Upload folder using huggingface_hub
Browse files- config.json +64 -0
- experiment_cfg/metadata.json +363 -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 +1784 -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 +1,363 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"new_embodiment": {
|
| 3 |
+
"statistics": {
|
| 4 |
+
"state": {
|
| 5 |
+
"left_arm": {
|
| 6 |
+
"max": [
|
| 7 |
+
78.35006713867188,
|
| 8 |
+
73.41722106933594,
|
| 9 |
+
100.0,
|
| 10 |
+
75.20030212402344,
|
| 11 |
+
99.85347747802734
|
| 12 |
+
],
|
| 13 |
+
"min": [
|
| 14 |
+
-21.25709342956543,
|
| 15 |
+
-99.91614532470703,
|
| 16 |
+
-34.46580123901367,
|
| 17 |
+
-98.9417953491211,
|
| 18 |
+
-95.9951171875
|
| 19 |
+
],
|
| 20 |
+
"mean": [
|
| 21 |
+
22.03197479248047,
|
| 22 |
+
-36.339046478271484,
|
| 23 |
+
68.31412506103516,
|
| 24 |
+
5.61252498626709,
|
| 25 |
+
-22.224000930786133
|
| 26 |
+
],
|
| 27 |
+
"std": [
|
| 28 |
+
16.480056762695312,
|
| 29 |
+
58.42597961425781,
|
| 30 |
+
33.54707336425781,
|
| 31 |
+
54.855201721191406,
|
| 32 |
+
32.123435974121094
|
| 33 |
+
],
|
| 34 |
+
"q01": [
|
| 35 |
+
-6.453950185775753,
|
| 36 |
+
-99.24528503417969,
|
| 37 |
+
-16.04827880859375,
|
| 38 |
+
-98.9417953491211,
|
| 39 |
+
-63.19560352325439
|
| 40 |
+
],
|
| 41 |
+
"q99": [
|
| 42 |
+
78.35006713867188,
|
| 43 |
+
62.767295837402344,
|
| 44 |
+
99.63948059082031,
|
| 45 |
+
74.149658203125,
|
| 46 |
+
99.85347747802734
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"gripper1": {
|
| 50 |
+
"max": [
|
| 51 |
+
97.7272720336914
|
| 52 |
+
],
|
| 53 |
+
"min": [
|
| 54 |
+
0.14204545319080353
|
| 55 |
+
],
|
| 56 |
+
"mean": [
|
| 57 |
+
7.044379234313965
|
| 58 |
+
],
|
| 59 |
+
"std": [
|
| 60 |
+
14.734519004821777
|
| 61 |
+
],
|
| 62 |
+
"q01": [
|
| 63 |
+
0.14204545319080353
|
| 64 |
+
],
|
| 65 |
+
"q99": [
|
| 66 |
+
74.32031150817873
|
| 67 |
+
]
|
| 68 |
+
},
|
| 69 |
+
"right_arm": {
|
| 70 |
+
"max": [
|
| 71 |
+
7.535795211791992,
|
| 72 |
+
84.6780776977539,
|
| 73 |
+
99.46284484863281,
|
| 74 |
+
74.34782409667969,
|
| 75 |
+
30.256410598754883
|
| 76 |
+
],
|
| 77 |
+
"min": [
|
| 78 |
+
-31.602048873901367,
|
| 79 |
+
-98.00664520263672,
|
| 80 |
+
-39.03837585449219,
|
| 81 |
+
-98.17805480957031,
|
| 82 |
+
-67.23494720458984
|
| 83 |
+
],
|
| 84 |
+
"mean": [
|
| 85 |
+
-3.5733110904693604,
|
| 86 |
+
-42.38075637817383,
|
| 87 |
+
75.2816390991211,
|
| 88 |
+
9.783737182617188,
|
| 89 |
+
-44.31592559814453
|
| 90 |
+
],
|
| 91 |
+
"std": [
|
| 92 |
+
10.12617015838623,
|
| 93 |
+
69.33080291748047,
|
| 94 |
+
31.52765464782715,
|
| 95 |
+
71.03961181640625,
|
| 96 |
+
25.092512130737305
|
| 97 |
+
],
|
| 98 |
+
"q01": [
|
| 99 |
+
-24.43306541442871,
|
| 100 |
+
-97.91144561767578,
|
| 101 |
+
-7.278341293334961,
|
| 102 |
+
-97.5155258178711,
|
| 103 |
+
-61.98304748535156
|
| 104 |
+
],
|
| 105 |
+
"q99": [
|
| 106 |
+
6.932931423187256,
|
| 107 |
+
68.21515655517578,
|
| 108 |
+
99.46284484863281,
|
| 109 |
+
74.34782409667969,
|
| 110 |
+
20.407204513549807
|
| 111 |
+
]
|
| 112 |
+
},
|
| 113 |
+
"gripper2": {
|
| 114 |
+
"max": [
|
| 115 |
+
71.49090576171875
|
| 116 |
+
],
|
| 117 |
+
"min": [
|
| 118 |
+
1.3722126483917236
|
| 119 |
+
],
|
| 120 |
+
"mean": [
|
| 121 |
+
12.590181350708008
|
| 122 |
+
],
|
| 123 |
+
"std": [
|
| 124 |
+
11.938528060913086
|
| 125 |
+
],
|
| 126 |
+
"q01": [
|
| 127 |
+
1.5556938648223877
|
| 128 |
+
],
|
| 129 |
+
"q99": [
|
| 130 |
+
48.94545364379883
|
| 131 |
+
]
|
| 132 |
+
}
|
| 133 |
+
},
|
| 134 |
+
"action": {
|
| 135 |
+
"left_arm": {
|
| 136 |
+
"max": [
|
| 137 |
+
78.18477630615234,
|
| 138 |
+
81.9516372680664,
|
| 139 |
+
100.0,
|
| 140 |
+
78.87323760986328,
|
| 141 |
+
100.0
|
| 142 |
+
],
|
| 143 |
+
"min": [
|
| 144 |
+
-21.44892120361328,
|
| 145 |
+
-100.0,
|
| 146 |
+
-37.03024673461914,
|
| 147 |
+
-100.0,
|
| 148 |
+
-100.0
|
| 149 |
+
],
|
| 150 |
+
"mean": [
|
| 151 |
+
22.302133560180664,
|
| 152 |
+
-36.885658264160156,
|
| 153 |
+
67.65310668945312,
|
| 154 |
+
5.681365013122559,
|
| 155 |
+
-22.29690170288086
|
| 156 |
+
],
|
| 157 |
+
"std": [
|
| 158 |
+
16.664201736450195,
|
| 159 |
+
57.817176818847656,
|
| 160 |
+
34.0733642578125,
|
| 161 |
+
55.35456466674805,
|
| 162 |
+
32.176795959472656
|
| 163 |
+
],
|
| 164 |
+
"q01": [
|
| 165 |
+
-6.389906406402588,
|
| 166 |
+
-99.91173553466797,
|
| 167 |
+
-17.40612455368042,
|
| 168 |
+
-100.0,
|
| 169 |
+
-64.58546447753906
|
| 170 |
+
],
|
| 171 |
+
"q99": [
|
| 172 |
+
78.18477630615234,
|
| 173 |
+
62.26213584899905,
|
| 174 |
+
99.81668090820312,
|
| 175 |
+
76.96768951416016,
|
| 176 |
+
100.0
|
| 177 |
+
]
|
| 178 |
+
},
|
| 179 |
+
"gripper1": {
|
| 180 |
+
"max": [
|
| 181 |
+
97.76386260986328
|
| 182 |
+
],
|
| 183 |
+
"min": [
|
| 184 |
+
0.0
|
| 185 |
+
],
|
| 186 |
+
"mean": [
|
| 187 |
+
5.872297286987305
|
| 188 |
+
],
|
| 189 |
+
"std": [
|
| 190 |
+
15.169022560119629
|
| 191 |
+
],
|
| 192 |
+
"q01": [
|
| 193 |
+
0.0
|
| 194 |
+
],
|
| 195 |
+
"q99": [
|
| 196 |
+
74.85331123352056
|
| 197 |
+
]
|
| 198 |
+
},
|
| 199 |
+
"right_arm": {
|
| 200 |
+
"max": [
|
| 201 |
+
8.018280029296875,
|
| 202 |
+
84.90252685546875,
|
| 203 |
+
99.72960662841797,
|
| 204 |
+
74.73233032226562,
|
| 205 |
+
30.77525520324707
|
| 206 |
+
],
|
| 207 |
+
"min": [
|
| 208 |
+
-31.616119384765625,
|
| 209 |
+
-99.087890625,
|
| 210 |
+
-39.26325225830078,
|
| 211 |
+
-99.57283020019531,
|
| 212 |
+
-67.18055725097656
|
| 213 |
+
],
|
| 214 |
+
"mean": [
|
| 215 |
+
-3.350882053375244,
|
| 216 |
+
-42.753501892089844,
|
| 217 |
+
75.27497100830078,
|
| 218 |
+
9.843159675598145,
|
| 219 |
+
-44.773658752441406
|
| 220 |
+
],
|
| 221 |
+
"std": [
|
| 222 |
+
10.194260597229004,
|
| 223 |
+
68.84522247314453,
|
| 224 |
+
32.172489166259766,
|
| 225 |
+
71.25251770019531,
|
| 226 |
+
25.30291748046875
|
| 227 |
+
],
|
| 228 |
+
"q01": [
|
| 229 |
+
-24.304113388061523,
|
| 230 |
+
-97.5953598022461,
|
| 231 |
+
-8.574124078750605,
|
| 232 |
+
-97.60785675048828,
|
| 233 |
+
-61.92479705810547
|
| 234 |
+
],
|
| 235 |
+
"q99": [
|
| 236 |
+
7.213930130004883,
|
| 237 |
+
69.39029693603516,
|
| 238 |
+
99.72960662841797,
|
| 239 |
+
74.73233032226562,
|
| 240 |
+
20.75176239013672
|
| 241 |
+
]
|
| 242 |
+
},
|
| 243 |
+
"gripper2": {
|
| 244 |
+
"max": [
|
| 245 |
+
72.06661987304688
|
| 246 |
+
],
|
| 247 |
+
"min": [
|
| 248 |
+
0.0
|
| 249 |
+
],
|
| 250 |
+
"mean": [
|
| 251 |
+
11.008831977844238
|
| 252 |
+
],
|
| 253 |
+
"std": [
|
| 254 |
+
12.901287078857422
|
| 255 |
+
],
|
| 256 |
+
"q01": [
|
| 257 |
+
0.9084027409553528
|
| 258 |
+
],
|
| 259 |
+
"q99": [
|
| 260 |
+
48.82664489746094
|
| 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 |
+
"left": {
|
| 276 |
+
"resolution": [
|
| 277 |
+
640,
|
| 278 |
+
480
|
| 279 |
+
],
|
| 280 |
+
"channels": 3,
|
| 281 |
+
"fps": 30.0
|
| 282 |
+
},
|
| 283 |
+
"top_depth": {
|
| 284 |
+
"resolution": [
|
| 285 |
+
640,
|
| 286 |
+
480
|
| 287 |
+
],
|
| 288 |
+
"channels": 3,
|
| 289 |
+
"fps": 30.0
|
| 290 |
+
}
|
| 291 |
+
},
|
| 292 |
+
"state": {
|
| 293 |
+
"left_arm": {
|
| 294 |
+
"absolute": true,
|
| 295 |
+
"rotation_type": null,
|
| 296 |
+
"shape": [
|
| 297 |
+
5
|
| 298 |
+
],
|
| 299 |
+
"continuous": true
|
| 300 |
+
},
|
| 301 |
+
"gripper1": {
|
| 302 |
+
"absolute": true,
|
| 303 |
+
"rotation_type": null,
|
| 304 |
+
"shape": [
|
| 305 |
+
1
|
| 306 |
+
],
|
| 307 |
+
"continuous": true
|
