Upload folder using huggingface_hub
Browse files- checkpoints-v3.1-uni/checkpoint-9216/ema.safetensors +3 -0
- checkpoints-v3.1-uni/checkpoint-9216/model.safetensors +3 -0
- checkpoints-v3.1-uni/checkpoint-9216/optimizer.pt +3 -0
- checkpoints-v3.1-uni/checkpoint-9216/rng_state.pth +3 -0
- checkpoints-v3.1-uni/checkpoint-9216/scaler.pt +3 -0
- checkpoints-v3.1-uni/checkpoint-9216/scheduler.pt +3 -0
- checkpoints-v3.1-uni/checkpoint-9216/trainer_state.json +511 -0
- checkpoints-v3.1-uni/checkpoint-9216/training_args.bin +3 -0
checkpoints-v3.1-uni/checkpoint-9216/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a1ac4bb3999c6f5d40f434084e53a1bdd0a592c3883175a49d4d19a49539ddd
|
| 3 |
+
size 550088
|
checkpoints-v3.1-uni/checkpoint-9216/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bee0e89744e0270d7a75fea8e76c7c9968b1113e90a84f4ba3aa915cb486e0f2
|
| 3 |
+
size 19318680
|
checkpoints-v3.1-uni/checkpoint-9216/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2438b92b3d899903b1f95fb95585b4b01f721403b6af414db79edcfe157325cc
|
| 3 |
+
size 1176715
|
checkpoints-v3.1-uni/checkpoint-9216/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:720b47cab6dc34c6a2929f41e2532467f65fe93b04e5db09e6d13200325084af
|
| 3 |
+
size 14645
|
checkpoints-v3.1-uni/checkpoint-9216/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51747e333ccb561efb3e4ddba2406abc43ec169e287fbb5ac91b1c526ca7f263
|
| 3 |
+
size 1383
|
checkpoints-v3.1-uni/checkpoint-9216/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61ec26aa4e6dd5142e5325eee770e44f77900604245ac526990154cf3ed52246
|
| 3 |
+
size 1465
|
checkpoints-v3.1-uni/checkpoint-9216/trainer_state.json
ADDED
|
@@ -0,0 +1,511 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.4256616322571706,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 9216,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 0.07098234444856644,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 0.20588445663452148,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.22737464308738708,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 0.17480437457561493,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.12320689857006073,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 0.15027859807014465,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.20722734928131104,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 0.1415732204914093,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.04582402333362053,
|
| 43 |
+
"eval_loss": 0.09126683337333268,
|
| 44 |
+
"eval_mse_loss": 0.09126683337333268,
|
| 45 |
+
"flow/cos_sim": 0.9541759895135279,
|
| 46 |
+
"flow/improvement_ratio": 0.9938432111587698,
|
| 47 |
+
"flow/mag_ratio_mean": 0.8994761448200435,
|
| 48 |
+
"flow/mag_ratio_std": 0.11755070905094822,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.04582402333362053,
|
| 54 |
+
"eval_loss": 0.09126683337333268,
|
| 55 |
+
"eval_mse_loss": 0.09126683337333268,
|
| 56 |
+
"eval_runtime": 37.4141,
|
| 57 |
+
"eval_samples_per_second": 748.193,
|
| 58 |
+
"eval_steps_per_second": 11.707,
|
| 59 |
+
"flow/cos_sim": 0.9541759895135279,
|
| 60 |
+
"flow/improvement_ratio": 0.9938432111587698,
|
| 61 |
+
"flow/mag_ratio_mean": 0.8994761448200435,
|
| 62 |
+
"flow/mag_ratio_std": 0.11755070905094822,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.12740713357925415,
