Final model upload
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
- trainer_state.json +1093 -0
- training_args.bin +3 -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 |
+
45.07377624511719,
|
| 8 |
+
55.723270416259766,
|
| 9 |
+
100.0,
|
| 10 |
+
75.35292053222656,
|
| 11 |
+
-32.6773681640625
|
| 12 |
+
],
|
| 13 |
+
"min": [
|
| 14 |
+
-17.848644256591797,
|
| 15 |
+
-100.0,
|
| 16 |
+
-19.266876220703125,
|
| 17 |
+
-95.87943267822266,
|
| 18 |
+
-62.09217834472656
|
| 19 |
+
],
|
| 20 |
+
"mean": [
|
| 21 |
+
5.961390018463135,
|
| 22 |
+
-69.77558898925781,
|
| 23 |
+
86.302001953125,
|
| 24 |
+
44.21531295776367,
|
| 25 |
+
-43.49432373046875
|
| 26 |
+
],
|
| 27 |
+
"std": [
|
| 28 |
+
11.90206527709961,
|
| 29 |
+
51.40785598754883,
|
| 30 |
+
26.839374542236328,
|
| 31 |
+
52.70073699951172,
|
| 32 |
+
3.3679187297821045
|
| 33 |
+
],
|
| 34 |
+
"q01": [
|
| 35 |
+
-12.803426742553711,
|
| 36 |
+
-100.0,
|
| 37 |
+
4.336164474487305,
|
| 38 |
+
-86.26478576660156,
|
| 39 |
+
-53.236663818359375
|
| 40 |
+
],
|
| 41 |
+
"q99": [
|
| 42 |
+
40.98048400878906,
|
| 43 |
+
42.80922317504883,
|
| 44 |
+
100.0,
|
| 45 |
+
75.35292053222656,
|
| 46 |
+
-35.784568786621094
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"gripper1": {
|
| 50 |
+
"max": [
|
| 51 |
+
76.53194427490234
|
| 52 |
+
],
|
| 53 |
+
"min": [
|
| 54 |
+
0.0
|
| 55 |
+
],
|
| 56 |
+
"mean": [
|
| 57 |
+
9.278239250183105
|
| 58 |
+
],
|
| 59 |
+
"std": [
|
| 60 |
+
10.504193305969238
|
| 61 |
+
],
|
| 62 |
+
"q01": [
|
| 63 |
+
1.8904824256896973
|
| 64 |
+
],
|
| 65 |
+
"q99": [
|
| 66 |
+
57.62711715698242
|
| 67 |
+
]
|
| 68 |
+
},
|
| 69 |
+
"right_arm": {
|
| 70 |
+
"max": [
|
| 71 |
+
55.04941940307617,
|
| 72 |
+
87.9725112915039,
|
| 73 |
+
98.92520904541016,
|
| 74 |
+
78.59872436523438,
|
| 75 |
+
43.591060638427734
|
| 76 |
+
],
|
| 77 |
+
"min": [
|
| 78 |
+
-29.179201126098633,
|
| 79 |
+
-98.71134185791016,
|
| 80 |
+
-56.38154983520508,
|
| 81 |
+
-97.8768539428711,
|
| 82 |
+
-70.25430297851562
|
| 83 |
+
],
|
| 84 |
+
"mean": [
|
| 85 |
+
4.257948875427246,
|
| 86 |
+
-6.147779941558838,
|
| 87 |
+
54.093971252441406,
|
| 88 |
+
-21.178335189819336,
|
| 89 |
+
-35.65965270996094
|
| 90 |
+
],
|
| 91 |
+
"std": [
|
| 92 |
+
21.212779998779297,
|
| 93 |
+
55.296844482421875,
|
| 94 |
+
32.8692741394043,
|
| 95 |
+
56.77510452270508,
|
| 96 |
+
26.9205322265625
|
| 97 |
+
],
|
| 98 |
+
"q01": [
|
| 99 |
+
-26.342931747436523,
|
| 100 |
+
-98.3676986694336,
|
| 101 |
+
-20.644872665405273,
|
| 102 |
+
-97.62207794189453,
|
| 103 |
+
-59.15746307373047
|
| 104 |
+
],
|
| 105 |
+
"q99": [
|
| 106 |
+
49.63472366333008,
|
| 107 |
+
69.58763122558594,
|
| 108 |
+
98.56694793701172,
|
| 109 |
+
77.74946594238281,
|
| 110 |
+
37.63164520263672
|
| 111 |
+
]
|
| 112 |
+
},
|
| 113 |
+
"gripper2": {
|
| 114 |
+
"max": [
|
| 115 |
+
84.19355010986328
|
| 116 |
+
],
|
| 117 |
+
"min": [
|
| 118 |
+
0.9032257795333862
|
| 119 |
+
],
|
| 120 |
+
"mean": [
|
| 121 |
+
14.522881507873535
|
| 122 |
+
],
|
| 123 |
+
"std": [
|
| 124 |
+
17.216049194335938
|
| 125 |
+
],
|
| 126 |
+
"q01": [
|
| 127 |
+
1.0322580337524414
|
| 128 |
+
],
|
| 129 |
+
"q99": [
|
| 130 |
+
66.38710021972656
|
| 131 |
+
]
|
| 132 |
+
}
|
| 133 |
+
},
|
| 134 |
+
"action": {
|
| 135 |
+
"left_arm": {
|
| 136 |
+
"max": [
|
| 137 |
+
46.59775924682617,
|
| 138 |
+
55.16328430175781,
|
| 139 |
+
100.0,
|
| 140 |
+
76.70845031738281,
|
| 141 |
+
-32.421207427978516
|
| 142 |
+
],
|
| 143 |
+
"min": [
|
| 144 |
+
-18.173988342285156,
|
| 145 |
+
-100.0,
|
| 146 |
+
-21.63153076171875,
|
| 147 |
+
-97.25030517578125,
|
| 148 |
+
-64.23161315917969
|
| 149 |
+
],
|
| 150 |
+
"mean": [
|
| 151 |
+
6.162662506103516,
|
| 152 |
+
-70.2553939819336,
|
| 153 |
+
86.26774597167969,
|
| 154 |
+
44.83501052856445,
|
| 155 |
+
-43.4974365234375
|
| 156 |
+
],
|
| 157 |
+
"std": [
|
| 158 |
+
12.038820266723633,
|
| 159 |
+
51.182594299316406,
|
| 160 |
+
27.53900146484375,
|
| 161 |
+
53.55598449707031,
|
| 162 |
+
3.376429557800293
|
| 163 |
+
],
|
| 164 |
+
"q01": [
|
| 165 |
+
-12.6614990234375,
|
| 166 |
+
-100.0,
|
| 167 |
+
2.4747936725616455,
|
| 168 |
+
-86.89850616455078,
|
| 169 |
+
-53.18836975097656
|
| 170 |
+
],
|
| 171 |
+
"q99": [
|
| 172 |
+
41.77433395385742,
|
| 173 |
+
41.83583450317383,
|
| 174 |
+
100.0,
|
| 175 |
+
76.70845031738281,
|
| 176 |
+
-35.93940734863281
|
| 177 |
+
]
|
| 178 |
+
},
|
| 179 |
+
"gripper1": {
|
| 180 |
+
"max": [
|
| 181 |
+
77.81721496582031
|
| 182 |
+
],
|
| 183 |
+
"min": [
|
| 184 |
+
0.0
|
| 185 |
+
],
|
| 186 |
+
"mean": [
|
| 187 |
+
9.255863189697266
|
| 188 |
+
],
|
| 189 |
+
"std": [