| 308 |
+
},
|
| 309 |
+
"right_arm": {
|
| 310 |
+
"absolute": true,
|
| 311 |
+
"rotation_type": null,
|
| 312 |
+
"shape": [
|
| 313 |
+
5
|
| 314 |
+
],
|
| 315 |
+
"continuous": true
|
| 316 |
+
},
|
| 317 |
+
"gripper2": {
|
| 318 |
+
"absolute": true,
|
| 319 |
+
"rotation_type": null,
|
| 320 |
+
"shape": [
|
| 321 |
+
1
|
| 322 |
+
],
|
| 323 |
+
"continuous": true
|
| 324 |
+
}
|
| 325 |
+
},
|
| 326 |
+
"action": {
|
| 327 |
+
"left_arm": {
|
| 328 |
+
"absolute": true,
|
| 329 |
+
"rotation_type": null,
|
| 330 |
+
"shape": [
|
| 331 |
+
5
|
| 332 |
+
],
|
| 333 |
+
"continuous": true
|
| 334 |
+
},
|
| 335 |
+
"gripper1": {
|
| 336 |
+
"absolute": true,
|
| 337 |
+
"rotation_type": null,
|
| 338 |
+
"shape": [
|
| 339 |
+
1
|
| 340 |
+
],
|
| 341 |
+
"continuous": true
|
| 342 |
+
},
|
| 343 |
+
"right_arm": {
|
| 344 |
+
"absolute": true,
|
| 345 |
+
"rotation_type": null,
|
| 346 |
+
"shape": [
|
| 347 |
+
5
|
| 348 |
+
],
|
| 349 |
+
"continuous": true
|
| 350 |
+
},
|
| 351 |
+
"gripper2": {
|
| 352 |
+
"absolute": true,
|
| 353 |
+
"rotation_type": null,
|
| 354 |
+
"shape": [
|
| 355 |
+
1
|
| 356 |
+
],
|
| 357 |
+
"continuous": true
|
| 358 |
+
}
|
| 359 |
+
}
|
| 360 |
+
},
|
| 361 |
+
"embodiment_tag": "new_embodiment"
|
| 362 |
+
}
|
| 363 |
+
}
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dae0c19967fcdcd1b8c687028e215005cb1d11a596f80be77056c6c720d274f6
|
| 3 |
+
size 4999367032
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:494201c95f43b84fb822c750056d5d0e023742800eb14922e3b5b480d23e9381
|
| 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:eff09be57e724db3a4553ae0c9a099e01ec7550953b37bddb32027c324398147
|
| 3 |
+
size 8550719959
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a65d51e972099d01f624474088e776989a2e2a7b2a795f7b76c5dff41a84a52b
|
| 3 |
+
size 14645
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e79479860d3115689987260f01630c7535af7b264b07b02c59547f0d87a1984
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1784 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.36997299197158606,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0014798919678863443,
|
| 14 |
+
"grad_norm": 7.47622537612915,
|
| 15 |
+
"learning_rate": 7.2e-06,
|
| 16 |
+
"loss": 1.1232,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0029597839357726886,
|
| 21 |
+
"grad_norm": 1.880067229270935,
|
| 22 |
+
"learning_rate": 1.52e-05,
|
| 23 |
+
"loss": 0.5006,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.004439675903659033,
|
| 28 |
+
"grad_norm": 1.9676536321640015,
|
| 29 |
+
"learning_rate": 2.32e-05,
|
| 30 |
+
"loss": 0.2438,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.005919567871545377,
|
| 35 |
+
"grad_norm": 1.628406047821045,
|
| 36 |
+
"learning_rate": 3.12e-05,
|
| 37 |
+
"loss": 0.1641,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.007399459839431722,
|
| 42 |
+
"grad_norm": 2.5903799533843994,
|
| 43 |
+
"learning_rate": 3.9200000000000004e-05,
|
| 44 |
+
"loss": 0.1585,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.008879351807318065,
|
| 49 |
+
"grad_norm": 1.520799160003662,
|
| 50 |
+
"learning_rate": 4.72e-05,
|
| 51 |
+
"loss": 0.1282,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.01035924377520441,
|
| 56 |
+
"grad_norm": 1.7377681732177734,
|
| 57 |
+
"learning_rate": 5.520000000000001e-05,
|
| 58 |
+
"loss": 0.1256,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.011839135743090754,
|
| 63 |
+
"grad_norm": 1.098388433456421,
|
| 64 |
+
"learning_rate": 6.32e-05,
|
| 65 |
+
"loss": 0.1031,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.013319027710977099,
|
| 70 |
+
"grad_norm": 1.4509732723236084,
|
| 71 |
+
"learning_rate": 7.12e-05,
|
| 72 |
+
"loss": 0.0973,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.014798919678863444,
|
| 77 |
+
"grad_norm": 1.9007303714752197,
|
| 78 |
+
"learning_rate": 7.920000000000001e-05,
|
| 79 |
+
"loss": 0.1152,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.016278811646749786,
|
| 84 |
+
"grad_norm": 1.5740575790405273,
|
| 85 |
+
"learning_rate": 8.72e-05,
|
| 86 |
+
"loss": 0.1151,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.01775870361463613,
|
| 91 |
+
"grad_norm": 1.2793081998825073,
|
| 92 |
+
"learning_rate": 9.52e-05,
|
| 93 |
+
"loss": 0.1182,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.019238595582522475,
|
| 98 |
+
"grad_norm": 1.5838185548782349,
|
| 99 |
+
"learning_rate": 9.999930010724872e-05,
|
| 100 |
+
"loss": 0.0955,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.02071848755040882,
|
| 105 |
+
"grad_norm": 1.8031917810440063,
|
| 106 |
+
"learning_rate": 9.999142653881985e-05,
|
| 107 |
+
"loss": 0.1141,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.022198379518295164,
|
| 112 |
+
"grad_norm": 1.5219581127166748,
|
| 113 |
+
"learning_rate": 9.997480591826183e-05,
|
| 114 |
+
"loss": 0.1069,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.02367827148618151,
|
| 119 |
+
"grad_norm": 1.1764310598373413,
|
| 120 |
+
"learning_rate": 9.994944115370199e-05,
|
| 121 |
+
"loss": 0.0964,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.025158163454067854,
|
| 126 |
+
"grad_norm": 1.0616095066070557,
|
| 127 |
+
"learning_rate": 9.991533668323974e-05,
|
| 128 |
+
"loss": 0.0761,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.026638055421954198,
|
| 133 |
+
"grad_norm": 1.1210483312606812,
|
| 134 |
+
"learning_rate": 9.987249847416987e-05,
|
| 135 |
+
"loss": 0.087,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.028117947389840543,
|
| 140 |
+
"grad_norm": 1.1077345609664917,
|
| 141 |
+
"learning_rate": 9.982093402193857e-05,
|
| 142 |
+
"loss": 0.0931,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.029597839357726887,
|
| 147 |
+
"grad_norm": 1.352004051208496,
|
| 148 |
+
"learning_rate": 9.976065234883193e-05,
|
| 149 |
+
"loss": 0.083,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.03107773132561323,
|
| 154 |
+
"grad_norm": 1.0884106159210205,
|
| 155 |
+
"learning_rate": 9.969166400239726e-05,
|
| 156 |
+
"loss": 0.0783,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.03255762329349957,
|
| 161 |
+
"grad_norm": 1.1031562089920044,
|
| 162 |
+
"learning_rate": 9.961398105359764e-05,
|
| 163 |
+
"loss": 0.1047,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.03403751526138592,
|
| 168 |
+
"grad_norm": 1.0741270780563354,
|
| 169 |
+
"learning_rate": 9.952761709469975e-05,
|
| 170 |
+
"loss": 0.1017,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.03551740722927226,
|
| 175 |
+
"grad_norm": 1.2804232835769653,
|
| 176 |
+
"learning_rate": 9.94325872368957e-05,
|
| 177 |
+
"loss": 0.0968,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.036997299197158606,
|
| 182 |
+
"grad_norm": 1.0856305360794067,
|
| 183 |
+
"learning_rate": 9.932890810765902e-05,
|
| 184 |
+
"loss": 0.0644,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.03847719116504495,
|
| 189 |
+
"grad_norm": 1.1068660020828247,
|
| 190 |
+
"learning_rate": 9.921659784783526e-05,
|
| 191 |
+
"loss": 0.0644,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.039957083132931295,
|
| 196 |
+
"grad_norm": 1.043481469154358,
|
| 197 |
+
"learning_rate": 9.909567610846788e-05,
|
| 198 |
+
"loss": 0.0667,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.04143697510081764,
|
| 203 |
+
"grad_norm": 1.3862495422363281,
|
| 204 |
+
"learning_rate": 9.896616404736001e-05,
|
| 205 |
+
