add
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- adapter_config.json +17 -0
- adapter_model.bin +3 -0
- checkpoint-1000/optimizer.pt +3 -0
- checkpoint-1000/pytorch_model.bin +3 -0
- checkpoint-1000/rng_state.pth +3 -0
- checkpoint-1000/scaler.pt +3 -0
- checkpoint-1000/scheduler.pt +3 -0
- checkpoint-1000/trainer_state.json +316 -0
- checkpoint-1000/training_args.bin +3 -0
- checkpoint-1500/optimizer.pt +3 -0
- checkpoint-1500/pytorch_model.bin +3 -0
- checkpoint-1500/rng_state.pth +3 -0
- checkpoint-1500/scaler.pt +3 -0
- checkpoint-1500/scheduler.pt +3 -0
- checkpoint-1500/trainer_state.json +466 -0
- checkpoint-1500/training_args.bin +3 -0
- checkpoint-2000/optimizer.pt +3 -0
- checkpoint-2000/pytorch_model.bin +3 -0
- checkpoint-2000/rng_state.pth +3 -0
- checkpoint-2000/scaler.pt +3 -0
- checkpoint-2000/scheduler.pt +3 -0
- checkpoint-2000/trainer_state.json +616 -0
- checkpoint-2000/training_args.bin +3 -0
- checkpoint-2500/optimizer.pt +3 -0
- checkpoint-2500/pytorch_model.bin +3 -0
- checkpoint-2500/rng_state.pth +3 -0
- checkpoint-2500/scaler.pt +3 -0
- checkpoint-2500/scheduler.pt +3 -0
- checkpoint-2500/trainer_state.json +766 -0
- checkpoint-2500/training_args.bin +3 -0
- checkpoint-3000/optimizer.pt +3 -0
- checkpoint-3000/pytorch_model.bin +3 -0
- checkpoint-3000/rng_state.pth +3 -0
- checkpoint-3000/scaler.pt +3 -0
- checkpoint-3000/scheduler.pt +3 -0
- checkpoint-3000/trainer_state.json +916 -0
- checkpoint-3000/training_args.bin +3 -0
- checkpoint-3500/optimizer.pt +3 -0
- checkpoint-3500/pytorch_model.bin +3 -0
- checkpoint-3500/rng_state.pth +3 -0
- checkpoint-3500/scaler.pt +3 -0
- checkpoint-3500/scheduler.pt +3 -0
- checkpoint-3500/trainer_state.json +1066 -0
- checkpoint-3500/training_args.bin +3 -0
- checkpoint-4000/optimizer.pt +3 -0
- checkpoint-4000/pytorch_model.bin +3 -0
- checkpoint-4000/rng_state.pth +3 -0
- checkpoint-4000/scaler.pt +3 -0
- checkpoint-4000/scheduler.pt +3 -0
- checkpoint-4000/trainer_state.json +1216 -0
adapter_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "decapoda-research/llama-7b-hf",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"lora_alpha": 32,
|
| 8 |
+
"lora_dropout": 0.1,
|
| 9 |
+
"modules_to_save": null,
|
| 10 |
+
"peft_type": "LORA",
|
| 11 |
+
"r": 8,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"q_proj",
|
| 14 |
+
"v_proj"
|
| 15 |
+
],
|
| 16 |
+
"task_type": "CAUSAL_LM"
|
| 17 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c4b955087f09aace7a5d755a95a6792cb0db82f4a084b97b2efb972c9f4c0af
|
| 3 |
+
size 16822989
|
checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:979ac3bc6cbe88f0a465b45aa635de5ec1b976163ed6db6d50f7c7d439253edd
|
| 3 |
+
size 33629893
|
checkpoint-1000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8234c8c2a5f460f54a4623c898674414ebe417073721e170c173183c3e51cd6d
|
| 3 |
+
size 7548185429
|
checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9da99b2666c082cb0531c92fbc81ab3d3e8224b5a766766fe7f009db89afea6
|
| 3 |
+
size 14575
|
checkpoint-1000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:616f782e508bf93abe9de82fb1a8777069847068afdada0050c6f94df6e0661f
|
| 3 |
+
size 557
|
checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:428f2f0ea9063d04e8798a20929f1746aee4194634d5af1d36fa2ea04d562359
|
| 3 |
+
size 627
|
checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 7.774538386783285,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
}
|
| 310 |
+
],
|
| 311 |
+
"max_steps": 4096,
|
| 312 |
+
"num_train_epochs": 32,
|
| 313 |
+
"total_flos": 1.2995638935552e+18,
|
| 314 |
+
"trial_name": null,
|
| 315 |
+
"trial_params": null
|
| 316 |
+
}
|
checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b61f1babbbf314354bf942366de8dc63aa97ebdccb27573b76080ab5b50935
|
| 3 |
+
size 3963
|
checkpoint-1500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7be759d8e3c9449c68b4477fc8e78894f12eabe3ed4d3408bfdf307b7212790
|
| 3 |
+
size 33629893
|
checkpoint-1500/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d589a9f67ecd0e153fc208ba616150a644e7bd4d2a0266aca2d392beaadd9ad9
|
| 3 |
+
size 7548185429
|
checkpoint-1500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eff331b9eab2cd38e6a0fadd8adffff1a8b68b2fc0667f5221befa41e70a8b55
|
| 3 |
+
size 14575
|
checkpoint-1500/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:356b72cd9ae87e2bf8febf01c2dbebe9ae4ee48b6e3610acb7c32587cc971085
|
| 3 |
+
size 557
|
checkpoint-1500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:764d4d2c60f25d225c9497dc7347e571e56aa3fdf9303049b653f92a6d91c06c
|
| 3 |
+
size 627
|
checkpoint-1500/trainer_state.json
ADDED
|
@@ -0,0 +1,466 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 11.661807580174926,
|
| 5 |
+
"global_step": 1500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.93,
|
| 312 |
+
"learning_rate": 1.50341796875e-05,
|
| 313 |
+
"loss": 0.4488,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 8.09,
|
| 318 |
+
"learning_rate": 1.4936523437500001e-05,
|
| 319 |
+
"loss": 0.4443,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.24,
|
| 324 |
+
"learning_rate": 1.4838867187500002e-05,
|
| 325 |
+
"loss": 0.4646,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 8.4,
|
| 330 |
+
"learning_rate": 1.47412109375e-05,
|
| 331 |
+
"loss": 0.4513,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.55,
|
| 336 |
+
"learning_rate": 1.4643554687500001e-05,
|
| 337 |
+
"loss": 0.4394,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.71,
|
| 342 |
+
"learning_rate": 1.4545898437500002e-05,
|
| 343 |
+
"loss": 0.4337,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.86,
|
| 348 |
+
"learning_rate": 1.44482421875e-05,
|
| 349 |
+
"loss": 0.5037,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 9.02,
|
| 354 |
+
"learning_rate": 1.4350585937500001e-05,
|
| 355 |
+
"loss": 0.4209,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.17,
|
| 360 |
+
"learning_rate": 1.42529296875e-05,
|
| 361 |
+
"loss": 0.4073,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 9.33,
|
| 366 |
+
"learning_rate": 1.4155273437500001e-05,
|
| 367 |
+
"loss": 0.3879,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.48,
|
| 372 |
+
"learning_rate": 1.4057617187500002e-05,
|
| 373 |
+
"loss": 0.4434,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.64,
|
| 378 |
+
"learning_rate": 1.39599609375e-05,
|
| 379 |
+
"loss": 0.4062,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.8,
|
| 384 |
+
"learning_rate": 1.3862304687500001e-05,
|
| 385 |
+
"loss": 0.39,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.95,
|
| 390 |
+
"learning_rate": 1.3764648437500002e-05,
|
| 391 |
+
"loss": 0.4466,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 10.11,
|
| 396 |
+
"learning_rate": 1.36669921875e-05,
|
| 397 |
+
"loss": 0.4208,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 10.26,
|
| 402 |
+
"learning_rate": 1.3569335937500001e-05,
|
| 403 |
+
"loss": 0.3662,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.42,
|
| 408 |
+
"learning_rate": 1.3471679687500002e-05,
|
| 409 |
+
"loss": 0.4049,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 10.57,
|
| 414 |
+
"learning_rate": 1.33740234375e-05,
|
| 415 |
+
"loss": 0.3989,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.73,
|
| 420 |
+
"learning_rate": 1.3276367187500001e-05,
|
| 421 |
+
"loss": 0.3839,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.88,
|
| 426 |
+
"learning_rate": 1.3178710937500002e-05,
|
| 427 |
+
"loss": 0.4065,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 11.04,
|
| 432 |
+
"learning_rate": 1.30810546875e-05,
|
| 433 |
+
"loss": 0.3877,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 11.2,
|
| 438 |
+
"learning_rate": 1.2983398437500001e-05,
|
| 439 |
+
"loss": 0.4048,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.35,
|
| 444 |
+
"learning_rate": 1.2885742187500002e-05,
|
| 445 |
+
"loss": 0.3715,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.51,
|
| 450 |
+
"learning_rate": 1.27880859375e-05,
|
| 451 |
+
"loss": 0.3752,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 11.66,
|
| 456 |
+
"learning_rate": 1.2690429687500001e-05,
|
| 457 |
+
"loss": 0.3401,
|
| 458 |
+
"step": 1500
|
| 459 |
+
}
|
| 460 |
+
],
|
| 461 |
+
"max_steps": 4096,
|
| 462 |
+
"num_train_epochs": 32,
|
| 463 |
+
"total_flos": 1.9493458403328e+18,
|
| 464 |
+
"trial_name": null,
|
| 465 |
+
"trial_params": null
|
| 466 |
+
}
|
checkpoint-1500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b61f1babbbf314354bf942366de8dc63aa97ebdccb27573b76080ab5b50935
|
| 3 |
+
size 3963
|
checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dacbee21340912bc1b19620e6764f9e2dee7c229253a61caafea94ddd974a741
|
| 3 |
+
size 33629893
|
checkpoint-2000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92317da0dff52efc988b6d544fa6f157b2456b354033ff665573c79d2ee31eeb
|
| 3 |
+
size 7548185429
|
checkpoint-2000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d02125204d43e1df0083f5b0bffbdd3804358954e6cd266141cb9fedb325cdd4
|
| 3 |
+
size 14575
|
checkpoint-2000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f8b56bdf768c57d7f26c9f4a2bafcaeec7eb86582a790424fd2601003bfffdd
|
| 3 |
+
size 557
|
checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d318b1f56c85e8a7fb6ddaaa52d7cff0c567edba8a10f0e94e1e4ee35da82c8
|
| 3 |
+
size 627
|
checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 15.54907677356657,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.93,
|
| 312 |
+
"learning_rate": 1.50341796875e-05,
|
| 313 |
+
"loss": 0.4488,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 8.09,
|
| 318 |
+
"learning_rate": 1.4936523437500001e-05,
|
| 319 |
+
"loss": 0.4443,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.24,
|
| 324 |
+
"learning_rate": 1.4838867187500002e-05,
|
| 325 |
+
"loss": 0.4646,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 8.4,
|
| 330 |
+
"learning_rate": 1.47412109375e-05,
|
| 331 |
+
"loss": 0.4513,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.55,
|
| 336 |
+
"learning_rate": 1.4643554687500001e-05,
|
| 337 |
+
"loss": 0.4394,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.71,
|
| 342 |
+
"learning_rate": 1.4545898437500002e-05,
|
| 343 |
+
"loss": 0.4337,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.86,
|
| 348 |
+
"learning_rate": 1.44482421875e-05,
|
| 349 |
+
"loss": 0.5037,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 9.02,
|
| 354 |
+
"learning_rate": 1.4350585937500001e-05,
|
| 355 |
+
"loss": 0.4209,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.17,
|
| 360 |
+
"learning_rate": 1.42529296875e-05,
|
| 361 |
+
"loss": 0.4073,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 9.33,
|
| 366 |
+
"learning_rate": 1.4155273437500001e-05,
|
| 367 |
+
"loss": 0.3879,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.48,
|
| 372 |
+
"learning_rate": 1.4057617187500002e-05,
|
| 373 |
+
"loss": 0.4434,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.64,
|
| 378 |
+
"learning_rate": 1.39599609375e-05,
|
| 379 |
+
"loss": 0.4062,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.8,
|
| 384 |
+
"learning_rate": 1.3862304687500001e-05,
|
| 385 |
+
"loss": 0.39,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.95,
|
| 390 |
+
"learning_rate": 1.3764648437500002e-05,
|
| 391 |
+
"loss": 0.4466,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 10.11,
|
| 396 |
+
"learning_rate": 1.36669921875e-05,
|
| 397 |
+
"loss": 0.4208,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 10.26,
|
| 402 |
+
"learning_rate": 1.3569335937500001e-05,
|
| 403 |
+
"loss": 0.3662,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.42,
|
| 408 |
+
"learning_rate": 1.3471679687500002e-05,
|
| 409 |
+
"loss": 0.4049,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 10.57,
|
| 414 |
+
"learning_rate": 1.33740234375e-05,
|
| 415 |
+
"loss": 0.3989,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.73,
|
| 420 |
+
"learning_rate": 1.3276367187500001e-05,
|
| 421 |
+
"loss": 0.3839,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.88,
|
| 426 |
+
"learning_rate": 1.3178710937500002e-05,
|
| 427 |
+
"loss": 0.4065,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 11.04,
|
| 432 |
+
"learning_rate": 1.30810546875e-05,
|
| 433 |
+
"loss": 0.3877,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 11.2,
|
| 438 |
+
"learning_rate": 1.2983398437500001e-05,
|
| 439 |
+
"loss": 0.4048,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.35,
|
| 444 |
+
"learning_rate": 1.2885742187500002e-05,
|
