Upload folder using huggingface_hub
Browse files- checkpoint-3000/config.json +64 -0
- checkpoint-3000/experiment_cfg/metadata.json +195 -0
- checkpoint-3000/model-00001-of-00002.safetensors +3 -0
- checkpoint-3000/model-00002-of-00002.safetensors +3 -0
- checkpoint-3000/model.safetensors.index.json +0 -0
- checkpoint-3000/optimizer.pt +3 -0
- checkpoint-3000/rng_state.pth +3 -0
- checkpoint-3000/scheduler.pt +3 -0
- checkpoint-3000/trainer_state.json +2134 -0
checkpoint-3000/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 |
+
}
|
checkpoint-3000/experiment_cfg/metadata.json
ADDED
|
@@ -0,0 +1,195 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"new_embodiment": {
|
| 3 |
+
"statistics": {
|
| 4 |
+
"state": {
|
| 5 |
+
"single_arm": {
|
| 6 |
+
"max": [
|
| 7 |
+
72.44120788574219,
|
| 8 |
+
50.95326232910156,
|
| 9 |
+
100.00331115722656,
|
| 10 |
+
99.04779052734375,
|
| 11 |
+
-29.83800506591797
|
| 12 |
+
],
|
| 13 |
+
"min": [
|
| 14 |
+
-67.01437377929688,
|
| 15 |
+
-100.02632141113281,
|
| 16 |
+
-51.69486618041992,
|
| 17 |
+
-64.30675506591797,
|
| 18 |
+
-65.67131042480469
|
| 19 |
+
],
|
| 20 |
+
"mean": [
|
| 21 |
+
10.735069274902344,
|
| 22 |
+
-20.40676498413086,
|
| 23 |
+
33.425880432128906,
|
| 24 |
+
45.713680267333984,
|
| 25 |
+
-55.623966217041016
|
| 26 |
+
],
|
| 27 |
+
"std": [
|
| 28 |
+
21.354379653930664,
|
| 29 |
+
33.67801284790039,
|
| 30 |
+
38.5892333984375,
|
| 31 |
+
25.043033599853516,
|
| 32 |
+
4.391514301300049
|
| 33 |
+
],
|
| 34 |
+
"q01": [
|
| 35 |
+
-41.56745010375977,
|
| 36 |
+
-97.62313850402832,
|
| 37 |
+
-34.71778793334961,
|
| 38 |
+
-12.331634063720703,
|
| 39 |
+
-64.83274955749512
|
| 40 |
+
],
|
| 41 |
+
"q99": [
|
| 42 |
+
42.352006530761706,
|
| 43 |
+
38.30383804321288,
|
| 44 |
+
99.99868515014649,
|
| 45 |
+
94.28368362426758,
|
| 46 |
+
-35.80953811645508
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
"gripper": {
|
| 50 |
+
"max": [
|
| 51 |
+
84.74613952636719
|
| 52 |
+
],
|
| 53 |
+
"min": [
|
| 54 |
+
0.24420477449893951
|
| 55 |
+
],
|
| 56 |
+
"mean": [
|
| 57 |
+
29.522899627685547
|
| 58 |
+
],
|
| 59 |
+
"std": [
|
| 60 |
+
19.160173416137695
|
| 61 |
+
],
|
| 62 |
+
"q01": [
|
| 63 |
+
0.25476316541433336
|
| 64 |
+
],
|
| 65 |
+
"q99": [
|
| 66 |
+
70.32193046569812
|
| 67 |
+
]
|
| 68 |
+
}
|
| 69 |
+
},
|
| 70 |
+
"action": {
|
| 71 |
+
"single_arm": {
|
| 72 |
+
"max": [
|
| 73 |
+
72.62216186523438,
|
| 74 |
+
48.09161376953125,
|
| 75 |
+
100.0,
|
| 76 |
+
99.56253051757812,
|
| 77 |
+
-47.78969192504883
|
| 78 |
+
],
|
| 79 |
+
"min": [
|
| 80 |
+
-66.57963562011719,
|
| 81 |
+
-99.15181732177734,
|
| 82 |
+
-52.43735122680664,
|
| 83 |
+
-65.52931213378906,
|
| 84 |
+
-65.681396484375
|
| 85 |
+
],
|
| 86 |
+
"mean": [
|
| 87 |
+
11.098306655883789,
|
| 88 |
+
-21.99810028076172,
|
| 89 |
+
31.94671630859375,
|
| 90 |
+
45.533084869384766,
|
| 91 |
+
-56.312156677246094
|
| 92 |
+
],
|
| 93 |
+
"std": [
|
| 94 |
+
21.259096145629883,
|
| 95 |
+
33.51719284057617,
|
| 96 |
+
39.30229568481445,
|
| 97 |
+
25.75189208984375,
|
| 98 |
+
3.206831216812134
|
| 99 |
+
],
|
| 100 |
+
"q01": [
|
| 101 |
+
-38.75792385101318,
|
| 102 |
+
-98.982177734375,
|
| 103 |
+
-36.32983898162842,
|
| 104 |
+
-12.773406982421875,
|
| 105 |
+
-64.8443603515625
|
| 106 |
+
],
|
| 107 |
+
"q99": [
|
| 108 |
+
42.11114501953125,
|
| 109 |
+
36.01441940307615,
|
| 110 |
+
100.0,
|
| 111 |
+
94.92562866210938,
|
| 112 |
+
-49.72534942626953
|
| 113 |
+
]
|
| 114 |
+
},
|
| 115 |
+
"gripper": {
|
| 116 |
+
"max": [
|
| 117 |
+
84.87325286865234
|
| 118 |
+
],
|
| 119 |
+
"min": [
|
| 120 |
+
8.669766202729079e-07
|
| 121 |
+
],
|
| 122 |
+
"mean": [
|
| 123 |
+
26.466516494750977
|
| 124 |
+
],
|
| 125 |
+
"std": [
|
| 126 |
+
19.58903694152832
|
| 127 |
+
],
|
| 128 |
+
"q01": [
|
| 129 |
+
0.24529890716075897
|
| 130 |
+
],
|
| 131 |
+
"q99": [
|
| 132 |
+
73.16679931640617
|
| 133 |
+
]
|
| 134 |
+
}
|
| 135 |
+
}
|
| 136 |
+
},
|
| 137 |
+
"modalities": {
|
| 138 |
+
"video": {
|
| 139 |
+
"wrist": {
|
| 140 |
+
"resolution": [
|
| 141 |
+
640,
|
| 142 |
+
480
|
| 143 |
+
],
|
| 144 |
+
"channels": 3,
|
| 145 |
+
"fps": 30.0
|
| 146 |
+
},
|
| 147 |
+
"front": {
|
| 148 |
+
"resolution": [
|
| 149 |
+
640,
|
| 150 |
+
480
|
| 151 |
+
],
|
| 152 |
+
"channels": 3,
|
| 153 |
+
"fps": 30.0
|
| 154 |
+
}
|
| 155 |
+
},
|
| 156 |
+
"state": {
|
| 157 |
+
"single_arm": {
|
| 158 |
+
"absolute": true,
|
| 159 |
+
"rotation_type": null,
|
| 160 |
+
"shape": [
|
| 161 |
+
5
|
| 162 |
+
],
|
| 163 |
+
"continuous": true
|
| 164 |
+
},
|
| 165 |
+
"gripper": {
|
| 166 |
+
"absolute": true,
|
| 167 |
+
"rotation_type": null,
|
| 168 |
+
"shape": [
|
| 169 |
+
1
|
| 170 |
+
],
|
| 171 |
+
"continuous": true
|
| 172 |
+
}
|
| 173 |
+
},
|
| 174 |
+
"action": {
|
| 175 |
+
"single_arm": {
|
| 176 |
+
"absolute": true,
|
| 177 |
+
"rotation_type": null,
|
| 178 |
+
"shape": [
|
| 179 |
+
5
|
| 180 |
+
],
|
| 181 |
+
"continuous": true
|
| 182 |
+
},
|
| 183 |
+
"gripper": {
|
| 184 |
+
"absolute": true,
|
| 185 |
+
"rotation_type": null,
|
| 186 |
+
"shape": [
|
| 187 |
+
1
|
| 188 |
+
],
|
| 189 |
+
"continuous": true
|
| 190 |
+
}
|
| 191 |
+
}
|
| 192 |
+
},
|
| 193 |
+
"embodiment_tag": "new_embodiment"
|
| 194 |
+
}
|
| 195 |
+
}
|
checkpoint-3000/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cafb7503696ecf72cdb19b9c8c85c72d445788e5dcc42f460ceb60966effe420
|
| 3 |
+
size 4999367032
|
checkpoint-3000/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b6f89f98771f52f98395fa053304dbfa198e7ce159d40aa60ec70aeae19a659
|
| 3 |
+
size 2586705312
|
checkpoint-3000/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a4031e7f97a07063d08456aec5e885dc443333adc1aa743f8aeec2490934d4c8
|
| 3 |
+
size 8550720062
|
checkpoint-3000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34402c3b8b13e503a65545d394bc3ab52abbd330f236249734fd7ff80d59288f
|
| 3 |
+
size 14244
|
checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd8c35ca0a04d63f5ecbb378fad36530757075db084afa00aeae07333e2344cd
|
| 3 |
+
size 1064
|
checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,2134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.912621359223301,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.009708737864077669,
|
| 14 |
+
"grad_norm": 3.141545295715332,
|
| 15 |
+
"learning_rate": 1.5e-06,
|
| 16 |
+
"loss": 0.3406,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.019417475728155338,
|
| 21 |
+
"grad_norm": 3.8397164344787598,
|
| 22 |
+
"learning_rate": 3.166666666666667e-06,
|
| 23 |
+
"loss": 0.3078,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.02912621359223301,
|
| 28 |
+
"grad_norm": 1.7836706638336182,
|
| 29 |
+
"learning_rate": 4.833333333333333e-06,
|
| 30 |
+
"loss": 0.1926,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.038834951456310676,
|
| 35 |
+
"grad_norm": 1.257520079612732,
|
| 36 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 37 |
+
"loss": 0.1476,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04854368932038835,
|
| 42 |
+
"grad_norm": 1.0694407224655151,
|
| 43 |
+
"learning_rate": 8.166666666666668e-06,
|
| 44 |
+
"loss": 0.135,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05825242718446602,
|
| 49 |
+
"grad_norm": 0.9249787926673889,
|
| 50 |
+
"learning_rate": 9.833333333333333e-06,
|
| 51 |
+
"loss": 0.1516,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06796116504854369,
|
| 56 |
+
"grad_norm": 1.0610548257827759,
|
| 57 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 58 |
+
"loss": 0.1073,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07766990291262135,
|
| 63 |
+
"grad_norm": 0.9249606132507324,
|
| 64 |
+
"learning_rate": 1.3166666666666665e-05,
|
| 65 |
+
"loss": 0.0894,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08737864077669903,
|
| 70 |
+
"grad_norm": 0.6027079224586487,
|
| 71 |
+
"learning_rate": 1.4833333333333336e-05,
|
| 72 |
+
"loss": 0.0814,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.0970873786407767,
|
| 77 |
+
"grad_norm": 0.8849421739578247,
|
| 78 |
+
"learning_rate": 1.65e-05,
|
| 79 |
+
"loss": 0.0963,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.10679611650485436,
|
| 84 |
+
"grad_norm": 1.2120862007141113,
|
| 85 |
+
