Spaces:
Paused
Paused
Upload 16 files
Browse files- combined/adapter_config.json +16 -0
- combined/adapter_model.bin +3 -0
- combined/all_results.json +7 -0
- combined/checkpoint-1000/adapter_config.json +16 -0
- combined/checkpoint-1000/adapter_model.bin +3 -0
- combined/checkpoint-1000/finetuning_args.bin +3 -0
- combined/checkpoint-1000/optimizer.pt +3 -0
- combined/checkpoint-1000/rng_state.pth +3 -0
- combined/checkpoint-1000/scaler.pt +3 -0
- combined/checkpoint-1000/scheduler.pt +3 -0
- combined/checkpoint-1000/trainer_state.json +616 -0
- combined/checkpoint-1000/training_args.bin +3 -0
- combined/finetuning_args.bin +3 -0
- combined/train_results.json +7 -0
- combined/trainer_state.json +1033 -0
- combined/training_args.bin +3 -0
combined/adapter_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "THUDM/chatglm-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"lora_alpha": 32.0,
|
| 8 |
+
"lora_dropout": 0.1,
|
| 9 |
+
"modules_to_save": null,
|
| 10 |
+
"peft_type": "LORA",
|
| 11 |
+
"r": 8,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"query_key_value"
|
| 14 |
+
],
|
| 15 |
+
"task_type": "CAUSAL_LM"
|
| 16 |
+
}
|
combined/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a8ce5eac7fed9691eb91f1f21aad3a7dd0a50110a14b25a982d3410eeed62b9
|
| 3 |
+
size 14700825
|
combined/all_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 29.73,
|
| 3 |
+
"train_loss": 1.0390962482385693,
|
| 4 |
+
"train_runtime": 7981.5401,
|
| 5 |
+
"train_samples_per_second": 3.39,
|
| 6 |
+
"train_steps_per_second": 0.21
|
| 7 |
+
}
|
combined/checkpoint-1000/adapter_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "THUDM/chatglm-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"lora_alpha": 32.0,
|
| 8 |
+
"lora_dropout": 0.1,
|
| 9 |
+
"modules_to_save": null,
|
| 10 |
+
"peft_type": "LORA",
|
| 11 |
+
"r": 8,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"query_key_value"
|
| 14 |
+
],
|
| 15 |
+
"task_type": "CAUSAL_LM"
|
| 16 |
+
}
|
combined/checkpoint-1000/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebaa144e838f9c76dbd05121edf3b286f5a8dfa5557f62022e212a50a1a62020
|
| 3 |
+
size 14700825
|
combined/checkpoint-1000/finetuning_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0d21be8c105704d26792b0b5ff588414e9de8cdb4d00194854d7cec1b515a17
|
| 3 |
+
size 895
|
combined/checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c22adb6c5f28409a09e7a27846d82aef020ca654ee3e6e8ffd098f413150357
|
| 3 |
+
size 29407173
|
combined/checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9056600d69ce9b551ea75b28276fdb0d3874dd189f1d13daa04badbeb22eac78
|
| 3 |
+
size 14575
|
combined/checkpoint-1000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:309ab98ab5bd8db4af4aa2487c1eaff20632ec9a829c1243494a63392707c335
|
| 3 |
+
size 557
|
combined/checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0fd446562e7913451026214d0ba8e90c5e67d0e8e1d8a96b6d79282158a45bb8
|
| 3 |
+
size 627
|
combined/checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 17.699115044247787,
|
| 5 |
+
"global_step": 1000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.18,
|
| 12 |
+
"learning_rate": 0.000999929189777604,
|
| 13 |
+
"loss": 4.7362,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.35,
|
| 18 |
+
"learning_rate": 0.0009996844394494005,
|
| 19 |
+
"loss": 4.2612,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.53,
|
| 24 |
+
"learning_rate": 0.0009992649603785806,
|
| 25 |
+
"loss": 4.2216,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.71,
|
| 30 |
+
"learning_rate": 0.0009986708992478202,
|
| 31 |
+
"loss": 4.1602,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.88,
|
| 36 |
+
"learning_rate": 0.0009979024637873308,
|
| 37 |
+
"loss": 4.0982,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.06,
|
| 42 |
+
"learning_rate": 0.0009969599227022219,
|
| 43 |
+
"loss": 4.0312,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.24,
|
| 48 |
+
"learning_rate": 0.000995843605578539,
|
| 49 |
+
"loss": 3.7638,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.42,
|
| 54 |
+
"learning_rate": 0.0009945539027680166,
|
| 55 |
+
"loss": 3.7872,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.59,
|
| 60 |
+
"learning_rate": 0.0009930912652515787,
|
| 61 |
+
"loss": 3.7943,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.77,
|
| 66 |
+
"learning_rate": 0.0009914562044816423,
|
| 67 |
+
"loss": 3.753,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.95,
|
| 72 |
+
"learning_rate": 0.0009900243933150229,
|
| 73 |
+
"loss": 3.7757,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.12,
|
| 78 |
+
"learning_rate": 0.000988080451875917,
|
| 79 |
+
"loss": 3.4788,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.3,
|
| 84 |
+
"learning_rate": 0.0009859658393553863,
|
| 85 |
+
"loss": 3.2514,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.48,
|
| 90 |
+
"learning_rate": 0.0009836812951872773,
|
| 91 |
+
"loss": 3.3447,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.65,
|
| 96 |
+
"learning_rate": 0.0009812276182268236,
|
| 97 |
+
"loss": 3.4213,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.83,
|
| 102 |
+
"learning_rate": 0.000978605666471306,
|
| 103 |
+
