Upload 11 files
Browse files- README.md +1 -1
- adapter_config.json +4 -4
- adapter_model.safetensors +1 -1
- all_results.json +6 -6
- train_results.json +6 -6
- trainer_log.jsonl +202 -156
- trainer_state.json +798 -476
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -44,7 +44,7 @@ The following hyperparameters were used during training:
|
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: cosine
|
| 46 |
- lr_scheduler_warmup_ratio: 0.1
|
| 47 |
-
- num_epochs:
|
| 48 |
- mixed_precision_training: Native AMP
|
| 49 |
|
| 50 |
### Training results
|
|
|
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: cosine
|
| 46 |
- lr_scheduler_warmup_ratio: 0.1
|
| 47 |
+
- num_epochs: 65.0
|
| 48 |
- mixed_precision_training: Native AMP
|
| 49 |
|
| 50 |
### Training results
|
adapter_config.json
CHANGED
|
@@ -20,13 +20,13 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
| 23 |
"up_proj",
|
|
|
|
| 24 |
"o_proj",
|
| 25 |
-
"gate_proj",
|
| 26 |
-
"k_proj",
|
| 27 |
"down_proj",
|
| 28 |
-
"v_proj"
|
| 29 |
-
"q_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"k_proj",
|
| 24 |
+
"gate_proj",
|
| 25 |
"up_proj",
|
| 26 |
+
"q_proj",
|
| 27 |
"o_proj",
|
|
|
|
|
|
|
| 28 |
"down_proj",
|
| 29 |
+
"v_proj"
|
|
|
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 83945296
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebb031c13017ac3ec5e54ea625c9ed18656d90880c5a70962e8f756b23ffcc40
|
| 3 |
size 83945296
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 64.48,
|
| 3 |
+
"total_flos": 2.3325606118844006e+17,
|
| 4 |
+
"train_loss": 0.1495482857003993,
|
| 5 |
+
"train_runtime": 6882.5617,
|
| 6 |
+
"train_samples_per_second": 4.722,
|
| 7 |
+
"train_steps_per_second": 0.586
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 64.48,
|
| 3 |
+
"total_flos": 2.3325606118844006e+17,
|
| 4 |
+
"train_loss": 0.1495482857003993,
|
| 5 |
+
"train_runtime": 6882.5617,
|
| 6 |
+
"train_samples_per_second": 4.722,
|
| 7 |
+
"train_steps_per_second": 0.586
|
| 8 |
}
|
trainer_log.jsonl
CHANGED
|
@@ -1,156 +1,202 @@
|
|
| 1 |
-
{"current_steps": 20, "total_steps":
|
| 2 |
-
{"current_steps": 40, "total_steps":
|
| 3 |
-
{"current_steps": 60, "total_steps":
|
| 4 |
-
{"current_steps": 80, "total_steps":
|
| 5 |
-
{"current_steps": 100, "total_steps":
|
| 6 |
-
{"current_steps": 120, "total_steps":
|
| 7 |
-
{"current_steps": 140, "total_steps":
|
| 8 |
-
{"current_steps": 160, "total_steps":
|
| 9 |
-
{"current_steps": 180, "total_steps":
|
| 10 |
-
{"current_steps": 200, "total_steps":
|
| 11 |
-
{"current_steps": 220, "total_steps":
|
| 12 |
-
{"current_steps": 240, "total_steps":
|
| 13 |
-
{"current_steps": 260, "total_steps":
|
| 14 |
-
{"current_steps": 280, "total_steps":
|
| 15 |
-
{"current_steps": 300, "total_steps":
|
| 16 |
-
{"current_steps": 320, "total_steps":
|
| 17 |
-
{"current_steps": 340, "total_steps":
|
| 18 |
-
{"current_steps": 360, "total_steps":
|
| 19 |
-
{"current_steps": 380, "total_steps":
|
| 20 |
-
{"current_steps": 400, "total_steps":
|
| 21 |
-
{"current_steps": 420, "total_steps":
|
| 22 |
-
{"current_steps": 440, "total_steps":
|
| 23 |
-
{"current_steps": 460, "total_steps":
|
| 24 |
-
{"current_steps": 480, "total_steps":
|
| 25 |
-
{"current_steps": 500, "total_steps":
|
| 26 |
-
{"current_steps": 520, "total_steps":
|
| 27 |
-
{"current_steps": 540, "total_steps":
|
| 28 |
-
{"current_steps": 560, "total_steps":
|
| 29 |
-
{"current_steps": 580, "total_steps":
|
| 30 |
-
{"current_steps": 600, "total_steps":
|
| 31 |
-
{"current_steps": 620, "total_steps":
|
| 32 |
-
{"current_steps": 640, "total_steps":
|
| 33 |
-
{"current_steps": 660, "total_steps":
|
| 34 |
-
{"current_steps": 680, "total_steps":
|
| 35 |
-
{"current_steps": 700, "total_steps":
|
| 36 |
-
{"current_steps": 720, "total_steps":
|
| 37 |
-
{"current_steps": 740, "total_steps":
|
| 38 |
-
{"current_steps": 760, "total_steps":
|
| 39 |
-
{"current_steps": 780, "total_steps":
|
| 40 |
-
{"current_steps": 800, "total_steps":
|
| 41 |
-
{"current_steps": 820, "total_steps":
|
| 42 |
-
{"current_steps": 840, "total_steps":
|
| 43 |
-
{"current_steps": 860, "total_steps":
|
| 44 |
-
{"current_steps": 880, "total_steps":
|
| 45 |
-
{"current_steps": 900, "total_steps":
|
| 46 |
-
{"current_steps": 920, "total_steps":
|
| 47 |
-
{"current_steps": 940, "total_steps":
|
| 48 |
-
{"current_steps": 960, "total_steps":
|
| 49 |
-
{"current_steps": 980, "total_steps":
|
| 50 |
-
{"current_steps": 1000, "total_steps":
|
| 51 |
-
{"current_steps": 1020, "total_steps":
|
| 52 |
-
{"current_steps": 1040, "total_steps":
|
| 53 |
-
{"current_steps": 1060, "total_steps":
|
| 54 |
-
{"current_steps": 1080, "total_steps":
|
| 55 |
-
{"current_steps": 1100, "total_steps":
|
| 56 |
-
{"current_steps": 1120, "total_steps":
|
| 57 |
-
{"current_steps": 1140, "total_steps":
|
| 58 |
-
{"current_steps": 1160, "total_steps":
|
| 59 |
-
{"current_steps": 1180, "total_steps":
|
| 60 |
-
{"current_steps": 1200, "total_steps":
|
| 61 |
-
{"current_steps": 1220, "total_steps":
|
| 62 |
-
{"current_steps": 1240, "total_steps":
|
| 63 |
-
{"current_steps": 1260, "total_steps":
|
| 64 |
-
{"current_steps": 1280, "total_steps":
|
| 65 |
-
{"current_steps": 1300, "total_steps":
|
| 66 |
-
{"current_steps": 1320, "total_steps":
|
| 67 |
-
{"current_steps": 1340, "total_steps":
|
| 68 |
-
{"current_steps": 1360, "total_steps":
|
| 69 |
-
{"current_steps": 1380, "total_steps":
|
| 70 |
-
{"current_steps": 1400, "total_steps":
|
| 71 |
-
{"current_steps": 1420, "total_steps":
|
| 72 |
-
{"current_steps": 1440, "total_steps":
|
| 73 |
-
{"current_steps": 1460, "total_steps":
|
| 74 |
-
{"current_steps": 1480, "total_steps":
|
| 75 |
-
{"current_steps": 1500, "total_steps":
|
| 76 |
-
{"current_steps": 1520, "total_steps":
|
| 77 |
-
{"current_steps": 1540, "total_steps":
|
| 78 |
-
{"current_steps": 1560, "total_steps":
|
| 79 |
-
{"current_steps": 1580, "total_steps":
|
| 80 |
-
{"current_steps": 1600, "total_steps":
|
| 81 |
-
{"current_steps": 1620, "total_steps":
|
| 82 |
-
{"current_steps": 1640, "total_steps":
|
| 83 |
-
{"current_steps": 1660, "total_steps":
|
| 84 |
-
{"current_steps": 1680, "total_steps":
|
| 85 |
-
{"current_steps": 1700, "total_steps":
|
| 86 |
-
{"current_steps": 1720, "total_steps":
|
| 87 |
-
{"current_steps": 1740, "total_steps":
|
| 88 |
-
{"current_steps": 1760, "total_steps":
|
| 89 |
-
{"current_steps": 1780, "total_steps":
|
| 90 |
-
{"current_steps": 1800, "total_steps":
|
| 91 |
-
{"current_steps": 1820, "total_steps":
|
| 92 |
-
{"current_steps": 1840, "total_steps":
|
| 93 |
-
{"current_steps": 1860, "total_steps":
|
| 94 |
-
{"current_steps": 1880, "total_steps":
|
| 95 |
-
{"current_steps": 1900, "total_steps":
|
| 96 |
-
{"current_steps": 1920, "total_steps":
|
| 97 |
-
{"current_steps": 1940, "total_steps":
|
| 98 |
-
{"current_steps": 1960, "total_steps":
|
| 99 |
-
{"current_steps": 1980, "total_steps":
|
| 100 |
-
{"current_steps": 2000, "total_steps":
|
| 101 |
-
{"current_steps": 2020, "total_steps":
|
| 102 |
-
{"current_steps": 2040, "total_steps":
|
| 103 |
-
{"current_steps": 2060, "total_steps":
|
| 104 |
-
{"current_steps": 2080, "total_steps":
|
| 105 |
-
{"current_steps": 2100, "total_steps":
|
| 106 |
-
{"current_steps": 2120, "total_steps":
|
| 107 |
-
{"current_steps": 2140, "total_steps":
|
| 108 |
-
{"current_steps": 2160, "total_steps":
|
| 109 |
-
{"current_steps": 2180, "total_steps":
|
| 110 |
-
{"current_steps": 2200, "total_steps":
|
| 111 |
-
{"current_steps": 2220, "total_steps":
|
| 112 |
-
{"current_steps": 2240, "total_steps":
|
| 113 |
-
{"current_steps": 2260, "total_steps":
|
| 114 |
-
{"current_steps": 2280, "total_steps":
|
| 115 |
-
{"current_steps": 2300, "total_steps":
|
| 116 |
-
{"current_steps": 2320, "total_steps":
|
| 117 |
-
{"current_steps": 2340, "total_steps":
|
| 118 |
-
{"current_steps": 2360, "total_steps":
|
| 119 |
-
{"current_steps": 2380, "total_steps":
|
| 120 |
-
{"current_steps": 2400, "total_steps":
|
| 121 |
-
{"current_steps": 2420, "total_steps":
|
| 122 |
-
{"current_steps": 2440, "total_steps":
|
| 123 |
-
{"current_steps": 2460, "total_steps":
|
| 124 |
-
{"current_steps": 2480, "total_steps":
|
| 125 |
-
{"current_steps": 2500, "total_steps":
|
| 126 |
-
{"current_steps": 2520, "total_steps":
|
| 127 |
-
{"current_steps": 2540, "total_steps":
|
| 128 |
-
{"current_steps": 2560, "total_steps":
|
| 129 |
-
{"current_steps": 2580, "total_steps":
|
| 130 |
-
{"current_steps": 2600, "total_steps":
|
| 131 |
-
{"current_steps": 2620, "total_steps":
|
| 132 |
-
{"current_steps": 2640, "total_steps":
|
| 133 |
-
{"current_steps": 2660, "total_steps":
|
| 134 |
-
{"current_steps": 2680, "total_steps":
|
| 135 |
-
{"current_steps": 2700, "total_steps":
|
| 136 |
-
{"current_steps": 2720, "total_steps":
|
| 137 |
-
{"current_steps": 2740, "total_steps":
|
| 138 |
-
{"current_steps": 2760, "total_steps":
|
| 139 |
-
{"current_steps": 2780, "total_steps":
|
| 140 |
-
{"current_steps": 2800, "total_steps":
|
| 141 |
-
{"current_steps": 2820, "total_steps":
|
| 142 |
-
{"current_steps": 2840, "total_steps":
|
| 143 |
-
{"current_steps": 2860, "total_steps":
|
| 144 |
-
{"current_steps": 2880, "total_steps":
|
| 145 |
-
{"current_steps": 2900, "total_steps":
|
| 146 |
-
{"current_steps": 2920, "total_steps":
|
| 147 |
-
{"current_steps": 2940, "total_steps":
|
| 148 |
-
{"current_steps": 2960, "total_steps":
|
| 149 |
-
{"current_steps": 2980, "total_steps":
|
| 150 |
-
{"current_steps": 3000, "total_steps":
|
| 151 |
-
{"current_steps": 3020, "total_steps":
|
| 152 |
-
{"current_steps": 3040, "total_steps":
|
| 153 |
-
{"current_steps": 3060, "total_steps":
|
| 154 |
-
{"current_steps": 3080, "total_steps":
|
| 155 |
-
{"current_steps": 3100, "total_steps":
|
| 156 |
-
{"current_steps":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"current_steps": 20, "total_steps": 4030, "loss": 2.826, "learning_rate": 2.3573200992555833e-06, "epoch": 0.32, "percentage": 0.5, "elapsed_time": "0:00:34", "remaining_time": "1:55:11", "throughput": "0.00", "total_tokens": 0}
|
| 2 |
+
{"current_steps": 40, "total_steps": 4030, "loss": 2.72, "learning_rate": 4.838709677419355e-06, "epoch": 0.64, "percentage": 0.99, "elapsed_time": "0:01:06", "remaining_time": "1:51:00", "throughput": "0.00", "total_tokens": 0}
|
| 3 |
+
{"current_steps": 60, "total_steps": 4030, "loss": 2.4912, "learning_rate": 7.320099255583126e-06, "epoch": 0.96, "percentage": 1.49, "elapsed_time": "0:01:40", "remaining_time": "1:50:44", "throughput": "0.00", "total_tokens": 0}
|
| 4 |
+
{"current_steps": 80, "total_steps": 4030, "loss": 2.0561, "learning_rate": 9.801488833746898e-06, "epoch": 1.28, "percentage": 1.99, "elapsed_time": "0:02:15", "remaining_time": "1:51:10", "throughput": "0.00", "total_tokens": 0}
|
| 5 |
+
{"current_steps": 100, "total_steps": 4030, "loss": 1.7744, "learning_rate": 1.2282878411910669e-05, "epoch": 1.6, "percentage": 2.48, "elapsed_time": "0:02:51", "remaining_time": "1:52:01", "throughput": "0.00", "total_tokens": 0}
|
| 6 |
+
{"current_steps": 120, "total_steps": 4030, "loss": 1.8387, "learning_rate": 1.4764267990074444e-05, "epoch": 1.92, "percentage": 2.98, "elapsed_time": "0:03:25", "remaining_time": "1:51:28", "throughput": "0.00", "total_tokens": 0}
|
| 7 |
+
{"current_steps": 140, "total_steps": 4030, "loss": 1.4478, "learning_rate": 1.7245657568238215e-05, "epoch": 2.24, "percentage": 3.47, "elapsed_time": "0:03:59", "remaining_time": "1:50:52", "throughput": "0.00", "total_tokens": 0}
|
| 8 |
+
{"current_steps": 160, "total_steps": 4030, "loss": 1.285, "learning_rate": 1.9727047146401986e-05, "epoch": 2.56, "percentage": 3.97, "elapsed_time": "0:04:34", "remaining_time": "1:50:27", "throughput": "0.00", "total_tokens": 0}
|
| 9 |
+
{"current_steps": 180, "total_steps": 4030, "loss": 1.2235, "learning_rate": 2.2208436724565757e-05, "epoch": 2.88, "percentage": 4.47, "elapsed_time": "0:05:08", "remaining_time": "1:50:02", "throughput": "0.00", "total_tokens": 0}
|
| 10 |
+
{"current_steps": 200, "total_steps": 4030, "loss": 0.9575, "learning_rate": 2.468982630272953e-05, "epoch": 3.2, "percentage": 4.96, "elapsed_time": "0:05:40", "remaining_time": "1:48:48", "throughput": "0.00", "total_tokens": 0}
|
| 11 |
+
{"current_steps": 220, "total_steps": 4030, "loss": 0.7086, "learning_rate": 2.7171215880893302e-05, "epoch": 3.52, "percentage": 5.46, "elapsed_time": "0:06:15", "remaining_time": "1:48:24", "throughput": "0.00", "total_tokens": 0}
|
| 12 |
+
{"current_steps": 240, "total_steps": 4030, "loss": 0.8587, "learning_rate": 2.9652605459057077e-05, "epoch": 3.84, "percentage": 5.96, "elapsed_time": "0:06:50", "remaining_time": "1:48:03", "throughput": "0.00", "total_tokens": 0}
|
| 13 |
+
{"current_steps": 260, "total_steps": 4030, "loss": 0.5978, "learning_rate": 3.2133995037220844e-05, "epoch": 4.16, "percentage": 6.45, "elapsed_time": "0:07:25", "remaining_time": "1:47:33", "throughput": "0.00", "total_tokens": 0}
|
| 14 |
+
{"current_steps": 280, "total_steps": 4030, "loss": 0.4668, "learning_rate": 3.461538461538462e-05, "epoch": 4.48, "percentage": 6.95, "elapsed_time": "0:07:58", "remaining_time": "1:46:54", "throughput": "0.00", "total_tokens": 0}
|
| 15 |
+
{"current_steps": 300, "total_steps": 4030, "loss": 0.5667, "learning_rate": 3.7096774193548386e-05, "epoch": 4.8, "percentage": 7.44, "elapsed_time": "0:08:32", "remaining_time": "1:46:10", "throughput": "0.00", "total_tokens": 0}
|
| 16 |
+
{"current_steps": 320, "total_steps": 4030, "loss": 0.4373, "learning_rate": 3.957816377171216e-05, "epoch": 5.12, "percentage": 7.94, "elapsed_time": "0:09:06", "remaining_time": "1:45:31", "throughput": "0.00", "total_tokens": 0}
|
| 17 |
+
{"current_steps": 340, "total_steps": 4030, "loss": 0.3492, "learning_rate": 4.205955334987593e-05, "epoch": 5.44, "percentage": 8.44, "elapsed_time": "0:09:39", "remaining_time": "1:44:54", "throughput": "0.00", "total_tokens": 0}
|
| 18 |
+
{"current_steps": 360, "total_steps": 4030, "loss": 0.4018, "learning_rate": 4.45409429280397e-05, "epoch": 5.76, "percentage": 8.93, "elapsed_time": "0:10:14", "remaining_time": "1:44:19", "throughput": "0.00", "total_tokens": 0}
|
| 19 |
+
{"current_steps": 380, "total_steps": 4030, "loss": 0.279, "learning_rate": 4.702233250620348e-05, "epoch": 6.08, "percentage": 9.43, "elapsed_time": "0:10:48", "remaining_time": "1:43:45", "throughput": "0.00", "total_tokens": 0}
|
| 20 |
+
{"current_steps": 400, "total_steps": 4030, "loss": 0.2362, "learning_rate": 4.950372208436725e-05, "epoch": 6.4, "percentage": 9.93, "elapsed_time": "0:11:22", "remaining_time": "1:43:09", "throughput": "0.00", "total_tokens": 0}
|
| 21 |
+
{"current_steps": 420, "total_steps": 4030, "loss": 0.2665, "learning_rate": 4.99975992459978e-05, "epoch": 6.72, "percentage": 10.42, "elapsed_time": "0:11:55", "remaining_time": "1:42:27", "throughput": "0.00", "total_tokens": 0}
|
| 22 |
+
{"current_steps": 440, "total_steps": 4030, "loss": 0.2029, "learning_rate": 4.9987846973104825e-05, "epoch": 7.04, "percentage": 10.92, "elapsed_time": "0:12:29", "remaining_time": "1:41:58", "throughput": "0.00", "total_tokens": 0}
|
| 23 |
+
{"current_steps": 460, "total_steps": 4030, "loss": 0.1747, "learning_rate": 4.9970596058519116e-05, "epoch": 7.36, "percentage": 11.41, "elapsed_time": "0:13:05", "remaining_time": "1:41:37", "throughput": "0.00", "total_tokens": 0}
|
| 24 |
+
{"current_steps": 480, "total_steps": 4030, "loss": 0.1486, "learning_rate": 4.994585167909436e-05, "epoch": 7.68, "percentage": 11.91, "elapsed_time": "0:13:41", "remaining_time": "1:41:16", "throughput": "0.00", "total_tokens": 0}
|
| 25 |
+
{"current_steps": 500, "total_steps": 4030, "loss": 0.1866, "learning_rate": 4.9913621260409695e-05, "epoch": 8.0, "percentage": 12.41, "elapsed_time": "0:14:17", "remaining_time": "1:40:57", "throughput": "0.00", "total_tokens": 0}
|
| 26 |
+
{"current_steps": 520, "total_steps": 4030, "loss": 0.1476, "learning_rate": 4.987391447454136e-05, "epoch": 8.32, "percentage": 12.9, "elapsed_time": "0:14:54", "remaining_time": "1:40:37", "throughput": "0.00", "total_tokens": 0}
|
| 27 |
+
{"current_steps": 540, "total_steps": 4030, "loss": 0.1403, "learning_rate": 4.982674323716023e-05, "epoch": 8.64, "percentage": 13.4, "elapsed_time": "0:15:31", "remaining_time": "1:40:18", "throughput": "0.00", "total_tokens": 0}
|
| 28 |
+
{"current_steps": 560, "total_steps": 4030, "loss": 0.1018, "learning_rate": 4.977212170395598e-05, "epoch": 8.96, "percentage": 13.9, "elapsed_time": "0:16:08", "remaining_time": "1:40:00", "throughput": "0.00", "total_tokens": 0}
|
| 29 |
+