"loss": 0.0694,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.042916867068703984,
|
| 210 |
+
"grad_norm": 1.405779480934143,
|
| 211 |
+
"learning_rate": 9.882808432537224e-05,
|
| 212 |
+
"loss": 0.0823,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.04439675903659033,
|
| 217 |
+
"grad_norm": 1.4005481004714966,
|
| 218 |
+
"learning_rate": 9.86814611024578e-05,
|
| 219 |
+
"loss": 0.0812,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.04587665100447667,
|
| 224 |
+
"grad_norm": 0.8486709594726562,
|
| 225 |
+
"learning_rate": 9.852632003343518e-05,
|
| 226 |
+
"loss": 0.0966,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.04735654297236302,
|
| 231 |
+
"grad_norm": 0.6197394132614136,
|
| 232 |
+
"learning_rate": 9.836268826349933e-05,
|
| 233 |
+
"loss": 0.0595,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.04883643494024936,
|
| 238 |
+
"grad_norm": 0.776345431804657,
|
| 239 |
+
"learning_rate": 9.819059442347193e-05,
|
| 240 |
+
"loss": 0.0599,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.05031632690813571,
|
| 245 |
+
"grad_norm": 0.7929945588111877,
|
| 246 |
+
"learning_rate": 9.801006862479202e-05,
|
| 247 |
+
"loss": 0.07,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.05179621887602205,
|
| 252 |
+
"grad_norm": 0.9411725997924805,
|
| 253 |
+
"learning_rate": 9.782114245424718e-05,
|
| 254 |
+
"loss": 0.0637,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.053276110843908396,
|
| 259 |
+
"grad_norm": 0.825547993183136,
|
| 260 |
+
"learning_rate": 9.762384896844684e-05,
|
| 261 |
+
"loss": 0.0628,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.05475600281179474,
|
| 266 |
+
"grad_norm": 1.1270116567611694,
|
| 267 |
+
"learning_rate": 9.741822268803833e-05,
|
| 268 |
+
"loss": 0.0702,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.056235894779681085,
|
| 273 |
+
"grad_norm": 0.947913408279419,
|
| 274 |
+
"learning_rate": 9.720429959166675e-05,
|
| 275 |
+
"loss": 0.073,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.05771578674756743,
|
| 280 |
+
"grad_norm": 1.0599220991134644,
|
| 281 |
+
"learning_rate": 9.69821171096798e-05,
|
| 282 |
+
"loss": 0.0678,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.059195678715453774,
|
| 287 |
+
"grad_norm": 1.0569720268249512,
|
| 288 |
+
"learning_rate": 9.675171411757842e-05,
|
| 289 |
+
"loss": 0.0653,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.06067557068334012,
|
| 294 |
+
"grad_norm": 1.1427547931671143,
|
| 295 |
+
"learning_rate": 9.65131309292149e-05,
|
| 296 |
+
"loss": 0.0739,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.06215546265122646,
|
| 301 |
+
"grad_norm": 0.816104531288147,
|
| 302 |
+
"learning_rate": 9.626640928973892e-05,
|
| 303 |
+
"loss": 0.0518,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.06363535461911281,
|
| 308 |
+
"grad_norm": 0.7173371911048889,
|
| 309 |
+
"learning_rate": 9.601159236829352e-05,
|
| 310 |
+
"loss": 0.0687,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.06511524658699915,
|
| 315 |
+
"grad_norm": 0.8344293236732483,
|
| 316 |
+
"learning_rate": 9.574872475046166e-05,
|
| 317 |
+
"loss": 0.059,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.0665951385548855,
|
| 322 |
+
"grad_norm": 0.6896253824234009,
|
| 323 |
+
"learning_rate": 9.547785243046505e-05,
|
| 324 |
+
"loss": 0.046,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.06807503052277183,
|
| 329 |
+
"grad_norm": 1.0698304176330566,
|
| 330 |
+
"learning_rate": 9.519902280311653e-05,
|
| 331 |
+
"loss": 0.0511,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.06955492249065819,
|
| 336 |
+
"grad_norm": 0.7512239217758179,
|
| 337 |
+
"learning_rate": 9.491228465552726e-05,
|
| 338 |
+
"loss": 0.0578,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.07103481445854452,
|
| 343 |
+
"grad_norm": 1.0382441282272339,
|
| 344 |
+
"learning_rate": 9.461768815857053e-05,
|
| 345 |
+
"loss": 0.0618,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.07251470642643087,
|
| 350 |
+
"grad_norm": 0.7745275497436523,
|
| 351 |
+
"learning_rate": 9.431528485810316e-05,
|
| 352 |
+
"loss": 0.069,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.07399459839431721,
|
| 357 |
+
"grad_norm": 0.7567397952079773,
|
| 358 |
+
"learning_rate": 9.400512766594659e-05,
|
| 359 |
+
"loss": 0.047,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.07547449036220356,
|
| 364 |
+
"grad_norm": 0.7035396695137024,
|
| 365 |
+
"learning_rate": 9.368727085062872e-05,
|
| 366 |
+
"loss": 0.0563,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.0769543823300899,
|
| 371 |
+
"grad_norm": 0.8480479717254639,
|
| 372 |
+
"learning_rate": 9.336177002788862e-05,
|
| 373 |
+
"loss": 0.0498,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.07843427429797625,
|
| 378 |
+
"grad_norm": 0.6571253538131714,
|
| 379 |
+
"learning_rate": 9.302868215094534e-05,
|
| 380 |
+
"loss": 0.0579,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.07991416626586259,
|
| 385 |
+
"grad_norm": 0.6242689490318298,
|
| 386 |
+
"learning_rate": 9.268806550053264e-05,
|
| 387 |
+
"loss": 0.0465,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.08139405823374894,
|
| 392 |
+
"grad_norm": 1.0013155937194824,
|
| 393 |
+
"learning_rate": 9.233997967470174e-05,
|
| 394 |
+
"loss": 0.0792,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.08287395020163528,
|
| 399 |
+
"grad_norm": 1.0697152614593506,
|
| 400 |
+
"learning_rate": 9.198448557839321e-05,
|
| 401 |
+
"loss": 0.0707,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.08435384216952163,
|
| 406 |
+
"grad_norm": 0.9719803333282471,
|
| 407 |
+
"learning_rate": 9.162164541278051e-05,
|
| 408 |
+
"loss": 0.0586,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.08583373413740797,
|
| 413 |
+
"grad_norm": 1.0201226472854614,
|
| 414 |
+
"learning_rate": 9.125152266438649e-05,
|
| 415 |
+
"loss": 0.0535,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.08731362610529432,
|
| 420 |
+
"grad_norm": 0.8870009779930115,
|
| 421 |
+
"learning_rate": 9.087418209397506e-05,
|
| 422 |
+
"loss": 0.058,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.08879351807318066,
|
| 427 |
+
"grad_norm": 0.7464373707771301,
|
| 428 |
+
"learning_rate": 9.04896897252201e-05,
|
| 429 |
+
"loss": 0.0676,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.090273410041067,
|
| 434 |
+
"grad_norm": 0.6516630053520203,
|
| 435 |
+
"learning_rate": 9.009811283315304e-05,
|
| 436 |
+
"loss": 0.0506,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.09175330200895335,
|
| 441 |
+
"grad_norm": 0.6615219712257385,
|
| 442 |
+
"learning_rate": 8.969951993239177e-05,
|
| 443 |
+
"loss": 0.0479,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.09323319397683968,
|
| 448 |
+
"grad_norm": 0.6776785850524902,
|
| 449 |
+
"learning_rate": 8.929398076515259e-05,
|
| 450 |
+
"loss": 0.0459,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.09471308594472604,
|
| 455 |
+
"grad_norm": 0.7998208403587341,
|
| 456 |
+
"learning_rate": 8.888156628904724e-05,
|
| 457 |
+
"loss": 0.0493,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.09619297791261237,
|
| 462 |
+
"grad_norm": 0.8668547868728638,
|
| 463 |
+
"learning_rate": 8.846234866466747e-05,
|
| 464 |
+
"loss": 0.0431,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.09767286988049872,
|
| 469 |
+
"grad_norm": 0.7501780390739441,
|
| 470 |
+
"learning_rate": 8.803640124295902e-05,
|
| 471 |
+