"loss": 3.3397,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.01,
|
| 108 |
+
"learning_rate": 0.0009758163567600282,
|
| 109 |
+
"loss": 3.3895,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.19,
|
| 114 |
+
"learning_rate": 0.0009728606644537177,
|
| 115 |
+
"loss": 2.8609,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 3.36,
|
| 120 |
+
"learning_rate": 0.0009697396230934648,
|
| 121 |
+
"loss": 2.775,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.54,
|
| 126 |
+
"learning_rate": 0.0009664543240393149,
|
| 127 |
+
"loss": 2.9453,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.72,
|
| 132 |
+
"learning_rate": 0.0009630059160886439,
|
| 133 |
+
"loss": 2.9014,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.89,
|
| 138 |
+
"learning_rate": 0.0009593956050744492,
|
| 139 |
+
"loss": 3.0505,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 4.07,
|
| 144 |
+
"learning_rate": 0.0009556246534436953,
|
| 145 |
+
"loss": 2.6717,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 4.25,
|
| 150 |
+
"learning_rate": 0.0009516943798158648,
|
| 151 |
+
"loss": 2.5435,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 4.42,
|
| 156 |
+
"learning_rate": 0.0009476061585218652,
|
| 157 |
+
"loss": 2.4999,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.6,
|
| 162 |
+
"learning_rate": 0.0009433614191234554,
|
| 163 |
+
"loss": 2.4868,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.78,
|
| 168 |
+
"learning_rate": 0.0009389616459133597,
|
| 169 |
+
"loss": 2.4924,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.96,
|
| 174 |
+
"learning_rate": 0.0009344083773962423,
|
| 175 |
+
"loss": 2.6032,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 5.13,
|
| 180 |
+
"learning_rate": 0.0009297032057507264,
|
| 181 |
+
"loss": 2.1659,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 5.31,
|
| 186 |
+
"learning_rate": 0.0009248477762726437,
|
| 187 |
+
"loss": 2.222,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 5.49,
|
| 192 |
+
"learning_rate": 0.0009198437867997105,
|
| 193 |
+
"loss": 2.0641,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 5.66,
|
| 198 |
+
"learning_rate": 0.0009146929871178311,
|
| 199 |
+
"loss": 2.2005,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.84,
|
| 204 |
+
"learning_rate": 0.0009093971783492354,
|
| 205 |
+
"loss": 2.2747,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 6.02,
|
| 210 |
+
"learning_rate": 0.000903958212322666,
|
| 211 |
+
"loss": 2.2078,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 6.19,
|
| 216 |
+
"learning_rate": 0.0008983779909258347,
|
| 217 |
+
"loss": 1.7158,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 6.37,
|
| 222 |
+
"learning_rate": 0.0008926584654403724,
|
| 223 |
+
"loss": 1.7397,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 6.55,
|
| 228 |
+
"learning_rate": 0.0008868016358595103,
|
| 229 |
+
"loss": 2.0346,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 6.73,
|
| 234 |
+
"learning_rate": 0.0008808095501887243,
|
| 235 |
+
"loss": 1.8981,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.9,
|
| 240 |
+
"learning_rate": 0.0008746843037295936,
|
| 241 |
+
"loss": 1.9037,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 7.08,
|
| 246 |
+
"learning_rate": 0.0008684280383471181,
|
| 247 |
+
"loss": 1.7414,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 7.26,
|
| 252 |
+
"learning_rate": 0.0008620429417207562,
|
| 253 |
+
"loss": 1.5215,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 7.43,
|
| 258 |
+
"learning_rate": 0.0008555312465794402,
|
| 259 |
+
"loss": 1.5522,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 7.61,
|
| 264 |
+
"learning_rate": 0.0008488952299208401,
|
| 265 |
+
"loss": 1.5898,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 7.79,
|
| 270 |
+
"learning_rate": 0.0008421372122151469,
|
| 271 |
+
"loss": 1.692,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.96,
|
| 276 |
+
"learning_rate": 0.0008352595565936554,
|
| 277 |
+
"loss": 1.6656,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 8.14,
|
| 282 |
+
"learning_rate": 0.0008282646680224282,
|
| 283 |
+
"loss": 1.3681,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 8.32,
|
| 288 |
+
"learning_rate": 0.0008211549924613316,
|
| 289 |
+
"loss": 1.323,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 8.5,
|
| 294 |
+
"learning_rate": 0.0008139330160087374,
|
| 295 |
+
"loss": 1.301,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 8.67,
|
| 300 |
+
"learning_rate": 0.0008066012640321876,
|
| 301 |
+
"loss": 1.4134,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 8.85,
|
| 306 |
+
"learning_rate": 0.0007991623002853296,
|
| 307 |
+
"loss": 1.4661,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 9.03,
|
| 312 |
+
"learning_rate": 0.0007916187260114262,
|
| 313 |
+
"loss": 1.4292,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 9.2,
|
| 318 |
+
"learning_rate": 0.0007839731790337593,
|
| 319 |
+
"loss": 0.8971,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 9.38,
|
| 324 |
+
"learning_rate": 0.0007762283328332388,
|
| 325 |
+
"loss": 1.2906,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 9.56,
|
| 330 |
+
"learning_rate": 0.000768386895613546,
|
| 331 |
+
