.42984575033187866
|
| 45 |
+
],
|
| 46 |
+
"q01": [
|
| 47 |
+
-0.6606351971626282,
|
| 48 |
+
0.0,
|
| 49 |
+
-1.0912253677845,
|
| 50 |
+
-2.3692572116851807,
|
| 51 |
+
-0.8572364151477814,
|
| 52 |
+
-0.9572039842605591,
|
| 53 |
+
-0.7632963478565216,
|
| 54 |
+
0.0
|
| 55 |
+
],
|
| 56 |
+
"q99": [
|
| 57 |
+
1.5600794732570658,
|
| 58 |
+
1.3574411869049072,
|
| 59 |
+
1.2358606159687053,
|
| 60 |
+
-0.8429062426090235,
|
| 61 |
+
1.4359354257583623,
|
| 62 |
+
0.8057965099811555,
|
| 63 |
+
0.25932800322771077,
|
| 64 |
+
1.0960297584533691
|
| 65 |
+
]
|
| 66 |
+
},
|
| 67 |
+
"right_arm": {
|
| 68 |
+
"max": [
|
| 69 |
+
1.8287568092346191,
|
| 70 |
+
0.0,
|
| 71 |
+
1.3382184505462646,
|
| 72 |
+
-0.22242721915245056,
|
| 73 |
+
0.8595685362815857,
|
| 74 |
+
0.9241275787353516,
|
| 75 |
+
1.0220441818237305,
|
| 76 |
+
1.1034554243087769
|
| 77 |
+
],
|
| 78 |
+
"min": [
|
| 79 |
+
-0.9815319776535034,
|
| 80 |
+
-1.6567351818084717,
|
| 81 |
+
-1.280957818031311,
|
| 82 |
+
-2.376268148422241,
|
| 83 |
+
-1.5769321918487549,
|
| 84 |
+
-1.1178884506225586,
|
| 85 |
+
-0.7857963442802429,
|
| 86 |
+
-0.0014851351734250784
|
| 87 |
+
],
|
| 88 |
+
"mean": [
|
| 89 |
+
0.7112464904785156,
|
| 90 |
+
-0.38431429862976074,
|
| 91 |
+
-0.05038423091173172,
|
| 92 |
+
-1.9969016313552856,
|
| 93 |
+
-0.1571054607629776,
|
| 94 |
+
-0.23348413407802582,
|
| 95 |
+
0.05030986666679382,
|
| 96 |
+
0.315399706363678
|
| 97 |
+
],
|
| 98 |
+
"std": [
|
| 99 |
+
0.5694007873535156,
|
| 100 |
+
0.2913043200969696,
|
| 101 |
+
0.3403535485267639,
|
| 102 |
+
0.3016273081302643,
|
| 103 |
+
0.32125237584114075,
|
| 104 |
+
0.44166308641433716,
|
| 105 |
+
0.19431881606578827,
|
| 106 |
+
0.4312712550163269
|
| 107 |
+
],
|
| 108 |
+
"q01": [
|
| 109 |
+
-0.5425132811069489,
|
| 110 |
+
-1.2065567553043366,
|
| 111 |
+
-1.0298469722270966,
|
| 112 |
+
-2.3741109371185303,
|
| 113 |
+
-1.2513364136219025,
|
| 114 |
+
-0.8991026788949966,
|
| 115 |
+
-0.3514146715402603,
|
| 116 |
+
0.0
|
| 117 |
+
],
|
| 118 |
+
"q99": [
|
| 119 |
+
1.3983804643154145,
|
| 120 |
+
0.0,
|
| 121 |
+
0.6476736575365069,
|
| 122 |
+
-0.846702840924263,
|
| 123 |
+
0.4488235995173456,
|
| 124 |
+
0.7369561165571222,
|
| 125 |
+
0.5709165930747986,
|
| 126 |
+
1.0930594205856323
|
| 127 |
+
]
|
| 128 |
+
}
|
| 129 |
+
},
|
| 130 |
+
"action": {
|
| 131 |
+
"left_arm": {
|
| 132 |
+
"max": [
|
| 133 |
+
2.1087193489074707,
|
| 134 |
+
1.628348708152771,
|
| 135 |
+
1.7296408414840698,
|
| 136 |
+
-0.2546408176422119,
|
| 137 |
+
1.7066199779510498,
|
| 138 |
+
1.1449086666107178,
|
| 139 |
+
0.6062183976173401,
|
| 140 |
+
1.2699494361877441
|
| 141 |
+
],
|
| 142 |
+
"min": [
|
| 143 |
+
-1.1152039766311646,
|
| 144 |
+
0.0,
|
| 145 |
+
-1.3621749877929688,
|
| 146 |
+
-2.3715343475341797,
|
| 147 |
+
-1.5892040729522705,
|
| 148 |
+
-1.0783884525299072,
|
| 149 |
+
-1.5401166677474976,
|
| 150 |
+
-1.1618837118148804
|
| 151 |
+
],
|
| 152 |
+
"mean": [
|
| 153 |
+
0.583679735660553,
|
| 154 |
+
0.35692107677459717,
|
| 155 |
+
-0.08536829054355621,
|
| 156 |
+
-1.8764104843139648,
|
| 157 |
+
0.27062463760375977,
|
| 158 |
+
-0.29642942547798157,
|
| 159 |
+
-0.10094120353460312,
|
| 160 |
+
0.28250041604042053
|
| 161 |
+
],
|
| 162 |
+
"std": [
|
| 163 |
+
0.5405545234680176,
|
| 164 |
+
0.38556116819381714,
|
| 165 |
+
0.3860621154308319,
|
| 166 |
+
0.270335853099823,
|
| 167 |
+
0.4224850833415985,
|
| 168 |
+
0.431160569190979,
|
| 169 |
+
0.20703385770320892,
|
| 170 |
+
0.5268236994743347
|
| 171 |
+
],
|
| 172 |
+
"q01": [
|
| 173 |
+
-0.6596117615699768,
|
| 174 |
+
0.0,
|
| 175 |
+
-1.0891263484954834,
|
| 176 |
+
-2.3700003623962402,
|
| 177 |
+
-0.8598729252815246,
|
| 178 |
+
-0.9572039842605591,
|
| 179 |
+
-0.7639224529266357,
|
| 180 |
+
-0.37648552656173706
|
| 181 |
+
],
|
| 182 |
+
"q99": [
|
| 183 |
+
1.5623552799224854,
|
| 184 |
+
1.3590802788734482,
|
| 185 |
+
1.2385272979736328,
|
| 186 |
+
-0.8390874862670898,
|
| 187 |
+
1.4349727630615234,
|
| 188 |
+
0.8072680830955505,
|
| 189 |
+
0.2593694031238556,
|
| 190 |
+
1.187074065208435
|
| 191 |
+
]
|
| 192 |
+
},
|
| 193 |
+
"right_arm": {
|
| 194 |
+
"max": [
|
| 195 |
+
1.8293983936309814,
|
| 196 |
+
0.0,
|
| 197 |
+
1.3382847309112549,
|
| 198 |
+
-0.22242721915245056,
|
| 199 |
+
0.8609836101531982,
|
| 200 |
+
0.9239075779914856,
|
| 201 |
+
1.0221302509307861,
|
| 202 |
+
1.3436164855957031
|
| 203 |
+
],
|
| 204 |
+
"min": [
|
| 205 |
+
-0.9817476868629456,
|
| 206 |
+
-1.656699299812317,
|
| 207 |
+
-1.2824078798294067,
|
| 208 |
+
-2.376136302947998,
|
| 209 |
+
-1.5769321918487549,
|
| 210 |
+
-1.0952622890472412,
|
| 211 |
+
-0.7853981852531433,
|
| 212 |
+
-0.7415729761123657
|
| 213 |
+
],
|
| 214 |
+
"mean": [
|
| 215 |
+
0.7114920020103455,
|
| 216 |
+
-0.38386180996894836,
|
| 217 |
+
-0.0503559336066246,
|
| 218 |
+
-1.9966758489608765,
|
| 219 |
+
-0.15680620074272156,
|
| 220 |
+
-0.23381134867668152,
|
| 221 |
+
0.05017194524407387,
|
| 222 |
+
0.36734870076179504
|
| 223 |
+
],
|
| 224 |
+
"std": [
|
| 225 |
+
0.5695649981498718,
|
| 226 |
+
0.2916468679904938,
|
| 227 |
+
0.34082460403442383,
|
| 228 |
+
0.3023609220981598,
|
| 229 |
+
0.3213924467563629,
|
| 230 |
+
0.4416179060935974,
|
| 231 |
+
0.19426511228084564,
|
| 232 |
+
0.5313277244567871
|
| 233 |
+
],
|
| 234 |
+
"q01": [
|
| 235 |
+
-0.5430291891098022,
|
| 236 |
+
-1.208776831626892,
|
| 237 |
+
-1.0308350324630737,
|
| 238 |
+
-2.3746023178100586,
|
| 239 |
+
-1.2501943111419678,
|
| 240 |
+
-0.8989127278327942,
|
| 241 |
+
-0.35128161311149597,
|
| 242 |
+
-0.08502917736768723
|
| 243 |
+
],
|
| 244 |
+
"q99": [
|
| 245 |
+
1.398139238357544,
|
| 246 |
+
0.0,
|
| 247 |
+
0.6491908431053162,
|
| 248 |
+
-0.8406214714050293,
|
| 249 |
+
0.44967594742774963,
|
| 250 |
+
0.7366704940795898,
|
| 251 |
+
0.5709196329116821,
|
