Upload folder using huggingface_hub
Browse files
coco_checkpoints/checkpoint-1445/adapter.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72c4d6f5d9a4fa1041f97cad75da6f005dddef6561ef9a6f5ee1588d4baa58ef
|
| 3 |
+
size 17064856
|
coco_checkpoints/checkpoint-1445/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7deb029965a5f3aa5628ab7bf742d8f256b2428ddf7503f6827f9734b76cf91
|
| 3 |
+
size 8714492
|
coco_checkpoints/checkpoint-1445/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb028321b15740df4ceb8773af2e92f09aeb64177e0d03281800bfd793bbc2a7
|
| 3 |
+
size 14244
|
coco_checkpoints/checkpoint-1445/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac3188b75baca9bbfe53e3e08b7e090c4f46683e7933bbae216c3ba16f204ab2
|
| 3 |
+
size 1064
|
coco_checkpoints/checkpoint-1445/trainer_state.json
ADDED
|
@@ -0,0 +1,758 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 250,
|
| 6 |
+
"global_step": 1445,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.011072664359861591,
|
| 13 |
+
"grad_norm": 7.3027024269104,
|
| 14 |
+
"learning_rate": 0.00036363636363636367,
|
| 15 |
+
"loss": 7.4688,
|
| 16 |
+
"step": 16
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.022145328719723183,
|
| 20 |
+
"grad_norm": 3.441852331161499,
|
| 21 |
+
"learning_rate": 0.0006363636363636364,
|
| 22 |
+
"loss": 6.1486,
|
| 23 |
+
"step": 32
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03321799307958478,
|
| 27 |
+
"grad_norm": 0.9797470569610596,
|
| 28 |
+
"learning_rate": 0.001,
|
| 29 |
+
"loss": 5.1898,
|
| 30 |
+
"step": 48
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.044290657439446365,
|
| 34 |
+
"grad_norm": 0.4113083481788635,
|
| 35 |
+
"learning_rate": 0.0009996782216198338,
|
| 36 |
+
"loss": 5.0792,
|
| 37 |
+
"step": 64
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05536332179930796,
|
| 41 |
+
"grad_norm": 0.23998336493968964,
|
| 42 |
+
"learning_rate": 0.0009987133006446386,
|
| 43 |
+
"loss": 4.9765,
|
| 44 |
+
"step": 80
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06643598615916955,
|
| 48 |
+
"grad_norm": 0.34719282388687134,
|
| 49 |
+
"learning_rate": 0.0009971064790372482,
|
| 50 |
+
"loss": 5.0589,
|
| 51 |
+
"step": 96
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07750865051903114,
|
| 55 |
+
"grad_norm": 0.28703269362449646,
|
| 56 |
+
"learning_rate": 0.0009948598249594788,
|
| 57 |
+
"loss": 5.0273,
|
| 58 |
+
"step": 112
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08858131487889273,
|
| 62 |
+
"grad_norm": 0.3285124599933624,
|
| 63 |
+
"learning_rate": 0.00099197623011017,
|
| 64 |
+
"loss": 4.9965,
|
| 65 |
+
"step": 128
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09965397923875433,
|
| 69 |
+
"grad_norm": 0.21146929264068604,
|
| 70 |
+
"learning_rate": 0.0009884594060032405,
|
| 71 |
+
"loss": 5.0386,
|
| 72 |
+
"step": 144
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.11072664359861592,
|
| 76 |
+
"grad_norm": 0.2574313282966614,
|
| 77 |
+
"learning_rate": 0.0009843138791905482,
|
| 78 |
+
"loss": 4.9744,
|
| 79 |
+
"step": 160
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.12179930795847752,
|
| 83 |
+
"grad_norm": 0.2936093211174011,
|
| 84 |
+
"learning_rate": 0.000979544985435704,
|
| 85 |
+
"loss": 5.0448,
|
| 86 |
+
"step": 176
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1328719723183391,
|
| 90 |
+
"grad_norm": 0.2813468277454376,
|
| 91 |
+
"learning_rate": 0.0009741588628463382,
|
| 92 |
+
"loss": 5.0399,
|
| 93 |
+
"step": 192
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1439446366782007,
