Upload folder using huggingface_hub
Browse files- config.json +64 -0
- experiment_cfg/metadata.json +355 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +0 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +734 -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 +1,355 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"new_embodiment": {
|
| 3 |
+
"statistics": {
|
| 4 |
+
"state": {
|
| 5 |
+
"left_arm": {
|
| 6 |
+
"max": [
|
| 7 |
+
2.187028646469116,
|
| 8 |
+
-99.21630096435547,
|
| 9 |
+
98.77407836914062,
|
| 10 |
+
74.72486114501953,
|
| 11 |
+
46.27594757080078
|
| 12 |
+
],
|
| 13 |
+
"min": [
|
| 14 |
+
-12.066365242004395,
|
| 15 |
+
-99.45140838623047,
|
| 16 |
+
98.33625030517578,
|
| 17 |
+
73.35863494873047,
|
| 18 |
+
-52.869354248046875
|
| 19 |
+
],
|
| 20 |
+
"mean": [
|
| 21 |
+
-6.931744575500488,
|
| 22 |
+
-99.3890380859375,
|
| 23 |
+
98.62979888916016,
|
| 24 |
+
74.25790405273438,
|
| 25 |
+
1.3063249588012695
|
| 26 |
+
],
|
| 27 |
+
"std": [
|
| 28 |
+
4.007203578948975,
|
| 29 |
+
0.0767870619893074,
|
| 30 |
+
0.12424216419458389,
|
| 31 |
+
0.4313095211982727,
|
| 32 |
+
48.00337600708008
|
| 33 |
+
],
|
| 34 |
+
"q01": [
|
| 35 |
+
-12.066365242004395,
|
| 36 |
+
-99.45140838623047,
|
| 37 |
+
98.51138305664062,
|
| 38 |
+
73.58633422851562,
|
| 39 |
+
-52.869354248046875
|
| 40 |
+
],
|
| 41 |
+
"q99": [
|
| 42 |
+
0.15082956850528717,
|
| 43 |
+
-99.21630096435547,
|
| 44 |
+
98.77407836914062,
|
| 45 |
+
74.72486114501953,
|
| 46 |
+
45.88522720336914
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"gripper1": {
|
| 50 |
+
"max": [
|
| 51 |
+
17.847938537597656
|
| 52 |
+
],
|
| 53 |
+
"min": [
|
| 54 |
+
0.4510309398174286
|
| 55 |
+
],
|
| 56 |
+
"mean": [
|
| 57 |
+
10.498392105102539
|
| 58 |
+
],
|
| 59 |
+
"std": [
|
| 60 |
+
7.273200035095215
|
| 61 |
+
],
|
| 62 |
+
"q01": [
|
| 63 |
+
0.4510309398174286
|
| 64 |
+
],
|
| 65 |
+
"q99": [
|
| 66 |
+
17.847938537597656
|
| 67 |
+
]
|
| 68 |
+
},
|
| 69 |
+
"right_arm": {
|
| 70 |
+
"max": [
|
| 71 |
+
31.229597091674805,
|
| 72 |
+
71.30260467529297,
|
| 73 |
+
98.0138168334961,
|
| 74 |
+
75.76736450195312,
|
| 75 |
+
41.3212776184082
|
| 76 |
+
],
|
| 77 |
+
"min": [
|
| 78 |
+
-27.022125244140625,
|
| 79 |
+
-97.51502990722656,
|
| 80 |
+
-31.865285873413086,
|
| 81 |
+
-96.60742950439453,
|
| 82 |
+
-65.18462371826172
|
| 83 |
+
],
|
| 84 |
+
"mean": [
|
| 85 |
+
6.173414707183838,
|
| 86 |
+
8.965652465820312,
|
| 87 |
+
43.04948425292969,
|
| 88 |
+
-32.647727966308594,
|
| 89 |
+
-44.51104736328125
|
| 90 |
+
],
|
| 91 |
+
"std": [
|
| 92 |
+
17.221529006958008,
|
| 93 |
+
47.10391616821289,
|
| 94 |
+
32.55530548095703,
|
| 95 |
+
45.9875373840332,
|
| 96 |
+
20.5490779876709
|
| 97 |
+
],
|
| 98 |
+
"q01": [
|
| 99 |
+
-24.338048934936523,
|
| 100 |
+
-97.35470581054688,
|
| 101 |
+
-17.356649856567383,
|
| 102 |
+
-96.44587707519531,
|
| 103 |
+
-59.05551528930664
|
| 104 |
+
],
|
| 105 |
+
"q99": [
|
| 106 |
+
29.63365936279297,
|
| 107 |
+
62.324649810791016,
|
| 108 |
+
97.75475311279297,
|
| 109 |
+
73.66720581054688,
|
| 110 |
+
29.715649795532308
|
| 111 |
+
]
|
| 112 |
+
},
|
| 113 |
+
"gripper2": {
|
| 114 |
+
"max": [
|
| 115 |
+
71.65532684326172
|
| 116 |
+
],
|
| 117 |
+
"min": [
|
| 118 |
+
2.324263095855713
|
| 119 |
+
],
|
| 120 |
+
"mean": [
|
| 121 |
+
23.636430740356445
|
| 122 |
+
],
|
| 123 |
+
"std": [
|
| 124 |
+
15.703285217285156
|
| 125 |
+
],
|
| 126 |
+
"q01": [
|
| 127 |
+
2.4376416206359863
|
| 128 |
+
],
|
| 129 |
+
"q99": [
|
| 130 |
+
63.20068164825567
|
| 131 |
+
]
|
| 132 |
+
}
|
| 133 |
+
},
|
| 134 |
+
"action": {
|
| 135 |
+
"left_arm": {
|
| 136 |
+
"max": [
|
| 137 |
+
2.2628026008605957,
|
| 138 |
+
-100.0,
|
| 139 |
+
100.0,