| 445 |
+
"loss": 0.3715,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.51,
|
| 450 |
+
"learning_rate": 1.27880859375e-05,
|
| 451 |
+
"loss": 0.3752,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 11.66,
|
| 456 |
+
"learning_rate": 1.2690429687500001e-05,
|
| 457 |
+
"loss": 0.3401,
|
| 458 |
+
"step": 1500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 11.82,
|
| 462 |
+
"learning_rate": 1.25927734375e-05,
|
| 463 |
+
"loss": 0.3545,
|
| 464 |
+
"step": 1520
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.97,
|
| 468 |
+
"learning_rate": 1.24951171875e-05,
|
| 469 |
+
"loss": 0.3718,
|
| 470 |
+
"step": 1540
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 12.13,
|
| 474 |
+
"learning_rate": 1.2397460937500001e-05,
|
| 475 |
+
"loss": 0.3755,
|
| 476 |
+
"step": 1560
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 12.28,
|
| 480 |
+
"learning_rate": 1.22998046875e-05,
|
| 481 |
+
"loss": 0.3865,
|
| 482 |
+
"step": 1580
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 12.44,
|
| 486 |
+
"learning_rate": 1.2202148437500001e-05,
|
| 487 |
+
"loss": 0.3237,
|
| 488 |
+
"step": 1600
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 12.59,
|
| 492 |
+
"learning_rate": 1.2104492187500001e-05,
|
| 493 |
+
"loss": 0.3702,
|
| 494 |
+
"step": 1620
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 12.75,
|
| 498 |
+
"learning_rate": 1.20068359375e-05,
|
| 499 |
+
"loss": 0.3238,
|
| 500 |
+
"step": 1640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 12.91,
|
| 504 |
+
"learning_rate": 1.1909179687500001e-05,
|
| 505 |
+
"loss": 0.3373,
|
| 506 |
+
"step": 1660
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 13.06,
|
| 510 |
+
"learning_rate": 1.1811523437500002e-05,
|
| 511 |
+
"loss": 0.3486,
|
| 512 |
+
"step": 1680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 13.22,
|
| 516 |
+
"learning_rate": 1.17138671875e-05,
|
| 517 |
+
"loss": 0.362,
|
| 518 |
+
"step": 1700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 13.37,
|
| 522 |
+
"learning_rate": 1.1616210937500001e-05,
|
| 523 |
+
"loss": 0.3257,
|
| 524 |
+
"step": 1720
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 13.53,
|
| 528 |
+
"learning_rate": 1.1518554687500002e-05,
|
| 529 |
+
"loss": 0.3414,
|
| 530 |
+
"step": 1740
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 13.68,
|
| 534 |
+
"learning_rate": 1.14208984375e-05,
|
| 535 |
+
"loss": 0.3121,
|
| 536 |
+
"step": 1760
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 13.84,
|
| 540 |
+
"learning_rate": 1.1323242187500001e-05,
|
| 541 |
+
"loss": 0.3598,
|
| 542 |
+
"step": 1780
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 13.99,
|
| 546 |
+
"learning_rate": 1.1225585937500002e-05,
|
| 547 |
+
"loss": 0.2906,
|
| 548 |
+
"step": 1800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 14.15,
|
| 552 |
+
"learning_rate": 1.11279296875e-05,
|
| 553 |
+
"loss": 0.3356,
|
| 554 |
+
"step": 1820
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 14.31,
|
| 558 |
+
"learning_rate": 1.1030273437500001e-05,
|
| 559 |
+
"loss": 0.2991,
|
| 560 |
+
"step": 1840
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 14.46,
|
| 564 |
+
"learning_rate": 1.09326171875e-05,
|
| 565 |
+
"loss": 0.2987,
|
| 566 |
+
"step": 1860
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 14.62,
|
| 570 |
+
"learning_rate": 1.08349609375e-05,
|
| 571 |
+
"loss": 0.3249,
|
| 572 |
+
"step": 1880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 14.77,
|
| 576 |
+
"learning_rate": 1.0737304687500001e-05,
|
| 577 |
+
"loss": 0.3425,
|
| 578 |
+
"step": 1900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 14.93,
|
| 582 |
+
"learning_rate": 1.06396484375e-05,
|
| 583 |
+
"loss": 0.3398,
|
| 584 |
+
"step": 1920
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 15.08,
|
| 588 |
+
"learning_rate": 1.05419921875e-05,
|
| 589 |
+
"loss": 0.3018,
|
| 590 |
+
"step": 1940
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 15.24,
|
| 594 |
+
"learning_rate": 1.0444335937500001e-05,
|
| 595 |
+
"loss": 0.2475,
|
| 596 |
+
"step": 1960
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 15.39,
|
| 600 |
+
"learning_rate": 1.03466796875e-05,
|
| 601 |
+
"loss": 0.3507,
|
| 602 |
+
"step": 1980
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 15.55,
|
| 606 |
+
"learning_rate": 1.02490234375e-05,
|
| 607 |
+
"loss": 0.3084,
|
| 608 |
+
"step": 2000
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 4096,
|
| 612 |
+
"num_train_epochs": 32,
|
| 613 |
+
"total_flos": 2.5991277871104e+18,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b61f1babbbf314354bf942366de8dc63aa97ebdccb27573b76080ab5b50935
|
| 3 |
+
size 3963
|
checkpoint-2500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d0ce503ac3cfc0fb0bf9d5b0ba86746a0e747395397e162110c0b3055fa7b49
|
| 3 |
+
size 33629893
|
checkpoint-2500/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b4ea21b4a203b57985d8fcd5debaaac82022e8ea9bf67b93fae1b484be32371
|
| 3 |
+
size 7548185429
|
checkpoint-2500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bca9221408904a641ab27bbba78b92b50bc20ac10c411059ee7210723c151689
|
| 3 |
+
size 14575
|
checkpoint-2500/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94c8197602c1dd7f85c5171081176d8d544489fff7e8448e92707aa6eca14087
|
| 3 |
+
size 557
|
checkpoint-2500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e266ba6a4bc8c85cbd44824c164d6a53bd5711d4872f2290a82858dd687d6778
|
| 3 |
+
size 627
|
checkpoint-2500/trainer_state.json
ADDED
|
@@ -0,0 +1,766 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.43634596695821,
|
| 5 |
+
"global_step": 2500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.93,
|
| 312 |
+
"learning_rate": 1.50341796875e-05,
|
| 313 |
+
"loss": 0.4488,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 8.09,
|
| 318 |
+
"learning_rate": 1.4936523437500001e-05,
|
| 319 |
+
"loss": 0.4443,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.24,
|
| 324 |
+
"learning_rate": 1.4838867187500002e-05,
|
| 325 |
+
"loss": 0.4646,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 8.4,
|
| 330 |
+
"learning_rate": 1.47412109375e-05,
|
| 331 |
+
"loss": 0.4513,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.55,
|
| 336 |
+
"learning_rate": 1.4643554687500001e-05,
|
| 337 |
+
"loss": 0.4394,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.71,
|
| 342 |
+
"learning_rate": 1.4545898437500002e-05,
|
| 343 |
+
"loss": 0.4337,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.86,
|
| 348 |
+
"learning_rate": 1.44482421875e-05,
|
| 349 |
+
"loss": 0.5037,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 9.02,
|
| 354 |
+
"learning_rate": 1.4350585937500001e-05,
|
| 355 |
+
"loss": 0.4209,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.17,
|
| 360 |
+
"learning_rate": 1.42529296875e-05,
|
| 361 |
+
"loss": 0.4073,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 9.33,
|
| 366 |
+
"learning_rate": 1.4155273437500001e-05,
|
| 367 |
+
"loss": 0.3879,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.48,
|
| 372 |
+
"learning_rate": 1.4057617187500002e-05,
|
| 373 |
+
"loss": 0.4434,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.64,
|
| 378 |
+
"learning_rate": 1.39599609375e-05,
|
| 379 |
+
"loss": 0.4062,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.8,
|
| 384 |
+
"learning_rate": 1.3862304687500001e-05,
|
| 385 |
+
"loss": 0.39,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.95,
|
| 390 |
+
"learning_rate": 1.3764648437500002e-05,
|
| 391 |
+
"loss": 0.4466,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 10.11,
|
| 396 |
+
"learning_rate": 1.36669921875e-05,
|
| 397 |
+
"loss": 0.4208,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 10.26,
|
| 402 |
+
"learning_rate": 1.3569335937500001e-05,
|
| 403 |
+
"loss": 0.3662,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.42,
|
| 408 |
+
"learning_rate": 1.3471679687500002e-05,
|
| 409 |
+
"loss": 0.4049,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 10.57,
|
| 414 |
+
"learning_rate": 1.33740234375e-05,
|
| 415 |
+
"loss": 0.3989,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.73,
|
| 420 |
+
"learning_rate": 1.3276367187500001e-05,
|
| 421 |
+
"loss": 0.3839,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.88,
|
| 426 |
+
"learning_rate": 1.3178710937500002e-05,
|
| 427 |
+
"loss": 0.4065,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 11.04,
|
| 432 |
+
"learning_rate": 1.30810546875e-05,
|
| 433 |
+
"loss": 0.3877,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 11.2,
|
| 438 |
+
"learning_rate": 1.2983398437500001e-05,
|
| 439 |
+
"loss": 0.4048,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.35,
|
| 444 |
+
"learning_rate": 1.2885742187500002e-05,
|
| 445 |
+
"loss": 0.3715,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.51,
|
| 450 |
+
"learning_rate": 1.27880859375e-05,
|
| 451 |
+
"loss": 0.3752,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 11.66,
|
| 456 |
+
"learning_rate": 1.2690429687500001e-05,
|
| 457 |
+
"loss": 0.3401,
|
| 458 |
+
"step": 1500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 11.82,
|
| 462 |
+
"learning_rate": 1.25927734375e-05,
|
| 463 |
+
"loss": 0.3545,
|
| 464 |
+
"step": 1520
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.97,
|
| 468 |
+
"learning_rate": 1.24951171875e-05,
|
| 469 |
+
"loss": 0.3718,
|
| 470 |
+
"step": 1540
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 12.13,
|
| 474 |
+
"learning_rate": 1.2397460937500001e-05,
|
| 475 |
+
"loss": 0.3755,
|
| 476 |
+
"step": 1560
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 12.28,
|
| 480 |
+
"learning_rate": 1.22998046875e-05,
|
| 481 |
+
"loss": 0.3865,
|
| 482 |
+
"step": 1580
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 12.44,
|
| 486 |
+
"learning_rate": 1.2202148437500001e-05,
|
| 487 |
+
"loss": 0.3237,
|
| 488 |
+
"step": 1600
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 12.59,
|
| 492 |
+
"learning_rate": 1.2104492187500001e-05,
|
| 493 |
+
"loss": 0.3702,
|
| 494 |
+
"step": 1620
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 12.75,
|
| 498 |
+
"learning_rate": 1.20068359375e-05,
|
| 499 |
+
"loss": 0.3238,
|
| 500 |
+
"step": 1640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 12.91,
|
| 504 |
+
"learning_rate": 1.1909179687500001e-05,
|
| 505 |
+
"loss": 0.3373,
|
| 506 |
+
"step": 1660
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 13.06,
|
| 510 |
+
"learning_rate": 1.1811523437500002e-05,
|
| 511 |
+
"loss": 0.3486,
|
| 512 |
+
"step": 1680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 13.22,
|
| 516 |
+
"learning_rate": 1.17138671875e-05,
|
| 517 |
+
"loss": 0.362,
|
| 518 |
+
"step": 1700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 13.37,
|
| 522 |
+
"learning_rate": 1.1616210937500001e-05,
|
| 523 |
+
"loss": 0.3257,
|
| 524 |
+
"step": 1720
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 13.53,
|
| 528 |
+
"learning_rate": 1.1518554687500002e-05,
|
| 529 |
+
"loss": 0.3414,
|
| 530 |
+
"step": 1740
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 13.68,
|
| 534 |
+
"learning_rate": 1.14208984375e-05,
|
| 535 |
+
"loss": 0.3121,
|
| 536 |
+
"step": 1760
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 13.84,
|
| 540 |
+
"learning_rate": 1.1323242187500001e-05,
|
| 541 |
+
"loss": 0.3598,
|
| 542 |
+
"step": 1780
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 13.99,
|
| 546 |
+
"learning_rate": 1.1225585937500002e-05,
|
| 547 |
+
"loss": 0.2906,
|
| 548 |
+
"step": 1800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 14.15,
|
| 552 |
+
"learning_rate": 1.11279296875e-05,
|
| 553 |
+
"loss": 0.3356,
|
| 554 |
+
"step": 1820
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 14.31,
|
| 558 |
+
"learning_rate": 1.1030273437500001e-05,
|
| 559 |
+
"loss": 0.2991,
|
| 560 |
+
"step": 1840
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 14.46,
|
| 564 |
+
"learning_rate": 1.09326171875e-05,
|
| 565 |
+
"loss": 0.2987,
|
| 566 |
+
"step": 1860
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 14.62,
|
| 570 |
+
"learning_rate": 1.08349609375e-05,
|
| 571 |
+
"loss": 0.3249,
|
| 572 |
+
"step": 1880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 14.77,
|
| 576 |
+
"learning_rate": 1.0737304687500001e-05,
|
| 577 |
+
"loss": 0.3425,
|
| 578 |
+
"step": 1900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 14.93,
|
| 582 |
+
"learning_rate": 1.06396484375e-05,
|
| 583 |
+
"loss": 0.3398,
|
| 584 |
+
"step": 1920