{"current_steps": 580, "total_steps": 4030, "loss": 0.0992, "learning_rate": 4.9710066266389074e-05, "epoch": 9.28, "percentage": 14.39, "elapsed_time": "0:16:45", "remaining_time": "1:39:39", "throughput": "0.00", "total_tokens": 0}
|
| 30 |
+
{"current_steps": 600, "total_steps": 4030, "loss": 0.1134, "learning_rate": 4.964059554677187e-05, "epoch": 9.6, "percentage": 14.89, "elapsed_time": "0:17:21", "remaining_time": "1:39:14", "throughput": "0.00", "total_tokens": 0}
|
| 31 |
+
{"current_steps": 620, "total_steps": 4030, "loss": 0.0781, "learning_rate": 4.956373039268022e-05, "epoch": 9.92, "percentage": 15.38, "elapsed_time": "0:17:57", "remaining_time": "1:38:45", "throughput": "0.00", "total_tokens": 0}
|
| 32 |
+
{"current_steps": 640, "total_steps": 4030, "loss": 0.0892, "learning_rate": 4.947949387069721e-05, "epoch": 10.24, "percentage": 15.88, "elapsed_time": "0:18:33", "remaining_time": "1:38:16", "throughput": "0.00", "total_tokens": 0}
|
| 33 |
+
{"current_steps": 660, "total_steps": 4030, "loss": 0.0499, "learning_rate": 4.938791125949119e-05, "epoch": 10.56, "percentage": 16.38, "elapsed_time": "0:19:06", "remaining_time": "1:37:35", "throughput": "0.00", "total_tokens": 0}
|
| 34 |
+
{"current_steps": 680, "total_steps": 4030, "loss": 0.0831, "learning_rate": 4.9289010042229765e-05, "epoch": 10.88, "percentage": 16.87, "elapsed_time": "0:19:40", "remaining_time": "1:36:56", "throughput": "0.00", "total_tokens": 0}
|
| 35 |
+
{"current_steps": 700, "total_steps": 4030, "loss": 0.0715, "learning_rate": 4.918281989833238e-05, "epoch": 11.2, "percentage": 17.37, "elapsed_time": "0:20:14", "remaining_time": "1:36:19", "throughput": "0.00", "total_tokens": 0}
|
| 36 |
+
{"current_steps": 720, "total_steps": 4030, "loss": 0.0718, "learning_rate": 4.9069372694563756e-05, "epoch": 11.52, "percentage": 17.87, "elapsed_time": "0:20:48", "remaining_time": "1:35:41", "throughput": "0.00", "total_tokens": 0}
|
| 37 |
+
{"current_steps": 740, "total_steps": 4030, "loss": 0.0849, "learning_rate": 4.8948702475470933e-05, "epoch": 11.84, "percentage": 18.36, "elapsed_time": "0:21:22", "remaining_time": "1:35:03", "throughput": "0.00", "total_tokens": 0}
|
| 38 |
+
{"current_steps": 760, "total_steps": 4030, "loss": 0.0683, "learning_rate": 4.882084545316684e-05, "epoch": 12.16, "percentage": 18.86, "elapsed_time": "0:21:56", "remaining_time": "1:34:25", "throughput": "0.00", "total_tokens": 0}
|
| 39 |
+
{"current_steps": 780, "total_steps": 4030, "loss": 0.0808, "learning_rate": 4.868583999646329e-05, "epoch": 12.48, "percentage": 19.35, "elapsed_time": "0:22:30", "remaining_time": "1:33:48", "throughput": "0.00", "total_tokens": 0}
|
| 40 |
+
{"current_steps": 800, "total_steps": 4030, "loss": 0.0607, "learning_rate": 4.8543726619356846e-05, "epoch": 12.8, "percentage": 19.85, "elapsed_time": "0:23:04", "remaining_time": "1:33:09", "throughput": "0.00", "total_tokens": 0}
|
| 41 |
+
{"current_steps": 820, "total_steps": 4030, "loss": 0.062, "learning_rate": 4.83945479688709e-05, "epoch": 13.12, "percentage": 20.35, "elapsed_time": "0:23:38", "remaining_time": "1:32:32", "throughput": "0.00", "total_tokens": 0}
|
| 42 |
+
{"current_steps": 840, "total_steps": 4030, "loss": 0.0461, "learning_rate": 4.8238348812257684e-05, "epoch": 13.44, "percentage": 20.84, "elapsed_time": "0:24:12", "remaining_time": "1:31:54", "throughput": "0.00", "total_tokens": 0}
|
| 43 |
+
{"current_steps": 860, "total_steps": 4030, "loss": 0.0482, "learning_rate": 4.808349953928184e-05, "epoch": 13.76, "percentage": 21.34, "elapsed_time": "0:24:46", "remaining_time": "1:31:17", "throughput": "0.00", "total_tokens": 0}
|
| 44 |
+
{"current_steps": 880, "total_steps": 4030, "loss": 0.0388, "learning_rate": 4.791374712344622e-05, "epoch": 14.08, "percentage": 21.84, "elapsed_time": "0:25:20", "remaining_time": "1:30:41", "throughput": "0.00", "total_tokens": 0}
|
| 45 |
+
{"current_steps": 900, "total_steps": 4030, "loss": 0.0251, "learning_rate": 4.7737118485753564e-05, "epoch": 14.4, "percentage": 22.33, "elapsed_time": "0:25:54", "remaining_time": "1:30:05", "throughput": "0.00", "total_tokens": 0}
|
| 46 |
+
{"current_steps": 920, "total_steps": 4030, "loss": 0.0515, "learning_rate": 4.75536666309653e-05, "epoch": 14.72, "percentage": 22.83, "elapsed_time": "0:26:27", "remaining_time": "1:29:27", "throughput": "0.00", "total_tokens": 0}
|
| 47 |
+
{"current_steps": 940, "total_steps": 4030, "loss": 0.0536, "learning_rate": 4.73634466114326e-05, "epoch": 15.04, "percentage": 23.33, "elapsed_time": "0:27:01", "remaining_time": "1:28:51", "throughput": "0.00", "total_tokens": 0}
|
| 48 |
+
{"current_steps": 960, "total_steps": 4030, "loss": 0.0392, "learning_rate": 4.7166515510575676e-05, "epoch": 15.36, "percentage": 23.82, "elapsed_time": "0:27:36", "remaining_time": "1:28:17", "throughput": "0.00", "total_tokens": 0}
|
| 49 |
+
{"current_steps": 980, "total_steps": 4030, "loss": 0.0369, "learning_rate": 4.696293242575356e-05, "epoch": 15.68, "percentage": 24.32, "elapsed_time": "0:28:10", "remaining_time": "1:27:41", "throughput": "0.00", "total_tokens": 0}
|
| 50 |
+
{"current_steps": 1000, "total_steps": 4030, "loss": 0.0651, "learning_rate": 4.675275845052942e-05, "epoch": 16.0, "percentage": 24.81, "elapsed_time": "0:28:45", "remaining_time": "1:27:07", "throughput": "0.00", "total_tokens": 0}
|
| 51 |
+
{"current_steps": 1020, "total_steps": 4030, "loss": 0.037, "learning_rate": 4.6536056656336947e-05, "epoch": 16.32, "percentage": 25.31, "elapsed_time": "0:29:20", "remaining_time": "1:26:34", "throughput": "0.00", "total_tokens": 0}
|
| 52 |
+
{"current_steps": 1040, "total_steps": 4030, "loss": 0.0272, "learning_rate": 4.631289207355313e-05, "epoch": 16.64, "percentage": 25.81, "elapsed_time": "0:29:53", "remaining_time": "1:25:57", "throughput": "0.00", "total_tokens": 0}
|
| 53 |
+
{"current_steps": 1060, "total_steps": 4030, "loss": 0.0507, "learning_rate": 4.6083331671983185e-05, "epoch": 16.96, "percentage": 26.3, "elapsed_time": "0:30:27", "remaining_time": "1:25:19", "throughput": "0.00", "total_tokens": 0}
|
| 54 |
+
{"current_steps": 1080, "total_steps": 4030, "loss": 0.0274, "learning_rate": 4.584744434076352e-05, "epoch": 17.28, "percentage": 26.8, "elapsed_time": "0:31:00", "remaining_time": "1:24:42", "throughput": "0.00", "total_tokens": 0}
|
| 55 |
+
{"current_steps": 1100, "total_steps": 4030, "loss": 0.0565, "learning_rate": 4.560530086768863e-05, "epoch": 17.6, "percentage": 27.3, "elapsed_time": "0:31:33", "remaining_time": "1:24:04", "throughput": "0.00", "total_tokens": 0}
|
| 56 |
+
{"current_steps": 1120, "total_steps": 4030, "loss": 0.0425, "learning_rate": 4.535697391796832e-05, "epoch": 17.92, "percentage": 27.79, "elapsed_time": "0:32:07", "remaining_time": "1:23:28", "throughput": "0.00", "total_tokens": 0}
|
| 57 |
+
{"current_steps": 1140, "total_steps": 4030, "loss": 0.0273, "learning_rate": 4.510253801242147e-05, "epoch": 18.24, "percentage": 28.29, "elapsed_time": "0:32:41", "remaining_time": "1:22:52", "throughput": "0.00", "total_tokens": 0}
|
| 58 |
+
{"current_steps": 1160, "total_steps": 4030, "loss": 0.0438, "learning_rate": 4.4842069505112984e-05, "epoch": 18.56, "percentage": 28.78, "elapsed_time": "0:33:15", "remaining_time": "1:22:17", "throughput": "0.00", "total_tokens": 0}
|
| 59 |
+
{"current_steps": 1180, "total_steps": 4030, "loss": 0.0544, "learning_rate": 4.457564656044056e-05, "epoch": 18.88, "percentage": 29.28, "elapsed_time": "0:33:49", "remaining_time": "1:21:40", "throughput": "0.00", "total_tokens": 0}
|
| 60 |
+
{"current_steps": 1200, "total_steps": 4030, "loss": 0.0283, "learning_rate": 4.430334912967824e-05, "epoch": 19.2, "percentage": 29.78, "elapsed_time": "0:34:23", "remaining_time": "1:21:05", "throughput": "0.00", "total_tokens": 0}
|
| 61 |
+
{"current_steps": 1220, "total_steps": 4030, "loss": 0.0393, "learning_rate": 4.402525892698367e-05, "epoch": 19.52, "percentage": 30.27, "elapsed_time": "0:34:57", "remaining_time": "1:20:30", "throughput": "0.00", "total_tokens": 0}
|
| 62 |
+
{"current_steps": 1240, "total_steps": 4030, "loss": 0.0249, "learning_rate": 4.374145940487641e-05, "epoch": 19.84, "percentage": 30.77, "elapsed_time": "0:35:30", "remaining_time": "1:19:53", "throughput": "0.00", "total_tokens": 0}
|
| 63 |
+
{"current_steps": 1260, "total_steps": 4030, "loss": 0.0293, "learning_rate": 4.345203572919454e-05, "epoch": 20.16, "percentage": 31.27, "elapsed_time": "0:36:04", "remaining_time": "1:19:18", "throughput": "0.00", "total_tokens": 0}
|
| 64 |
+
{"current_steps": 1280, "total_steps": 4030, "loss": 0.0287, "learning_rate": 4.315707475353706e-05, "epoch": 20.48, "percentage": 31.76, "elapsed_time": "0:36:38", "remaining_time": "1:18:43", "throughput": "0.00", "total_tokens": 0}
|
| 65 |
+
{"current_steps": 1300, "total_steps": 4030, "loss": 0.0521, "learning_rate": 4.285666499319992e-05, "epoch": 20.8, "percentage": 32.26, "elapsed_time": "0:37:12", "remaining_time": "1:18:07", "throughput": "0.00", "total_tokens": 0}
|
| 66 |
+
{"current_steps": 1320, "total_steps": 4030, "loss": 0.0285, "learning_rate": 4.25508965986133e-05, "epoch": 21.12, "percentage": 32.75, "elapsed_time": "0:37:46", "remaining_time": "1:17:32", "throughput": "0.00", "total_tokens": 0}
|
| 67 |
+
{"current_steps": 1340, "total_steps": 4030, "loss": 0.0346, "learning_rate": 4.2239861328288214e-05, "epoch": 21.44, "percentage": 33.25, "elapsed_time": "0:38:19", "remaining_time": "1:16:56", "throughput": "0.00", "total_tokens": 0}
|
| 68 |
+
{"current_steps": 1360, "total_steps": 4030, "loss": 0.022, "learning_rate": 4.1923652521280585e-05, "epoch": 21.76, "percentage": 33.75, "elapsed_time": "0:38:54", "remaining_time": "1:16:22", "throughput": "0.00", "total_tokens": 0}
|
| 69 |
+
{"current_steps": 1380, "total_steps": 4030, "loss": 0.0482, "learning_rate": 4.160236506918098e-05, "epoch": 22.08, "percentage": 34.24, "elapsed_time": "0:39:28", "remaining_time": "1:15:48", "throughput": "0.00", "total_tokens": 0}
|
| 70 |
+
{"current_steps": 1400, "total_steps": 4030, "loss": 0.019, "learning_rate": 4.127609538763842e-05, "epoch": 22.4, "percentage": 34.74, "elapsed_time": "0:40:02", "remaining_time": "1:15:12", "throughput": "0.00", "total_tokens": 0}
|
| 71 |
+
{"current_steps": 1420, "total_steps": 4030, "loss": 0.0312, "learning_rate": 4.094494138742685e-05, "epoch": 22.72, "percentage": 35.24, "elapsed_time": "0:40:36", "remaining_time": "1:14:37", "throughput": "0.00", "total_tokens": 0}
|
| 72 |
+
{"current_steps": 1440, "total_steps": 4030, "loss": 0.0377, "learning_rate": 4.0609002445063036e-05, "epoch": 23.04, "percentage": 35.73, "elapsed_time": "0:41:09", "remaining_time": "1:14:02", "throughput": "0.00", "total_tokens": 0}
|
| 73 |
+
{"current_steps": 1460, "total_steps": 4030, "loss": 0.0307, "learning_rate": 4.02683793729844e-05, "epoch": 23.36, "percentage": 36.23, "elapsed_time": "0:41:43", "remaining_time": "1:13:26", "throughput": "0.00", "total_tokens": 0}
|
| 74 |
+
{"current_steps": 1480, "total_steps": 4030, "loss": 0.0419, "learning_rate": 3.9923174389296085e-05, "epoch": 23.68, "percentage": 36.72, "elapsed_time": "0:42:16", "remaining_time": "1:12:51", "throughput": "0.00", "total_tokens": 0}
|
| 75 |
+
{"current_steps": 1500, "total_steps": 4030, "loss": 0.0223, "learning_rate": 3.957349108709623e-05, "epoch": 24.0, "percentage": 37.22, "elapsed_time": "0:42:50", "remaining_time": "1:12:16", "throughput": "0.00", "total_tokens": 0}
|
| 76 |
+
{"current_steps": 1520, "total_steps": 4030, "loss": 0.0209, "learning_rate": 3.921943440338849e-05, "epoch": 24.32, "percentage": 37.72, "elapsed_time": "0:43:24", "remaining_time": "1:11:40", "throughput": "0.00", "total_tokens": 0}
|
| 77 |
+
{"current_steps": 1540, "total_steps": 4030, "loss": 0.0491, "learning_rate": 3.886111058759132e-05, "epoch": 24.64, "percentage": 38.21, "elapsed_time": "0:43:58", "remaining_time": "1:11:06", "throughput": "0.00", "total_tokens": 0}
|
| 78 |
+
{"current_steps": 1560, "total_steps": 4030, "loss": 0.0298, "learning_rate": 3.849862716965352e-05, "epoch": 24.96, "percentage": 38.71, "elapsed_time": "0:44:32", "remaining_time": "1:10:31", "throughput": "0.00", "total_tokens": 0}
|
| 79 |
+
{"current_steps": 1580, "total_steps": 4030, "loss": 0.0319, "learning_rate": 3.813209292778527e-05, "epoch": 25.28, "percentage": 39.21, "elapsed_time": "0:45:06", "remaining_time": "1:09:56", "throughput": "0.00", "total_tokens": 0}
|
| 80 |
+
{"current_steps": 1600, "total_steps": 4030, "loss": 0.0302, "learning_rate": 3.776161785581481e-05, "epoch": 25.6, "percentage": 39.7, "elapsed_time": "0:45:40", "remaining_time": "1:09:22", "throughput": "0.00", "total_tokens": 0}
|
| 81 |
+
{"current_steps": 1620, "total_steps": 4030, "loss": 0.04, "learning_rate": 3.738731313018019e-05, "epoch": 25.92, "percentage": 40.2, "elapsed_time": "0:46:14", "remaining_time": "1:08:47", "throughput": "0.00", "total_tokens": 0}
|
| 82 |
+
{"current_steps": 1640, "total_steps": 4030, "loss": 0.0354, "learning_rate": 3.700929107656614e-05, "epoch": 26.24, "percentage": 40.69, "elapsed_time": "0:46:48", "remaining_time": "1:08:12", "throughput": "0.00", "total_tokens": 0}
|
| 83 |
+
{"current_steps": 1660, "total_steps": 4030, "loss": 0.0186, "learning_rate": 3.662766513619611e-05, "epoch": 26.56, "percentage": 41.19, "elapsed_time": "0:47:22", "remaining_time": "1:07:38", "throughput": "0.00", "total_tokens": 0}
|
| 84 |
+
{"current_steps": 1680, "total_steps": 4030, "loss": 0.022, "learning_rate": 3.62425498317895e-05, "epoch": 26.88, "percentage": 41.69, "elapsed_time": "0:47:57", "remaining_time": "1:07:04", "throughput": "0.00", "total_tokens": 0}
|
| 85 |
+
{"current_steps": 1700, "total_steps": 4030, "loss": 0.015, "learning_rate": 3.585406073319439e-05, "epoch": 27.2, "percentage": 42.18, "elapsed_time": "0:48:31", "remaining_time": "1:06:30", "throughput": "0.00", "total_tokens": 0}
|
| 86 |
+
{"current_steps": 1720, "total_steps": 4030, "loss": 0.0381, "learning_rate": 3.546231442270596e-05, "epoch": 27.52, "percentage": 42.68, "elapsed_time": "0:49:05", "remaining_time": "1:05:55", "throughput": "0.00", "total_tokens": 0}
|
| 87 |
+
{"current_steps": 1740, "total_steps": 4030, "loss": 0.0277, "learning_rate": 3.506742846008116e-05, "epoch": 27.84, "percentage": 43.18, "elapsed_time": "0:49:39", "remaining_time": "1:05:20", "throughput": "0.00", "total_tokens": 0}
|
| 88 |
+
{"current_steps": 1760, "total_steps": 4030, "loss": 0.0423, "learning_rate": 3.4669521347259996e-05, "epoch": 28.16, "percentage": 43.67, "elapsed_time": "0:50:12", "remaining_time": "1:04:45", "throughput": "0.00", "total_tokens": 0}
|
| 89 |
+
{"current_steps": 1780, "total_steps": 4030, "loss": 0.0115, "learning_rate": 3.426871249280414e-05, "epoch": 28.48, "percentage": 44.17, "elapsed_time": "0:50:46", "remaining_time": "1:04:10", "throughput": "0.00", "total_tokens": 0}
|
| 90 |
+
{"current_steps": 1800, "total_steps": 4030, "loss": 0.0275, "learning_rate": 3.386512217606339e-05, "epoch": 28.8, "percentage": 44.67, "elapsed_time": "0:51:19", "remaining_time": "1:03:35", "throughput": "0.00", "total_tokens": 0}
|
| 91 |
+
{"current_steps": 1820, "total_steps": 4030, "loss": 0.0309, "learning_rate": 3.345887151108087e-05, "epoch": 29.12, "percentage": 45.16, "elapsed_time": "0:51:53", "remaining_time": "1:03:00", "throughput": "0.00", "total_tokens": 0}
|
| 92 |
+
{"current_steps": 1840, "total_steps": 4030, "loss": 0.0294, "learning_rate": 3.305008241024774e-05, "epoch": 29.44, "percentage": 45.66, "elapsed_time": "0:52:26", "remaining_time": "1:02:25", "throughput": "0.00", "total_tokens": 0}
|
| 93 |
+
{"current_steps": 1860, "total_steps": 4030, "loss": 0.0213, "learning_rate": 3.2638877547718264e-05, "epoch": 29.76, "percentage": 46.15, "elapsed_time": "0:53:00", "remaining_time": "1:01:50", "throughput": "0.00", "total_tokens": 0}
|
| 94 |
+
{"current_steps": 1880, "total_steps": 4030, "loss": 0.0326, "learning_rate": 3.222538032259643e-05, "epoch": 30.08, "percentage": 46.65, "elapsed_time": "0:53:34", "remaining_time": "1:01:16", "throughput": "0.00", "total_tokens": 0}
|
| 95 |
+
{"current_steps": 1900, "total_steps": 4030, "loss": 0.0249, "learning_rate": 3.1809714821904834e-05, "epoch": 30.4, "percentage": 47.15, "elapsed_time": "0:54:08", "remaining_time": "1:00:41", "throughput": "0.00", "total_tokens": 0}
|
| 96 |
+
{"current_steps": 1920, "total_steps": 4030, "loss": 0.0115, "learning_rate": 3.1392005783347244e-05, "epoch": 30.72, "percentage": 47.64, "elapsed_time": "0:54:42", "remaining_time": "1:00:07", "throughput": "0.00", "total_tokens": 0}
|
| 97 |
+