|
| 190 |
+
10.7840576171875
|
| 191 |
+
],
|
| 192 |
+
"q01": [
|
| 193 |
+
0.0
|
| 194 |
+
],
|
| 195 |
+
"q99": [
|
| 196 |
+
57.67524337768555
|
| 197 |
+
]
|
| 198 |
+
},
|
| 199 |
+
"right_arm": {
|
| 200 |
+
"max": [
|
| 201 |
+
55.28523635864258,
|
| 202 |
+
86.01165771484375,
|
| 203 |
+
99.54934692382812,
|
| 204 |
+
79.18629455566406,
|
| 205 |
+
44.62266540527344
|
| 206 |
+
],
|
| 207 |
+
"min": [
|
| 208 |
+
-29.614093780517578,
|
| 209 |
+
-99.1673583984375,
|
| 210 |
+
-60.34249496459961,
|
| 211 |
+
-99.82869720458984,
|
| 212 |
+
-70.54956817626953
|
| 213 |
+
],
|
| 214 |
+
"mean": [
|
| 215 |
+
4.253830909729004,
|
| 216 |
+
-8.125250816345215,
|
| 217 |
+
52.800907135009766,
|
| 218 |
+
-21.562973022460938,
|
| 219 |
+
-35.52308654785156
|
| 220 |
+
],
|
| 221 |
+
"std": [
|
| 222 |
+
21.214542388916016,
|
| 223 |
+
54.50907516479492,
|
| 224 |
+
34.04475784301758,
|
| 225 |
+
57.11996841430664,
|
| 226 |
+
26.99753761291504
|
| 227 |
+
],
|
| 228 |
+
"q01": [
|
| 229 |
+
-26.42617416381836,
|
| 230 |
+
-98.08493041992188,
|
| 231 |
+
-24.200090408325195,
|
| 232 |
+
-98.80085754394531,
|
| 233 |
+
-58.927162170410156
|
| 234 |
+
],
|
| 235 |
+
"q99": [
|
| 236 |
+
49.6644287109375,
|
| 237 |
+
67.44379425048828,
|
| 238 |
+
98.9184341430664,
|
| 239 |
+
78.32976531982422,
|
| 240 |
+
38.25926971435547
|
| 241 |
+
]
|
| 242 |
+
},
|
| 243 |
+
"gripper2": {
|
| 244 |
+
"max": [
|
| 245 |
+
84.55223846435547
|
| 246 |
+
],
|
| 247 |
+
"min": [
|
| 248 |
+
0.1492537260055542
|
| 249 |
+
],
|
| 250 |
+
"mean": [
|
| 251 |
+
12.350263595581055
|
| 252 |
+
],
|
| 253 |
+
"std": [
|
| 254 |
+
18.221548080444336
|
| 255 |
+
],
|
| 256 |
+
"q01": [
|
| 257 |
+
0.8208954930305481
|
| 258 |
+
],
|
| 259 |
+
"q99": [
|
| 260 |
+
66.3432846069336
|
| 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 sha256:5e5f2f9badfbaa9d9f6ca0b6a2cc1ef9b6484a3012bcf64b9460179119af0a61
|
| 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:a2443a691eaba4c2ffec5c2e5543a994cb5a90b0f667fcaa9d982a1fd460cdd2
|
| 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 +1,1093 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.8142589118198873,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01876172607879925,
|
| 14 |
+
"grad_norm": 3.62505841255188,
|
| 15 |
+
"learning_rate": 1.2e-05,
|
| 16 |
+
"loss": 1.0239,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0375234521575985,
|
| 21 |
+
"grad_norm": 1.075655221939087,
|
| 22 |
+
"learning_rate": 2.5333333333333337e-05,
|
| 23 |
+
"loss": 0.2792,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05628517823639775,
|
| 28 |
+
"grad_norm": 0.6459623575210571,
|
| 29 |
+
"learning_rate": 3.866666666666667e-05,
|
| 30 |
+
"loss": 0.1393,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.075046904315197,
|
| 35 |
+
"grad_norm": 0.5845428705215454,
|
| 36 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 37 |
+
"loss": 0.1033,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09380863039399624,
|
| 42 |
+
"grad_norm": 0.5566798448562622,
|
| 43 |
+
"learning_rate": 6.533333333333334e-05,
|
| 44 |
+
"loss": 0.086,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1125703564727955,
|
| 49 |
+
"grad_norm": 0.5197991132736206,
|
| 50 |
+
"learning_rate": 7.866666666666666e-05,
|
| 51 |
+
"loss": 0.0709,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.13133208255159476,
|
| 56 |
+
"grad_norm": 0.5038473010063171,
|
| 57 |
+
"learning_rate": 9.200000000000001e-05,
|
| 58 |
+
"loss": 0.0693,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.150093808630394,
|
| 63 |
+
"grad_norm": 0.6459262371063232,
|
| 64 |
+
"learning_rate": 9.999805586153205e-05,
|
| 65 |
+
"loss": 0.0631,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16885553470919323,
|
| 70 |
+
"grad_norm": 0.41942092776298523,
|
| 71 |
+
"learning_rate": 9.997618604001829e-05,
|
| 72 |
+
"loss": 0.0583,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.18761726078799248,
|
| 77 |
+
"grad_norm": 0.5155685544013977,
|
| 78 |
+
"learning_rate": 9.993002688846913e-05,
|
| 79 |
+
"loss": 0.0551,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.20637898686679174,
|
| 84 |
+
"grad_norm": 0.35105767846107483,
|
| 85 |
+
"learning_rate": 9.985960084106682e-05,
|
| 86 |
+
"loss": 0.0477,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.225140712945591,
|
| 91 |
+
"grad_norm": 0.37585997581481934,
|
| 92 |
+
"learning_rate": 9.976494212614377e-05,
|
| 93 |
+
"loss": 0.0469,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.24390243902439024,
|
| 98 |
+
"grad_norm": 0.3624650835990906,
|
| 99 |
+
"learning_rate": 9.964609674954696e-05,
|
| 100 |
+