"loss": 1.2389,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 9.73,
|
| 336 |
+
"learning_rate": 0.0007604516093541333,
|
| 337 |
+
"loss": 1.2767,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 9.91,
|
| 342 |
+
"learning_rate": 0.0007524252488514144,
|
| 343 |
+
"loss": 1.1868,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 10.09,
|
| 348 |
+
"learning_rate": 0.0007443106207484776,
|
| 349 |
+
"loss": 1.0139,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 10.27,
|
| 354 |
+
"learning_rate": 0.0007361105625536646,
|
| 355 |
+
"loss": 1.021,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 10.44,
|
| 360 |
+
"learning_rate": 0.0007278279416483561,
|
| 361 |
+
"loss": 0.9439,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 10.62,
|
| 366 |
+
"learning_rate": 0.0007194656542843102,
|
| 367 |
+
"loss": 0.9772,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 10.8,
|
| 372 |
+
"learning_rate": 0.0007110266245709072,
|
| 373 |
+
"loss": 1.0697,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 10.97,
|
| 378 |
+
"learning_rate": 0.0007025138034526517,
|
| 379 |
+
"loss": 1.1471,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 11.15,
|
| 384 |
+
"learning_rate": 0.0006939301676772927,
|
| 385 |
+
"loss": 0.8037,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 11.33,
|
| 390 |
+
"learning_rate": 0.0006852787187549182,
|
| 391 |
+
"loss": 0.8488,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 11.5,
|
| 396 |
+
"learning_rate": 0.0006765624819083944,
|
| 397 |
+
"loss": 0.8035,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 11.68,
|
| 402 |
+
"learning_rate": 0.0006677845050155106,
|
| 403 |
+
"loss": 0.8825,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 11.86,
|
| 408 |
+
"learning_rate": 0.0006589478575432023,
|
| 409 |
+
"loss": 0.8639,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 12.04,
|
| 414 |
+
"learning_rate": 0.0006500556294742265,
|
| 415 |
+
"loss": 0.9365,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 12.21,
|
| 420 |
+
"learning_rate": 0.0006411109302266615,
|
| 421 |
+
"loss": 0.704,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 12.39,
|
| 426 |
+
"learning_rate": 0.0006321168875666118,
|
| 427 |
+
"loss": 0.7477,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 12.57,
|
| 432 |
+
"learning_rate": 0.0006230766465144965,
|
| 433 |
+
"loss": 0.6058,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 12.74,
|
| 438 |
+
"learning_rate": 0.0006139933682453035,
|
| 439 |
+
"loss": 0.809,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 12.92,
|
| 444 |
+
"learning_rate": 0.0006048702289831964,
|
| 445 |
+
"loss": 0.8475,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 13.1,
|
| 450 |
+
"learning_rate": 0.0005957104188908586,
|
| 451 |
+
"loss": 0.6654,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 13.27,
|
| 456 |
+
"learning_rate": 0.0005865171409539613,
|
| 457 |
+
"loss": 0.6185,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 13.45,
|
| 462 |
+
"learning_rate": 0.0005772936098611519,
|
| 463 |
+
"loss": 0.4949,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 13.63,
|
| 468 |
+
"learning_rate": 0.0005680430508799453,
|
| 469 |
+
"loss": 0.5332,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 13.81,
|
| 474 |
+
"learning_rate": 0.0005587686987289189,
|
| 475 |
+
"loss": 0.6924,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 13.98,
|
| 480 |
+
"learning_rate": 0.0005494737964466028,
|
| 481 |
+
"loss": 0.6796,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 14.16,
|
| 486 |
+
"learning_rate": 0.0005401615942574595,
|
| 487 |
+
"loss": 0.5147,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 14.34,
|
| 492 |
+
"learning_rate": 0.0005308353484353508,
|
| 493 |
+
"loss": 0.4158,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 14.51,
|
| 498 |
+
"learning_rate": 0.0005214983201648908,
|
| 499 |
+
"loss": 0.5493,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 14.69,
|
| 504 |
+
"learning_rate": 0.0005121537744010808,
|
| 505 |
+
"loss": 0.5508,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 14.87,
|
| 510 |
+
"learning_rate": 0.0005028049787276249,
|
| 511 |
+
"loss": 0.513,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 15.04,
|
| 516 |
+
"learning_rate": 0.0004934552022143279,
|
| 517 |
+
"loss": 0.5027,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 15.22,
|
| 522 |
+
"learning_rate": 0.00048410771427397304,
|
| 523 |
+
"loss": 0.3954,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 15.4,
|
| 528 |
+
"learning_rate": 0.0004747657835190795,
|
| 529 |
+
"loss": 0.4376,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 15.58,
|
| 534 |
+
"learning_rate": 0.0004654326766189404,
|
| 535 |
+
"loss": 0.4182,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 15.75,
|
| 540 |
+
"learning_rate": 0.0004561116571573388,
|
| 541 |
+
"loss": 0.4889,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 15.93,
|
| 546 |
+
"learning_rate": 0.00044680598449134434,
|
| 547 |
+
"loss": 0.3596,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 16.11,
|
| 552 |
+
"learning_rate": 0.00043751891261158783,
|
| 553 |
+
"loss": 0.3129,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 16.28,
|
| 558 |
+
