Add files using upload-large-folder tool
Browse files- config.json +64 -0
- experiment_cfg/metadata.json +307 -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 +1,307 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"new_embodiment": {
|
| 3 |
+
"statistics": {
|
| 4 |
+
"state": {
|
| 5 |
+
"left_arm": {
|
| 6 |
+
"max": [
|
| 7 |
+
1.9294003248214722,
|
| 8 |
+
1.602099061012268,
|
| 9 |
+
1.6264151334762573,
|
| 10 |
+
-0.25446105003356934,
|
| 11 |
+
1.5196356773376465,
|
| 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 |
+
-0.9786677360534668,
|
| 23 |
+
-1.5404508113861084,
|
| 24 |
+
-0.0014851351734250784
|
| 25 |
+
],
|
| 26 |
+
"mean": [
|
| 27 |
+
0.593299925327301,
|
| 28 |
+
0.32747721672058105,
|
| 29 |
+
-0.05066496506333351,
|
| 30 |
+
-1.9063727855682373,
|
| 31 |
+
0.17272311449050903,
|
| 32 |
+
-0.23221395909786224,
|
| 33 |
+
-0.10456155240535736,
|
| 34 |
+
0.31981971859931946
|
| 35 |
+
],
|
| 36 |
+
"std": [
|
| 37 |
+
0.5220098495483398,
|
| 38 |
+
0.37464016675949097,
|
| 39 |
+
0.35901498794555664,
|
| 40 |
+
0.2830471694469452,
|
| 41 |
+
0.3904763460159302,
|
| 42 |
+
0.4544326961040497,
|
| 43 |
+
0.24206174910068512,
|
| 44 |
+
0.42259541153907776
|
| 45 |
+
],
|
| 46 |
+
"q01": [
|
| 47 |
+
-0.6101756036281586,
|
| 48 |
+
0.0,
|
| 49 |
+
-0.9561625659465789,
|
| 50 |
+
-2.369952440261841,
|
| 51 |
+
-1.06134934425354,
|
| 52 |
+
-0.8941370129585267,
|
| 53 |
+
-0.9334145128726958,
|
| 54 |
+
0.0
|
| 55 |
+
],
|
| 56 |
+
"q99": [
|
| 57 |
+
1.5445568442344675,
|
| 58 |
+
1.3273332118988044,
|
| 59 |
+
1.254622507095338,
|
| 60 |
+
-0.7825890600681288,
|
| 61 |
+
1.1786281228065492,
|
| 62 |
+
0.8640626072883606,
|
| 63 |
+
0.27465743422508326,
|
| 64 |
+
1.0960297584533691
|
| 65 |
+
]
|
| 66 |
+
},
|
| 67 |
+
"right_arm": {
|
| 68 |
+
"max": [
|
| 69 |
+
1.8287568092346191,
|
| 70 |
+
0.0,
|
| 71 |
+
1.3382184505462646,
|
| 72 |
+
-0.22242721915245056,
|
| 73 |
+
0.6614453196525574,
|
| 74 |
+
0.851934552192688,
|
| 75 |
+
0.7879786491394043,
|
| 76 |
+
1.0871188640594482
|
| 77 |
+
],
|
| 78 |
+
"min": [
|
| 79 |
+
-0.9815319776535034,
|
| 80 |
+
-1.6567351818084717,
|
| 81 |
+
-1.280957818031311,
|
| 82 |
+
-2.376268148422241,
|
| 83 |
+
-1.5769321918487549,
|
| 84 |
+
-1.1178884506225586,
|
| 85 |
+
-0.38204869627952576,
|
| 86 |
+
-0.0014851351734250784
|
| 87 |
+
],
|
| 88 |
+
"mean": [
|
| 89 |
+
0.7194070816040039,
|
| 90 |
+
-0.4011678397655487,
|
| 91 |
+
-0.04745709151029587,
|
| 92 |
+
-1.989113688468933,
|
| 93 |
+
-0.16602793335914612,
|
| 94 |
+
-0.24818535149097443,
|
| 95 |
+
0.047956615686416626,
|
| 96 |
+
0.3058697283267975
|
| 97 |
+
],
|
| 98 |
+
"std": [
|
| 99 |
+
0.5864532589912415,
|
| 100 |
+
0.30040380358695984,
|
| 101 |
+
0.3263898193836212,
|
| 102 |
+
0.2959355115890503,
|
| 103 |
+
0.32285091280937195,
|
| 104 |
+
0.43125075101852417,
|
| 105 |
+
0.18399831652641296,
|
| 106 |
+
0.427061527967453
|
| 107 |
+
],
|
| 108 |
+
"q01": [
|
| 109 |
+
-0.5737208127975464,
|
| 110 |
+
-1.260634982585907,
|
| 111 |
+
-1.029768466949463,
|
| 112 |
+
-2.3731162548065186,
|
| 113 |
+
-1.2456367015838623,
|
| 114 |
+
-0.8758910298347473,
|
| 115 |
+
-0.27340284287929534,
|
| 116 |
+
0.0
|
| 117 |
+
],
|
| 118 |
+
"q99": [
|
| 119 |
+
1.4089793443679814,
|
| 120 |
+
0.0,
|
| 121 |
+
0.6310521304607393,
|
| 122 |
+
-0.8161724567413328,
|
| 123 |
+
0.36215848326683053,
|
| 124 |
+
0.7335819780826569,
|
| 125 |
+
0.5693755149841309,
|
| 126 |
+
1.078208088874817
|
| 127 |
+
]
|
| 128 |
+
}
|
| 129 |
+
},
|
| 130 |
+
"action": {
|
| 131 |
+
"left_arm": {
|
| 132 |
+
"max": [
|
| 133 |
+
1.932225227355957,
|
| 134 |
+
1.6037930250167847,
|
| 135 |
+
1.6268140077590942,
|
| 136 |
+
-0.2546408176422119,
|
| 137 |
+
1.5193828344345093,
|
| 138 |
+
1.1449086666107178,
|
| 139 |
+
0.6062183976173401,
|
| 140 |
+
1.217768669128418
|
| 141 |
+
],
|
| 142 |
+
"min": [
|
| 143 |
+
-1.1152039766311646,
|
| 144 |
+
0.0,
|
| 145 |
+
-1.3621749877929688,
|
| 146 |
+
-2.3715343475341797,
|
| 147 |
+
-1.5892040729522705,
|
| 148 |
+
-0.9786797165870667,
|
| 149 |
+
-1.5401166677474976,
|
| 150 |
+
-0.45471855998039246
|
| 151 |
+
],
|
| 152 |
+
"mean": [
|
| 153 |
+
0.5945148468017578,
|
| 154 |
+
0.3266119956970215,
|
| 155 |
+
-0.0502079613506794,
|
| 156 |
+
-1.9065903425216675,
|
| 157 |
+
0.17195096611976624,
|
| 158 |
+
-0.2329772710800171,
|
| 159 |
+
-0.10447312891483307,
|
| 160 |
+
0.31009721755981445
|
| 161 |
+
],
|
| 162 |
+
"std": [
|
| 163 |
+
0.5217796564102173,
|
| 164 |
+
0.3746543228626251,
|
| 165 |
+
0.35901448130607605,
|
| 166 |
+
0.2825615406036377,
|
| 167 |
+
0.3902038037776947,
|
| 168 |
+
0.4548611342906952,
|
| 169 |
+
0.24220769107341766,
|
| 170 |
+
0.4963742792606354
|
| 171 |
+
],
|
| 172 |
+
"q01": [
|
| 173 |
+
-0.6089903712272644,
|
| 174 |
+
0.0,
|
| 175 |
+
-0.9526020884513855,
|
| 176 |
+
-2.3700003623962402,
|
| 177 |
+
-1.0650429010391234,
|
| 178 |
+
-0.8943107724189758,
|
| 179 |
+
-0.9341943264007568,
|
| 180 |
+
-0.34887388348579407
|
| 181 |
+
],
|
| 182 |
+
"q99": [
|
| 183 |
+
1.5470080375671387,
|
| 184 |
+
1.3290762901306152,
|
| 185 |
+
1.2554092407226562,
|
| 186 |
+
-0.7900000810623169,
|
| 187 |
+
1.177138090133667,
|
| 188 |
+
0.8640530705451965,
|
| 189 |
+
0.2762514352798462,
|
| 190 |
+
1.187074065208435
|
| 191 |
+
]
|
| 192 |
+
},
|
| 193 |
+
"right_arm": {
|
| 194 |
+
"max": [
|
| 195 |
+
1.8293983936309814,
|
| 196 |
+
0.0,
|
| 197 |
+
1.3382847309112549,
|
| 198 |
+
-0.22242721915245056,
|
| 199 |
+
0.6614686846733093,
|
| 200 |
+
0.8517752289772034,
|
| 201 |
+
0.7873165607452393,
|
| 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 |
+
-0.38196122646331787,
|
| 212 |
+
-0.20314569771289825
|
| 213 |
+
],
|
| 214 |
+
"mean": [
|
| 215 |
+
0.7209159731864929,
|
| 216 |
+
-0.4004747271537781,
|
| 217 |
+
-0.047843363136053085,
|
| 218 |
+
-1.9898872375488281,
|
| 219 |
+
-0.16549433767795563,
|
| 220 |
+
-0.24894820153713226,
|
| 221 |
+
0.048004526644945145,
|
| 222 |
+