"loss": 0.0423,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2626641651031895,
|
| 105 |
+
"grad_norm": 0.41389772295951843,
|
| 106 |
+
"learning_rate": 9.950312247227825e-05,
|
| 107 |
+
"loss": 0.0429,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.28142589118198874,
|
| 112 |
+
"grad_norm": 0.2967846691608429,
|
| 113 |
+
"learning_rate": 9.933608878242153e-05,
|
| 114 |
+
"loss": 0.0432,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.300187617260788,
|
| 119 |
+
"grad_norm": 0.3088073134422302,
|
| 120 |
+
"learning_rate": 9.914507686137019e-05,
|
| 121 |
+
"loss": 0.0392,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.31894934333958724,
|
| 126 |
+
"grad_norm": 0.2572576701641083,
|
| 127 |
+
"learning_rate": 9.893017954437156e-05,
|
| 128 |
+
"loss": 0.0377,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33771106941838647,
|
| 133 |
+
"grad_norm": 0.40623533725738525,
|
| 134 |
+
"learning_rate": 9.869150127540727e-05,
|
| 135 |
+
"loss": 0.0347,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.35647279549718575,
|
| 140 |
+
"grad_norm": 0.41246405243873596,
|
| 141 |
+
"learning_rate": 9.842915805643155e-05,
|
| 142 |
+
"loss": 0.0349,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.37523452157598497,
|
| 147 |
+
"grad_norm": 0.34366941452026367,
|
| 148 |
+
"learning_rate": 9.81432773909923e-05,
|
| 149 |
+
"loss": 0.0348,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.39399624765478425,
|
| 154 |
+
"grad_norm": 0.41731664538383484,
|
| 155 |
+
"learning_rate": 9.783399822226189e-05,
|
| 156 |
+
"loss": 0.0313,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.41275797373358347,
|
| 161 |
+
"grad_norm": 0.39282381534576416,
|
| 162 |
+
"learning_rate": 9.750147086550844e-05,
|
| 163 |
+
"loss": 0.0361,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.43151969981238275,
|
| 168 |
+
"grad_norm": 0.2867172956466675,
|
| 169 |
+
"learning_rate": 9.714585693503974e-05,
|
| 170 |
+
"loss": 0.0347,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.450281425891182,
|
| 175 |
+
"grad_norm": 0.2782723903656006,
|
| 176 |
+
"learning_rate": 9.676732926565585e-05,
|
| 177 |
+
"loss": 0.0302,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.46904315196998125,
|
| 182 |
+
"grad_norm": 0.34300780296325684,
|
| 183 |
+
"learning_rate": 9.636607182864827e-05,
|
| 184 |
+
"loss": 0.0286,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.4878048780487805,
|
| 189 |
+
"grad_norm": 0.31700846552848816,
|
| 190 |
+
"learning_rate": 9.594227964238653e-05,
|
| 191 |
+
"loss": 0.032,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5065666041275797,
|
| 196 |
+
"grad_norm": 0.48065921664237976,
|
| 197 |
+
"learning_rate": 9.549615867753573e-05,
|
| 198 |
+
"loss": 0.0303,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.525328330206379,
|
| 203 |
+
"grad_norm": 0.3679749369621277,
|
| 204 |
+
"learning_rate": 9.502792575695112e-05,
|
| 205 |
+
"loss": 0.0313,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5440900562851783,
|
| 210 |
+
"grad_norm": 0.3818820118904114,
|
| 211 |
+
"learning_rate": 9.453780845029821e-05,
|
| 212 |
+
"loss": 0.0287,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5628517823639775,
|
| 217 |
+
"grad_norm": 0.2918836176395416,
|
| 218 |
+
"learning_rate": 9.402604496344984e-05,
|
| 219 |
+
"loss": 0.0281,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5816135084427767,
|
| 224 |
+
"grad_norm": 0.35366421937942505,
|
| 225 |
+
"learning_rate": 9.349288402271388e-05,
|
| 226 |
+
"loss": 0.0286,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.600375234521576,
|
| 231 |
+
"grad_norm": 0.3901827931404114,
|
| 232 |
+
"learning_rate": 9.293858475394754e-05,
|
| 233 |
+
"loss": 0.0286,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6191369606003753,
|
| 238 |
+
"grad_norm": 0.4397772252559662,
|
| 239 |
+
"learning_rate": 9.236341655661778e-05,
|
| 240 |
+
"loss": 0.0267,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6378986866791745,
|
| 245 |
+
"grad_norm": 0.3393397033214569,
|
| 246 |
+
"learning_rate": 9.176765897286813e-05,
|
| 247 |
+
"loss": 0.0284,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.6566604127579737,
|
| 252 |
+
"grad_norm": 0.2746170163154602,
|
| 253 |
+
"learning_rate": 9.115160155165614e-05,
|
| 254 |
+
"loss": 0.0251,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6754221388367729,
|
| 259 |
+
"grad_norm": 0.36951932311058044,
|
| 260 |
+
"learning_rate": 9.05155437080275e-05,
|
| 261 |
+
"loss": 0.0263,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6941838649155723,
|
| 266 |
+
"grad_norm": 0.32861998677253723,
|
| 267 |
+
"learning_rate": 8.98597945775948e-05,