|
| 140 |
+
76.97608184814453,
|
| 141 |
+
46.617828369140625
|
| 142 |
+
],
|
| 143 |
+
"min": [
|
| 144 |
+
-12.107979774475098,
|
| 145 |
+
-100.0,
|
| 146 |
+
99.9084243774414,
|
| 147 |
+
75.35468292236328,
|
| 148 |
+
-52.625152587890625
|
| 149 |
+
],
|
| 150 |
+
"mean": [
|
| 151 |
+
-6.822534084320068,
|
| 152 |
+
-100.0,
|
| 153 |
+
99.92558288574219,
|
| 154 |
+
76.36930084228516,
|
| 155 |
+
1.283119797706604
|
| 156 |
+
],
|
| 157 |
+
"std": [
|
| 158 |
+
4.120677471160889,
|
| 159 |
+
0.0,
|
| 160 |
+
0.0450742170214653,
|
| 161 |
+
0.6158173680305481,
|
| 162 |
+
47.87895584106445
|
| 163 |
+
],
|
| 164 |
+
"q01": [
|
| 165 |
+
-12.107979774475098,
|
| 166 |
+
-100.0,
|
| 167 |
+
99.9084243774414,
|
| 168 |
+
75.35468292236328,
|
| 169 |
+
-52.625152587890625
|
| 170 |
+
],
|
| 171 |
+
"q99": [
|
| 172 |
+
0.43668121099472046,
|
| 173 |
+
-100.0,
|
| 174 |
+
100.0,
|
| 175 |
+
76.97608184814453,
|
| 176 |
+
45.738704681396484
|
| 177 |
+
]
|
| 178 |
+
},
|
| 179 |
+
"gripper1": {
|
| 180 |
+
"max": [
|
| 181 |
+
18.17367935180664
|
| 182 |
+
],
|
| 183 |
+
"min": [
|
| 184 |
+
0.17905102670192719
|
| 185 |
+
],
|
| 186 |
+
"mean": [
|
| 187 |
+
10.681233406066895
|
| 188 |
+
],
|
| 189 |
+
"std": [
|
| 190 |
+
7.212050914764404
|
| 191 |
+
],
|
| 192 |
+
"q01": [
|
| 193 |
+
0.2685765326023102
|
| 194 |
+
],
|
| 195 |
+
"q99": [
|
| 196 |
+
17.99462890625
|
| 197 |
+
]
|
| 198 |
+
},
|
| 199 |
+
"right_arm": {
|
| 200 |
+
"max": [
|
| 201 |
+
31.447315216064453,
|
| 202 |
+
71.15780639648438,
|
| 203 |
+
99.20035552978516,
|
| 204 |
+
77.47068786621094,
|
| 205 |
+
42.69340896606445
|
| 206 |
+
],
|
| 207 |
+
"min": [
|
| 208 |
+
-27.511274337768555,
|
| 209 |
+
-99.25834655761719,
|
| 210 |
+
-32.56330490112305,
|
| 211 |
+
-98.40870666503906,
|
| 212 |
+
-66.81427764892578
|
| 213 |
+
],
|
| 214 |
+
"mean": [
|
| 215 |
+
6.259029865264893,
|
| 216 |
+
7.35891580581665,
|
| 217 |
+
42.082435607910156,
|
| 218 |
+
-33.01646423339844,
|
| 219 |
+
-44.4665641784668
|
| 220 |
+
],
|
| 221 |
+
"std": [
|
| 222 |
+
17.27552604675293,
|
| 223 |
+
47.02023696899414,
|
| 224 |
+
33.24121856689453,
|
| 225 |
+
46.35660171508789,
|
| 226 |
+
20.584171295166016
|
| 227 |
+
],
|
| 228 |
+
"q01": [
|
| 229 |
+
-24.231243133544922,
|
| 230 |
+
-97.77503204345703,
|
| 231 |
+
-18.96934700012207,
|
| 232 |
+
-97.82244873046875,
|
| 233 |
+
-59.10393142700195
|
| 234 |
+
],
|
| 235 |
+
"q99": [
|
| 236 |
+
30.052480850219858,
|
| 237 |
+
62.34033966064453,
|
| 238 |
+
98.48956298828125,
|
| 239 |
+
74.70687103271484,
|
| 240 |
+
30.398540496826172
|
| 241 |
+
]
|
| 242 |
+
},
|
| 243 |
+
"gripper2": {
|
| 244 |
+
"max": [
|
| 245 |
+
71.90612030029297
|
| 246 |
+
],
|
| 247 |
+
"min": [
|
| 248 |
+
0.4267425239086151
|
| 249 |
+
],
|
| 250 |
+
"mean": [
|
| 251 |
+
20.33241081237793
|
| 252 |
+
],
|
| 253 |
+
"std": [
|
| 254 |
+
17.951961517333984
|
| 255 |
+
],
|
| 256 |
+
"q01": [
|
| 257 |
+
1.8492176532745361
|
| 258 |
+
],
|
| 259 |
+
"q99": [
|
| 260 |
+
64.36272933960029
|
| 261 |
+
]
|
| 262 |
+
}
|
| 263 |
+
}
|
| 264 |
+
},
|
| 265 |
+
"modalities": {
|
| 266 |
+
"video": {
|
| 267 |
+
"right": {
|
| 268 |
+
"resolution": [
|
| 269 |
+
640,
|
| 270 |
+
480
|
| 271 |
+
],
|
| 272 |
+
"channels": 3,
|
| 273 |
+
"fps": 30.0
|
| 274 |
+
},
|
| 275 |
+
"top_rgb": {
|
| 276 |
+
"resolution": [
|
| 277 |
+
640,
|
| 278 |
+
480
|
| 279 |
+
],
|
| 280 |
+
"channels": 3,
|
| 281 |
+
"fps": 30.0
|
| 282 |
+
}
|
| 283 |
+
},
|
| 284 |
+
"state": {
|
| 285 |
+
"left_arm": {
|
| 286 |
+
"absolute": true,
|
| 287 |
+
"rotation_type": null,
|
| 288 |
+
"shape": [
|
| 289 |
+
5
|
| 290 |
+
],
|
| 291 |
+
"continuous": true
|
| 292 |
+
},
|
| 293 |
+
"gripper1": {
|
| 294 |
+
"absolute": true,