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 15.08,
|
| 588 |
+
"learning_rate": 1.05419921875e-05,
|
| 589 |
+
"loss": 0.3018,
|
| 590 |
+
"step": 1940
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 15.24,
|
| 594 |
+
"learning_rate": 1.0444335937500001e-05,
|
| 595 |
+
"loss": 0.2475,
|
| 596 |
+
"step": 1960
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 15.39,
|
| 600 |
+
"learning_rate": 1.03466796875e-05,
|
| 601 |
+
"loss": 0.3507,
|
| 602 |
+
"step": 1980
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 15.55,
|
| 606 |
+
"learning_rate": 1.02490234375e-05,
|
| 607 |
+
"loss": 0.3084,
|
| 608 |
+
"step": 2000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 15.7,
|
| 612 |
+
"learning_rate": 1.0151367187500001e-05,
|
| 613 |
+
"loss": 0.3212,
|
| 614 |
+
"step": 2020
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 15.86,
|
| 618 |
+
"learning_rate": 1.00537109375e-05,
|
| 619 |
+
"loss": 0.2831,
|
| 620 |
+
"step": 2040
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 16.02,
|
| 624 |
+
"learning_rate": 9.956054687500001e-06,
|
| 625 |
+
"loss": 0.3072,
|
| 626 |
+
"step": 2060
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 16.17,
|
| 630 |
+
"learning_rate": 9.858398437500002e-06,
|
| 631 |
+
"loss": 0.3293,
|
| 632 |
+
"step": 2080
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 16.33,
|
| 636 |
+
"learning_rate": 9.7607421875e-06,
|
| 637 |
+
"loss": 0.2738,
|
| 638 |
+
"step": 2100
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 16.48,
|
| 642 |
+
"learning_rate": 9.663085937500001e-06,
|
| 643 |
+
"loss": 0.3245,
|
| 644 |
+
"step": 2120
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 16.64,
|
| 648 |
+
"learning_rate": 9.565429687500002e-06,
|
| 649 |
+
"loss": 0.2846,
|
| 650 |
+
"step": 2140
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 16.79,
|
| 654 |
+
"learning_rate": 9.4677734375e-06,
|
| 655 |
+
"loss": 0.2906,
|
| 656 |
+
"step": 2160
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 16.95,
|
| 660 |
+
"learning_rate": 9.370117187500001e-06,
|
| 661 |
+
"loss": 0.263,
|
| 662 |
+
"step": 2180
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 17.1,
|
| 666 |
+
"learning_rate": 9.2724609375e-06,
|
| 667 |
+
"loss": 0.286,
|
| 668 |
+
"step": 2200
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 17.26,
|
| 672 |
+
"learning_rate": 9.1748046875e-06,
|
| 673 |
+
"loss": 0.3152,
|
| 674 |
+
"step": 2220
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 17.41,
|
| 678 |
+
"learning_rate": 9.077148437500001e-06,
|
| 679 |
+
"loss": 0.2811,
|
| 680 |
+
"step": 2240
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 17.57,
|
| 684 |
+
"learning_rate": 8.9794921875e-06,
|
| 685 |
+
"loss": 0.257,
|
| 686 |
+
"step": 2260
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 17.73,
|
| 690 |
+
"learning_rate": 8.8818359375e-06,
|
| 691 |
+
"loss": 0.267,
|
| 692 |
+
"step": 2280
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 17.88,
|
| 696 |
+
"learning_rate": 8.784179687500001e-06,
|
| 697 |
+
"loss": 0.3056,
|
| 698 |
+
"step": 2300
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 18.04,
|
| 702 |
+
"learning_rate": 8.6865234375e-06,
|
| 703 |
+
"loss": 0.2522,
|
| 704 |
+
"step": 2320
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 18.19,
|
| 708 |
+
"learning_rate": 8.5888671875e-06,
|
| 709 |
+
"loss": 0.2806,
|
| 710 |
+
"step": 2340
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 18.35,
|
| 714 |
+
"learning_rate": 8.491210937500001e-06,
|
| 715 |
+
"loss": 0.266,
|
| 716 |
+
"step": 2360
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 18.5,
|
| 720 |
+
"learning_rate": 8.3935546875e-06,
|
| 721 |
+
"loss": 0.288,
|
| 722 |
+
"step": 2380
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 18.66,
|
| 726 |
+
"learning_rate": 8.2958984375e-06,
|
| 727 |
+
"loss": 0.2712,
|
| 728 |
+
"step": 2400
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 18.81,
|
| 732 |
+
"learning_rate": 8.198242187500001e-06,
|
| 733 |
+
"loss": 0.3099,
|
| 734 |
+
"step": 2420
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 18.97,
|
| 738 |
+
"learning_rate": 8.1005859375e-06,
|
| 739 |
+
"loss": 0.2205,
|
| 740 |
+
"step": 2440
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 19.13,
|
| 744 |
+
"learning_rate": 8.002929687500001e-06,
|
| 745 |
+
"loss": 0.253,
|
| 746 |
+
"step": 2460
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 19.28,
|
| 750 |
+
"learning_rate": 7.905273437500001e-06,
|
| 751 |
+
"loss": 0.2885,
|
| 752 |
+
"step": 2480
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 19.44,
|
| 756 |
+
"learning_rate": 7.8076171875e-06,
|
| 757 |
+
"loss": 0.2326,
|
| 758 |
+
"step": 2500
|
| 759 |
+
}
|
| 760 |
+
],
|
| 761 |
+
"max_steps": 4096,
|
| 762 |
+
"num_train_epochs": 32,
|
| 763 |
+
"total_flos": 3.248909733888e+18,
|
| 764 |
+
"trial_name": null,
|
| 765 |
+
"trial_params": null
|
| 766 |
+
}
|
checkpoint-2500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b61f1babbbf314354bf942366de8dc63aa97ebdccb27573b76080ab5b50935
|
| 3 |
+
size 3963
|
checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abd6096034801a8462add69f797380a513b314f24469c357222535bef86c3c49
|
| 3 |
+
size 33629893
|
checkpoint-3000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fadea24b4342ba15391579a69c31e51d973a4033c689ba35538b1ae86a64416
|
| 3 |
+
size 7548185429
|
checkpoint-3000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:57b99a395afb04fbb4ea183752c061c3a4f2c1c2097078e1a5e8562b178e2802
|
| 3 |
+
size 14575
|
checkpoint-3000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:513a9b45b859c887bcff749f25fc1ee060b7a5baf665003a7040bbf3fcbb9016
|
| 3 |
+
size 557
|
checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc629afeb6334b604b054705c97bd270f739029d4e5678293dbec8b8ac0c9565
|
| 3 |
+
size 627
|
checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,916 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 23.323615160349853,
|
| 5 |
+
"global_step": 3000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.93,
|
| 312 |
+
"learning_rate": 1.50341796875e-05,
|
| 313 |
+
"loss": 0.4488,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 8.09,
|
| 318 |
+
"learning_rate": 1.4936523437500001e-05,
|
| 319 |
+
"loss": 0.4443,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.24,
|
| 324 |
+
"learning_rate": 1.4838867187500002e-05,
|
| 325 |
+
"loss": 0.4646,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 8.4,
|
| 330 |
+
"learning_rate": 1.47412109375e-05,
|
| 331 |
+
"loss": 0.4513,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.55,
|
| 336 |
+
"learning_rate": 1.4643554687500001e-05,
|
| 337 |
+
"loss": 0.4394,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.71,
|
| 342 |
+
"learning_rate": 1.4545898437500002e-05,
|
| 343 |
+
"loss": 0.4337,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.86,
|
| 348 |
+
"learning_rate": 1.44482421875e-05,
|
| 349 |
+
"loss": 0.5037,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 9.02,
|
| 354 |
+
"learning_rate": 1.4350585937500001e-05,
|
| 355 |
+
"loss": 0.4209,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.17,
|
| 360 |
+
"learning_rate": 1.42529296875e-05,
|
| 361 |
+
"loss": 0.4073,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 9.33,
|
| 366 |
+
"learning_rate": 1.4155273437500001e-05,
|
| 367 |
+
"loss": 0.3879,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.48,
|
| 372 |
+
"learning_rate": 1.4057617187500002e-05,
|
| 373 |
+
"loss": 0.4434,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.64,
|
| 378 |
+
"learning_rate": 1.39599609375e-05,
|
| 379 |
+
"loss": 0.4062,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.8,
|
| 384 |
+
"learning_rate": 1.3862304687500001e-05,
|
| 385 |
+
"loss": 0.39,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.95,
|
| 390 |
+
"learning_rate": 1.3764648437500002e-05,
|
| 391 |
+
"loss": 0.4466,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 10.11,
|
| 396 |
+
"learning_rate": 1.36669921875e-05,
|
| 397 |
+
"loss": 0.4208,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 10.26,
|
| 402 |
+
"learning_rate": 1.3569335937500001e-05,
|
| 403 |
+
"loss": 0.3662,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.42,
|
| 408 |
+
"learning_rate": 1.3471679687500002e-05,
|
| 409 |
+
"loss": 0.4049,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 10.57,
|
| 414 |
+
"learning_rate": 1.33740234375e-05,
|
| 415 |
+
"loss": 0.3989,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.73,
|
| 420 |
+
"learning_rate": 1.3276367187500001e-05,
|
| 421 |
+
"loss": 0.3839,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.88,
|
| 426 |
+
"learning_rate": 1.3178710937500002e-05,
|
| 427 |
+
"loss": 0.4065,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 11.04,
|
| 432 |
+
"learning_rate": 1.30810546875e-05,
|
| 433 |
+
"loss": 0.3877,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 11.2,
|
| 438 |
+
"learning_rate": 1.2983398437500001e-05,
|
| 439 |
+
"loss": 0.4048,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.35,
|
| 444 |
+
"learning_rate": 1.2885742187500002e-05,
|
| 445 |
+
"loss": 0.3715,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.51,
|
| 450 |
+
"learning_rate": 1.27880859375e-05,
|
| 451 |
+
"loss": 0.3752,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 11.66,
|
| 456 |
+
"learning_rate": 1.2690429687500001e-05,
|
| 457 |
+
"loss": 0.3401,
|
| 458 |
+
"step": 1500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 11.82,
|
| 462 |
+
"learning_rate": 1.25927734375e-05,
|
| 463 |
+
"loss": 0.3545,
|
| 464 |
+
"step": 1520
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.97,
|
| 468 |
+
"learning_rate": 1.24951171875e-05,
|
| 469 |
+
"loss": 0.3718,
|
| 470 |
+
"step": 1540
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 12.13,
|
| 474 |
+
"learning_rate": 1.2397460937500001e-05,
|
| 475 |
+
"loss": 0.3755,
|
| 476 |
+
"step": 1560
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 12.28,
|
| 480 |
+
"learning_rate": 1.22998046875e-05,
|
| 481 |
+
"loss": 0.3865,
|
| 482 |
+
"step": 1580
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 12.44,
|
| 486 |
+
"learning_rate": 1.2202148437500001e-05,
|
| 487 |
+
"loss": 0.3237,
|
| 488 |
+
"step": 1600
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 12.59,
|
| 492 |
+
"learning_rate": 1.2104492187500001e-05,
|
| 493 |
+
"loss": 0.3702,
|
| 494 |
+
"step": 1620
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 12.75,
|
| 498 |
+
"learning_rate": 1.20068359375e-05,
|
| 499 |
+
"loss": 0.3238,
|
| 500 |
+
"step": 1640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 12.91,
|
| 504 |
+
"learning_rate": 1.1909179687500001e-05,
|
| 505 |
+
"loss": 0.3373,
|
| 506 |
+
"step": 1660
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 13.06,
|
| 510 |
+
"learning_rate": 1.1811523437500002e-05,
|
| 511 |
+
"loss": 0.3486,
|
| 512 |
+
"step": 1680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 13.22,
|
| 516 |
+
"learning_rate": 1.17138671875e-05,
|
| 517 |
+
"loss": 0.362,
|
| 518 |
+
"step": 1700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 13.37,
|
| 522 |
+
"learning_rate": 1.1616210937500001e-05,
|
| 523 |
+
"loss": 0.3257,
|
| 524 |
+
"step": 1720
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 13.53,
|
| 528 |
+
"learning_rate": 1.1518554687500002e-05,
|
| 529 |
+
"loss": 0.3414,
|
| 530 |
+
"step": 1740
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 13.68,
|
| 534 |
+
"learning_rate": 1.14208984375e-05,
|
| 535 |
+
"loss": 0.3121,
|
| 536 |
+
"step": 1760
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 13.84,
|
| 540 |
+
"learning_rate": 1.1323242187500001e-05,
|
| 541 |
+
"loss": 0.3598,
|
| 542 |
+
"step": 1780
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 13.99,
|
| 546 |
+
"learning_rate": 1.1225585937500002e-05,
|
| 547 |
+
"loss": 0.2906,
|
| 548 |
+
"step": 1800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 14.15,
|
| 552 |
+
"learning_rate": 1.11279296875e-05,
|
| 553 |
+
"loss": 0.3356,
|
| 554 |
+
"step": 1820
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 14.31,
|
| 558 |
+
"learning_rate": 1.1030273437500001e-05,
|
| 559 |
+
"loss": 0.2991,
|
| 560 |
+
"step": 1840
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 14.46,
|
| 564 |
+
"learning_rate": 1.09326171875e-05,
|
| 565 |
+
"loss": 0.2987,
|
| 566 |
+
"step": 1860
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 14.62,
|
| 570 |
+
"learning_rate": 1.08349609375e-05,