"learning_rate": 0.0004282536890044104,
|
| 559 |
+
"loss": 0.3088,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 16.46,
|
| 564 |
+
"learning_rate": 0.0004190135535162894,
|
| 565 |
+
"loss": 0.3212,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 16.64,
|
| 570 |
+
"learning_rate": 0.00040980173722093496,
|
| 571 |
+
"loss": 0.4326,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 16.81,
|
| 576 |
+
"learning_rate": 0.00040062146128945275,
|
| 577 |
+
"loss": 0.3264,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 16.99,
|
| 582 |
+
"learning_rate": 0.0003914759358639719,
|
| 583 |
+
"loss": 0.3636,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 17.17,
|
| 588 |
+
"learning_rate": 0.0003823683589351284,
|
| 589 |
+
"loss": 0.2175,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 17.35,
|
| 594 |
+
"learning_rate": 0.0003733019152237965,
|
| 595 |
+
"loss": 0.208,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 17.52,
|
| 600 |
+
"learning_rate": 0.0003642797750674629,
|
| 601 |
+
"loss": 0.2717,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 17.7,
|
| 606 |
+
"learning_rate": 0.0003553050933116273,
|
| 607 |
+
"loss": 0.329,
|
| 608 |
+
"step": 1000
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 1680,
|
| 612 |
+
"num_train_epochs": 30,
|
| 613 |
+
"total_flos": 1.2761491789676544e+17,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
combined/checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5e57e44e7215d66f6f4fb7ed026c0ebedcf7be9b514db88f2ec922ec682c4f3
|
| 3 |
+
size 4027
|
combined/finetuning_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0d21be8c105704d26792b0b5ff588414e9de8cdb4d00194854d7cec1b515a17
|
| 3 |
+
size 895
|
combined/train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 29.73,
|
| 3 |
+
"train_loss": 1.0390962482385693,
|
| 4 |
+
"train_runtime": 7981.5401,
|
| 5 |
+
"train_samples_per_second": 3.39,
|
| 6 |
+
"train_steps_per_second": 0.21
|
| 7 |
+
}
|
combined/trainer_state.json
ADDED
|
@@ -0,0 +1,1033 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 29.734513274336283,
|
| 5 |
+
"global_step": 1680,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.18,
|
| 12 |
+
"learning_rate": 0.000999929189777604,
|
| 13 |
+
"loss": 4.7362,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.35,
|
| 18 |
+
"learning_rate": 0.0009996844394494005,
|
| 19 |
+
"loss": 4.2612,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.53,
|
| 24 |
+
"learning_rate": 0.0009992649603785806,
|
| 25 |
+
"loss": 4.2216,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.71,
|
| 30 |
+
"learning_rate": 0.0009986708992478202,
|
| 31 |
+
"loss": 4.1602,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.88,
|
| 36 |
+
"learning_rate": 0.0009979024637873308,
|
| 37 |
+
"loss": 4.0982,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.06,
|
| 42 |
+
"learning_rate": 0.0009969599227022219,
|
| 43 |
+
"loss": 4.0312,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.24,
|
| 48 |
+
"learning_rate": 0.000995843605578539,
|
| 49 |
+
"loss": 3.7638,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.42,
|
| 54 |
+
"learning_rate": 0.0009945539027680166,
|
| 55 |
+
"loss": 3.7872,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.59,
|
| 60 |
+
"learning_rate": 0.0009930912652515787,
|
| 61 |
+
"loss": 3.7943,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.77,
|
| 66 |
+
"learning_rate": 0.0009914562044816423,
|
| 67 |
+
"loss": 3.753,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.95,
|
| 72 |
+
"learning_rate": 0.0009900243933150229,
|
| 73 |
+
"loss": 3.7757,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.12,
|
| 78 |
+
"learning_rate": 0.000988080451875917,
|
| 79 |
+
"loss": 3.4788,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.3,
|
| 84 |
+
"learning_rate": 0.0009859658393553863,
|
| 85 |
+
"loss": 3.2514,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.48,
|
| 90 |
+
"learning_rate": 0.0009836812951872773,
|
| 91 |
+
"loss": 3.3447,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.65,
|
| 96 |
+
"learning_rate": 0.0009812276182268236,
|
| 97 |
+
"loss": 3.4213,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.83,
|
| 102 |
+
"learning_rate": 0.000978605666471306,
|
| 103 |
+
"loss": 3.3397,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.01,
|
| 108 |
+
"learning_rate": 0.0009758163567600282,
|
| 109 |
+
"loss": 3.3895,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.19,
|
| 114 |
+
"learning_rate": 0.0009728606644537177,
|
| 115 |
+
"loss": 2.8609,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 3.36,
|
| 120 |
+
"learning_rate": 0.0009697396230934648,
|
| 121 |
+
"loss": 2.775,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 3.54,
|
| 126 |
+
"learning_rate": 0.0009664543240393149,
|
| 127 |
+
"loss": 2.9453,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.72,
|
| 132 |
+
"learning_rate": 0.0009630059160886439,
|
| 133 |
+
"loss": 2.9014,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.89,
|
| 138 |
+
"learning_rate": 0.0009593956050744492,
|
| 139 |
+
"loss": 3.0505,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 4.07,
|
| 144 |
+
"learning_rate": 0.0009556246534436953,
|
| 145 |
+
