Upload folder using huggingface_hub
Browse files- 20250501-1421/checkpoint-111/optimizer.pt +3 -0
- 20250501-1421/checkpoint-111/pytorch_model.bin +3 -0
- 20250501-1421/checkpoint-111/rng_state.pth +3 -0
- 20250501-1421/checkpoint-111/scheduler.pt +3 -0
- 20250501-1421/checkpoint-111/trainer_state.json +811 -0
- 20250501-1421/checkpoint-111/training_args.bin +3 -0
- 20250501-1421/runs/May01_14-21-16_d55355693cce/events.out.tfevents.1746109277.d55355693cce.100.0 +3 -0
20250501-1421/checkpoint-111/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5711efa0fa29e3b31af8605d63f8159b42b8a498d41a4d4bd1c5b1367a42a3e1
|
| 3 |
+
size 33824762
|
20250501-1421/checkpoint-111/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4fb6ff2c422748c278df1a02d53926882dc259ab62a1aafcdaed9b41b69dcb17
|
| 3 |
+
size 368596590
|
20250501-1421/checkpoint-111/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:238523785dc8a6bd63d8f51ae72844d04988b9978d2050f234191a39fe7b1141
|
| 3 |
+
size 14244
|
20250501-1421/checkpoint-111/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e42e66dcf545600cbaa1c353643c7beb3f3d693939da1e7ad003b79b8ca066d
|
| 3 |
+
size 1064
|
20250501-1421/checkpoint-111/trainer_state.json
ADDED
|
@@ -0,0 +1,811 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9955156950672646,
|
| 6 |
+
"eval_steps": 200,
|
| 7 |
+
"global_step": 111,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.008968609865470852,
|
| 14 |
+
"grad_norm": 2.960822105407715,
|
| 15 |
+
"learning_rate": 0.0001,
|
| 16 |
+
"loss": 2.8706,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.017937219730941704,
|
| 21 |
+
"grad_norm": 0.7307156324386597,
|
| 22 |
+
"learning_rate": 9.90990990990991e-05,
|
| 23 |
+
"loss": 1.3,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.026905829596412557,
|
| 28 |
+
"grad_norm": 0.3430079519748688,
|
| 29 |
+
"learning_rate": 9.81981981981982e-05,
|
| 30 |
+
"loss": 1.2556,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03587443946188341,
|
| 35 |
+
"grad_norm": 3.3454463481903076,
|
| 36 |
+
"learning_rate": 9.729729729729731e-05,
|
| 37 |
+
"loss": 5.9789,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04484304932735426,
|
| 42 |
+
"grad_norm": 0.40624433755874634,
|
| 43 |
+
"learning_rate": 9.639639639639641e-05,
|
| 44 |
+
"loss": 0.9036,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.053811659192825115,
|
| 49 |
+
"grad_norm": 0.52606600522995,
|
| 50 |
+
"learning_rate": 9.54954954954955e-05,
|
| 51 |
+
"loss": 2.2173,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06278026905829596,
|
| 56 |
+
"grad_norm": 1.4110300540924072,
|
| 57 |
+
"learning_rate": 9.45945945945946e-05,
|
| 58 |
+
"loss": 1.0893,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07174887892376682,
|
| 63 |
+
"grad_norm": 0.46715906262397766,
|
| 64 |
+
"learning_rate": 9.36936936936937e-05,
|
| 65 |
+
"loss": 2.6939,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08071748878923767,
|
| 70 |
+
"grad_norm": 0.6949467658996582,
|
| 71 |
+
"learning_rate": 9.279279279279279e-05,
|
| 72 |
+
"loss": 1.0149,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.08968609865470852,
|
| 77 |
+
"grad_norm": 0.488537460565567,
|
| 78 |
+
"learning_rate": 9.18918918918919e-05,
|
| 79 |
+
"loss": 1.3068,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09865470852017937,
|
| 84 |
+
"grad_norm": 2.622659206390381,
|
| 85 |
+
"learning_rate": 9.0990990990991e-05,
|
| 86 |
+