|
| 68 |
+
"learning_rate": 0.0009967551747861387,
|
| 69 |
+
"loss": 0.13867560029029846,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.49607324600219727,
|
| 75 |
+
"learning_rate": 0.000994232528651847,
|
| 76 |
+
"loss": 0.13481874763965607,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.26182159781455994,
|
| 82 |
+
"learning_rate": 0.0009909945800260092,
|
| 83 |
+
"loss": 0.13546937704086304,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.12674923241138458,
|
| 89 |
+
"learning_rate": 0.0009870460151900522,
|
| 90 |
+
"loss": 0.13297873735427856,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.04454593354470381,
|
| 96 |
+
"eval_loss": 0.08455971572292995,
|
| 97 |
+
"eval_mse_loss": 0.08455971572292995,
|
| 98 |
+
"flow/cos_sim": 0.9554540837464267,
|
| 99 |
+
"flow/improvement_ratio": 0.9942142926666835,
|
| 100 |
+
"flow/mag_ratio_mean": 0.9197173737771979,
|
| 101 |
+
"flow/mag_ratio_std": 0.1105972074448518,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.04454593354470381,
|
| 107 |
+
"eval_loss": 0.08455971572292995,
|
| 108 |
+
"eval_mse_loss": 0.08455971572292995,
|
| 109 |
+
"eval_runtime": 36.8159,
|
| 110 |
+
"eval_samples_per_second": 760.352,
|
| 111 |
+
"eval_steps_per_second": 11.897,
|
| 112 |
+
"flow/cos_sim": 0.9554540837464267,
|
| 113 |
+
"flow/improvement_ratio": 0.9942142926666835,
|
| 114 |
+
"flow/mag_ratio_mean": 0.9197173737771979,
|
| 115 |
+
"flow/mag_ratio_std": 0.1105972074448518,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.2195156067609787,
|
| 121 |
+
"learning_rate": 0.0009823925488998885,
|
| 122 |
+
"loss": 0.1306387186050415,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.13101470470428467,
|
| 128 |
+
"learning_rate": 0.0009770409161149525,
|
| 129 |
+
"loss": 0.12852290272712708,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.1463838666677475,
|
| 135 |
+
"learning_rate": 0.0009709988622506973,
|
| 136 |
+
"loss": 0.1289902925491333,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.1557082235813141,
|
| 142 |
+
"learning_rate": 0.000964275131968659,
|
| 143 |
+
"loss": 0.12751106917858124,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.04429665551474105,
|
| 149 |
+
"eval_loss": 0.08277356523391047,
|
| 150 |
+
"eval_mse_loss": 0.08277356523391047,
|
| 151 |
+
"flow/cos_sim": 0.9557033510512958,
|
| 152 |
+
"flow/improvement_ratio": 0.9939037547808259,
|
| 153 |
+
"flow/mag_ratio_mean": 0.92809867015168,
|
| 154 |
+
"flow/mag_ratio_std": 0.10807815949394278,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.04429665551474105,
|
| 160 |
+
"eval_loss": 0.08277356523391047,
|
| 161 |
+
"eval_mse_loss": 0.08277356523391047,
|
| 162 |
+
"eval_runtime": 36.8301,
|
| 163 |
+
"eval_samples_per_second": 760.058,
|
| 164 |
+
"eval_steps_per_second": 11.892,
|
| 165 |
+
"flow/cos_sim": 0.9557033510512958,
|
| 166 |
+
"flow/improvement_ratio": 0.9939037547808259,
|
| 167 |
+
"flow/mag_ratio_mean": 0.92809867015168,
|
| 168 |
+
"flow/mag_ratio_std": 0.10807815949394278,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.165770024061203,
|
| 174 |
+
"learning_rate": 0.0009568794565203123,
|
| 175 |
+
"loss": 0.12700329720973969,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.17231467366218567,
|
| 181 |
+
"learning_rate": 0.0009488225396630347,