|
| 571 |
+
"loss": 0.3249,
|
| 572 |
+
"step": 1880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 14.77,
|
| 576 |
+
"learning_rate": 1.0737304687500001e-05,
|
| 577 |
+
"loss": 0.3425,
|
| 578 |
+
"step": 1900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 14.93,
|
| 582 |
+
"learning_rate": 1.06396484375e-05,
|
| 583 |
+
"loss": 0.3398,
|
| 584 |
+
"step": 1920
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 15.08,
|
| 588 |
+
"learning_rate": 1.05419921875e-05,
|
| 589 |
+
"loss": 0.3018,
|
| 590 |
+
"step": 1940
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 15.24,
|
| 594 |
+
"learning_rate": 1.0444335937500001e-05,
|
| 595 |
+
"loss": 0.2475,
|
| 596 |
+
"step": 1960
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 15.39,
|
| 600 |
+
"learning_rate": 1.03466796875e-05,
|
| 601 |
+
"loss": 0.3507,
|
| 602 |
+
"step": 1980
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 15.55,
|
| 606 |
+
"learning_rate": 1.02490234375e-05,
|
| 607 |
+
"loss": 0.3084,
|
| 608 |
+
"step": 2000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 15.7,
|
| 612 |
+
"learning_rate": 1.0151367187500001e-05,
|
| 613 |
+
"loss": 0.3212,
|
| 614 |
+
"step": 2020
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 15.86,
|
| 618 |
+
"learning_rate": 1.00537109375e-05,
|
| 619 |
+
"loss": 0.2831,
|
| 620 |
+
"step": 2040
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 16.02,
|
| 624 |
+
"learning_rate": 9.956054687500001e-06,
|
| 625 |
+
"loss": 0.3072,
|
| 626 |
+
"step": 2060
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 16.17,
|
| 630 |
+
"learning_rate": 9.858398437500002e-06,
|
| 631 |
+
"loss": 0.3293,
|
| 632 |
+
"step": 2080
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 16.33,
|
| 636 |
+
"learning_rate": 9.7607421875e-06,
|
| 637 |
+
"loss": 0.2738,
|
| 638 |
+
"step": 2100
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 16.48,
|
| 642 |
+
"learning_rate": 9.663085937500001e-06,
|
| 643 |
+
"loss": 0.3245,
|
| 644 |
+
"step": 2120
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 16.64,
|
| 648 |
+
"learning_rate": 9.565429687500002e-06,
|
| 649 |
+
"loss": 0.2846,
|
| 650 |
+
"step": 2140
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 16.79,
|
| 654 |
+
"learning_rate": 9.4677734375e-06,
|
| 655 |
+
"loss": 0.2906,
|
| 656 |
+
"step": 2160
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 16.95,
|
| 660 |
+
"learning_rate": 9.370117187500001e-06,
|
| 661 |
+
"loss": 0.263,
|
| 662 |
+
"step": 2180
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 17.1,
|
| 666 |
+
"learning_rate": 9.2724609375e-06,
|
| 667 |
+
"loss": 0.286,
|
| 668 |
+
"step": 2200
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 17.26,
|
| 672 |
+
"learning_rate": 9.1748046875e-06,
|
| 673 |
+
"loss": 0.3152,
|
| 674 |
+
"step": 2220
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 17.41,
|
| 678 |
+
"learning_rate": 9.077148437500001e-06,
|
| 679 |
+
"loss": 0.2811,
|
| 680 |
+
"step": 2240
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 17.57,
|
| 684 |
+
"learning_rate": 8.9794921875e-06,
|
| 685 |
+
"loss": 0.257,
|
| 686 |
+
"step": 2260
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 17.73,
|
| 690 |
+
"learning_rate": 8.8818359375e-06,
|
| 691 |
+
"loss": 0.267,
|
| 692 |
+
"step": 2280
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 17.88,
|
| 696 |
+
"learning_rate": 8.784179687500001e-06,
|
| 697 |
+
"loss": 0.3056,
|
| 698 |
+
"step": 2300
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 18.04,
|
| 702 |
+
"learning_rate": 8.6865234375e-06,
|
| 703 |
+
"loss": 0.2522,
|
| 704 |
+
"step": 2320
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 18.19,
|
| 708 |
+
"learning_rate": 8.5888671875e-06,
|
| 709 |
+
"loss": 0.2806,
|
| 710 |
+
"step": 2340
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 18.35,
|
| 714 |
+
"learning_rate": 8.491210937500001e-06,
|
| 715 |
+
"loss": 0.266,
|
| 716 |
+
"step": 2360
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 18.5,
|
| 720 |
+
"learning_rate": 8.3935546875e-06,
|
| 721 |
+
"loss": 0.288,
|
| 722 |
+
"step": 2380
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 18.66,
|
| 726 |
+
"learning_rate": 8.2958984375e-06,
|
| 727 |
+
"loss": 0.2712,
|
| 728 |
+
"step": 2400
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 18.81,
|
| 732 |
+
"learning_rate": 8.198242187500001e-06,
|
| 733 |
+
"loss": 0.3099,
|
| 734 |
+
"step": 2420
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 18.97,
|
| 738 |
+
"learning_rate": 8.1005859375e-06,
|
| 739 |
+
"loss": 0.2205,
|
| 740 |
+
"step": 2440
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 19.13,
|
| 744 |
+
"learning_rate": 8.002929687500001e-06,
|
| 745 |
+
"loss": 0.253,
|
| 746 |
+
"step": 2460
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 19.28,
|
| 750 |
+
"learning_rate": 7.905273437500001e-06,
|
| 751 |
+
"loss": 0.2885,
|
| 752 |
+
"step": 2480
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 19.44,
|
| 756 |
+
"learning_rate": 7.8076171875e-06,
|
| 757 |
+
"loss": 0.2326,
|
| 758 |
+
"step": 2500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 19.59,
|
| 762 |
+
"learning_rate": 7.709960937500001e-06,
|
| 763 |
+
"loss": 0.255,
|
| 764 |
+
"step": 2520
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 19.75,
|
| 768 |
+
"learning_rate": 7.612304687500001e-06,
|
| 769 |
+
"loss": 0.2698,
|
| 770 |
+
"step": 2540
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 19.9,
|
| 774 |
+
"learning_rate": 7.5146484375000004e-06,
|
| 775 |
+
"loss": 0.2532,
|
| 776 |
+
"step": 2560
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 20.06,
|
| 780 |
+
"learning_rate": 7.4169921875e-06,
|
| 781 |
+
"loss": 0.2562,
|
| 782 |
+
"step": 2580
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 20.21,
|
| 786 |
+
"learning_rate": 7.319335937500001e-06,
|
| 787 |
+
"loss": 0.2717,
|
| 788 |
+
"step": 2600
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 20.37,
|
| 792 |
+
"learning_rate": 7.2216796875000005e-06,
|
| 793 |
+
"loss": 0.2652,
|
| 794 |
+
"step": 2620
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 20.52,
|
| 798 |
+
"learning_rate": 7.1240234375e-06,
|
| 799 |
+
"loss": 0.2654,
|
| 800 |
+
"step": 2640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 20.68,
|
| 804 |
+
"learning_rate": 7.026367187500001e-06,
|
| 805 |
+
"loss": 0.2766,
|
| 806 |
+
"step": 2660
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 20.84,
|
| 810 |
+
"learning_rate": 6.928710937500001e-06,
|
| 811 |
+
"loss": 0.2246,
|
| 812 |
+
"step": 2680
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 20.99,
|
| 816 |
+
"learning_rate": 6.8310546875e-06,
|
| 817 |
+
"loss": 0.2546,
|
| 818 |
+
"step": 2700
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 21.15,
|
| 822 |
+
"learning_rate": 6.733398437500001e-06,
|
| 823 |
+
"loss": 0.2708,
|
| 824 |
+
"step": 2720
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 21.3,
|
| 828 |
+
"learning_rate": 6.635742187500001e-06,
|
| 829 |
+
"loss": 0.2385,
|
| 830 |
+
"step": 2740
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 21.46,
|
| 834 |
+
"learning_rate": 6.5380859375e-06,
|
| 835 |
+
"loss": 0.2271,
|
| 836 |
+
"step": 2760
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 21.61,
|
| 840 |
+
"learning_rate": 6.4404296875e-06,
|
| 841 |
+
"loss": 0.2874,
|
| 842 |
+
"step": 2780
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 21.77,
|
| 846 |
+
"learning_rate": 6.342773437500001e-06,
|
| 847 |
+
"loss": 0.2447,
|
| 848 |
+
"step": 2800
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 21.92,
|
| 852 |
+
"learning_rate": 6.2451171875000005e-06,
|
| 853 |
+
"loss": 0.2309,
|
| 854 |
+
"step": 2820
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 22.08,
|
| 858 |
+
"learning_rate": 6.1474609375e-06,
|
| 859 |
+
"loss": 0.2587,
|
| 860 |
+
"step": 2840
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 22.24,
|
| 864 |
+
"learning_rate": 6.049804687500001e-06,
|
| 865 |
+
"loss": 0.2303,
|
| 866 |
+
"step": 2860
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 22.39,
|
| 870 |
+
"learning_rate": 5.952148437500001e-06,
|
| 871 |
+
"loss": 0.2376,
|
| 872 |
+
"step": 2880
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 22.55,
|
| 876 |
+
"learning_rate": 5.8544921875e-06,
|
| 877 |
+
"loss": 0.2696,
|
| 878 |
+
"step": 2900
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 22.7,
|
| 882 |
+
"learning_rate": 5.7568359375e-06,
|
| 883 |
+
"loss": 0.2459,
|
| 884 |
+
"step": 2920
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 22.86,
|
| 888 |
+
"learning_rate": 5.659179687500001e-06,
|
| 889 |
+
"loss": 0.2032,
|
| 890 |
+
"step": 2940
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 23.01,
|
| 894 |
+
"learning_rate": 5.5615234375e-06,
|
| 895 |
+
"loss": 0.2491,
|
| 896 |
+
"step": 2960
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 23.17,
|
| 900 |
+
"learning_rate": 5.4638671875e-06,
|
| 901 |
+
"loss": 0.2395,
|
| 902 |
+
"step": 2980
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 23.32,
|
| 906 |
+
"learning_rate": 5.366210937500001e-06,
|
| 907 |
+
"loss": 0.2573,
|
| 908 |
+
"step": 3000
|
| 909 |
+
}
|
| 910 |
+
],
|
| 911 |
+
"max_steps": 4096,
|
| 912 |
+
"num_train_epochs": 32,
|
| 913 |
+
"total_flos": 3.8986916806656e+18,
|
| 914 |
+
"trial_name": null,
|
| 915 |
+
"trial_params": null
|
| 916 |
+
}
|
checkpoint-3000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b61f1babbbf314354bf942366de8dc63aa97ebdccb27573b76080ab5b50935
|
| 3 |
+
size 3963
|
checkpoint-3500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d4004f9fccd306f50b7742214a698c4f5527d7017c24d5545c4d9b3db6a85977
|
| 3 |
+
size 33629893
|
checkpoint-3500/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dee3137d50f8ccd0d3b7e6e53eba049a5066b881ce7dedb55584a335b4f505fa
|
| 3 |
+
size 7548185429
|
checkpoint-3500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b252559e55f1cebc645f9e33d786a9d51b1c14fae33d7291871ffe2763bda82e
|
| 3 |
+
size 14575
|
checkpoint-3500/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32a90a65317216d30151b19ae7ffa5171734912b4676ce34dfcb16507da25212
|
| 3 |
+
size 557
|
checkpoint-3500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f059ff70a4abf1019a77d47012bc09d5d465a3c63eeb7b984cf7d7d16ed46e4
|
| 3 |
+
size 627
|
checkpoint-3500/trainer_state.json
ADDED
|
@@ -0,0 +1,1066 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 27.2108843537415,
|
| 5 |
+
"global_step": 3500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.93,
|
| 312 |
+
"learning_rate": 1.50341796875e-05,
|
| 313 |
+
"loss": 0.4488,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 8.09,
|
| 318 |
+
"learning_rate": 1.4936523437500001e-05,
|
| 319 |
+
"loss": 0.4443,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.24,
|
| 324 |
+
"learning_rate": 1.4838867187500002e-05,
|
| 325 |
+
"loss": 0.4646,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 8.4,
|
| 330 |
+
"learning_rate": 1.47412109375e-05,
|
| 331 |
+
"loss": 0.4513,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.55,
|
| 336 |
+
"learning_rate": 1.4643554687500001e-05,
|
| 337 |
+
"loss": 0.4394,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.71,
|
| 342 |
+
"learning_rate": 1.4545898437500002e-05,
|
| 343 |
+
"loss": 0.4337,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.86,
|
| 348 |
+
"learning_rate": 1.44482421875e-05,
|
| 349 |
+
"loss": 0.5037,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 9.02,
|
| 354 |
+
"learning_rate": 1.4350585937500001e-05,
|
| 355 |
+
"loss": 0.4209,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.17,
|
| 360 |
+
"learning_rate": 1.42529296875e-05,
|
| 361 |
+
"loss": 0.4073,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 9.33,
|
| 366 |
+
"learning_rate": 1.4155273437500001e-05,
|
| 367 |
+
"loss": 0.3879,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.48,
|
| 372 |
+
"learning_rate": 1.4057617187500002e-05,
|
| 373 |
+
"loss": 0.4434,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.64,
|
| 378 |
+
"learning_rate": 1.39599609375e-05,
|
| 379 |
+
"loss": 0.4062,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.8,
|
| 384 |
+
"learning_rate": 1.3862304687500001e-05,
|
| 385 |
+
"loss": 0.39,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.95,
|
| 390 |
+
"learning_rate": 1.3764648437500002e-05,
|
| 391 |
+
"loss": 0.4466,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 10.11,
|
| 396 |
+