| 252 |
+
1.3221303224563599
|
| 253 |
+
]
|
| 254 |
+
},
|
| 255 |
+
"conditional_joint": {
|
| 256 |
+
"max": [
|
| 257 |
+
1.0
|
| 258 |
+
],
|
| 259 |
+
"min": [
|
| 260 |
+
0.0
|
| 261 |
+
],
|
| 262 |
+
"mean": [
|
| 263 |
+
0.5180045366287231
|
| 264 |
+
],
|
| 265 |
+
"std": [
|
| 266 |
+
0.49958887696266174
|
| 267 |
+
],
|
| 268 |
+
"q01": [
|
| 269 |
+
0.0
|
| 270 |
+
],
|
| 271 |
+
"q99": [
|
| 272 |
+
1.0
|
| 273 |
+
]
|
| 274 |
+
}
|
| 275 |
+
}
|
| 276 |
+
},
|
| 277 |
+
"modalities": {
|
| 278 |
+
"video": {
|
| 279 |
+
"cam_head": {
|
| 280 |
+
"resolution": [
|
| 281 |
+
672,
|
| 282 |
+
376
|
| 283 |
+
],
|
| 284 |
+
"channels": 3,
|
| 285 |
+
"fps": 10.0
|
| 286 |
+
}
|
| 287 |
+
},
|
| 288 |
+
"state": {
|
| 289 |
+
"left_arm": {
|
| 290 |
+
"absolute": true,
|
| 291 |
+
"rotation_type": null,
|
| 292 |
+
"shape": [
|
| 293 |
+
8
|
| 294 |
+
],
|
| 295 |
+
"continuous": true
|
| 296 |
+
},
|
| 297 |
+
"right_arm": {
|
| 298 |
+
"absolute": true,
|
| 299 |
+
"rotation_type": null,
|
| 300 |
+
"shape": [
|
| 301 |
+
8
|
| 302 |
+
],
|
| 303 |
+
"continuous": true
|
| 304 |
+
}
|
| 305 |
+
},
|
| 306 |
+
"action": {
|
| 307 |
+
"left_arm": {
|
| 308 |
+
"absolute": true,
|
| 309 |
+
"rotation_type": null,
|
| 310 |
+
"shape": [
|
| 311 |
+
8
|
| 312 |
+
],
|
| 313 |
+
"continuous": true
|
| 314 |
+
},
|
| 315 |
+
"right_arm": {
|
| 316 |
+
"absolute": true,
|
| 317 |
+
"rotation_type": null,
|
| 318 |
+
"shape": [
|
| 319 |
+
8
|
| 320 |
+
],
|
| 321 |
+
"continuous": true
|
| 322 |
+
},
|
| 323 |
+
"conditional_joint": {
|
| 324 |
+
"absolute": true,
|
| 325 |
+
"rotation_type": null,
|
| 326 |
+
"shape": [
|
| 327 |
+
1
|
| 328 |
+
],
|
| 329 |
+
"continuous": true
|
| 330 |
+
}
|
| 331 |
+
}
|
| 332 |
+
},
|
| 333 |
+
"embodiment_tag": "new_embodiment"
|
| 334 |
+
}
|
| 335 |
+
}
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1be1608bd2c53e44f77a10d9f29d85c1300c24fd525b095cca989dcf2ff48800
|
| 3 |
+
size 4999367032
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf1c313d73239809b47dc3ffb5766de5d94cf7a9d23d521cb575112d4c2641e2
|
| 3 |
+
size 2586705312
|
model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1434 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 64.51612903225806,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.3225806451612903,
|
| 14 |
+
"grad_norm": 4.83341646194458,
|
| 15 |
+
"learning_rate": 1.8e-07,
|
| 16 |
+
"loss": 0.6146,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.6451612903225806,
|
| 21 |
+
"grad_norm": 4.601212978363037,
|
| 22 |
+
"learning_rate": 3.8e-07,
|
| 23 |
+
"loss": 0.6156,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.967741935483871,
|
| 28 |
+
"grad_norm": 4.288501739501953,
|
| 29 |
+
"learning_rate": 5.8e-07,
|
| 30 |
+
"loss": 0.6041,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.2903225806451613,
|
| 35 |
+
"grad_norm": 3.9819118976593018,
|
| 36 |
+
"learning_rate": 7.8e-07,
|
| 37 |
+
"loss": 0.5691,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.6129032258064515,
|
| 42 |
+
"grad_norm": 3.2057559490203857,
|
| 43 |
+
"learning_rate": 9.8e-07,
|
| 44 |
+
"loss": 0.4867,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.935483870967742,
|
| 49 |
+
"grad_norm": 1.8752663135528564,
|
| 50 |
+
"learning_rate": 1.18e-06,
|
| 51 |
+
"loss": 0.4109,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 2.258064516129032,
|
| 56 |
+
"grad_norm": 1.5188192129135132,
|
| 57 |
+
"learning_rate": 1.3800000000000001e-06,
|
| 58 |
+
"loss": 0.3164,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 2.5806451612903225,
|
| 63 |
+
"grad_norm": 0.7869158983230591,
|
| 64 |
+
"learning_rate": 1.5800000000000003e-06,
|
| 65 |
+
"loss": 0.2642,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 2.903225806451613,
|
| 70 |
+
"grad_norm": 0.499815970659256,
|
| 71 |
+
"learning_rate": 1.7800000000000001e-06,
|
| 72 |
+
"loss": 0.2078,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 3.225806451612903,
|
| 77 |
+
"grad_norm": 0.36157462000846863,
|
| 78 |
+
"learning_rate": 1.98e-06,
|
| 79 |
+
"loss": 0.1876,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 3.5483870967741935,
|
| 84 |
+
"grad_norm": 0.4228762686252594,
|
| 85 |
+
"learning_rate": 2.1800000000000003e-06,
|
| 86 |
+
"loss": 0.167,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 3.870967741935484,
|
| 91 |
+
"grad_norm": 0.22709406912326813,
|
| 92 |
+
"learning_rate": 2.38e-06,
|
| 93 |
+
"loss": 0.1528,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 4.193548387096774,
|
| 98 |
+
"grad_norm": 0.2506711184978485,
|
| 99 |
+
"learning_rate": 2.5800000000000003e-06,
|
| 100 |
+
"loss": 0.141,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 4.516129032258064,
|
| 105 |
+
"grad_norm": 0.17586937546730042,
|
| 106 |
+
"learning_rate": 2.78e-06,
|
| 107 |
+
"loss": 0.1301,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 4.838709677419355,
|
| 112 |
+
"grad_norm": 0.2080131471157074,
|
| 113 |
+
"learning_rate": 2.9800000000000003e-06,
|
| 114 |
+
"loss": 0.1252,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 5.161290322580645,
|
| 119 |
+
"grad_norm": 0.14357304573059082,
|
| 120 |
+
"learning_rate": 3.1800000000000005e-06,
|
| 121 |
+
"loss": 0.1174,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 5.483870967741936,
|
| 126 |
+
"grad_norm": 0.15768013894557953,
|
| 127 |
+
"learning_rate": 3.38e-06,
|
| 128 |
+
"loss": 0.1123,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 5.806451612903226,
|
| 133 |
+
"grad_norm": 0.1478148102760315,
|
| 134 |
+
"learning_rate": 3.58e-06,
|
| 135 |
+
"loss": 0.1059,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 6.129032258064516,
|
| 140 |
+
"grad_norm": 0.11791401356458664,
|
| 141 |
+
"learning_rate": 3.7800000000000002e-06,
|
| 142 |
+
"loss": 0.1028,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 6.451612903225806,