"learning_rate": 1.8166666666666667e-05,
|
| 86 |
+
"loss": 0.0817,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.11650485436893204,
|
| 91 |
+
"grad_norm": 0.781768262386322,
|
| 92 |
+
"learning_rate": 1.9833333333333335e-05,
|
| 93 |
+
"loss": 0.087,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1262135922330097,
|
| 98 |
+
"grad_norm": 0.5335855484008789,
|
| 99 |
+
"learning_rate": 2.15e-05,
|
| 100 |
+
"loss": 0.0502,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.13592233009708737,
|
| 105 |
+
"grad_norm": 0.6909231543540955,
|
| 106 |
+
"learning_rate": 2.3166666666666666e-05,
|
| 107 |
+
"loss": 0.0813,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.14563106796116504,
|
| 112 |
+
"grad_norm": 0.8511887192726135,
|
| 113 |
+
"learning_rate": 2.4833333333333335e-05,
|
| 114 |
+
"loss": 0.0823,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1553398058252427,
|
| 119 |
+
"grad_norm": 1.1109554767608643,
|
| 120 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 121 |
+
"loss": 0.0638,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.1650485436893204,
|
| 126 |
+
"grad_norm": 0.7318443655967712,
|
| 127 |
+
"learning_rate": 2.816666666666667e-05,
|
| 128 |
+
"loss": 0.0649,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.17475728155339806,
|
| 133 |
+
"grad_norm": 0.7571411728858948,
|
| 134 |
+
"learning_rate": 2.9833333333333335e-05,
|
| 135 |
+
"loss": 0.064,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.18446601941747573,
|
| 140 |
+
"grad_norm": 0.8281847834587097,
|
| 141 |
+
"learning_rate": 3.15e-05,
|
| 142 |
+
"loss": 0.0731,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1941747572815534,
|
| 147 |
+
"grad_norm": 0.7228130102157593,
|
| 148 |
+
"learning_rate": 3.316666666666667e-05,
|
| 149 |
+
"loss": 0.0674,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.20388349514563106,
|
| 154 |
+
"grad_norm": 0.9129646420478821,
|
| 155 |
+
"learning_rate": 3.483333333333334e-05,
|
| 156 |
+
"loss": 0.0674,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.21359223300970873,
|
| 161 |
+
"grad_norm": 0.7561999559402466,
|
| 162 |
+
"learning_rate": 3.65e-05,
|
| 163 |
+
"loss": 0.0632,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.22330097087378642,
|
| 168 |
+
"grad_norm": 0.6391080617904663,
|
| 169 |
+
"learning_rate": 3.816666666666667e-05,
|
| 170 |
+
"loss": 0.0748,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.23300970873786409,
|
| 175 |
+
"grad_norm": 0.8094485402107239,
|
| 176 |
+
"learning_rate": 3.983333333333333e-05,
|
| 177 |
+
"loss": 0.0822,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.24271844660194175,
|
| 182 |
+
"grad_norm": 0.5360986590385437,
|
| 183 |
+
"learning_rate": 4.15e-05,
|
| 184 |
+
"loss": 0.0617,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.2524271844660194,
|
| 189 |
+
"grad_norm": 0.7553185224533081,
|
| 190 |
+
"learning_rate": 4.316666666666667e-05,
|
| 191 |
+
"loss": 0.0666,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.2621359223300971,
|
| 196 |
+
"grad_norm": 0.6815693378448486,
|
| 197 |
+
"learning_rate": 4.483333333333333e-05,
|
| 198 |
+
"loss": 0.061,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.27184466019417475,
|
| 203 |
+
"grad_norm": 0.5073156356811523,
|
| 204 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 205 |
+
"loss": 0.0473,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.2815533980582524,
|
| 210 |
+
"grad_norm": 0.8138278722763062,
|
| 211 |
+
"learning_rate": 4.8166666666666674e-05,
|
| 212 |
+
"loss": 0.0652,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2912621359223301,
|
| 217 |
+
"grad_norm": 0.8348583579063416,
|
| 218 |
+
"learning_rate": 4.9833333333333336e-05,
|
| 219 |
+
"loss": 0.0746,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.30097087378640774,
|
| 224 |
+
"grad_norm": 1.0741016864776611,
|
| 225 |
+
"learning_rate": 4.999969242985639e-05,
|
| 226 |
+
"loss": 0.0718,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.3106796116504854,
|
| 231 |
+
"grad_norm": 1.182652473449707,
|
| 232 |
+
"learning_rate": 4.999862923413781e-05,
|
| 233 |
+
"loss": 0.06,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.32038834951456313,
|
| 238 |
+
"grad_norm": 0.835841953754425,
|
| 239 |
+
"learning_rate": 4.999680664797127e-05,
|
| 240 |
+
"loss": 0.0607,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.3300970873786408,
|
| 245 |
+
"grad_norm": 0.6821876764297485,
|
| 246 |
+
"learning_rate": 4.999422472672202e-05,
|
| 247 |
+
"loss": 0.0554,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.33980582524271846,
|
| 252 |
+
"grad_norm": 0.9348113536834717,
|
| 253 |
+
"learning_rate": 4.99908835488218e-05,
|
| 254 |
+
"loss": 0.0629,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.34951456310679613,
|
| 259 |
+
"grad_norm": 0.5926679372787476,
|
| 260 |
+
"learning_rate": 4.998678321576651e-05,
|
| 261 |
+
"loss": 0.0506,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3592233009708738,
|
| 266 |
+
"grad_norm": 1.0646535158157349,
|
| 267 |
+
"learning_rate": 4.9981923852113145e-05,
|
| 268 |
+
"loss": 0.0782,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.36893203883495146,
|
| 273 |
+
"grad_norm": 1.0149277448654175,
|
| 274 |
+
"learning_rate": 4.997630560547597e-05,
|
| 275 |
+
"loss": 0.0665,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3786407766990291,
|
| 280 |
+
"grad_norm": 0.9700633883476257,
|
| 281 |
+
"learning_rate": 4.996992864652204e-05,
|
| 282 |
+
"loss": 0.067,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3883495145631068,
|
| 287 |
+
"grad_norm": 1.1201013326644897,
|
| 288 |
+
"learning_rate": 4.996279316896606e-05,
|
| 289 |
+
"loss": 0.0656,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.39805825242718446,
|
| 294 |
+
"grad_norm": 0.7025376558303833,
|
| 295 |
+
"learning_rate": 4.9954899389564455e-05,
|
| 296 |
+
"loss": 0.0711,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.4077669902912621,
|
| 301 |
+
"grad_norm": 0.6064212918281555,
|
| 302 |
+
"learning_rate": 4.9946247548108794e-05,
|
| 303 |
+
"loss": 0.0572,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.4174757281553398,
|
| 308 |
+
"grad_norm": 0.5238557457923889,
|
| 309 |
+
"learning_rate": 4.993683790741852e-05,
|
| 310 |
+
"loss": 0.0633,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.42718446601941745,
|
| 315 |
+
"grad_norm": 0.599403977394104,
|
| 316 |
+
"learning_rate": 4.992667075333296e-05,
|
| 317 |
+
"loss": 0.0508,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.4368932038834951,
|
| 322 |
+
"grad_norm": 0.55897057056427,
|
| 323 |
+
"learning_rate": 4.991574639470263e-05,
|
| 324 |
+
"loss": 0.0539,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.44660194174757284,
|
| 329 |
+
"grad_norm": 0.628349244594574,
|
| 330 |
+
"learning_rate": 4.990406516337987e-05,
|
| 331 |
+
"loss": 0.0496,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.4563106796116505,
|
| 336 |
+
"grad_norm": 0.7520539164543152,
|
| 337 |
+
"learning_rate": 4.989162741420876e-05,
|
| 338 |
+
"loss": 0.0504,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.46601941747572817,
|
| 343 |
+
"grad_norm": 1.0347453355789185,
|
| 344 |
+
"learning_rate": 4.9878433525014335e-05,
|
| 345 |
+
"loss": 0.0646,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.47572815533980584,
|
| 350 |
+
"grad_norm": 1.04426109790802,
|
| 351 |
+
"learning_rate": 4.9864483896591094e-05,
|
| 352 |
+
"loss": 0.0679,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4854368932038835,
|
| 357 |
+
"grad_norm": 0.9814009070396423,
|
| 358 |
+
"learning_rate": 4.984977895269087e-05,
|
| 359 |
+
"loss": 0.0607,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.49514563106796117,
|
| 364 |
+
"grad_norm": 0.9229623079299927,
|
| 365 |
+
"learning_rate": 4.983431914000991e-05,
|
| 366 |
+
"loss": 0.0667,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.5048543689320388,
|
| 371 |
+
"grad_norm": 1.0306048393249512,
|
| 372 |
+
"learning_rate": 4.981810492817532e-05,
|
| 373 |
+
"loss": 0.0658,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.5145631067961165,
|
| 378 |
+
"grad_norm": 0.9046841859817505,
|
| 379 |
+
"learning_rate": 4.980113680973082e-05,
|
| 380 |
+