|
| 295 |
+
"rotation_type": null,
|
| 296 |
+
"shape": [
|
| 297 |
+
1
|
| 298 |
+
],
|
| 299 |
+
"continuous": true
|
| 300 |
+
},
|
| 301 |
+
"right_arm": {
|
| 302 |
+
"absolute": true,
|
| 303 |
+
"rotation_type": null,
|
| 304 |
+
"shape": [
|
| 305 |
+
5
|
| 306 |
+
],
|
| 307 |
+
"continuous": true
|
| 308 |
+
},
|
| 309 |
+
"gripper2": {
|
| 310 |
+
"absolute": true,
|
| 311 |
+
"rotation_type": null,
|
| 312 |
+
"shape": [
|
| 313 |
+
1
|
| 314 |
+
],
|
| 315 |
+
"continuous": true
|
| 316 |
+
}
|
| 317 |
+
},
|
| 318 |
+
"action": {
|
| 319 |
+
"left_arm": {
|
| 320 |
+
"absolute": true,
|
| 321 |
+
"rotation_type": null,
|
| 322 |
+
"shape": [
|
| 323 |
+
5
|
| 324 |
+
],
|
| 325 |
+
"continuous": true
|
| 326 |
+
},
|
| 327 |
+
"gripper1": {
|
| 328 |
+
"absolute": true,
|
| 329 |
+
"rotation_type": null,
|
| 330 |
+
"shape": [
|
| 331 |
+
1
|
| 332 |
+
],
|
| 333 |
+
"continuous": true
|
| 334 |
+
},
|
| 335 |
+
"right_arm": {
|
| 336 |
+
"absolute": true,
|
| 337 |
+
"rotation_type": null,
|
| 338 |
+
"shape": [
|
| 339 |
+
5
|
| 340 |
+
],
|
| 341 |
+
"continuous": true
|
| 342 |
+
},
|
| 343 |
+
"gripper2": {
|
| 344 |
+
"absolute": true,
|
| 345 |
+
"rotation_type": null,
|
| 346 |
+
"shape": [
|
| 347 |
+
1
|
| 348 |
+
],
|
| 349 |
+
"continuous": true
|
| 350 |
+
}
|
| 351 |
+
}
|
| 352 |
+
},
|
| 353 |
+
"embodiment_tag": "new_embodiment"
|
| 354 |
+
}
|
| 355 |
+
}
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2db511ed728aaf36eecfd28fa48f0d77139bf02c6b27d1f169399b53d84ebeba
|
| 3 |
+
size 4999367032
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f04a59b96816a7a820663a91a327c11bb5a3b1ab34a2d059d64325e693869ea
|
| 3 |
+
size 2586705312
|
model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec2cbe99dd138b6621aa07c959f498cfba8ab74b1d1ed7d8344b9e45016b4819
|
| 3 |
+
size 8550720062
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0291d1666eaed8b00f4187bc1c1b5e7de8ae43108df5b7b87ac4e0e74b66318
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0a428aa401f171b5e4e33fd1981cc4dd012b028cd7da79c75568021691b3e83
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,734 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.4945054945054945,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.054945054945054944,
|
| 14 |
+
"grad_norm": 3.1129469871520996,
|
| 15 |
+
"learning_rate": 1.3846153846153847e-05,
|
| 16 |
+
"loss": 1.2218,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.10989010989010989,
|
| 21 |
+
"grad_norm": 1.2952975034713745,
|
| 22 |
+
"learning_rate": 2.9230769230769234e-05,
|
| 23 |
+
"loss": 0.4319,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16483516483516483,
|
| 28 |
+
"grad_norm": 2.0939278602600098,
|
| 29 |
+
"learning_rate": 4.461538461538462e-05,
|
| 30 |
+
"loss": 0.2786,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.21978021978021978,
|
| 35 |
+
"grad_norm": 1.0714941024780273,
|
| 36 |
+
"learning_rate": 6e-05,
|
| 37 |
+
"loss": 0.2013,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.27472527472527475,
|
| 42 |
+
"grad_norm": 1.486087441444397,
|
| 43 |
+
"learning_rate": 7.538461538461539e-05,
|
| 44 |
+
"loss": 0.1555,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.32967032967032966,
|
| 49 |
+
"grad_norm": 0.8458685874938965,
|
| 50 |
+
"learning_rate": 9.076923076923078e-05,
|
| 51 |
+
"loss": 0.1299,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.38461538461538464,
|
| 56 |
+
"grad_norm": 0.5655759572982788,
|
| 57 |
+
"learning_rate": 9.999741165552688e-05,
|
| 58 |
+
"loss": 0.1234,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.43956043956043955,
|
| 63 |
+
"grad_norm": 0.7847583293914795,
|
| 64 |
+
"learning_rate": 9.996829585771195e-05,
|
| 65 |
+