|
| 147 |
+
"grad_norm": 0.12674356997013092,
|
| 148 |
+
"learning_rate": 3.98e-06,
|
| 149 |
+
"loss": 0.1002,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 6.774193548387097,
|
| 154 |
+
"grad_norm": 0.10726606845855713,
|
| 155 |
+
"learning_rate": 4.18e-06,
|
| 156 |
+
"loss": 0.0936,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 7.096774193548387,
|
| 161 |
+
"grad_norm": 0.13341137766838074,
|
| 162 |
+
"learning_rate": 4.38e-06,
|
| 163 |
+
"loss": 0.093,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 7.419354838709677,
|
| 168 |
+
"grad_norm": 0.09624793380498886,
|
| 169 |
+
"learning_rate": 4.58e-06,
|
| 170 |
+
"loss": 0.0891,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 7.741935483870968,
|
| 175 |
+
"grad_norm": 0.14156094193458557,
|
| 176 |
+
"learning_rate": 4.780000000000001e-06,
|
| 177 |
+
"loss": 0.0861,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 8.064516129032258,
|
| 182 |
+
"grad_norm": 0.10326485335826874,
|
| 183 |
+
"learning_rate": 4.98e-06,
|
| 184 |
+
"loss": 0.0835,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 8.387096774193548,
|
| 189 |
+
"grad_norm": 0.08262120187282562,
|
| 190 |
+
"learning_rate": 5.18e-06,
|
| 191 |
+
"loss": 0.0805,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 8.709677419354838,
|
| 196 |
+
"grad_norm": 0.1273198425769806,
|
| 197 |
+
"learning_rate": 5.38e-06,
|
| 198 |
+
"loss": 0.0779,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 9.03225806451613,
|
| 203 |
+
"grad_norm": 0.12214396893978119,
|
| 204 |
+
"learning_rate": 5.580000000000001e-06,
|
| 205 |
+
"loss": 0.0757,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 9.35483870967742,
|
| 210 |
+
"grad_norm": 0.1394161581993103,
|
| 211 |
+
"learning_rate": 5.78e-06,
|
| 212 |
+
"loss": 0.0728,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 9.67741935483871,
|
| 217 |
+
"grad_norm": 0.12285201996564865,
|
| 218 |
+
"learning_rate": 5.98e-06,
|
| 219 |
+
"loss": 0.0722,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 10.0,
|
| 224 |
+
"grad_norm": 0.15410321950912476,
|
| 225 |
+
"learning_rate": 6.18e-06,
|
| 226 |
+
"loss": 0.0692,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 10.32258064516129,
|
| 231 |
+
"grad_norm": 0.07469774037599564,
|
| 232 |
+
"learning_rate": 6.38e-06,
|
| 233 |
+
"loss": 0.0676,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 10.64516129032258,
|
| 238 |
+
"grad_norm": 0.16188545525074005,
|
| 239 |
+
"learning_rate": 6.58e-06,
|
| 240 |
+
"loss": 0.0643,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 10.967741935483872,
|
| 245 |
+
"grad_norm": 0.07787156105041504,
|
| 246 |
+
"learning_rate": 6.78e-06,
|
| 247 |
+
"loss": 0.0632,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 11.290322580645162,
|
| 252 |
+
"grad_norm": 0.13963359594345093,
|
| 253 |
+
"learning_rate": 6.98e-06,
|
| 254 |
+
"loss": 0.0593,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 11.612903225806452,
|
| 259 |
+
"grad_norm": 0.12019138038158417,
|
| 260 |
+
"learning_rate": 7.180000000000001e-06,
|
| 261 |
+
"loss": 0.0594,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 11.935483870967742,
|
| 266 |
+
"grad_norm": 0.13999758660793304,
|
| 267 |
+
"learning_rate": 7.3800000000000005e-06,
|
| 268 |
+
"loss": 0.0578,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 12.258064516129032,
|
| 273 |
+
"grad_norm": 0.11083265393972397,
|
| 274 |
+
"learning_rate": 7.580000000000001e-06,
|
| 275 |
+
"loss": 0.0544,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 12.580645161290322,
|
| 280 |
+
"grad_norm": 0.10924753546714783,
|
| 281 |
+
"learning_rate": 7.78e-06,
|
| 282 |
+
"loss": 0.0529,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 12.903225806451612,
|
| 287 |
+
"grad_norm": 0.10007922351360321,
|
| 288 |
+
"learning_rate": 7.98e-06,
|
| 289 |
+
"loss": 0.0519,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 13.225806451612904,
|
| 294 |
+
"grad_norm": 0.09409403055906296,
|
| 295 |
+
"learning_rate": 8.18e-06,
|
| 296 |
+
"loss": 0.0499,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 13.548387096774194,
|
| 301 |
+
"grad_norm": 0.1096971333026886,
|
| 302 |
+
"learning_rate": 8.380000000000001e-06,
|
| 303 |
+
"loss": 0.049,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 13.870967741935484,
|
| 308 |
+
"grad_norm": 0.10034970194101334,
|
| 309 |
+
"learning_rate": 8.580000000000001e-06,
|
| 310 |
+
"loss": 0.0462,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 14.193548387096774,
|
| 315 |
+
"grad_norm": 0.13287365436553955,
|
| 316 |
+
"learning_rate": 8.78e-06,
|
| 317 |
+
"loss": 0.0454,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 14.516129032258064,
|
| 322 |
+
"grad_norm": 0.10294642299413681,
|
| 323 |
+
"learning_rate": 8.98e-06,
|
| 324 |
+
"loss": 0.0436,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 14.838709677419354,
|
| 329 |
+
"grad_norm": 0.11570604145526886,
|
| 330 |
+
"learning_rate": 9.180000000000002e-06,
|
| 331 |
+
"loss": 0.0404,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 15.161290322580646,
|
| 336 |
+
"grad_norm": 0.08866477012634277,
|
| 337 |
+
"learning_rate": 9.38e-06,
|
| 338 |
+
"loss": 0.0395,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 15.483870967741936,
|
| 343 |
+
"grad_norm": 0.1237388625741005,
|
| 344 |
+
"learning_rate": 9.58e-06,
|
| 345 |
+
"loss": 0.0377,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 15.806451612903226,
|
| 350 |
+
"grad_norm": 0.09724581241607666,
|
| 351 |
+
"learning_rate": 9.78e-06,
|
| 352 |
+
"loss": 0.0364,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 16.129032258064516,
|
| 357 |
+
"grad_norm": 0.1216067299246788,
|
| 358 |
+
"learning_rate": 9.980000000000001e-06,
|
| 359 |
+
"loss": 0.0369,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 16.451612903225808,
|
| 364 |
+