"learning_rate": 1.36669921875e-05,
|
| 397 |
+
"loss": 0.4208,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 10.26,
|
| 402 |
+
"learning_rate": 1.3569335937500001e-05,
|
| 403 |
+
"loss": 0.3662,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.42,
|
| 408 |
+
"learning_rate": 1.3471679687500002e-05,
|
| 409 |
+
"loss": 0.4049,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 10.57,
|
| 414 |
+
"learning_rate": 1.33740234375e-05,
|
| 415 |
+
"loss": 0.3989,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.73,
|
| 420 |
+
"learning_rate": 1.3276367187500001e-05,
|
| 421 |
+
"loss": 0.3839,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.88,
|
| 426 |
+
"learning_rate": 1.3178710937500002e-05,
|
| 427 |
+
"loss": 0.4065,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 11.04,
|
| 432 |
+
"learning_rate": 1.30810546875e-05,
|
| 433 |
+
"loss": 0.3877,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 11.2,
|
| 438 |
+
"learning_rate": 1.2983398437500001e-05,
|
| 439 |
+
"loss": 0.4048,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.35,
|
| 444 |
+
"learning_rate": 1.2885742187500002e-05,
|
| 445 |
+
"loss": 0.3715,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.51,
|
| 450 |
+
"learning_rate": 1.27880859375e-05,
|
| 451 |
+
"loss": 0.3752,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 11.66,
|
| 456 |
+
"learning_rate": 1.2690429687500001e-05,
|
| 457 |
+
"loss": 0.3401,
|
| 458 |
+
"step": 1500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 11.82,
|
| 462 |
+
"learning_rate": 1.25927734375e-05,
|
| 463 |
+
"loss": 0.3545,
|
| 464 |
+
"step": 1520
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.97,
|
| 468 |
+
"learning_rate": 1.24951171875e-05,
|
| 469 |
+
"loss": 0.3718,
|
| 470 |
+
"step": 1540
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 12.13,
|
| 474 |
+
"learning_rate": 1.2397460937500001e-05,
|
| 475 |
+
"loss": 0.3755,
|
| 476 |
+
"step": 1560
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 12.28,
|
| 480 |
+
"learning_rate": 1.22998046875e-05,
|
| 481 |
+
"loss": 0.3865,
|
| 482 |
+
"step": 1580
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 12.44,
|
| 486 |
+
"learning_rate": 1.2202148437500001e-05,
|
| 487 |
+
"loss": 0.3237,
|
| 488 |
+
"step": 1600
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 12.59,
|
| 492 |
+
"learning_rate": 1.2104492187500001e-05,
|
| 493 |
+
"loss": 0.3702,
|
| 494 |
+
"step": 1620
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 12.75,
|
| 498 |
+
"learning_rate": 1.20068359375e-05,
|
| 499 |
+
"loss": 0.3238,
|
| 500 |
+
"step": 1640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 12.91,
|
| 504 |
+
"learning_rate": 1.1909179687500001e-05,
|
| 505 |
+
"loss": 0.3373,
|
| 506 |
+
"step": 1660
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 13.06,
|
| 510 |
+
"learning_rate": 1.1811523437500002e-05,
|
| 511 |
+
"loss": 0.3486,
|
| 512 |
+
"step": 1680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 13.22,
|
| 516 |
+
"learning_rate": 1.17138671875e-05,
|
| 517 |
+
"loss": 0.362,
|
| 518 |
+
"step": 1700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 13.37,
|
| 522 |
+
"learning_rate": 1.1616210937500001e-05,
|
| 523 |
+
"loss": 0.3257,
|
| 524 |
+
"step": 1720
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 13.53,
|
| 528 |
+
"learning_rate": 1.1518554687500002e-05,
|
| 529 |
+
"loss": 0.3414,
|
| 530 |
+
"step": 1740
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 13.68,
|
| 534 |
+
"learning_rate": 1.14208984375e-05,
|
| 535 |
+
"loss": 0.3121,
|
| 536 |
+
"step": 1760
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 13.84,
|
| 540 |
+
"learning_rate": 1.1323242187500001e-05,
|
| 541 |
+
"loss": 0.3598,
|
| 542 |
+
"step": 1780
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 13.99,
|
| 546 |
+
"learning_rate": 1.1225585937500002e-05,
|
| 547 |
+
"loss": 0.2906,
|
| 548 |
+
"step": 1800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 14.15,
|
| 552 |
+
"learning_rate": 1.11279296875e-05,
|
| 553 |
+
"loss": 0.3356,
|
| 554 |
+
"step": 1820
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 14.31,
|
| 558 |
+
"learning_rate": 1.1030273437500001e-05,
|
| 559 |
+
"loss": 0.2991,
|
| 560 |
+
"step": 1840
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 14.46,
|
| 564 |
+
"learning_rate": 1.09326171875e-05,
|
| 565 |
+
"loss": 0.2987,
|
| 566 |
+
"step": 1860
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 14.62,
|
| 570 |
+
"learning_rate": 1.08349609375e-05,
|
| 571 |
+
"loss": 0.3249,
|
| 572 |
+
"step": 1880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 14.77,
|
| 576 |
+
"learning_rate": 1.0737304687500001e-05,
|
| 577 |
+
"loss": 0.3425,
|
| 578 |
+
"step": 1900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 14.93,
|
| 582 |
+
"learning_rate": 1.06396484375e-05,
|
| 583 |
+
"loss": 0.3398,
|
| 584 |
+
"step": 1920
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 15.08,
|
| 588 |
+
"learning_rate": 1.05419921875e-05,
|
| 589 |
+
"loss": 0.3018,
|
| 590 |
+
"step": 1940
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 15.24,
|
| 594 |
+
"learning_rate": 1.0444335937500001e-05,
|
| 595 |
+
"loss": 0.2475,
|
| 596 |
+
"step": 1960
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 15.39,
|
| 600 |
+
"learning_rate": 1.03466796875e-05,
|
| 601 |
+
"loss": 0.3507,
|
| 602 |
+
"step": 1980
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 15.55,
|
| 606 |
+
"learning_rate": 1.02490234375e-05,
|
| 607 |
+
"loss": 0.3084,
|
| 608 |
+
"step": 2000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 15.7,
|
| 612 |
+
"learning_rate": 1.0151367187500001e-05,
|
| 613 |
+
"loss": 0.3212,
|
| 614 |
+
"step": 2020
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 15.86,
|
| 618 |
+
"learning_rate": 1.00537109375e-05,
|
| 619 |
+
"loss": 0.2831,
|
| 620 |
+
"step": 2040
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 16.02,
|
| 624 |
+
"learning_rate": 9.956054687500001e-06,
|
| 625 |
+
"loss": 0.3072,
|
| 626 |
+
"step": 2060
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 16.17,
|
| 630 |
+
"learning_rate": 9.858398437500002e-06,
|
| 631 |
+
"loss": 0.3293,
|
| 632 |
+
"step": 2080
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 16.33,
|
| 636 |
+
"learning_rate": 9.7607421875e-06,
|
| 637 |
+
"loss": 0.2738,
|
| 638 |
+
"step": 2100
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 16.48,
|
| 642 |
+
"learning_rate": 9.663085937500001e-06,
|
| 643 |
+
"loss": 0.3245,
|
| 644 |
+
"step": 2120
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 16.64,
|
| 648 |
+
"learning_rate": 9.565429687500002e-06,
|
| 649 |
+
"loss": 0.2846,
|
| 650 |
+
"step": 2140
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 16.79,
|
| 654 |
+
"learning_rate": 9.4677734375e-06,
|
| 655 |
+
"loss": 0.2906,
|
| 656 |
+
"step": 2160
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 16.95,
|
| 660 |
+
"learning_rate": 9.370117187500001e-06,
|
| 661 |
+
"loss": 0.263,
|
| 662 |
+
"step": 2180
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 17.1,
|
| 666 |
+
"learning_rate": 9.2724609375e-06,
|
| 667 |
+
"loss": 0.286,
|
| 668 |
+
"step": 2200
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 17.26,
|
| 672 |
+
"learning_rate": 9.1748046875e-06,
|
| 673 |
+
"loss": 0.3152,
|
| 674 |
+
"step": 2220
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 17.41,
|
| 678 |
+
"learning_rate": 9.077148437500001e-06,
|
| 679 |
+
"loss": 0.2811,
|
| 680 |
+
"step": 2240
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 17.57,
|
| 684 |
+
"learning_rate": 8.9794921875e-06,
|
| 685 |
+
"loss": 0.257,
|
| 686 |
+
"step": 2260
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 17.73,
|
| 690 |
+
"learning_rate": 8.8818359375e-06,
|
| 691 |
+
"loss": 0.267,
|
| 692 |
+
"step": 2280
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 17.88,
|
| 696 |
+
"learning_rate": 8.784179687500001e-06,
|
| 697 |
+
"loss": 0.3056,
|
| 698 |
+
"step": 2300
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 18.04,
|
| 702 |
+
"learning_rate": 8.6865234375e-06,
|
| 703 |
+
"loss": 0.2522,
|
| 704 |
+
"step": 2320
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 18.19,
|
| 708 |
+
"learning_rate": 8.5888671875e-06,
|
| 709 |
+
"loss": 0.2806,
|
| 710 |
+
"step": 2340
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 18.35,
|
| 714 |
+
"learning_rate": 8.491210937500001e-06,
|
| 715 |
+
"loss": 0.266,
|
| 716 |
+
"step": 2360
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 18.5,
|
| 720 |
+
"learning_rate": 8.3935546875e-06,
|
| 721 |
+
"loss": 0.288,
|
| 722 |
+
"step": 2380
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 18.66,
|
| 726 |
+
"learning_rate": 8.2958984375e-06,
|
| 727 |
+
"loss": 0.2712,
|
| 728 |
+
"step": 2400
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 18.81,
|
| 732 |
+
"learning_rate": 8.198242187500001e-06,
|
| 733 |
+
"loss": 0.3099,
|
| 734 |
+
"step": 2420
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 18.97,
|
| 738 |
+
"learning_rate": 8.1005859375e-06,
|
| 739 |
+
"loss": 0.2205,
|
| 740 |
+
"step": 2440
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 19.13,
|
| 744 |
+
"learning_rate": 8.002929687500001e-06,
|
| 745 |
+
"loss": 0.253,
|
| 746 |
+
"step": 2460
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 19.28,
|
| 750 |
+
"learning_rate": 7.905273437500001e-06,
|
| 751 |
+
"loss": 0.2885,
|
| 752 |
+
"step": 2480
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 19.44,
|
| 756 |
+
"learning_rate": 7.8076171875e-06,
|
| 757 |
+
"loss": 0.2326,
|
| 758 |
+
"step": 2500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 19.59,
|
| 762 |
+
"learning_rate": 7.709960937500001e-06,
|
| 763 |
+
"loss": 0.255,
|
| 764 |
+
"step": 2520
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 19.75,
|
| 768 |
+
"learning_rate": 7.612304687500001e-06,
|
| 769 |
+
"loss": 0.2698,
|
| 770 |
+
"step": 2540
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 19.9,
|
| 774 |
+
"learning_rate": 7.5146484375000004e-06,
|
| 775 |
+
"loss": 0.2532,
|
| 776 |
+
"step": 2560
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 20.06,
|
| 780 |
+
"learning_rate": 7.4169921875e-06,
|
| 781 |
+
"loss": 0.2562,
|
| 782 |
+
"step": 2580
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 20.21,
|
| 786 |
+
"learning_rate": 7.319335937500001e-06,
|
| 787 |
+
"loss": 0.2717,
|
| 788 |
+
"step": 2600
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 20.37,
|
| 792 |
+
"learning_rate": 7.2216796875000005e-06,
|
| 793 |
+
"loss": 0.2652,
|
| 794 |
+
"step": 2620
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 20.52,
|
| 798 |
+
"learning_rate": 7.1240234375e-06,
|
| 799 |
+
"loss": 0.2654,
|
| 800 |
+
"step": 2640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 20.68,
|
| 804 |
+
"learning_rate": 7.026367187500001e-06,
|
| 805 |
+
"loss": 0.2766,
|
| 806 |
+
"step": 2660
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 20.84,
|
| 810 |
+
"learning_rate": 6.928710937500001e-06,
|
| 811 |
+
"loss": 0.2246,
|
| 812 |
+
"step": 2680
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 20.99,
|
| 816 |
+
"learning_rate": 6.8310546875e-06,
|
| 817 |
+
"loss": 0.2546,
|
| 818 |
+
"step": 2700
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 21.15,
|
| 822 |
+
"learning_rate": 6.733398437500001e-06,
|
| 823 |
+
"loss": 0.2708,
|
| 824 |
+
"step": 2720
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 21.3,
|
| 828 |
+
"learning_rate": 6.635742187500001e-06,
|
| 829 |
+
"loss": 0.2385,
|
| 830 |
+
"step": 2740
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 21.46,
|
| 834 |
+
"learning_rate": 6.5380859375e-06,
|
| 835 |
+
"loss": 0.2271,
|
| 836 |
+
"step": 2760
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 21.61,
|
| 840 |
+
"learning_rate": 6.4404296875e-06,
|
| 841 |
+
"loss": 0.2874,
|
| 842 |
+
"step": 2780
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 21.77,
|
| 846 |
+
"learning_rate": 6.342773437500001e-06,
|
| 847 |
+
"loss": 0.2447,
|
| 848 |
+
"step": 2800
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 21.92,
|
| 852 |
+
"learning_rate": 6.2451171875000005e-06,
|
| 853 |
+
"loss": 0.2309,
|
| 854 |
+
"step": 2820
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 22.08,
|
| 858 |
+
"learning_rate": 6.1474609375e-06,
|
| 859 |
+
"loss": 0.2587,
|
| 860 |
+
"step": 2840
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 22.24,
|
| 864 |
+
"learning_rate": 6.049804687500001e-06,
|