|
| 97 |
+
"grad_norm": 0.27917245030403137,
|
| 98 |
+
"learning_rate": 0.0009681624439736599,
|
| 99 |
+
"loss": 5.0001,
|
| 100 |
+
"step": 208
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.15501730103806227,
|
| 104 |
+
"grad_norm": 0.23994863033294678,
|
| 105 |
+
"learning_rate": 0.0009615634468894752,
|
| 106 |
+
"loss": 4.9829,
|
| 107 |
+
"step": 224
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.16608996539792387,
|
| 111 |
+
"grad_norm": 0.18816818296909332,
|
| 112 |
+
"learning_rate": 0.0009543703652521542,
|
| 113 |
+
"loss": 4.9348,
|
| 114 |
+
"step": 240
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.17301038062283736,
|
| 118 |
+
"eval_bleu": 0.07082991922728418,
|
| 119 |
+
"eval_cap_loss": 1.862730330052497,
|
| 120 |
+
"eval_con_loss": 2.0789193881875687,
|
| 121 |
+
"eval_loss": 3.941649715827423,
|
| 122 |
+
"step": 250
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.17301038062283736,
|
| 126 |
+
"eval_bleu": 0.07082991922728418,
|
| 127 |
+
"eval_cap_loss": 1.862730330052497,
|
| 128 |
+
"eval_con_loss": 2.0789193881875687,
|
| 129 |
+
"eval_loss": 3.941649715827423,
|
| 130 |
+
"eval_runtime": 809.5614,
|
| 131 |
+
"eval_samples_per_second": 19.036,
|
| 132 |
+
"eval_steps_per_second": 2.38,
|
| 133 |
+
"step": 250
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.17716262975778546,
|
| 137 |
+
"grad_norm": 0.24259746074676514,
|
| 138 |
+
"learning_rate": 0.0009465924573743279,
|
| 139 |
+
"loss": 4.9856,
|
| 140 |
+
"step": 256
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.18823529411764706,
|
| 144 |
+
"grad_norm": 0.2658279836177826,
|
| 145 |
+
"learning_rate": 0.0009382397343063877,
|
| 146 |
+
"loss": 4.9553,
|
| 147 |
+
"step": 272
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.19930795847750865,
|
| 151 |
+
"grad_norm": 0.18191872537136078,
|
| 152 |
+
"learning_rate": 0.0009293229469511293,
|
| 153 |
+
"loss": 4.9995,
|
| 154 |
+
"step": 288
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.21038062283737025,
|
| 158 |
+
"grad_norm": 0.29174473881721497,
|
| 159 |
+
"learning_rate": 0.000919853572226118,
|
| 160 |
+
"loss": 5.0197,
|
| 161 |
+
"step": 304
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.22145328719723184,
|
| 165 |
+
"grad_norm": 0.21258646249771118,
|
| 166 |
+
"learning_rate": 0.0009098437982915953,
|
| 167 |
+
"loss": 4.9636,
|
| 168 |
+
"step": 320
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.23252595155709344,
|
| 172 |
+
"grad_norm": 0.18101181089878082,
|
| 173 |
+
"learning_rate": 0.0008993065088629304,
|
| 174 |
+
"loss": 4.9274,
|
| 175 |
+
"step": 336
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.24359861591695503,
|
| 179 |
+
"grad_norm": 0.2041468620300293,
|
| 180 |
+
"learning_rate": 0.0008882552666278186,
|
| 181 |
+
"loss": 5.0077,
|
| 182 |
+
"step": 352
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.2546712802768166,
|
| 186 |
+
"grad_norm": 0.3073960244655609,
|
| 187 |
+
"learning_rate": 0.0008767042957895606,
|
| 188 |
+
"loss": 4.9781,
|
| 189 |
+
"step": 368
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.2657439446366782,
|
| 193 |
+
"grad_norm": 0.17274291813373566,
|
| 194 |
+
"learning_rate": 0.0008646684637588991,
|
| 195 |
+
"loss": 4.911,
|
| 196 |
+
"step": 384
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.2768166089965398,
|
| 200 |
+
"grad_norm": 0.20907478034496307,
|
| 201 |
+
"learning_rate": 0.0008521632620179735,
|
| 202 |
+
"loss": 5.0089,
|
| 203 |
+
"step": 400
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.2878892733564014,