"grad_norm": 0.14183978736400604,
|
| 365 |
+
"learning_rate": 1.018e-05,
|
| 366 |
+
"loss": 0.0335,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 16.774193548387096,
|
| 371 |
+
"grad_norm": 0.10971418768167496,
|
| 372 |
+
"learning_rate": 1.038e-05,
|
| 373 |
+
"loss": 0.0337,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 17.096774193548388,
|
| 378 |
+
"grad_norm": 0.10205938667058945,
|
| 379 |
+
"learning_rate": 1.058e-05,
|
| 380 |
+
"loss": 0.0314,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 17.419354838709676,
|
| 385 |
+
"grad_norm": 0.08642183989286423,
|
| 386 |
+
"learning_rate": 1.0780000000000002e-05,
|
| 387 |
+
"loss": 0.0309,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 17.741935483870968,
|
| 392 |
+
"grad_norm": 0.11878576874732971,
|
| 393 |
+
"learning_rate": 1.098e-05,
|
| 394 |
+
"loss": 0.0313,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 18.06451612903226,
|
| 399 |
+
"grad_norm": 0.11662951111793518,
|
| 400 |
+
"learning_rate": 1.118e-05,
|
| 401 |
+
"loss": 0.03,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 18.387096774193548,
|
| 406 |
+
"grad_norm": 0.13768725097179413,
|
| 407 |
+
"learning_rate": 1.1380000000000001e-05,
|
| 408 |
+
"loss": 0.0285,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 18.70967741935484,
|
| 413 |
+
"grad_norm": 0.1288958489894867,
|
| 414 |
+
"learning_rate": 1.1580000000000001e-05,
|
| 415 |
+
"loss": 0.0288,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 19.032258064516128,
|
| 420 |
+
"grad_norm": 0.11418294161558151,
|
| 421 |
+
"learning_rate": 1.178e-05,
|
| 422 |
+
"loss": 0.0289,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 19.35483870967742,
|
| 427 |
+
"grad_norm": 0.12079351395368576,
|
| 428 |
+
"learning_rate": 1.198e-05,
|
| 429 |
+
"loss": 0.0287,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 19.677419354838708,
|
| 434 |
+
"grad_norm": 0.2023964524269104,
|
| 435 |
+
"learning_rate": 1.2180000000000002e-05,
|
| 436 |
+
"loss": 0.0279,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 20.0,
|
| 441 |
+
"grad_norm": 0.19349494576454163,
|
| 442 |
+
"learning_rate": 1.238e-05,
|
| 443 |
+
"loss": 0.0278,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 20.322580645161292,
|
| 448 |
+
"grad_norm": 0.17760895192623138,
|
| 449 |
+
"learning_rate": 1.258e-05,
|
| 450 |
+
"loss": 0.0273,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 20.64516129032258,
|
| 455 |
+
"grad_norm": 0.1515321135520935,
|
| 456 |
+
"learning_rate": 1.278e-05,
|
| 457 |
+
"loss": 0.0257,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 20.967741935483872,
|
| 462 |
+
"grad_norm": 0.10156595706939697,
|
| 463 |
+
"learning_rate": 1.2980000000000001e-05,
|
| 464 |
+
"loss": 0.0257,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 21.29032258064516,
|
| 469 |
+
"grad_norm": 0.12676751613616943,
|
| 470 |
+
"learning_rate": 1.3180000000000001e-05,
|
| 471 |
+
"loss": 0.0249,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 21.612903225806452,
|
| 476 |
+
"grad_norm": 0.13134433329105377,
|
| 477 |
+
"learning_rate": 1.338e-05,
|
| 478 |
+
"loss": 0.0239,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 21.93548387096774,
|
| 483 |
+
"grad_norm": 0.13195613026618958,
|
| 484 |
+
"learning_rate": 1.358e-05,
|
| 485 |
+
"loss": 0.0242,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 22.258064516129032,
|
| 490 |
+
"grad_norm": 0.12307468056678772,
|
| 491 |
+
"learning_rate": 1.3780000000000002e-05,
|
| 492 |
+
"loss": 0.0238,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 22.580645161290324,
|
| 497 |
+
"grad_norm": 0.1110624298453331,
|
| 498 |
+
"learning_rate": 1.3980000000000002e-05,
|
| 499 |
+
"loss": 0.0253,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 22.903225806451612,
|
| 504 |
+
"grad_norm": 0.13264206051826477,
|
| 505 |
+
"learning_rate": 1.4180000000000001e-05,
|
| 506 |
+
"loss": 0.0239,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 23.225806451612904,
|
| 511 |
+
"grad_norm": 0.11219354718923569,
|
| 512 |
+
"learning_rate": 1.4380000000000001e-05,
|
| 513 |
+
"loss": 0.0234,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 23.548387096774192,
|
| 518 |
+
"grad_norm": 0.14356958866119385,
|
| 519 |
+
"learning_rate": 1.4580000000000003e-05,
|
| 520 |
+
"loss": 0.0225,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 23.870967741935484,
|
| 525 |
+
"grad_norm": 0.13088181614875793,
|
| 526 |
+
"learning_rate": 1.4779999999999999e-05,
|
| 527 |
+
"loss": 0.0221,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 24.193548387096776,
|
| 532 |
+
"grad_norm": 0.15026481449604034,
|
| 533 |
+
"learning_rate": 1.4979999999999999e-05,
|
| 534 |
+
"loss": 0.0225,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 24.516129032258064,
|
| 539 |
+
"grad_norm": 0.20722658932209015,
|
| 540 |
+
"learning_rate": 1.518e-05,
|
| 541 |
+
"loss": 0.0225,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 24.838709677419356,
|
| 546 |
+
"grad_norm": 0.13866043090820312,
|
| 547 |
+
"learning_rate": 1.538e-05,
|
| 548 |
+
"loss": 0.0221,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 25.161290322580644,
|
| 553 |
+
"grad_norm": 0.11805257201194763,
|
| 554 |
+
"learning_rate": 1.558e-05,
|
| 555 |
+
"loss": 0.0214,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 25.483870967741936,
|
| 560 |
+
"grad_norm": 0.1360214352607727,
|
| 561 |
+
"learning_rate": 1.578e-05,
|
| 562 |
+
"loss": 0.0205,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 25.806451612903224,
|
| 567 |
+
"grad_norm": 0.15285663306713104,
|
| 568 |
+
"learning_rate": 1.598e-05,
|
| 569 |
+
"loss": 0.0215,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 26.129032258064516,
|
| 574 |
+
"grad_norm": 0.13831627368927002,
|
| 575 |
+
"learning_rate": 1.618e-05,
|
| 576 |
+