"loss": 0.1084,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.4945054945054945,
|
| 70 |
+
"grad_norm": 0.5260426998138428,
|
| 71 |
+
"learning_rate": 9.990684773387432e-05,
|
| 72 |
+
"loss": 0.0915,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5494505494505495,
|
| 77 |
+
"grad_norm": 0.6431106925010681,
|
| 78 |
+
"learning_rate": 9.981310704444089e-05,
|
| 79 |
+
"loss": 0.0908,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.6043956043956044,
|
| 84 |
+
"grad_norm": 0.4315970838069916,
|
| 85 |
+
"learning_rate": 9.968713444496294e-05,
|
| 86 |
+
"loss": 0.0793,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6593406593406593,
|
| 91 |
+
"grad_norm": 0.4451838731765747,
|
| 92 |
+
"learning_rate": 9.952901144686857e-05,
|
| 93 |
+
"loss": 0.0728,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7142857142857143,
|
| 98 |
+
"grad_norm": 0.44112035632133484,
|
| 99 |
+
"learning_rate": 9.933884036472022e-05,
|
| 100 |
+
"loss": 0.0668,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.7692307692307693,
|
| 105 |
+
"grad_norm": 0.31693196296691895,
|
| 106 |
+
"learning_rate": 9.911674425001134e-05,
|
| 107 |
+
"loss": 0.0648,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8241758241758241,
|
| 112 |
+
"grad_norm": 0.5056548118591309,
|
| 113 |
+
"learning_rate": 9.886286681154495e-05,
|
| 114 |
+
"loss": 0.0644,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.8791208791208791,
|
| 119 |
+
"grad_norm": 0.370319664478302,
|
| 120 |
+
"learning_rate": 9.857737232244604e-05,
|
| 121 |
+
"loss": 0.0578,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9340659340659341,
|
| 126 |
+
"grad_norm": 0.6004385352134705,
|
| 127 |
+
"learning_rate": 9.826044551386744e-05,
|
| 128 |
+
"loss": 0.0561,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.989010989010989,
|
| 133 |
+
"grad_norm": 0.6260236501693726,
|
| 134 |
+
"learning_rate": 9.791229145545831e-05,
|
| 135 |
+
"loss": 0.0511,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.043956043956044,
|
| 140 |
+
"grad_norm": 0.3883954882621765,
|
| 141 |
+
"learning_rate": 9.753313542267241e-05,
|
| 142 |
+
"loss": 0.0528,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.098901098901099,
|
| 147 |
+
"grad_norm": 0.4612903594970703,
|
| 148 |
+
"learning_rate": 9.712322275100208e-05,
|
| 149 |
+
"loss": 0.0539,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.1538461538461537,
|
| 154 |
+
"grad_norm": 0.41645845770835876,
|
| 155 |
+
"learning_rate": 9.668281867723223e-05,
|
| 156 |
+
"loss": 0.0467,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.2087912087912087,
|
| 161 |
+
"grad_norm": 0.45377564430236816,
|
| 162 |
+
"learning_rate": 9.621220816781708e-05,
|
| 163 |
+
"loss": 0.0489,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.2637362637362637,
|
| 168 |
+
"grad_norm": 0.33625757694244385,
|
| 169 |
+
"learning_rate": 9.571169573449075e-05,
|
| 170 |
+
"loss": 0.0467,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.3186813186813187,
|
| 175 |
+
"grad_norm": 0.2890590727329254,
|
| 176 |
+
"learning_rate": 9.51816052372307e-05,
|
| 177 |
+
"loss": 0.0392,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.3736263736263736,
|
| 182 |
+
"grad_norm": 0.5078686475753784,
|
| 183 |
+
"learning_rate": 9.46222796747021e-05,
|
| 184 |
+
"loss": 0.0404,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.4285714285714286,
|
| 189 |
+
"grad_norm": 0.3641613721847534,
|
| 190 |
+
"learning_rate": 9.403408096231812e-05,
|
| 191 |
+
"loss": 0.0407,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.4835164835164836,
|
| 196 |
+
"grad_norm": 0.30617615580558777,
|
| 197 |
+
"learning_rate": 9.341738969806011e-05,
|
| 198 |
+
"loss": 0.0393,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.5384615384615383,