{"current_steps": 1940, "total_steps": 4030, "loss": 0.0322, "learning_rate": 3.0972378557875884e-05, "epoch": 31.04, "percentage": 48.14, "elapsed_time": "0:55:16", "remaining_time": "0:59:32", "throughput": "0.00", "total_tokens": 0}
|
| 98 |
+
{"current_steps": 1960, "total_steps": 4030, "loss": 0.0316, "learning_rate": 3.055095907207465e-05, "epoch": 31.36, "percentage": 48.64, "elapsed_time": "0:55:50", "remaining_time": "0:58:58", "throughput": "0.00", "total_tokens": 0}
|
| 99 |
+
{"current_steps": 1980, "total_steps": 4030, "loss": 0.0248, "learning_rate": 3.0127873790369627e-05, "epoch": 31.68, "percentage": 49.13, "elapsed_time": "0:56:24", "remaining_time": "0:58:24", "throughput": "0.00", "total_tokens": 0}
|
| 100 |
+
{"current_steps": 2000, "total_steps": 4030, "loss": 0.0234, "learning_rate": 2.9703249677078156e-05, "epoch": 32.0, "percentage": 49.63, "elapsed_time": "0:56:58", "remaining_time": "0:57:49", "throughput": "0.00", "total_tokens": 0}
|
| 101 |
+
{"current_steps": 2020, "total_steps": 4030, "loss": 0.0277, "learning_rate": 2.9277214158307937e-05, "epoch": 32.32, "percentage": 50.12, "elapsed_time": "0:57:33", "remaining_time": "0:57:16", "throughput": "0.00", "total_tokens": 0}
|
| 102 |
+
{"current_steps": 2040, "total_steps": 4030, "loss": 0.0162, "learning_rate": 2.8849895083717537e-05, "epoch": 32.64, "percentage": 50.62, "elapsed_time": "0:58:07", "remaining_time": "0:56:41", "throughput": "0.00", "total_tokens": 0}
|
| 103 |
+
{"current_steps": 2060, "total_steps": 4030, "loss": 0.022, "learning_rate": 2.842142068814977e-05, "epoch": 32.96, "percentage": 51.12, "elapsed_time": "0:58:41", "remaining_time": "0:56:08", "throughput": "0.00", "total_tokens": 0}
|
| 104 |
+
{"current_steps": 2080, "total_steps": 4030, "loss": 0.0278, "learning_rate": 2.7991919553149497e-05, "epoch": 33.28, "percentage": 51.61, "elapsed_time": "0:59:15", "remaining_time": "0:55:33", "throughput": "0.00", "total_tokens": 0}
|
| 105 |
+
{"current_steps": 2100, "total_steps": 4030, "loss": 0.0189, "learning_rate": 2.756152056837743e-05, "epoch": 33.6, "percentage": 52.11, "elapsed_time": "0:59:49", "remaining_time": "0:54:59", "throughput": "0.00", "total_tokens": 0}
|
| 106 |
+
{"current_steps": 2120, "total_steps": 4030, "loss": 0.0228, "learning_rate": 2.7130352892931388e-05, "epoch": 33.92, "percentage": 52.61, "elapsed_time": "1:00:23", "remaining_time": "0:54:24", "throughput": "0.00", "total_tokens": 0}
|
| 107 |
+
{"current_steps": 2140, "total_steps": 4030, "loss": 0.0319, "learning_rate": 2.669854591658679e-05, "epoch": 34.24, "percentage": 53.1, "elapsed_time": "1:00:56", "remaining_time": "0:53:49", "throughput": "0.00", "total_tokens": 0}
|
| 108 |
+
{"current_steps": 2160, "total_steps": 4030, "loss": 0.0153, "learning_rate": 2.6266229220967818e-05, "epoch": 34.56, "percentage": 53.6, "elapsed_time": "1:01:29", "remaining_time": "0:53:14", "throughput": "0.00", "total_tokens": 0}
|
| 109 |
+
{"current_steps": 2180, "total_steps": 4030, "loss": 0.0267, "learning_rate": 2.5833532540661127e-05, "epoch": 34.88, "percentage": 54.09, "elapsed_time": "1:02:03", "remaining_time": "0:52:39", "throughput": "0.00", "total_tokens": 0}
|
| 110 |
+
{"current_steps": 2200, "total_steps": 4030, "loss": 0.0178, "learning_rate": 2.540058572428356e-05, "epoch": 35.2, "percentage": 54.59, "elapsed_time": "1:02:37", "remaining_time": "0:52:05", "throughput": "0.00", "total_tokens": 0}
|
| 111 |
+
{"current_steps": 2220, "total_steps": 4030, "loss": 0.0217, "learning_rate": 2.496751869551567e-05, "epoch": 35.52, "percentage": 55.09, "elapsed_time": "1:03:10", "remaining_time": "0:51:30", "throughput": "0.00", "total_tokens": 0}
|
| 112 |
+
{"current_steps": 2240, "total_steps": 4030, "loss": 0.017, "learning_rate": 2.453446141411273e-05, "epoch": 35.84, "percentage": 55.58, "elapsed_time": "1:03:44", "remaining_time": "0:50:56", "throughput": "0.00", "total_tokens": 0}
|
| 113 |
+
{"current_steps": 2260, "total_steps": 4030, "loss": 0.0257, "learning_rate": 2.4101543836904938e-05, "epoch": 36.16, "percentage": 56.08, "elapsed_time": "1:04:18", "remaining_time": "0:50:21", "throughput": "0.00", "total_tokens": 0}
|
| 114 |
+
{"current_steps": 2280, "total_steps": 4030, "loss": 0.0237, "learning_rate": 2.3668895878798424e-05, "epoch": 36.48, "percentage": 56.58, "elapsed_time": "1:04:52", "remaining_time": "0:49:47", "throughput": "0.00", "total_tokens": 0}
|
| 115 |
+
{"current_steps": 2300, "total_steps": 4030, "loss": 0.024, "learning_rate": 2.32366473737889e-05, "epoch": 36.8, "percentage": 57.07, "elapsed_time": "1:05:25", "remaining_time": "0:49:12", "throughput": "0.00", "total_tokens": 0}
|
| 116 |
+
{"current_steps": 2320, "total_steps": 4030, "loss": 0.0225, "learning_rate": 2.2804928035999594e-05, "epoch": 37.12, "percentage": 57.57, "elapsed_time": "1:06:00", "remaining_time": "0:48:39", "throughput": "0.00", "total_tokens": 0}
|
| 117 |
+
{"current_steps": 2340, "total_steps": 4030, "loss": 0.0239, "learning_rate": 2.23738674207551e-05, "epoch": 37.44, "percentage": 58.06, "elapsed_time": "1:06:35", "remaining_time": "0:48:05", "throughput": "0.00", "total_tokens": 0}
|
| 118 |
+
{"current_steps": 2360, "total_steps": 4030, "loss": 0.0235, "learning_rate": 2.1943594885702984e-05, "epoch": 37.76, "percentage": 58.56, "elapsed_time": "1:07:09", "remaining_time": "0:47:31", "throughput": "0.00", "total_tokens": 0}
|
| 119 |
+
{"current_steps": 2380, "total_steps": 4030, "loss": 0.0286, "learning_rate": 2.151423955199456e-05, "epoch": 38.08, "percentage": 59.06, "elapsed_time": "1:07:43", "remaining_time": "0:46:57", "throughput": "0.00", "total_tokens": 0}
|
| 120 |
+
{"current_steps": 2400, "total_steps": 4030, "loss": 0.0323, "learning_rate": 2.108593026553681e-05, "epoch": 38.4, "percentage": 59.55, "elapsed_time": "1:08:18", "remaining_time": "0:46:23", "throughput": "0.00", "total_tokens": 0}
|
| 121 |
+
{"current_steps": 2420, "total_steps": 4030, "loss": 0.0241, "learning_rate": 2.0658795558326743e-05, "epoch": 38.72, "percentage": 60.05, "elapsed_time": "1:08:52", "remaining_time": "0:45:48", "throughput": "0.00", "total_tokens": 0}
|
| 122 |
+
{"current_steps": 2440, "total_steps": 4030, "loss": 0.0158, "learning_rate": 2.0232963609880093e-05, "epoch": 39.04, "percentage": 60.55, "elapsed_time": "1:09:26", "remaining_time": "0:45:14", "throughput": "0.00", "total_tokens": 0}
|
| 123 |
+
{"current_steps": 2460, "total_steps": 4030, "loss": 0.0241, "learning_rate": 1.9808562208765667e-05, "epoch": 39.36, "percentage": 61.04, "elapsed_time": "1:10:00", "remaining_time": "0:44:40", "throughput": "0.00", "total_tokens": 0}
|
| 124 |
+
{"current_steps": 2480, "total_steps": 4030, "loss": 0.0174, "learning_rate": 1.938571871425715e-05, "epoch": 39.68, "percentage": 61.54, "elapsed_time": "1:10:34", "remaining_time": "0:44:06", "throughput": "0.00", "total_tokens": 0}
|
| 125 |
+
{"current_steps": 2500, "total_steps": 4030, "loss": 0.0183, "learning_rate": 1.896456001811357e-05, "epoch": 40.0, "percentage": 62.03, "elapsed_time": "1:11:08", "remaining_time": "0:43:32", "throughput": "0.00", "total_tokens": 0}
|
| 126 |
+
{"current_steps": 2520, "total_steps": 4030, "loss": 0.012, "learning_rate": 1.854521250650026e-05, "epoch": 40.32, "percentage": 62.53, "elapsed_time": "1:11:43", "remaining_time": "0:42:58", "throughput": "0.00", "total_tokens": 0}
|
| 127 |
+
{"current_steps": 2540, "total_steps": 4030, "loss": 0.0225, "learning_rate": 1.8127802022061334e-05, "epoch": 40.64, "percentage": 63.03, "elapsed_time": "1:12:16", "remaining_time": "0:42:24", "throughput": "0.00", "total_tokens": 0}
|
| 128 |
+
{"current_steps": 2560, "total_steps": 4030, "loss": 0.0391, "learning_rate": 1.7712453826155457e-05, "epoch": 40.96, "percentage": 63.52, "elapsed_time": "1:12:50", "remaining_time": "0:41:49", "throughput": "0.00", "total_tokens": 0}
|
| 129 |
+
{"current_steps": 2580, "total_steps": 4030, "loss": 0.0229, "learning_rate": 1.72992925612659e-05, "epoch": 41.28, "percentage": 64.02, "elapsed_time": "1:13:24", "remaining_time": "0:41:15", "throughput": "0.00", "total_tokens": 0}
|
| 130 |
+
{"current_steps": 2600, "total_steps": 4030, "loss": 0.015, "learning_rate": 1.688844221359645e-05, "epoch": 41.6, "percentage": 64.52, "elapsed_time": "1:13:59", "remaining_time": "0:40:41", "throughput": "0.00", "total_tokens": 0}
|
| 131 |
+
{"current_steps": 2620, "total_steps": 4030, "loss": 0.0287, "learning_rate": 1.6480026075864163e-05, "epoch": 41.92, "percentage": 65.01, "elapsed_time": "1:14:33", "remaining_time": "0:40:07", "throughput": "0.00", "total_tokens": 0}
|
| 132 |
+
{"current_steps": 2640, "total_steps": 4030, "loss": 0.0229, "learning_rate": 1.6074166710300247e-05, "epoch": 42.24, "percentage": 65.51, "elapsed_time": "1:15:08", "remaining_time": "0:39:33", "throughput": "0.00", "total_tokens": 0}
|
| 133 |
+
{"current_steps": 2660, "total_steps": 4030, "loss": 0.0352, "learning_rate": 1.567098591187021e-05, "epoch": 42.56, "percentage": 66.0, "elapsed_time": "1:15:42", "remaining_time": "0:38:59", "throughput": "0.00", "total_tokens": 0}
|
| 134 |
+
{"current_steps": 2680, "total_steps": 4030, "loss": 0.0242, "learning_rate": 1.5270604671724188e-05, "epoch": 42.88, "percentage": 66.5, "elapsed_time": "1:16:16", "remaining_time": "0:38:25", "throughput": "0.00", "total_tokens": 0}
|
| 135 |
+
{"current_steps": 2700, "total_steps": 4030, "loss": 0.0165, "learning_rate": 1.4873143140888538e-05, "epoch": 43.2, "percentage": 67.0, "elapsed_time": "1:16:50", "remaining_time": "0:37:51", "throughput": "0.00", "total_tokens": 0}
|
| 136 |
+
{"current_steps": 2720, "total_steps": 4030, "loss": 0.0274, "learning_rate": 1.4478720594209532e-05, "epoch": 43.52, "percentage": 67.49, "elapsed_time": "1:17:24", "remaining_time": "0:37:16", "throughput": "0.00", "total_tokens": 0}
|
| 137 |
+
{"current_steps": 2740, "total_steps": 4030, "loss": 0.0185, "learning_rate": 1.4087455394559984e-05, "epoch": 43.84, "percentage": 67.99, "elapsed_time": "1:17:58", "remaining_time": "0:36:42", "throughput": "0.00", "total_tokens": 0}
|
| 138 |
+
{"current_steps": 2760, "total_steps": 4030, "loss": 0.0509, "learning_rate": 1.369946495731954e-05, "epoch": 44.16, "percentage": 68.49, "elapsed_time": "1:18:31", "remaining_time": "0:36:08", "throughput": "0.00", "total_tokens": 0}
|
| 139 |
+
{"current_steps": 2780, "total_steps": 4030, "loss": 0.027, "learning_rate": 1.3314865715139346e-05, "epoch": 44.48, "percentage": 68.98, "elapsed_time": "1:19:05", "remaining_time": "0:35:33", "throughput": "0.00", "total_tokens": 0}
|
| 140 |
+
{"current_steps": 2800, "total_steps": 4030, "loss": 0.0163, "learning_rate": 1.2933773083001517e-05, "epoch": 44.8, "percentage": 69.48, "elapsed_time": "1:19:39", "remaining_time": "0:34:59", "throughput": "0.00", "total_tokens": 0}
|
| 141 |
+
{"current_steps": 2820, "total_steps": 4030, "loss": 0.0125, "learning_rate": 1.255630142358421e-05, "epoch": 45.12, "percentage": 69.98, "elapsed_time": "1:20:14", "remaining_time": "0:34:25", "throughput": "0.00", "total_tokens": 0}
|
| 142 |
+
{"current_steps": 2840, "total_steps": 4030, "loss": 0.0327, "learning_rate": 1.2182564012942193e-05, "epoch": 45.44, "percentage": 70.47, "elapsed_time": "1:20:49", "remaining_time": "0:33:51", "throughput": "0.00", "total_tokens": 0}
|
| 143 |
+
{"current_steps": 2860, "total_steps": 4030, "loss": 0.0302, "learning_rate": 1.1812673006513789e-05, "epoch": 45.76, "percentage": 70.97, "elapsed_time": "1:21:23", "remaining_time": "0:33:17", "throughput": "0.00", "total_tokens": 0}
|
| 144 |
+
{"current_steps": 2880, "total_steps": 4030, "loss": 0.0209, "learning_rate": 1.14467394054639e-05, "epoch": 46.08, "percentage": 71.46, "elapsed_time": "1:21:57", "remaining_time": "0:32:43", "throughput": "0.00", "total_tokens": 0}
|
| 145 |
+
{"current_steps": 2900, "total_steps": 4030, "loss": 0.025, "learning_rate": 1.108487302337353e-05, "epoch": 46.4, "percentage": 71.96, "elapsed_time": "1:22:32", "remaining_time": "0:32:09", "throughput": "0.00", "total_tokens": 0}
|
| 146 |
+
{"current_steps": 2920, "total_steps": 4030, "loss": 0.0284, "learning_rate": 1.0727182453285647e-05, "epoch": 46.72, "percentage": 72.46, "elapsed_time": "1:23:06", "remaining_time": "0:31:35", "throughput": "0.00", "total_tokens": 0}
|
| 147 |
+
{"current_steps": 2940, "total_steps": 4030, "loss": 0.0174, "learning_rate": 1.0373775035117305e-05, "epoch": 47.04, "percentage": 72.95, "elapsed_time": "1:23:40", "remaining_time": "0:31:01", "throughput": "0.00", "total_tokens": 0}
|
| 148 |
+
{"current_steps": 2960, "total_steps": 4030, "loss": 0.0115, "learning_rate": 1.002475682344792e-05, "epoch": 47.36, "percentage": 73.45, "elapsed_time": "1:24:15", "remaining_time": "0:30:27", "throughput": "0.00", "total_tokens": 0}
|
| 149 |
+
{"current_steps": 2980, "total_steps": 4030, "loss": 0.0238, "learning_rate": 9.680232555693067e-06, "epoch": 47.68, "percentage": 73.95, "elapsed_time": "1:24:49", "remaining_time": "0:29:53", "throughput": "0.00", "total_tokens": 0}
|
| 150 |
+
{"current_steps": 3000, "total_steps": 4030, "loss": 0.0294, "learning_rate": 9.340305620673778e-06, "epoch": 48.0, "percentage": 74.44, "elapsed_time": "1:25:23", "remaining_time": "0:29:19", "throughput": "0.00", "total_tokens": 0}
|
| 151 |
+
{"current_steps": 3020, "total_steps": 4030, "loss": 0.0226, "learning_rate": 9.005078027590375e-06, "epoch": 48.32, "percentage": 74.94, "elapsed_time": "1:25:58", "remaining_time": "0:28:45", "throughput": "0.00", "total_tokens": 0}
|
| 152 |
+
{"current_steps": 3040, "total_steps": 4030, "loss": 0.0196, "learning_rate": 8.67465037541038e-06, "epoch": 48.64, "percentage": 75.43, "elapsed_time": "1:26:32", "remaining_time": "0:28:11", "throughput": "0.00", "total_tokens": 0}
|
| 153 |
+
{"current_steps": 3060, "total_steps": 4030, "loss": 0.0175, "learning_rate": 8.34912182267959e-06, "epoch": 48.96, "percentage": 75.93, "elapsed_time": "1:27:07", "remaining_time": "0:27:36", "throughput": "0.00", "total_tokens": 0}
|
| 154 |
+
{"current_steps": 3080, "total_steps": 4030, "loss": 0.015, "learning_rate": 8.028590057765523e-06, "epoch": 49.28, "percentage": 76.43, "elapsed_time": "1:27:41", "remaining_time": "0:27:02", "throughput": "0.00", "total_tokens": 0}
|
| 155 |
+
{"current_steps": 3100, "total_steps": 4030, "loss": 0.0221, "learning_rate": 7.713151269541844e-06, "epoch": 49.6, "percentage": 76.92, "elapsed_time": "1:28:15", "remaining_time": "0:26:28", "throughput": "0.00", "total_tokens": 0}
|
| 156 |
+
{"current_steps": 3120, "total_steps": 4030, "loss": 0.0161, "learning_rate": 7.402900118522979e-06, "epoch": 49.92, "percentage": 77.42, "elapsed_time": "1:28:48", "remaining_time": "0:25:54", "throughput": "0.00", "total_tokens": 0}
|
| 157 |
+
{"current_steps": 3140, "total_steps": 4030, "loss": 0.0237, "learning_rate": 7.097929708457282e-06, "epoch": 50.24, "percentage": 77.92, "elapsed_time": "1:29:23", "remaining_time": "0:25:20", "throughput": "0.00", "total_tokens": 0}
|
| 158 |
+
{"current_steps": 3160, "total_steps": 4030, "loss": 0.0172, "learning_rate": 6.7983315583873695e-06, "epoch": 50.56, "percentage": 78.41, "elapsed_time": "1:29:56", "remaining_time": "0:24:45", "throughput": "0.00", "total_tokens": 0}
|
| 159 |
+
{"current_steps": 3180, "total_steps": 4030, "loss": 0.0198, "learning_rate": 6.504195575186009e-06, "epoch": 50.88, "percentage": 78.91, "elapsed_time": "1:30:30", "remaining_time": "0:24:11", "throughput": "0.00", "total_tokens": 0}
|
| 160 |
+
{"current_steps": 3200, "total_steps": 4030, "loss": 0.0227, "learning_rate": 6.215610026575916e-06, "epoch": 51.2, "percentage": 79.4, "elapsed_time": "1:31:04", "remaining_time": "0:23:37", "throughput": "0.00", "total_tokens": 0}
|
| 161 |
+
{"current_steps": 3220, "total_steps": 4030, "loss": 0.0156, "learning_rate": 5.93266151464123e-06, "epoch": 51.52, "percentage": 79.9, "elapsed_time": "1:31:37", "remaining_time": "0:23:03", "throughput": "0.00", "total_tokens": 0}
|
| 162 |
+
{"current_steps": 3240, "total_steps": 4030, "loss": 0.0268, "learning_rate": 5.655434949839061e-06, "epoch": 51.84, "percentage": 80.4, "elapsed_time": "1:32:12", "remaining_time": "0:22:28", "throughput": "0.00", "total_tokens": 0}
|
| 163 |
+
{"current_steps": 3260, "total_steps": 4030, "loss": 0.0209, "learning_rate": 5.384013525518541e-06, "epoch": 52.16, "percentage": 80.89, "elapsed_time": "1:32:46", "remaining_time": "0:21:54", "throughput": "0.00", "total_tokens": 0}
|
| 164 |
+
{"current_steps": 3280, "total_steps": 4030, "loss": 0.0202, "learning_rate": 5.118478692955194e-06, "epoch": 52.48, "percentage": 81.39, "elapsed_time": "1:33:19", "remaining_time": "0:21:20", "throughput": "0.00", "total_tokens": 0}
|
| 165 |
+