"loss": 0.021,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 26.451612903225808,
|
| 581 |
+
"grad_norm": 0.1219727173447609,
|
| 582 |
+
"learning_rate": 1.6380000000000002e-05,
|
| 583 |
+
"loss": 0.0208,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 26.774193548387096,
|
| 588 |
+
"grad_norm": 0.14780209958553314,
|
| 589 |
+
"learning_rate": 1.658e-05,
|
| 590 |
+
"loss": 0.0212,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 27.096774193548388,
|
| 595 |
+
"grad_norm": 0.13057461380958557,
|
| 596 |
+
"learning_rate": 1.6780000000000002e-05,
|
| 597 |
+
"loss": 0.0206,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 27.419354838709676,
|
| 602 |
+
"grad_norm": 0.13308848440647125,
|
| 603 |
+
"learning_rate": 1.698e-05,
|
| 604 |
+
"loss": 0.021,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 27.741935483870968,
|
| 609 |
+
"grad_norm": 0.1406048834323883,
|
| 610 |
+
"learning_rate": 1.718e-05,
|
| 611 |
+
"loss": 0.0204,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 28.06451612903226,
|
| 616 |
+
"grad_norm": 0.1681479662656784,
|
| 617 |
+
"learning_rate": 1.7380000000000003e-05,
|
| 618 |
+
"loss": 0.0203,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 28.387096774193548,
|
| 623 |
+
"grad_norm": 0.15317517518997192,
|
| 624 |
+
"learning_rate": 1.758e-05,
|
| 625 |
+
"loss": 0.0192,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 28.70967741935484,
|
| 630 |
+
"grad_norm": 0.17327739298343658,
|
| 631 |
+
"learning_rate": 1.7780000000000003e-05,
|
| 632 |
+
"loss": 0.0191,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 29.032258064516128,
|
| 637 |
+
"grad_norm": 0.13825540244579315,
|
| 638 |
+
"learning_rate": 1.798e-05,
|
| 639 |
+
"loss": 0.0194,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 29.35483870967742,
|
| 644 |
+
"grad_norm": 0.14294251799583435,
|
| 645 |
+
"learning_rate": 1.818e-05,
|
| 646 |
+
"loss": 0.0191,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 29.677419354838708,
|
| 651 |
+
"grad_norm": 0.1148597002029419,
|
| 652 |
+
"learning_rate": 1.838e-05,
|
| 653 |
+
"loss": 0.0194,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 30.0,
|
| 658 |
+
"grad_norm": 0.10468383878469467,
|
| 659 |
+
"learning_rate": 1.858e-05,
|
| 660 |
+
"loss": 0.019,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 30.322580645161292,
|
| 665 |
+
"grad_norm": 0.16523204743862152,
|
| 666 |
+
"learning_rate": 1.878e-05,
|
| 667 |
+
"loss": 0.0183,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 30.64516129032258,
|
| 672 |
+
"grad_norm": 0.1483301818370819,
|
| 673 |
+
"learning_rate": 1.898e-05,
|
| 674 |
+
"loss": 0.0183,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 30.967741935483872,
|
| 679 |
+
"grad_norm": 0.19092631340026855,
|
| 680 |
+
"learning_rate": 1.918e-05,
|
| 681 |
+
"loss": 0.0188,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 31.29032258064516,
|
| 686 |
+
"grad_norm": 0.19763915240764618,
|
| 687 |
+
"learning_rate": 1.938e-05,
|
| 688 |
+
"loss": 0.0181,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 31.612903225806452,
|
| 693 |
+
"grad_norm": 0.18357335031032562,
|
| 694 |
+
"learning_rate": 1.9580000000000002e-05,
|
| 695 |
+
"loss": 0.0179,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 31.93548387096774,
|
| 700 |
+
"grad_norm": 0.1674022376537323,
|
| 701 |
+
"learning_rate": 1.978e-05,
|
| 702 |
+
"loss": 0.0181,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 32.25806451612903,
|
| 707 |
+
"grad_norm": 0.1773853302001953,
|
| 708 |
+
"learning_rate": 1.9980000000000002e-05,
|
| 709 |
+
"loss": 0.0187,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 32.58064516129032,
|
| 714 |
+
"grad_norm": 0.17372146248817444,
|
| 715 |
+
"learning_rate": 2.0180000000000003e-05,
|
| 716 |
+
"loss": 0.0175,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 32.903225806451616,
|
| 721 |
+
"grad_norm": 0.16912385821342468,
|
| 722 |
+
"learning_rate": 2.038e-05,
|
| 723 |
+
"loss": 0.0177,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 33.225806451612904,
|
| 728 |
+
"grad_norm": 0.1924871802330017,
|
| 729 |
+
"learning_rate": 2.0580000000000003e-05,
|
| 730 |
+
"loss": 0.017,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 33.54838709677419,
|
| 735 |
+
"grad_norm": 0.1980140060186386,
|
| 736 |
+
"learning_rate": 2.078e-05,
|
| 737 |
+
"loss": 0.0174,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 33.87096774193548,
|
| 742 |
+
"grad_norm": 0.15361692011356354,
|
| 743 |
+
"learning_rate": 2.098e-05,
|
| 744 |
+
"loss": 0.0169,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 34.193548387096776,
|
| 749 |
+
"grad_norm": 0.15744690597057343,
|
| 750 |
+
"learning_rate": 2.118e-05,
|
| 751 |
+
"loss": 0.0174,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 34.516129032258064,
|
| 756 |
+
"grad_norm": 0.18941748142242432,
|
| 757 |
+
"learning_rate": 2.138e-05,
|
| 758 |
+
"loss": 0.0173,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 34.83870967741935,
|
| 763 |
+
"grad_norm": 0.18401823937892914,
|
| 764 |
+
"learning_rate": 2.158e-05,
|
| 765 |
+
"loss": 0.0176,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 35.16129032258065,
|
| 770 |
+
"grad_norm": 0.1347324103116989,
|
| 771 |
+
"learning_rate": 2.178e-05,
|
| 772 |
+
"loss": 0.017,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 35.483870967741936,
|
| 777 |
+
"grad_norm": 0.13733318448066711,
|
| 778 |
+
"learning_rate": 2.198e-05,
|
| 779 |
+
"loss": 0.0163,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 35.806451612903224,
|
| 784 |
+
"grad_norm": 0.17490868270397186,
|
| 785 |
+
"learning_rate": 2.218e-05,
|
| 786 |
+
"loss": 0.0167,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 36.12903225806452,
|
| 791 |
+
"grad_norm": 0.17697739601135254,
|
| 792 |
+
"learning_rate": 2.2380000000000003e-05,
|
| 793 |
+