|
| 182 |
+
"loss": 0.12523098289966583,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.1858324557542801,
|
| 188 |
+
"learning_rate": 0.0009401160421685646,
|
| 189 |
+
"loss": 0.12655843794345856,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 0.11940455436706543,
|
| 195 |
+
"learning_rate": 0.0009307725649463714,
|
| 196 |
+
"loss": 0.12418897449970245,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.04446321669710825,
|
| 202 |
+
"eval_loss": 0.08204046752490954,
|
| 203 |
+
"eval_mse_loss": 0.08204046752490954,
|
| 204 |
+
"flow/cos_sim": 0.9555367968942476,
|
| 205 |
+
"flow/improvement_ratio": 0.9933926257640804,
|
| 206 |
+
"flow/mag_ratio_mean": 0.9328070380916335,
|
| 207 |
+
"flow/mag_ratio_std": 0.10609594972370422,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.04446321669710825,
|
| 213 |
+
"eval_loss": 0.08204046752490954,
|
| 214 |
+
"eval_mse_loss": 0.08204046752490954,
|
| 215 |
+
"eval_runtime": 36.8821,
|
| 216 |
+
"eval_samples_per_second": 758.987,
|
| 217 |
+
"eval_steps_per_second": 11.876,
|
| 218 |
+
"flow/cos_sim": 0.9555367968942476,
|
| 219 |
+
"flow/improvement_ratio": 0.9933926257640804,
|
| 220 |
+
"flow/mag_ratio_mean": 0.9328070380916335,
|
| 221 |
+
"flow/mag_ratio_std": 0.10609594972370422,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.136522576212883,
|
| 227 |
+
"learning_rate": 0.0009208056308063659,
|
| 228 |
+
"loss": 0.12340260297060013,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.18442140519618988,
|
| 234 |
+
"learning_rate": 0.0009102296648873445,
|
| 235 |
+
"loss": 0.12370757013559341,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.16970551013946533,
|
| 241 |
+
"learning_rate": 0.0008990599737794927,
|
| 242 |
+
"loss": 0.12366028130054474,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.13654212653636932,
|
| 248 |
+
"learning_rate": 0.0008873127233711644,
|
| 249 |
+
"loss": 0.12185485661029816,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.043456991479549234,
|
| 255 |
+
"eval_loss": 0.0809854866945308,
|
| 256 |
+
"eval_mse_loss": 0.0809854866945308,
|
| 257 |
+
"flow/cos_sim": 0.9565430256329714,
|
| 258 |
+
"flow/improvement_ratio": 0.9940090940151041,
|
| 259 |
+
"flow/mag_ratio_mean": 0.9294795245612593,
|
| 260 |
+
"flow/mag_ratio_std": 0.10000736579367014,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.043456991479549234,
|
| 266 |
+
"eval_loss": 0.0809854866945308,
|
| 267 |
+
"eval_mse_loss": 0.0809854866945308,
|
| 268 |
+
"eval_runtime": 37.5627,
|
| 269 |
+
"eval_samples_per_second": 745.234,
|
| 270 |
+
"eval_steps_per_second": 11.661,
|
| 271 |
+
"flow/cos_sim": 0.9565430256329714,
|
| 272 |
+
"flow/improvement_ratio": 0.9940090940151041,
|
| 273 |
+
"flow/mag_ratio_mean": 0.9294795245612593,
|
| 274 |
+
"flow/mag_ratio_std": 0.10000736579367014,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.1422841101884842,
|
| 280 |
+
"learning_rate": 0.0008750049154520011,
|
| 281 |
+
"loss": 0.12234608083963394,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.24812348186969757,
|
| 287 |
+
"learning_rate": 0.0008621543631062487,
|
| 288 |
+
"loss": 0.12298071384429932,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.16056349873542786,
|
| 294 |
+
"learning_rate": 0.0008487796649318904,
|
| 295 |
+