{"current_steps": 3300, "total_steps": 4030, "loss": 0.0192, "learning_rate": 4.858910136908123e-06, "epoch": 52.8, "percentage": 81.89, "elapsed_time": "1:33:53", "remaining_time": "0:20:46", "throughput": "0.00", "total_tokens": 0}
|
| 166 |
+
{"current_steps": 3320, "total_steps": 4030, "loss": 0.0205, "learning_rate": 4.605385751707248e-06, "epoch": 53.12, "percentage": 82.38, "elapsed_time": "1:34:27", "remaining_time": "0:20:12", "throughput": "0.00", "total_tokens": 0}
|
| 167 |
+
{"current_steps": 3340, "total_steps": 4030, "loss": 0.0129, "learning_rate": 4.357981617877932e-06, "epoch": 53.44, "percentage": 82.88, "elapsed_time": "1:35:01", "remaining_time": "0:19:37", "throughput": "0.00", "total_tokens": 0}
|
| 168 |
+
{"current_steps": 3360, "total_steps": 4030, "loss": 0.0258, "learning_rate": 4.116771979309797e-06, "epoch": 53.76, "percentage": 83.37, "elapsed_time": "1:35:35", "remaining_time": "0:19:03", "throughput": "0.00", "total_tokens": 0}
|
| 169 |
+
{"current_steps": 3380, "total_steps": 4030, "loss": 0.0306, "learning_rate": 3.881829220976807e-06, "epoch": 54.08, "percentage": 83.87, "elapsed_time": "1:36:09", "remaining_time": "0:18:29", "throughput": "0.00", "total_tokens": 0}
|
| 170 |
+
{"current_steps": 3400, "total_steps": 4030, "loss": 0.0198, "learning_rate": 3.653223847215126e-06, "epoch": 54.4, "percentage": 84.37, "elapsed_time": "1:36:43", "remaining_time": "0:17:55", "throughput": "0.00", "total_tokens": 0}
|
| 171 |
+
{"current_steps": 3420, "total_steps": 4030, "loss": 0.0257, "learning_rate": 3.4310244605653797e-06, "epoch": 54.72, "percentage": 84.86, "elapsed_time": "1:37:17", "remaining_time": "0:17:21", "throughput": "0.00", "total_tokens": 0}
|
| 172 |
+
{"current_steps": 3440, "total_steps": 4030, "loss": 0.0125, "learning_rate": 3.215297741185572e-06, "epoch": 55.04, "percentage": 85.36, "elapsed_time": "1:37:52", "remaining_time": "0:16:47", "throughput": "0.00", "total_tokens": 0}
|
| 173 |
+
{"current_steps": 3460, "total_steps": 4030, "loss": 0.0124, "learning_rate": 3.0061084268410006e-06, "epoch": 55.36, "percentage": 85.86, "elapsed_time": "1:38:27", "remaining_time": "0:16:13", "throughput": "0.00", "total_tokens": 0}
|
| 174 |
+
{"current_steps": 3480, "total_steps": 4030, "loss": 0.023, "learning_rate": 2.8035192934769362e-06, "epoch": 55.68, "percentage": 86.35, "elapsed_time": "1:39:01", "remaining_time": "0:15:39", "throughput": "0.00", "total_tokens": 0}
|
| 175 |
+
{"current_steps": 3500, "total_steps": 4030, "loss": 0.0194, "learning_rate": 2.607591136380122e-06, "epoch": 56.0, "percentage": 86.85, "elapsed_time": "1:39:36", "remaining_time": "0:15:04", "throughput": "0.00", "total_tokens": 0}
|
| 176 |
+
{"current_steps": 3520, "total_steps": 4030, "loss": 0.0162, "learning_rate": 2.4183827519346308e-06, "epoch": 56.32, "percentage": 87.34, "elapsed_time": "1:40:10", "remaining_time": "0:14:30", "throughput": "0.00", "total_tokens": 0}
|
| 177 |
+
{"current_steps": 3540, "total_steps": 4030, "loss": 0.0337, "learning_rate": 2.235950919977545e-06, "epoch": 56.64, "percentage": 87.84, "elapsed_time": "1:40:44", "remaining_time": "0:13:56", "throughput": "0.00", "total_tokens": 0}
|
| 178 |
+
{"current_steps": 3560, "total_steps": 4030, "loss": 0.0139, "learning_rate": 2.0603503867598182e-06, "epoch": 56.96, "percentage": 88.34, "elapsed_time": "1:41:18", "remaining_time": "0:13:22", "throughput": "0.00", "total_tokens": 0}
|
| 179 |
+
{"current_steps": 3580, "total_steps": 4030, "loss": 0.0193, "learning_rate": 1.8916338485173823e-06, "epoch": 57.28, "percentage": 88.83, "elapsed_time": "1:41:52", "remaining_time": "0:12:48", "throughput": "0.00", "total_tokens": 0}
|
| 180 |
+
{"current_steps": 3600, "total_steps": 4030, "loss": 0.0203, "learning_rate": 1.7298519356574727e-06, "epoch": 57.6, "percentage": 89.33, "elapsed_time": "1:42:28", "remaining_time": "0:12:14", "throughput": "0.00", "total_tokens": 0}
|
| 181 |
+
{"current_steps": 3620, "total_steps": 4030, "loss": 0.0212, "learning_rate": 1.5750531975648324e-06, "epoch": 57.92, "percentage": 89.83, "elapsed_time": "1:43:02", "remaining_time": "0:11:40", "throughput": "0.00", "total_tokens": 0}
|
| 182 |
+
{"current_steps": 3640, "total_steps": 4030, "loss": 0.0173, "learning_rate": 1.4272840880324934e-06, "epoch": 58.24, "percentage": 90.32, "elapsed_time": "1:43:36", "remaining_time": "0:11:06", "throughput": "0.00", "total_tokens": 0}
|
| 183 |
+
{"current_steps": 3660, "total_steps": 4030, "loss": 0.0139, "learning_rate": 1.286588951321363e-06, "epoch": 58.56, "percentage": 90.82, "elapsed_time": "1:44:10", "remaining_time": "0:10:31", "throughput": "0.00", "total_tokens": 0}
|
| 184 |
+
{"current_steps": 3680, "total_steps": 4030, "loss": 0.0268, "learning_rate": 1.1530100088528867e-06, "epoch": 58.88, "percentage": 91.32, "elapsed_time": "1:44:45", "remaining_time": "0:09:57", "throughput": "0.00", "total_tokens": 0}
|
| 185 |
+
{"current_steps": 3700, "total_steps": 4030, "loss": 0.0191, "learning_rate": 1.0265873465387516e-06, "epoch": 59.2, "percentage": 91.81, "elapsed_time": "1:45:18", "remaining_time": "0:09:23", "throughput": "0.00", "total_tokens": 0}
|
| 186 |
+
{"current_steps": 3720, "total_steps": 4030, "loss": 0.0168, "learning_rate": 9.073589027514789e-07, "epoch": 59.52, "percentage": 92.31, "elapsed_time": "1:45:53", "remaining_time": "0:08:49", "throughput": "0.00", "total_tokens": 0}
|
| 187 |
+
{"current_steps": 3740, "total_steps": 4030, "loss": 0.0246, "learning_rate": 7.953604569393841e-07, "epoch": 59.84, "percentage": 92.8, "elapsed_time": "1:46:27", "remaining_time": "0:08:15", "throughput": "0.00", "total_tokens": 0}
|
| 188 |
+
{"current_steps": 3760, "total_steps": 4030, "loss": 0.019, "learning_rate": 6.906256188895038e-07, "epoch": 60.16, "percentage": 93.3, "elapsed_time": "1:47:00", "remaining_time": "0:07:41", "throughput": "0.00", "total_tokens": 0}
|
| 189 |
+
{"current_steps": 3780, "total_steps": 4030, "loss": 0.0168, "learning_rate": 5.931858186415756e-07, "epoch": 60.48, "percentage": 93.8, "elapsed_time": "1:47:34", "remaining_time": "0:07:06", "throughput": "0.00", "total_tokens": 0}
|
| 190 |
+
{"current_steps": 3800, "total_steps": 4030, "loss": 0.0197, "learning_rate": 5.03070297056149e-07, "epoch": 60.8, "percentage": 94.29, "elapsed_time": "1:48:08", "remaining_time": "0:06:32", "throughput": "0.00", "total_tokens": 0}
|
| 191 |
+
{"current_steps": 3820, "total_steps": 4030, "loss": 0.0207, "learning_rate": 4.203060970396383e-07, "epoch": 61.12, "percentage": 94.79, "elapsed_time": "1:48:42", "remaining_time": "0:05:58", "throughput": "0.00", "total_tokens": 0}
|
| 192 |
+
{"current_steps": 3840, "total_steps": 4030, "loss": 0.0224, "learning_rate": 3.4491805542899157e-07, "epoch": 61.44, "percentage": 95.29, "elapsed_time": "1:49:16", "remaining_time": "0:05:24", "throughput": "0.00", "total_tokens": 0}
|
| 193 |
+
{"current_steps": 3860, "total_steps": 4030, "loss": 0.0151, "learning_rate": 2.769287955383532e-07, "epoch": 61.76, "percentage": 95.78, "elapsed_time": "1:49:51", "remaining_time": "0:04:50", "throughput": "0.00", "total_tokens": 0}
|
| 194 |
+
{"current_steps": 3880, "total_steps": 4030, "loss": 0.0284, "learning_rate": 2.1635872037001626e-07, "epoch": 62.08, "percentage": 96.28, "elapsed_time": "1:50:25", "remaining_time": "0:04:16", "throughput": "0.00", "total_tokens": 0}
|
| 195 |
+
{"current_steps": 3900, "total_steps": 4030, "loss": 0.0217, "learning_rate": 1.6322600649162356e-07, "epoch": 62.4, "percentage": 96.77, "elapsed_time": "1:50:59", "remaining_time": "0:03:41", "throughput": "0.00", "total_tokens": 0}
|
| 196 |
+
{"current_steps": 3920, "total_steps": 4030, "loss": 0.0103, "learning_rate": 1.1754659858156659e-07, "epoch": 62.72, "percentage": 97.27, "elapsed_time": "1:51:34", "remaining_time": "0:03:07", "throughput": "0.00", "total_tokens": 0}
|
| 197 |
+
{"current_steps": 3940, "total_steps": 4030, "loss": 0.0333, "learning_rate": 7.933420464410201e-08, "epoch": 63.04, "percentage": 97.77, "elapsed_time": "1:52:08", "remaining_time": "0:02:33", "throughput": "0.00", "total_tokens": 0}
|
| 198 |
+
{"current_steps": 3960, "total_steps": 4030, "loss": 0.0231, "learning_rate": 4.860029189569237e-08, "epoch": 63.36, "percentage": 98.26, "elapsed_time": "1:52:42", "remaining_time": "0:01:59", "throughput": "0.00", "total_tokens": 0}
|
| 199 |
+
{"current_steps": 3980, "total_steps": 4030, "loss": 0.0226, "learning_rate": 2.535408332381417e-08, "epoch": 63.68, "percentage": 98.76, "elapsed_time": "1:53:16", "remaining_time": "0:01:25", "throughput": "0.00", "total_tokens": 0}
|
| 200 |
+
{"current_steps": 4000, "total_steps": 4030, "loss": 0.0218, "learning_rate": 9.60255491919415e-09, "epoch": 64.0, "percentage": 99.26, "elapsed_time": "1:53:50", "remaining_time": "0:00:51", "throughput": "0.00", "total_tokens": 0}
|
| 201 |
+
{"current_steps": 4020, "total_steps": 4030, "loss": 0.0219, "learning_rate": 1.3504335823810722e-09, "epoch": 64.32, "percentage": 99.75, "elapsed_time": "1:54:25", "remaining_time": "0:00:17", "throughput": "0.00", "total_tokens": 0}
|
| 202 |
+
{"current_steps": 4030, "total_steps": 4030, "epoch": 64.48, "percentage": 100.0, "elapsed_time": "1:54:42", "remaining_time": "0:00:00", "throughput": "0.00", "total_tokens": 0}
|
trainer_state.json
CHANGED
|
@@ -1,1112 +1,1434 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.32,
|
| 13 |
-
"grad_norm": 2.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 20
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.64,
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 2.
|
| 23 |
"step": 40
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.96,
|
| 27 |
-
"grad_norm": 2.
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 2.
|
| 30 |
"step": 60
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 1.28,
|
| 34 |
-
"grad_norm": 1.
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss":
|
| 37 |
"step": 80
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 1.6,
|
| 41 |
-
"grad_norm": 1.
|
| 42 |
-
"learning_rate": 1.
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 100
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 1.92,
|
| 48 |
-
"grad_norm": 3.
|
| 49 |
-
"learning_rate": 1.
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 120
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 2.24,
|
| 55 |
-
"grad_norm": 2.
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 140
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 2.56,
|
| 62 |
-
"grad_norm": 1.
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 160
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 2.88,
|
| 69 |
-
"grad_norm": 4.
|
| 70 |
-
"learning_rate": 2.
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 180
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 3.2,
|
| 76 |
-
"grad_norm": 2.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 200
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 3.52,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 220
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 3.84,
|
| 90 |
-
"grad_norm": 3.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 240
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 4.16,
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 260
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 4.48,
|
| 104 |
-
"grad_norm": 1.
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 280
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 4.8,
|
| 111 |
-
"grad_norm": 2.
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 300
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 5.12,
|
| 118 |
-
"grad_norm": 1.
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 320
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 5.44,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 4.
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 340
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 5.76,
|
| 132 |
-
"grad_norm": 2.
|
| 133 |
-
"learning_rate": 4.
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 360
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 6.08,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate": 4.
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 380
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 6.4,
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate": 4.
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 400
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 6.72,
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate": 4.
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 420
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 7.04,
|
| 160 |
-
"grad_norm": 1.
|
| 161 |
-
"learning_rate": 4.
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 440
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 7.36,
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate": 4.
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 460
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 7.68,
|
| 174 |
-
"grad_norm":
|
| 175 |
-
"learning_rate": 4.
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 480
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 8.0,
|
| 181 |
-
"grad_norm": 1.
|
| 182 |
-
"learning_rate": 4.
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 8.32,
|
| 188 |
-
"grad_norm":
|
| 189 |
-
"learning_rate": 4.
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 520
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 8.64,
|
| 195 |
-
"grad_norm": 2.
|
| 196 |
-
"learning_rate": 4.
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 540
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 8.96,
|
| 202 |
-
"grad_norm":
|
| 203 |
-
"learning_rate": 4.
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 560
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 9.28,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate": 4.
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 580
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 9.6,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate": 4.
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 600
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 9.92,
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate": 4.
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 620
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 10.24,
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate": 4.
|
| 232 |
-
"loss": 0.
|
| 233 |
"step": 640
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 10.56,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate": 4.
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 660
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 10.88,
|
| 244 |
-
"grad_norm":
|
| 245 |
-
"learning_rate": 4.
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 680
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 11.2,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
-
"learning_rate": 4.
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 700
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 11.52,
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate": 4.
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 720
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 11.84,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
-
"learning_rate": 4.
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 740
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 12.16,
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate": 4.
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 760
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 12.48,
|
| 279 |
-
"grad_norm":
|
| 280 |
-
"learning_rate": 4.
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 780
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 12.8,
|
| 286 |
-
"grad_norm": 1.
|
| 287 |
-
"learning_rate": 4.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 800
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 13.12,
|
| 293 |
-
"grad_norm": 1.
|
| 294 |
-
"learning_rate": 4.
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 820
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 13.44,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
-
"learning_rate": 4.
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 840
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 13.76,
|
| 307 |
-
"grad_norm": 1.
|
| 308 |
-
"learning_rate": 4.
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 860
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 14.08,
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
-
"learning_rate": 4.
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 880