"loss": 0.0173,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 36.45161290322581,
|
| 798 |
+
"grad_norm": 0.19351543486118317,
|
| 799 |
+
"learning_rate": 2.258e-05,
|
| 800 |
+
"loss": 0.0174,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 36.774193548387096,
|
| 805 |
+
"grad_norm": 0.17104071378707886,
|
| 806 |
+
"learning_rate": 2.2780000000000002e-05,
|
| 807 |
+
"loss": 0.0159,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 37.096774193548384,
|
| 812 |
+
"grad_norm": 0.1536385416984558,
|
| 813 |
+
"learning_rate": 2.298e-05,
|
| 814 |
+
"loss": 0.016,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 37.41935483870968,
|
| 819 |
+
"grad_norm": 0.14215905964374542,
|
| 820 |
+
"learning_rate": 2.318e-05,
|
| 821 |
+
"loss": 0.0167,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 37.74193548387097,
|
| 826 |
+
"grad_norm": 0.17754510045051575,
|
| 827 |
+
"learning_rate": 2.3380000000000003e-05,
|
| 828 |
+
"loss": 0.0163,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 38.064516129032256,
|
| 833 |
+
"grad_norm": 0.15013834834098816,
|
| 834 |
+
"learning_rate": 2.358e-05,
|
| 835 |
+
"loss": 0.0152,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 38.38709677419355,
|
| 840 |
+
"grad_norm": 0.17872834205627441,
|
| 841 |
+
"learning_rate": 2.3780000000000003e-05,
|
| 842 |
+
"loss": 0.0152,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 38.70967741935484,
|
| 847 |
+
"grad_norm": 0.16451333463191986,
|
| 848 |
+
"learning_rate": 2.398e-05,
|
| 849 |
+
"loss": 0.015,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 39.03225806451613,
|
| 854 |
+
"grad_norm": 0.17044350504875183,
|
| 855 |
+
"learning_rate": 2.418e-05,
|
| 856 |
+
"loss": 0.0149,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 39.354838709677416,
|
| 861 |
+
"grad_norm": 0.16590650379657745,
|
| 862 |
+
"learning_rate": 2.438e-05,
|
| 863 |
+
"loss": 0.0155,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 39.67741935483871,
|
| 868 |
+
"grad_norm": 0.1371670961380005,
|
| 869 |
+
"learning_rate": 2.4580000000000002e-05,
|
| 870 |
+
"loss": 0.015,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 40.0,
|
| 875 |
+
"grad_norm": 0.11834592372179031,
|
| 876 |
+
"learning_rate": 2.478e-05,
|
| 877 |
+
"loss": 0.0158,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 40.32258064516129,
|
| 882 |
+
"grad_norm": 0.17337927222251892,
|
| 883 |
+
"learning_rate": 2.498e-05,
|
| 884 |
+
"loss": 0.0148,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 40.645161290322584,
|
| 889 |
+
"grad_norm": 0.17919541895389557,
|
| 890 |
+
"learning_rate": 2.5180000000000003e-05,
|
| 891 |
+
"loss": 0.0146,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 40.96774193548387,
|
| 896 |
+
"grad_norm": 0.1446114331483841,
|
| 897 |
+
"learning_rate": 2.5380000000000004e-05,
|
| 898 |
+
"loss": 0.0144,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 41.29032258064516,
|
| 903 |
+
"grad_norm": 0.12958332896232605,
|
| 904 |
+
"learning_rate": 2.5580000000000002e-05,
|
| 905 |
+
"loss": 0.0143,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 41.61290322580645,
|
| 910 |
+
"grad_norm": 0.208790123462677,
|
| 911 |
+
"learning_rate": 2.5779999999999997e-05,
|
| 912 |
+
"loss": 0.0145,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 41.935483870967744,
|
| 917 |
+
"grad_norm": 0.13380534946918488,
|
| 918 |
+
"learning_rate": 2.598e-05,
|
| 919 |
+
"loss": 0.0144,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 42.25806451612903,
|
| 924 |
+
"grad_norm": 0.202678382396698,
|
| 925 |
+
"learning_rate": 2.618e-05,
|
| 926 |
+
"loss": 0.0137,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 42.58064516129032,
|
| 931 |
+
"grad_norm": 0.19495125114917755,
|
| 932 |
+
"learning_rate": 2.6379999999999998e-05,
|
| 933 |
+
"loss": 0.0147,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 42.903225806451616,
|
| 938 |
+
"grad_norm": 0.17326615750789642,
|
| 939 |
+
"learning_rate": 2.658e-05,
|
| 940 |
+
"loss": 0.0143,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 43.225806451612904,
|
| 945 |
+
"grad_norm": 0.17888951301574707,
|
| 946 |
+
"learning_rate": 2.678e-05,
|
| 947 |
+
"loss": 0.0143,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 43.54838709677419,
|
| 952 |
+
"grad_norm": 0.1571645885705948,
|
| 953 |
+
"learning_rate": 2.698e-05,
|
| 954 |
+
"loss": 0.0136,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 43.87096774193548,
|
| 959 |
+
"grad_norm": 0.13882234692573547,
|
| 960 |
+
"learning_rate": 2.718e-05,
|
| 961 |
+
"loss": 0.0136,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 44.193548387096776,
|
| 966 |
+
"grad_norm": 0.12413440644741058,
|
| 967 |
+
"learning_rate": 2.738e-05,
|
| 968 |
+
"loss": 0.0132,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 44.516129032258064,
|
| 973 |
+
"grad_norm": 0.13164107501506805,
|
| 974 |
+
"learning_rate": 2.758e-05,
|
| 975 |
+
"loss": 0.0138,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 44.83870967741935,
|
| 980 |
+
"grad_norm": 0.14705562591552734,
|
| 981 |
+
"learning_rate": 2.778e-05,
|
| 982 |
+
"loss": 0.0135,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 45.16129032258065,
|
| 987 |
+
"grad_norm": 0.17040109634399414,
|
| 988 |
+
"learning_rate": 2.798e-05,
|
| 989 |
+
"loss": 0.0136,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 45.483870967741936,
|
| 994 |
+
"grad_norm": 0.17892764508724213,
|
| 995 |
+
"learning_rate": 2.818e-05,
|
| 996 |
+
"loss": 0.0135,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 45.806451612903224,
|
| 1001 |
+
"grad_norm": 0.14825959503650665,
|
| 1002 |
+
"learning_rate": 2.8380000000000003e-05,
|
| 1003 |
+
"loss": 0.013,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 46.12903225806452,
|
| 1008 |
+
"grad_norm": 0.11769016087055206,
|
| 1009 |
+
"learning_rate": 2.858e-05,
|
| 1010 |
+