|
| 207 |
+
"grad_norm": 0.2937226891517639,
|
| 208 |
+
"learning_rate": 0.0008392047861810229,
|
| 209 |
+
"loss": 4.9306,
|
| 210 |
+
"step": 416
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.29896193771626295,
|
| 214 |
+
"grad_norm": 0.35890915989875793,
|
| 215 |
+
"learning_rate": 0.0008258097152775044,
|
| 216 |
+
"loss": 4.9363,
|
| 217 |
+
"step": 432
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.31003460207612454,
|
| 221 |
+
"grad_norm": 0.19902297854423523,
|
| 222 |
+
"learning_rate": 0.0008119952902842882,
|
| 223 |
+
"loss": 4.9872,
|
| 224 |
+
"step": 448
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.32110726643598614,
|
| 228 |
+
"grad_norm": 0.2420748621225357,
|
| 229 |
+
"learning_rate": 0.0007977792919345632,
|
| 230 |
+
"loss": 5.0178,
|
| 231 |
+
"step": 464
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.33217993079584773,
|
| 235 |
+
"grad_norm": 0.21633224189281464,
|
| 236 |
+
"learning_rate": 0.0007831800178320152,
|
| 237 |
+
"loss": 4.9588,
|
| 238 |
+
"step": 480
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.34325259515570933,
|
| 242 |
+
"grad_norm": 0.2581658363342285,
|
| 243 |
+
"learning_rate": 0.0007682162588997332,
|
| 244 |
+
"loss": 4.9214,
|
| 245 |
+
"step": 496
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.3460207612456747,
|
| 249 |
+
"eval_bleu": 0.0743834461108624,
|
| 250 |
+
"eval_cap_loss": 1.8318203130285544,
|
| 251 |
+
"eval_con_loss": 2.0786952485553205,
|
| 252 |
+
"eval_loss": 3.9105155644914187,
|
| 253 |
+
"step": 500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.3460207612456747,
|
| 257 |
+
"eval_bleu": 0.0743834461108624,
|
| 258 |
+
"eval_cap_loss": 1.8318203130285544,
|
| 259 |
+
"eval_con_loss": 2.0786952485553205,
|
| 260 |
+
"eval_loss": 3.9105155644914187,
|
| 261 |
+
"eval_runtime": 820.0211,
|
| 262 |
+
"eval_samples_per_second": 18.793,
|
| 263 |
+
"eval_steps_per_second": 2.35,
|
| 264 |
+
"step": 500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.3543252595155709,
|
| 268 |
+
"grad_norm": 0.24933111667633057,
|
| 269 |
+
"learning_rate": 0.0007529072751941595,
|
| 270 |
+
"loss": 4.9951,
|
| 271 |
+
"step": 512
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.3653979238754325,
|
| 275 |
+
"grad_norm": 0.1792280375957489,
|
| 276 |
+
"learning_rate": 0.0007372727711152087,
|
| 277 |
+
"loss": 4.8813,
|
| 278 |
+
"step": 528
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.3764705882352941,
|
| 282 |
+
"grad_norm": 0.13637550175189972,
|
| 283 |
+
"learning_rate": 0.0007213328700444696,
|
| 284 |
+
"loss": 4.9125,
|
| 285 |
+
"step": 544
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.3875432525951557,
|
| 289 |
+
"grad_norm": 0.3409987688064575,
|
| 290 |
+
"learning_rate": 0.0007051080884441287,
|
| 291 |
+
"loss": 4.9073,
|
| 292 |
+
"step": 560
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.3986159169550173,
|
| 296 |
+
"grad_norm": 0.2467418760061264,
|
| 297 |
+
"learning_rate": 0.0006886193094499536,
|
| 298 |
+
"loss": 4.9227,
|
| 299 |
+
"step": 576
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.4096885813148789,
|
| 303 |
+
"grad_norm": 0.2751147747039795,
|
| 304 |
+
"learning_rate": 0.000671887755992327,
|
| 305 |
+
"loss": 4.9871,
|
| 306 |
+
"step": 592
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.4207612456747405,
|
| 310 |
+
"grad_norm": 0.21447473764419556,
|
| 311 |
+
"learning_rate": 0.000654934963479926,
|
| 312 |
+
"loss": 4.9484,
|
| 313 |
+
"step": 608
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.4318339100346021,