"loss": 0.12491145730018616,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.1158125102519989,
|
| 301 |
+
"learning_rate": 0.0008349001781229053,
|
| 302 |
+
"loss": 0.12146689742803574,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.04404348295743335,
|
| 308 |
+
"eval_loss": 0.08143461094191086,
|
| 309 |
+
"eval_mse_loss": 0.08143461094191086,
|
| 310 |
+
"flow/cos_sim": 0.955956531578003,
|
| 311 |
+
"flow/improvement_ratio": 0.9938269595032958,
|
| 312 |
+
"flow/mag_ratio_mean": 0.9270812613234672,
|
| 313 |
+
"flow/mag_ratio_std": 0.10172502801128447,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.04404348295743335,
|
| 319 |
+
"eval_loss": 0.08143461094191086,
|
| 320 |
+
"eval_mse_loss": 0.08143461094191086,
|
| 321 |
+
"eval_runtime": 37.2365,
|
| 322 |
+
"eval_samples_per_second": 751.762,
|
| 323 |
+
"eval_steps_per_second": 11.763,
|
| 324 |
+
"flow/cos_sim": 0.955956531578003,
|
| 325 |
+
"flow/improvement_ratio": 0.9938269595032958,
|
| 326 |
+
"flow/mag_ratio_mean": 0.9270812613234672,
|
| 327 |
+
"flow/mag_ratio_std": 0.10172502801128447,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.12894846498966217,
|
| 333 |
+
"learning_rate": 0.0008205359904536107,
|
| 334 |
+
"loss": 0.12104764580726624,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.1334332525730133,
|
| 340 |
+
"learning_rate": 0.0008057078912056363,
|
| 341 |
+
"loss": 0.12087016552686691,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.134778693318367,
|
| 347 |
+
"learning_rate": 0.0007904373410796086,
|
| 348 |
+
"loss": 0.12076316028833389,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.11102022230625153,
|
| 354 |
+
"learning_rate": 0.0007747464411350876,
|
| 355 |
+
"loss": 0.12042346596717834,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.04294020341938897,
|
| 361 |
+
"eval_loss": 0.07952374037882509,
|
| 362 |
+
"eval_mse_loss": 0.07952374037882509,
|
| 363 |
+
"flow/cos_sim": 0.9570598111032895,
|
| 364 |
+
"flow/improvement_ratio": 0.9943427937488033,
|
| 365 |
+
"flow/mag_ratio_mean": 0.9337734204449065,
|
| 366 |
+
"flow/mag_ratio_std": 0.09960645821676951,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.04294020341938897,
|
| 372 |
+
"eval_loss": 0.07952374037882509,
|
| 373 |
+
"eval_mse_loss": 0.07952374037882509,
|
| 374 |
+
"eval_runtime": 37.5482,
|
| 375 |
+
"eval_samples_per_second": 745.522,
|
| 376 |
+
"eval_steps_per_second": 11.665,
|
| 377 |
+
"flow/cos_sim": 0.9570598111032895,
|
| 378 |
+
"flow/improvement_ratio": 0.9943427937488033,
|
| 379 |
+
"flow/mag_ratio_mean": 0.9337734204449065,
|
| 380 |
+
"flow/mag_ratio_std": 0.09960645821676951,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.7088926434516907,
|
| 386 |
+
"learning_rate": 0.000758657900803716,
|
| 387 |
+
"loss": 0.12144383788108826,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.12531471252441406,
|
| 393 |
+
"learning_rate": 0.000742195005021869,
|
| 394 |
+
"loss": 0.12000911682844162,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.12283795326948166,
|
| 400 |
+
"learning_rate": 0.0007253815805303786,
|
| 401 |
+
"loss": 0.12044861167669296,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.10739121586084366,
|
| 407 |
+
"learning_rate": 0.0007082419613901028,
|
| 408 |
+
"loss": 0.12056493014097214,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.04426976554435954,