"loss": 0.0746,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.5242718446601942,
|
| 385 |
+
"grad_norm": 0.6024751663208008,
|
| 386 |
+
"learning_rate": 4.978341530012175e-05,
|
| 387 |
+
"loss": 0.0525,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.5339805825242718,
|
| 392 |
+
"grad_norm": 1.0224015712738037,
|
| 393 |
+
"learning_rate": 4.976494093767943e-05,
|
| 394 |
+
"loss": 0.0725,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5436893203883495,
|
| 399 |
+
"grad_norm": 1.0598809719085693,
|
| 400 |
+
"learning_rate": 4.9745714283604803e-05,
|
| 401 |
+
"loss": 0.0853,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5533980582524272,
|
| 406 |
+
"grad_norm": 0.6167446970939636,
|
| 407 |
+
"learning_rate": 4.972573592195139e-05,
|
| 408 |
+
"loss": 0.0612,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5631067961165048,
|
| 413 |
+
"grad_norm": 0.6139333844184875,
|
| 414 |
+
"learning_rate": 4.970500645960756e-05,
|
| 415 |
+
"loss": 0.0525,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5728155339805825,
|
| 420 |
+
"grad_norm": 0.7935957908630371,
|
| 421 |
+
"learning_rate": 4.968352652627806e-05,
|
| 422 |
+
"loss": 0.0526,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5825242718446602,
|
| 427 |
+
"grad_norm": 0.5704646706581116,
|
| 428 |
+
"learning_rate": 4.966129677446492e-05,
|
| 429 |
+
"loss": 0.0614,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5922330097087378,
|
| 434 |
+
"grad_norm": 0.5081838369369507,
|
| 435 |
+
"learning_rate": 4.9638317879447606e-05,
|
| 436 |
+
"loss": 0.0549,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.6019417475728155,
|
| 441 |
+
"grad_norm": 0.5273293256759644,
|
| 442 |
+
"learning_rate": 4.961459053926252e-05,
|
| 443 |
+
"loss": 0.0523,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.6116504854368932,
|
| 448 |
+
"grad_norm": 0.423949271440506,
|
| 449 |
+
"learning_rate": 4.9590115474681816e-05,
|
| 450 |
+
"loss": 0.0571,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.6213592233009708,
|
| 455 |
+
"grad_norm": 0.8299949169158936,
|
| 456 |
+
"learning_rate": 4.956489342919147e-05,
|
| 457 |
+
"loss": 0.0548,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.6310679611650486,
|
| 462 |
+
"grad_norm": 0.8283652067184448,
|
| 463 |
+
"learning_rate": 4.95389251689687e-05,
|
| 464 |
+
"loss": 0.0495,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.6407766990291263,
|
| 469 |
+
"grad_norm": 0.828658938407898,
|
| 470 |
+
"learning_rate": 4.9512211482858714e-05,
|
| 471 |
+
"loss": 0.0642,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6504854368932039,
|
| 476 |
+
"grad_norm": 0.7801154851913452,
|
| 477 |
+
"learning_rate": 4.948475318235073e-05,
|
| 478 |
+
"loss": 0.0467,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6601941747572816,
|
| 483 |
+
"grad_norm": 0.7108743786811829,
|
| 484 |
+
"learning_rate": 4.945655110155333e-05,
|
| 485 |
+
"loss": 0.0624,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6699029126213593,
|
| 490 |
+
"grad_norm": 0.5172573328018188,
|
| 491 |
+
"learning_rate": 4.9427606097169117e-05,
|
| 492 |
+
"loss": 0.0601,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6796116504854369,
|
| 497 |
+
"grad_norm": 0.7694447636604309,
|
| 498 |
+
"learning_rate": 4.939791904846869e-05,
|
| 499 |
+
"loss": 0.0585,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6893203883495146,
|
| 504 |
+
"grad_norm": 0.651713490486145,
|
| 505 |
+
"learning_rate": 4.9367490857263944e-05,
|
| 506 |
+
"loss": 0.0521,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6990291262135923,
|
| 511 |
+
"grad_norm": 1.2775520086288452,
|
| 512 |
+
"learning_rate": 4.9336322447880676e-05,
|
| 513 |
+
"loss": 0.0687,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.7087378640776699,
|
| 518 |
+
"grad_norm": 0.6819589138031006,
|
| 519 |
+
"learning_rate": 4.930441476713049e-05,
|
| 520 |
+
"loss": 0.047,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.7184466019417476,
|
| 525 |
+
"grad_norm": 0.8026332855224609,
|
| 526 |
+
"learning_rate": 4.927176878428206e-05,
|
| 527 |
+
"loss": 0.0419,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.7281553398058253,
|
| 532 |
+
"grad_norm": 0.7379328608512878,
|
| 533 |
+
"learning_rate": 4.923838549103166e-05,
|
| 534 |
+
"loss": 0.0547,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.7378640776699029,
|
| 539 |
+
"grad_norm": 0.5019276738166809,
|
| 540 |
+
"learning_rate": 4.920426590147304e-05,
|
| 541 |
+
"loss": 0.0492,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.7475728155339806,
|
| 546 |
+
"grad_norm": 0.8244302272796631,
|
| 547 |
+
"learning_rate": 4.916941105206666e-05,
|
| 548 |
+
"loss": 0.0542,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.7572815533980582,
|
| 553 |
+
"grad_norm": 0.4618038535118103,
|
| 554 |
+
"learning_rate": 4.9133822001608164e-05,
|
| 555 |
+
"loss": 0.0676,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7669902912621359,
|
| 560 |
+
"grad_norm": 0.5262795686721802,
|
| 561 |
+
"learning_rate": 4.9097499831196216e-05,
|
| 562 |
+
"loss": 0.0415,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7766990291262136,
|
| 567 |
+
"grad_norm": 0.537287175655365,
|
| 568 |
+
"learning_rate": 4.906044564419969e-05,
|
| 569 |
+
"loss": 0.0511,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.7864077669902912,
|
| 574 |
+
"grad_norm": 0.9852665066719055,
|
| 575 |
+
"learning_rate": 4.902266056622414e-05,
|
| 576 |
+
"loss": 0.055,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7961165048543689,
|
| 581 |
+
"grad_norm": 0.6320510506629944,
|
| 582 |
+
"learning_rate": 4.8984145745077584e-05,
|
| 583 |
+
"loss": 0.0554,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.8058252427184466,
|
| 588 |
+
"grad_norm": 0.6188237071037292,
|
| 589 |
+
"learning_rate": 4.894490235073566e-05,
|
| 590 |
+
"loss": 0.0783,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.8155339805825242,
|
| 595 |
+
"grad_norm": 0.6136628985404968,
|
| 596 |
+
"learning_rate": 4.890493157530609e-05,
|
| 597 |
+
"loss": 0.0532,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.8252427184466019,
|
| 602 |
+
"grad_norm": 0.6563002467155457,
|
| 603 |
+
"learning_rate": 4.8864234632992457e-05,
|
| 604 |
+
"loss": 0.0583,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.8349514563106796,
|
| 609 |
+
"grad_norm": 0.779821515083313,
|
| 610 |
+
"learning_rate": 4.88228127600573e-05,
|
| 611 |
+
"loss": 0.0608,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.8446601941747572,
|
| 616 |
+
"grad_norm": 0.3783950209617615,
|
| 617 |
+
"learning_rate": 4.878066721478461e-05,
|
| 618 |
+
"loss": 0.0599,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.8543689320388349,
|
| 623 |
+
"grad_norm": 0.7626412510871887,
|
| 624 |
+
"learning_rate": 4.8737799277441566e-05,
|
| 625 |
+
"loss": 0.0526,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.8640776699029126,
|
| 630 |
+
"grad_norm": 0.5785266160964966,
|
| 631 |
+
"learning_rate": 4.869421025023965e-05,
|
| 632 |
+
"loss": 0.0623,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8737864077669902,
|
| 637 |
+
"grad_norm": 0.959281861782074,
|
| 638 |
+
"learning_rate": 4.8649901457295096e-05,
|
| 639 |
+
"loss": 0.0509,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.883495145631068,
|
| 644 |
+
"grad_norm": 0.6331967115402222,
|
| 645 |
+
"learning_rate": 4.860487424458867e-05,
|
| 646 |
+
"loss": 0.0498,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8932038834951457,
|
| 651 |
+
"grad_norm": 0.7830513119697571,
|
| 652 |
+
"learning_rate": 4.8559129979924787e-05,
|
| 653 |
+
"loss": 0.0457,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.9029126213592233,
|
| 658 |
+
"grad_norm": 0.8294515013694763,
|
| 659 |
+
"learning_rate": 4.8512670052889955e-05,
|
| 660 |
+
"loss": 0.0528,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.912621359223301,
|
| 665 |
+
"grad_norm": 0.7075780034065247,
|
| 666 |
+
"learning_rate": 4.846549587481052e-05,
|
| 667 |
+
"loss": 0.0474,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.9223300970873787,
|
| 672 |
+
"grad_norm": 0.4232042729854584,
|
| 673 |
+
"learning_rate": 4.841760887870988e-05,
|
| 674 |
+
"loss": 0.042,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.9320388349514563,