|
| 268 |
+
"loss": 0.0254,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7129455909943715,
|
| 273 |
+
"grad_norm": 0.2612588703632355,
|
| 274 |
+
"learning_rate": 8.9184672866292e-05,
|
| 275 |
+
"loss": 0.0224,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7317073170731707,
|
| 280 |
+
"grad_norm": 0.27696406841278076,
|
| 281 |
+
"learning_rate": 8.849050669547768e-05,
|
| 282 |
+
"loss": 0.0222,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7504690431519699,
|
| 287 |
+
"grad_norm": 0.32792651653289795,
|
| 288 |
+
"learning_rate": 8.77776334424621e-05,
|
| 289 |
+
"loss": 0.0235,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7692307692307693,
|
| 294 |
+
"grad_norm": 0.30844274163246155,
|
| 295 |
+
"learning_rate": 8.704639957653567e-05,
|
| 296 |
+
"loss": 0.0235,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7879924953095685,
|
| 301 |
+
"grad_norm": 0.1702653169631958,
|
| 302 |
+
"learning_rate": 8.629716049057872e-05,
|
| 303 |
+
"loss": 0.0245,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8067542213883677,
|
| 308 |
+
"grad_norm": 0.2871593236923218,
|
| 309 |
+
"learning_rate": 8.553028032833397e-05,
|
| 310 |
+
"loss": 0.0219,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8255159474671669,
|
| 315 |
+
"grad_norm": 0.27894458174705505,
|
| 316 |
+
"learning_rate": 8.474613180742628e-05,
|
| 317 |
+
"loss": 0.0187,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8442776735459663,
|
| 322 |
+
"grad_norm": 0.2544475197792053,
|
| 323 |
+
"learning_rate": 8.394509603821499e-05,
|
| 324 |
+
"loss": 0.0192,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8630393996247655,
|
| 329 |
+
"grad_norm": 0.22808410227298737,
|
| 330 |
+
"learning_rate": 8.31275623385675e-05,
|
| 331 |
+
"loss": 0.0217,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8818011257035647,
|
| 336 |
+
"grad_norm": 0.23319919407367706,
|
| 337 |
+
"learning_rate": 8.229392804464362e-05,
|
| 338 |
+
"loss": 0.0194,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.900562851782364,
|
| 343 |
+
"grad_norm": 0.3450873792171478,
|
| 344 |
+
"learning_rate": 8.14445983177832e-05,
|
| 345 |
+
"loss": 0.0219,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9193245778611632,
|
| 350 |
+
"grad_norm": 0.32911404967308044,
|
| 351 |
+
"learning_rate": 8.057998594759022e-05,
|
| 352 |
+
"loss": 0.0204,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9380863039399625,
|
| 357 |
+
"grad_norm": 0.31988227367401123,
|
| 358 |
+
"learning_rate": 7.970051115130966e-05,
|
| 359 |
+
"loss": 0.0214,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9568480300187617,
|
| 364 |
+
"grad_norm": 0.3178263306617737,
|
| 365 |
+
"learning_rate": 7.880660136959428e-05,
|
| 366 |
+
"loss": 0.0187,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.975609756097561,
|
| 371 |
+
"grad_norm": 0.27726197242736816,
|
| 372 |
+
"learning_rate": 7.789869105876083e-05,
|
| 373 |
+
"loss": 0.0176,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9943714821763602,
|
| 378 |
+
"grad_norm": 0.2947522699832916,
|
| 379 |
+
"learning_rate": 7.697722147963626e-05,
|
| 380 |
+
"loss": 0.0184,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0131332082551594,
|
| 385 |
+
"grad_norm": 0.3290204405784607,
|
| 386 |
+
"learning_rate": 7.604264048309717e-05,
|
| 387 |
+
"loss": 0.0186,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0318949343339587,
|
| 392 |
+
"grad_norm": 0.30177435278892517,
|
| 393 |
+
"learning_rate": 7.509540229240601e-05,
|
| 394 |
+
"loss": 0.0186,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.050656660412758,
|
| 399 |
+
"grad_norm": 0.2899879813194275,
|
| 400 |
+
"learning_rate": 7.413596728245054e-05,
|
| 401 |
+
"loss": 0.0186,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0694183864915572,
|
| 406 |
+
"grad_norm": 0.29170823097229004,
|
| 407 |
+
"learning_rate": 7.316480175599309e-05,
|
| 408 |
+
"loss": 0.019,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0881801125703565,
|
| 413 |
+
"grad_norm": 0.2978293299674988,
|
| 414 |
+
"learning_rate": 7.218237771703921e-05,
|
| 415 |
+
"loss": 0.0189,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1069418386491556,
|
| 420 |
+
"grad_norm": 0.2766323685646057,
|
| 421 |
+
"learning_rate": 7.118917264143501e-05,
|
| 422 |
+
"loss": 0.0171,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.125703564727955,
|
| 427 |
+
"grad_norm": 0.2202291339635849,
|
| 428 |
+
"learning_rate": 7.018566924480543e-05,
|
| 429 |
+
"loss": 0.0185,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.1444652908067543,
|
| 434 |
+
"grad_norm": 0.27024173736572266,
|
| 435 |
+
"learning_rate": 6.917235524794558e-05,