|
| 317 |
+
"grad_norm": 0.26142337918281555,
|
| 318 |
+
"learning_rate": 0.0006377827520812061,
|
| 319 |
+
"loss": 4.9204,
|
| 320 |
+
"step": 624
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.4429065743944637,
|
| 324 |
+
"grad_norm": 0.3584180176258087,
|
| 325 |
+
"learning_rate": 0.0006204531986393678,
|
| 326 |
+
"loss": 5.0,
|
| 327 |
+
"step": 640
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.4539792387543253,
|
| 331 |
+
"grad_norm": 0.2877948582172394,
|
| 332 |
+
"learning_rate": 0.0006029686082569537,
|
| 333 |
+
"loss": 5.0003,
|
| 334 |
+
"step": 656
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.46505190311418687,
|
| 338 |
+
"grad_norm": 0.2049325406551361,
|
| 339 |
+
"learning_rate": 0.000585351485586648,
|
| 340 |
+
"loss": 4.9349,
|
| 341 |
+
"step": 672
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 0.47612456747404847,
|
| 345 |
+
"grad_norm": 0.200492262840271,
|
| 346 |
+
"learning_rate": 0.0005676245058652349,
|
| 347 |
+
"loss": 4.9385,
|
| 348 |
+
"step": 688
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.48719723183391006,
|
| 352 |
+
"grad_norm": 0.24712397158145905,
|
| 353 |
+
"learning_rate": 0.000549810485727994,
|
| 354 |
+
"loss": 4.9516,
|
| 355 |
+
"step": 704
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.4982698961937716,
|
| 359 |
+
"grad_norm": 0.380741149187088,
|
| 360 |
+
"learning_rate": 0.0005319323538411021,
|
| 361 |
+
"loss": 4.9121,
|
| 362 |
+
"step": 720
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.5093425605536333,
|
| 366 |
+
"grad_norm": 0.2339814305305481,
|
| 367 |
+
"learning_rate": 0.0005140131213898345,
|
| 368 |
+
"loss": 4.9058,
|
| 369 |
+
"step": 736
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.5190311418685121,
|
| 373 |
+
"eval_bleu": 0.0755992902414665,
|
| 374 |
+
"eval_cap_loss": 1.8195727369484656,
|
| 375 |
+
"eval_con_loss": 2.0776209530585406,
|
| 376 |
+
"eval_loss": 3.897193689945144,
|
| 377 |
+
"step": 750
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.5190311418685121,
|
| 381 |
+
"eval_bleu": 0.0755992902414665,
|
| 382 |
+
"eval_cap_loss": 1.8195727369484656,
|
| 383 |
+
"eval_con_loss": 2.0776209530585406,
|
| 384 |
+
"eval_loss": 3.897193689945144,
|
| 385 |
+
"eval_runtime": 819.7019,
|
| 386 |
+
"eval_samples_per_second": 18.801,
|
| 387 |
+
"eval_steps_per_second": 2.351,
|
| 388 |
+
"step": 750
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.5204152249134948,
|
| 392 |
+
"grad_norm": 0.23871256411075592,
|
| 393 |
+
"learning_rate": 0.0004960758524605593,
|
| 394 |
+
"loss": 4.9217,
|
| 395 |
+
"step": 752
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5314878892733564,
|
| 399 |
+
"grad_norm": 0.19021867215633392,
|
| 400 |
+
"learning_rate": 0.0004781436343546391,
|
| 401 |
+
"loss": 4.9383,
|
| 402 |
+
"step": 768
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.542560553633218,
|
| 406 |
+
"grad_norm": 0.39012765884399414,
|
| 407 |
+
"learning_rate": 0.0004602395478724539,
|
| 408 |
+
"loss": 4.8852,
|
| 409 |
+
"step": 784
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5536332179930796,
|
| 413 |
+
"grad_norm": 0.2299482375383377,
|
| 414 |
+
"learning_rate": 0.00044238663760578963,
|
| 415 |
+
"loss": 4.9461,
|
| 416 |
+
"step": 800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5647058823529412,
|
| 420 |
+
"grad_norm": 0.20332716405391693,
|
| 421 |
+
"learning_rate": 0.0004246078822768339,
|
| 422 |
+
"loss": 4.8672,
|
| 423 |
+
"step": 816
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5757785467128028,