"loss": 6.1029,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.10762331838565023,
|
| 91 |
+
"grad_norm": 1.1801209449768066,
|
| 92 |
+
"learning_rate": 9.009009009009009e-05,
|
| 93 |
+
"loss": 1.7577,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.11659192825112108,
|
| 98 |
+
"grad_norm": 2.3658924102783203,
|
| 99 |
+
"learning_rate": 8.918918918918919e-05,
|
| 100 |
+
"loss": 0.7523,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.12556053811659193,
|
| 105 |
+
"grad_norm": 2.378786325454712,
|
| 106 |
+
"learning_rate": 8.82882882882883e-05,
|
| 107 |
+
"loss": 5.655,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.13452914798206278,
|
| 112 |
+
"grad_norm": 1.1980235576629639,
|
| 113 |
+
"learning_rate": 8.738738738738738e-05,
|
| 114 |
+
"loss": 1.9926,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.14349775784753363,
|
| 119 |
+
"grad_norm": 1.1755034923553467,
|
| 120 |
+
"learning_rate": 8.64864864864865e-05,
|
| 121 |
+
"loss": 2.6233,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.15246636771300448,
|
| 126 |
+
"grad_norm": 0.9538990259170532,
|
| 127 |
+
"learning_rate": 8.55855855855856e-05,
|
| 128 |
+
"loss": 0.936,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.16143497757847533,
|
| 133 |
+
"grad_norm": 10.638568878173828,
|
| 134 |
+
"learning_rate": 8.468468468468469e-05,
|
| 135 |
+
"loss": 18.6878,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.17040358744394618,
|
| 140 |
+
"grad_norm": 1.5819883346557617,
|
| 141 |
+
"learning_rate": 8.378378378378379e-05,
|
| 142 |
+
"loss": 0.6895,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.17937219730941703,
|
| 147 |
+
"grad_norm": 1.8580043315887451,
|
| 148 |
+
"learning_rate": 8.288288288288289e-05,
|
| 149 |
+
"loss": 0.9961,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.18834080717488788,
|
| 154 |
+
"grad_norm": 4.262594223022461,
|
| 155 |
+
"learning_rate": 8.198198198198198e-05,
|
| 156 |
+
"loss": 1.0639,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.19730941704035873,
|
| 161 |
+
"grad_norm": 10.581626892089844,
|
| 162 |
+
"learning_rate": 8.108108108108109e-05,
|
| 163 |
+
"loss": 11.8574,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2062780269058296,
|
| 168 |
+
"grad_norm": 1.8223183155059814,
|
| 169 |
+
"learning_rate": 8.018018018018019e-05,
|
| 170 |
+
"loss": 2.4031,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.21524663677130046,
|
| 175 |
+
"grad_norm": 0.7646092772483826,
|
| 176 |
+
"learning_rate": 7.927927927927928e-05,
|
| 177 |
+
"loss": 1.0303,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.2242152466367713,
|
| 182 |
+
"grad_norm": 1.9194566011428833,
|
| 183 |
+
"learning_rate": 7.837837837837838e-05,
|
| 184 |
+
"loss": 0.9733,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.23318385650224216,
|
| 189 |
+
"grad_norm": 1.6878315210342407,
|
| 190 |
+
"learning_rate": 7.747747747747748e-05,
|
| 191 |
+
"loss": 2.3391,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.242152466367713,
|
| 196 |
+
"grad_norm": 1.3996132612228394,
|
| 197 |
+
"learning_rate": 7.657657657657657e-05,
|
| 198 |
+
"loss": 0.5793,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.25112107623318386,
|
| 203 |
+
"grad_norm": 2.891868829727173,
|
| 204 |
+
"learning_rate": 7.567567567567568e-05,
|
| 205 |
+
"loss": 2.4065,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.2600896860986547,
|
| 210 |
+
"grad_norm": 1.5177550315856934,