|
| 679 |
+
"grad_norm": 0.6930996179580688,
|
| 680 |
+
"learning_rate": 4.836901051926489e-05,
|
| 681 |
+
"loss": 0.0616,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.941747572815534,
|
| 686 |
+
"grad_norm": 1.1047744750976562,
|
| 687 |
+
"learning_rate": 4.831970227276171e-05,
|
| 688 |
+
"loss": 0.0606,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.9514563106796117,
|
| 693 |
+
"grad_norm": 0.6250813603401184,
|
| 694 |
+
"learning_rate": 4.82696856370509e-05,
|
| 695 |
+
"loss": 0.0419,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.9611650485436893,
|
| 700 |
+
"grad_norm": 0.5649757385253906,
|
| 701 |
+
"learning_rate": 4.8218962131502e-05,
|
| 702 |
+
"loss": 0.0434,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.970873786407767,
|
| 707 |
+
"grad_norm": 0.5142524242401123,
|
| 708 |
+
"learning_rate": 4.81675332969573e-05,
|
| 709 |
+
"loss": 0.0455,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.9805825242718447,
|
| 714 |
+
"grad_norm": 0.437963604927063,
|
| 715 |
+
"learning_rate": 4.811540069568512e-05,
|
| 716 |
+
"loss": 0.0477,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9902912621359223,
|
| 721 |
+
"grad_norm": 0.9865050315856934,
|
| 722 |
+
"learning_rate": 4.8062565911332235e-05,
|
| 723 |
+
"loss": 0.0636,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.0,
|
| 728 |
+
"grad_norm": 0.4108821749687195,
|
| 729 |
+
"learning_rate": 4.8009030548875896e-05,
|
| 730 |
+
"loss": 0.0386,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.0097087378640777,
|
| 735 |
+
"grad_norm": 0.40395405888557434,
|
| 736 |
+
"learning_rate": 4.795479623457497e-05,
|
| 737 |
+
"loss": 0.0406,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.0194174757281553,
|
| 742 |
+
"grad_norm": 0.6007186770439148,
|
| 743 |
+
"learning_rate": 4.789986461592061e-05,
|
| 744 |
+
"loss": 0.05,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.029126213592233,
|
| 749 |
+
"grad_norm": 0.7217828631401062,
|
| 750 |
+
"learning_rate": 4.784423736158616e-05,
|
| 751 |
+
"loss": 0.051,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.0388349514563107,
|
| 756 |
+
"grad_norm": 0.6370289325714111,
|
| 757 |
+
"learning_rate": 4.7787916161376515e-05,
|
| 758 |
+
"loss": 0.0641,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.0485436893203883,
|
| 763 |
+
"grad_norm": 0.6220316886901855,
|
| 764 |
+
"learning_rate": 4.773090272617672e-05,
|
| 765 |
+
"loss": 0.0624,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.058252427184466,
|
| 770 |
+
"grad_norm": 0.7248903512954712,
|
| 771 |
+
"learning_rate": 4.7673198787900063e-05,
|
| 772 |
+
"loss": 0.0567,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.0679611650485437,
|
| 777 |
+
"grad_norm": 0.7164710164070129,
|
| 778 |
+
"learning_rate": 4.761480609943546e-05,
|
| 779 |
+
"loss": 0.0424,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0776699029126213,
|
| 784 |
+
"grad_norm": 0.6365995407104492,
|
| 785 |
+
"learning_rate": 4.755572643459414e-05,
|
| 786 |
+
"loss": 0.062,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.087378640776699,
|
| 791 |
+
"grad_norm": 0.5840019583702087,
|
| 792 |
+
"learning_rate": 4.7495961588055836e-05,
|
| 793 |
+
"loss": 0.0344,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.0970873786407767,
|
| 798 |
+
"grad_norm": 0.6124367713928223,
|
| 799 |
+
"learning_rate": 4.7435513375314253e-05,
|
| 800 |
+
"loss": 0.0524,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.1067961165048543,
|
| 805 |
+
"grad_norm": 0.6590617299079895,
|
| 806 |
+
"learning_rate": 4.737438363262187e-05,
|
| 807 |
+
"loss": 0.0615,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.116504854368932,
|
| 812 |
+
"grad_norm": 0.6731277108192444,
|
| 813 |
+
"learning_rate": 4.7312574216934225e-05,
|
| 814 |
+
"loss": 0.0536,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.1262135922330097,
|
| 819 |
+
"grad_norm": 0.4859127700328827,
|
| 820 |
+
"learning_rate": 4.7250087005853446e-05,
|
| 821 |
+
"loss": 0.0456,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.1359223300970873,
|
| 826 |
+
"grad_norm": 0.48708751797676086,
|
| 827 |
+
"learning_rate": 4.718692389757128e-05,
|
| 828 |
+
"loss": 0.0499,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.145631067961165,
|
| 833 |
+
"grad_norm": 0.7826804518699646,
|
| 834 |
+
"learning_rate": 4.7123086810811356e-05,
|
| 835 |
+
"loss": 0.0437,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.1553398058252426,
|
| 840 |
+
"grad_norm": 0.550841748714447,
|
| 841 |
+
"learning_rate": 4.705857768477098e-05,
|
| 842 |
+
"loss": 0.0507,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.1650485436893203,
|
| 847 |
+
"grad_norm": 0.7365569472312927,
|
| 848 |
+
"learning_rate": 4.699339847906215e-05,
|
| 849 |
+
"loss": 0.0624,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.174757281553398,
|
| 854 |
+
"grad_norm": 1.0098744630813599,
|
| 855 |
+
"learning_rate": 4.6927551173652075e-05,
|
| 856 |
+
"loss": 0.0619,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.1844660194174756,
|
| 861 |
+
"grad_norm": 0.738837718963623,
|
| 862 |
+
"learning_rate": 4.6861037768803016e-05,
|
| 863 |
+
"loss": 0.0503,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.1941747572815533,
|
| 868 |
+
"grad_norm": 0.5268725752830505,
|
| 869 |
+
"learning_rate": 4.679386028501156e-05,
|
| 870 |
+
"loss": 0.0489,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.203883495145631,
|
| 875 |
+
"grad_norm": 0.8031307458877563,
|
| 876 |
+
"learning_rate": 4.672602076294714e-05,
|
| 877 |
+
"loss": 0.0595,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.2135922330097086,
|
| 882 |
+
"grad_norm": 0.6441630721092224,
|
| 883 |
+
"learning_rate": 4.665752126339018e-05,
|
| 884 |
+
"loss": 0.0425,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.2233009708737863,
|
| 889 |
+
"grad_norm": 0.5169068574905396,
|
| 890 |
+
"learning_rate": 4.658836386716938e-05,
|
| 891 |
+
"loss": 0.0536,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.233009708737864,
|
| 896 |
+
"grad_norm": 0.6084913015365601,
|
| 897 |
+
"learning_rate": 4.65185506750986e-05,
|
| 898 |
+
"loss": 0.0349,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.2427184466019416,
|
| 903 |
+
"grad_norm": 0.7514814734458923,
|
| 904 |
+
"learning_rate": 4.6448083807912934e-05,
|
| 905 |
+
"loss": 0.053,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.2524271844660193,
|
| 910 |
+
"grad_norm": 0.49506863951683044,
|
| 911 |
+
"learning_rate": 4.637696540620441e-05,
|
| 912 |
+
"loss": 0.0497,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.262135922330097,
|
| 917 |
+
"grad_norm": 0.6199485659599304,
|
| 918 |
+
"learning_rate": 4.630519763035687e-05,
|
| 919 |
+
"loss": 0.0517,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.2718446601941746,
|
| 924 |
+
"grad_norm": 0.6022672653198242,
|
| 925 |
+
"learning_rate": 4.623278266048039e-05,
|
| 926 |
+
"loss": 0.0439,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.2815533980582523,
|
| 931 |
+
"grad_norm": 0.7462307810783386,
|
| 932 |
+
"learning_rate": 4.6159722696345045e-05,
|
| 933 |
+
"loss": 0.0454,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.29126213592233,
|
| 938 |
+
"grad_norm": 0.40218812227249146,
|
| 939 |
+
"learning_rate": 4.608601995731407e-05,
|
| 940 |
+
"loss": 0.0326,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.3009708737864076,
|
| 945 |
+
"grad_norm": 0.7571138739585876,
|
| 946 |
+
"learning_rate": 4.601167668227648e-05,
|
| 947 |
+
"loss": 0.0364,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.3106796116504853,
|
| 952 |
+
"grad_norm": 0.77698814868927,
|
| 953 |
+
"learning_rate": 4.593669512957901e-05,
|
| 954 |
+
"loss": 0.058,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.3203883495145632,
|
| 959 |
+
"grad_norm": 0.7244229316711426,
|
| 960 |
+
"learning_rate": 4.586107757695755e-05,
|
| 961 |
+
"loss": 0.056,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.3300970873786409,
|
| 966 |
+
"grad_norm": 0.5994431972503662,
|
| 967 |
+
"learning_rate": 4.578482632146793e-05,
|
| 968 |
+
"loss": 0.038,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.3398058252427185,
|
| 973 |
+
"grad_norm": 0.7894623279571533,