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 14.4,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate": 4.
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 900
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 14.72,
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
-
"learning_rate": 4.
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 920
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 15.04,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate": 4.
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 940
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 15.36,
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
-
"learning_rate": 4.
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 960
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 15.68,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
-
"learning_rate": 4.
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 980
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 16.0,
|
| 356 |
-
"grad_norm":
|
| 357 |
-
"learning_rate": 4.
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 1000
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 16.32,
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
-
"learning_rate": 4.
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 1020
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 16.64,
|
| 370 |
-
"grad_norm":
|
| 371 |
-
"learning_rate": 4.
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 1040
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 16.96,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
-
"learning_rate": 4.
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 1060
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 17.28,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate": 4.
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 1080
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 17.6,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
-
"learning_rate": 4.
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 1100
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 17.92,
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
-
"learning_rate": 4.
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 1120
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 18.24,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate":
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 1140
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 18.56,
|
| 412 |
-
"grad_norm": 1.
|
| 413 |
-
"learning_rate":
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 1160
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 18.88,
|
| 419 |
-
"grad_norm":
|
| 420 |
-
"learning_rate":
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 1180
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 19.2,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 1200
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 19.52,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate":
|
| 435 |
-
"loss": 0.
|
| 436 |
"step": 1220
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 19.84,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate":
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 1240
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 20.16,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
-
"learning_rate":
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 1260
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 20.48,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
-
"learning_rate":
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 1280
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 20.8,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 1300
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 21.12,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 1320
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 21.44,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
-
"learning_rate":
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 1340
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 21.76,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
-
"learning_rate":
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 1360
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 22.08,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
-
"learning_rate":
|
| 491 |
-
"loss": 0.
|
| 492 |
"step": 1380
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 22.4,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
-
"learning_rate":
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 1400
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 22.72,
|
| 503 |
-
"grad_norm":
|
| 504 |
-
"learning_rate":
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 1420
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 23.04,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 1440
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 23.36,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
-
"learning_rate":
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 1460
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 23.68,
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
-
"learning_rate": 3.
|
| 526 |
-
"loss": 0.
|
| 527 |
"step": 1480
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 24.0,
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
-
"learning_rate": 3.
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 1500
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 24.32,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
-
"learning_rate": 3.
|
| 540 |
-
"loss": 0.
|
| 541 |
"step": 1520
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 24.64,
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 1540
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 24.96,
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
-
"learning_rate":
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 1560
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 25.28,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
-
"learning_rate":
|
| 561 |
-
"loss": 0.
|
| 562 |
"step": 1580
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 25.6,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
-
"learning_rate":
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 1600
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 25.92,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
-
"learning_rate":
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 1620
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 26.24,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
-
"learning_rate":
|
| 582 |
"loss": 0.0354,
|
| 583 |
"step": 1640
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 26.56,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 1660
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 26.88,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 1680
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 27.2,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
-
"learning_rate":
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 1700
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 27.52,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
-
"learning_rate":
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 1720
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 27.84,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 1740
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 28.16,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
-
"learning_rate":
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 1760
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 28.48,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 1780
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 28.8,
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 1800
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 29.12,
|
| 643 |
-
"grad_norm": 0.
|
| 644 |
-
"learning_rate":
|
| 645 |
-
"loss": 0.
|
| 646 |
"step": 1820
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 29.44,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
-
"learning_rate":
|
| 652 |
-
"loss": 0.
|
| 653 |
"step": 1840
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 29.76,
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
-
"learning_rate":
|
| 659 |
-
"loss": 0.
|
| 660 |
"step": 1860
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 30.08,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
-
"learning_rate":
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 1880
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 30.4,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
-
"learning_rate":
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 1900
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 30.72,
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
-
"learning_rate":
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 1920
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 31.04,
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
-
"learning_rate":
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 1940
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 31.36,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
-
"learning_rate":
|
| 694 |
-
"loss": 0.
|
| 695 |
"step": 1960
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 31.68,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 1980
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 32.0,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
-
"learning_rate":
|
| 708 |
-
"loss": 0.
|
| 709 |
"step": 2000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 32.32,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
-
"learning_rate":
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 2020
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 32.64,
|
| 720 |
-
"grad_norm": 0.
|
| 721 |
-
"learning_rate":
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 2040
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 32.96,
|
| 727 |
-
"grad_norm": 0.
|
| 728 |
-
"learning_rate":
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 2060
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 33.28,
|
| 734 |
-
"grad_norm": 0.
|
| 735 |
-
"learning_rate":
|
| 736 |
-
"loss": 0.
|
| 737 |
"step": 2080
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 33.6,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
-
"learning_rate":
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 2100
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 33.92,
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
-
"learning_rate":
|
| 750 |
-
"loss": 0.
|
| 751 |
"step": 2120
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 34.24,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
-
"learning_rate":
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 2140
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 34.56,
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 2160
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 34.88,
|
| 769 |
-
"grad_norm": 0.
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 2180
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 35.2,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
-
"learning_rate":
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 2200
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 35.52,
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
-
"learning_rate":
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 2220
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 35.84,
|
| 790 |
-
"grad_norm": 0.
|
| 791 |
-
"learning_rate":
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 2240
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 36.16,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
-
"learning_rate":
|
| 799 |
-
"loss": 0.
|
| 800 |
"step": 2260
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 36.48,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
-
"learning_rate":
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 2280
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 36.8,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
-
"learning_rate":
|
| 813 |
-
"loss": 0.
|
| 814 |
"step": 2300
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 37.12,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
-
"learning_rate":
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 2320
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 37.44,
|
| 825 |
-
"grad_norm": 0.
|
| 826 |
-
"learning_rate":
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 2340
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 37.76,
|
| 832 |
-
"grad_norm": 0.
|
| 833 |
-
"learning_rate":
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 2360
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 38.08,
|
| 839 |
-
"grad_norm": 0.
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 2380
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 38.4,
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
-
"learning_rate":
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 2400
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 38.72,
|
| 853 |
-
"grad_norm": 0.
|
| 854 |
-
"learning_rate":
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 2420
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 39.04,
|
| 860 |
-
"grad_norm": 0.
|
| 861 |
-
"learning_rate":
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 2440
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 39.36,
|
| 867 |
-
"grad_norm": 0.
|
| 868 |
-
"learning_rate":
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 2460
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 39.68,
|
| 874 |
-
"grad_norm": 0.
|
| 875 |
-
"learning_rate":
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 2480
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 40.0,
|
| 881 |
-
"grad_norm": 0.
|
| 882 |
-
"learning_rate":
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 2500
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 40.32,
|
| 888 |
-
"grad_norm": 0.
|
| 889 |
-
"learning_rate":
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 2520
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 40.64,
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
-
"learning_rate":
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 2540
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 40.96,
|
| 902 |
-
"grad_norm": 0.
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 2560
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 41.28,
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 2580
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 41.6,
|
| 916 |
-
"grad_norm": 0.
|
| 917 |
-
"learning_rate":
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 2600
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 41.92,
|
| 923 |
-
"grad_norm": 0.
|
| 924 |
-
"learning_rate":
|
| 925 |
-
"loss": 0.
|
| 926 |
"step": 2620
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 42.24,
|
| 930 |
-
"grad_norm": 0.
|
| 931 |
-
"learning_rate":
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 2640
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 42.56,
|
| 937 |
-
"grad_norm": 0.
|
| 938 |
-
"learning_rate":
|
| 939 |
-
"loss": 0.
|
| 940 |
"step": 2660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 42.88,
|
| 944 |
-
"grad_norm": 0.
|
| 945 |
-
"learning_rate":
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 2680
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 43.2,
|
| 951 |
-
"grad_norm": 0.
|
| 952 |
-
"learning_rate":
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 2700
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 43.52,
|
| 958 |
-
"grad_norm": 0.
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 2720
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 43.84,
|
| 965 |
-
"grad_norm": 0.
|
| 966 |
-
"learning_rate":
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 2740
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 44.16,
|
| 972 |
-
"grad_norm": 0.
|
| 973 |
-
"learning_rate": 1.
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 2760
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 44.48,
|
| 979 |
-
"grad_norm": 0.
|
| 980 |
-
"learning_rate": 1.
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 2780
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 44.8,
|
| 986 |
-
"grad_norm": 0.
|
| 987 |
-
"learning_rate": 1.
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 2800
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 45.12,
|
| 993 |
-
"grad_norm": 0.
|
| 994 |
-
"learning_rate": 1.