"loss": 0.0664,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.09915276184838506,
|
| 476 |
+
"grad_norm": 0.5983700156211853,
|
| 477 |
+
"learning_rate": 8.760379855238723e-05,
|
| 478 |
+
"loss": 0.0424,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.10063265381627141,
|
| 483 |
+
"grad_norm": 0.7453112006187439,
|
| 484 |
+
"learning_rate": 8.716461628589683e-05,
|
| 485 |
+
"loss": 0.0508,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.10211254578415775,
|
| 490 |
+
"grad_norm": 0.7227299809455872,
|
| 491 |
+
"learning_rate": 8.671893128766784e-05,
|
| 492 |
+
"loss": 0.045,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.1035924377520441,
|
| 497 |
+
"grad_norm": 0.7978153824806213,
|
| 498 |
+
"learning_rate": 8.626682153967001e-05,
|
| 499 |
+
"loss": 0.0473,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.10507232971993044,
|
| 504 |
+
"grad_norm": 0.6989890336990356,
|
| 505 |
+
"learning_rate": 8.580836614801827e-05,
|
| 506 |
+
"loss": 0.046,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.10655222168781679,
|
| 511 |
+
"grad_norm": 0.8369282484054565,
|
| 512 |
+
"learning_rate": 8.534364532913144e-05,
|
| 513 |
+
"loss": 0.049,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.10803211365570313,
|
| 518 |
+
"grad_norm": 0.5447561144828796,
|
| 519 |
+
"learning_rate": 8.487274039569675e-05,
|
| 520 |
+
"loss": 0.0428,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.10951200562358948,
|
| 525 |
+
"grad_norm": 0.7491251826286316,
|
| 526 |
+
"learning_rate": 8.439573374244237e-05,
|
| 527 |
+
"loss": 0.0368,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.11099189759147582,
|
| 532 |
+
"grad_norm": 0.48374220728874207,
|
| 533 |
+
"learning_rate": 8.391270883172073e-05,
|
| 534 |
+
"loss": 0.0403,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.11247178955936217,
|
| 539 |
+
"grad_norm": 0.9518898129463196,
|
| 540 |
+
"learning_rate": 8.342375017890512e-05,
|
| 541 |
+
"loss": 0.038,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.11395168152724851,
|
| 546 |
+
"grad_norm": 0.8046047687530518,
|
| 547 |
+
"learning_rate": 8.292894333760186e-05,
|
| 548 |
+
"loss": 0.0414,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.11543157349513486,
|
| 553 |
+
"grad_norm": 0.6393342614173889,
|
| 554 |
+
"learning_rate": 8.242837488468087e-05,
|
| 555 |
+
"loss": 0.0615,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.1169114654630212,
|
| 560 |
+
"grad_norm": 0.6403966546058655,
|
| 561 |
+
"learning_rate": 8.192213240512737e-05,
|
| 562 |
+
"loss": 0.0403,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.11839135743090755,
|
| 567 |
+
"grad_norm": 0.5846046805381775,
|
| 568 |
+
"learning_rate": 8.141030447671686e-05,
|
| 569 |
+
"loss": 0.0376,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.11987124939879389,
|
| 574 |
+
"grad_norm": 1.1184839010238647,
|
| 575 |
+
"learning_rate": 8.089298065451672e-05,
|
| 576 |
+
"loss": 0.042,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.12135114136668024,
|
| 581 |
+
"grad_norm": 0.804345965385437,
|
| 582 |
+
"learning_rate": 8.037025145521657e-05,
|
| 583 |
+
"loss": 0.0448,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.12283103333456658,
|
| 588 |
+
"grad_norm": 0.40974166989326477,
|
| 589 |
+
"learning_rate": 7.984220834129052e-05,
|
| 590 |
+
"loss": 0.0691,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.12431092530245293,
|
| 595 |
+
"grad_norm": 0.5297871232032776,
|
| 596 |
+
"learning_rate": 7.93089437049939e-05,
|
| 597 |
+
"loss": 0.0477,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.12579081727033928,
|
| 602 |
+
"grad_norm": 0.6452186107635498,
|
| 603 |
+
"learning_rate": 7.877055085219721e-05,
|
| 604 |
+
"loss": 0.0505,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.12727070923822562,
|
| 609 |
+
"grad_norm": 0.6497986316680908,
|
| 610 |
+
"learning_rate": 7.82271239860604e-05,
|
| 611 |
+
"loss": 0.0471,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.12875060120611195,
|
| 616 |
+
"grad_norm": 0.6113291382789612,
|
| 617 |
+
"learning_rate": 7.767875819054997e-05,
|
| 618 |
+
"loss": 0.0485,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.1302304931739983,
|
| 623 |
+
"grad_norm": 0.8844181895256042,
|
| 624 |
+
"learning_rate": 7.712554941380206e-05,
|
| 625 |
+
"loss": 0.0429,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.13171038514188466,
|
| 630 |
+
"grad_norm": 0.7872292995452881,
|
| 631 |
+
"learning_rate": 7.656759445133428e-05,
|
| 632 |
+
"loss": 0.0471,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.133190277109771,
|
| 637 |
+
"grad_norm": 0.8881542682647705,
|
| 638 |
+
"learning_rate": 7.600499092910934e-05,
|
| 639 |
+
"loss": 0.0498,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.13467016907765733,
|
| 644 |
+
"grad_norm": 0.6990298628807068,
|
| 645 |
+
"learning_rate": 7.543783728645328e-05,
|
| 646 |
+
"loss": 0.0385,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.13615006104554367,
|
| 651 |
+
"grad_norm": 0.8354172110557556,
|
| 652 |
+
"learning_rate": 7.486623275883151e-05,
|
| 653 |
+
"loss": 0.0347,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.13762995301343,
|
| 658 |
+
"grad_norm": 0.8185162544250488,
|
| 659 |
+
"learning_rate": 7.429027736048535e-05,
|
| 660 |
+
"loss": 0.0409,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.13910984498131637,
|
| 665 |
+
"grad_norm": 0.9761926531791687,
|
| 666 |
+
"learning_rate": 7.37100718669326e-05,
|
| 667 |
+
"loss": 0.0422,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.1405897369492027,
|
| 672 |
+
"grad_norm": 0.5239071249961853,
|
| 673 |
+
"learning_rate": 7.312571779733463e-05,
|
| 674 |
+
"loss": 0.0269,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.14206962891708905,
|
| 679 |
+
"grad_norm": 0.7226734757423401,
|
| 680 |
+
"learning_rate": 7.253731739673349e-05,
|
| 681 |
+
"loss": 0.0366,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.14354952088497538,
|
| 686 |
+
"grad_norm": 0.6310585141181946,
|
| 687 |
+
"learning_rate": 7.194497361816196e-05,
|
| 688 |
+
"loss": 0.0527,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.14502941285286175,
|
| 693 |
+
"grad_norm": 0.5752474665641785,
|
| 694 |
+
"learning_rate": 7.134879010462988e-05,
|
| 695 |
+
"loss": 0.0312,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.1465093048207481,
|
| 700 |
+
"grad_norm": 0.5619791150093079,
|
| 701 |
+
"learning_rate": 7.07488711709894e-05,
|
| 702 |
+
"loss": 0.0314,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.14798919678863443,
|
| 707 |
+
"grad_norm": 0.6874368190765381,
|
| 708 |
+
"learning_rate": 7.014532178568314e-05,
|
| 709 |
+
"loss": 0.0294,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.14946908875652076,
|
| 714 |
+
"grad_norm": 0.6269967555999756,
|
| 715 |
+
"learning_rate": 6.953824755237756e-05,
|
| 716 |
+
"loss": 0.0357,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.15094898072440713,
|
| 721 |
+
"grad_norm": 0.7482361793518066,
|
| 722 |
+
"learning_rate": 6.892775469148553e-05,
|
| 723 |
+
"loss": 0.043,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.15242887269229347,
|
| 728 |
+
"grad_norm": 0.7447315454483032,
|
| 729 |
+
"learning_rate": 6.831395002158067e-05,
|
| 730 |
+
"loss": 0.0319,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.1539087646601798,
|
| 735 |
+
"grad_norm": 0.5281906127929688,
|
| 736 |
+
"learning_rate": 6.76969409407074e-05,
|
| 737 |
+