|
| 211 |
+
"learning_rate": 7.477477477477479e-05,
|
| 212 |
+
"loss": 0.8909,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.26905829596412556,
|
| 217 |
+
"grad_norm": 4.505558967590332,
|
| 218 |
+
"learning_rate": 7.387387387387387e-05,
|
| 219 |
+
"loss": 5.6894,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.27802690582959644,
|
| 224 |
+
"grad_norm": 3.0119218826293945,
|
| 225 |
+
"learning_rate": 7.297297297297297e-05,
|
| 226 |
+
"loss": 0.6294,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.28699551569506726,
|
| 231 |
+
"grad_norm": 3.9027931690216064,
|
| 232 |
+
"learning_rate": 7.207207207207208e-05,
|
| 233 |
+
"loss": 0.8345,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.29596412556053814,
|
| 238 |
+
"grad_norm": 5.136641502380371,
|
| 239 |
+
"learning_rate": 7.117117117117116e-05,
|
| 240 |
+
"loss": 2.685,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.30493273542600896,
|
| 245 |
+
"grad_norm": 2.472689390182495,
|
| 246 |
+
"learning_rate": 7.027027027027028e-05,
|
| 247 |
+
"loss": 1.1706,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.31390134529147984,
|
| 252 |
+
"grad_norm": 7.271293640136719,
|
| 253 |
+
"learning_rate": 6.936936936936938e-05,
|
| 254 |
+
"loss": 5.7874,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.32286995515695066,
|
| 259 |
+
"grad_norm": 1.4728788137435913,
|
| 260 |
+
"learning_rate": 6.846846846846847e-05,
|
| 261 |
+
"loss": 0.832,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.33183856502242154,
|
| 266 |
+
"grad_norm": 1.3111985921859741,
|
| 267 |
+
"learning_rate": 6.756756756756757e-05,
|
| 268 |
+
"loss": 1.7645,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.34080717488789236,
|
| 273 |
+
"grad_norm": 3.0567102432250977,
|
| 274 |
+
"learning_rate": 6.666666666666667e-05,
|
| 275 |
+
"loss": 1.5445,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.34977578475336324,
|
| 280 |
+
"grad_norm": 8.974732398986816,
|
| 281 |
+
"learning_rate": 6.576576576576577e-05,
|
| 282 |
+
"loss": 9.2399,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.35874439461883406,
|
| 287 |
+
"grad_norm": 2.831979990005493,
|
| 288 |
+
"learning_rate": 6.486486486486487e-05,
|
| 289 |
+
"loss": 3.8836,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.36771300448430494,
|
| 294 |
+
"grad_norm": 14.11927604675293,
|
| 295 |
+
"learning_rate": 6.396396396396397e-05,
|
| 296 |
+
"loss": 11.535,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.37668161434977576,
|
| 301 |
+
"grad_norm": 4.956366062164307,
|
| 302 |
+
"learning_rate": 6.306306306306306e-05,
|
| 303 |
+
"loss": 1.1701,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.38565022421524664,
|
| 308 |
+
"grad_norm": 3.545992136001587,
|
| 309 |
+
"learning_rate": 6.216216216216216e-05,
|
| 310 |
+
"loss": 3.0184,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.39461883408071746,
|
| 315 |
+
"grad_norm": 2.012129068374634,
|
| 316 |
+
"learning_rate": 6.126126126126126e-05,
|
| 317 |
+
"loss": 4.641,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.40358744394618834,
|
| 322 |
+
"grad_norm": 1.8412423133850098,
|
| 323 |
+
"learning_rate": 6.0360360360360365e-05,
|
| 324 |
+
"loss": 1.1749,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.4125560538116592,
|
| 329 |
+
"grad_norm": 2.2313199043273926,
|
| 330 |
+
"learning_rate": 5.9459459459459466e-05,
|
| 331 |
+