0.3641880452632904
|
| 223 |
+
],
|
| 224 |
+
"std": [
|
| 225 |
+
0.5858996510505676,
|
| 226 |
+
0.3006748855113983,
|
| 227 |
+
0.32654568552970886,
|
| 228 |
+
0.2949144244194031,
|
| 229 |
+
0.3225865364074707,
|
| 230 |
+
0.43120241165161133,
|
| 231 |
+
0.18383744359016418,
|
| 232 |
+
0.534224271774292
|
| 233 |
+
],
|
| 234 |
+
"q01": [
|
| 235 |
+
-0.5737088322639465,
|
| 236 |
+
-1.2640001773834229,
|
| 237 |
+
-1.0308350324630737,
|
| 238 |
+
-2.373068332672119,
|
| 239 |
+
-1.2455923557281494,
|
| 240 |
+
-0.8759030103683472,
|
| 241 |
+
-0.2730485796928406,
|
| 242 |
+
-0.0788932517170906
|
| 243 |
+
],
|
| 244 |
+
"q99": [
|
| 245 |
+
1.410417079925537,
|
| 246 |
+
0.0,
|
| 247 |
+
0.6307740807533264,
|
| 248 |
+
-0.8191457390785217,
|
| 249 |
+
0.3637310564517975,
|
| 250 |
+
0.7336010336875916,
|
| 251 |
+
0.5693848729133606,
|
| 252 |
+
1.3221303224563599
|
| 253 |
+
]
|
| 254 |
+
}
|
| 255 |
+
}
|
| 256 |
+
},
|
| 257 |
+
"modalities": {
|
| 258 |
+
"video": {
|
| 259 |
+
"cam_head": {
|
| 260 |
+
"resolution": [
|
| 261 |
+
672,
|
| 262 |
+
376
|
| 263 |
+
],
|
| 264 |
+
"channels": 3,
|
| 265 |
+
"fps": 10.0
|
| 266 |
+
}
|
| 267 |
+
},
|
| 268 |
+
"state": {
|
| 269 |
+
"left_arm": {
|
| 270 |
+
"absolute": true,
|
| 271 |
+
"rotation_type": null,
|
| 272 |
+
"shape": [
|
| 273 |
+
8
|
| 274 |
+
],
|
| 275 |
+
"continuous": true
|
| 276 |
+
},
|
| 277 |
+
"right_arm": {
|
| 278 |
+
"absolute": true,
|
| 279 |
+
"rotation_type": null,
|
| 280 |
+
"shape": [
|
| 281 |
+
8
|
| 282 |
+
],
|
| 283 |
+
"continuous": true
|
| 284 |
+
}
|
| 285 |
+
},
|
| 286 |
+
"action": {
|
| 287 |
+
"left_arm": {
|
| 288 |
+
"absolute": true,
|
| 289 |
+
"rotation_type": null,
|
| 290 |
+
"shape": [
|
| 291 |
+
8
|
| 292 |
+
],
|
| 293 |
+
"continuous": true
|
| 294 |
+
},
|
| 295 |
+
"right_arm": {
|
| 296 |
+
"absolute": true,
|
| 297 |
+
"rotation_type": null,
|
| 298 |
+
"shape": [
|
| 299 |
+
8
|
| 300 |
+
],
|
| 301 |
+
"continuous": true
|
| 302 |
+
}
|
| 303 |
+
}
|
| 304 |
+
},
|
| 305 |
+
"embodiment_tag": "new_embodiment"
|
| 306 |
+
}
|
| 307 |
+
}
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8dc9fc08499c881bbb0b95233a234222b3584d6d74f92e7787d6b3ef00b479f
|
| 3 |
+
size 4999367032
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72063132af90c9d887d4efd2f6323b8e725abdbf501b33e20adfc7829075e512
|
| 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": 15.748031496062993,
|
| 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": 0.07874015748031496,
|
| 14 |
+
"grad_norm": 5.119835376739502,
|
| 15 |
+
"learning_rate": 1.8e-06,
|
| 16 |
+
"loss": 0.8008,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.15748031496062992,
|
| 21 |
+
"grad_norm": 2.7846670150756836,
|
| 22 |
+
"learning_rate": 3.8e-06,
|
| 23 |
+
"loss": 0.6753,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.23622047244094488,
|
| 28 |
+
"grad_norm": 0.9429484605789185,
|
| 29 |
+
"learning_rate": 5.8e-06,
|
| 30 |
+
"loss": 0.3854,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.31496062992125984,
|
| 35 |
+
"grad_norm": 0.4289324879646301,
|
| 36 |
+
"learning_rate": 7.8e-06,
|
| 37 |
+
"loss": 0.2374,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.3937007874015748,
|
| 42 |
+
"grad_norm": 0.7212084531784058,
|
| 43 |
+
"learning_rate": 9.800000000000001e-06,
|
| 44 |
+
"loss": 0.1937,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.47244094488188976,
|
| 49 |
+
"grad_norm": 0.3743739426136017,
|
| 50 |
+
"learning_rate": 1.18e-05,
|
| 51 |
+
"loss": 0.169,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.5511811023622047,
|
| 56 |
+
"grad_norm": 0.33906233310699463,
|
| 57 |
+
"learning_rate": 1.3800000000000002e-05,
|
| 58 |
+
"loss": 0.1513,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.6299212598425197,
|
| 63 |
+
"grad_norm": 0.4938454031944275,
|
| 64 |
+
"learning_rate": 1.58e-05,
|
| 65 |
+
"loss": 0.1333,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.7086614173228346,
|
| 70 |
+
"grad_norm": 0.3791239261627197,
|
| 71 |
+
"learning_rate": 1.78e-05,
|
| 72 |
+
"loss": 0.1209,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.7874015748031497,
|
| 77 |
+
"grad_norm": 0.31115081906318665,
|
| 78 |
+
"learning_rate": 1.9800000000000004e-05,
|
| 79 |
+
"loss": 0.108,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.8661417322834646,
|
| 84 |
+
"grad_norm": 0.3266952633857727,
|
| 85 |
+
"learning_rate": 2.18e-05,
|
| 86 |
+
"loss": 0.105,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.9448818897637795,
|
| 91 |
+
"grad_norm": 0.42559316754341125,
|
| 92 |
+
"learning_rate": 2.38e-05,
|
| 93 |
+
"loss": 0.0931,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.0236220472440944,
|
| 98 |
+
"grad_norm": 0.40235716104507446,
|
| 99 |
+
"learning_rate": 2.58e-05,
|
| 100 |
+
"loss": 0.0867,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.1023622047244095,
|
| 105 |
+
"grad_norm": 0.3455784022808075,
|
| 106 |
+
"learning_rate": 2.7800000000000005e-05,
|
| 107 |
+
"loss": 0.0843,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.1811023622047245,
|
| 112 |
+
"grad_norm": 0.39895790815353394,
|
| 113 |
+
"learning_rate": 2.98e-05,
|
| 114 |
+
"loss": 0.0803,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.2598425196850394,
|
| 119 |
+
"grad_norm": 0.32412493228912354,
|
| 120 |
+
"learning_rate": 3.18e-05,
|
| 121 |
+
"loss": 0.075,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.3385826771653544,
|
| 126 |
+
"grad_norm": 0.3238581717014313,
|
| 127 |
+
"learning_rate": 3.38e-05,
|
| 128 |
+
"loss": 0.069,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.4173228346456692,
|
| 133 |
+
"grad_norm": 0.35713112354278564,
|
| 134 |
+
"learning_rate": 3.58e-05,
|
| 135 |
+
"loss": 0.0632,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.4960629921259843,
|
| 140 |
+
"grad_norm": 0.31078988313674927,
|
| 141 |
+
"learning_rate": 3.7800000000000004e-05,
|
| 142 |
+
"loss": 0.0606,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.574803149606299,