|
| 436 |
+
"loss": 0.0176,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1632270168855534,
|
| 441 |
+
"grad_norm": 0.3646942675113678,
|
| 442 |
+
"learning_rate": 6.814972313977967e-05,
|
| 443 |
+
"loss": 0.018,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1819887429643527,
|
| 448 |
+
"grad_norm": 0.28385287523269653,
|
| 449 |
+
"learning_rate": 6.711826993800248e-05,
|
| 450 |
+
"loss": 0.0175,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.200750469043152,
|
| 455 |
+
"grad_norm": 0.21348656713962555,
|
| 456 |
+
"learning_rate": 6.607849694751977e-05,
|
| 457 |
+
"loss": 0.0174,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.2195121951219512,
|
| 462 |
+
"grad_norm": 0.23341545462608337,
|
| 463 |
+
"learning_rate": 6.503090951680512e-05,
|
| 464 |
+
"loss": 0.0163,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2382739212007505,
|
| 469 |
+
"grad_norm": 0.21429835259914398,
|
| 470 |
+
"learning_rate": 6.397601679229126e-05,
|
| 471 |
+
"loss": 0.0173,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2570356472795496,
|
| 476 |
+
"grad_norm": 0.22369572520256042,
|
| 477 |
+
"learning_rate": 6.291433147091583e-05,
|
| 478 |
+
"loss": 0.0138,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.275797373358349,
|
| 483 |
+
"grad_norm": 0.2960052788257599,
|
| 484 |
+
"learning_rate": 6.184636955094138e-05,
|
| 485 |
+
"loss": 0.0147,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.2945590994371483,
|
| 490 |
+
"grad_norm": 0.2096583992242813,
|
| 491 |
+
"learning_rate": 6.077265008117081e-05,
|
| 492 |
+
"loss": 0.0164,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.3133208255159474,
|
| 497 |
+
"grad_norm": 0.16649533808231354,
|
| 498 |
+
"learning_rate": 5.969369490868042e-05,
|
| 499 |
+
"loss": 0.0154,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3320825515947468,
|
| 504 |
+
"grad_norm": 0.21355998516082764,
|
| 505 |
+
"learning_rate": 5.861002842519259e-05,
|
| 506 |
+
"loss": 0.0135,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3508442776735459,
|
| 511 |
+
"grad_norm": 0.203876331448555,
|
| 512 |
+
"learning_rate": 5.75221773122121e-05,
|
| 513 |
+
"loss": 0.0149,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3696060037523452,
|
| 518 |
+
"grad_norm": 0.20695096254348755,
|
| 519 |
+
"learning_rate": 5.6430670285049314e-05,
|
| 520 |
+
"loss": 0.0154,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3883677298311445,
|
| 525 |
+
"grad_norm": 0.2057274878025055,
|
| 526 |
+
"learning_rate": 5.533603783585496e-05,
|
| 527 |
+
"loss": 0.0159,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4071294559099436,
|
| 532 |
+
"grad_norm": 0.20767205953598022,
|
| 533 |
+
"learning_rate": 5.423881197579144e-05,
|
| 534 |
+
"loss": 0.0168,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.425891181988743,
|
| 539 |
+
"grad_norm": 0.16897110641002655,
|
| 540 |
+
"learning_rate": 5.313952597646568e-05,
|
| 541 |
+
"loss": 0.016,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.444652908067542,
|
| 546 |
+
"grad_norm": 0.2598036527633667,
|
| 547 |
+
"learning_rate": 5.203871411074954e-05,
|
| 548 |
+
"loss": 0.015,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4634146341463414,
|
| 553 |
+
"grad_norm": 0.2077157348394394,
|
| 554 |
+
"learning_rate": 5.093691139311356e-05,
|
| 555 |
+
"loss": 0.0138,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4821763602251408,
|
| 560 |
+
"grad_norm": 0.24033433198928833,
|
| 561 |
+
"learning_rate": 4.9834653319600246e-05,
|
| 562 |
+
"loss": 0.0135,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.50093808630394,
|
| 567 |
+
"grad_norm": 0.29492080211639404,
|
| 568 |
+
"learning_rate": 4.873247560756324e-05,
|
| 569 |
+
"loss": 0.0129,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.5196998123827392,
|
| 574 |
+
"grad_norm": 0.11821957677602768,
|
| 575 |
+
"learning_rate": 4.7630913935299066e-05,
|
| 576 |
+
"loss": 0.013,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.5384615384615383,
|
| 581 |
+
"grad_norm": 0.14121685922145844,
|
| 582 |
+
"learning_rate": 4.65305036816978e-05,
|
| 583 |
+
"loss": 0.0148,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5572232645403377,
|
| 588 |
+
"grad_norm": 0.17172347009181976,
|
| 589 |
+
"learning_rate": 4.543177966603925e-05,
|
| 590 |
+
"loss": 0.014,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.575984990619137,
|
| 595 |
+
"grad_norm": 0.17106831073760986,
|
| 596 |
+
"learning_rate": 4.433527588806103e-05,
|
| 597 |
+
"loss": 0.0119,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5947467166979363,
|
| 602 |
+
"grad_norm": 0.20430134236812592,
|
| 603 |
+