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 2820
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 45.44,
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
-
"learning_rate": 1.
|
| 1002 |
-
"loss": 0.
|
| 1003 |
"step": 2840
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 45.76,
|
| 1007 |
-
"grad_norm": 0.
|
| 1008 |
-
"learning_rate":
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 2860
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 46.08,
|
| 1014 |
-
"grad_norm": 0.
|
| 1015 |
-
"learning_rate":
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 2880
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 46.4,
|
| 1021 |
-
"grad_norm": 0.
|
| 1022 |
-
"learning_rate":
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 2900
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 46.72,
|
| 1028 |
-
"grad_norm": 0.
|
| 1029 |
-
"learning_rate":
|
| 1030 |
-
"loss": 0.
|
| 1031 |
"step": 2920
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 47.04,
|
| 1035 |
-
"grad_norm": 0.
|
| 1036 |
-
"learning_rate":
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 2940
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 47.36,
|
| 1042 |
-
"grad_norm": 0.
|
| 1043 |
-
"learning_rate":
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 2960
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 47.68,
|
| 1049 |
-
"grad_norm": 0.
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 2980
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 48.0,
|
| 1056 |
-
"grad_norm": 0.
|
| 1057 |
-
"learning_rate":
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 3000
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 48.32,
|
| 1063 |
-
"grad_norm": 0.
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 3020
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 48.64,
|
| 1070 |
-
"grad_norm": 0.
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 3040
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 48.96,
|
| 1077 |
-
"grad_norm": 0.
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 3060
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 49.28,
|
| 1084 |
-
"grad_norm": 0.
|
| 1085 |
-
"learning_rate":
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 3080
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 49.6,
|
| 1091 |
-
"grad_norm": 0.
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 3100
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch": 49.
|
| 1098 |
-
"
|
| 1099 |
-
"
|
| 1100 |
-
"
|
| 1101 |
-
"
|
| 1102 |
-
|
| 1103 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1104 |
}
|
| 1105 |
],
|
| 1106 |
"logging_steps": 20,
|
| 1107 |
-
"max_steps":
|
| 1108 |
"num_input_tokens_seen": 0,
|
| 1109 |
-
"num_train_epochs":
|
| 1110 |
"save_steps": 1000,
|
| 1111 |
"stateful_callbacks": {
|
| 1112 |
"TrainerControl": {
|
|
@@ -1120,7 +1442,7 @@
|
|
| 1120 |
"attributes": {}
|
| 1121 |
}
|
| 1122 |
},
|
| 1123 |
-
"total_flos":
|
| 1124 |
"train_batch_size": 2,
|
| 1125 |
"trial_name": null,
|
| 1126 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 64.48,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 4030,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.32,
|
| 13 |
+
"grad_norm": 2.246424436569214,
|
| 14 |
+
"learning_rate": 2.3573200992555833e-06,
|
| 15 |
+
"loss": 2.826,
|
| 16 |
"step": 20
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.64,
|
| 20 |
+
"grad_norm": 0.9050242900848389,
|
| 21 |
+
"learning_rate": 4.838709677419355e-06,
|
| 22 |
+
"loss": 2.72,
|
| 23 |
"step": 40
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.96,
|
| 27 |
+
"grad_norm": 2.6034655570983887,
|
| 28 |
+
"learning_rate": 7.320099255583126e-06,
|
| 29 |
+
"loss": 2.4912,
|
| 30 |
"step": 60
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 1.28,
|
| 34 |
+
"grad_norm": 1.3487274646759033,
|
| 35 |
+
"learning_rate": 9.801488833746898e-06,
|
| 36 |
+
"loss": 2.0561,
|
| 37 |
"step": 80
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 1.6,
|
| 41 |
+
"grad_norm": 1.6185756921768188,
|
| 42 |
+
"learning_rate": 1.2282878411910669e-05,
|
| 43 |
+
"loss": 1.7744,
|
| 44 |
"step": 100
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 1.92,
|
| 48 |
+
"grad_norm": 3.017139196395874,
|
| 49 |
+
"learning_rate": 1.4764267990074444e-05,
|
| 50 |
+
"loss": 1.8387,
|
| 51 |
"step": 120
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 2.24,
|
| 55 |
+
"grad_norm": 2.2100813388824463,
|
| 56 |
+
"learning_rate": 1.7245657568238215e-05,
|
| 57 |
+
"loss": 1.4478,
|
| 58 |
"step": 140
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 2.56,
|
| 62 |
+
"grad_norm": 1.574629545211792,
|
| 63 |
+
"learning_rate": 1.9727047146401986e-05,
|
| 64 |
+
"loss": 1.285,
|
| 65 |
"step": 160
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 2.88,
|
| 69 |
+
"grad_norm": 4.586638450622559,
|
| 70 |
+
"learning_rate": 2.2208436724565757e-05,
|
| 71 |
+
"loss": 1.2235,
|
| 72 |
"step": 180
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 3.2,
|
| 76 |
+
"grad_norm": 2.7081515789031982,
|
| 77 |
+
"learning_rate": 2.468982630272953e-05,
|
| 78 |
+
"loss": 0.9575,
|
| 79 |
"step": 200
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 3.52,
|
| 83 |
+
"grad_norm": 0.9670729041099548,
|
| 84 |
+
"learning_rate": 2.7171215880893302e-05,
|
| 85 |
+
"loss": 0.7086,
|
| 86 |
"step": 220
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 3.84,
|
| 90 |
+
"grad_norm": 3.229243040084839,
|
| 91 |
+
"learning_rate": 2.9652605459057077e-05,
|
| 92 |
+
"loss": 0.8587,
|
| 93 |
"step": 240
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 4.16,
|
| 97 |
+
"grad_norm": 1.1293463706970215,
|
| 98 |
+
"learning_rate": 3.2133995037220844e-05,
|
| 99 |
+
"loss": 0.5978,
|
| 100 |
"step": 260
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 4.48,
|
| 104 |
+
"grad_norm": 1.7043830156326294,
|
| 105 |
+
"learning_rate": 3.461538461538462e-05,
|
| 106 |
+
"loss": 0.4668,
|
| 107 |
"step": 280
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 4.8,
|
| 111 |
+
"grad_norm": 2.565268039703369,
|
| 112 |
+
"learning_rate": 3.7096774193548386e-05,
|
| 113 |
+
"loss": 0.5667,
|
| 114 |
"step": 300
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 5.12,
|
| 118 |
+
"grad_norm": 1.158849835395813,
|
| 119 |
+
"learning_rate": 3.957816377171216e-05,
|
| 120 |
+
"loss": 0.4373,
|
| 121 |
"step": 320
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 5.44,
|
| 125 |
+
"grad_norm": 2.714164972305298,
|
| 126 |
+
"learning_rate": 4.205955334987593e-05,
|
| 127 |
+
"loss": 0.3492,
|
| 128 |
"step": 340
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 5.76,
|
| 132 |
+
"grad_norm": 2.2089672088623047,
|
| 133 |
+
"learning_rate": 4.45409429280397e-05,
|
| 134 |
+
"loss": 0.4018,
|
| 135 |
"step": 360
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 6.08,
|
| 139 |
+
"grad_norm": 1.8179335594177246,
|
| 140 |
+
"learning_rate": 4.702233250620348e-05,
|
| 141 |
+
"loss": 0.279,
|
| 142 |
"step": 380
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 6.4,
|
| 146 |
+
"grad_norm": 1.4858269691467285,
|
| 147 |
+
"learning_rate": 4.950372208436725e-05,
|
| 148 |
+
"loss": 0.2362,
|
| 149 |
"step": 400
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 6.72,
|
| 153 |
+
"grad_norm": 1.7704375982284546,
|
| 154 |
+
"learning_rate": 4.99975992459978e-05,
|
| 155 |
+
"loss": 0.2665,
|
| 156 |
"step": 420
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 7.04,
|
| 160 |
+
"grad_norm": 1.2611212730407715,
|
| 161 |
+
"learning_rate": 4.9987846973104825e-05,
|
| 162 |
+
"loss": 0.2029,
|
| 163 |
"step": 440
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 7.36,
|
| 167 |
+
"grad_norm": 2.994542360305786,
|
| 168 |
+
"learning_rate": 4.9970596058519116e-05,
|
| 169 |
+
"loss": 0.1747,
|
| 170 |
"step": 460
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 7.68,
|
| 174 |
+
"grad_norm": 2.7456889152526855,
|
| 175 |
+
"learning_rate": 4.994585167909436e-05,
|
| 176 |
+
"loss": 0.1486,
|
| 177 |
"step": 480
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 8.0,
|
| 181 |
+
"grad_norm": 1.8236416578292847,
|
| 182 |
+
"learning_rate": 4.9913621260409695e-05,
|
| 183 |
+
"loss": 0.1866,
|
| 184 |
"step": 500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 8.32,
|
| 188 |
+
"grad_norm": 2.636003017425537,
|
| 189 |
+
"learning_rate": 4.987391447454136e-05,
|
| 190 |
+
"loss": 0.1476,
|
| 191 |
"step": 520
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 8.64,
|
| 195 |
+
"grad_norm": 2.879154920578003,
|
| 196 |
+
"learning_rate": 4.982674323716023e-05,
|
| 197 |
+
"loss": 0.1403,
|
| 198 |
"step": 540
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 8.96,
|
| 202 |
+
"grad_norm": 0.9377075433731079,
|
| 203 |
+
"learning_rate": 4.977212170395598e-05,
|
| 204 |
+
"loss": 0.1018,
|
| 205 |
"step": 560
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 9.28,
|
| 209 |
+
"grad_norm": 0.311233788728714,
|
| 210 |
+
"learning_rate": 4.9710066266389074e-05,
|
| 211 |
+
"loss": 0.0992,
|
| 212 |
"step": 580
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 9.6,
|
| 216 |
+
"grad_norm": 0.8316205739974976,
|
| 217 |
+
"learning_rate": 4.964059554677187e-05,
|
| 218 |
+
"loss": 0.1134,
|
| 219 |
"step": 600
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 9.92,
|
| 223 |
+
"grad_norm": 2.567354679107666,
|
| 224 |
+
"learning_rate": 4.956373039268022e-05,
|
| 225 |
+
"loss": 0.0781,
|
| 226 |
"step": 620
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 10.24,
|
| 230 |
+
"grad_norm": 0.0829504132270813,
|
| 231 |
+
"learning_rate": 4.947949387069721e-05,
|
| 232 |
+
"loss": 0.0892,
|
| 233 |
"step": 640
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 10.56,
|
| 237 |
+
"grad_norm": 0.8588472008705139,
|
| 238 |
+
"learning_rate": 4.938791125949119e-05,
|
| 239 |
+
"loss": 0.0499,
|
| 240 |
"step": 660
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 10.88,
|
| 244 |
+
"grad_norm": 1.2792423963546753,
|
| 245 |
+
"learning_rate": 4.9289010042229765e-05,
|
| 246 |
+
"loss": 0.0831,
|
| 247 |
"step": 680
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 11.2,
|
| 251 |
+
"grad_norm": 0.4728279709815979,
|
| 252 |
+
"learning_rate": 4.918281989833238e-05,
|
| 253 |
+
"loss": 0.0715,
|
| 254 |
"step": 700
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 11.52,
|
| 258 |
+
"grad_norm": 2.5855355262756348,
|
| 259 |
+
"learning_rate": 4.9069372694563756e-05,
|
| 260 |
+
"loss": 0.0718,
|
| 261 |
"step": 720
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 11.84,
|
| 265 |
+
"grad_norm": 0.8059779405593872,
|
| 266 |
+
"learning_rate": 4.8948702475470933e-05,
|
| 267 |
+
"loss": 0.0849,
|
| 268 |
"step": 740
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 12.16,
|
| 272 |
+
"grad_norm": 1.2841193675994873,
|
| 273 |
+
"learning_rate": 4.882084545316684e-05,
|
| 274 |
+
"loss": 0.0683,
|
| 275 |
"step": 760
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 12.48,
|
| 279 |
+
"grad_norm": 1.3422589302062988,
|
| 280 |
+
"learning_rate": 4.868583999646329e-05,
|
| 281 |
+
"loss": 0.0808,
|
| 282 |
"step": 780
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 12.8,
|
| 286 |
+
"grad_norm": 1.3376965522766113,
|
| 287 |
+
"learning_rate": 4.8543726619356846e-05,
|
| 288 |
+
"loss": 0.0607,
|
| 289 |
"step": 800
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 13.12,
|
| 293 |
+
"grad_norm": 1.008899450302124,
|
| 294 |
+
"learning_rate": 4.83945479688709e-05,
|
| 295 |
+
"loss": 0.062,
|
| 296 |
"step": 820
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 13.44,
|
| 300 |
+
"grad_norm": 0.441413551568985,
|
| 301 |
+
"learning_rate": 4.8238348812257684e-05,
|
| 302 |
+
"loss": 0.0461,
|
| 303 |
"step": 840
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 13.76,
|
| 307 |
+
"grad_norm": 1.296985149383545,
|
| 308 |
+
"learning_rate": 4.808349953928184e-05,
|
| 309 |
+
"loss": 0.0482,
|
| 310 |
"step": 860
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 14.08,
|
| 314 |
+
"grad_norm": 0.035805635154247284,
|
| 315 |
+
"learning_rate": 4.791374712344622e-05,
|
| 316 |
+
"loss": 0.0388,
|
| 317 |
"step": 880
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 14.4,
|
| 321 |
+
"grad_norm": 0.10618308186531067,
|
| 322 |
+
"learning_rate": 4.7737118485753564e-05,
|
| 323 |
+
"loss": 0.0251,
|
| 324 |
"step": 900
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 14.72,
|
| 328 |
+
"grad_norm": 0.866423487663269,
|
| 329 |
+
"learning_rate": 4.75536666309653e-05,
|
| 330 |
+
"loss": 0.0515,
|
| 331 |
"step": 920
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 15.04,
|
| 335 |
+
"grad_norm": 0.5916399955749512,
|
| 336 |
+
"learning_rate": 4.73634466114326e-05,
|
| 337 |
+
"loss": 0.0536,
|
| 338 |
"step": 940
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 15.36,
|
| 342 |
+
"grad_norm": 0.1653570532798767,
|
| 343 |
+
"learning_rate": 4.7166515510575676e-05,
|
| 344 |
+
"loss": 0.0392,
|
| 345 |
"step": 960
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 15.68,
|
| 349 |
+
"grad_norm": 0.027391331270337105,
|
| 350 |
+
"learning_rate": 4.696293242575356e-05,
|
| 351 |
+
"loss": 0.0369,
|
| 352 |
"step": 980
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 16.0,
|
| 356 |
+
"grad_norm": 2.17256760597229,
|
| 357 |
+
"learning_rate": 4.675275845052942e-05,
|
| 358 |
+
"loss": 0.0651,
|
| 359 |
"step": 1000
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 16.32,
|
| 363 |
+
"grad_norm": 0.8612786531448364,
|
| 364 |
+
"learning_rate": 4.6536056656336947e-05,
|
| 365 |
+
"loss": 0.037,
|
| 366 |
"step": 1020
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 16.64,
|
| 370 |
+
"grad_norm": 4.489969253540039,
|
| 371 |
+
"learning_rate": 4.631289207355313e-05,
|
| 372 |
+
"loss": 0.0272,
|
| 373 |
"step": 1040
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 16.96,
|
| 377 |
+
"grad_norm": 0.4311043918132782,
|
| 378 |
+
"learning_rate": 4.6083331671983185e-05,
|
| 379 |
+
"loss": 0.0507,
|
| 380 |
"step": 1060
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 17.28,
|
| 384 |
+
"grad_norm": 0.4327545762062073,
|
| 385 |
+
"learning_rate": 4.584744434076352e-05,
|
| 386 |
+
"loss": 0.0274,
|
| 387 |
"step": 1080
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 17.6,
|
| 391 |
+
"grad_norm": 0.12099918723106384,
|
| 392 |
+
"learning_rate": 4.560530086768863e-05,
|
| 393 |
+
"loss": 0.0565,
|
| 394 |
"step": 1100
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 17.92,
|
| 398 |
+
"grad_norm": 0.103216253221035,
|
| 399 |
+
"learning_rate": 4.535697391796832e-05,
|
| 400 |
+
"loss": 0.0425,
|
| 401 |
"step": 1120
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 18.24,
|
| 405 |
+
"grad_norm": 0.419209748506546,
|
| 406 |
+
"learning_rate": 4.510253801242147e-05,
|
| 407 |
+
"loss": 0.0273,
|
| 408 |
"step": 1140
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 18.56,
|
| 412 |
+
"grad_norm": 1.3193784952163696,
|
| 413 |
+
"learning_rate": 4.4842069505112984e-05,
|
| 414 |
+
"loss": 0.0438,
|
| 415 |
"step": 1160
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 18.88,
|
| 419 |
+
"grad_norm": 1.5185387134552002,
|
| 420 |
+
"learning_rate": 4.457564656044056e-05,
|
| 421 |
+
"loss": 0.0544,
|
| 422 |
"step": 1180
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 19.2,
|
| 426 |
+
"grad_norm": 0.4024270474910736,
|
| 427 |
+
"learning_rate": 4.430334912967824e-05,