"loss": 2.6717,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 4.25,
|
| 150 |
+
"learning_rate": 0.0009516943798158648,
|
| 151 |
+
"loss": 2.5435,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 4.42,
|
| 156 |
+
"learning_rate": 0.0009476061585218652,
|
| 157 |
+
"loss": 2.4999,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.6,
|
| 162 |
+
"learning_rate": 0.0009433614191234554,
|
| 163 |
+
"loss": 2.4868,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 4.78,
|
| 168 |
+
"learning_rate": 0.0009389616459133597,
|
| 169 |
+
"loss": 2.4924,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.96,
|
| 174 |
+
"learning_rate": 0.0009344083773962423,
|
| 175 |
+
"loss": 2.6032,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 5.13,
|
| 180 |
+
"learning_rate": 0.0009297032057507264,
|
| 181 |
+
"loss": 2.1659,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 5.31,
|
| 186 |
+
"learning_rate": 0.0009248477762726437,
|
| 187 |
+
"loss": 2.222,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 5.49,
|
| 192 |
+
"learning_rate": 0.0009198437867997105,
|
| 193 |
+
"loss": 2.0641,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 5.66,
|
| 198 |
+
"learning_rate": 0.0009146929871178311,
|
| 199 |
+
"loss": 2.2005,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.84,
|
| 204 |
+
"learning_rate": 0.0009093971783492354,
|
| 205 |
+
"loss": 2.2747,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 6.02,
|
| 210 |
+
"learning_rate": 0.000903958212322666,
|
| 211 |
+
"loss": 2.2078,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 6.19,
|
| 216 |
+
"learning_rate": 0.0008983779909258347,
|
| 217 |
+
"loss": 1.7158,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 6.37,
|
| 222 |
+
"learning_rate": 0.0008926584654403724,
|
| 223 |
+
"loss": 1.7397,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 6.55,
|
| 228 |
+
"learning_rate": 0.0008868016358595103,
|
| 229 |
+
"loss": 2.0346,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 6.73,
|
| 234 |
+
"learning_rate": 0.0008808095501887243,
|
| 235 |
+
"loss": 1.8981,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 6.9,
|
| 240 |
+
"learning_rate": 0.0008746843037295936,
|
| 241 |
+
"loss": 1.9037,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 7.08,
|
| 246 |
+
"learning_rate": 0.0008684280383471181,
|
| 247 |
+
"loss": 1.7414,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 7.26,
|
| 252 |
+
"learning_rate": 0.0008620429417207562,
|
| 253 |
+
"loss": 1.5215,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 7.43,
|
| 258 |
+
"learning_rate": 0.0008555312465794402,
|
| 259 |
+
"loss": 1.5522,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 7.61,
|
| 264 |
+
"learning_rate": 0.0008488952299208401,
|
| 265 |
+
"loss": 1.5898,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 7.79,
|
| 270 |
+
"learning_rate": 0.0008421372122151469,
|
| 271 |
+
"loss": 1.692,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 7.96,
|
| 276 |
+
"learning_rate": 0.0008352595565936554,
|
| 277 |
+
"loss": 1.6656,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 8.14,
|
| 282 |
+
"learning_rate": 0.0008282646680224282,
|
| 283 |
+
"loss": 1.3681,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 8.32,
|
| 288 |
+
"learning_rate": 0.0008211549924613316,
|
| 289 |
+
"loss": 1.323,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 8.5,
|
| 294 |
+
"learning_rate": 0.0008139330160087374,
|
| 295 |
+
"loss": 1.301,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 8.67,
|
| 300 |
+
"learning_rate": 0.0008066012640321876,
|
| 301 |
+
"loss": 1.4134,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 8.85,
|
| 306 |
+
"learning_rate": 0.0007991623002853296,
|
| 307 |
+
"loss": 1.4661,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 9.03,
|
| 312 |
+
"learning_rate": 0.0007916187260114262,
|
| 313 |
+
"loss": 1.4292,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 9.2,
|
| 318 |
+
"learning_rate": 0.0007839731790337593,
|
| 319 |
+
"loss": 0.8971,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 9.38,
|
| 324 |
+
"learning_rate": 0.0007762283328332388,
|
| 325 |
+
"loss": 1.2906,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 9.56,
|
| 330 |
+
"learning_rate": 0.000768386895613546,
|
| 331 |
+
"loss": 1.2389,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 9.73,
|
| 336 |
+
"learning_rate": 0.0007604516093541333,
|
| 337 |
+
"loss": 1.2767,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 9.91,
|
| 342 |
+
"learning_rate": 0.0007524252488514144,
|
| 343 |
+
"loss": 1.1868,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 10.09,
|
| 348 |
+
"learning_rate": 0.0007443106207484776,
|
| 349 |
+
"loss": 1.0139,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 10.27,
|
| 354 |
+
"learning_rate": 0.0007361105625536646,
|
| 355 |
+
"loss": 1.021,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 10.44,
|
| 360 |
+
"learning_rate": 0.0007278279416483561,
|
| 361 |
+
"loss": 0.9439,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 10.62,
|
| 366 |
+
"learning_rate": 0.0007194656542843102,
|
| 367 |
+
"loss": 0.9772,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 10.8,
|
| 372 |
+
"learning_rate": 0.0007110266245709072,