| 865 |
+
"loss": 0.2303,
|
| 866 |
+
"step": 2860
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 22.39,
|
| 870 |
+
"learning_rate": 5.952148437500001e-06,
|
| 871 |
+
"loss": 0.2376,
|
| 872 |
+
"step": 2880
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 22.55,
|
| 876 |
+
"learning_rate": 5.8544921875e-06,
|
| 877 |
+
"loss": 0.2696,
|
| 878 |
+
"step": 2900
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 22.7,
|
| 882 |
+
"learning_rate": 5.7568359375e-06,
|
| 883 |
+
"loss": 0.2459,
|
| 884 |
+
"step": 2920
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 22.86,
|
| 888 |
+
"learning_rate": 5.659179687500001e-06,
|
| 889 |
+
"loss": 0.2032,
|
| 890 |
+
"step": 2940
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 23.01,
|
| 894 |
+
"learning_rate": 5.5615234375e-06,
|
| 895 |
+
"loss": 0.2491,
|
| 896 |
+
"step": 2960
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 23.17,
|
| 900 |
+
"learning_rate": 5.4638671875e-06,
|
| 901 |
+
"loss": 0.2395,
|
| 902 |
+
"step": 2980
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 23.32,
|
| 906 |
+
"learning_rate": 5.366210937500001e-06,
|
| 907 |
+
"loss": 0.2573,
|
| 908 |
+
"step": 3000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 23.48,
|
| 912 |
+
"learning_rate": 5.2685546875000005e-06,
|
| 913 |
+
"loss": 0.2016,
|
| 914 |
+
"step": 3020
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 23.63,
|
| 918 |
+
"learning_rate": 5.1708984375e-06,
|
| 919 |
+
"loss": 0.2486,
|
| 920 |
+
"step": 3040
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 23.79,
|
| 924 |
+
"learning_rate": 5.073242187500001e-06,
|
| 925 |
+
"loss": 0.2209,
|
| 926 |
+
"step": 3060
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 23.95,
|
| 930 |
+
"learning_rate": 4.9755859375000006e-06,
|
| 931 |
+
"loss": 0.2373,
|
| 932 |
+
"step": 3080
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 24.1,
|
| 936 |
+
"learning_rate": 4.8779296875e-06,
|
| 937 |
+
"loss": 0.2433,
|
| 938 |
+
"step": 3100
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 24.26,
|
| 942 |
+
"learning_rate": 4.7802734375e-06,
|
| 943 |
+
"loss": 0.1869,
|
| 944 |
+
"step": 3120
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 24.41,
|
| 948 |
+
"learning_rate": 4.682617187500001e-06,
|
| 949 |
+
"loss": 0.2299,
|
| 950 |
+
"step": 3140
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 24.57,
|
| 954 |
+
"learning_rate": 4.5849609375e-06,
|
| 955 |
+
"loss": 0.2556,
|
| 956 |
+
"step": 3160
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 24.72,
|
| 960 |
+
"learning_rate": 4.4873046875e-06,
|
| 961 |
+
"loss": 0.2569,
|
| 962 |
+
"step": 3180
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 24.88,
|
| 966 |
+
"learning_rate": 4.389648437500001e-06,
|
| 967 |
+
"loss": 0.2298,
|
| 968 |
+
"step": 3200
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 25.03,
|
| 972 |
+
"learning_rate": 4.2919921875000005e-06,
|
| 973 |
+
"loss": 0.2368,
|
| 974 |
+
"step": 3220
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 25.19,
|
| 978 |
+
"learning_rate": 4.1943359375e-06,
|
| 979 |
+
"loss": 0.1928,
|
| 980 |
+
"step": 3240
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 25.34,
|
| 984 |
+
"learning_rate": 4.0966796875e-06,
|
| 985 |
+
"loss": 0.2253,
|
| 986 |
+
"step": 3260
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 25.5,
|
| 990 |
+
"learning_rate": 3.9990234375000005e-06,
|
| 991 |
+
"loss": 0.2369,
|
| 992 |
+
"step": 3280
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 25.66,
|
| 996 |
+
"learning_rate": 3.9013671875e-06,
|
| 997 |
+
"loss": 0.2117,
|
| 998 |
+
"step": 3300
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 25.81,
|
| 1002 |
+
"learning_rate": 3.8037109375000004e-06,
|
| 1003 |
+
"loss": 0.2407,
|
| 1004 |
+
"step": 3320
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 25.97,
|
| 1008 |
+
"learning_rate": 3.7060546875e-06,
|
| 1009 |
+
"loss": 0.2451,
|
| 1010 |
+
"step": 3340
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 26.12,
|
| 1014 |
+
"learning_rate": 3.6083984375000004e-06,
|
| 1015 |
+
"loss": 0.235,
|
| 1016 |
+
"step": 3360
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 26.28,
|
| 1020 |
+
"learning_rate": 3.5107421875e-06,
|
| 1021 |
+
"loss": 0.2446,
|
| 1022 |
+
"step": 3380
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 26.43,
|
| 1026 |
+
"learning_rate": 3.4130859375000003e-06,
|
| 1027 |
+
"loss": 0.1815,
|
| 1028 |
+
"step": 3400
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 26.59,
|
| 1032 |
+
"learning_rate": 3.3154296875000004e-06,
|
| 1033 |
+
"loss": 0.2283,
|
| 1034 |
+
"step": 3420
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 26.74,
|
| 1038 |
+
"learning_rate": 3.2177734375e-06,
|
| 1039 |
+
"loss": 0.2136,
|
| 1040 |
+
"step": 3440
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 26.9,
|
| 1044 |
+
"learning_rate": 3.1201171875000003e-06,
|
| 1045 |
+
"loss": 0.2438,
|
| 1046 |
+
"step": 3460
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 27.06,
|
| 1050 |
+
"learning_rate": 3.0224609375e-06,
|
| 1051 |
+
"loss": 0.2097,
|
| 1052 |
+
"step": 3480
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 27.21,
|
| 1056 |
+
"learning_rate": 2.9248046875000003e-06,
|
| 1057 |
+
"loss": 0.2361,
|
| 1058 |
+
"step": 3500
|
| 1059 |
+
}
|
| 1060 |
+
],
|
| 1061 |
+
"max_steps": 4096,
|
| 1062 |
+
"num_train_epochs": 32,
|
| 1063 |
+
"total_flos": 4.5484736274432e+18,
|
| 1064 |
+
"trial_name": null,
|
| 1065 |
+
"trial_params": null
|
| 1066 |
+
}
|
checkpoint-3500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b61f1babbbf314354bf942366de8dc63aa97ebdccb27573b76080ab5b50935
|
| 3 |
+
size 3963
|
checkpoint-4000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a15a4ebf3c224f4983f72caa0e87190a30aa620985e31ed1520e454a4150d9b
|
| 3 |
+
size 33629893
|
checkpoint-4000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:720b9d77e0ce851d217c7a957a6f612f365e218c624e061d9ea510dc8315b3f7
|
| 3 |
+
size 7548185429
|
checkpoint-4000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3bd42de1b0b08a74a025ad71c27122cfef9f6672e215dce9e7f12d408f8f11f1
|
| 3 |
+
size 14575
|
checkpoint-4000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ee8f29e0fd7b704c75c284f9871facc4dbb4d208c091248f943fdf0558bb403
|
| 3 |
+
size 557
|
checkpoint-4000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3c19d74521b72228fcdac77949603cb2eeeff72403dc8a36c02e5088c9d2841
|
| 3 |
+
size 627
|
checkpoint-4000/trainer_state.json
ADDED
|
@@ -0,0 +1,1216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 31.09815354713314,
|
| 5 |
+
"global_step": 4000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.16,
|
| 12 |
+
"learning_rate": 1.9912109375000002e-05,
|
| 13 |
+
"loss": 1.9656,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.31,
|
| 18 |
+
"learning_rate": 1.98193359375e-05,
|
| 19 |
+
"loss": 1.8445,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.47,
|
| 24 |
+
"learning_rate": 1.97216796875e-05,
|
| 25 |
+
"loss": 1.59,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.62,
|
| 30 |
+
"learning_rate": 1.9624023437500002e-05,
|
| 31 |
+
"loss": 1.6275,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.78,
|
| 36 |
+
"learning_rate": 1.9526367187500002e-05,
|
| 37 |
+
"loss": 1.473,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.93,
|
| 42 |
+
"learning_rate": 1.9428710937500003e-05,
|
| 43 |
+
"loss": 1.3701,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.09,
|
| 48 |
+
"learning_rate": 1.93310546875e-05,
|
| 49 |
+
"loss": 1.3224,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.24,
|
| 54 |
+
"learning_rate": 1.92333984375e-05,
|
| 55 |
+
"loss": 1.1423,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.4,
|
| 60 |
+
"learning_rate": 1.91357421875e-05,
|
| 61 |
+
"loss": 1.1652,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.55,
|
| 66 |
+
"learning_rate": 1.9038085937500002e-05,
|
| 67 |
+
"loss": 1.1422,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.71,
|
| 72 |
+
"learning_rate": 1.8940429687500002e-05,
|
| 73 |
+
"loss": 1.002,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.87,
|
| 78 |
+
"learning_rate": 1.8842773437500003e-05,
|
| 79 |
+
"loss": 1.0779,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.02,
|
| 84 |
+
"learning_rate": 1.87451171875e-05,
|
| 85 |
+
"loss": 0.9887,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.18,
|
| 90 |
+
"learning_rate": 1.86474609375e-05,
|
| 91 |
+
"loss": 0.9543,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.33,
|
| 96 |
+
"learning_rate": 1.85498046875e-05,
|
| 97 |
+
"loss": 0.9371,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.49,
|
| 102 |
+
"learning_rate": 1.8452148437500002e-05,
|
| 103 |
+
"loss": 0.8701,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.64,
|
| 108 |
+
"learning_rate": 1.8354492187500003e-05,
|
| 109 |
+
"loss": 0.875,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.8,
|
| 114 |
+
"learning_rate": 1.82568359375e-05,
|
| 115 |
+
"loss": 0.7843,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.95,
|
| 120 |
+
"learning_rate": 1.81591796875e-05,
|
| 121 |
+
"loss": 0.7945,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.11,
|
| 126 |
+
"learning_rate": 1.80615234375e-05,
|
| 127 |
+
"loss": 0.799,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.27,
|
| 132 |
+
"learning_rate": 1.79638671875e-05,
|
| 133 |
+
"loss": 0.7623,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.42,
|
| 138 |
+
"learning_rate": 1.7866210937500002e-05,
|
| 139 |
+
"loss": 0.7263,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.58,
|
| 144 |
+
"learning_rate": 1.7768554687500003e-05,
|
| 145 |
+
"loss": 0.7779,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.73,
|
| 150 |
+
"learning_rate": 1.76708984375e-05,
|
| 151 |
+
"loss": 0.695,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.89,
|
| 156 |
+
"learning_rate": 1.75732421875e-05,
|
| 157 |
+
"loss": 0.7344,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.04,
|
| 162 |
+
"learning_rate": 1.74755859375e-05,
|
| 163 |
+
"loss": 0.695,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.2,
|
| 168 |
+
"learning_rate": 1.7377929687500002e-05,
|
| 169 |
+
"loss": 0.6504,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.35,
|
| 174 |
+
"learning_rate": 1.7280273437500002e-05,
|
| 175 |
+
"loss": 0.6447,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.51,
|
| 180 |
+
"learning_rate": 1.7182617187500003e-05,
|
| 181 |
+
"loss": 0.6931,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.66,
|
| 186 |
+
"learning_rate": 1.70849609375e-05,
|
| 187 |
+
"loss": 0.6256,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.82,
|
| 192 |
+
"learning_rate": 1.69873046875e-05,
|
| 193 |
+
"loss": 0.6132,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.98,
|
| 198 |
+
"learning_rate": 1.68896484375e-05,
|
| 199 |
+
"loss": 0.6001,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.13,
|
| 204 |
+
"learning_rate": 1.6791992187500002e-05,
|
| 205 |
+
"loss": 0.6176,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 5.29,
|
| 210 |
+
"learning_rate": 1.6694335937500002e-05,
|
| 211 |
+
"loss": 0.5709,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.44,
|
| 216 |
+
"learning_rate": 1.65966796875e-05,
|
| 217 |
+
"loss": 0.564,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.6,
|
| 222 |
+
"learning_rate": 1.64990234375e-05,
|
| 223 |
+
"loss": 0.5969,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.75,
|
| 228 |
+
"learning_rate": 1.64013671875e-05,
|
| 229 |
+
"loss": 0.5484,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.91,
|
| 234 |
+
"learning_rate": 1.63037109375e-05,
|
| 235 |
+
"loss": 0.5667,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.06,
|
| 240 |
+
"learning_rate": 1.6206054687500002e-05,
|
| 241 |
+
"loss": 0.5442,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 6.22,
|
| 246 |
+
"learning_rate": 1.6108398437500003e-05,
|
| 247 |
+
"loss": 0.4857,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 6.38,
|
| 252 |
+
"learning_rate": 1.60107421875e-05,
|
| 253 |
+
"loss": 0.5225,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.53,
|
| 258 |
+
"learning_rate": 1.59130859375e-05,
|
| 259 |
+
"loss": 0.5457,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.69,
|
| 264 |
+
"learning_rate": 1.58154296875e-05,
|
| 265 |
+
"loss": 0.5315,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.84,
|
| 270 |
+
"learning_rate": 1.57177734375e-05,
|
| 271 |
+