"loss": 0.0311,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.15538865662806614,
|
| 742 |
+
"grad_norm": 0.5784289240837097,
|
| 743 |
+
"learning_rate": 6.707683540758915e-05,
|
| 744 |
+
"loss": 0.0362,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.1568685485959525,
|
| 749 |
+
"grad_norm": 0.5917581915855408,
|
| 750 |
+
"learning_rate": 6.645374192273894e-05,
|
| 751 |
+
"loss": 0.0406,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.15834844056383884,
|
| 756 |
+
"grad_norm": 0.8599121570587158,
|
| 757 |
+
"learning_rate": 6.582776950947474e-05,
|
| 758 |
+
"loss": 0.0468,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.15982833253172518,
|
| 763 |
+
"grad_norm": 0.5944121479988098,
|
| 764 |
+
"learning_rate": 6.519902769484368e-05,
|
| 765 |
+
"loss": 0.0464,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.16130822449961152,
|
| 770 |
+
"grad_norm": 0.6295680999755859,
|
| 771 |
+
"learning_rate": 6.456762649045796e-05,
|
| 772 |
+
"loss": 0.0371,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.16278811646749788,
|
| 777 |
+
"grad_norm": 1.0905753374099731,
|
| 778 |
+
"learning_rate": 6.393367637324593e-05,
|
| 779 |
+
"loss": 0.0465,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.16426800843538422,
|
| 784 |
+
"grad_norm": 0.7028170228004456,
|
| 785 |
+
"learning_rate": 6.329728826612192e-05,
|
| 786 |
+
"loss": 0.0493,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.16574790040327056,
|
| 791 |
+
"grad_norm": 0.6126824617385864,
|
| 792 |
+
"learning_rate": 6.265857351857788e-05,
|
| 793 |
+
"loss": 0.0369,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.1672277923711569,
|
| 798 |
+
"grad_norm": 0.7399412393569946,
|
| 799 |
+
"learning_rate": 6.201764388720049e-05,
|
| 800 |
+
"loss": 0.0412,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.16870768433904326,
|
| 805 |
+
"grad_norm": 0.5751072764396667,
|
| 806 |
+
"learning_rate": 6.137461151611692e-05,
|
| 807 |
+
"loss": 0.0446,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.1701875763069296,
|
| 812 |
+
"grad_norm": 0.4830611050128937,
|
| 813 |
+
"learning_rate": 6.072958891737296e-05,
|
| 814 |
+
"loss": 0.0396,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.17166746827481594,
|
| 819 |
+
"grad_norm": 0.5445456504821777,
|
| 820 |
+
"learning_rate": 6.00826889512466e-05,
|
| 821 |
+
"loss": 0.0367,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.17314736024270228,
|
| 826 |
+
"grad_norm": 0.7560216188430786,
|
| 827 |
+
"learning_rate": 5.943402480650071e-05,
|
| 828 |
+
"loss": 0.0328,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.17462725221058864,
|
| 833 |
+
"grad_norm": 0.6100801229476929,
|
| 834 |
+
"learning_rate": 5.8783709980578414e-05,
|
| 835 |
+
"loss": 0.031,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.17610714417847498,
|
| 840 |
+
"grad_norm": 0.5339049696922302,
|
| 841 |
+
"learning_rate": 5.813185825974419e-05,
|
| 842 |
+
"loss": 0.0365,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.17758703614636132,
|
| 847 |
+
"grad_norm": 0.7650025486946106,
|
| 848 |
+
"learning_rate": 5.747858369917465e-05,
|
| 849 |
+
"loss": 0.034,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.17906692811424765,
|
| 854 |
+
"grad_norm": 0.6139253377914429,
|
| 855 |
+
"learning_rate": 5.682400060300213e-05,
|
| 856 |
+
"loss": 0.0399,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.180546820082134,
|
| 861 |
+
"grad_norm": 0.514707088470459,
|
| 862 |
+
"learning_rate": 5.6168223504314863e-05,
|
| 863 |
+
"loss": 0.0371,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.18202671205002036,
|
| 868 |
+
"grad_norm": 0.5504834651947021,
|
| 869 |
+
"learning_rate": 5.551136714511691e-05,
|
| 870 |
+
"loss": 0.0338,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.1835066040179067,
|
| 875 |
+
"grad_norm": 0.6514374017715454,
|
| 876 |
+
"learning_rate": 5.485354645625167e-05,
|
| 877 |
+
"loss": 0.0392,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.18498649598579303,
|
| 882 |
+
"grad_norm": 0.5081560015678406,
|
| 883 |
+
"learning_rate": 5.419487653729234e-05,
|
| 884 |
+
"loss": 0.0332,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.18646638795367937,
|
| 889 |
+
"grad_norm": 0.904439389705658,
|
| 890 |
+
"learning_rate": 5.353547263640273e-05,
|
| 891 |
+
"loss": 0.0382,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.18794627992156573,
|
| 896 |
+
"grad_norm": 0.5332797765731812,
|
| 897 |
+
"learning_rate": 5.2875450130172324e-05,
|
| 898 |
+
"loss": 0.0221,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.18942617188945207,
|
| 903 |
+
"grad_norm": 0.6516315937042236,
|
| 904 |
+
"learning_rate": 5.221492450342856e-05,
|
| 905 |
+
"loss": 0.0371,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.1909060638573384,
|
| 910 |
+
"grad_norm": 0.7484680414199829,
|
| 911 |
+
"learning_rate": 5.155401132903045e-05,
|
| 912 |
+
"loss": 0.0377,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.19238595582522475,
|
| 917 |
+
"grad_norm": 0.9511647820472717,
|
| 918 |
+
"learning_rate": 5.089282624764654e-05,
|
| 919 |
+
"loss": 0.0386,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.1938658477931111,
|
| 924 |
+
"grad_norm": 0.3802145719528198,
|
| 925 |
+
"learning_rate": 5.0231484947521336e-05,
|
| 926 |
+
"loss": 0.0324,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.19534573976099745,
|
| 931 |
+
"grad_norm": 0.5430207252502441,
|
| 932 |
+
"learning_rate": 4.9570103144233024e-05,
|
| 933 |
+
"loss": 0.0296,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.1968256317288838,
|
| 938 |
+
"grad_norm": 0.631252646446228,
|
| 939 |
+
"learning_rate": 4.890879656044669e-05,
|
| 940 |
+
"loss": 0.0241,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.19830552369677013,
|
| 945 |
+
"grad_norm": 0.4435235261917114,
|
| 946 |
+
"learning_rate": 4.824768090566618e-05,
|
| 947 |
+
"loss": 0.0285,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.1997854156646565,
|
| 952 |
+
"grad_norm": 0.7757057547569275,
|
| 953 |
+
"learning_rate": 4.7586871855988326e-05,
|
| 954 |
+
"loss": 0.0419,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.20126530763254283,
|
| 959 |
+
"grad_norm": 0.6761030554771423,
|
| 960 |
+
"learning_rate": 4.692648503386289e-05,
|
| 961 |
+
"loss": 0.0338,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.20274519960042917,
|
| 966 |
+
"grad_norm": 0.3639812767505646,
|
| 967 |
+
"learning_rate": 4.6266635987862086e-05,
|
| 968 |
+
"loss": 0.0244,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.2042250915683155,
|
| 973 |
+
"grad_norm": 0.3727477490901947,
|
| 974 |
+
"learning_rate": 4.560744017246284e-05,
|
| 975 |
+
"loss": 0.025,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.20570498353620187,
|
| 980 |
+
"grad_norm": 0.5406703948974609,
|
| 981 |
+
"learning_rate": 4.4949012927845676e-05,
|
| 982 |
+
"loss": 0.031,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.2071848755040882,
|
| 987 |
+
"grad_norm": 0.45045414566993713,
|
| 988 |
+
"learning_rate": 4.429146945971346e-05,
|
| 989 |
+
"loss": 0.0236,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.20866476747197454,
|
| 994 |
+
"grad_norm": 0.549584686756134,
|
| 995 |
+
"learning_rate": 4.3634924819133746e-05,
|
| 996 |
+
"loss": 0.0286,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.21014465943986088,
|
| 1001 |
+
"grad_norm": 0.7233926653862,
|
| 1002 |
+
"learning_rate": 4.297949388240823e-05,
|