|
| 427 |
+
"grad_norm": 0.2742558717727661,
|
| 428 |
+
"learning_rate": 0.00040692616516195134,
|
| 429 |
+
"loss": 5.0016,
|
| 430 |
+
"step": 832
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5868512110726644,
|
| 434 |
+
"grad_norm": 0.18639664351940155,
|
| 435 |
+
"learning_rate": 0.0003893642446383089,
|
| 436 |
+
"loss": 4.8718,
|
| 437 |
+
"step": 848
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5979238754325259,
|
| 441 |
+
"grad_norm": 0.14423900842666626,
|
| 442 |
+
"learning_rate": 0.00037194472489126174,
|
| 443 |
+
"loss": 4.9254,
|
| 444 |
+
"step": 864
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.6089965397923875,
|
| 448 |
+
"grad_norm": 0.25753453373908997,
|
| 449 |
+
"learning_rate": 0.00035469002682019933,
|
| 450 |
+
"loss": 5.0226,
|
| 451 |
+
"step": 880
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.6200692041522491,
|
| 455 |
+
"grad_norm": 0.29433926939964294,
|
| 456 |
+
"learning_rate": 0.00033762235918030425,
|
| 457 |
+
"loss": 4.9075,
|
| 458 |
+
"step": 896
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.6311418685121107,
|
| 462 |
+
"grad_norm": 0.16973747313022614,
|
| 463 |
+
"learning_rate": 0.0003207636899973617,
|
| 464 |
+
"loss": 4.9367,
|
| 465 |
+
"step": 912
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.6422145328719723,
|
| 469 |
+
"grad_norm": 0.23007385432720184,
|
| 470 |
+
"learning_rate": 0.0003041357182924178,
|
| 471 |
+
"loss": 4.8899,
|
| 472 |
+
"step": 928
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6532871972318339,
|
| 476 |
+
"grad_norm": 0.23643143475055695,
|
| 477 |
+
"learning_rate": 0.000287759846152675,
|
| 478 |
+
"loss": 4.8813,
|
| 479 |
+
"step": 944
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6643598615916955,
|
| 483 |
+
"grad_norm": 0.17182067036628723,
|
| 484 |
+
"learning_rate": 0.00027165715118457735,
|
| 485 |
+
"loss": 4.9201,
|
| 486 |
+
"step": 960
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6754325259515571,
|
| 490 |
+
"grad_norm": 0.29925912618637085,
|
| 491 |
+
"learning_rate": 0.0002558483593845372,
|
| 492 |
+
"loss": 4.898,
|
| 493 |
+
"step": 976
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6865051903114187,
|
| 497 |
+
"grad_norm": 0.17474640905857086,
|
| 498 |
+
"learning_rate": 0.00024035381846222555,
|
| 499 |
+
"loss": 4.8839,
|
| 500 |
+
"step": 992
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6920415224913494,
|
| 504 |
+
"eval_bleu": 0.07638727198557162,
|
| 505 |
+
"eval_cap_loss": 1.797022891217832,
|
| 506 |
+
"eval_con_loss": 2.0767276909416923,
|
| 507 |
+
"eval_loss": 3.87375058110786,
|
| 508 |
+
"step": 1000
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.6920415224913494,
|
| 512 |
+
"eval_bleu": 0.07638727198557162,
|
| 513 |
+
"eval_cap_loss": 1.797022891217832,
|
| 514 |
+
"eval_con_loss": 2.0767276909416923,
|
| 515 |
+
"eval_loss": 3.87375058110786,
|
| 516 |
+
"eval_runtime": 815.983,
|
| 517 |
+
"eval_samples_per_second": 18.886,
|
| 518 |
+
"eval_steps_per_second": 2.362,
|
| 519 |
+
"step": 1000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.6975778546712803,
|
| 523 |
+
"grad_norm": 0.2200097143650055,
|
| 524 |
+
"learning_rate": 0.00022519347165076065,
|
| 525 |
+
"loss": 4.9202,
|
| 526 |
+
"step": 1008
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.7086505190311418,
|
| 530 |
+
"grad_norm": 0.2346457690000534,
|
| 531 |
+
"learning_rate": 0.00021038683203750092,
|
| 532 |
+
"loss": 4.8698,
|
| 533 |
+
"step": 1024
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.7197231833910035,