|
| 203 |
+
"grad_norm": 0.3459275960922241,
|
| 204 |
+
"learning_rate": 9.277260491620907e-05,
|
| 205 |
+
"loss": 0.0401,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.5934065934065935,
|
| 210 |
+
"grad_norm": 0.2681010663509369,
|
| 211 |
+
"learning_rate": 9.210014382914784e-05,
|
| 212 |
+
"loss": 0.0363,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.6483516483516483,
|
| 217 |
+
"grad_norm": 0.35169124603271484,
|
| 218 |
+
"learning_rate": 9.140044155740101e-05,
|
| 219 |
+
"loss": 0.0341,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.7032967032967035,
|
| 224 |
+
"grad_norm": 0.28844815492630005,
|
| 225 |
+
"learning_rate": 9.067395084808709e-05,
|
| 226 |
+
"loss": 0.0353,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.7582417582417582,
|
| 231 |
+
"grad_norm": 0.3767380714416504,
|
| 232 |
+
"learning_rate": 8.992114178196558e-05,
|
| 233 |
+
"loss": 0.0339,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.8131868131868132,
|
| 238 |
+
"grad_norm": 0.37167835235595703,
|
| 239 |
+
"learning_rate": 8.914250146926788e-05,
|
| 240 |
+
"loss": 0.032,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.8681318681318682,
|
| 245 |
+
"grad_norm": 0.2863631844520569,
|
| 246 |
+
"learning_rate": 8.833853373450936e-05,
|
| 247 |
+
"loss": 0.0339,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.9230769230769231,
|
| 252 |
+
"grad_norm": 0.25606516003608704,
|
| 253 |
+
"learning_rate": 8.750975879048629e-05,
|
| 254 |
+
"loss": 0.0302,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.978021978021978,
|
| 259 |
+
"grad_norm": 0.28373125195503235,
|
| 260 |
+
"learning_rate": 8.665671290166853e-05,
|
| 261 |
+
"loss": 0.031,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.032967032967033,
|
| 266 |
+
"grad_norm": 0.41454407572746277,
|
| 267 |
+
"learning_rate": 8.577994803720606e-05,
|
| 268 |
+
"loss": 0.0426,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.087912087912088,
|
| 273 |
+
"grad_norm": 0.4597318470478058,
|
| 274 |
+
"learning_rate": 8.488003151377351e-05,
|
| 275 |
+
"loss": 0.0366,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.142857142857143,
|
| 280 |
+
"grad_norm": 0.3471977412700653,
|
| 281 |
+
"learning_rate": 8.395754562848408e-05,
|
| 282 |
+
"loss": 0.0358,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.197802197802198,
|
| 287 |
+
"grad_norm": 0.37275147438049316,
|
| 288 |
+
"learning_rate": 8.301308728211017e-05,
|
| 289 |
+
"loss": 0.0321,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.2527472527472527,
|
| 294 |
+
"grad_norm": 0.29526907205581665,
|
| 295 |
+
"learning_rate": 8.20472675928548e-05,
|
| 296 |
+
"loss": 0.0318,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.3076923076923075,
|
| 301 |
+
"grad_norm": 0.31538689136505127,
|
| 302 |
+
"learning_rate": 8.10607115009232e-05,
|
| 303 |
+
"loss": 0.0312,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.3626373626373627,
|
| 308 |
+
"grad_norm": 0.3267456889152527,
|
| 309 |
+
"learning_rate": 8.005405736415126e-05,
|
| 310 |
+
"loss": 0.0292,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.4175824175824174,
|
| 315 |
+
"grad_norm": 0.23045669496059418,
|
| 316 |
+
"learning_rate": 7.902795654495154e-05,
|
| 317 |
+
"loss": 0.0268,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.4725274725274726,
|
| 322 |
+
"grad_norm": 0.28116142749786377,
|
| 323 |
+
"learning_rate": 7.798307298884487e-05,
|
| 324 |
+
"loss": 0.0268,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.5274725274725274,
|
| 329 |
+
"grad_norm": 0.3379570245742798,
|
| 330 |
+
"learning_rate": 7.692008279484989e-05,
|
| 331 |
+
"loss": 0.0271,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.5824175824175826,
|
| 336 |