| 1003 |
+
"loss": 0.0297,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.21162455140774725,
|
| 1008 |
+
"grad_norm": 0.5810157656669617,
|
| 1009 |
+
"learning_rate": 4.2325291330972664e-05,
|
| 1010 |
+
"loss": 0.0189,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.21310444337563358,
|
| 1015 |
+
"grad_norm": 0.5259445905685425,
|
| 1016 |
+
"learning_rate": 4.167243163133094e-05,
|
| 1017 |
+
"loss": 0.0271,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.21458433534351992,
|
| 1022 |
+
"grad_norm": 0.5740777254104614,
|
| 1023 |
+
"learning_rate": 4.1021029015026736e-05,
|
| 1024 |
+
"loss": 0.0367,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.21606422731140626,
|
| 1029 |
+
"grad_norm": 0.6542416214942932,
|
| 1030 |
+
"learning_rate": 4.037119745865641e-05,
|
| 1031 |
+
"loss": 0.0323,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.21754411927929262,
|
| 1036 |
+
"grad_norm": 0.5732387900352478,
|
| 1037 |
+
"learning_rate": 3.972305066392626e-05,
|
| 1038 |
+
"loss": 0.0295,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.21902401124717896,
|
| 1043 |
+
"grad_norm": 0.8444030284881592,
|
| 1044 |
+
"learning_rate": 3.9076702037758076e-05,
|
| 1045 |
+
"loss": 0.027,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.2205039032150653,
|
| 1050 |
+
"grad_norm": 0.36182349920272827,
|
| 1051 |
+
"learning_rate": 3.8432264672446293e-05,
|
| 1052 |
+
"loss": 0.0306,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.22198379518295164,
|
| 1057 |
+
"grad_norm": 0.367953360080719,
|
| 1058 |
+
"learning_rate": 3.778985132586995e-05,
|
| 1059 |
+
"loss": 0.0258,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.22346368715083798,
|
| 1064 |
+
"grad_norm": 0.43248283863067627,
|
| 1065 |
+
"learning_rate": 3.714957440176345e-05,
|
| 1066 |
+
"loss": 0.0237,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.22494357911872434,
|
| 1071 |
+
"grad_norm": 0.5691545605659485,
|
| 1072 |
+
"learning_rate": 3.651154593004911e-05,
|
| 1073 |
+
"loss": 0.0257,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.22642347108661068,
|
| 1078 |
+
"grad_norm": 0.4153839945793152,
|
| 1079 |
+
"learning_rate": 3.587587754723523e-05,
|
| 1080 |
+
"loss": 0.0256,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.22790336305449702,
|
| 1085 |
+
"grad_norm": 0.4405042827129364,
|
| 1086 |
+
"learning_rate": 3.5242680476882815e-05,
|
| 1087 |
+
"loss": 0.0243,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.22938325502238335,
|
| 1092 |
+
"grad_norm": 0.5467635989189148,
|
| 1093 |
+
"learning_rate": 3.461206551014481e-05,
|
| 1094 |
+
"loss": 0.0242,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.23086314699026972,
|
| 1099 |
+
"grad_norm": 0.4557804763317108,
|
| 1100 |
+
"learning_rate": 3.3984142986380764e-05,
|
| 1101 |
+
"loss": 0.0338,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.23234303895815606,
|
| 1106 |
+
"grad_norm": 0.39460909366607666,
|
| 1107 |
+
"learning_rate": 3.335902277385067e-05,
|
| 1108 |
+
"loss": 0.0205,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.2338229309260424,
|
| 1113 |
+
"grad_norm": 0.5078433752059937,
|
| 1114 |
+
"learning_rate": 3.2736814250491196e-05,
|
| 1115 |
+
"loss": 0.0248,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.23530282289392873,
|
| 1120 |
+
"grad_norm": 0.6240681409835815,
|
| 1121 |
+
"learning_rate": 3.211762628477771e-05,
|
| 1122 |
+
"loss": 0.0312,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.2367827148618151,
|
| 1127 |
+
"grad_norm": 0.37351563572883606,
|
| 1128 |
+
"learning_rate": 3.150156721667547e-05,
|
| 1129 |
+
"loss": 0.0245,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.23826260682970143,
|
| 1134 |
+
"grad_norm": 0.3536587953567505,
|
| 1135 |
+
"learning_rate": 3.088874483868325e-05,
|
| 1136 |
+
"loss": 0.021,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.23974249879758777,
|
| 1141 |
+
"grad_norm": 0.30240142345428467,
|
| 1142 |
+
"learning_rate": 3.0279266376972715e-05,
|
| 1143 |
+
"loss": 0.025,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.2412223907654741,
|
| 1148 |
+
"grad_norm": 0.4821987450122833,
|
| 1149 |
+
"learning_rate": 2.96732384726271e-05,
|
| 1150 |
+
"loss": 0.0427,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.24270228273336047,
|
| 1155 |
+
"grad_norm": 0.6150904297828674,
|
| 1156 |
+
"learning_rate": 2.907076716298196e-05,
|
| 1157 |
+
"loss": 0.0297,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.2441821747012468,
|
| 1162 |
+
"grad_norm": 0.49226638674736023,
|
| 1163 |
+
"learning_rate": 2.847195786307174e-05,
|
| 1164 |
+
"loss": 0.0267,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.24566206666913315,
|
| 1169 |
+
"grad_norm": 0.49682047963142395,
|
| 1170 |
+
"learning_rate": 2.7876915347185227e-05,
|
| 1171 |
+
"loss": 0.0326,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.2471419586370195,
|
| 1176 |
+
"grad_norm": 0.42303502559661865,
|
| 1177 |
+
"learning_rate": 2.7285743730533143e-05,
|
| 1178 |
+
"loss": 0.0212,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.24862185060490585,
|
| 1183 |
+
"grad_norm": 0.5483497977256775,
|
| 1184 |
+
"learning_rate": 2.6698546451030826e-05,
|
| 1185 |
+
"loss": 0.0202,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.25010174257279216,
|
| 1190 |
+
"grad_norm": 0.5056328177452087,
|
| 1191 |
+
"learning_rate": 2.611542625119975e-05,
|
| 1192 |
+
"loss": 0.0228,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.25158163454067856,
|
| 1197 |
+
"grad_norm": 0.32099300622940063,
|
| 1198 |
+
"learning_rate": 2.5536485160190482e-05,
|
| 1199 |
+
"loss": 0.0224,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.2530615265085649,
|
| 1204 |
+
"grad_norm": 0.5951581597328186,
|
| 1205 |
+
"learning_rate": 2.496182447593055e-05,
|
| 1206 |
+
"loss": 0.0272,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.25454141847645123,
|
| 1211 |
+
"grad_norm": 0.45417091250419617,
|
| 1212 |
+
"learning_rate": 2.4391544747400252e-05,
|
| 1213 |
+
"loss": 0.0227,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.25602131044433757,
|
| 1218 |
+
"grad_norm": 0.5122156739234924,
|
| 1219 |
+
"learning_rate": 2.3825745757039452e-05,
|
| 1220 |
+
"loss": 0.024,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.2575012024122239,
|
| 1225 |
+
"grad_norm": 0.4841585159301758,
|
| 1226 |
+
"learning_rate": 2.3264526503288642e-05,
|
| 1227 |
+
"loss": 0.0225,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.25898109438011024,
|
| 1232 |
+
"grad_norm": 0.43821659684181213,
|
| 1233 |
+
"learning_rate": 2.2707985183266978e-05,
|
| 1234 |
+
"loss": 0.0215,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.2604609863479966,
|
| 1239 |
+
"grad_norm": 0.42495104670524597,
|
| 1240 |
+
"learning_rate": 2.215621917559062e-05,
|
| 1241 |
+
"loss": 0.0213,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.2619408783158829,
|
| 1246 |
+
"grad_norm": 0.4166795015335083,
|
| 1247 |
+
"learning_rate": 2.1609325023334377e-05,
|
| 1248 |
+
"loss": 0.018,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.2634207702837693,
|
| 1253 |
+
"grad_norm": 0.29304540157318115,
|
| 1254 |
+
"learning_rate": 2.1067398417139466e-05,
|
| 1255 |
+
"loss": 0.029,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.26490066225165565,
|
| 1260 |
+
"grad_norm": 0.6335829496383667,
|
| 1261 |
+
"learning_rate": 2.0530534178470322e-05,
|
| 1262 |
+