"learning_rate": 4.324152526842517e-05,
|
| 604 |
+
"loss": 0.0145,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.6135084427767354,
|
| 609 |
+
"grad_norm": 0.19569315016269684,
|
| 610 |
+
"learning_rate": 4.215105938970889e-05,
|
| 611 |
+
"loss": 0.0137,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6322701688555346,
|
| 616 |
+
"grad_norm": 0.16233587265014648,
|
| 617 |
+
"learning_rate": 4.1064408238045994e-05,
|
| 618 |
+
"loss": 0.0134,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6510318949343339,
|
| 623 |
+
"grad_norm": 0.16989213228225708,
|
| 624 |
+
"learning_rate": 3.9982099945543945e-05,
|
| 625 |
+
"loss": 0.0126,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6697936210131332,
|
| 630 |
+
"grad_norm": 0.18682365119457245,
|
| 631 |
+
"learning_rate": 3.890466053360211e-05,
|
| 632 |
+
"loss": 0.0115,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.6885553470919326,
|
| 637 |
+
"grad_norm": 0.20883384346961975,
|
| 638 |
+
"learning_rate": 3.783261365725592e-05,
|
| 639 |
+
"loss": 0.0124,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7073170731707317,
|
| 644 |
+
"grad_norm": 0.21428252756595612,
|
| 645 |
+
"learning_rate": 3.676648035067093e-05,
|
| 646 |
+
"loss": 0.0122,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.726078799249531,
|
| 651 |
+
"grad_norm": 0.20880082249641418,
|
| 652 |
+
"learning_rate": 3.570677877391092e-05,
|
| 653 |
+
"loss": 0.0125,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7448405253283301,
|
| 658 |
+
"grad_norm": 0.16143251955509186,
|
| 659 |
+
"learning_rate": 3.465402396110269e-05,
|
| 660 |
+
"loss": 0.012,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7636022514071295,
|
| 665 |
+
"grad_norm": 0.1445058137178421,
|
| 666 |
+
"learning_rate": 3.360872757012011e-05,
|
| 667 |
+
"loss": 0.0123,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.7823639774859288,
|
| 672 |
+
"grad_norm": 0.21280893683433533,
|
| 673 |
+
"learning_rate": 3.257139763390925e-05,
|
| 674 |
+
"loss": 0.0125,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.8011257035647281,
|
| 679 |
+
"grad_norm": 0.16495130956172943,
|
| 680 |
+
"learning_rate": 3.1542538313575035e-05,
|
| 681 |
+
"loss": 0.0116,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.8198874296435272,
|
| 686 |
+
"grad_norm": 0.19219523668289185,
|
| 687 |
+
"learning_rate": 3.052264965335e-05,
|
| 688 |
+
"loss": 0.0133,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8386491557223263,
|
| 693 |
+
"grad_norm": 0.173895001411438,
|
| 694 |
+
"learning_rate": 2.9512227337563604e-05,
|
| 695 |
+
"loss": 0.0116,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8574108818011257,
|
| 700 |
+
"grad_norm": 0.16958042979240417,
|
| 701 |
+
"learning_rate": 2.8511762449730795e-05,
|
| 702 |
+
"loss": 0.0125,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.876172607879925,
|
| 707 |
+
"grad_norm": 0.1639089733362198,
|
| 708 |
+
"learning_rate": 2.7521741233876496e-05,
|
| 709 |
+
"loss": 0.0122,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8949343339587243,
|
| 714 |
+
"grad_norm": 0.16494053602218628,
|
| 715 |
+
"learning_rate": 2.654264485821214e-05,
|
| 716 |
+
"loss": 0.011,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9136960600375235,
|
| 721 |
+
"grad_norm": 0.17180362343788147,
|
| 722 |
+
"learning_rate": 2.55749491812794e-05,
|
| 723 |
+
"loss": 0.0113,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9324577861163226,
|
| 728 |
+
"grad_norm": 0.19289036095142365,
|
| 729 |
+
"learning_rate": 2.4619124520674146e-05,
|
| 730 |
+
"loss": 0.0131,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.951219512195122,
|
| 735 |
+
"grad_norm": 0.1478220820426941,
|
| 736 |
+
"learning_rate": 2.3675635424463754e-05,
|
| 737 |
+
"loss": 0.0097,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.9699812382739212,
|
| 742 |
+
"grad_norm": 0.18471956253051758,
|
| 743 |
+
"learning_rate": 2.2744940445408202e-05,
|
| 744 |
+
"loss": 0.0118,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.9887429643527206,
|
| 749 |
+
"grad_norm": 0.16719312965869904,
|
| 750 |
+
"learning_rate": 2.182749191809518e-05,
|
| 751 |
+
"loss": 0.0104,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.00750469043152,
|
| 756 |
+
"grad_norm": 0.15944059193134308,
|
| 757 |
+
"learning_rate": 2.09237357390974e-05,
|
| 758 |
+
"loss": 0.0109,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 2.026266416510319,
|
| 763 |
+
"grad_norm": 0.20150484144687653,
|
| 764 |
+
"learning_rate": 2.0034111150258666e-05,
|
| 765 |
+
"loss": 0.0129,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.045028142589118,
|
| 770 |
+
"grad_norm": 0.15957467257976532,