|
| 974 |
+
"learning_rate": 4.570794367941616e-05,
|
| 975 |
+
"loss": 0.0425,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.3495145631067962,
|
| 980 |
+
"grad_norm": 0.7884004712104797,
|
| 981 |
+
"learning_rate": 4.563043198628806e-05,
|
| 982 |
+
"loss": 0.0495,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.3592233009708738,
|
| 987 |
+
"grad_norm": 0.6530587673187256,
|
| 988 |
+
"learning_rate": 4.5552293596678294e-05,
|
| 989 |
+
"loss": 0.0414,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.3689320388349515,
|
| 994 |
+
"grad_norm": 0.8399489521980286,
|
| 995 |
+
"learning_rate": 4.5473530884218886e-05,
|
| 996 |
+
"loss": 0.0493,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.3786407766990292,
|
| 1001 |
+
"grad_norm": 0.6782814860343933,
|
| 1002 |
+
"learning_rate": 4.539414624150708e-05,
|
| 1003 |
+
"loss": 0.0472,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.3883495145631068,
|
| 1008 |
+
"grad_norm": 0.6846131682395935,
|
| 1009 |
+
"learning_rate": 4.5314142080032696e-05,
|
| 1010 |
+
"loss": 0.0335,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.3980582524271845,
|
| 1015 |
+
"grad_norm": 0.4258232116699219,
|
| 1016 |
+
"learning_rate": 4.5233520830104805e-05,
|
| 1017 |
+
"loss": 0.0576,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.4077669902912622,
|
| 1022 |
+
"grad_norm": 0.6838470101356506,
|
| 1023 |
+
"learning_rate": 4.515228494077798e-05,
|
| 1024 |
+
"loss": 0.0606,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.4174757281553398,
|
| 1029 |
+
"grad_norm": 0.5958576798439026,
|
| 1030 |
+
"learning_rate": 4.5070436879777865e-05,
|
| 1031 |
+
"loss": 0.0444,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.4271844660194175,
|
| 1036 |
+
"grad_norm": 0.48623862862586975,
|
| 1037 |
+
"learning_rate": 4.4987979133426215e-05,
|
| 1038 |
+
"loss": 0.0449,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.4368932038834952,
|
| 1043 |
+
"grad_norm": 0.5873615741729736,
|
| 1044 |
+
"learning_rate": 4.490491420656537e-05,
|
| 1045 |
+
"loss": 0.0442,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.4466019417475728,
|
| 1050 |
+
"grad_norm": 0.4631931781768799,
|
| 1051 |
+
"learning_rate": 4.482124462248217e-05,
|
| 1052 |
+
"loss": 0.0423,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.4563106796116505,
|
| 1057 |
+
"grad_norm": 0.5900189280509949,
|
| 1058 |
+
"learning_rate": 4.473697292283129e-05,
|
| 1059 |
+
"loss": 0.0485,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.4660194174757282,
|
| 1064 |
+
"grad_norm": 0.36097192764282227,
|
| 1065 |
+
"learning_rate": 4.465210166755803e-05,
|
| 1066 |
+
"loss": 0.0364,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.4757281553398058,
|
| 1071 |
+
"grad_norm": 0.6405830383300781,
|
| 1072 |
+
"learning_rate": 4.456663343482059e-05,
|
| 1073 |
+
"loss": 0.0416,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.4854368932038835,
|
| 1078 |
+
"grad_norm": 0.5862817168235779,
|
| 1079 |
+
"learning_rate": 4.44805708209117e-05,
|
| 1080 |
+
"loss": 0.0475,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.4951456310679612,
|
| 1085 |
+
"grad_norm": 0.7803816199302673,
|
| 1086 |
+
"learning_rate": 4.4393916440179786e-05,
|
| 1087 |
+
"loss": 0.0472,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.5048543689320388,
|
| 1092 |
+
"grad_norm": 0.5544291138648987,
|
| 1093 |
+
"learning_rate": 4.430667292494955e-05,
|
| 1094 |
+
"loss": 0.0361,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.5145631067961165,
|
| 1099 |
+
"grad_norm": 0.5777389407157898,
|
| 1100 |
+
"learning_rate": 4.4218842925441966e-05,
|
| 1101 |
+
"loss": 0.0439,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.5242718446601942,
|
| 1106 |
+
"grad_norm": 0.6668662428855896,
|
| 1107 |
+
"learning_rate": 4.413042910969385e-05,
|
| 1108 |
+
"loss": 0.0312,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.5339805825242718,
|
| 1113 |
+
"grad_norm": 0.8618906736373901,
|
| 1114 |
+
"learning_rate": 4.404143416347675e-05,
|
| 1115 |
+
"loss": 0.0481,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.5436893203883495,
|
| 1120 |
+
"grad_norm": 0.5840690732002258,
|
| 1121 |
+
"learning_rate": 4.395186079021537e-05,
|
| 1122 |
+
"loss": 0.0485,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.5533980582524272,
|
| 1127 |
+
"grad_norm": 0.3441118001937866,
|
| 1128 |
+
"learning_rate": 4.386171171090547e-05,
|
| 1129 |
+
"loss": 0.0401,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.5631067961165048,
|
| 1134 |
+
"grad_norm": 0.27396291494369507,
|
| 1135 |
+
"learning_rate": 4.37709896640312e-05,
|
| 1136 |
+
"loss": 0.0379,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.5728155339805825,
|
| 1141 |
+
"grad_norm": 0.4788952171802521,
|
| 1142 |
+
"learning_rate": 4.367969740548189e-05,
|
| 1143 |
+
"loss": 0.0398,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.5825242718446602,
|
| 1148 |
+
"grad_norm": 0.6418083906173706,
|
| 1149 |
+
"learning_rate": 4.358783770846836e-05,
|
| 1150 |
+
"loss": 0.0657,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.5922330097087378,
|
| 1155 |
+
"grad_norm": 0.6678286790847778,
|
| 1156 |
+
"learning_rate": 4.349541336343867e-05,
|
| 1157 |
+
"loss": 0.0466,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.6019417475728155,
|
| 1162 |
+
"grad_norm": 1.0702563524246216,
|
| 1163 |
+
"learning_rate": 4.3402427177993366e-05,
|
| 1164 |
+
"loss": 0.0452,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.6116504854368932,
|
| 1169 |
+
"grad_norm": 0.6941045522689819,
|
| 1170 |
+
"learning_rate": 4.3308881976800146e-05,
|
| 1171 |
+
"loss": 0.0521,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.6213592233009708,
|
| 1176 |
+
"grad_norm": 0.5572572350502014,
|
| 1177 |
+
"learning_rate": 4.321478060150813e-05,
|
| 1178 |
+
"loss": 0.0336,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.6310679611650487,
|
| 1183 |
+
"grad_norm": 0.5476228594779968,
|
| 1184 |
+
"learning_rate": 4.312012591066146e-05,
|
| 1185 |
+
"loss": 0.0375,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.6407766990291264,
|
| 1190 |
+
"grad_norm": 0.6187303066253662,
|
| 1191 |
+
"learning_rate": 4.302492077961253e-05,
|
| 1192 |
+
"loss": 0.041,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.650485436893204,
|
| 1197 |
+
"grad_norm": 0.8198221921920776,
|
| 1198 |
+
"learning_rate": 4.292916810043459e-05,
|
| 1199 |
+
"loss": 0.0463,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.6601941747572817,
|
| 1204 |
+
"grad_norm": 0.6056732535362244,
|
| 1205 |
+
"learning_rate": 4.283287078183392e-05,
|
| 1206 |
+
"loss": 0.0494,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.6699029126213594,
|
| 1211 |
+
"grad_norm": 0.6936320662498474,
|
| 1212 |
+
"learning_rate": 4.273603174906149e-05,
|
| 1213 |
+
"loss": 0.0413,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.679611650485437,
|
| 1218 |
+
"grad_norm": 0.6517147421836853,
|
| 1219 |
+
"learning_rate": 4.2638653943824026e-05,
|
| 1220 |
+
"loss": 0.0462,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.6893203883495147,
|
| 1225 |
+
"grad_norm": 0.5960354804992676,
|
| 1226 |
+
"learning_rate": 4.254074032419474e-05,
|
| 1227 |
+
"loss": 0.0403,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.6990291262135924,
|
| 1232 |
+
"grad_norm": 0.5907045602798462,
|
| 1233 |
+
"learning_rate": 4.244229386452342e-05,
|
| 1234 |
+
"loss": 0.0378,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.70873786407767,
|
| 1239 |
+
"grad_norm": 0.5899525284767151,
|
| 1240 |
+
"learning_rate": 4.2343317555346084e-05,
|
| 1241 |
+
"loss": 0.0457,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.7184466019417477,
|
| 1246 |
+
"grad_norm": 0.4403168261051178,
|
| 1247 |
+
"learning_rate": 4.2243814403294126e-05,
|
| 1248 |
+
"loss": 0.0317,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.7281553398058254,
|
| 1253 |
+
"grad_norm": 0.4933370351791382,
|
| 1254 |
+
"learning_rate": 4.214378743100302e-05,
|
| 1255 |
+
"loss": 0.0497,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.737864077669903,
|
| 1260 |
+
"grad_norm": 0.6774694919586182,
|
| 1261 |
+
"learning_rate": 4.204323967702045e-05,
|
| 1262 |
+