"loss": 0.021,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.266380554219542,
|
| 1267 |
+
"grad_norm": 0.35307204723358154,
|
| 1268 |
+
"learning_rate": 1.9998826243023666e-05,
|
| 1269 |
+
"loss": 0.021,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.2678604461874283,
|
| 1274 |
+
"grad_norm": 0.4115915894508362,
|
| 1275 |
+
"learning_rate": 1.9472367644292457e-05,
|
| 1276 |
+
"loss": 0.0212,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.26934033815531466,
|
| 1281 |
+
"grad_norm": 0.5536908507347107,
|
| 1282 |
+
"learning_rate": 1.8951250497287716e-05,
|
| 1283 |
+
"loss": 0.0291,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.270820230123201,
|
| 1288 |
+
"grad_norm": 0.6118968725204468,
|
| 1289 |
+
"learning_rate": 1.843556598242109e-05,
|
| 1290 |
+
"loss": 0.0351,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.27230012209108734,
|
| 1295 |
+
"grad_norm": 0.5278725624084473,
|
| 1296 |
+
"learning_rate": 1.792540432955087e-05,
|
| 1297 |
+
"loss": 0.0243,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.2737800140589737,
|
| 1302 |
+
"grad_norm": 0.365327388048172,
|
| 1303 |
+
"learning_rate": 1.742085480219449e-05,
|
| 1304 |
+
"loss": 0.0303,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.27525990602686,
|
| 1309 |
+
"grad_norm": 0.5491426587104797,
|
| 1310 |
+
"learning_rate": 1.6922005681909843e-05,
|
| 1311 |
+
"loss": 0.0321,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.2767397979947464,
|
| 1316 |
+
"grad_norm": 0.5193561315536499,
|
| 1317 |
+
"learning_rate": 1.642894425284867e-05,
|
| 1318 |
+
"loss": 0.0207,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.27821968996263274,
|
| 1323 |
+
"grad_norm": 0.39915916323661804,
|
| 1324 |
+
"learning_rate": 1.5941756786484335e-05,
|
| 1325 |
+
"loss": 0.0197,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.2796995819305191,
|
| 1330 |
+
"grad_norm": 0.4479086101055145,
|
| 1331 |
+
"learning_rate": 1.5460528526516804e-05,
|
| 1332 |
+
"loss": 0.0233,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.2811794738984054,
|
| 1337 |
+
"grad_norm": 0.3311406373977661,
|
| 1338 |
+
"learning_rate": 1.498534367395748e-05,
|
| 1339 |
+
"loss": 0.0202,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.28265936586629176,
|
| 1344 |
+
"grad_norm": 0.2992391586303711,
|
| 1345 |
+
"learning_rate": 1.4516285372396437e-05,
|
| 1346 |
+
"loss": 0.0264,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.2841392578341781,
|
| 1351 |
+
"grad_norm": 0.480135440826416,
|
| 1352 |
+
"learning_rate": 1.4053435693454775e-05,
|
| 1353 |
+
"loss": 0.0221,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.28561914980206443,
|
| 1358 |
+
"grad_norm": 0.27029678225517273,
|
| 1359 |
+
"learning_rate": 1.359687562242437e-05,
|
| 1360 |
+
"loss": 0.0153,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.28709904176995077,
|
| 1365 |
+
"grad_norm": 0.5432844758033752,
|
| 1366 |
+
"learning_rate": 1.314668504409779e-05,
|
| 1367 |
+
"loss": 0.0207,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.28857893373783716,
|
| 1372 |
+
"grad_norm": 0.4192567765712738,
|
| 1373 |
+
"learning_rate": 1.2702942728790895e-05,
|
| 1374 |
+
"loss": 0.0168,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.2900588257057235,
|
| 1379 |
+
"grad_norm": 0.5646419525146484,
|
| 1380 |
+
"learning_rate": 1.2265726318560172e-05,
|
| 1381 |
+
"loss": 0.0269,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.29153871767360984,
|
| 1386 |
+
"grad_norm": 0.47364342212677,
|
| 1387 |
+
"learning_rate": 1.1835112313617697e-05,
|
| 1388 |
+
"loss": 0.0158,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.2930186096414962,
|
| 1393 |
+
"grad_norm": 0.3461420238018036,
|
| 1394 |
+
"learning_rate": 1.1411176058945771e-05,
|
| 1395 |
+
"loss": 0.0212,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.2944985016093825,
|
| 1400 |
+
"grad_norm": 0.34054312109947205,
|
| 1401 |
+
"learning_rate": 1.0993991731113817e-05,
|
| 1402 |
+
"loss": 0.0143,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.29597839357726885,
|
| 1407 |
+
"grad_norm": 0.48517104983329773,
|
| 1408 |
+
"learning_rate": 1.058363232529948e-05,
|
| 1409 |
+
"loss": 0.0256,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.2974582855451552,
|
| 1414 |
+
"grad_norm": 0.4162018299102783,
|
| 1415 |
+
"learning_rate": 1.0180169642516718e-05,
|
| 1416 |
+
"loss": 0.033,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.2989381775130415,
|
| 1421 |
+
"grad_norm": 0.4353783130645752,
|
| 1422 |
+
"learning_rate": 9.783674277052667e-06,
|
| 1423 |
+
"loss": 0.018,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.3004180694809279,
|
| 1428 |
+
"grad_norm": 0.4821432828903198,
|
| 1429 |
+
"learning_rate": 9.394215604115641e-06,
|
| 1430 |
+
"loss": 0.0136,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.30189796144881426,
|
| 1435 |
+
"grad_norm": 0.3058512806892395,
|
| 1436 |
+
"learning_rate": 9.011861767696522e-06,
|
| 1437 |
+
"loss": 0.0217,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.3033778534167006,
|
| 1442 |
+
"grad_norm": 0.4392576813697815,
|
| 1443 |
+
"learning_rate": 8.636679668645536e-06,
|
| 1444 |
+
"loss": 0.0182,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.30485774538458693,
|
| 1449 |
+
"grad_norm": 0.3896658718585968,
|
| 1450 |
+
"learning_rate": 8.268734952966505e-06,
|
| 1451 |
+
"loss": 0.0189,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.30633763735247327,
|
| 1456 |
+
"grad_norm": 0.4406259059906006,
|
| 1457 |
+
"learning_rate": 7.908092000330747e-06,
|
| 1458 |
+
"loss": 0.0145,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.3078175293203596,
|
| 1463 |
+
"grad_norm": 0.357083261013031,
|
| 1464 |
+
"learning_rate": 7.5548139128124364e-06,
|
| 1465 |
+
"loss": 0.0177,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.30929742128824594,
|
| 1470 |
+
"grad_norm": 0.2061689794063568,
|
| 1471 |
+
"learning_rate": 7.2089625038476606e-06,
|
| 1472 |
+
"loss": 0.0173,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.3107773132561323,
|
| 1477 |
+
"grad_norm": 0.4433155953884125,
|
| 1478 |
+
"learning_rate": 6.87059828741875e-06,
|
| 1479 |
+
"loss": 0.0188,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.3122572052240186,
|
| 1484 |
+
"grad_norm": 0.31645211577415466,
|
| 1485 |
+
"learning_rate": 6.539780467466172e-06,
|
| 1486 |
+
"loss": 0.0322,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.313737097191905,
|
| 1491 |
+
"grad_norm": 0.3037683367729187,
|
| 1492 |
+
"learning_rate": 6.216566927529455e-06,
|
| 1493 |
+
"loss": 0.0177,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.31521698915979135,
|
| 1498 |
+
"grad_norm": 0.2664813995361328,
|
| 1499 |
+
"learning_rate": 5.9010142206194e-06,
|
| 1500 |
+
"loss": 0.018,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.3166968811276777,
|
| 1505 |
+
"grad_norm": 0.22976835072040558,
|
| 1506 |
+
"learning_rate": 5.593177559322777e-06,
|
| 1507 |
+
"loss": 0.0145,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.318176773095564,
|
| 1512 |
+
"grad_norm": 0.3369593024253845,
|
| 1513 |
+
"learning_rate": 5.293110806141832e-06,
|
| 1514 |
+
"loss": 0.0218,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.31965666506345036,
|
| 1519 |
+
"grad_norm": 0.3195848762989044,
|
| 1520 |
+
"learning_rate": 5.000866464069842e-06,
|
| 1521 |
+
"loss": 0.0194,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.3211365570313367,
|
| 1526 |
+