+
"grad_norm": 0.2906380295753479,
|
| 337 |
+
"learning_rate": 7.583967377800853e-05,
|
| 338 |
+
"loss": 0.0259,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.6373626373626373,
|
| 343 |
+
"grad_norm": 0.3166893422603607,
|
| 344 |
+
"learning_rate": 7.474254502433064e-05,
|
| 345 |
+
"loss": 0.028,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.6923076923076925,
|
| 350 |
+
"grad_norm": 0.3112421929836273,
|
| 351 |
+
"learning_rate": 7.362940643844564e-05,
|
| 352 |
+
"loss": 0.0284,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.7472527472527473,
|
| 357 |
+
"grad_norm": 0.3011311888694763,
|
| 358 |
+
"learning_rate": 7.250097828425384e-05,
|
| 359 |
+
"loss": 0.0276,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.802197802197802,
|
| 364 |
+
"grad_norm": 0.21055926382541656,
|
| 365 |
+
"learning_rate": 7.135799071887486e-05,
|
| 366 |
+
"loss": 0.0238,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.857142857142857,
|
| 371 |
+
"grad_norm": 0.3269054591655731,
|
| 372 |
+
"learning_rate": 7.020118332019438e-05,
|
| 373 |
+
"loss": 0.0233,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.912087912087912,
|
| 378 |
+
"grad_norm": 0.2849093973636627,
|
| 379 |
+
"learning_rate": 6.903130460831539e-05,
|
| 380 |
+
"loss": 0.0249,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.967032967032967,
|
| 385 |
+
"grad_norm": 0.2573489844799042,
|
| 386 |
+
"learning_rate": 6.784911156122305e-05,
|
| 387 |
+
"loss": 0.0241,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 3.021978021978022,
|
| 392 |
+
"grad_norm": 0.3901364505290985,
|
| 393 |
+
"learning_rate": 6.665536912497687e-05,
|
| 394 |
+
"loss": 0.026,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.076923076923077,
|
| 399 |
+
"grad_norm": 0.41034600138664246,
|
| 400 |
+
"learning_rate": 6.545084971874738e-05,
|
| 401 |
+
"loss": 0.0287,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.131868131868132,
|
| 406 |
+
"grad_norm": 0.24250033497810364,
|
| 407 |
+
"learning_rate": 6.42363327350168e-05,
|
| 408 |
+
"loss": 0.0282,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.186813186813187,
|
| 413 |
+
"grad_norm": 0.3414716124534607,
|
| 414 |
+
"learning_rate": 6.301260403526801e-05,
|
| 415 |
+
"loss": 0.028,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.241758241758242,
|
| 420 |
+
"grad_norm": 0.22624026238918304,
|
| 421 |
+
"learning_rate": 6.178045544148739e-05,
|
| 422 |
+
"loss": 0.0244,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.2967032967032965,
|
| 427 |
+
"grad_norm": 0.21665652096271515,
|
| 428 |
+
"learning_rate": 6.054068422381112e-05,
|
| 429 |
+
"loss": 0.0242,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 3.3516483516483517,
|
| 434 |
+
"grad_norm": 0.19114629924297333,
|
| 435 |
+
"learning_rate": 5.929409258464606e-05,
|
| 436 |
+
"loss": 0.0237,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.4065934065934065,
|
| 441 |
+
"grad_norm": 0.21987110376358032,
|
| 442 |
+
"learning_rate": 5.8041487139599216e-05,
|
| 443 |
+
"loss": 0.025,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.4615384615384617,
|
| 448 |
+
"grad_norm": 0.29730021953582764,
|
| 449 |
+
"learning_rate": 5.678367839555163e-05,
|
| 450 |
+
"loss": 0.0226,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 3.5164835164835164,
|
| 455 |
+
"grad_norm": 0.2368212342262268,
|
| 456 |
+
"learning_rate": 5.5521480226214405e-05,
|
| 457 |
+
"loss": 0.0241,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.571428571428571,
|
| 462 |
+
"grad_norm": 0.23423047363758087,
|
| 463 |
+
"learning_rate": 5.425570934550617e-05,
|
| 464 |
+
"loss": 0.0207,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.6263736263736264,
|
| 469 |
+
"grad_norm": 0.18969936668872833,