"loss": 0.0134,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 46.45161290322581,
|
| 1015 |
+
"grad_norm": 0.16785204410552979,
|
| 1016 |
+
"learning_rate": 2.8780000000000002e-05,
|
| 1017 |
+
"loss": 0.0131,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 46.774193548387096,
|
| 1022 |
+
"grad_norm": 0.12980397045612335,
|
| 1023 |
+
"learning_rate": 2.898e-05,
|
| 1024 |
+
"loss": 0.0141,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 47.096774193548384,
|
| 1029 |
+
"grad_norm": 0.19282320141792297,
|
| 1030 |
+
"learning_rate": 2.9180000000000002e-05,
|
| 1031 |
+
"loss": 0.014,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 47.41935483870968,
|
| 1036 |
+
"grad_norm": 0.17180287837982178,
|
| 1037 |
+
"learning_rate": 2.9380000000000003e-05,
|
| 1038 |
+
"loss": 0.0133,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 47.74193548387097,
|
| 1043 |
+
"grad_norm": 0.1480427086353302,
|
| 1044 |
+
"learning_rate": 2.958e-05,
|
| 1045 |
+
"loss": 0.0133,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 48.064516129032256,
|
| 1050 |
+
"grad_norm": 0.18842412531375885,
|
| 1051 |
+
"learning_rate": 2.9780000000000003e-05,
|
| 1052 |
+
"loss": 0.0135,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 48.38709677419355,
|
| 1057 |
+
"grad_norm": 0.2088242620229721,
|
| 1058 |
+
"learning_rate": 2.998e-05,
|
| 1059 |
+
"loss": 0.0142,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 48.70967741935484,
|
| 1064 |
+
"grad_norm": 0.1986250877380371,
|
| 1065 |
+
"learning_rate": 3.0180000000000002e-05,
|
| 1066 |
+
"loss": 0.0128,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 49.03225806451613,
|
| 1071 |
+
"grad_norm": 0.1687200963497162,
|
| 1072 |
+
"learning_rate": 3.0380000000000004e-05,
|
| 1073 |
+
"loss": 0.0127,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 49.354838709677416,
|
| 1078 |
+
"grad_norm": 0.1558302789926529,
|
| 1079 |
+
"learning_rate": 3.058e-05,
|
| 1080 |
+
"loss": 0.0131,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 49.67741935483871,
|
| 1085 |
+
"grad_norm": 0.17784525454044342,
|
| 1086 |
+
"learning_rate": 3.078e-05,
|
| 1087 |
+
"loss": 0.0125,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 50.0,
|
| 1092 |
+
"grad_norm": 0.22291983664035797,
|
| 1093 |
+
"learning_rate": 3.0980000000000005e-05,
|
| 1094 |
+
"loss": 0.0121,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 50.32258064516129,
|
| 1099 |
+
"grad_norm": 0.21824175119400024,
|
| 1100 |
+
"learning_rate": 3.118e-05,
|
| 1101 |
+
"loss": 0.012,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 50.645161290322584,
|
| 1106 |
+
"grad_norm": 0.2027992457151413,
|
| 1107 |
+
"learning_rate": 3.138e-05,
|
| 1108 |
+
"loss": 0.0127,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 50.96774193548387,
|
| 1113 |
+
"grad_norm": 0.2533424496650696,
|
| 1114 |
+
"learning_rate": 3.1580000000000006e-05,
|
| 1115 |
+
"loss": 0.0121,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 51.29032258064516,
|
| 1120 |
+
"grad_norm": 0.15423865616321564,
|
| 1121 |
+
"learning_rate": 3.1780000000000004e-05,
|
| 1122 |
+
"loss": 0.0125,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 51.61290322580645,
|
| 1127 |
+
"grad_norm": 0.13885457813739777,
|
| 1128 |
+
"learning_rate": 3.198e-05,
|
| 1129 |
+
"loss": 0.013,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 51.935483870967744,
|
| 1134 |
+
"grad_norm": 0.1417354941368103,
|
| 1135 |
+
"learning_rate": 3.218e-05,
|
| 1136 |
+
"loss": 0.0119,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 52.25806451612903,
|
| 1141 |
+
"grad_norm": 0.17335128784179688,
|
| 1142 |
+
"learning_rate": 3.238e-05,
|
| 1143 |
+
"loss": 0.012,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 52.58064516129032,
|
| 1148 |
+
"grad_norm": 0.19661465287208557,
|
| 1149 |
+
"learning_rate": 3.2579999999999996e-05,
|
| 1150 |
+
"loss": 0.0119,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 52.903225806451616,
|
| 1155 |
+
"grad_norm": 0.2602691650390625,
|
| 1156 |
+
"learning_rate": 3.278e-05,
|
| 1157 |
+
"loss": 0.0122,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 53.225806451612904,
|
| 1162 |
+
"grad_norm": 0.21529579162597656,
|
| 1163 |
+
"learning_rate": 3.298e-05,
|
| 1164 |
+
"loss": 0.0117,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 53.54838709677419,
|
| 1169 |
+
"grad_norm": 0.1529514640569687,
|
| 1170 |
+
"learning_rate": 3.318e-05,
|
| 1171 |
+
"loss": 0.0116,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 53.87096774193548,
|
| 1176 |
+
"grad_norm": 0.1834433674812317,
|
| 1177 |
+
"learning_rate": 3.338e-05,
|
| 1178 |
+
"loss": 0.0125,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 54.193548387096776,
|
| 1183 |
+
"grad_norm": 0.24152186512947083,
|
| 1184 |
+
"learning_rate": 3.358e-05,
|
| 1185 |
+
"loss": 0.0122,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 54.516129032258064,
|
| 1190 |
+
"grad_norm": 0.1319815069437027,
|
| 1191 |
+
"learning_rate": 3.378e-05,
|
| 1192 |
+
"loss": 0.0116,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 54.83870967741935,
|
| 1197 |
+
"grad_norm": 0.16622133553028107,
|
| 1198 |
+
"learning_rate": 3.398e-05,
|
| 1199 |
+
"loss": 0.0123,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 55.16129032258065,
|
| 1204 |
+
"grad_norm": 0.12881875038146973,
|
| 1205 |
+
"learning_rate": 3.418e-05,
|
| 1206 |
+
"loss": 0.0115,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 55.483870967741936,
|
| 1211 |
+
"grad_norm": 0.16083110868930817,
|
| 1212 |
+
"learning_rate": 3.438e-05,
|
| 1213 |
+
"loss": 0.0121,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 55.806451612903224,
|
| 1218 |
+
"grad_norm": 0.17213866114616394,
|
| 1219 |
+
"learning_rate": 3.4580000000000004e-05,
|
| 1220 |
+
"loss": 0.0111,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 56.12903225806452,
|