|
| 537 |
+
"grad_norm": 0.24073997139930725,
|
| 538 |
+
"learning_rate": 0.00019595295744848825,
|
| 539 |
+
"loss": 5.01,
|
| 540 |
+
"step": 1040
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.730795847750865,
|
| 544 |
+
"grad_norm": 0.2730049192905426,
|
| 545 |
+
"learning_rate": 0.00018191042591886197,
|
| 546 |
+
"loss": 4.8847,
|
| 547 |
+
"step": 1056
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.7418685121107267,
|
| 551 |
+
"grad_norm": 0.21411365270614624,
|
| 552 |
+
"learning_rate": 0.00016827731178081822,
|
| 553 |
+
"loss": 4.9802,
|
| 554 |
+
"step": 1072
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.7529411764705882,
|
| 558 |
+
"grad_norm": 0.2260352373123169,
|
| 559 |
+
"learning_rate": 0.0001550711623998926,
|
| 560 |
+
"loss": 4.9593,
|
| 561 |
+
"step": 1088
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.7640138408304499,
|
| 565 |
+
"grad_norm": 0.2655484974384308,
|
| 566 |
+
"learning_rate": 0.0001423089755895095,
|
| 567 |
+
"loss": 4.8097,
|
| 568 |
+
"step": 1104
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.7750865051903114,
|
| 572 |
+
"grad_norm": 0.2067459672689438,
|
| 573 |
+
"learning_rate": 0.0001300071777328658,
|
| 574 |
+
"loss": 4.947,
|
| 575 |
+
"step": 1120
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.7861591695501731,
|
| 579 |
+
"grad_norm": 0.21391142904758453,
|
| 580 |
+
"learning_rate": 0.00011818160264031097,
|
| 581 |
+
"loss": 4.8673,
|
| 582 |
+
"step": 1136
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.7972318339100346,
|
| 586 |
+
"grad_norm": 0.3288532793521881,
|
| 587 |
+
"learning_rate": 0.00010684747116943683,
|
| 588 |
+
"loss": 4.9652,
|
| 589 |
+
"step": 1152
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 0.8083044982698961,
|
| 593 |
+
"grad_norm": 0.1693842113018036,
|
| 594 |
+
"learning_rate": 9.60193716341039e-05,
|
| 595 |
+
"loss": 4.9003,
|
| 596 |
+
"step": 1168
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.8193771626297578,
|
| 600 |
+
"grad_norm": 0.3056645691394806,
|
| 601 |
+
"learning_rate": 8.571124102762767e-05,
|
| 602 |
+
"loss": 4.9436,
|
| 603 |
+
"step": 1184
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.8304498269896193,
|
| 607 |
+
"grad_norm": 0.3365015387535095,
|
| 608 |
+
"learning_rate": 7.593634708428437e-05,
|
| 609 |
+
"loss": 5.0349,
|
| 610 |
+
"step": 1200
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.841522491349481,
|
| 614 |
+
"grad_norm": 0.2918214201927185,
|
| 615 |
+
"learning_rate": 6.670727120223142e-05,
|
| 616 |
+
"loss": 4.8812,
|
| 617 |
+
"step": 1216
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.8525951557093425,
|
| 621 |
+
"grad_norm": 0.2237936407327652,
|
| 622 |
+
"learning_rate": 5.8035892249820085e-05,
|
| 623 |
+
"loss": 4.8842,
|
| 624 |
+
"step": 1232
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.8636678200692042,
|
| 628 |
+
"grad_norm": 0.23957201838493347,
|
| 629 |
+
"learning_rate": 4.993337127614273e-05,
|
| 630 |
+
"loss": 4.9281,
|
| 631 |
+
"step": 1248
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 0.8650519031141869,
|
| 635 |
+
"eval_bleu": 0.07725825664704841,
|
| 636 |
+
"eval_cap_loss": 1.781457468055355,
|
| 637 |
+
"eval_con_loss": 2.0761843878924693,
|
| 638 |
+
"eval_loss": 3.8576418577727716,
|
| 639 |
+
"step": 1250
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.8650519031141869,
|
| 643 |
+
"eval_bleu": 0.07725825664704841,
|
| 644 |
+
"eval_cap_loss": 1.781457468055355,
|
| 645 |
+
"eval_con_loss": 2.0761843878924693,