|
| 428 |
+
"loss": 0.0283,
|
| 429 |
"step": 1200
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 19.52,
|
| 433 |
+
"grad_norm": 0.16141988337039948,
|
| 434 |
+
"learning_rate": 4.402525892698367e-05,
|
| 435 |
+
"loss": 0.0393,
|
| 436 |
"step": 1220
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 19.84,
|
| 440 |
+
"grad_norm": 0.07228437811136246,
|
| 441 |
+
"learning_rate": 4.374145940487641e-05,
|
| 442 |
+
"loss": 0.0249,
|
| 443 |
"step": 1240
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 20.16,
|
| 447 |
+
"grad_norm": 0.7919737696647644,
|
| 448 |
+
"learning_rate": 4.345203572919454e-05,
|
| 449 |
+
"loss": 0.0293,
|
| 450 |
"step": 1260
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 20.48,
|
| 454 |
+
"grad_norm": 0.26585039496421814,
|
| 455 |
+
"learning_rate": 4.315707475353706e-05,
|
| 456 |
+
"loss": 0.0287,
|
| 457 |
"step": 1280
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 20.8,
|
| 461 |
+
"grad_norm": 0.5761149525642395,
|
| 462 |
+
"learning_rate": 4.285666499319992e-05,
|
| 463 |
+
"loss": 0.0521,
|
| 464 |
"step": 1300
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 21.12,
|
| 468 |
+
"grad_norm": 0.018601374700665474,
|
| 469 |
+
"learning_rate": 4.25508965986133e-05,
|
| 470 |
+
"loss": 0.0285,
|
| 471 |
"step": 1320
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 21.44,
|
| 475 |
+
"grad_norm": 0.00528874434530735,
|
| 476 |
+
"learning_rate": 4.2239861328288214e-05,
|
| 477 |
+
"loss": 0.0346,
|
| 478 |
"step": 1340
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 21.76,
|
| 482 |
+
"grad_norm": 0.3073647618293762,
|
| 483 |
+
"learning_rate": 4.1923652521280585e-05,
|
| 484 |
+
"loss": 0.022,
|
| 485 |
"step": 1360
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 22.08,
|
| 489 |
+
"grad_norm": 0.42911043763160706,
|
| 490 |
+
"learning_rate": 4.160236506918098e-05,
|
| 491 |
+
"loss": 0.0482,
|
| 492 |
"step": 1380
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 22.4,
|
| 496 |
+
"grad_norm": 0.6457176804542542,
|
| 497 |
+
"learning_rate": 4.127609538763842e-05,
|
| 498 |
+
"loss": 0.019,
|
| 499 |
"step": 1400
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 22.72,
|
| 503 |
+
"grad_norm": 2.3716557025909424,
|
| 504 |
+
"learning_rate": 4.094494138742685e-05,
|
| 505 |
+
"loss": 0.0312,
|
| 506 |
"step": 1420
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 23.04,
|
| 510 |
+
"grad_norm": 0.01667410507798195,
|
| 511 |
+
"learning_rate": 4.0609002445063036e-05,
|
| 512 |
+
"loss": 0.0377,
|
| 513 |
"step": 1440
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 23.36,
|
| 517 |
+
"grad_norm": 0.6381007432937622,
|
| 518 |
+
"learning_rate": 4.02683793729844e-05,
|
| 519 |
+
"loss": 0.0307,
|
| 520 |
"step": 1460
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 23.68,
|
| 524 |
+
"grad_norm": 0.42919328808784485,
|
| 525 |
+
"learning_rate": 3.9923174389296085e-05,
|
| 526 |
+
"loss": 0.0419,
|
| 527 |
"step": 1480
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 24.0,
|
| 531 |
+
"grad_norm": 0.01456019002944231,
|
| 532 |
+
"learning_rate": 3.957349108709623e-05,
|
| 533 |
+
"loss": 0.0223,
|
| 534 |
"step": 1500
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 24.32,
|
| 538 |
+
"grad_norm": 0.31073492765426636,
|
| 539 |
+
"learning_rate": 3.921943440338849e-05,
|
| 540 |
+
"loss": 0.0209,
|
| 541 |
"step": 1520
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 24.64,
|
| 545 |
+
"grad_norm": 0.38279736042022705,
|
| 546 |
+
"learning_rate": 3.886111058759132e-05,
|
| 547 |
+
"loss": 0.0491,
|
| 548 |
"step": 1540
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 24.96,
|
| 552 |
+
"grad_norm": 0.30651962757110596,
|
| 553 |
+
"learning_rate": 3.849862716965352e-05,
|
| 554 |
+
"loss": 0.0298,
|
| 555 |
"step": 1560
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 25.28,
|
| 559 |
+
"grad_norm": 0.4538489580154419,
|
| 560 |
+
"learning_rate": 3.813209292778527e-05,
|
| 561 |
+
"loss": 0.0319,
|
| 562 |
"step": 1580
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 25.6,
|
| 566 |
+
"grad_norm": 0.11643072962760925,
|
| 567 |
+
"learning_rate": 3.776161785581481e-05,
|
| 568 |
+
"loss": 0.0302,
|
| 569 |
"step": 1600
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 25.92,
|
| 573 |
+
"grad_norm": 0.008515519089996815,
|
| 574 |
+
"learning_rate": 3.738731313018019e-05,
|
| 575 |
+
"loss": 0.04,
|
| 576 |
"step": 1620
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 26.24,
|
| 580 |
+
"grad_norm": 0.002214708598330617,
|
| 581 |
+
"learning_rate": 3.700929107656614e-05,
|
| 582 |
"loss": 0.0354,
|
| 583 |
"step": 1640
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 26.56,
|
| 587 |
+
"grad_norm": 0.02200801856815815,
|
| 588 |
+
"learning_rate": 3.662766513619611e-05,
|
| 589 |
+
"loss": 0.0186,
|
| 590 |
"step": 1660
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 26.88,
|
| 594 |
+
"grad_norm": 0.1882447600364685,
|
| 595 |
+
"learning_rate": 3.62425498317895e-05,
|
| 596 |
+
"loss": 0.022,
|
| 597 |
"step": 1680
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 27.2,
|
| 601 |
+
"grad_norm": 0.004948125686496496,
|
| 602 |
+
"learning_rate": 3.585406073319439e-05,
|
| 603 |
+
"loss": 0.015,
|
| 604 |
"step": 1700
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 27.52,
|
| 608 |
+
"grad_norm": 0.3387264013290405,
|
| 609 |
+
"learning_rate": 3.546231442270596e-05,
|
| 610 |
+
"loss": 0.0381,
|
| 611 |
"step": 1720
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 27.84,
|
| 615 |
+
"grad_norm": 0.09048642963171005,
|
| 616 |
+
"learning_rate": 3.506742846008116e-05,
|
| 617 |
+
"loss": 0.0277,
|
| 618 |
"step": 1740
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 28.16,
|
| 622 |
+
"grad_norm": 0.6405784487724304,
|
| 623 |
+
"learning_rate": 3.4669521347259996e-05,
|
| 624 |
+
"loss": 0.0423,
|
| 625 |
"step": 1760
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 28.48,
|
| 629 |
+
"grad_norm": 0.16012047231197357,
|
| 630 |
+
"learning_rate": 3.426871249280414e-05,
|
| 631 |
+
"loss": 0.0115,
|
| 632 |
"step": 1780
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 28.8,
|
| 636 |
+
"grad_norm": 0.3279825448989868,
|
| 637 |
+
"learning_rate": 3.386512217606339e-05,
|
| 638 |
+
"loss": 0.0275,
|
| 639 |
"step": 1800
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 29.12,
|
| 643 |
+
"grad_norm": 0.005494344513863325,
|
| 644 |
+
"learning_rate": 3.345887151108087e-05,
|
| 645 |
+
"loss": 0.0309,
|
| 646 |
"step": 1820
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 29.44,
|
| 650 |
+
"grad_norm": 0.0037028896622359753,
|
| 651 |
+
"learning_rate": 3.305008241024774e-05,
|
| 652 |
+
"loss": 0.0294,
|
| 653 |
"step": 1840
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 29.76,
|
| 657 |
+
"grad_norm": 0.003084386931732297,
|
| 658 |
+
"learning_rate": 3.2638877547718264e-05,
|
| 659 |
+
"loss": 0.0213,
|
| 660 |
"step": 1860
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 30.08,
|
| 664 |
+
"grad_norm": 0.0017954249633476138,
|
| 665 |
+
"learning_rate": 3.222538032259643e-05,
|
| 666 |
+
"loss": 0.0326,
|
| 667 |
"step": 1880
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 30.4,
|
| 671 |
+
"grad_norm": 0.26840922236442566,
|
| 672 |
+
"learning_rate": 3.1809714821904834e-05,
|
| 673 |
+
"loss": 0.0249,
|
| 674 |
"step": 1900
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 30.72,
|
| 678 |
+
"grad_norm": 0.7214370965957642,
|
| 679 |
+
"learning_rate": 3.1392005783347244e-05,
|
| 680 |
+
"loss": 0.0115,
|
| 681 |
"step": 1920
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 31.04,
|
| 685 |
+
"grad_norm": 0.1613769233226776,
|
| 686 |
+
"learning_rate": 3.0972378557875884e-05,
|
| 687 |
+
"loss": 0.0322,
|
| 688 |
"step": 1940
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 31.36,
|
| 692 |
+
"grad_norm": 0.18066717684268951,
|
| 693 |
+
"learning_rate": 3.055095907207465e-05,
|
| 694 |
+
"loss": 0.0316,
|
| 695 |
"step": 1960
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 31.68,
|
| 699 |
+
"grad_norm": 0.24756371974945068,
|
| 700 |
+
"learning_rate": 3.0127873790369627e-05,
|
| 701 |
+
"loss": 0.0248,
|
| 702 |
"step": 1980
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 32.0,
|
| 706 |
+
"grad_norm": 0.08604203909635544,
|
| 707 |
+
"learning_rate": 2.9703249677078156e-05,
|
| 708 |
+
"loss": 0.0234,
|
| 709 |
"step": 2000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 32.32,
|
| 713 |
+
"grad_norm": 0.0022385423071682453,
|
| 714 |
+
"learning_rate": 2.9277214158307937e-05,
|
| 715 |
+
"loss": 0.0277,
|
| 716 |
"step": 2020
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 32.64,
|
| 720 |
+
"grad_norm": 0.0020592950750142336,
|
| 721 |
+
"learning_rate": 2.8849895083717537e-05,
|
| 722 |
+
"loss": 0.0162,
|
| 723 |
"step": 2040
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 32.96,
|
| 727 |
+
"grad_norm": 0.20633552968502045,
|
| 728 |
+
"learning_rate": 2.842142068814977e-05,
|
| 729 |
+
"loss": 0.022,
|
| 730 |
"step": 2060
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 33.28,
|
| 734 |
+
"grad_norm": 0.0019172705942764878,
|
| 735 |
+
"learning_rate": 2.7991919553149497e-05,
|
| 736 |
+
"loss": 0.0278,
|
| 737 |
"step": 2080
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 33.6,
|
| 741 |
+
"grad_norm": 0.0013098755152896047,
|
| 742 |
+
"learning_rate": 2.756152056837743e-05,
|
| 743 |
+
"loss": 0.0189,
|
| 744 |
"step": 2100
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 33.92,
|
| 748 |
+
"grad_norm": 0.09349821507930756,
|
| 749 |
+
"learning_rate": 2.7130352892931388e-05,
|
| 750 |
+
"loss": 0.0228,
|
| 751 |
"step": 2120
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 34.24,
|
| 755 |
+
"grad_norm": 0.0017231553792953491,
|
| 756 |
+
"learning_rate": 2.669854591658679e-05,
|
| 757 |
+
"loss": 0.0319,
|
| 758 |
"step": 2140
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 34.56,
|
| 762 |
+
"grad_norm": 0.047173839062452316,
|
| 763 |
+
"learning_rate": 2.6266229220967818e-05,
|
| 764 |
+
"loss": 0.0153,
|
| 765 |
"step": 2160
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 34.88,
|
| 769 |
+
"grad_norm": 0.2877206802368164,
|
| 770 |
+
"learning_rate": 2.5833532540661127e-05,
|
| 771 |
+
"loss": 0.0267,
|
| 772 |
"step": 2180
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 35.2,
|
| 776 |
+
"grad_norm": 0.25823402404785156,
|
| 777 |
+
"learning_rate": 2.540058572428356e-05,
|
| 778 |
+
"loss": 0.0178,
|
| 779 |
"step": 2200
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 35.52,
|
| 783 |
+
"grad_norm": 0.23003694415092468,
|
| 784 |
+
"learning_rate": 2.496751869551567e-05,
|
| 785 |
+
"loss": 0.0217,
|
| 786 |
"step": 2220
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 35.84,
|
| 790 |
+
"grad_norm": 0.23193888366222382,
|
| 791 |
+
"learning_rate": 2.453446141411273e-05,
|
| 792 |
+
"loss": 0.017,
|
| 793 |
"step": 2240
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 36.16,
|
| 797 |
+
"grad_norm": 0.1941184252500534,
|
| 798 |
+
"learning_rate": 2.4101543836904938e-05,
|
| 799 |
+
"loss": 0.0257,
|
| 800 |
"step": 2260
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 36.48,
|
| 804 |
+
"grad_norm": 0.012731954455375671,
|
| 805 |
+
"learning_rate": 2.3668895878798424e-05,
|
| 806 |
+
"loss": 0.0237,
|
| 807 |
"step": 2280
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 36.8,
|
| 811 |
+
"grad_norm": 0.18219026923179626,
|
| 812 |
+
"learning_rate": 2.32366473737889e-05,
|
| 813 |
+
"loss": 0.024,
|
| 814 |
"step": 2300
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 37.12,
|
| 818 |
+
"grad_norm": 0.256547212600708,
|
| 819 |
+
"learning_rate": 2.2804928035999594e-05,
|
| 820 |
+
"loss": 0.0225,
|
| 821 |
"step": 2320
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 37.44,
|
| 825 |
+
"grad_norm": 0.45314905047416687,
|
| 826 |
+
"learning_rate": 2.23738674207551e-05,
|
| 827 |
+
"loss": 0.0239,
|
| 828 |
"step": 2340
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 37.76,
|
| 832 |
+
"grad_norm": 0.3919714689254761,
|
| 833 |
+
"learning_rate": 2.1943594885702984e-05,
|
| 834 |
+
"loss": 0.0235,
|
| 835 |
"step": 2360
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 38.08,
|
| 839 |
+
"grad_norm": 0.0769328773021698,
|
| 840 |
+
"learning_rate": 2.151423955199456e-05,
|
| 841 |
+
"loss": 0.0286,
|
| 842 |
"step": 2380
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 38.4,
|
| 846 |
+
"grad_norm": 0.3520802855491638,
|
| 847 |
+
"learning_rate": 2.108593026553681e-05,
|
| 848 |
+
"loss": 0.0323,
|
| 849 |
"step": 2400
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 38.72,
|
| 853 |
+
"grad_norm": 0.3691672384738922,
|
| 854 |
+
"learning_rate": 2.0658795558326743e-05,
|
| 855 |
+
"loss": 0.0241,
|
| 856 |
"step": 2420
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 39.04,
|
| 860 |
+
"grad_norm": 0.001480752951465547,
|
| 861 |
+
"learning_rate": 2.0232963609880093e-05,
|
| 862 |
+
"loss": 0.0158,
|
| 863 |
"step": 2440
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 39.36,
|
| 867 |
+
"grad_norm": 0.31921085715293884,
|
| 868 |
+
"learning_rate": 1.9808562208765667e-05,
|
| 869 |
+
"loss": 0.0241,
|
| 870 |
"step": 2460
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 39.68,
|
| 874 |
+
"grad_norm": 0.20936931669712067,
|
| 875 |
+
"learning_rate": 1.938571871425715e-05,
|
| 876 |
+
"loss": 0.0174,
|
| 877 |
"step": 2480
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 40.0,
|
| 881 |
+
"grad_norm": 0.0011563162552192807,
|
| 882 |
+
"learning_rate": 1.896456001811357e-05,
|
| 883 |
+
"loss": 0.0183,
|
| 884 |
"step": 2500
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 40.32,
|
| 888 |
+
"grad_norm": 0.19230084121227264,
|
| 889 |
+
"learning_rate": 1.854521250650026e-05,
|
| 890 |
+
"loss": 0.012,
|
| 891 |
"step": 2520
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 40.64,
|
| 895 |
+
"grad_norm": 0.32013317942619324,
|
| 896 |
+
"learning_rate": 1.8127802022061334e-05,
|
| 897 |
+
"loss": 0.0225,
|
| 898 |
"step": 2540
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 40.96,
|
| 902 |
+
"grad_norm": 0.11989307403564453,
|
| 903 |
+
"learning_rate": 1.7712453826155457e-05,
|
| 904 |
+
"loss": 0.0391,
|
| 905 |
"step": 2560
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 41.28,
|
| 909 |
+
"grad_norm": 0.0009496643324382603,
|
| 910 |
+
"learning_rate": 1.72992925612659e-05,
|
| 911 |
+
"loss": 0.0229,
|
| 912 |
"step": 2580
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 41.6,
|
| 916 |
+
"grad_norm": 0.0012078011641278863,
|
| 917 |
+
"learning_rate": 1.688844221359645e-05,
|
| 918 |
+
"loss": 0.015,
|
| 919 |
"step": 2600
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 41.92,
|
| 923 |
+
"grad_norm": 0.0012093032710254192,
|
| 924 |
+
"learning_rate": 1.6480026075864163e-05,
|
| 925 |
+
"loss": 0.0287,
|
| 926 |
"step": 2620
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 42.24,
|
| 930 |
+
"grad_norm": 0.2027181088924408,
|
| 931 |
+
"learning_rate": 1.6074166710300247e-05,
|
| 932 |
+
"loss": 0.0229,
|
| 933 |
"step": 2640
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 42.56,
|
| 937 |
+
"grad_norm": 0.2977555990219116,
|
| 938 |
+
"learning_rate": 1.567098591187021e-05,
|
| 939 |
+
"loss": 0.0352,
|
| 940 |
"step": 2660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 42.88,
|
| 944 |
+