|
| 373 |
+
"loss": 1.0697,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 10.97,
|
| 378 |
+
"learning_rate": 0.0007025138034526517,
|
| 379 |
+
"loss": 1.1471,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 11.15,
|
| 384 |
+
"learning_rate": 0.0006939301676772927,
|
| 385 |
+
"loss": 0.8037,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 11.33,
|
| 390 |
+
"learning_rate": 0.0006852787187549182,
|
| 391 |
+
"loss": 0.8488,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 11.5,
|
| 396 |
+
"learning_rate": 0.0006765624819083944,
|
| 397 |
+
"loss": 0.8035,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 11.68,
|
| 402 |
+
"learning_rate": 0.0006677845050155106,
|
| 403 |
+
"loss": 0.8825,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 11.86,
|
| 408 |
+
"learning_rate": 0.0006589478575432023,
|
| 409 |
+
"loss": 0.8639,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 12.04,
|
| 414 |
+
"learning_rate": 0.0006500556294742265,
|
| 415 |
+
"loss": 0.9365,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 12.21,
|
| 420 |
+
"learning_rate": 0.0006411109302266615,
|
| 421 |
+
"loss": 0.704,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 12.39,
|
| 426 |
+
"learning_rate": 0.0006321168875666118,
|
| 427 |
+
"loss": 0.7477,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 12.57,
|
| 432 |
+
"learning_rate": 0.0006230766465144965,
|
| 433 |
+
"loss": 0.6058,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 12.74,
|
| 438 |
+
"learning_rate": 0.0006139933682453035,
|
| 439 |
+
"loss": 0.809,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 12.92,
|
| 444 |
+
"learning_rate": 0.0006048702289831964,
|
| 445 |
+
"loss": 0.8475,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 13.1,
|
| 450 |
+
"learning_rate": 0.0005957104188908586,
|
| 451 |
+
"loss": 0.6654,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 13.27,
|
| 456 |
+
"learning_rate": 0.0005865171409539613,
|
| 457 |
+
"loss": 0.6185,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 13.45,
|
| 462 |
+
"learning_rate": 0.0005772936098611519,
|
| 463 |
+
"loss": 0.4949,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 13.63,
|
| 468 |
+
"learning_rate": 0.0005680430508799453,
|
| 469 |
+
"loss": 0.5332,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 13.81,
|
| 474 |
+
"learning_rate": 0.0005587686987289189,
|
| 475 |
+
"loss": 0.6924,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 13.98,
|
| 480 |
+
"learning_rate": 0.0005494737964466028,
|
| 481 |
+
"loss": 0.6796,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 14.16,
|
| 486 |
+
"learning_rate": 0.0005401615942574595,
|
| 487 |
+
"loss": 0.5147,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 14.34,
|
| 492 |
+
"learning_rate": 0.0005308353484353508,
|
| 493 |
+
"loss": 0.4158,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 14.51,
|
| 498 |
+
"learning_rate": 0.0005214983201648908,
|
| 499 |
+
"loss": 0.5493,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 14.69,
|
| 504 |
+
"learning_rate": 0.0005121537744010808,
|
| 505 |
+
"loss": 0.5508,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 14.87,
|
| 510 |
+
"learning_rate": 0.0005028049787276249,
|
| 511 |
+
"loss": 0.513,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 15.04,
|
| 516 |
+
"learning_rate": 0.0004934552022143279,
|
| 517 |
+
"loss": 0.5027,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 15.22,
|
| 522 |
+
"learning_rate": 0.00048410771427397304,
|
| 523 |
+
"loss": 0.3954,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 15.4,
|
| 528 |
+
"learning_rate": 0.0004747657835190795,
|
| 529 |
+
"loss": 0.4376,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 15.58,
|
| 534 |
+
"learning_rate": 0.0004654326766189404,
|
| 535 |
+
"loss": 0.4182,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 15.75,
|
| 540 |
+
"learning_rate": 0.0004561116571573388,
|
| 541 |
+
"loss": 0.4889,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 15.93,
|
| 546 |
+
"learning_rate": 0.00044680598449134434,
|
| 547 |
+
"loss": 0.3596,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 16.11,
|
| 552 |
+
"learning_rate": 0.00043751891261158783,
|
| 553 |
+
"loss": 0.3129,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 16.28,
|
| 558 |
+
"learning_rate": 0.0004282536890044104,
|
| 559 |
+
"loss": 0.3088,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 16.46,
|
| 564 |
+
"learning_rate": 0.0004190135535162894,
|
| 565 |
+
"loss": 0.3212,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 16.64,
|
| 570 |
+
"learning_rate": 0.00040980173722093496,
|
| 571 |
+
"loss": 0.4326,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 16.81,
|
| 576 |
+
"learning_rate": 0.00040062146128945275,
|
| 577 |
+
"loss": 0.3264,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 16.99,
|
| 582 |
+
"learning_rate": 0.0003914759358639719,
|
| 583 |
+
"loss": 0.3636,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 17.17,
|
| 588 |
+
"learning_rate": 0.0003823683589351284,
|
| 589 |
+
"loss": 0.2175,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 17.35,
|
| 594 |
+
"learning_rate": 0.0003733019152237965,
|
| 595 |
+
"loss": 0.208,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 17.52,