|
| 646 |
+
"eval_loss": 3.8576418577727716,
|
| 647 |
+
"eval_runtime": 819.4117,
|
| 648 |
+
"eval_samples_per_second": 18.807,
|
| 649 |
+
"eval_steps_per_second": 2.352,
|
| 650 |
+
"step": 1250
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.8747404844290657,
|
| 654 |
+
"grad_norm": 0.09429222345352173,
|
| 655 |
+
"learning_rate": 4.2410137145495964e-05,
|
| 656 |
+
"loss": 4.8982,
|
| 657 |
+
"step": 1264
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.8858131487889274,
|
| 661 |
+
"grad_norm": 0.23660314083099365,
|
| 662 |
+
"learning_rate": 3.54758731142486e-05,
|
| 663 |
+
"loss": 4.9127,
|
| 664 |
+
"step": 1280
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 0.8968858131487889,
|
| 668 |
+
"grad_norm": 0.2092629075050354,
|
| 669 |
+
"learning_rate": 2.9139504367391158e-05,
|
| 670 |
+
"loss": 5.0183,
|
| 671 |
+
"step": 1296
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.9079584775086506,
|
| 675 |
+
"grad_norm": 0.2870592772960663,
|
| 676 |
+
"learning_rate": 2.3409186530809423e-05,
|
| 677 |
+
"loss": 4.9319,
|
| 678 |
+
"step": 1312
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.9190311418685121,
|
| 682 |
+
"grad_norm": 0.23483090102672577,
|
| 683 |
+
"learning_rate": 1.8292295174068717e-05,
|
| 684 |
+
"loss": 4.836,
|
| 685 |
+
"step": 1328
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.9301038062283737,
|
| 689 |
+
"grad_norm": 0.24700886011123657,
|
| 690 |
+
"learning_rate": 1.3795416317218035e-05,
|
| 691 |
+
"loss": 4.9113,
|
| 692 |
+
"step": 1344
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.9411764705882353,
|
| 696 |
+
"grad_norm": 0.20898666977882385,
|
| 697 |
+
"learning_rate": 9.924337953834795e-06,
|
| 698 |
+
"loss": 4.9519,
|
| 699 |
+
"step": 1360
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.9522491349480969,
|
| 703 |
+
"grad_norm": 0.14755718410015106,
|
| 704 |
+
"learning_rate": 6.684042601220186e-06,
|
| 705 |
+
"loss": 4.8827,
|
| 706 |
+
"step": 1376
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.9633217993079585,
|
| 710 |
+
"grad_norm": 0.2121778130531311,
|
| 711 |
+
"learning_rate": 4.078700887333364e-06,
|
| 712 |
+
"loss": 4.861,
|
| 713 |
+
"step": 1392
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.9743944636678201,
|
| 717 |
+
"grad_norm": 0.2525366544723511,
|
| 718 |
+
"learning_rate": 2.1116661827202956e-06,
|
| 719 |
+
"loss": 4.8674,
|
| 720 |
+
"step": 1408
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.9854671280276817,
|
| 724 |
+
"grad_norm": 0.278390109539032,
|
| 725 |
+
"learning_rate": 7.854702843449468e-07,
|
| 726 |
+
"loss": 4.9493,
|
| 727 |
+
"step": 1424
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.9965397923875432,
|
| 731 |
+
"grad_norm": 0.22038041055202484,
|
| 732 |
+
"learning_rate": 1.0182015687909552e-07,
|
| 733 |
+
"loss": 4.9065,
|
| 734 |
+
"step": 1440
|
| 735 |
+
}
|
| 736 |
+
],
|
| 737 |
+
"logging_steps": 16,
|
| 738 |
+
"max_steps": 1445,
|
| 739 |
+
"num_input_tokens_seen": 0,
|
| 740 |
+
"num_train_epochs": 1,
|
| 741 |
+
"save_steps": 250,
|
| 742 |
+
"stateful_callbacks": {
|
| 743 |
+
"TrainerControl": {
|
| 744 |
+
"args": {
|
| 745 |
+
"should_epoch_stop": false,
|
| 746 |
+
"should_evaluate": false,
|
| 747 |
+
"should_log": false,
|
| 748 |
+
"should_save": true,
|
| 749 |
+
"should_training_stop": true
|
| 750 |
+
},
|
| 751 |
+
"attributes": {}
|
| 752 |
+
}
|
| 753 |
+
},
|
| 754 |
+
"total_flos": 0.0,
|
| 755 |
+
"train_batch_size": 32,
|
| 756 |
+
"trial_name": null,
|
| 757 |
+
"trial_params": null
|
| 758 |
+
}
|