|
| 147 |
+
"grad_norm": 0.23168951272964478,
|
| 148 |
+
"learning_rate": 3.9800000000000005e-05,
|
| 149 |
+
"loss": 0.0566,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.6535433070866141,
|
| 154 |
+
"grad_norm": 0.2528112828731537,
|
| 155 |
+
"learning_rate": 4.18e-05,
|
| 156 |
+
"loss": 0.0553,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.7322834645669292,
|
| 161 |
+
"grad_norm": 0.31132972240448,
|
| 162 |
+
"learning_rate": 4.38e-05,
|
| 163 |
+
"loss": 0.0516,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.811023622047244,
|
| 168 |
+
"grad_norm": 0.3506482243537903,
|
| 169 |
+
"learning_rate": 4.58e-05,
|
| 170 |
+
"loss": 0.051,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.889763779527559,
|
| 175 |
+
"grad_norm": 0.30420321226119995,
|
| 176 |
+
"learning_rate": 4.78e-05,
|
| 177 |
+
"loss": 0.0498,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.968503937007874,
|
| 182 |
+
"grad_norm": 0.27608105540275574,
|
| 183 |
+
"learning_rate": 4.9800000000000004e-05,
|
| 184 |
+
"loss": 0.0425,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 2.047244094488189,
|
| 189 |
+
"grad_norm": 0.24153359234333038,
|
| 190 |
+
"learning_rate": 5.1800000000000005e-05,
|
| 191 |
+
"loss": 0.0425,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 2.1259842519685037,
|
| 196 |
+
"grad_norm": 0.2684983015060425,
|
| 197 |
+
"learning_rate": 5.380000000000001e-05,
|
| 198 |
+
"loss": 0.0397,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 2.204724409448819,
|
| 203 |
+
"grad_norm": 0.2812291979789734,
|
| 204 |
+
"learning_rate": 5.580000000000001e-05,
|
| 205 |
+
"loss": 0.0347,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.283464566929134,
|
| 210 |
+
"grad_norm": 0.272079199552536,
|
| 211 |
+
"learning_rate": 5.7799999999999995e-05,
|
| 212 |
+
"loss": 0.0339,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 2.362204724409449,
|
| 217 |
+
"grad_norm": 0.30601683259010315,
|
| 218 |
+
"learning_rate": 5.9800000000000003e-05,
|
| 219 |
+
"loss": 0.0359,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 2.440944881889764,
|
| 224 |
+
"grad_norm": 0.3129172921180725,
|
| 225 |
+
"learning_rate": 6.18e-05,
|
| 226 |
+
"loss": 0.0351,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 2.5196850393700787,
|
| 231 |
+
"grad_norm": 0.27252131700515747,
|
| 232 |
+
"learning_rate": 6.38e-05,
|
| 233 |
+
"loss": 0.0287,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.5984251968503935,
|
| 238 |
+
"grad_norm": 0.2653070390224457,
|
| 239 |
+
"learning_rate": 6.58e-05,
|
| 240 |
+
"loss": 0.0313,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.677165354330709,
|
| 245 |
+
"grad_norm": 0.35808777809143066,
|
| 246 |
+
"learning_rate": 6.780000000000001e-05,
|
| 247 |
+
"loss": 0.0356,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.7559055118110236,
|
| 252 |
+
"grad_norm": 0.26742085814476013,
|
| 253 |
+
"learning_rate": 6.98e-05,
|
| 254 |
+
"loss": 0.0299,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 2.8346456692913384,
|
| 259 |
+
"grad_norm": 0.4106348156929016,
|
| 260 |
+
"learning_rate": 7.18e-05,
|
| 261 |
+
"loss": 0.0324,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.9133858267716537,
|
| 266 |
+
"grad_norm": 0.213535338640213,
|
| 267 |
+
"learning_rate": 7.38e-05,
|
| 268 |
+
"loss": 0.0273,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.9921259842519685,
|
| 273 |
+
"grad_norm": 0.26808497309684753,
|
| 274 |
+
"learning_rate": 7.58e-05,
|
| 275 |
+
"loss": 0.0254,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 3.0708661417322833,
|
| 280 |
+
"grad_norm": 0.18177832663059235,
|
| 281 |
+
"learning_rate": 7.780000000000001e-05,
|
| 282 |
+
"loss": 0.0261,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 3.1496062992125986,
|
| 287 |
+
"grad_norm": 0.2706851065158844,
|
| 288 |
+
"learning_rate": 7.98e-05,
|
| 289 |
+
"loss": 0.0268,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.2283464566929134,
|
| 294 |
+
"grad_norm": 0.29524528980255127,
|
| 295 |
+
"learning_rate": 8.18e-05,
|
| 296 |
+
"loss": 0.0285,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 3.3070866141732282,
|
| 301 |
+
"grad_norm": 0.21399272978305817,
|
| 302 |
+
"learning_rate": 8.38e-05,
|
| 303 |
+
"loss": 0.0252,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 3.3858267716535435,
|
| 308 |
+
"grad_norm": 0.290097177028656,
|
| 309 |
+
"learning_rate": 8.58e-05,
|
| 310 |
+
"loss": 0.0279,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 3.4645669291338583,
|
| 315 |
+
"grad_norm": 0.2789689004421234,
|
| 316 |
+
"learning_rate": 8.78e-05,
|
| 317 |
+
"loss": 0.0236,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 3.543307086614173,
|
| 322 |
+
"grad_norm": 0.3307545781135559,
|
| 323 |
+
"learning_rate": 8.98e-05,
|
| 324 |
+
"loss": 0.0256,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 3.622047244094488,
|
| 329 |
+
"grad_norm": 0.2919306457042694,
|
| 330 |
+
"learning_rate": 9.180000000000001e-05,
|
| 331 |
+
"loss": 0.0227,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.7007874015748032,
|
| 336 |
+
"grad_norm": 0.27534034848213196,
|
| 337 |
+
"learning_rate": 9.38e-05,
|
| 338 |
+
"loss": 0.0219,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 3.779527559055118,
|
| 343 |
+
"grad_norm": 0.26348116993904114,
|
| 344 |
+
"learning_rate": 9.58e-05,
|
| 345 |
+
"loss": 0.0242,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 3.8582677165354333,
|
| 350 |
+
"grad_norm": 0.29468125104904175,
|
| 351 |
+
"learning_rate": 9.78e-05,
|
| 352 |
+
"loss": 0.0224,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 3.937007874015748,
|
| 357 |
+
"grad_norm": 0.20534993708133698,
|
| 358 |
+
"learning_rate": 9.98e-05,
|
| 359 |
+
"loss": 0.0235,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.015748031496063,
|
| 364 |
+
"grad_norm": 0.2911393642425537,
|
| 365 |
+
"learning_rate": 9.9999778549206e-05,
|
| 366 |
+