|
| 600 |
+
"learning_rate": 0.0003642797750674629,
|
| 601 |
+
"loss": 0.2717,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 17.7,
|
| 606 |
+
"learning_rate": 0.0003553050933116273,
|
| 607 |
+
"loss": 0.329,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 17.88,
|
| 612 |
+
"learning_rate": 0.0003463810082066231,
|
| 613 |
+
"loss": 0.33,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 18.05,
|
| 618 |
+
"learning_rate": 0.00033751064031023887,
|
| 619 |
+
"loss": 0.2679,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 18.23,
|
| 624 |
+
"learning_rate": 0.0003286970913965275,
|
| 625 |
+
"loss": 0.224,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 18.41,
|
| 630 |
+
"learning_rate": 0.00031994344337118454,
|
| 631 |
+
"loss": 0.2048,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 18.58,
|
| 636 |
+
"learning_rate": 0.0003112527571938717,
|
| 637 |
+
"loss": 0.2078,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 18.76,
|
| 642 |
+
"learning_rate": 0.00030262807180786645,
|
| 643 |
+
"loss": 0.2303,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 18.94,
|
| 648 |
+
"learning_rate": 0.0002940724030774104,
|
| 649 |
+
"loss": 0.2142,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 19.12,
|
| 654 |
+
"learning_rate": 0.0002855887427331267,
|
| 655 |
+
"loss": 0.2067,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 19.29,
|
| 660 |
+
"learning_rate": 0.0002771800573258781,
|
| 661 |
+
"loss": 0.1458,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 19.47,
|
| 666 |
+
"learning_rate": 0.0002688492871894276,
|
| 667 |
+
"loss": 0.1774,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 19.65,
|
| 672 |
+
"learning_rate": 0.0002605993454122687,
|
| 673 |
+
"loss": 0.1741,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 19.82,
|
| 678 |
+
"learning_rate": 0.0002524331168189817,
|
| 679 |
+
"loss": 0.1883,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 20.0,
|
| 684 |
+
"learning_rate": 0.00024435345696147403,
|
| 685 |
+
"loss": 0.1689,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 20.18,
|
| 690 |
+
"learning_rate": 0.00023636319112045495,
|
| 691 |
+
"loss": 0.1238,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 20.35,
|
| 696 |
+
"learning_rate": 0.00022846511331749792,
|
| 697 |
+
"loss": 0.1372,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 20.53,
|
| 702 |
+
"learning_rate": 0.00022066198533803106,
|
| 703 |
+
"loss": 0.1298,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 20.71,
|
| 708 |
+
"learning_rate": 0.00021295653576560165,
|
| 709 |
+
"loss": 0.1445,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 20.88,
|
| 714 |
+
"learning_rate": 0.00020535145902774978,
|
| 715 |
+
"loss": 0.1454,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 21.06,
|
| 720 |
+
"learning_rate": 0.00019784941445382642,
|
| 721 |
+
"loss": 0.1417,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 21.24,
|
| 726 |
+
"learning_rate": 0.00019045302534508295,
|
| 727 |
+
"loss": 0.1099,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 21.42,
|
| 732 |
+
"learning_rate": 0.0001831648780573612,
|
| 733 |
+
"loss": 0.0922,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 21.59,
|
| 738 |
+
"learning_rate": 0.00017598752109670014,
|
| 739 |
+
"loss": 0.1258,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 21.77,
|
| 744 |
+
"learning_rate": 0.00016892346422817944,
|
| 745 |
+
"loss": 0.123,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 21.95,
|
| 750 |
+
"learning_rate": 0.00016197517759830948,
|
| 751 |
+
"loss": 0.1047,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 22.12,
|
| 756 |
+
"learning_rate": 0.00015514509087127522,
|
| 757 |
+
"loss": 0.0991,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 22.3,
|
| 762 |
+
"learning_rate": 0.00014843559237933475,
|
| 763 |
+
"loss": 0.1057,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 22.48,
|
| 768 |
+
"learning_rate": 0.00014184902828767287,
|
| 769 |
+
"loss": 0.0796,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 22.65,
|
| 774 |
+
"learning_rate": 0.0001353877017739974,
|
| 775 |
+
"loss": 0.0747,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 22.83,
|
| 780 |
+
"learning_rate": 0.00012905387222316822,
|
| 781 |
+
"loss": 0.0807,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 23.01,
|
| 786 |
+
"learning_rate": 0.00012284975443714076,
|
| 787 |
+
"loss": 0.1032,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 23.19,
|
| 792 |
+
"learning_rate": 0.00011677751786049661,
|
| 793 |
+
"loss": 0.0743,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 23.36,
|
| 798 |
+
"learning_rate": 0.0001108392858218371,
|
| 799 |
+
"loss": 0.0823,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 23.54,
|
| 804 |
+
"learning_rate": 0.00010503713479130123,
|
| 805 |
+
"loss": 0.0758,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 23.72,
|
| 810 |
+
"learning_rate": 9.937309365446973e-05,
|
| 811 |
+
"loss": 0.0642,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 23.89,
|
| 816 |
+
"learning_rate": 9.384914300290748e-05,
|
| 817 |
+
"loss": 0.0791,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 24.07,
|
| 822 |
+