"loss": 1.5577,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.42152466367713004,
|
| 336 |
+
"grad_norm": 10.823288917541504,
|
| 337 |
+
"learning_rate": 5.855855855855856e-05,
|
| 338 |
+
"loss": 15.2545,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.4304932735426009,
|
| 343 |
+
"grad_norm": 4.433194637298584,
|
| 344 |
+
"learning_rate": 5.765765765765766e-05,
|
| 345 |
+
"loss": 1.6109,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.43946188340807174,
|
| 350 |
+
"grad_norm": 2.7911524772644043,
|
| 351 |
+
"learning_rate": 5.6756756756756757e-05,
|
| 352 |
+
"loss": 0.8831,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4484304932735426,
|
| 357 |
+
"grad_norm": 7.617527008056641,
|
| 358 |
+
"learning_rate": 5.585585585585585e-05,
|
| 359 |
+
"loss": 1.4399,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.45739910313901344,
|
| 364 |
+
"grad_norm": 7.5390777587890625,
|
| 365 |
+
"learning_rate": 5.4954954954954966e-05,
|
| 366 |
+
"loss": 6.7722,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4663677130044843,
|
| 371 |
+
"grad_norm": 9.198465347290039,
|
| 372 |
+
"learning_rate": 5.405405405405406e-05,
|
| 373 |
+
"loss": 1.1899,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.47533632286995514,
|
| 378 |
+
"grad_norm": 13.445169448852539,
|
| 379 |
+
"learning_rate": 5.3153153153153155e-05,
|
| 380 |
+
"loss": 17.8221,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.484304932735426,
|
| 385 |
+
"grad_norm": 2.501037359237671,
|
| 386 |
+
"learning_rate": 5.2252252252252256e-05,
|
| 387 |
+
"loss": 0.7443,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.49327354260089684,
|
| 392 |
+
"grad_norm": 20.265993118286133,
|
| 393 |
+
"learning_rate": 5.135135135135135e-05,
|
| 394 |
+
"loss": 18.0624,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5022421524663677,
|
| 399 |
+
"grad_norm": 2.8558311462402344,
|
| 400 |
+
"learning_rate": 5.0450450450450445e-05,
|
| 401 |
+
"loss": 2.4351,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5112107623318386,
|
| 406 |
+
"grad_norm": 9.033570289611816,
|
| 407 |
+
"learning_rate": 4.954954954954955e-05,
|
| 408 |
+
"loss": 1.05,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5201793721973094,
|
| 413 |
+
"grad_norm": 12.937838554382324,
|
| 414 |
+
"learning_rate": 4.8648648648648654e-05,
|
| 415 |
+
"loss": 1.2733,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5291479820627802,
|
| 420 |
+
"grad_norm": 4.5761942863464355,
|
| 421 |
+
"learning_rate": 4.774774774774775e-05,
|
| 422 |
+
"loss": 1.7739,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5381165919282511,
|
| 427 |
+
"grad_norm": 10.550862312316895,
|
| 428 |
+
"learning_rate": 4.684684684684685e-05,
|
| 429 |
+
"loss": 0.7909,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.547085201793722,
|
| 434 |
+
"grad_norm": 13.009169578552246,
|
| 435 |
+
"learning_rate": 4.594594594594595e-05,
|
| 436 |
+
"loss": 0.9064,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5560538116591929,
|
| 441 |
+
"grad_norm": 35.53451919555664,
|
| 442 |
+
"learning_rate": 4.5045045045045046e-05,
|
| 443 |
+
"loss": 12.5267,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5650224215246636,
|
| 448 |
+
"grad_norm": 3.889129161834717,
|
| 449 |
+
"learning_rate": 4.414414414414415e-05,
|
| 450 |
+
"loss": 0.6209,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5739910313901345,
|
| 455 |
+
"grad_norm": 33.15275573730469,
|
| 456 |
+