|
| 414 |
+
"eval_loss": 0.08079966172984202,
|
| 415 |
+
"eval_mse_loss": 0.08079966172984202,
|
| 416 |
+
"flow/cos_sim": 0.9557302452657865,
|
| 417 |
+
"flow/improvement_ratio": 0.9934861994769475,
|
| 418 |
+
"flow/mag_ratio_mean": 0.9359486946776577,
|
| 419 |
+
"flow/mag_ratio_std": 0.10338757488213174,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.04426976554435954,
|
| 425 |
+
"eval_loss": 0.08079966172984202,
|
| 426 |
+
"eval_mse_loss": 0.08079966172984202,
|
| 427 |
+
"eval_runtime": 37.3561,
|
| 428 |
+
"eval_samples_per_second": 749.356,
|
| 429 |
+
"eval_steps_per_second": 11.725,
|
| 430 |
+
"flow/cos_sim": 0.9557302452657865,
|
| 431 |
+
"flow/improvement_ratio": 0.9934861994769475,
|
| 432 |
+
"flow/mag_ratio_mean": 0.9359486946776577,
|
| 433 |
+
"flow/mag_ratio_std": 0.10338757488213174,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.12729094922542572,
|
| 439 |
+
"learning_rate": 0.0006908009537632514,
|
| 440 |
+
"loss": 0.11962540447711945,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.16566377878189087,
|
| 446 |
+
"learning_rate": 0.0006730838000114403,
|
| 447 |
+
"loss": 0.12100031971931458,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.14468614757061005,
|
| 453 |
+
"learning_rate": 0.0006551161421624341,
|
| 454 |
+
"loss": 0.12011279165744781,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 0.1833762377500534,
|
| 460 |
+
"learning_rate": 0.0006369239847984517,
|
| 461 |
+
"loss": 0.11977369338274002,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.04329353226901598,
|
| 467 |
+
"eval_loss": 0.07977866525145154,
|
| 468 |
+
"eval_mse_loss": 0.07977866525145154,
|
| 469 |
+
"flow/cos_sim": 0.9567064815732442,
|
| 470 |
+
"flow/improvement_ratio": 0.9943556535189555,
|
| 471 |
+
"flow/mag_ratio_mean": 0.9365287250307597,
|
| 472 |
+
"flow/mag_ratio_std": 0.10248907555115822,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.04329353226901598,
|
| 478 |
+
"eval_loss": 0.07977866525145154,
|
| 479 |
+
"eval_mse_loss": 0.07977866525145154,
|
| 480 |
+
"eval_runtime": 37.6413,
|
| 481 |
+
"eval_samples_per_second": 743.678,
|
| 482 |
+
"eval_steps_per_second": 11.636,
|
| 483 |
+
"flow/cos_sim": 0.9567064815732442,
|
| 484 |
+
"flow/improvement_ratio": 0.9943556535189555,
|
| 485 |
+
"flow/mag_ratio_mean": 0.9365287250307597,
|
| 486 |
+
"flow/mag_ratio_std": 0.10248907555115822,
|
| 487 |
+
"step": 9216
|
| 488 |
+
}
|
| 489 |
+
],
|
| 490 |
+
"logging_steps": 256,
|
| 491 |
+
"max_steps": 21651,
|
| 492 |
+
"num_input_tokens_seen": 0,
|
| 493 |
+
"num_train_epochs": 1,
|
| 494 |
+
"save_steps": 1024,
|
| 495 |
+
"stateful_callbacks": {
|
| 496 |
+
"TrainerControl": {
|
| 497 |
+
"args": {
|
| 498 |
+
"should_epoch_stop": false,
|
| 499 |
+
"should_evaluate": false,
|
| 500 |
+
"should_log": false,
|
| 501 |
+
"should_save": true,
|
| 502 |
+
"should_training_stop": false
|
| 503 |
+
},
|
| 504 |
+
"attributes": {}
|
| 505 |
+
}
|
| 506 |
+
},
|
| 507 |
+
"total_flos": 0.0,
|
| 508 |
+
"train_batch_size": 64,
|
| 509 |
+
"trial_name": null,
|
| 510 |
+
"trial_params": null
|
| 511 |
+
}
|
checkpoints-v3.1-uni/checkpoint-9216/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee9cb7ebeac20164f12c8a5bf77bbe01722a7871151494aba288a393bf7de081
|
| 3 |
+
size 5137
|