"loss": 0.0206,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 4.094488188976378,
|
| 371 |
+
"grad_norm": 0.2478438913822174,
|
| 372 |
+
"learning_rate": 9.999901304280685e-05,
|
| 373 |
+
"loss": 0.0219,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.173228346456693,
|
| 378 |
+
"grad_norm": 0.32605063915252686,
|
| 379 |
+
"learning_rate": 9.999770075521164e-05,
|
| 380 |
+
"loss": 0.0232,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 4.251968503937007,
|
| 385 |
+
"grad_norm": 0.22585000097751617,
|
| 386 |
+
"learning_rate": 9.99958417007713e-05,
|
| 387 |
+
"loss": 0.0228,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.330708661417323,
|
| 392 |
+
"grad_norm": 0.3227289319038391,
|
| 393 |
+
"learning_rate": 9.999343589981615e-05,
|
| 394 |
+
"loss": 0.018,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 4.409448818897638,
|
| 399 |
+
"grad_norm": 0.2523372769355774,
|
| 400 |
+
"learning_rate": 9.999048337865568e-05,
|
| 401 |
+
"loss": 0.0215,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 4.488188976377953,
|
| 406 |
+
"grad_norm": 0.34720173478126526,
|
| 407 |
+
"learning_rate": 9.998698416957815e-05,
|
| 408 |
+
"loss": 0.0242,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.566929133858268,
|
| 413 |
+
"grad_norm": 0.24005654454231262,
|
| 414 |
+
"learning_rate": 9.998293831085037e-05,
|
| 415 |
+
"loss": 0.0213,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.645669291338582,
|
| 420 |
+
"grad_norm": 0.28940242528915405,
|
| 421 |
+
"learning_rate": 9.997834584671719e-05,
|
| 422 |
+
"loss": 0.0204,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 4.724409448818898,
|
| 427 |
+
"grad_norm": 0.2654191255569458,
|
| 428 |
+
"learning_rate": 9.997320682740107e-05,
|
| 429 |
+
"loss": 0.0217,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.803149606299213,
|
| 434 |
+
"grad_norm": 0.2912241816520691,
|
| 435 |
+
"learning_rate": 9.996752130910149e-05,
|
| 436 |
+
"loss": 0.0197,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 4.881889763779528,
|
| 441 |
+
"grad_norm": 0.23718924820423126,
|
| 442 |
+
"learning_rate": 9.99612893539944e-05,
|
| 443 |
+
"loss": 0.0209,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 4.960629921259843,
|
| 448 |
+
"grad_norm": 0.2647818326950073,
|
| 449 |
+
"learning_rate": 9.995451103023144e-05,
|
| 450 |
+
"loss": 0.0222,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 5.039370078740157,
|
| 455 |
+
"grad_norm": 0.35324886441230774,
|
| 456 |
+
"learning_rate": 9.994718641193928e-05,
|
| 457 |
+
"loss": 0.0224,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 5.118110236220472,
|
| 462 |
+
"grad_norm": 0.2671961188316345,
|
| 463 |
+
"learning_rate": 9.993931557921874e-05,
|
| 464 |
+
"loss": 0.0219,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 5.196850393700787,
|
| 469 |
+
"grad_norm": 0.2596529722213745,
|
| 470 |
+
"learning_rate": 9.993089861814402e-05,
|
| 471 |
+
"loss": 0.0203,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 5.275590551181103,
|
| 476 |
+
"grad_norm": 0.25885483622550964,
|
| 477 |
+
"learning_rate": 9.992193562076166e-05,
|
| 478 |
+
"loss": 0.0188,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 5.354330708661418,
|
| 483 |
+
"grad_norm": 0.24976016581058502,
|
| 484 |
+
"learning_rate": 9.991242668508954e-05,
|
| 485 |
+
"loss": 0.0175,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 5.433070866141732,
|
| 490 |
+
"grad_norm": 0.24121227860450745,
|
| 491 |
+
"learning_rate": 9.990237191511587e-05,
|
| 492 |
+
"loss": 0.0158,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 5.511811023622047,
|
| 497 |
+
"grad_norm": 0.22227917611598969,
|
| 498 |
+
"learning_rate": 9.989177142079802e-05,
|
| 499 |
+
"loss": 0.0177,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 5.590551181102362,
|
| 504 |
+
"grad_norm": 0.231464222073555,
|
| 505 |
+
"learning_rate": 9.988062531806126e-05,
|
| 506 |
+
"loss": 0.0183,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 5.669291338582677,
|
| 511 |
+
"grad_norm": 0.16609017550945282,
|
| 512 |
+
"learning_rate": 9.986893372879762e-05,
|
| 513 |
+
"loss": 0.018,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 5.748031496062993,
|
| 518 |
+
"grad_norm": 0.19624024629592896,
|
| 519 |
+
"learning_rate": 9.985669678086443e-05,
|
| 520 |
+
"loss": 0.018,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 5.826771653543307,
|
| 525 |
+
"grad_norm": 0.22255055606365204,
|
| 526 |
+
"learning_rate": 9.984391460808298e-05,
|
| 527 |
+
"loss": 0.0199,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 5.905511811023622,
|
| 532 |
+
"grad_norm": 0.22765639424324036,
|
| 533 |
+
"learning_rate": 9.983058735023709e-05,
|
| 534 |
+
"loss": 0.0191,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 5.984251968503937,
|
| 539 |
+
"grad_norm": 0.23915418982505798,
|
| 540 |
+
"learning_rate": 9.98167151530715e-05,
|
| 541 |
+
"loss": 0.0178,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 6.062992125984252,
|
| 546 |
+
"grad_norm": 0.2489311248064041,
|
| 547 |
+
"learning_rate": 9.980229816829034e-05,
|
| 548 |
+
"loss": 0.0202,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 6.141732283464567,
|
| 553 |
+
"grad_norm": 0.22865547239780426,
|
| 554 |
+
"learning_rate": 9.978733655355544e-05,
|
| 555 |
+
"loss": 0.0187,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 6.2204724409448815,
|
| 560 |
+
"grad_norm": 0.19393905997276306,
|
| 561 |
+
"learning_rate": 9.977183047248464e-05,
|
| 562 |
+
"loss": 0.0168,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 6.299212598425197,
|
| 567 |
+
"grad_norm": 0.20525363087654114,
|
| 568 |
+
"learning_rate": 9.975578009464992e-05,
|
| 569 |
+
"loss": 0.018,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 6.377952755905512,
|
| 574 |
+
"grad_norm": 0.2537108063697815,
|
| 575 |
+
"learning_rate": 9.97391855955757e-05,
|
| 576 |
+
"loss": 0.0143,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 6.456692913385827,
|
| 581 |
+