"learning_rate": 4.324324324324325e-05,
|
| 457 |
+
"loss": 5.5689,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5829596412556054,
|
| 462 |
+
"grad_norm": 10.819781303405762,
|
| 463 |
+
"learning_rate": 4.234234234234234e-05,
|
| 464 |
+
"loss": 0.4751,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5919282511210763,
|
| 469 |
+
"grad_norm": 27.67494773864746,
|
| 470 |
+
"learning_rate": 4.1441441441441444e-05,
|
| 471 |
+
"loss": 6.5307,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.600896860986547,
|
| 476 |
+
"grad_norm": 2.997436285018921,
|
| 477 |
+
"learning_rate": 4.0540540540540545e-05,
|
| 478 |
+
"loss": 1.2605,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6098654708520179,
|
| 483 |
+
"grad_norm": 36.24047088623047,
|
| 484 |
+
"learning_rate": 3.963963963963964e-05,
|
| 485 |
+
"loss": 7.8812,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6188340807174888,
|
| 490 |
+
"grad_norm": 24.230558395385742,
|
| 491 |
+
"learning_rate": 3.873873873873874e-05,
|
| 492 |
+
"loss": 4.867,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6278026905829597,
|
| 497 |
+
"grad_norm": 8.829228401184082,
|
| 498 |
+
"learning_rate": 3.783783783783784e-05,
|
| 499 |
+
"loss": 0.7367,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6367713004484304,
|
| 504 |
+
"grad_norm": 6.5498199462890625,
|
| 505 |
+
"learning_rate": 3.693693693693694e-05,
|
| 506 |
+
"loss": 0.6523,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6457399103139013,
|
| 511 |
+
"grad_norm": 12.350178718566895,
|
| 512 |
+
"learning_rate": 3.603603603603604e-05,
|
| 513 |
+
"loss": 0.9404,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6547085201793722,
|
| 518 |
+
"grad_norm": 1.8611844778060913,
|
| 519 |
+
"learning_rate": 3.513513513513514e-05,
|
| 520 |
+
"loss": 0.3853,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6636771300448431,
|
| 525 |
+
"grad_norm": 7.657431125640869,
|
| 526 |
+
"learning_rate": 3.4234234234234234e-05,
|
| 527 |
+
"loss": 3.3681,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.672645739910314,
|
| 532 |
+
"grad_norm": 13.772859573364258,
|
| 533 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 534 |
+
"loss": 1.3095,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6816143497757847,
|
| 539 |
+
"grad_norm": 5.472783088684082,
|
| 540 |
+
"learning_rate": 3.2432432432432436e-05,
|
| 541 |
+
"loss": 0.6071,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6905829596412556,
|
| 546 |
+
"grad_norm": 14.440363883972168,
|
| 547 |
+
"learning_rate": 3.153153153153153e-05,
|
| 548 |
+
"loss": 7.5165,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6995515695067265,
|
| 553 |
+
"grad_norm": 13.61552906036377,
|
| 554 |
+
"learning_rate": 3.063063063063063e-05,
|
| 555 |
+
"loss": 3.1452,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7085201793721974,
|
| 560 |
+
"grad_norm": 48.33591842651367,
|
| 561 |
+
"learning_rate": 2.9729729729729733e-05,
|
| 562 |
+
"loss": 16.6477,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7174887892376681,
|
| 567 |
+
"grad_norm": 7.3193464279174805,
|
| 568 |
+
"learning_rate": 2.882882882882883e-05,
|
| 569 |
+
"loss": 2.8526,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.726457399103139,
|
| 574 |
+
"grad_norm": 13.097620010375977,
|
| 575 |
+
"learning_rate": 2.7927927927927926e-05,
|
| 576 |
+
"loss": 0.87,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7354260089686099,
|