|
| 771 |
+
"learning_rate": 1.9159050525214452e-05,
|
| 772 |
+
"loss": 0.011,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 2.0637898686679175,
|
| 777 |
+
"grad_norm": 0.133390411734581,
|
| 778 |
+
"learning_rate": 1.8298979159250557e-05,
|
| 779 |
+
"loss": 0.0103,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.082551594746717,
|
| 784 |
+
"grad_norm": 0.13808146119117737,
|
| 785 |
+
"learning_rate": 1.745431506260173e-05,
|
| 786 |
+
"loss": 0.0115,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 2.101313320825516,
|
| 791 |
+
"grad_norm": 0.16296911239624023,
|
| 792 |
+
"learning_rate": 1.662546875729138e-05,
|
| 793 |
+
"loss": 0.0112,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.120075046904315,
|
| 798 |
+
"grad_norm": 0.13522660732269287,
|
| 799 |
+
"learning_rate": 1.581284307761024e-05,
|
| 800 |
+
"loss": 0.0106,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.1388367729831144,
|
| 805 |
+
"grad_norm": 0.19323870539665222,
|
| 806 |
+
"learning_rate": 1.5016832974331724e-05,
|
| 807 |
+
"loss": 0.0123,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.1575984990619137,
|
| 812 |
+
"grad_norm": 0.1635001301765442,
|
| 813 |
+
"learning_rate": 1.4237825322758736e-05,
|
| 814 |
+
"loss": 0.013,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 2.176360225140713,
|
| 819 |
+
"grad_norm": 0.17513315379619598,
|
| 820 |
+
"learning_rate": 1.3476198734695122e-05,
|
| 821 |
+
"loss": 0.0104,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.1951219512195124,
|
| 826 |
+
"grad_norm": 0.10277465730905533,
|
| 827 |
+
"learning_rate": 1.2732323374433707e-05,
|
| 828 |
+
"loss": 0.0105,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 2.2138836772983113,
|
| 833 |
+
"grad_norm": 0.1144745722413063,
|
| 834 |
+
"learning_rate": 1.2006560778849578e-05,
|
| 835 |
+
"loss": 0.0103,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.2326454033771106,
|
| 840 |
+
"grad_norm": 0.1264655739068985,
|
| 841 |
+
"learning_rate": 1.1299263681686706e-05,
|
| 842 |
+
"loss": 0.0119,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 2.25140712945591,
|
| 847 |
+
"grad_norm": 0.10604596883058548,
|
| 848 |
+
"learning_rate": 1.0610775842122972e-05,
|
| 849 |
+
"loss": 0.0109,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.2701688555347093,
|
| 854 |
+
"grad_norm": 0.14508432149887085,
|
| 855 |
+
"learning_rate": 9.941431877696955e-06,
|
| 856 |
+
"loss": 0.0109,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.2889305816135086,
|
| 861 |
+
"grad_norm": 0.12284737080335617,
|
| 862 |
+
"learning_rate": 9.291557101677784e-06,
|
| 863 |
+
"loss": 0.0093,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.3076923076923075,
|
| 868 |
+
"grad_norm": 0.1272381991147995,
|
| 869 |
+
"learning_rate": 8.661467364956993e-06,
|
| 870 |
+
"loss": 0.0116,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.326454033771107,
|
| 875 |
+
"grad_norm": 0.14646074175834656,
|
| 876 |
+
"learning_rate": 8.051468902539272e-06,
|
| 877 |
+
"loss": 0.0102,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.345215759849906,
|
| 882 |
+
"grad_norm": 0.11431489139795303,
|
| 883 |
+
"learning_rate": 7.461858184706777e-06,
|
| 884 |
+
"loss": 0.0091,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.3639774859287055,
|
| 889 |
+
"grad_norm": 0.1557149738073349,
|
| 890 |
+
"learning_rate": 6.892921772929112e-06,
|
| 891 |
+
"loss": 0.0097,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.382739212007505,
|
| 896 |
+
"grad_norm": 0.14029847085475922,
|
| 897 |
+
"learning_rate": 6.344936180589351e-06,
|
| 898 |
+
"loss": 0.0108,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.401500938086304,
|
| 903 |
+
"grad_norm": 0.11800851672887802,
|
| 904 |
+
"learning_rate": 5.818167738593505e-06,
|
| 905 |
+
"loss": 0.0098,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.420262664165103,
|
| 910 |
+
"grad_norm": 0.11177805066108704,
|
| 911 |
+
"learning_rate": 5.312872465928881e-06,
|
| 912 |
+
"loss": 0.0093,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.4390243902439024,
|
| 917 |
+
"grad_norm": 0.12152400612831116,
|
| 918 |
+
"learning_rate": 4.829295945234258e-06,
|
| 919 |
+
"loss": 0.0099,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.4577861163227017,
|
| 924 |
+
"grad_norm": 0.10167639702558517,
|
| 925 |
+
"learning_rate": 4.367673203442241e-06,
|
| 926 |
+
"loss": 0.0096,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.476547842401501,
|
| 931 |
+
"grad_norm": 0.13529738783836365,
|
| 932 |
+
"learning_rate": 3.928228597551947e-06,
|
| 933 |
+
"loss": 0.0099,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.4953095684803,