"learning_rate": 8.846721444159461e-05,
|
| 823 |
+
"loss": 0.0595,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 24.25,
|
| 828 |
+
"learning_rate": 8.32291899134856e-05,
|
| 829 |
+
"loss": 0.0676,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 24.42,
|
| 834 |
+
"learning_rate": 7.813690104143555e-05,
|
| 835 |
+
"loss": 0.0611,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 24.6,
|
| 840 |
+
"learning_rate": 7.319212848772117e-05,
|
| 841 |
+
"loss": 0.0642,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 24.78,
|
| 846 |
+
"learning_rate": 6.83966013313826e-05,
|
| 847 |
+
"loss": 0.0602,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 24.96,
|
| 852 |
+
"learning_rate": 6.375199646360142e-05,
|
| 853 |
+
"loss": 0.066,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 25.13,
|
| 858 |
+
"learning_rate": 5.9259938001329336e-05,
|
| 859 |
+
"loss": 0.0581,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 25.31,
|
| 864 |
+
"learning_rate": 5.492199671936915e-05,
|
| 865 |
+
"loss": 0.0527,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 25.49,
|
| 870 |
+
"learning_rate": 5.073968950110941e-05,
|
| 871 |
+
"loss": 0.0634,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 25.66,
|
| 876 |
+
"learning_rate": 4.6714478808103314e-05,
|
| 877 |
+
"loss": 0.0545,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 25.84,
|
| 882 |
+
"learning_rate": 4.2847772168678246e-05,
|
| 883 |
+
"loss": 0.0549,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 26.02,
|
| 888 |
+
"learning_rate": 3.9140921685753064e-05,
|
| 889 |
+
"loss": 0.0529,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 26.19,
|
| 894 |
+
"learning_rate": 3.559522356403788e-05,
|
| 895 |
+
"loss": 0.0558,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 26.37,
|
| 900 |
+
"learning_rate": 3.221191765677928e-05,
|
| 901 |
+
"loss": 0.0529,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 26.55,
|
| 906 |
+
"learning_rate": 2.8992187032210516e-05,
|
| 907 |
+
"loss": 0.0552,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 26.73,
|
| 912 |
+
"learning_rate": 2.5937157559859093e-05,
|
| 913 |
+
"loss": 0.0499,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 26.9,
|
| 918 |
+
"learning_rate": 2.3047897516853767e-05,
|
| 919 |
+
"loss": 0.0451,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 27.08,
|
| 924 |
+
"learning_rate": 2.032541721437209e-05,
|
| 925 |
+
"loss": 0.0511,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 27.26,
|
| 930 |
+
"learning_rate": 1.777066864435628e-05,
|
| 931 |
+
"loss": 0.0491,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 27.43,
|
| 936 |
+
"learning_rate": 1.538454514662285e-05,
|
| 937 |
+
"loss": 0.045,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 27.61,
|
| 942 |
+
"learning_rate": 1.3167881096480372e-05,
|
| 943 |
+
"loss": 0.0594,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 27.79,
|
| 948 |
+
"learning_rate": 1.1121451612967626e-05,
|
| 949 |
+
"loss": 0.0456,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 27.96,
|
| 954 |
+
"learning_rate": 9.245972287810544e-06,
|
| 955 |
+
"loss": 0.0444,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 28.14,
|
| 960 |
+
"learning_rate": 7.542098935195918e-06,
|
| 961 |
+
"loss": 0.0615,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 28.32,
|
| 966 |
+
"learning_rate": 6.010427362447224e-06,
|
| 967 |
+
"loss": 0.0456,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 28.5,
|
| 972 |
+
"learning_rate": 4.6514931616840066e-06,
|
| 973 |
+
"loss": 0.0445,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 28.67,
|
| 978 |
+
"learning_rate": 3.4657715225368535e-06,
|
| 979 |
+
"loss": 0.0494,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 28.85,
|
| 984 |
+
"learning_rate": 2.453677065984039e-06,
|
| 985 |
+
"loss": 0.0411,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 29.03,
|
| 990 |
+
"learning_rate": 1.615563699367495e-06,
|
| 991 |
+
"loss": 0.0558,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 29.2,
|
| 996 |
+
"learning_rate": 9.517244926393609e-07,
|
| 997 |
+
"loss": 0.0457,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 29.38,
|
| 1002 |
+
"learning_rate": 4.623915758816266e-07,
|
| 1003 |
+
"loss": 0.0524,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 29.56,
|
| 1008 |
+
"learning_rate": 1.4773605813522827e-07,
|
| 1009 |
+
"loss": 0.0454,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 29.73,
|
| 1014 |
+
"learning_rate": 7.867967567354306e-09,
|
| 1015 |
+
"loss": 0.055,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 29.73,
|
| 1020 |
+
"step": 1680,
|
| 1021 |
+
"total_flos": 2.133691717066752e+17,
|
| 1022 |
+
"train_loss": 1.0390962482385693,
|
| 1023 |
+
"train_runtime": 7981.5401,
|
| 1024 |
+
"train_samples_per_second": 3.39,
|
| 1025 |
+
"train_steps_per_second": 0.21
|
| 1026 |
+
}
|
| 1027 |
+
],
|
| 1028 |
+
"max_steps": 1680,
|
| 1029 |
+
"num_train_epochs": 30,
|
| 1030 |
+
"total_flos": 2.133691717066752e+17,
|
| 1031 |
+
"trial_name": null,
|
| 1032 |
+
"trial_params": null
|
| 1033 |
+
}
|
combined/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5e57e44e7215d66f6f4fb7ed026c0ebedcf7be9b514db88f2ec922ec682c4f3
|
| 3 |
+
size 4027
|