| 581 |
+
"grad_norm": 11.424261093139648,
|
| 582 |
+
"learning_rate": 2.702702702702703e-05,
|
| 583 |
+
"loss": 0.7039,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7443946188340808,
|
| 588 |
+
"grad_norm": 4.838938236236572,
|
| 589 |
+
"learning_rate": 2.6126126126126128e-05,
|
| 590 |
+
"loss": 0.5254,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7533632286995515,
|
| 595 |
+
"grad_norm": 17.895553588867188,
|
| 596 |
+
"learning_rate": 2.5225225225225222e-05,
|
| 597 |
+
"loss": 5.5567,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7623318385650224,
|
| 602 |
+
"grad_norm": 5.642916202545166,
|
| 603 |
+
"learning_rate": 2.4324324324324327e-05,
|
| 604 |
+
"loss": 0.8884,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7713004484304933,
|
| 609 |
+
"grad_norm": 11.742246627807617,
|
| 610 |
+
"learning_rate": 2.3423423423423425e-05,
|
| 611 |
+
"loss": 4.378,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7802690582959642,
|
| 616 |
+
"grad_norm": 15.400406837463379,
|
| 617 |
+
"learning_rate": 2.2522522522522523e-05,
|
| 618 |
+
"loss": 1.1752,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7892376681614349,
|
| 623 |
+
"grad_norm": 6.748379707336426,
|
| 624 |
+
"learning_rate": 2.1621621621621624e-05,
|
| 625 |
+
"loss": 0.6377,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7982062780269058,
|
| 630 |
+
"grad_norm": 13.948418617248535,
|
| 631 |
+
"learning_rate": 2.0720720720720722e-05,
|
| 632 |
+
"loss": 4.6146,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8071748878923767,
|
| 637 |
+
"grad_norm": 5.0134687423706055,
|
| 638 |
+
"learning_rate": 1.981981981981982e-05,
|
| 639 |
+
"loss": 0.9851,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.8161434977578476,
|
| 644 |
+
"grad_norm": 2.1885986328125,
|
| 645 |
+
"learning_rate": 1.891891891891892e-05,
|
| 646 |
+
"loss": 1.7911,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8251121076233184,
|
| 651 |
+
"grad_norm": 19.37745475769043,
|
| 652 |
+
"learning_rate": 1.801801801801802e-05,
|
| 653 |
+
"loss": 13.3176,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8340807174887892,
|
| 658 |
+
"grad_norm": 25.039756774902344,
|
| 659 |
+
"learning_rate": 1.7117117117117117e-05,
|
| 660 |
+
"loss": 6.4343,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8430493273542601,
|
| 665 |
+
"grad_norm": 20.313587188720703,
|
| 666 |
+
"learning_rate": 1.6216216216216218e-05,
|
| 667 |
+
"loss": 9.2506,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.852017937219731,
|
| 672 |
+
"grad_norm": 17.610029220581055,
|
| 673 |
+
"learning_rate": 1.5315315315315316e-05,
|
| 674 |
+
"loss": 7.9287,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.8609865470852018,
|
| 679 |
+
"grad_norm": 6.322761058807373,
|
| 680 |
+
"learning_rate": 1.4414414414414416e-05,
|
| 681 |
+
"loss": 3.0756,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.8699551569506726,
|
| 686 |
+
"grad_norm": 8.840217590332031,
|
| 687 |
+
"learning_rate": 1.3513513513513515e-05,
|
| 688 |
+
"loss": 1.0194,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8789237668161435,
|
| 693 |
+
"grad_norm": 1.8822249174118042,
|
| 694 |
+
"learning_rate": 1.2612612612612611e-05,
|
| 695 |
+
"loss": 1.5888,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8878923766816144,
|
| 700 |
+
"grad_norm": 15.456230163574219,
|
| 701 |
+
"learning_rate": 1.1711711711711713e-05,
|
| 702 |
+