"grad_norm": 0.2665018141269684,
|
| 582 |
+
"learning_rate": 9.972204715673669e-05,
|
| 583 |
+
"loss": 0.0165,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 6.535433070866142,
|
| 588 |
+
"grad_norm": 0.18383699655532837,
|
| 589 |
+
"learning_rate": 9.970436496555617e-05,
|
| 590 |
+
"loss": 0.0164,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 6.6141732283464565,
|
| 595 |
+
"grad_norm": 0.3430931270122528,
|
| 596 |
+
"learning_rate": 9.968613921540373e-05,
|
| 597 |
+
"loss": 0.0176,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 6.692913385826771,
|
| 602 |
+
"grad_norm": 0.2601425349712372,
|
| 603 |
+
"learning_rate": 9.966737010559326e-05,
|
| 604 |
+
"loss": 0.0175,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 6.771653543307087,
|
| 609 |
+
"grad_norm": 0.19988982379436493,
|
| 610 |
+
"learning_rate": 9.964805784138072e-05,
|
| 611 |
+
"loss": 0.0172,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 6.850393700787402,
|
| 616 |
+
"grad_norm": 0.18660953640937805,
|
| 617 |
+
"learning_rate": 9.962820263396195e-05,
|
| 618 |
+
"loss": 0.0158,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 6.929133858267717,
|
| 623 |
+
"grad_norm": 0.22756962478160858,
|
| 624 |
+
"learning_rate": 9.960780470047033e-05,
|
| 625 |
+
"loss": 0.0185,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 7.0078740157480315,
|
| 630 |
+
"grad_norm": 0.14548353850841522,
|
| 631 |
+
"learning_rate": 9.958686426397437e-05,
|
| 632 |
+
"loss": 0.0164,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 7.086614173228346,
|
| 637 |
+
"grad_norm": 0.20737145841121674,
|
| 638 |
+
"learning_rate": 9.956538155347534e-05,
|
| 639 |
+
"loss": 0.0182,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 7.165354330708661,
|
| 644 |
+
"grad_norm": 0.20689648389816284,
|
| 645 |
+
"learning_rate": 9.95433568039047e-05,
|
| 646 |
+
"loss": 0.0145,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 7.244094488188976,
|
| 651 |
+
"grad_norm": 0.26220783591270447,
|
| 652 |
+
"learning_rate": 9.952079025612162e-05,
|
| 653 |
+
"loss": 0.0145,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 7.322834645669292,
|
| 658 |
+
"grad_norm": 0.23523452877998352,
|
| 659 |
+
"learning_rate": 9.949768215691022e-05,
|
| 660 |
+
"loss": 0.0168,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 7.4015748031496065,
|
| 665 |
+
"grad_norm": 0.207063227891922,
|
| 666 |
+
"learning_rate": 9.9474032758977e-05,
|
| 667 |
+
"loss": 0.0154,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 7.480314960629921,
|
| 672 |
+
"grad_norm": 0.2092580646276474,
|
| 673 |
+
"learning_rate": 9.944984232094794e-05,
|
| 674 |
+
"loss": 0.0169,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 7.559055118110236,
|
| 679 |
+
"grad_norm": 0.1808154582977295,
|
| 680 |
+
"learning_rate": 9.942511110736584e-05,
|
| 681 |
+
"loss": 0.0157,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 7.637795275590551,
|
| 686 |
+
"grad_norm": 0.2190985083580017,
|
| 687 |
+
"learning_rate": 9.939983938868726e-05,
|
| 688 |
+
"loss": 0.0155,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 7.716535433070866,
|
| 693 |
+
"grad_norm": 0.1607908308506012,
|
| 694 |
+
"learning_rate": 9.93740274412797e-05,
|
| 695 |
+
"loss": 0.0136,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 7.7952755905511815,
|
| 700 |
+
"grad_norm": 0.20882774889469147,
|
| 701 |
+
"learning_rate": 9.934767554741846e-05,
|
| 702 |
+
"loss": 0.0192,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 7.874015748031496,
|
| 707 |
+
"grad_norm": 0.18141894042491913,
|
| 708 |
+
"learning_rate": 9.932078399528361e-05,
|
| 709 |
+
"loss": 0.0134,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 7.952755905511811,
|
| 714 |
+
"grad_norm": 0.1842644363641739,
|
| 715 |
+
"learning_rate": 9.929335307895689e-05,
|
| 716 |
+
"loss": 0.0145,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 8.031496062992126,
|
| 721 |
+
"grad_norm": 0.19102592766284943,
|
| 722 |
+
"learning_rate": 9.926538309841839e-05,
|
| 723 |
+
"loss": 0.0179,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 8.11023622047244,
|
| 728 |
+
"grad_norm": 0.2554001212120056,
|
| 729 |
+
"learning_rate": 9.923687435954334e-05,
|
| 730 |
+
"loss": 0.0145,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 8.188976377952756,
|
| 735 |
+
"grad_norm": 0.2188219279050827,
|
| 736 |
+
"learning_rate": 9.920782717409873e-05,
|
| 737 |
+
"loss": 0.0133,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 8.26771653543307,
|
| 742 |
+
"grad_norm": 0.19668325781822205,
|
| 743 |
+
"learning_rate": 9.917824185973994e-05,
|
| 744 |
+
"loss": 0.013,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 8.346456692913385,
|
| 749 |
+
"grad_norm": 0.19224300980567932,
|
| 750 |
+
"learning_rate": 9.914811874000723e-05,
|
| 751 |
+
"loss": 0.012,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 8.4251968503937,
|
| 756 |
+
"grad_norm": 0.2617517113685608,
|
| 757 |
+
"learning_rate": 9.911745814432218e-05,
|
| 758 |
+
"loss": 0.0144,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 8.503937007874015,
|
| 763 |
+
"grad_norm": 0.340850293636322,
|
| 764 |
+
"learning_rate": 9.90862604079842e-05,
|
| 765 |
+
"loss": 0.0163,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 8.582677165354331,
|
| 770 |
+
"grad_norm": 0.24036389589309692,
|
| 771 |
+
"learning_rate": 9.90545258721667e-05,
|
| 772 |
+
"loss": 0.0143,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 8.661417322834646,
|
| 777 |
+
"grad_norm": 0.2523621916770935,
|
| 778 |
+
"learning_rate": 9.90222548839135e-05,
|
| 779 |
+
"loss": 0.0137,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 8.740157480314961,
|
| 784 |
+
"grad_norm": 0.25303855538368225,
|
| 785 |
+
"learning_rate": 9.898944779613495e-05,
|
| 786 |
+
"loss": 0.0124,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 8.818897637795276,
|
| 791 |
+
"grad_norm": 0.2672367990016937,
|
| 792 |
+
"learning_rate": 9.89561049676041e-05,
|