|
| 470 |
+
"learning_rate": 5.2987184779092715e-05,
|
| 471 |
+
"loss": 0.0228,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 3.6813186813186816,
|
| 476 |
+
"grad_norm": 0.2447623908519745,
|
| 477 |
+
"learning_rate": 5.171672733443098e-05,
|
| 478 |
+
"loss": 0.019,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 3.7362637362637363,
|
| 483 |
+
"grad_norm": 0.16867715120315552,
|
| 484 |
+
"learning_rate": 5.044515906965988e-05,
|
| 485 |
+
"loss": 0.0193,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 3.791208791208791,
|
| 490 |
+
"grad_norm": 0.15809307992458344,
|
| 491 |
+
"learning_rate": 4.917330276168208e-05,
|
| 492 |
+
"loss": 0.0218,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 3.8461538461538463,
|
| 497 |
+
"grad_norm": 0.2224949151277542,
|
| 498 |
+
"learning_rate": 4.790198137378056e-05,
|
| 499 |
+
"loss": 0.0185,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.901098901098901,
|
| 504 |
+
"grad_norm": 0.23750196397304535,
|
| 505 |
+
"learning_rate": 4.663201752311461e-05,
|
| 506 |
+
"loss": 0.0194,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.956043956043956,
|
| 511 |
+
"grad_norm": 0.16117140650749207,
|
| 512 |
+
"learning_rate": 4.536423294843978e-05,
|
| 513 |
+
"loss": 0.0185,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 4.010989010989011,
|
| 518 |
+
"grad_norm": 0.21286968886852264,
|
| 519 |
+
"learning_rate": 4.409944797839635e-05,
|
| 520 |
+
"loss": 0.0186,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 4.065934065934066,
|
| 525 |
+
"grad_norm": 0.2058119922876358,
|
| 526 |
+
"learning_rate": 4.283848100070988e-05,
|
| 527 |
+
"loss": 0.0198,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.1208791208791204,
|
| 532 |
+
"grad_norm": 0.18852297961711884,
|
| 533 |
+
"learning_rate": 4.1582147932648074e-05,
|
| 534 |
+
"loss": 0.0236,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 4.175824175824176,
|
| 539 |
+
"grad_norm": 0.2005940079689026,
|
| 540 |
+
"learning_rate": 4.033126169307584e-05,
|
| 541 |
+
"loss": 0.0193,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 4.230769230769231,
|
| 546 |
+
"grad_norm": 0.26670485734939575,
|
| 547 |
+
"learning_rate": 3.9086631676450586e-05,
|
| 548 |
+
"loss": 0.0197,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.285714285714286,
|
| 553 |
+
"grad_norm": 0.14271725714206696,
|
| 554 |
+
"learning_rate": 3.784906322909813e-05,
|
| 555 |
+
"loss": 0.0175,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.34065934065934,
|
| 560 |
+
"grad_norm": 0.2008945196866989,
|
| 561 |
+
"learning_rate": 3.661935712810779e-05,
|
| 562 |
+
"loss": 0.0191,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.395604395604396,
|
| 567 |
+
"grad_norm": 0.14711880683898926,
|
| 568 |
+
"learning_rate": 3.5398309063184146e-05,
|
| 569 |
+
"loss": 0.018,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 4.450549450549451,
|
| 574 |
+
"grad_norm": 0.2107602059841156,
|
| 575 |
+
"learning_rate": 3.418670912179057e-05,
|
| 576 |
+
"loss": 0.0173,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 4.5054945054945055,
|
| 581 |
+
"grad_norm": 0.16945497691631317,
|
| 582 |
+
"learning_rate": 3.298534127791785e-05,
|
| 583 |
+
"loss": 0.0178,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.56043956043956,
|
| 588 |
+
"grad_norm": 0.14888402819633484,
|
| 589 |
+
"learning_rate": 3.179498288480834e-05,
|
| 590 |
+
"loss": 0.0181,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.615384615384615,
|
| 595 |
+
"grad_norm": 0.1862138956785202,
|
| 596 |
+
"learning_rate": 3.061640417196433e-05,
|
| 597 |
+
"loss": 0.0171,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.670329670329671,
|
| 602 |
+
"grad_norm": 0.17378169298171997,