|
| 938 |
+
"grad_norm": 0.10373085737228394,
|
| 939 |
+
"learning_rate": 3.511175705587433e-06,
|
| 940 |
+
"loss": 0.0088,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.5140712945590993,
|
| 945 |
+
"grad_norm": 0.13128240406513214,
|
| 946 |
+
"learning_rate": 3.1167172227949347e-06,
|
| 947 |
+
"loss": 0.0113,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.5328330206378986,
|
| 952 |
+
"grad_norm": 0.11484728753566742,
|
| 953 |
+
"learning_rate": 2.7450448631293036e-06,
|
| 954 |
+
"loss": 0.0088,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.551594746716698,
|
| 959 |
+
"grad_norm": 0.1032821536064148,
|
| 960 |
+
"learning_rate": 2.3963392660775575e-06,
|
| 961 |
+
"loss": 0.0088,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.5703564727954973,
|
| 966 |
+
"grad_norm": 0.10594484210014343,
|
| 967 |
+
"learning_rate": 2.0707699088647836e-06,
|
| 968 |
+
"loss": 0.0099,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.5891181988742966,
|
| 973 |
+
"grad_norm": 0.0838296115398407,
|
| 974 |
+
"learning_rate": 1.7684950240852372e-06,
|
| 975 |
+
"loss": 0.0088,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.607879924953096,
|
| 980 |
+
"grad_norm": 0.09543336927890778,
|
| 981 |
+
"learning_rate": 1.4896615227983468e-06,
|
| 982 |
+
"loss": 0.0091,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.626641651031895,
|
| 987 |
+
"grad_norm": 0.1242263913154602,
|
| 988 |
+
"learning_rate": 1.2344049231273302e-06,
|
| 989 |
+
"loss": 0.0092,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.645403377110694,
|
| 994 |
+
"grad_norm": 0.1851985901594162,
|
| 995 |
+
"learning_rate": 1.00284928439493e-06,
|
| 996 |
+
"loss": 0.0106,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.6641651031894935,
|
| 1001 |
+
"grad_norm": 0.08041252940893173,
|
| 1002 |
+
"learning_rate": 7.951071468283167e-07,
|
| 1003 |
+
"loss": 0.0098,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.682926829268293,
|
| 1008 |
+
"grad_norm": 0.11750051379203796,
|
| 1009 |
+
"learning_rate": 6.11279476862553e-07,
|
| 1010 |
+
"loss": 0.0096,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.7016885553470917,
|
| 1015 |
+
"grad_norm": 0.08895982056856155,
|
| 1016 |
+
"learning_rate": 4.514556180690188e-07,
|
| 1017 |
+
"loss": 0.0093,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.720450281425891,
|
| 1022 |
+
"grad_norm": 0.11088193953037262,
|
| 1023 |
+
"learning_rate": 3.157132477328628e-07,
|
| 1024 |
+
"loss": 0.0108,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.7392120075046904,
|
| 1029 |
+
"grad_norm": 0.08702103048563004,
|
| 1030 |
+
"learning_rate": 2.041183391004453e-07,
|
| 1031 |
+
"loss": 0.0083,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.7579737335834897,
|
| 1036 |
+
"grad_norm": 0.10216435045003891,
|
| 1037 |
+
"learning_rate": 1.1672512931509283e-07,
|
| 1038 |
+
"loss": 0.0098,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.776735459662289,
|
| 1043 |
+
"grad_norm": 0.11117573082447052,
|
| 1044 |
+
"learning_rate": 5.3576093056922906e-08,
|
| 1045 |
+
"loss": 0.0097,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.7954971857410884,
|
| 1050 |
+
"grad_norm": 0.1127699688076973,
|
| 1051 |
+
"learning_rate": 1.4701921899362703e-08,
|
| 1052 |
+
"loss": 0.01,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.8142589118198873,
|
| 1057 |
+
"grad_norm": 0.15439969301223755,
|
| 1058 |
+
"learning_rate": 1.2150939247002058e-10,
|
| 1059 |
+
"loss": 0.0112,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.8142589118198873,
|
| 1064 |
+
"step": 1500,
|
| 1065 |
+
"total_flos": 0.0,
|
| 1066 |
+
"train_loss": 0.0293558024764061,
|
| 1067 |
+
"train_runtime": 1931.3019,
|
| 1068 |
+
"train_samples_per_second": 77.668,
|
| 1069 |
+
"train_steps_per_second": 0.777
|
| 1070 |
+
}
|
| 1071 |
+
],
|
| 1072 |
+
"logging_steps": 10,
|
| 1073 |
+
"max_steps": 1500,
|
| 1074 |
+
"num_input_tokens_seen": 0,
|
| 1075 |
+
"num_train_epochs": 3,
|
| 1076 |
+
"save_steps": 1000,
|
| 1077 |
+
"stateful_callbacks": {
|
| 1078 |
+
"TrainerControl": {
|
| 1079 |
+
"args": {
|
| 1080 |
+
"should_epoch_stop": false,
|
| 1081 |
+
"should_evaluate": false,
|
| 1082 |
+
"should_log": false,
|
| 1083 |
+
"should_save": true,
|
| 1084 |
+
"should_training_stop": true
|
| 1085 |
+
},
|
| 1086 |
+
"attributes": {}
|
| 1087 |
+
}
|
| 1088 |
+
},
|
| 1089 |
+
"total_flos": 0.0,
|
| 1090 |
+
"train_batch_size": 100,
|
| 1091 |
+
"trial_name": null,
|
| 1092 |
+
"trial_params": null
|
| 1093 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4bc0b3fd77b54b6fd8685a4e7088333a4dadd2c02c9b11dfbf3a3f4b4e9075b3
|
| 3 |
+
size 5304
|