| 793 |
+
"loss": 0.0135,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 8.89763779527559,
|
| 798 |
+
"grad_norm": 0.22292408347129822,
|
| 799 |
+
"learning_rate": 9.89222267629528e-05,
|
| 800 |
+
"loss": 0.0155,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 8.976377952755906,
|
| 805 |
+
"grad_norm": 0.2113981992006302,
|
| 806 |
+
"learning_rate": 9.888781355266763e-05,
|
| 807 |
+
"loss": 0.0139,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 9.05511811023622,
|
| 812 |
+
"grad_norm": 0.16752807796001434,
|
| 813 |
+
"learning_rate": 9.885286571308598e-05,
|
| 814 |
+
"loss": 0.0124,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 9.133858267716535,
|
| 819 |
+
"grad_norm": 0.1773703545331955,
|
| 820 |
+
"learning_rate": 9.881738362639182e-05,
|
| 821 |
+
"loss": 0.015,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 9.21259842519685,
|
| 826 |
+
"grad_norm": 0.26974138617515564,
|
| 827 |
+
"learning_rate": 9.878136768061154e-05,
|
| 828 |
+
"loss": 0.0162,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 9.291338582677165,
|
| 833 |
+
"grad_norm": 0.2184063196182251,
|
| 834 |
+
"learning_rate": 9.874481826960979e-05,
|
| 835 |
+
"loss": 0.0148,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 9.37007874015748,
|
| 840 |
+
"grad_norm": 0.1977306753396988,
|
| 841 |
+
"learning_rate": 9.870773579308503e-05,
|
| 842 |
+
"loss": 0.0123,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 9.448818897637794,
|
| 847 |
+
"grad_norm": 0.1981269121170044,
|
| 848 |
+
"learning_rate": 9.867012065656533e-05,
|
| 849 |
+
"loss": 0.0152,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 9.527559055118111,
|
| 854 |
+
"grad_norm": 0.17817805707454681,
|
| 855 |
+
"learning_rate": 9.863197327140376e-05,
|
| 856 |
+
"loss": 0.0123,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 9.606299212598426,
|
| 861 |
+
"grad_norm": 0.23420843482017517,
|
| 862 |
+
"learning_rate": 9.859329405477403e-05,
|
| 863 |
+
"loss": 0.0129,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 9.68503937007874,
|
| 868 |
+
"grad_norm": 0.25216200947761536,
|
| 869 |
+
"learning_rate": 9.855408342966585e-05,
|
| 870 |
+
"loss": 0.0138,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 9.763779527559056,
|
| 875 |
+
"grad_norm": 0.1990588754415512,
|
| 876 |
+
"learning_rate": 9.851434182488033e-05,
|
| 877 |
+
"loss": 0.0129,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 9.84251968503937,
|
| 882 |
+
"grad_norm": 0.27837619185447693,
|
| 883 |
+
"learning_rate": 9.84740696750253e-05,
|
| 884 |
+
"loss": 0.0124,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 9.921259842519685,
|
| 889 |
+
"grad_norm": 0.21090054512023926,
|
| 890 |
+
"learning_rate": 9.843326742051055e-05,
|
| 891 |
+
"loss": 0.013,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 10.0,
|
| 896 |
+
"grad_norm": 0.19581645727157593,
|
| 897 |
+
"learning_rate": 9.839193550754297e-05,
|
| 898 |
+
"loss": 0.0126,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 10.078740157480315,
|
| 903 |
+
"grad_norm": 0.21251627802848816,
|
| 904 |
+
"learning_rate": 9.835007438812177e-05,
|
| 905 |
+
"loss": 0.0148,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 10.15748031496063,
|
| 910 |
+
"grad_norm": 0.18511821329593658,
|
| 911 |
+
"learning_rate": 9.830768452003341e-05,
|
| 912 |
+
"loss": 0.0133,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 10.236220472440944,
|
| 917 |
+
"grad_norm": 0.18811464309692383,
|
| 918 |
+
"learning_rate": 9.826476636684671e-05,
|
| 919 |
+
"loss": 0.0126,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 10.31496062992126,
|
| 924 |
+
"grad_norm": 0.18782231211662292,
|
| 925 |
+
"learning_rate": 9.822132039790773e-05,
|
| 926 |
+
"loss": 0.0117,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 10.393700787401574,
|
| 931 |
+
"grad_norm": 0.16824057698249817,
|
| 932 |
+
"learning_rate": 9.817734708833461e-05,
|
| 933 |
+
"loss": 0.0106,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 10.472440944881889,
|
| 938 |
+
"grad_norm": 0.1814710795879364,
|
| 939 |
+
"learning_rate": 9.813284691901243e-05,
|
| 940 |
+
"loss": 0.0162,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 10.551181102362206,
|
| 945 |
+
"grad_norm": 0.2217687964439392,
|
| 946 |
+
"learning_rate": 9.808782037658792e-05,
|
| 947 |
+
"loss": 0.0155,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 10.62992125984252,
|
| 952 |
+
"grad_norm": 0.19781896471977234,
|
| 953 |
+
"learning_rate": 9.804226795346411e-05,
|
| 954 |
+
"loss": 0.0133,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 10.708661417322835,
|
| 959 |
+
"grad_norm": 0.24714171886444092,
|
| 960 |
+
"learning_rate": 9.799619014779503e-05,
|
| 961 |
+
"loss": 0.0129,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 10.78740157480315,
|
| 966 |
+
"grad_norm": 0.16805458068847656,
|
| 967 |
+
"learning_rate": 9.794958746348013e-05,
|
| 968 |
+
"loss": 0.0125,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 10.866141732283465,
|
| 973 |
+
"grad_norm": 0.18694327771663666,
|
| 974 |
+
"learning_rate": 9.790246041015896e-05,
|
| 975 |
+
"loss": 0.0112,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 10.94488188976378,
|
| 980 |
+
"grad_norm": 0.21768535673618317,
|
| 981 |
+
"learning_rate": 9.785480950320538e-05,
|
| 982 |
+
"loss": 0.0121,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 11.023622047244094,
|
| 987 |
+
"grad_norm": 0.16912485659122467,
|
| 988 |
+
"learning_rate": 9.78066352637221e-05,
|
| 989 |
+
"loss": 0.0109,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 11.10236220472441,
|
| 994 |
+
"grad_norm": 0.15913233160972595,
|
| 995 |
+
"learning_rate": 9.775793821853488e-05,
|
| 996 |
+
"loss": 0.0115,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 11.181102362204724,
|
| 1001 |
+
"grad_norm": 0.15250848233699799,
|
| 1002 |
+
"learning_rate": 9.77087189001868e-05,
|
| 1003 |
+
"loss": 0.0123,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 11.259842519685039,