"loss": 0.0383,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.7475728155339807,
|
| 1267 |
+
"grad_norm": 0.5610571503639221,
|
| 1268 |
+
"learning_rate": 4.1942174195714066e-05,
|
| 1269 |
+
"loss": 0.0469,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.7572815533980584,
|
| 1274 |
+
"grad_norm": 0.8438130617141724,
|
| 1275 |
+
"learning_rate": 4.184059405717863e-05,
|
| 1276 |
+
"loss": 0.0478,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.766990291262136,
|
| 1281 |
+
"grad_norm": 0.8638678789138794,
|
| 1282 |
+
"learning_rate": 4.173850234714282e-05,
|
| 1283 |
+
"loss": 0.0489,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.7766990291262137,
|
| 1288 |
+
"grad_norm": 0.8042917251586914,
|
| 1289 |
+
"learning_rate": 4.1635902166875456e-05,
|
| 1290 |
+
"loss": 0.0527,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.7864077669902914,
|
| 1295 |
+
"grad_norm": 1.2712494134902954,
|
| 1296 |
+
"learning_rate": 4.1532796633091296e-05,
|
| 1297 |
+
"loss": 0.0512,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.796116504854369,
|
| 1302 |
+
"grad_norm": 0.5284692645072937,
|
| 1303 |
+
"learning_rate": 4.142918887785638e-05,
|
| 1304 |
+
"loss": 0.061,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.8058252427184467,
|
| 1309 |
+
"grad_norm": 0.6059673428535461,
|
| 1310 |
+
"learning_rate": 4.1325082048492866e-05,
|
| 1311 |
+
"loss": 0.0513,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.8155339805825244,
|
| 1316 |
+
"grad_norm": 0.6848878860473633,
|
| 1317 |
+
"learning_rate": 4.122047930748343e-05,
|
| 1318 |
+
"loss": 0.0395,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.825242718446602,
|
| 1323 |
+
"grad_norm": 0.8983065485954285,
|
| 1324 |
+
"learning_rate": 4.1115383832375174e-05,
|
| 1325 |
+
"loss": 0.0334,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.8349514563106797,
|
| 1330 |
+
"grad_norm": 0.5672741532325745,
|
| 1331 |
+
"learning_rate": 4.100979881568316e-05,
|
| 1332 |
+
"loss": 0.0384,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.8446601941747574,
|
| 1337 |
+
"grad_norm": 0.3233070969581604,
|
| 1338 |
+
"learning_rate": 4.090372746479337e-05,
|
| 1339 |
+
"loss": 0.0416,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.854368932038835,
|
| 1344 |
+
"grad_norm": 0.6278699040412903,
|
| 1345 |
+
"learning_rate": 4.0797173001865305e-05,
|
| 1346 |
+
"loss": 0.0528,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.8640776699029127,
|
| 1351 |
+
"grad_norm": 0.7250891923904419,
|
| 1352 |
+
"learning_rate": 4.069013866373409e-05,
|
| 1353 |
+
"loss": 0.0361,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.8737864077669903,
|
| 1358 |
+
"grad_norm": 1.1977667808532715,
|
| 1359 |
+
"learning_rate": 4.058262770181217e-05,
|
| 1360 |
+
"loss": 0.0455,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.883495145631068,
|
| 1365 |
+
"grad_norm": 0.7614680528640747,
|
| 1366 |
+
"learning_rate": 4.0474643381990505e-05,
|
| 1367 |
+
"loss": 0.0453,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.8932038834951457,
|
| 1372 |
+
"grad_norm": 0.7502397298812866,
|
| 1373 |
+
"learning_rate": 4.036618898453941e-05,
|
| 1374 |
+
"loss": 0.0366,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.9029126213592233,
|
| 1379 |
+
"grad_norm": 1.1674546003341675,
|
| 1380 |
+
"learning_rate": 4.025726780400886e-05,
|
| 1381 |
+
"loss": 0.0536,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.912621359223301,
|
| 1386 |
+
"grad_norm": 0.6499914526939392,
|
| 1387 |
+
"learning_rate": 4.0147883149128433e-05,
|
| 1388 |
+
"loss": 0.0355,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.9223300970873787,
|
| 1393 |
+
"grad_norm": 0.5907668471336365,
|
| 1394 |
+
"learning_rate": 4.003803834270681e-05,
|
| 1395 |
+
"loss": 0.0436,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.9320388349514563,
|
| 1400 |
+
"grad_norm": 0.6319796442985535,
|
| 1401 |
+
"learning_rate": 3.9927736721530805e-05,
|
| 1402 |
+
"loss": 0.0356,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.941747572815534,
|
| 1407 |
+
"grad_norm": 0.49555259943008423,
|
| 1408 |
+
"learning_rate": 3.981698163626406e-05,
|
| 1409 |
+
"loss": 0.0394,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.9514563106796117,
|
| 1414 |
+
"grad_norm": 0.8911224603652954,
|
| 1415 |
+
"learning_rate": 3.970577645134519e-05,
|
| 1416 |
+
"loss": 0.0593,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.9611650485436893,
|
| 1421 |
+
"grad_norm": 0.5634008049964905,
|
| 1422 |
+
"learning_rate": 3.9594124544885615e-05,
|
| 1423 |
+
"loss": 0.0428,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.970873786407767,
|
| 1428 |
+
"grad_norm": 0.4893592596054077,
|
| 1429 |
+
"learning_rate": 3.948202930856697e-05,
|
| 1430 |
+
"loss": 0.0314,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.9805825242718447,
|
| 1435 |
+
"grad_norm": 0.3447858691215515,
|
| 1436 |
+
"learning_rate": 3.936949414753803e-05,
|
| 1437 |
+
"loss": 0.0402,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.9902912621359223,
|
| 1442 |
+
"grad_norm": 0.6254267692565918,
|
| 1443 |
+
"learning_rate": 3.925652248031127e-05,
|
| 1444 |
+
"loss": 0.0453,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.0,
|
| 1449 |
+
"grad_norm": 0.6977013945579529,
|
| 1450 |
+
"learning_rate": 3.914311773865909e-05,
|
| 1451 |
+
"loss": 0.0312,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.0097087378640777,
|
| 1456 |
+
"grad_norm": 1.1433523893356323,
|
| 1457 |
+
"learning_rate": 3.902928336750945e-05,
|
| 1458 |
+
"loss": 0.0346,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.0194174757281553,
|
| 1463 |
+
"grad_norm": 0.9643350839614868,
|
| 1464 |
+
"learning_rate": 3.891502282484132e-05,
|
| 1465 |
+
"loss": 0.0387,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.029126213592233,
|
| 1470 |
+
"grad_norm": 0.4589916169643402,
|
| 1471 |
+
"learning_rate": 3.8800339581579607e-05,
|
| 1472 |
+
"loss": 0.0361,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.0388349514563107,
|
| 1477 |
+
"grad_norm": 0.383705198764801,
|
| 1478 |
+
"learning_rate": 3.868523712148971e-05,
|
| 1479 |
+
"loss": 0.0324,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.0485436893203883,
|
| 1484 |
+
"grad_norm": 0.4596363604068756,
|
| 1485 |
+
"learning_rate": 3.8569718941071684e-05,
|
| 1486 |
+
"loss": 0.0507,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.058252427184466,
|
| 1491 |
+
"grad_norm": 0.7858505845069885,
|
| 1492 |
+
"learning_rate": 3.845378854945406e-05,
|
| 1493 |
+
"loss": 0.0401,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.0679611650485437,
|
| 1498 |
+
"grad_norm": 0.5478680729866028,
|
| 1499 |
+
"learning_rate": 3.83374494682872e-05,
|
| 1500 |
+
"loss": 0.0428,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.0776699029126213,
|
| 1505 |
+
"grad_norm": 0.4434937536716461,
|
| 1506 |
+
"learning_rate": 3.822070523163636e-05,
|
| 1507 |
+
"loss": 0.0239,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.087378640776699,
|
| 1512 |
+
"grad_norm": 0.7560930848121643,
|
| 1513 |
+
"learning_rate": 3.810355938587433e-05,
|
| 1514 |
+
"loss": 0.0518,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.0970873786407767,
|
| 1519 |
+
"grad_norm": 0.7263623476028442,
|
| 1520 |
+
"learning_rate": 3.798601548957366e-05,
|
| 1521 |
+
"loss": 0.0368,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.1067961165048543,
|
| 1526 |
+
"grad_norm": 0.43963274359703064,
|
| 1527 |
+
"learning_rate": 3.786807711339863e-05,
|
| 1528 |
+
"loss": 0.0346,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.116504854368932,
|
| 1533 |
+
"grad_norm": 0.5044754147529602,
|
| 1534 |
+
"learning_rate": 3.774974783999672e-05,
|
| 1535 |
+
"loss": 0.0426,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.1262135922330097,
|
| 1540 |
+
"grad_norm": 0.5606735348701477,
|
| 1541 |
+
"learning_rate": 3.763103126388984e-05,
|
| 1542 |
+
"loss": 0.0386,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.1359223300970873,
|
| 1547 |
+
"grad_norm": 0.40492603182792664,
|
| 1548 |
+
"learning_rate": 3.751193099136505e-05,
|
| 1549 |
+
"loss": 0.032,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.145631067961165,
|
| 1554 |
+
"grad_norm": 0.5373347401618958,
|
| 1555 |
+