"loss": 6.5939,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8968609865470852,
|
| 707 |
+
"grad_norm": 4.190560817718506,
|
| 708 |
+
"learning_rate": 1.0810810810810812e-05,
|
| 709 |
+
"loss": 2.3652,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.905829596412556,
|
| 714 |
+
"grad_norm": 4.274927139282227,
|
| 715 |
+
"learning_rate": 9.90990990990991e-06,
|
| 716 |
+
"loss": 2.2314,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9147982062780269,
|
| 721 |
+
"grad_norm": 10.192256927490234,
|
| 722 |
+
"learning_rate": 9.00900900900901e-06,
|
| 723 |
+
"loss": 0.9849,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9237668161434978,
|
| 728 |
+
"grad_norm": 6.9923095703125,
|
| 729 |
+
"learning_rate": 8.108108108108109e-06,
|
| 730 |
+
"loss": 0.8398,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9327354260089686,
|
| 735 |
+
"grad_norm": 18.946592330932617,
|
| 736 |
+
"learning_rate": 7.207207207207208e-06,
|
| 737 |
+
"loss": 15.561,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9417040358744395,
|
| 742 |
+
"grad_norm": 10.394827842712402,
|
| 743 |
+
"learning_rate": 6.306306306306306e-06,
|
| 744 |
+
"loss": 1.3364,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.9506726457399103,
|
| 749 |
+
"grad_norm": 7.969135284423828,
|
| 750 |
+
"learning_rate": 5.405405405405406e-06,
|
| 751 |
+
"loss": 4.5584,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.9596412556053812,
|
| 756 |
+
"grad_norm": 18.159500122070312,
|
| 757 |
+
"learning_rate": 4.504504504504505e-06,
|
| 758 |
+
"loss": 1.6239,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.968609865470852,
|
| 763 |
+
"grad_norm": 8.934094429016113,
|
| 764 |
+
"learning_rate": 3.603603603603604e-06,
|
| 765 |
+
"loss": 0.7876,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9775784753363229,
|
| 770 |
+
"grad_norm": 3.407855272293091,
|
| 771 |
+
"learning_rate": 2.702702702702703e-06,
|
| 772 |
+
"loss": 4.5206,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9865470852017937,
|
| 777 |
+
"grad_norm": 4.0875701904296875,
|
| 778 |
+
"learning_rate": 1.801801801801802e-06,
|
| 779 |
+
"loss": 2.6009,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9955156950672646,
|
| 784 |
+
"grad_norm": 3.929051637649536,
|
| 785 |
+
"learning_rate": 9.00900900900901e-07,
|
| 786 |
+
"loss": 6.0557,
|
| 787 |
+
"step": 111
|
| 788 |
+
}
|
| 789 |
+
],
|
| 790 |
+
"logging_steps": 1,
|
| 791 |
+
"max_steps": 111,
|
| 792 |
+
"num_input_tokens_seen": 0,
|
| 793 |
+
"num_train_epochs": 1,
|
| 794 |
+
"save_steps": 200,
|
| 795 |
+
"stateful_callbacks": {
|
| 796 |
+
"TrainerControl": {
|
| 797 |
+
"args": {
|
| 798 |
+
"should_epoch_stop": false,
|
| 799 |
+
"should_evaluate": false,
|
| 800 |
+
"should_log": false,
|
| 801 |
+
"should_save": true,
|
| 802 |
+
"should_training_stop": true
|
| 803 |
+
},
|
| 804 |
+
"attributes": {}
|
| 805 |
+
}
|
| 806 |
+
},
|
| 807 |
+
"total_flos": 0.0,
|
| 808 |
+
"train_batch_size": 24,
|
| 809 |
+
"trial_name": null,
|
| 810 |
+
"trial_params": null
|
| 811 |
+
}
|
20250501-1421/checkpoint-111/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5af9beeda97f152c735237547cc7b18512cf2624167015790c50fa65590e3a8b
|
| 3 |
+
size 5304
|
20250501-1421/runs/May01_14-21-16_d55355693cce/events.out.tfevents.1746109277.d55355693cce.100.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:733af7f3581ca84c24b5a6088a4614469a0ce481382b094ca7333e15a4280109
|
| 3 |
+
size 27539
|