| 1225 |
+
"grad_norm": 0.1291566640138626,
|
| 1226 |
+
"learning_rate": 3.478e-05,
|
| 1227 |
+
"loss": 0.0113,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 56.45161290322581,
|
| 1232 |
+
"grad_norm": 0.1637270748615265,
|
| 1233 |
+
"learning_rate": 3.498e-05,
|
| 1234 |
+
"loss": 0.011,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 56.774193548387096,
|
| 1239 |
+
"grad_norm": 0.1748403012752533,
|
| 1240 |
+
"learning_rate": 3.518e-05,
|
| 1241 |
+
"loss": 0.0108,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 57.096774193548384,
|
| 1246 |
+
"grad_norm": 0.18834678828716278,
|
| 1247 |
+
"learning_rate": 3.5380000000000003e-05,
|
| 1248 |
+
"loss": 0.0109,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 57.41935483870968,
|
| 1253 |
+
"grad_norm": 0.1650463044643402,
|
| 1254 |
+
"learning_rate": 3.558e-05,
|
| 1255 |
+
"loss": 0.0115,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 57.74193548387097,
|
| 1260 |
+
"grad_norm": 0.17039620876312256,
|
| 1261 |
+
"learning_rate": 3.578e-05,
|
| 1262 |
+
"loss": 0.0114,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 58.064516129032256,
|
| 1267 |
+
"grad_norm": 0.13908374309539795,
|
| 1268 |
+
"learning_rate": 3.5980000000000004e-05,
|
| 1269 |
+
"loss": 0.0107,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 58.38709677419355,
|
| 1274 |
+
"grad_norm": 0.1755029857158661,
|
| 1275 |
+
"learning_rate": 3.618e-05,
|
| 1276 |
+
"loss": 0.0104,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 58.70967741935484,
|
| 1281 |
+
"grad_norm": 0.16456303000450134,
|
| 1282 |
+
"learning_rate": 3.638e-05,
|
| 1283 |
+
"loss": 0.011,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 59.03225806451613,
|
| 1288 |
+
"grad_norm": 0.1215883269906044,
|
| 1289 |
+
"learning_rate": 3.6580000000000006e-05,
|
| 1290 |
+
"loss": 0.0106,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 59.354838709677416,
|
| 1295 |
+
"grad_norm": 0.13105925917625427,
|
| 1296 |
+
"learning_rate": 3.6780000000000004e-05,
|
| 1297 |
+
"loss": 0.0104,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 59.67741935483871,
|
| 1302 |
+
"grad_norm": 0.14026036858558655,
|
| 1303 |
+
"learning_rate": 3.698e-05,
|
| 1304 |
+
"loss": 0.0105,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 60.0,
|
| 1309 |
+
"grad_norm": 0.15687642991542816,
|
| 1310 |
+
"learning_rate": 3.7180000000000007e-05,
|
| 1311 |
+
"loss": 0.0105,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 60.32258064516129,
|
| 1316 |
+
"grad_norm": 0.1946769654750824,
|
| 1317 |
+
"learning_rate": 3.7380000000000005e-05,
|
| 1318 |
+
"loss": 0.011,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 60.645161290322584,
|
| 1323 |
+
"grad_norm": 0.2352798581123352,
|
| 1324 |
+
"learning_rate": 3.758e-05,
|
| 1325 |
+
"loss": 0.0108,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 60.96774193548387,
|
| 1330 |
+
"grad_norm": 0.2311507761478424,
|
| 1331 |
+
"learning_rate": 3.778000000000001e-05,
|
| 1332 |
+
"loss": 0.0119,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 61.29032258064516,
|
| 1337 |
+
"grad_norm": 0.24302849173545837,
|
| 1338 |
+
"learning_rate": 3.7980000000000006e-05,
|
| 1339 |
+
"loss": 0.0104,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 61.61290322580645,
|
| 1344 |
+
"grad_norm": 0.20787030458450317,
|
| 1345 |
+
"learning_rate": 3.818e-05,
|
| 1346 |
+
"loss": 0.011,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 61.935483870967744,
|
| 1351 |
+
"grad_norm": 0.21964043378829956,
|
| 1352 |
+
"learning_rate": 3.838e-05,
|
| 1353 |
+
"loss": 0.0105,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 62.25806451612903,
|
| 1358 |
+
"grad_norm": 0.17259149253368378,
|
| 1359 |
+
"learning_rate": 3.858e-05,
|
| 1360 |
+
"loss": 0.011,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 62.58064516129032,
|
| 1365 |
+
"grad_norm": 0.2735665440559387,
|
| 1366 |
+
"learning_rate": 3.878e-05,
|
| 1367 |
+
"loss": 0.0108,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 62.903225806451616,
|
| 1372 |
+
"grad_norm": 0.19282570481300354,
|
| 1373 |
+
"learning_rate": 3.898e-05,
|
| 1374 |
+
"loss": 0.0108,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 63.225806451612904,
|
| 1379 |
+
"grad_norm": 0.21735602617263794,
|
| 1380 |
+
"learning_rate": 3.918e-05,
|
| 1381 |
+
"loss": 0.0098,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 63.54838709677419,
|
| 1386 |
+
"grad_norm": 0.192997008562088,
|
| 1387 |
+
"learning_rate": 3.938e-05,
|
| 1388 |
+
"loss": 0.0101,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 63.87096774193548,
|
| 1393 |
+
"grad_norm": 0.21661442518234253,
|
| 1394 |
+
"learning_rate": 3.958e-05,
|
| 1395 |
+
"loss": 0.0107,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 64.19354838709677,
|
| 1400 |
+
"grad_norm": 0.23013737797737122,
|
| 1401 |
+
"learning_rate": 3.978e-05,
|
| 1402 |
+
"loss": 0.0106,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 64.51612903225806,
|
| 1407 |
+
"grad_norm": 0.17058345675468445,
|
| 1408 |
+
"learning_rate": 3.998e-05,
|
| 1409 |
+
"loss": 0.0101,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
}
|
| 1412 |
+
],
|
| 1413 |
+
"logging_steps": 10,
|
| 1414 |
+
"max_steps": 100000,
|
| 1415 |
+
"num_input_tokens_seen": 0,
|
| 1416 |
+
"num_train_epochs": 3226,
|
| 1417 |
+
"save_steps": 2000,
|
| 1418 |
+
"stateful_callbacks": {
|
| 1419 |
+
"TrainerControl": {
|
| 1420 |
+
"args": {
|
| 1421 |
+
"should_epoch_stop": false,
|
| 1422 |
+
"should_evaluate": false,
|
| 1423 |
+
"should_log": false,
|
| 1424 |
+
"should_save": true,
|
| 1425 |
+
"should_training_stop": false
|
| 1426 |
+
},
|
| 1427 |
+
"attributes": {}
|
| 1428 |
+
}
|
| 1429 |
+
},
|
| 1430 |
+
"total_flos": 0.0,
|
| 1431 |
+
"train_batch_size": 512,
|
| 1432 |
+
"trial_name": null,
|
| 1433 |
+
"trial_params": null
|
| 1434 |
+
}
|