"learning_rate": 3.739245064036511e-05,
|
| 1556 |
+
"loss": 0.0402,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.1553398058252426,
|
| 1561 |
+
"grad_norm": 0.5568259358406067,
|
| 1562 |
+
"learning_rate": 3.727259384037852e-05,
|
| 1563 |
+
"loss": 0.0355,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.1650485436893203,
|
| 1568 |
+
"grad_norm": 0.6880197525024414,
|
| 1569 |
+
"learning_rate": 3.715236423232928e-05,
|
| 1570 |
+
"loss": 0.027,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.174757281553398,
|
| 1575 |
+
"grad_norm": 0.6524577736854553,
|
| 1576 |
+
"learning_rate": 3.703176546846627e-05,
|
| 1577 |
+
"loss": 0.0332,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.1844660194174756,
|
| 1582 |
+
"grad_norm": 0.6772493720054626,
|
| 1583 |
+
"learning_rate": 3.6910801212252343e-05,
|
| 1584 |
+
"loss": 0.0449,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.1941747572815533,
|
| 1589 |
+
"grad_norm": 0.7550950050354004,
|
| 1590 |
+
"learning_rate": 3.678947513825299e-05,
|
| 1591 |
+
"loss": 0.0279,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.203883495145631,
|
| 1596 |
+
"grad_norm": 0.5255644917488098,
|
| 1597 |
+
"learning_rate": 3.666779093202479e-05,
|
| 1598 |
+
"loss": 0.0286,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.2135922330097086,
|
| 1603 |
+
"grad_norm": 0.48359280824661255,
|
| 1604 |
+
"learning_rate": 3.654575229000334e-05,
|
| 1605 |
+
"loss": 0.0295,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.2233009708737863,
|
| 1610 |
+
"grad_norm": 0.6473977565765381,
|
| 1611 |
+
"learning_rate": 3.642336291939109e-05,
|
| 1612 |
+
"loss": 0.0498,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.233009708737864,
|
| 1617 |
+
"grad_norm": 0.4669754207134247,
|
| 1618 |
+
"learning_rate": 3.6300626538044646e-05,
|
| 1619 |
+
"loss": 0.0231,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.2427184466019416,
|
| 1624 |
+
"grad_norm": 0.49186766147613525,
|
| 1625 |
+
"learning_rate": 3.6177546874361865e-05,
|
| 1626 |
+
"loss": 0.0358,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.2524271844660193,
|
| 1631 |
+
"grad_norm": 0.6987037658691406,
|
| 1632 |
+
"learning_rate": 3.6054127667168596e-05,
|
| 1633 |
+
"loss": 0.0308,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.262135922330097,
|
| 1638 |
+
"grad_norm": 0.6978657841682434,
|
| 1639 |
+
"learning_rate": 3.5930372665605064e-05,
|
| 1640 |
+
"loss": 0.0422,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.2718446601941746,
|
| 1645 |
+
"grad_norm": 0.7777918577194214,
|
| 1646 |
+
"learning_rate": 3.580628562901206e-05,
|
| 1647 |
+
"loss": 0.0405,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.2815533980582523,
|
| 1652 |
+
"grad_norm": 0.47653138637542725,
|
| 1653 |
+
"learning_rate": 3.568187032681667e-05,
|
| 1654 |
+
"loss": 0.0386,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.29126213592233,
|
| 1659 |
+
"grad_norm": 0.3695906102657318,
|
| 1660 |
+
"learning_rate": 3.55571305384178e-05,
|
| 1661 |
+
"loss": 0.041,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.3009708737864076,
|
| 1666 |
+
"grad_norm": 0.6194424629211426,
|
| 1667 |
+
"learning_rate": 3.543207005307138e-05,
|
| 1668 |
+
"loss": 0.0357,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.3106796116504853,
|
| 1673 |
+
"grad_norm": 0.826737642288208,
|
| 1674 |
+
"learning_rate": 3.530669266977521e-05,
|
| 1675 |
+
"loss": 0.043,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.320388349514563,
|
| 1680 |
+
"grad_norm": 0.5638214349746704,
|
| 1681 |
+
"learning_rate": 3.5181002197153627e-05,
|
| 1682 |
+
"loss": 0.0419,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.3300970873786406,
|
| 1687 |
+
"grad_norm": 0.7033189535140991,
|
| 1688 |
+
"learning_rate": 3.505500245334175e-05,
|
| 1689 |
+
"loss": 0.0403,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.3398058252427183,
|
| 1694 |
+
"grad_norm": 0.596515417098999,
|
| 1695 |
+
"learning_rate": 3.4928697265869515e-05,
|
| 1696 |
+
"loss": 0.0349,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.349514563106796,
|
| 1701 |
+
"grad_norm": 0.2086048573255539,
|
| 1702 |
+
"learning_rate": 3.4802090471545426e-05,
|
| 1703 |
+
"loss": 0.0465,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.3592233009708736,
|
| 1708 |
+
"grad_norm": 0.8158115148544312,
|
| 1709 |
+
"learning_rate": 3.467518591633995e-05,
|
| 1710 |
+
"loss": 0.0409,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.3689320388349513,
|
| 1715 |
+
"grad_norm": 0.771389901638031,
|
| 1716 |
+
"learning_rate": 3.454798745526876e-05,
|
| 1717 |
+
"loss": 0.0422,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.378640776699029,
|
| 1722 |
+
"grad_norm": 0.5441147685050964,
|
| 1723 |
+
"learning_rate": 3.4420498952275566e-05,
|
| 1724 |
+
"loss": 0.0317,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.3883495145631066,
|
| 1729 |
+
"grad_norm": 0.4860936999320984,
|
| 1730 |
+
"learning_rate": 3.429272428011476e-05,
|
| 1731 |
+
"loss": 0.0289,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.3980582524271843,
|
| 1736 |
+
"grad_norm": 0.39900827407836914,
|
| 1737 |
+
"learning_rate": 3.41646673202338e-05,
|
| 1738 |
+
"loss": 0.0364,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.407766990291262,
|
| 1743 |
+
"grad_norm": 0.8306066393852234,
|
| 1744 |
+
"learning_rate": 3.403633196265525e-05,
|
| 1745 |
+
"loss": 0.0422,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.4174757281553396,
|
| 1750 |
+
"grad_norm": 0.46274441480636597,
|
| 1751 |
+
"learning_rate": 3.390772210585866e-05,
|
| 1752 |
+
"loss": 0.0278,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.4271844660194173,
|
| 1757 |
+
"grad_norm": 0.31475892663002014,
|
| 1758 |
+
"learning_rate": 3.377884165666212e-05,
|
| 1759 |
+
"loss": 0.0321,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.436893203883495,
|
| 1764 |
+
"grad_norm": 0.46006783843040466,
|
| 1765 |
+
"learning_rate": 3.3649694530103563e-05,
|
| 1766 |
+
"loss": 0.0268,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.4466019417475726,
|
| 1771 |
+
"grad_norm": 0.696368396282196,
|
| 1772 |
+
"learning_rate": 3.352028464932188e-05,
|
| 1773 |
+
"loss": 0.0332,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.4563106796116507,
|
| 1778 |
+
"grad_norm": 0.5227698087692261,
|
| 1779 |
+
"learning_rate": 3.33906159454377e-05,
|
| 1780 |
+
"loss": 0.0267,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.466019417475728,
|
| 1785 |
+
"grad_norm": 0.44031020998954773,
|
| 1786 |
+
"learning_rate": 3.3260692357434e-05,
|
| 1787 |
+
"loss": 0.0306,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.475728155339806,
|
| 1792 |
+
"grad_norm": 0.2782779335975647,
|
| 1793 |
+
"learning_rate": 3.313051783203648e-05,
|
| 1794 |
+
"loss": 0.0399,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.4854368932038833,
|
| 1799 |
+
"grad_norm": 0.5851346850395203,
|
| 1800 |
+
"learning_rate": 3.300009632359357e-05,
|
| 1801 |
+
"loss": 0.0334,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.4951456310679614,
|
| 1806 |
+
"grad_norm": 0.6514892578125,
|
| 1807 |
+
"learning_rate": 3.2869431793956425e-05,
|
| 1808 |
+
"loss": 0.0308,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.5048543689320386,
|
| 1813 |
+
"grad_norm": 0.7347185015678406,
|
| 1814 |
+
"learning_rate": 3.2738528212358514e-05,
|
| 1815 |
+
"loss": 0.0344,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.5145631067961167,
|
| 1820 |
+
"grad_norm": 0.7562367916107178,
|
| 1821 |
+
"learning_rate": 3.260738955529504e-05,
|
| 1822 |
+
"loss": 0.0248,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.524271844660194,
|
| 1827 |
+
"grad_norm": 0.6510912775993347,
|
| 1828 |
+
"learning_rate": 3.247601980640217e-05,
|
| 1829 |
+
"loss": 0.0425,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.533980582524272,
|
| 1834 |
+
"grad_norm": 0.47440388798713684,
|
| 1835 |
+
"learning_rate": 3.234442295633602e-05,
|
| 1836 |
+
"loss": 0.0344,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.5436893203883493,
|
| 1841 |
+
"grad_norm": 0.3861897885799408,
|
| 1842 |
+
"learning_rate": 3.22126030026514e-05,
|
| 1843 |
+
"loss": 0.02,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.5533980582524274,