"grad_norm": 0.36129167675971985,
|
| 945 |
+
"learning_rate": 1.5270604671724188e-05,
|
| 946 |
+
"loss": 0.0242,
|
| 947 |
"step": 2680
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 43.2,
|
| 951 |
+
"grad_norm": 0.001115540275350213,
|
| 952 |
+
"learning_rate": 1.4873143140888538e-05,
|
| 953 |
+
"loss": 0.0165,
|
| 954 |
"step": 2700
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 43.52,
|
| 958 |
+
"grad_norm": 0.19148553907871246,
|
| 959 |
+
"learning_rate": 1.4478720594209532e-05,
|
| 960 |
+
"loss": 0.0274,
|
| 961 |
"step": 2720
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 43.84,
|
| 965 |
+
"grad_norm": 0.057757727801799774,
|
| 966 |
+
"learning_rate": 1.4087455394559984e-05,
|
| 967 |
+
"loss": 0.0185,
|
| 968 |
"step": 2740
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 44.16,
|
| 972 |
+
"grad_norm": 0.0009874219540506601,
|
| 973 |
+
"learning_rate": 1.369946495731954e-05,
|
| 974 |
+
"loss": 0.0509,
|
| 975 |
"step": 2760
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 44.48,
|
| 979 |
+
"grad_norm": 0.3896861672401428,
|
| 980 |
+
"learning_rate": 1.3314865715139346e-05,
|
| 981 |
+
"loss": 0.027,
|
| 982 |
"step": 2780
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 44.8,
|
| 986 |
+
"grad_norm": 0.19004037976264954,
|
| 987 |
+
"learning_rate": 1.2933773083001517e-05,
|
| 988 |
+
"loss": 0.0163,
|
| 989 |
"step": 2800
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 45.12,
|
| 993 |
+
"grad_norm": 0.0009183284710161388,
|
| 994 |
+
"learning_rate": 1.255630142358421e-05,
|
| 995 |
+
"loss": 0.0125,
|
| 996 |
"step": 2820
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 45.44,
|
| 1000 |
+
"grad_norm": 0.1238480657339096,
|
| 1001 |
+
"learning_rate": 1.2182564012942193e-05,
|
| 1002 |
+
"loss": 0.0327,
|
| 1003 |
"step": 2840
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 45.76,
|
| 1007 |
+
"grad_norm": 0.0009572324343025684,
|
| 1008 |
+
"learning_rate": 1.1812673006513789e-05,
|
| 1009 |
+
"loss": 0.0302,
|
| 1010 |
"step": 2860
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 46.08,
|
| 1014 |
+
"grad_norm": 0.0011610776418820024,
|
| 1015 |
+
"learning_rate": 1.14467394054639e-05,
|
| 1016 |
+
"loss": 0.0209,
|
| 1017 |
"step": 2880
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 46.4,
|
| 1021 |
+
"grad_norm": 0.04993343725800514,
|
| 1022 |
+
"learning_rate": 1.108487302337353e-05,
|
| 1023 |
+
"loss": 0.025,
|
| 1024 |
"step": 2900
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 46.72,
|
| 1028 |
+
"grad_norm": 0.1806841343641281,
|
| 1029 |
+
"learning_rate": 1.0727182453285647e-05,
|
| 1030 |
+
"loss": 0.0284,
|
| 1031 |
"step": 2920
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 47.04,
|
| 1035 |
+
"grad_norm": 0.0011777572799474,
|
| 1036 |
+
"learning_rate": 1.0373775035117305e-05,
|
| 1037 |
+
"loss": 0.0174,
|
| 1038 |
"step": 2940
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 47.36,
|
| 1042 |
+
"grad_norm": 0.14497865736484528,
|
| 1043 |
+
"learning_rate": 1.002475682344792e-05,
|
| 1044 |
+
"loss": 0.0115,
|
| 1045 |
"step": 2960
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 47.68,
|
| 1049 |
+
"grad_norm": 0.0014984839363023639,
|
| 1050 |
+
"learning_rate": 9.680232555693067e-06,
|
| 1051 |
+
"loss": 0.0238,
|
| 1052 |
"step": 2980
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 48.0,
|
| 1056 |
+
"grad_norm": 0.07430601865053177,
|
| 1057 |
+
"learning_rate": 9.340305620673778e-06,
|
| 1058 |
+
"loss": 0.0294,
|
| 1059 |
"step": 3000
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 48.32,
|
| 1063 |
+
"grad_norm": 0.07801785320043564,
|
| 1064 |
+
"learning_rate": 9.005078027590375e-06,
|
| 1065 |
+
"loss": 0.0226,
|
| 1066 |
"step": 3020
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 48.64,
|
| 1070 |
+
"grad_norm": 0.0007196432561613619,
|
| 1071 |
+
"learning_rate": 8.67465037541038e-06,
|
| 1072 |
+
"loss": 0.0196,
|
| 1073 |
"step": 3040
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 48.96,
|
| 1077 |
+
"grad_norm": 0.0008374506141990423,
|
| 1078 |
+
"learning_rate": 8.34912182267959e-06,
|
| 1079 |
+
"loss": 0.0175,
|
| 1080 |
"step": 3060
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 49.28,
|
| 1084 |
+
"grad_norm": 0.0010465418454259634,
|
| 1085 |
+
"learning_rate": 8.028590057765523e-06,
|
| 1086 |
+
"loss": 0.015,
|
| 1087 |
"step": 3080
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 49.6,
|
| 1091 |
+
"grad_norm": 0.0007761380402371287,
|
| 1092 |
+
"learning_rate": 7.713151269541844e-06,
|
| 1093 |
+
"loss": 0.0221,
|
| 1094 |
"step": 3100
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
+
"epoch": 49.92,
|
| 1098 |
+
"grad_norm": 0.0216947291046381,
|
| 1099 |
+
"learning_rate": 7.402900118522979e-06,
|
| 1100 |
+
"loss": 0.0161,
|
| 1101 |
+
"step": 3120
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 50.24,
|
| 1105 |
+
"grad_norm": 0.26546710729599,
|
| 1106 |
+
"learning_rate": 7.097929708457282e-06,
|
| 1107 |
+
"loss": 0.0237,
|
| 1108 |
+
"step": 3140
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 50.56,
|
| 1112 |
+
"grad_norm": 0.0011781662469729781,
|
| 1113 |
+
"learning_rate": 6.7983315583873695e-06,
|
| 1114 |
+
"loss": 0.0172,
|
| 1115 |
+
"step": 3160
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 50.88,
|
| 1119 |
+
"grad_norm": 0.39518535137176514,
|
| 1120 |
+
"learning_rate": 6.504195575186009e-06,
|
| 1121 |
+
"loss": 0.0198,
|
| 1122 |
+
"step": 3180
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 51.2,
|
| 1126 |
+
"grad_norm": 0.3506232500076294,
|
| 1127 |
+
"learning_rate": 6.215610026575916e-06,
|
| 1128 |
+
"loss": 0.0227,
|
| 1129 |
+
"step": 3200
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 51.52,
|
| 1133 |
+
"grad_norm": 0.31244903802871704,
|
| 1134 |
+
"learning_rate": 5.93266151464123e-06,
|
| 1135 |
+
"loss": 0.0156,
|
| 1136 |
+
"step": 3220
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 51.84,
|
| 1140 |
+
"grad_norm": 0.17840787768363953,
|
| 1141 |
+
"learning_rate": 5.655434949839061e-06,
|
| 1142 |
+
"loss": 0.0268,
|
| 1143 |
+
"step": 3240
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 52.16,
|
| 1147 |
+
"grad_norm": 0.1670505702495575,
|
| 1148 |
+
"learning_rate": 5.384013525518541e-06,
|
| 1149 |
+
"loss": 0.0209,
|
| 1150 |
+
"step": 3260
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 52.48,
|
| 1154 |
+
"grad_norm": 0.0010594127234071493,
|
| 1155 |
+
"learning_rate": 5.118478692955194e-06,
|
| 1156 |
+
"loss": 0.0202,
|
| 1157 |
+
"step": 3280
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 52.8,
|
| 1161 |
+
"grad_norm": 0.0015649694250896573,
|
| 1162 |
+
"learning_rate": 4.858910136908123e-06,
|
| 1163 |
+
"loss": 0.0192,
|
| 1164 |
+
"step": 3300
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 53.12,
|
| 1168 |
+
"grad_norm": 0.19762022793293,
|
| 1169 |
+
"learning_rate": 4.605385751707248e-06,
|
| 1170 |
+
"loss": 0.0205,
|
| 1171 |
+
"step": 3320
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 53.44,
|
| 1175 |
+
"grad_norm": 0.2010522186756134,
|
| 1176 |
+
"learning_rate": 4.357981617877932e-06,
|
| 1177 |
+
"loss": 0.0129,
|
| 1178 |
+
"step": 3340
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 53.76,
|
| 1182 |
+
"grad_norm": 0.19793441891670227,
|
| 1183 |
+
"learning_rate": 4.116771979309797e-06,
|
| 1184 |
+
"loss": 0.0258,
|
| 1185 |
+
"step": 3360
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 54.08,
|
| 1189 |
+
"grad_norm": 0.2605569064617157,
|
| 1190 |
+
"learning_rate": 3.881829220976807e-06,
|
| 1191 |
+
"loss": 0.0306,
|
| 1192 |
+
"step": 3380
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 54.4,
|
| 1196 |
+
"grad_norm": 0.037421807646751404,
|
| 1197 |
+
"learning_rate": 3.653223847215126e-06,
|
| 1198 |
+
"loss": 0.0198,
|
| 1199 |
+
"step": 3400
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 54.72,
|
| 1203 |
+
"grad_norm": 0.0007586870342493057,
|
| 1204 |
+
"learning_rate": 3.4310244605653797e-06,
|
| 1205 |
+
"loss": 0.0257,
|
| 1206 |
+
"step": 3420
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 55.04,
|
| 1210 |
+
"grad_norm": 0.27584579586982727,
|
| 1211 |
+
"learning_rate": 3.215297741185572e-06,
|
| 1212 |
+
"loss": 0.0125,
|
| 1213 |
+
"step": 3440
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 55.36,
|
| 1217 |
+
"grad_norm": 0.0007228174363262951,
|
| 1218 |
+
"learning_rate": 3.0061084268410006e-06,
|
| 1219 |
+
"loss": 0.0124,
|
| 1220 |
+
"step": 3460
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 55.68,
|
| 1224 |
+
"grad_norm": 0.04090801998972893,
|
| 1225 |
+
"learning_rate": 2.8035192934769362e-06,
|
| 1226 |
+
"loss": 0.023,
|
| 1227 |
+
"step": 3480
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 56.0,
|
| 1231 |
+
"grad_norm": 0.3518761694431305,
|
| 1232 |
+
"learning_rate": 2.607591136380122e-06,
|
| 1233 |
+
"loss": 0.0194,
|
| 1234 |
+
"step": 3500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 56.32,
|
| 1238 |
+
"grad_norm": 0.06331823766231537,
|
| 1239 |
+
"learning_rate": 2.4183827519346308e-06,
|
| 1240 |
+
"loss": 0.0162,
|
| 1241 |
+
"step": 3520
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 56.64,
|
| 1245 |
+
"grad_norm": 0.22303640842437744,
|
| 1246 |
+
"learning_rate": 2.235950919977545e-06,
|
| 1247 |
+
"loss": 0.0337,
|
| 1248 |
+
"step": 3540
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 56.96,
|
| 1252 |
+
"grad_norm": 0.08465743064880371,
|
| 1253 |
+
"learning_rate": 2.0603503867598182e-06,
|
| 1254 |
+
"loss": 0.0139,
|
| 1255 |
+
"step": 3560
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 57.28,
|
| 1259 |
+
"grad_norm": 0.20135080814361572,
|
| 1260 |
+
"learning_rate": 1.8916338485173823e-06,
|
| 1261 |
+
"loss": 0.0193,
|
| 1262 |
+
"step": 3580
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 57.6,
|
| 1266 |
+
"grad_norm": 0.0006721566896885633,
|
| 1267 |
+
"learning_rate": 1.7298519356574727e-06,
|
| 1268 |
+
"loss": 0.0203,
|
| 1269 |
+
"step": 3600
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 57.92,
|
| 1273 |
+
"grad_norm": 0.10799671709537506,
|
| 1274 |
+
"learning_rate": 1.5750531975648324e-06,
|
| 1275 |
+
"loss": 0.0212,
|
| 1276 |
+
"step": 3620
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 58.24,
|
| 1280 |
+
"grad_norm": 0.0010109569411724806,
|
| 1281 |
+
"learning_rate": 1.4272840880324934e-06,
|
| 1282 |
+
"loss": 0.0173,
|
| 1283 |
+
"step": 3640
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 58.56,
|
| 1287 |
+
"grad_norm": 0.0008448906592093408,
|
| 1288 |
+
"learning_rate": 1.286588951321363e-06,
|
| 1289 |
+
"loss": 0.0139,
|
| 1290 |
+
"step": 3660
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 58.88,
|
| 1294 |
+
"grad_norm": 0.0010856656590476632,
|
| 1295 |
+
"learning_rate": 1.1530100088528867e-06,
|
| 1296 |
+
"loss": 0.0268,
|
| 1297 |
+
"step": 3680
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 59.2,
|
| 1301 |
+
"grad_norm": 0.23958024382591248,
|
| 1302 |
+
"learning_rate": 1.0265873465387516e-06,
|
| 1303 |
+
"loss": 0.0191,
|
| 1304 |
+
"step": 3700
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 59.52,
|
| 1308 |
+
"grad_norm": 0.20584586262702942,
|
| 1309 |
+
"learning_rate": 9.073589027514789e-07,
|
| 1310 |
+
"loss": 0.0168,
|
| 1311 |
+
"step": 3720
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 59.84,
|
| 1315 |
+
"grad_norm": 0.031580936163663864,
|
| 1316 |
+
"learning_rate": 7.953604569393841e-07,
|
| 1317 |
+
"loss": 0.0246,
|
| 1318 |
+
"step": 3740
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 60.16,
|
| 1322 |
+
"grad_norm": 0.14215555787086487,
|
| 1323 |
+
"learning_rate": 6.906256188895038e-07,
|
| 1324 |
+
"loss": 0.019,
|
| 1325 |
+
"step": 3760
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 60.48,
|
| 1329 |
+
"grad_norm": 0.0012006442993879318,
|
| 1330 |
+
"learning_rate": 5.931858186415756e-07,
|
| 1331 |
+
"loss": 0.0168,
|
| 1332 |
+
"step": 3780
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 60.8,
|
| 1336 |
+
"grad_norm": 0.0063135698437690735,
|
| 1337 |
+
"learning_rate": 5.03070297056149e-07,
|
| 1338 |
+
"loss": 0.0197,
|
| 1339 |
+
"step": 3800
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 61.12,
|
| 1343 |
+
"grad_norm": 0.07496818155050278,
|
| 1344 |
+
"learning_rate": 4.203060970396383e-07,
|
| 1345 |
+
"loss": 0.0207,
|
| 1346 |
+
"step": 3820
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 61.44,
|
| 1350 |
+
"grad_norm": 0.16551247239112854,
|
| 1351 |
+
"learning_rate": 3.4491805542899157e-07,
|
| 1352 |
+
"loss": 0.0224,
|
| 1353 |
+
"step": 3840
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 61.76,
|
| 1357 |
+
"grad_norm": 0.0008456969517283142,
|
| 1358 |
+
"learning_rate": 2.769287955383532e-07,
|
| 1359 |
+
"loss": 0.0151,
|
| 1360 |
+
"step": 3860
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 62.08,
|
| 1364 |
+
"grad_norm": 0.0008134017698466778,
|
| 1365 |
+
"learning_rate": 2.1635872037001626e-07,
|
| 1366 |
+
"loss": 0.0284,
|
| 1367 |
+
"step": 3880
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 62.4,
|
| 1371 |
+
"grad_norm": 0.18878595530986786,
|
| 1372 |
+
"learning_rate": 1.6322600649162356e-07,
|
| 1373 |
+
"loss": 0.0217,
|
| 1374 |
+
"step": 3900
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 62.72,
|
| 1378 |
+
"grad_norm": 0.0008310906123369932,
|
| 1379 |
+
"learning_rate": 1.1754659858156659e-07,
|
| 1380 |
+
"loss": 0.0103,
|
| 1381 |
+
"step": 3920
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 63.04,
|
| 1385 |
+
"grad_norm": 0.38621172308921814,
|
| 1386 |
+
"learning_rate": 7.933420464410201e-08,
|
| 1387 |
+
"loss": 0.0333,
|
| 1388 |
+
"step": 3940
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 63.36,
|
| 1392 |
+
"grad_norm": 0.016794312745332718,
|
| 1393 |
+
"learning_rate": 4.860029189569237e-08,
|
| 1394 |
+
"loss": 0.0231,
|
| 1395 |
+
"step": 3960
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 63.68,
|
| 1399 |
+
"grad_norm": 0.16253815591335297,
|
| 1400 |
+
"learning_rate": 2.535408332381417e-08,
|
| 1401 |
+
"loss": 0.0226,
|
| 1402 |
+
"step": 3980
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 64.0,
|
| 1406 |
+
"grad_norm": 0.2387680560350418,
|
| 1407 |
+
"learning_rate": 9.60255491919415e-09,
|
| 1408 |
+
"loss": 0.0218,
|
| 1409 |
+
"step": 4000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 64.32,
|
| 1413 |
+
"grad_norm": 0.16293394565582275,
|
| 1414 |
+
"learning_rate": 1.3504335823810722e-09,
|
| 1415 |
+
"loss": 0.0219,
|
| 1416 |
+
"step": 4020
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 64.48,
|
| 1420 |
+
"step": 4030,
|
| 1421 |
+
"total_flos": 2.3325606118844006e+17,
|
| 1422 |
+
"train_loss": 0.1495482857003993,
|
| 1423 |
+
"train_runtime": 6882.5617,
|
| 1424 |
+
"train_samples_per_second": 4.722,
|
| 1425 |
+
"train_steps_per_second": 0.586
|
| 1426 |
}
|
| 1427 |
],
|
| 1428 |
"logging_steps": 20,
|
| 1429 |
+
"max_steps": 4030,
|
| 1430 |
"num_input_tokens_seen": 0,
|
| 1431 |
+
"num_train_epochs": 65,
|
| 1432 |
"save_steps": 1000,
|
| 1433 |
"stateful_callbacks": {
|
| 1434 |
"TrainerControl": {
|
|
|
|
| 1442 |
"attributes": {}
|
| 1443 |
}
|
| 1444 |
},
|
| 1445 |
+
"total_flos": 2.3325606118844006e+17,
|
| 1446 |
"train_batch_size": 2,
|
| 1447 |
"trial_name": null,
|
| 1448 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85bd8392ac04e1d95cc4bb0c1bc398edcef3309a98ea6502b653c72aeb29939f
|
| 3 |
size 5240
|