|
| 1008 |
+
"grad_norm": 0.17317131161689758,
|
| 1009 |
+
"learning_rate": 9.765897784693243e-05,
|
| 1010 |
+
"loss": 0.0117,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 11.338582677165354,
|
| 1015 |
+
"grad_norm": 0.23304998874664307,
|
| 1016 |
+
"learning_rate": 9.760871560273197e-05,
|
| 1017 |
+
"loss": 0.0107,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 11.417322834645669,
|
| 1022 |
+
"grad_norm": 0.2260117381811142,
|
| 1023 |
+
"learning_rate": 9.755793271724526e-05,
|
| 1024 |
+
"loss": 0.0113,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 11.496062992125985,
|
| 1029 |
+
"grad_norm": 0.20854035019874573,
|
| 1030 |
+
"learning_rate": 9.750662974582584e-05,
|
| 1031 |
+
"loss": 0.0156,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 11.5748031496063,
|
| 1036 |
+
"grad_norm": 0.18729598820209503,
|
| 1037 |
+
"learning_rate": 9.745480724951473e-05,
|
| 1038 |
+
"loss": 0.0115,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 11.653543307086615,
|
| 1043 |
+
"grad_norm": 0.1489574909210205,
|
| 1044 |
+
"learning_rate": 9.740246579503447e-05,
|
| 1045 |
+
"loss": 0.0122,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 11.73228346456693,
|
| 1050 |
+
"grad_norm": 0.16865724325180054,
|
| 1051 |
+
"learning_rate": 9.734960595478284e-05,
|
| 1052 |
+
"loss": 0.0121,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 11.811023622047244,
|
| 1057 |
+
"grad_norm": 0.1705121397972107,
|
| 1058 |
+
"learning_rate": 9.729622830682657e-05,
|
| 1059 |
+
"loss": 0.0117,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 11.88976377952756,
|
| 1064 |
+
"grad_norm": 0.12779462337493896,
|
| 1065 |
+
"learning_rate": 9.724233343489504e-05,
|
| 1066 |
+
"loss": 0.013,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 11.968503937007874,
|
| 1071 |
+
"grad_norm": 0.21109400689601898,
|
| 1072 |
+
"learning_rate": 9.718792192837396e-05,
|
| 1073 |
+
"loss": 0.0105,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 12.047244094488189,
|
| 1078 |
+
"grad_norm": 0.17350123822689056,
|
| 1079 |
+
"learning_rate": 9.713299438229886e-05,
|
| 1080 |
+
"loss": 0.0129,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 12.125984251968504,
|
| 1085 |
+
"grad_norm": 0.19555015861988068,
|
| 1086 |
+
"learning_rate": 9.707755139734855e-05,
|
| 1087 |
+
"loss": 0.0131,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 12.204724409448819,
|
| 1092 |
+
"grad_norm": 0.22949132323265076,
|
| 1093 |
+
"learning_rate": 9.702159357983866e-05,
|
| 1094 |
+
"loss": 0.0122,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 12.283464566929133,
|
| 1099 |
+
"grad_norm": 0.21299389004707336,
|
| 1100 |
+
"learning_rate": 9.696512154171492e-05,
|
| 1101 |
+
"loss": 0.013,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 12.362204724409448,
|
| 1106 |
+
"grad_norm": 0.2029636800289154,
|
| 1107 |
+
"learning_rate": 9.690813590054645e-05,
|
| 1108 |
+
"loss": 0.0127,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 12.440944881889763,
|
| 1113 |
+
"grad_norm": 0.2509428858757019,
|
| 1114 |
+
"learning_rate": 9.685063727951914e-05,
|
| 1115 |
+
"loss": 0.0115,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 12.519685039370078,
|
| 1120 |
+
"grad_norm": 0.17952832579612732,
|
| 1121 |
+
"learning_rate": 9.679262630742865e-05,
|
| 1122 |
+
"loss": 0.0123,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 12.598425196850394,
|
| 1127 |
+
"grad_norm": 0.17356553673744202,
|
| 1128 |
+
"learning_rate": 9.673410361867373e-05,
|
| 1129 |
+
"loss": 0.0133,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 12.67716535433071,
|
| 1134 |
+
"grad_norm": 0.2649160921573639,
|
| 1135 |
+
"learning_rate": 9.667506985324909e-05,
|
| 1136 |
+
"loss": 0.0116,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 12.755905511811024,
|
| 1141 |
+
"grad_norm": 0.209790900349617,
|
| 1142 |
+
"learning_rate": 9.661552565673855e-05,
|
| 1143 |
+
"loss": 0.0104,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 12.834645669291339,
|
| 1148 |
+
"grad_norm": 0.21641805768013,
|
| 1149 |
+
"learning_rate": 9.655547168030789e-05,
|
| 1150 |
+
"loss": 0.0129,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 12.913385826771654,
|
| 1155 |
+
"grad_norm": 0.2454116940498352,
|
| 1156 |
+
"learning_rate": 9.649490858069777e-05,
|
| 1157 |
+
"loss": 0.0104,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 12.992125984251969,
|
| 1162 |
+
"grad_norm": 0.21532991528511047,
|
| 1163 |
+
"learning_rate": 9.643383702021658e-05,
|
| 1164 |
+
"loss": 0.0111,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 13.070866141732283,
|
| 1169 |
+
"grad_norm": 0.22105859220027924,
|
| 1170 |
+
"learning_rate": 9.637225766673307e-05,
|
| 1171 |
+
"loss": 0.0098,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 13.149606299212598,
|
| 1176 |
+
"grad_norm": 0.16493800282478333,
|
| 1177 |
+
"learning_rate": 9.631017119366922e-05,
|
| 1178 |
+
"loss": 0.0107,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 13.228346456692913,
|
| 1183 |
+
"grad_norm": 0.18128368258476257,
|
| 1184 |
+
"learning_rate": 9.624757827999273e-05,
|
| 1185 |
+
"loss": 0.0117,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 13.307086614173228,
|
| 1190 |
+
"grad_norm": 0.2412339746952057,
|
| 1191 |
+
"learning_rate": 9.618447961020971e-05,
|
| 1192 |
+
"loss": 0.015,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 13.385826771653543,
|
| 1197 |
+
"grad_norm": 0.23829780519008636,
|
| 1198 |
+
"learning_rate": 9.612087587435707e-05,
|
| 1199 |
+
"loss": 0.0146,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 13.464566929133857,
|
| 1204 |
+
"grad_norm": 0.17435379326343536,
|
| 1205 |
+
"learning_rate": 9.605676776799508e-05,
|
| 1206 |
+
"loss": 0.012,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 13.543307086614174,
|
| 1211 |
+
"grad_norm": 0.26677370071411133,
|
| 1212 |
+
"learning_rate": 9.599215599219973e-05,
|
| 1213 |
+