"loss": 0.5345,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.0,
|
| 276 |
+
"learning_rate": 1.5620117187500002e-05,
|
| 277 |
+
"loss": 0.5169,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 7.15,
|
| 282 |
+
"learning_rate": 1.5522460937500003e-05,
|
| 283 |
+
"loss": 0.5115,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 7.31,
|
| 288 |
+
"learning_rate": 1.54248046875e-05,
|
| 289 |
+
"loss": 0.4913,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 7.46,
|
| 294 |
+
"learning_rate": 1.53271484375e-05,
|
| 295 |
+
"loss": 0.4868,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.62,
|
| 300 |
+
"learning_rate": 1.5229492187500001e-05,
|
| 301 |
+
"loss": 0.5226,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.77,
|
| 306 |
+
"learning_rate": 1.5131835937500002e-05,
|
| 307 |
+
"loss": 0.4061,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.93,
|
| 312 |
+
"learning_rate": 1.50341796875e-05,
|
| 313 |
+
"loss": 0.4488,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 8.09,
|
| 318 |
+
"learning_rate": 1.4936523437500001e-05,
|
| 319 |
+
"loss": 0.4443,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 8.24,
|
| 324 |
+
"learning_rate": 1.4838867187500002e-05,
|
| 325 |
+
"loss": 0.4646,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 8.4,
|
| 330 |
+
"learning_rate": 1.47412109375e-05,
|
| 331 |
+
"loss": 0.4513,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.55,
|
| 336 |
+
"learning_rate": 1.4643554687500001e-05,
|
| 337 |
+
"loss": 0.4394,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.71,
|
| 342 |
+
"learning_rate": 1.4545898437500002e-05,
|
| 343 |
+
"loss": 0.4337,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.86,
|
| 348 |
+
"learning_rate": 1.44482421875e-05,
|
| 349 |
+
"loss": 0.5037,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 9.02,
|
| 354 |
+
"learning_rate": 1.4350585937500001e-05,
|
| 355 |
+
"loss": 0.4209,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 9.17,
|
| 360 |
+
"learning_rate": 1.42529296875e-05,
|
| 361 |
+
"loss": 0.4073,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 9.33,
|
| 366 |
+
"learning_rate": 1.4155273437500001e-05,
|
| 367 |
+
"loss": 0.3879,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.48,
|
| 372 |
+
"learning_rate": 1.4057617187500002e-05,
|
| 373 |
+
"loss": 0.4434,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.64,
|
| 378 |
+
"learning_rate": 1.39599609375e-05,
|
| 379 |
+
"loss": 0.4062,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.8,
|
| 384 |
+
"learning_rate": 1.3862304687500001e-05,
|
| 385 |
+
"loss": 0.39,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.95,
|
| 390 |
+
"learning_rate": 1.3764648437500002e-05,
|
| 391 |
+
"loss": 0.4466,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 10.11,
|
| 396 |
+
"learning_rate": 1.36669921875e-05,
|
| 397 |
+
"loss": 0.4208,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 10.26,
|
| 402 |
+
"learning_rate": 1.3569335937500001e-05,
|
| 403 |
+
"loss": 0.3662,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.42,
|
| 408 |
+
"learning_rate": 1.3471679687500002e-05,
|
| 409 |
+
"loss": 0.4049,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 10.57,
|
| 414 |
+
"learning_rate": 1.33740234375e-05,
|
| 415 |
+
"loss": 0.3989,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 10.73,
|
| 420 |
+
"learning_rate": 1.3276367187500001e-05,
|
| 421 |
+
"loss": 0.3839,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.88,
|
| 426 |
+
"learning_rate": 1.3178710937500002e-05,
|
| 427 |
+
"loss": 0.4065,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 11.04,
|
| 432 |
+
"learning_rate": 1.30810546875e-05,
|
| 433 |
+
"loss": 0.3877,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 11.2,
|
| 438 |
+
"learning_rate": 1.2983398437500001e-05,
|
| 439 |
+
"loss": 0.4048,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.35,
|
| 444 |
+
"learning_rate": 1.2885742187500002e-05,
|
| 445 |
+
"loss": 0.3715,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.51,
|
| 450 |
+
"learning_rate": 1.27880859375e-05,
|
| 451 |
+
"loss": 0.3752,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 11.66,
|
| 456 |
+
"learning_rate": 1.2690429687500001e-05,
|
| 457 |
+
"loss": 0.3401,
|
| 458 |
+
"step": 1500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 11.82,
|
| 462 |
+
"learning_rate": 1.25927734375e-05,
|
| 463 |
+
"loss": 0.3545,
|
| 464 |
+
"step": 1520
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.97,
|
| 468 |
+
"learning_rate": 1.24951171875e-05,
|
| 469 |
+
"loss": 0.3718,
|
| 470 |
+
"step": 1540
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 12.13,
|
| 474 |
+
"learning_rate": 1.2397460937500001e-05,
|
| 475 |
+
"loss": 0.3755,
|
| 476 |
+
"step": 1560
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 12.28,
|
| 480 |
+
"learning_rate": 1.22998046875e-05,
|
| 481 |
+
"loss": 0.3865,
|
| 482 |
+
"step": 1580
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 12.44,
|
| 486 |
+
"learning_rate": 1.2202148437500001e-05,
|
| 487 |
+
"loss": 0.3237,
|
| 488 |
+
"step": 1600
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 12.59,
|
| 492 |
+
"learning_rate": 1.2104492187500001e-05,
|
| 493 |
+
"loss": 0.3702,
|
| 494 |
+
"step": 1620
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 12.75,
|
| 498 |
+
"learning_rate": 1.20068359375e-05,
|
| 499 |
+
"loss": 0.3238,
|
| 500 |
+
"step": 1640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 12.91,
|
| 504 |
+
"learning_rate": 1.1909179687500001e-05,
|
| 505 |
+
"loss": 0.3373,
|
| 506 |
+
"step": 1660
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 13.06,
|
| 510 |
+
"learning_rate": 1.1811523437500002e-05,
|
| 511 |
+
"loss": 0.3486,
|
| 512 |
+
"step": 1680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 13.22,
|
| 516 |
+
"learning_rate": 1.17138671875e-05,
|
| 517 |
+
"loss": 0.362,
|
| 518 |
+
"step": 1700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 13.37,
|
| 522 |
+
"learning_rate": 1.1616210937500001e-05,
|
| 523 |
+
"loss": 0.3257,
|
| 524 |
+
"step": 1720
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 13.53,
|
| 528 |
+
"learning_rate": 1.1518554687500002e-05,
|
| 529 |
+
"loss": 0.3414,
|
| 530 |
+
"step": 1740
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 13.68,
|
| 534 |
+
"learning_rate": 1.14208984375e-05,
|
| 535 |
+
"loss": 0.3121,
|
| 536 |
+
"step": 1760
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 13.84,
|
| 540 |
+
"learning_rate": 1.1323242187500001e-05,
|
| 541 |
+
"loss": 0.3598,
|
| 542 |
+
"step": 1780
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 13.99,
|
| 546 |
+
"learning_rate": 1.1225585937500002e-05,
|
| 547 |
+
"loss": 0.2906,
|
| 548 |
+
"step": 1800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 14.15,
|
| 552 |
+
"learning_rate": 1.11279296875e-05,
|
| 553 |
+
"loss": 0.3356,
|
| 554 |
+
"step": 1820
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 14.31,
|
| 558 |
+
"learning_rate": 1.1030273437500001e-05,
|
| 559 |
+
"loss": 0.2991,
|
| 560 |
+
"step": 1840
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 14.46,
|
| 564 |
+
"learning_rate": 1.09326171875e-05,
|
| 565 |
+
"loss": 0.2987,
|
| 566 |
+
"step": 1860
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 14.62,
|
| 570 |
+
"learning_rate": 1.08349609375e-05,
|
| 571 |
+
"loss": 0.3249,
|
| 572 |
+
"step": 1880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 14.77,
|
| 576 |
+
"learning_rate": 1.0737304687500001e-05,
|
| 577 |
+
"loss": 0.3425,
|
| 578 |
+
"step": 1900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 14.93,
|
| 582 |
+
"learning_rate": 1.06396484375e-05,
|
| 583 |
+
"loss": 0.3398,
|
| 584 |
+
"step": 1920
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 15.08,
|
| 588 |
+
"learning_rate": 1.05419921875e-05,
|
| 589 |
+
"loss": 0.3018,
|
| 590 |
+
"step": 1940
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 15.24,
|
| 594 |
+
"learning_rate": 1.0444335937500001e-05,
|
| 595 |
+
"loss": 0.2475,
|
| 596 |
+
"step": 1960
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 15.39,
|
| 600 |
+
"learning_rate": 1.03466796875e-05,
|
| 601 |
+
"loss": 0.3507,
|
| 602 |
+
"step": 1980
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 15.55,
|
| 606 |
+
"learning_rate": 1.02490234375e-05,
|
| 607 |
+
"loss": 0.3084,
|
| 608 |
+
"step": 2000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 15.7,
|
| 612 |
+
"learning_rate": 1.0151367187500001e-05,
|
| 613 |
+
"loss": 0.3212,
|
| 614 |
+
"step": 2020
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 15.86,
|
| 618 |
+
"learning_rate": 1.00537109375e-05,
|
| 619 |
+
"loss": 0.2831,
|
| 620 |
+
"step": 2040
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 16.02,
|
| 624 |
+
"learning_rate": 9.956054687500001e-06,
|
| 625 |
+
"loss": 0.3072,
|
| 626 |
+
"step": 2060
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 16.17,
|
| 630 |
+
"learning_rate": 9.858398437500002e-06,
|
| 631 |
+
"loss": 0.3293,
|
| 632 |
+
"step": 2080
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 16.33,
|
| 636 |
+
"learning_rate": 9.7607421875e-06,
|
| 637 |
+
"loss": 0.2738,
|
| 638 |
+
"step": 2100
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 16.48,
|
| 642 |
+
"learning_rate": 9.663085937500001e-06,
|
| 643 |
+
"loss": 0.3245,
|
| 644 |
+
"step": 2120
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 16.64,
|
| 648 |
+
"learning_rate": 9.565429687500002e-06,
|
| 649 |
+
"loss": 0.2846,
|
| 650 |
+
"step": 2140
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 16.79,
|
| 654 |
+
"learning_rate": 9.4677734375e-06,
|
| 655 |
+
"loss": 0.2906,
|
| 656 |
+
"step": 2160
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 16.95,
|
| 660 |
+
"learning_rate": 9.370117187500001e-06,
|
| 661 |
+
"loss": 0.263,
|
| 662 |
+
"step": 2180
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 17.1,
|
| 666 |
+
"learning_rate": 9.2724609375e-06,
|
| 667 |
+
"loss": 0.286,
|
| 668 |
+
"step": 2200
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 17.26,
|
| 672 |
+
"learning_rate": 9.1748046875e-06,
|
| 673 |
+
"loss": 0.3152,
|
| 674 |
+
"step": 2220
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 17.41,
|
| 678 |
+
"learning_rate": 9.077148437500001e-06,
|
| 679 |
+
"loss": 0.2811,
|
| 680 |
+
"step": 2240
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 17.57,
|
| 684 |
+
"learning_rate": 8.9794921875e-06,
|
| 685 |
+
"loss": 0.257,
|
| 686 |
+
"step": 2260
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 17.73,
|
| 690 |
+
"learning_rate": 8.8818359375e-06,
|
| 691 |
+
"loss": 0.267,
|
| 692 |
+
"step": 2280
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 17.88,
|
| 696 |
+
"learning_rate": 8.784179687500001e-06,
|
| 697 |
+
"loss": 0.3056,
|
| 698 |
+
"step": 2300
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 18.04,
|
| 702 |
+
"learning_rate": 8.6865234375e-06,
|
| 703 |
+
"loss": 0.2522,
|
| 704 |
+
"step": 2320
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 18.19,
|
| 708 |
+
"learning_rate": 8.5888671875e-06,
|
| 709 |
+
"loss": 0.2806,
|
| 710 |
+
"step": 2340
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 18.35,
|
| 714 |
+
"learning_rate": 8.491210937500001e-06,
|
| 715 |
+
"loss": 0.266,
|
| 716 |
+
"step": 2360
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 18.5,
|
| 720 |
+
"learning_rate": 8.3935546875e-06,
|
| 721 |
+
"loss": 0.288,
|
| 722 |
+
"step": 2380
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 18.66,
|
| 726 |
+
"learning_rate": 8.2958984375e-06,
|
| 727 |
+
"loss": 0.2712,
|
| 728 |
+
"step": 2400
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 18.81,
|
| 732 |
+
"learning_rate": 8.198242187500001e-06,
|
| 733 |
+
"loss": 0.3099,
|
| 734 |
+
"step": 2420
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 18.97,
|
| 738 |
+
"learning_rate": 8.1005859375e-06,
|
| 739 |
+
"loss": 0.2205,
|
| 740 |
+
"step": 2440
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 19.13,
|
| 744 |
+
"learning_rate": 8.002929687500001e-06,
|
| 745 |
+
"loss": 0.253,
|
| 746 |
+
"step": 2460
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 19.28,
|
| 750 |
+
"learning_rate": 7.905273437500001e-06,
|
| 751 |
+
"loss": 0.2885,
|
| 752 |
+
"step": 2480
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 19.44,
|
| 756 |
+
"learning_rate": 7.8076171875e-06,
|
| 757 |
+
"loss": 0.2326,
|
| 758 |
+
"step": 2500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 19.59,
|
| 762 |
+