|
| 1848 |
+
"grad_norm": 1.0448626279830933,
|
| 1849 |
+
"learning_rate": 3.208056394968043e-05,
|
| 1850 |
+
"loss": 0.0339,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.5631067961165046,
|
| 1855 |
+
"grad_norm": 0.5638805031776428,
|
| 1856 |
+
"learning_rate": 3.1948309808410866e-05,
|
| 1857 |
+
"loss": 0.0403,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.5728155339805827,
|
| 1862 |
+
"grad_norm": 0.3574475347995758,
|
| 1863 |
+
"learning_rate": 3.181584459636423e-05,
|
| 1864 |
+
"loss": 0.042,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.58252427184466,
|
| 1869 |
+
"grad_norm": 0.509300172328949,
|
| 1870 |
+
"learning_rate": 3.168317233747384e-05,
|
| 1871 |
+
"loss": 0.0448,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.592233009708738,
|
| 1876 |
+
"grad_norm": 0.3442770838737488,
|
| 1877 |
+
"learning_rate": 3.155029706196253e-05,
|
| 1878 |
+
"loss": 0.0432,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.6019417475728153,
|
| 1883 |
+
"grad_norm": 0.7308867573738098,
|
| 1884 |
+
"learning_rate": 3.141722280622021e-05,
|
| 1885 |
+
"loss": 0.0427,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.6116504854368934,
|
| 1890 |
+
"grad_norm": 0.4898527264595032,
|
| 1891 |
+
"learning_rate": 3.128395361268126e-05,
|
| 1892 |
+
"loss": 0.0392,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.6213592233009706,
|
| 1897 |
+
"grad_norm": 0.7416666746139526,
|
| 1898 |
+
"learning_rate": 3.115049352970177e-05,
|
| 1899 |
+
"loss": 0.0299,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.6310679611650487,
|
| 1904 |
+
"grad_norm": 0.7394320964813232,
|
| 1905 |
+
"learning_rate": 3.101684661143653e-05,
|
| 1906 |
+
"loss": 0.0309,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.6407766990291264,
|
| 1911 |
+
"grad_norm": 0.34724247455596924,
|
| 1912 |
+
"learning_rate": 3.088301691771585e-05,
|
| 1913 |
+
"loss": 0.045,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.650485436893204,
|
| 1918 |
+
"grad_norm": 0.6371554136276245,
|
| 1919 |
+
"learning_rate": 3.074900851392228e-05,
|
| 1920 |
+
"loss": 0.0331,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.6601941747572817,
|
| 1925 |
+
"grad_norm": 0.5572023391723633,
|
| 1926 |
+
"learning_rate": 3.061482547086712e-05,
|
| 1927 |
+
"loss": 0.0236,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.6699029126213594,
|
| 1932 |
+
"grad_norm": 0.42055222392082214,
|
| 1933 |
+
"learning_rate": 3.0480471864666687e-05,
|
| 1934 |
+
"loss": 0.0365,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.679611650485437,
|
| 1939 |
+
"grad_norm": 0.5603195428848267,
|
| 1940 |
+
"learning_rate": 3.0345951776618587e-05,
|
| 1941 |
+
"loss": 0.0292,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.6893203883495147,
|
| 1946 |
+
"grad_norm": 0.4651443064212799,
|
| 1947 |
+
"learning_rate": 3.021126929307766e-05,
|
| 1948 |
+
"loss": 0.0324,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.6990291262135924,
|
| 1953 |
+
"grad_norm": 0.6232860088348389,
|
| 1954 |
+
"learning_rate": 3.007642850533191e-05,
|
| 1955 |
+
"loss": 0.0408,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.70873786407767,
|
| 1960 |
+
"grad_norm": 0.5879865288734436,
|
| 1961 |
+
"learning_rate": 2.9941433509478156e-05,
|
| 1962 |
+
"loss": 0.0369,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.7184466019417477,
|
| 1967 |
+
"grad_norm": 0.4446578919887543,
|
| 1968 |
+
"learning_rate": 2.9806288406297676e-05,
|
| 1969 |
+
"loss": 0.0265,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.7281553398058254,
|
| 1974 |
+
"grad_norm": 0.6267426609992981,
|
| 1975 |
+
"learning_rate": 2.9670997301131586e-05,
|
| 1976 |
+
"loss": 0.0566,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.737864077669903,
|
| 1981 |
+
"grad_norm": 0.7768962979316711,
|
| 1982 |
+
"learning_rate": 2.9535564303756142e-05,
|
| 1983 |
+
"loss": 0.0355,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.7475728155339807,
|
| 1988 |
+
"grad_norm": 0.40637749433517456,
|
| 1989 |
+
"learning_rate": 2.9399993528257902e-05,
|
| 1990 |
+
"loss": 0.0372,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.7572815533980584,
|
| 1995 |
+
"grad_norm": 0.6524062752723694,
|
| 1996 |
+
"learning_rate": 2.9264289092908757e-05,
|
| 1997 |
+
"loss": 0.0421,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.766990291262136,
|
| 2002 |
+
"grad_norm": 0.3888689875602722,
|
| 2003 |
+
"learning_rate": 2.912845512004081e-05,
|
| 2004 |
+
"loss": 0.0372,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.7766990291262137,
|
| 2009 |
+
"grad_norm": 0.49758753180503845,
|
| 2010 |
+
"learning_rate": 2.8992495735921165e-05,
|
| 2011 |
+
"loss": 0.0374,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.7864077669902914,
|
| 2016 |
+
"grad_norm": 0.4377051591873169,
|
| 2017 |
+
"learning_rate": 2.8856415070626584e-05,
|
| 2018 |
+
"loss": 0.0333,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.796116504854369,
|
| 2023 |
+
"grad_norm": 0.4847152531147003,
|
| 2024 |
+
"learning_rate": 2.8720217257918032e-05,
|
| 2025 |
+
"loss": 0.0313,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.8058252427184467,
|
| 2030 |
+
"grad_norm": 0.5500916838645935,
|
| 2031 |
+
"learning_rate": 2.8583906435115047e-05,
|
| 2032 |
+
"loss": 0.0368,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.8155339805825244,
|
| 2037 |
+
"grad_norm": 0.4447888731956482,
|
| 2038 |
+
"learning_rate": 2.8447486742970176e-05,
|
| 2039 |
+
"loss": 0.0474,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.825242718446602,
|
| 2044 |
+
"grad_norm": 0.6298380494117737,
|
| 2045 |
+
"learning_rate": 2.8310962325543066e-05,
|
| 2046 |
+
"loss": 0.0238,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.8349514563106797,
|
| 2051 |
+
"grad_norm": 0.3452640175819397,
|
| 2052 |
+
"learning_rate": 2.817433733007466e-05,
|
| 2053 |
+
"loss": 0.0322,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.8446601941747574,
|
| 2058 |
+
"grad_norm": 0.8238608837127686,
|
| 2059 |
+
"learning_rate": 2.803761590686117e-05,
|
| 2060 |
+
"loss": 0.0265,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.854368932038835,
|
| 2065 |
+
"grad_norm": 0.6997093558311462,
|
| 2066 |
+
"learning_rate": 2.7900802209128058e-05,
|
| 2067 |
+
"loss": 0.0292,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.8640776699029127,
|
| 2072 |
+
"grad_norm": 0.549586832523346,
|
| 2073 |
+
"learning_rate": 2.776390039290378e-05,
|
| 2074 |
+
"loss": 0.0281,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.8737864077669903,
|
| 2079 |
+
"grad_norm": 0.35238906741142273,
|
| 2080 |
+
"learning_rate": 2.762691461689364e-05,
|
| 2081 |
+
"loss": 0.0352,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.883495145631068,
|
| 2086 |
+
"grad_norm": 0.6192727088928223,
|
| 2087 |
+
"learning_rate": 2.74898490423534e-05,
|
| 2088 |
+
"loss": 0.0287,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.8932038834951457,
|
| 2093 |
+
"grad_norm": 0.41445961594581604,
|
| 2094 |
+
"learning_rate": 2.7352707832962865e-05,
|
| 2095 |
+
"loss": 0.0318,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.9029126213592233,
|
| 2100 |
+
"grad_norm": 0.41043612360954285,
|
| 2101 |
+
"learning_rate": 2.7215495154699435e-05,
|
| 2102 |
+
"loss": 0.0303,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.912621359223301,
|
| 2107 |
+
"grad_norm": 0.4324229061603546,
|
| 2108 |
+
"learning_rate": 2.7078215175711546e-05,
|
| 2109 |
+
"loss": 0.0363,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
}
|
| 2112 |
+
],
|
| 2113 |
+
"logging_steps": 10,
|
| 2114 |
+
"max_steps": 6000,
|
| 2115 |
+
"num_input_tokens_seen": 0,
|
| 2116 |
+
"num_train_epochs": 6,
|
| 2117 |
+
"save_steps": 1000,
|
| 2118 |
+
"stateful_callbacks": {
|
| 2119 |
+
"TrainerControl": {
|
| 2120 |
+
"args": {
|
| 2121 |
+
"should_epoch_stop": false,
|
| 2122 |
+
"should_evaluate": false,
|
| 2123 |
+
"should_log": false,
|
| 2124 |
+
"should_save": true,
|
| 2125 |
+
"should_training_stop": false
|
| 2126 |
+
},
|
| 2127 |
+
"attributes": {}
|
| 2128 |
+
}
|
| 2129 |
+
},
|
| 2130 |
+
"total_flos": 0.0,
|
| 2131 |
+
"train_batch_size": 8,
|
| 2132 |
+
"trial_name": null,
|
| 2133 |
+
"trial_params": null
|
| 2134 |
+
}
|