"grad_norm": 0.19008630514144897,
|
| 1527 |
+
"learning_rate": 4.716495667404691e-06,
|
| 1528 |
+
"loss": 0.0174,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.32261644899922304,
|
| 1533 |
+
"grad_norm": 0.42658621072769165,
|
| 1534 |
+
"learning_rate": 4.440048172801725e-06,
|
| 1535 |
+
"loss": 0.0247,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.3240963409671094,
|
| 1540 |
+
"grad_norm": 0.6167186498641968,
|
| 1541 |
+
"learning_rate": 4.171572350567898e-06,
|
| 1542 |
+
"loss": 0.0193,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.32557623293499577,
|
| 1547 |
+
"grad_norm": 0.30169373750686646,
|
| 1548 |
+
"learning_rate": 3.9111151761983265e-06,
|
| 1549 |
+
"loss": 0.017,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.3270561249028821,
|
| 1554 |
+
"grad_norm": 0.3697860836982727,
|
| 1555 |
+
"learning_rate": 3.6587222221569075e-06,
|
| 1556 |
+
"loss": 0.0149,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.32853601687076844,
|
| 1561 |
+
"grad_norm": 0.43350812792778015,
|
| 1562 |
+
"learning_rate": 3.414437649902491e-06,
|
| 1563 |
+
"loss": 0.0142,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.3300159088386548,
|
| 1568 |
+
"grad_norm": 0.3299703896045685,
|
| 1569 |
+
"learning_rate": 3.1783042021619026e-06,
|
| 1570 |
+
"loss": 0.0161,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.3314958008065411,
|
| 1575 |
+
"grad_norm": 0.5686686038970947,
|
| 1576 |
+
"learning_rate": 2.9503631954511833e-06,
|
| 1577 |
+
"loss": 0.0204,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.33297569277442746,
|
| 1582 |
+
"grad_norm": 0.28825682401657104,
|
| 1583 |
+
"learning_rate": 2.7306545128464202e-06,
|
| 1584 |
+
"loss": 0.0225,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.3344555847423138,
|
| 1589 |
+
"grad_norm": 0.248334601521492,
|
| 1590 |
+
"learning_rate": 2.5192165970053307e-06,
|
| 1591 |
+
"loss": 0.0122,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.33593547671020013,
|
| 1596 |
+
"grad_norm": 0.4341728985309601,
|
| 1597 |
+
"learning_rate": 2.316086443440962e-06,
|
| 1598 |
+
"loss": 0.0164,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.3374153686780865,
|
| 1603 |
+
"grad_norm": 0.23554910719394684,
|
| 1604 |
+
"learning_rate": 2.1212995940485036e-06,
|
| 1605 |
+
"loss": 0.0171,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.33889526064597286,
|
| 1610 |
+
"grad_norm": 0.4276198744773865,
|
| 1611 |
+
"learning_rate": 1.9348901308864796e-06,
|
| 1612 |
+
"loss": 0.0246,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.3403751526138592,
|
| 1617 |
+
"grad_norm": 0.35541170835494995,
|
| 1618 |
+
"learning_rate": 1.7568906702134124e-06,
|
| 1619 |
+
"loss": 0.0112,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.34185504458174554,
|
| 1624 |
+
"grad_norm": 0.4070402681827545,
|
| 1625 |
+
"learning_rate": 1.5873323567808963e-06,
|
| 1626 |
+
"loss": 0.0202,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.3433349365496319,
|
| 1631 |
+
"grad_norm": 0.16610288619995117,
|
| 1632 |
+
"learning_rate": 1.4262448583841793e-06,
|
| 1633 |
+
"loss": 0.0129,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.3448148285175182,
|
| 1638 |
+
"grad_norm": 0.32411548495292664,
|
| 1639 |
+
"learning_rate": 1.2736563606711382e-06,
|
| 1640 |
+
"loss": 0.0173,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.34629472048540455,
|
| 1645 |
+
"grad_norm": 0.3122292160987854,
|
| 1646 |
+
"learning_rate": 1.1295935622106513e-06,
|
| 1647 |
+
"loss": 0.0183,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.3477746124532909,
|
| 1652 |
+
"grad_norm": 0.3919500708580017,
|
| 1653 |
+
"learning_rate": 9.94081669821062e-07,
|
| 1654 |
+
"loss": 0.0223,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.3492545044211773,
|
| 1659 |
+
"grad_norm": 0.3418029546737671,
|
| 1660 |
+
"learning_rate": 8.671443941597523e-07,
|
| 1661 |
+
"loss": 0.0172,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.3507343963890636,
|
| 1666 |
+
"grad_norm": 0.28939440846443176,
|
| 1667 |
+
"learning_rate": 7.488039455744611e-07,
|
| 1668 |
+
"loss": 0.0146,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.35221428835694996,
|
| 1673 |
+
"grad_norm": 0.407928466796875,
|
| 1674 |
+
"learning_rate": 6.390810302171146e-07,
|
| 1675 |
+
"loss": 0.0244,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.3536941803248363,
|
| 1680 |
+
"grad_norm": 0.32271912693977356,
|
| 1681 |
+
"learning_rate": 5.379948464208418e-07,
|
| 1682 |
+
"loss": 0.0214,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.35517407229272263,
|
| 1687 |
+
"grad_norm": 0.4125051498413086,
|
| 1688 |
+
"learning_rate": 4.455630813408329e-07,
|
| 1689 |
+
"loss": 0.0166,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.35665396426060897,
|
| 1694 |
+
"grad_norm": 0.24020278453826904,
|
| 1695 |
+
"learning_rate": 3.61801907859588e-07,
|
| 1696 |
+
"loss": 0.0174,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.3581338562284953,
|
| 1701 |
+
"grad_norm": 0.2490530014038086,
|
| 1702 |
+
"learning_rate": 2.867259817571355e-07,
|
| 1703 |
+
"loss": 0.0224,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.35961374819638164,
|
| 1708 |
+
"grad_norm": 0.1842329204082489,
|
| 1709 |
+
"learning_rate": 2.2034843914670588e-07,
|
| 1710 |
+
"loss": 0.0163,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.361093640164268,
|
| 1715 |
+
"grad_norm": 0.4745646119117737,
|
| 1716 |
+
"learning_rate": 1.626808941762703e-07,
|
| 1717 |
+
"loss": 0.0207,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.3625735321321544,
|
| 1722 |
+
"grad_norm": 0.4176884889602661,
|
| 1723 |
+
"learning_rate": 1.1373343699642158e-07,
|
| 1724 |
+
"loss": 0.0159,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.3640534241000407,
|
| 1729 |
+
"grad_norm": 0.280038982629776,
|
| 1730 |
+
"learning_rate": 7.351463199488651e-08,
|
| 1731 |
+
"loss": 0.017,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.36553331606792705,
|
| 1736 |
+
"grad_norm": 0.4648924767971039,
|
| 1737 |
+
"learning_rate": 4.203151629798563e-08,
|
| 1738 |
+
"loss": 0.0196,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.3670132080358134,
|
| 1743 |
+
"grad_norm": 0.42389336228370667,
|
| 1744 |
+
"learning_rate": 1.928959853936263e-08,
|
| 1745 |
+
"loss": 0.0275,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.3684931000036997,
|
| 1750 |
+
"grad_norm": 0.5853281021118164,
|
| 1751 |
+
"learning_rate": 5.292857896133097e-09,
|
| 1752 |
+
"loss": 0.0182,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.36997299197158606,
|
| 1757 |
+
"grad_norm": 0.2962166368961334,
|
| 1758 |
+
"learning_rate": 4.374339263035765e-11,
|
| 1759 |
+
"loss": 0.014,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
}
|
| 1762 |
+
],
|
| 1763 |
+
"logging_steps": 10,
|
| 1764 |
+
"max_steps": 2500,
|
| 1765 |
+
"num_input_tokens_seen": 0,
|
| 1766 |
+
"num_train_epochs": 1,
|
| 1767 |
+
"save_steps": 1000,
|
| 1768 |
+
"stateful_callbacks": {
|
| 1769 |
+
"TrainerControl": {
|
| 1770 |
+
"args": {
|
| 1771 |
+
"should_epoch_stop": false,
|
| 1772 |
+
"should_evaluate": false,
|
| 1773 |
+
"should_log": false,
|
| 1774 |
+
"should_save": true,
|
| 1775 |
+
"should_training_stop": true
|
| 1776 |
+
},
|
| 1777 |
+
"attributes": {}
|
| 1778 |
+
}
|
| 1779 |
+
},
|
| 1780 |
+
"total_flos": 0.0,
|
| 1781 |
+
"train_batch_size": 1,
|
| 1782 |
+
"trial_name": null,
|
| 1783 |
+
"trial_params": null
|
| 1784 |
+
}
|