"learning_rate": 7.709960937500001e-06,
|
| 763 |
+
"loss": 0.255,
|
| 764 |
+
"step": 2520
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 19.75,
|
| 768 |
+
"learning_rate": 7.612304687500001e-06,
|
| 769 |
+
"loss": 0.2698,
|
| 770 |
+
"step": 2540
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 19.9,
|
| 774 |
+
"learning_rate": 7.5146484375000004e-06,
|
| 775 |
+
"loss": 0.2532,
|
| 776 |
+
"step": 2560
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 20.06,
|
| 780 |
+
"learning_rate": 7.4169921875e-06,
|
| 781 |
+
"loss": 0.2562,
|
| 782 |
+
"step": 2580
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 20.21,
|
| 786 |
+
"learning_rate": 7.319335937500001e-06,
|
| 787 |
+
"loss": 0.2717,
|
| 788 |
+
"step": 2600
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 20.37,
|
| 792 |
+
"learning_rate": 7.2216796875000005e-06,
|
| 793 |
+
"loss": 0.2652,
|
| 794 |
+
"step": 2620
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 20.52,
|
| 798 |
+
"learning_rate": 7.1240234375e-06,
|
| 799 |
+
"loss": 0.2654,
|
| 800 |
+
"step": 2640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 20.68,
|
| 804 |
+
"learning_rate": 7.026367187500001e-06,
|
| 805 |
+
"loss": 0.2766,
|
| 806 |
+
"step": 2660
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 20.84,
|
| 810 |
+
"learning_rate": 6.928710937500001e-06,
|
| 811 |
+
"loss": 0.2246,
|
| 812 |
+
"step": 2680
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 20.99,
|
| 816 |
+
"learning_rate": 6.8310546875e-06,
|
| 817 |
+
"loss": 0.2546,
|
| 818 |
+
"step": 2700
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 21.15,
|
| 822 |
+
"learning_rate": 6.733398437500001e-06,
|
| 823 |
+
"loss": 0.2708,
|
| 824 |
+
"step": 2720
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 21.3,
|
| 828 |
+
"learning_rate": 6.635742187500001e-06,
|
| 829 |
+
"loss": 0.2385,
|
| 830 |
+
"step": 2740
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 21.46,
|
| 834 |
+
"learning_rate": 6.5380859375e-06,
|
| 835 |
+
"loss": 0.2271,
|
| 836 |
+
"step": 2760
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 21.61,
|
| 840 |
+
"learning_rate": 6.4404296875e-06,
|
| 841 |
+
"loss": 0.2874,
|
| 842 |
+
"step": 2780
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 21.77,
|
| 846 |
+
"learning_rate": 6.342773437500001e-06,
|
| 847 |
+
"loss": 0.2447,
|
| 848 |
+
"step": 2800
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 21.92,
|
| 852 |
+
"learning_rate": 6.2451171875000005e-06,
|
| 853 |
+
"loss": 0.2309,
|
| 854 |
+
"step": 2820
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 22.08,
|
| 858 |
+
"learning_rate": 6.1474609375e-06,
|
| 859 |
+
"loss": 0.2587,
|
| 860 |
+
"step": 2840
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 22.24,
|
| 864 |
+
"learning_rate": 6.049804687500001e-06,
|
| 865 |
+
"loss": 0.2303,
|
| 866 |
+
"step": 2860
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 22.39,
|
| 870 |
+
"learning_rate": 5.952148437500001e-06,
|
| 871 |
+
"loss": 0.2376,
|
| 872 |
+
"step": 2880
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 22.55,
|
| 876 |
+
"learning_rate": 5.8544921875e-06,
|
| 877 |
+
"loss": 0.2696,
|
| 878 |
+
"step": 2900
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 22.7,
|
| 882 |
+
"learning_rate": 5.7568359375e-06,
|
| 883 |
+
"loss": 0.2459,
|
| 884 |
+
"step": 2920
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 22.86,
|
| 888 |
+
"learning_rate": 5.659179687500001e-06,
|
| 889 |
+
"loss": 0.2032,
|
| 890 |
+
"step": 2940
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 23.01,
|
| 894 |
+
"learning_rate": 5.5615234375e-06,
|
| 895 |
+
"loss": 0.2491,
|
| 896 |
+
"step": 2960
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 23.17,
|
| 900 |
+
"learning_rate": 5.4638671875e-06,
|
| 901 |
+
"loss": 0.2395,
|
| 902 |
+
"step": 2980
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 23.32,
|
| 906 |
+
"learning_rate": 5.366210937500001e-06,
|
| 907 |
+
"loss": 0.2573,
|
| 908 |
+
"step": 3000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 23.48,
|
| 912 |
+
"learning_rate": 5.2685546875000005e-06,
|
| 913 |
+
"loss": 0.2016,
|
| 914 |
+
"step": 3020
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 23.63,
|
| 918 |
+
"learning_rate": 5.1708984375e-06,
|
| 919 |
+
"loss": 0.2486,
|
| 920 |
+
"step": 3040
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 23.79,
|
| 924 |
+
"learning_rate": 5.073242187500001e-06,
|
| 925 |
+
"loss": 0.2209,
|
| 926 |
+
"step": 3060
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 23.95,
|
| 930 |
+
"learning_rate": 4.9755859375000006e-06,
|
| 931 |
+
"loss": 0.2373,
|
| 932 |
+
"step": 3080
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 24.1,
|
| 936 |
+
"learning_rate": 4.8779296875e-06,
|
| 937 |
+
"loss": 0.2433,
|
| 938 |
+
"step": 3100
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 24.26,
|
| 942 |
+
"learning_rate": 4.7802734375e-06,
|
| 943 |
+
"loss": 0.1869,
|
| 944 |
+
"step": 3120
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 24.41,
|
| 948 |
+
"learning_rate": 4.682617187500001e-06,
|
| 949 |
+
"loss": 0.2299,
|
| 950 |
+
"step": 3140
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 24.57,
|
| 954 |
+
"learning_rate": 4.5849609375e-06,
|
| 955 |
+
"loss": 0.2556,
|
| 956 |
+
"step": 3160
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 24.72,
|
| 960 |
+
"learning_rate": 4.4873046875e-06,
|
| 961 |
+
"loss": 0.2569,
|
| 962 |
+
"step": 3180
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 24.88,
|
| 966 |
+
"learning_rate": 4.389648437500001e-06,
|
| 967 |
+
"loss": 0.2298,
|
| 968 |
+
"step": 3200
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 25.03,
|
| 972 |
+
"learning_rate": 4.2919921875000005e-06,
|
| 973 |
+
"loss": 0.2368,
|
| 974 |
+
"step": 3220
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 25.19,
|
| 978 |
+
"learning_rate": 4.1943359375e-06,
|
| 979 |
+
"loss": 0.1928,
|
| 980 |
+
"step": 3240
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 25.34,
|
| 984 |
+
"learning_rate": 4.0966796875e-06,
|
| 985 |
+
"loss": 0.2253,
|
| 986 |
+
"step": 3260
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 25.5,
|
| 990 |
+
"learning_rate": 3.9990234375000005e-06,
|
| 991 |
+
"loss": 0.2369,
|
| 992 |
+
"step": 3280
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 25.66,
|
| 996 |
+
"learning_rate": 3.9013671875e-06,
|
| 997 |
+
"loss": 0.2117,
|
| 998 |
+
"step": 3300
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 25.81,
|
| 1002 |
+
"learning_rate": 3.8037109375000004e-06,
|
| 1003 |
+
"loss": 0.2407,
|
| 1004 |
+
"step": 3320
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 25.97,
|
| 1008 |
+
"learning_rate": 3.7060546875e-06,
|
| 1009 |
+
"loss": 0.2451,
|
| 1010 |
+
"step": 3340
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 26.12,
|
| 1014 |
+
"learning_rate": 3.6083984375000004e-06,
|
| 1015 |
+
"loss": 0.235,
|
| 1016 |
+
"step": 3360
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 26.28,
|
| 1020 |
+
"learning_rate": 3.5107421875e-06,
|
| 1021 |
+
"loss": 0.2446,
|
| 1022 |
+
"step": 3380
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 26.43,
|
| 1026 |
+
"learning_rate": 3.4130859375000003e-06,
|
| 1027 |
+
"loss": 0.1815,
|
| 1028 |
+
"step": 3400
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 26.59,
|
| 1032 |
+
"learning_rate": 3.3154296875000004e-06,
|
| 1033 |
+
"loss": 0.2283,
|
| 1034 |
+
"step": 3420
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 26.74,
|
| 1038 |
+
"learning_rate": 3.2177734375e-06,
|
| 1039 |
+
"loss": 0.2136,
|
| 1040 |
+
"step": 3440
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 26.9,
|
| 1044 |
+
"learning_rate": 3.1201171875000003e-06,
|
| 1045 |
+
"loss": 0.2438,
|
| 1046 |
+
"step": 3460
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 27.06,
|
| 1050 |
+
"learning_rate": 3.0224609375e-06,
|
| 1051 |
+
"loss": 0.2097,
|
| 1052 |
+
"step": 3480
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 27.21,
|
| 1056 |
+
"learning_rate": 2.9248046875000003e-06,
|
| 1057 |
+
"loss": 0.2361,
|
| 1058 |
+
"step": 3500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 27.37,
|
| 1062 |
+
"learning_rate": 2.8271484375000004e-06,
|
| 1063 |
+
"loss": 0.2221,
|
| 1064 |
+
"step": 3520
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 27.52,
|
| 1068 |
+
"learning_rate": 2.7294921875e-06,
|
| 1069 |
+
"loss": 0.2324,
|
| 1070 |
+
"step": 3540
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 27.68,
|
| 1074 |
+
"learning_rate": 2.6318359375000003e-06,
|
| 1075 |
+
"loss": 0.231,
|
| 1076 |
+
"step": 3560
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 27.83,
|
| 1080 |
+
"learning_rate": 2.5341796875e-06,
|
| 1081 |
+
"loss": 0.1879,
|
| 1082 |
+
"step": 3580
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 27.99,
|
| 1086 |
+
"learning_rate": 2.4365234375000002e-06,
|
| 1087 |
+
"loss": 0.2024,
|
| 1088 |
+
"step": 3600
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 28.14,
|
| 1092 |
+
"learning_rate": 2.3388671875e-06,
|
| 1093 |
+
"loss": 0.2394,
|
| 1094 |
+
"step": 3620
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 28.3,
|
| 1098 |
+
"learning_rate": 2.2412109375e-06,
|
| 1099 |
+
"loss": 0.2174,
|
| 1100 |
+
"step": 3640
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 28.45,
|
| 1104 |
+
"learning_rate": 2.1435546875000003e-06,
|
| 1105 |
+
"loss": 0.2288,
|
| 1106 |
+
"step": 3660
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 28.61,
|
| 1110 |
+
"learning_rate": 2.0458984375e-06,
|
| 1111 |
+
"loss": 0.212,
|
| 1112 |
+
"step": 3680
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 28.77,
|
| 1116 |
+
"learning_rate": 1.9482421875000002e-06,
|
| 1117 |
+
"loss": 0.2082,
|
| 1118 |
+
"step": 3700
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 28.92,
|
| 1122 |
+
"learning_rate": 1.8505859375000002e-06,
|
| 1123 |
+
"loss": 0.1976,
|
| 1124 |
+
"step": 3720
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 29.08,
|
| 1128 |
+
"learning_rate": 1.7529296875000002e-06,
|
| 1129 |
+
"loss": 0.2332,
|
| 1130 |
+
"step": 3740
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 29.23,
|
| 1134 |
+
"learning_rate": 1.6552734375000001e-06,
|
| 1135 |
+
"loss": 0.2003,
|
| 1136 |
+
"step": 3760
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 29.39,
|
| 1140 |
+
"learning_rate": 1.5576171875e-06,
|
| 1141 |
+
"loss": 0.1978,
|
| 1142 |
+
"step": 3780
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 29.54,
|
| 1146 |
+
"learning_rate": 1.4599609375e-06,
|
| 1147 |
+
"loss": 0.2194,
|
| 1148 |
+
"step": 3800
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 29.7,
|
| 1152 |
+
"learning_rate": 1.3623046875000002e-06,
|
| 1153 |
+
"loss": 0.231,
|
| 1154 |
+
"step": 3820
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 29.85,
|
| 1158 |
+
"learning_rate": 1.2646484375000001e-06,
|
| 1159 |
+
"loss": 0.2196,
|
| 1160 |
+
"step": 3840
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 30.01,
|
| 1164 |
+
"learning_rate": 1.1669921875e-06,
|
| 1165 |
+
"loss": 0.2203,
|
| 1166 |
+
"step": 3860
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 30.17,
|
| 1170 |
+
"learning_rate": 1.0693359375e-06,
|
| 1171 |
+
"loss": 0.206,
|
| 1172 |
+
"step": 3880
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 30.32,
|
| 1176 |
+
"learning_rate": 9.716796875e-07,
|
| 1177 |
+
"loss": 0.2211,
|
| 1178 |
+
"step": 3900
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 30.48,
|
| 1182 |
+
"learning_rate": 8.740234375000001e-07,
|
| 1183 |
+
"loss": 0.2129,
|
| 1184 |
+
"step": 3920
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 30.63,
|
| 1188 |
+
"learning_rate": 7.763671875e-07,
|
| 1189 |
+
"loss": 0.2085,
|
| 1190 |
+
"step": 3940
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 30.79,
|
| 1194 |
+
"learning_rate": 6.787109375000001e-07,
|
| 1195 |
+
"loss": 0.1905,
|
| 1196 |
+
"step": 3960
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 30.94,
|
| 1200 |
+
"learning_rate": 5.810546875e-07,
|
| 1201 |
+
"loss": 0.2433,
|
| 1202 |
+
"step": 3980
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 31.1,
|
| 1206 |
+
"learning_rate": 4.833984375e-07,
|
| 1207 |
+
"loss": 0.1829,
|
| 1208 |
+
"step": 4000
|
| 1209 |
+
}
|
| 1210 |
+
],
|
| 1211 |
+
"max_steps": 4096,
|
| 1212 |
+
"num_train_epochs": 32,
|
| 1213 |
+
"total_flos": 5.1982555742208e+18,
|
| 1214 |
+
"trial_name": null,
|
| 1215 |
+
"trial_params": null
|
| 1216 |
+
}
|