"loss": 0.0119,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 13.622047244094489,
|
| 1218 |
+
"grad_norm": 0.17352107167243958,
|
| 1219 |
+
"learning_rate": 9.592704125355505e-05,
|
| 1220 |
+
"loss": 0.0119,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 13.700787401574804,
|
| 1225 |
+
"grad_norm": 0.1817910224199295,
|
| 1226 |
+
"learning_rate": 9.586142426414538e-05,
|
| 1227 |
+
"loss": 0.011,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 13.779527559055119,
|
| 1232 |
+
"grad_norm": 0.26779067516326904,
|
| 1233 |
+
"learning_rate": 9.57953057415476e-05,
|
| 1234 |
+
"loss": 0.0137,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 13.858267716535433,
|
| 1239 |
+
"grad_norm": 0.16992807388305664,
|
| 1240 |
+
"learning_rate": 9.572868640882328e-05,
|
| 1241 |
+
"loss": 0.0116,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 13.937007874015748,
|
| 1246 |
+
"grad_norm": 0.2475721836090088,
|
| 1247 |
+
"learning_rate": 9.56615669945108e-05,
|
| 1248 |
+
"loss": 0.013,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 14.015748031496063,
|
| 1253 |
+
"grad_norm": 0.30210572481155396,
|
| 1254 |
+
"learning_rate": 9.55939482326173e-05,
|
| 1255 |
+
"loss": 0.0124,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 14.094488188976378,
|
| 1260 |
+
"grad_norm": 0.19526968896389008,
|
| 1261 |
+
"learning_rate": 9.552583086261069e-05,
|
| 1262 |
+
"loss": 0.0135,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 14.173228346456693,
|
| 1267 |
+
"grad_norm": 0.1772489845752716,
|
| 1268 |
+
"learning_rate": 9.545721562941168e-05,
|
| 1269 |
+
"loss": 0.0119,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 14.251968503937007,
|
| 1274 |
+
"grad_norm": 0.20985430479049683,
|
| 1275 |
+
"learning_rate": 9.538810328338543e-05,
|
| 1276 |
+
"loss": 0.0106,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 14.330708661417322,
|
| 1281 |
+
"grad_norm": 0.2288864552974701,
|
| 1282 |
+
"learning_rate": 9.531849458033349e-05,
|
| 1283 |
+
"loss": 0.0121,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 14.409448818897637,
|
| 1288 |
+
"grad_norm": 0.14826878905296326,
|
| 1289 |
+
"learning_rate": 9.524839028148547e-05,
|
| 1290 |
+
"loss": 0.0109,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 14.488188976377952,
|
| 1295 |
+
"grad_norm": 0.24729447066783905,
|
| 1296 |
+
"learning_rate": 9.517779115349077e-05,
|
| 1297 |
+
"loss": 0.0122,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 14.566929133858268,
|
| 1302 |
+
"grad_norm": 0.23712359368801117,
|
| 1303 |
+
"learning_rate": 9.510669796841014e-05,
|
| 1304 |
+
"loss": 0.012,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 14.645669291338583,
|
| 1309 |
+
"grad_norm": 0.24393972754478455,
|
| 1310 |
+
"learning_rate": 9.503511150370727e-05,
|
| 1311 |
+
"loss": 0.014,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 14.724409448818898,
|
| 1316 |
+
"grad_norm": 0.16620883345603943,
|
| 1317 |
+
"learning_rate": 9.496303254224024e-05,
|
| 1318 |
+
"loss": 0.0124,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 14.803149606299213,
|
| 1323 |
+
"grad_norm": 0.19335606694221497,
|
| 1324 |
+
"learning_rate": 9.489046187225306e-05,
|
| 1325 |
+
"loss": 0.0116,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 14.881889763779528,
|
| 1330 |
+
"grad_norm": 0.17978379130363464,
|
| 1331 |
+
"learning_rate": 9.481740028736692e-05,
|
| 1332 |
+
"loss": 0.0127,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 14.960629921259843,
|
| 1337 |
+
"grad_norm": 0.15655072033405304,
|
| 1338 |
+
"learning_rate": 9.474384858657164e-05,
|
| 1339 |
+
"loss": 0.0112,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 15.039370078740157,
|
| 1344 |
+
"grad_norm": 0.13158245384693146,
|
| 1345 |
+
"learning_rate": 9.466980757421679e-05,
|
| 1346 |
+
"loss": 0.011,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 15.118110236220472,
|
| 1351 |
+
"grad_norm": 0.21858836710453033,
|
| 1352 |
+
"learning_rate": 9.459527806000305e-05,
|
| 1353 |
+
"loss": 0.0116,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 15.196850393700787,
|
| 1358 |
+
"grad_norm": 0.19522692263126373,
|
| 1359 |
+
"learning_rate": 9.452026085897325e-05,
|
| 1360 |
+
"loss": 0.0118,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 15.275590551181102,
|
| 1365 |
+
"grad_norm": 0.20890626311302185,
|
| 1366 |
+
"learning_rate": 9.444475679150348e-05,
|
| 1367 |
+
"loss": 0.0111,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 15.354330708661417,
|
| 1372 |
+
"grad_norm": 0.20746995508670807,
|
| 1373 |
+
"learning_rate": 9.436876668329411e-05,
|
| 1374 |
+
"loss": 0.0107,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 15.433070866141732,
|
| 1379 |
+
"grad_norm": 0.18878526985645294,
|
| 1380 |
+
"learning_rate": 9.429229136536079e-05,
|
| 1381 |
+
"loss": 0.0105,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 15.511811023622048,
|
| 1386 |
+
"grad_norm": 0.18786223232746124,
|
| 1387 |
+
"learning_rate": 9.421533167402534e-05,
|
| 1388 |
+
"loss": 0.0112,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 15.590551181102363,
|
| 1393 |
+
"grad_norm": 0.12698164582252502,
|
| 1394 |
+
"learning_rate": 9.413788845090666e-05,
|
| 1395 |
+
"loss": 0.011,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 15.669291338582678,
|
| 1400 |
+
"grad_norm": 0.22439169883728027,
|
| 1401 |
+
"learning_rate": 9.405996254291136e-05,
|
| 1402 |
+
"loss": 0.0113,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 15.748031496062993,
|
| 1407 |
+
"grad_norm": 0.16835476458072662,
|
| 1408 |
+
"learning_rate": 9.398155480222474e-05,
|
| 1409 |
+
"loss": 0.0111,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
}
|
| 1412 |
+
],
|
| 1413 |
+
"logging_steps": 10,
|
| 1414 |
+
"max_steps": 10000,
|
| 1415 |
+
"num_input_tokens_seen": 0,
|
| 1416 |
+
"num_train_epochs": 79,
|
| 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": 64,
|
| 1432 |
+
"trial_name": null,
|
| 1433 |
+
"trial_params": null
|
| 1434 |
+
}
|