|
| 603 |
+
"learning_rate": 2.945036774676587e-05,
|
| 604 |
+
"loss": 0.0156,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 4.725274725274725,
|
| 609 |
+
"grad_norm": 0.1787486970424652,
|
| 610 |
+
"learning_rate": 2.8297628101020322e-05,
|
| 611 |
+
"loss": 0.0205,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 4.78021978021978,
|
| 616 |
+
"grad_norm": 0.1538439691066742,
|
| 617 |
+
"learning_rate": 2.71589311227634e-05,
|
| 618 |
+
"loss": 0.0165,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 4.835164835164835,
|
| 623 |
+
"grad_norm": 0.16950446367263794,
|
| 624 |
+
"learning_rate": 2.6035013613627224e-05,
|
| 625 |
+
"loss": 0.0168,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.8901098901098905,
|
| 630 |
+
"grad_norm": 0.1429736614227295,
|
| 631 |
+
"learning_rate": 2.492660281208779e-05,
|
| 632 |
+
"loss": 0.0137,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.945054945054945,
|
| 637 |
+
"grad_norm": 0.15313905477523804,
|
| 638 |
+
"learning_rate": 2.3834415922900415e-05,
|
| 639 |
+
"loss": 0.0156,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 5.0,
|
| 644 |
+
"grad_norm": 0.5302070379257202,
|
| 645 |
+
"learning_rate": 2.2759159653027374e-05,
|
| 646 |
+
"loss": 0.0135,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 5.054945054945055,
|
| 651 |
+
"grad_norm": 0.18755768239498138,
|
| 652 |
+
"learning_rate": 2.170152975435859e-05,
|
| 653 |
+
"loss": 0.016,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 5.1098901098901095,
|
| 658 |
+
"grad_norm": 0.24148383736610413,
|
| 659 |
+
"learning_rate": 2.066221057352036e-05,
|
| 660 |
+
"loss": 0.015,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 5.164835164835165,
|
| 665 |
+
"grad_norm": 0.20809908211231232,
|
| 666 |
+
"learning_rate": 1.9641874609064443e-05,
|
| 667 |
+
"loss": 0.0174,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 5.21978021978022,
|
| 672 |
+
"grad_norm": 0.21820078790187836,
|
| 673 |
+
"learning_rate": 1.8641182076323148e-05,
|
| 674 |
+
"loss": 0.0157,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 5.274725274725275,
|
| 679 |
+
"grad_norm": 0.17557694017887115,
|
| 680 |
+
"learning_rate": 1.76607804802126e-05,
|
| 681 |
+
"loss": 0.0158,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 5.329670329670329,
|
| 686 |
+
"grad_norm": 0.2052188515663147,
|
| 687 |
+
"learning_rate": 1.6701304196260166e-05,
|
| 688 |
+
"loss": 0.0145,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 5.384615384615385,
|
| 693 |
+
"grad_norm": 0.19574584066867828,
|
| 694 |
+
"learning_rate": 1.5763374060127624e-05,
|
| 695 |
+
"loss": 0.0146,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.43956043956044,
|
| 700 |
+
"grad_norm": 0.15790286660194397,
|
| 701 |
+
"learning_rate": 1.4847596965895011e-05,
|
| 702 |
+
"loss": 0.0162,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 5.4945054945054945,
|
| 707 |
+
"grad_norm": 0.16527500748634338,
|
| 708 |
+
"learning_rate": 1.3954565473365844e-05,
|
| 709 |
+
"loss": 0.017,
|
| 710 |
+
"step": 1000
|
| 711 |
+
}
|
| 712 |
+
],
|
| 713 |
+
"logging_steps": 10,
|
| 714 |
+
"max_steps": 1300,
|
| 715 |
+
"num_input_tokens_seen": 0,
|
| 716 |
+
"num_train_epochs": 8,
|
| 717 |
+
"save_steps": 1000,
|
| 718 |
+
"stateful_callbacks": {
|
| 719 |
+
"TrainerControl": {
|
| 720 |
+
"args": {
|
| 721 |
+
"should_epoch_stop": false,
|
| 722 |
+
"should_evaluate": false,
|
| 723 |
+
"should_log": false,
|
| 724 |
+
"should_save": true,
|
| 725 |
+
"should_training_stop": false
|
| 726 |
+
},
|
| 727 |
+
"attributes": {}
|
| 728 |
+
}
|
| 729 |
+
},
|
| 730 |
+
"total_flos": 0.0,
|
| 731 |
+
"train_batch_size": 100,
|
| 732 |
+
"trial_name": null,
|
| 733 |
+
"trial_params": null
|
| 734 |
+
}
|