Upload folder using huggingface_hub
Browse files- head.pt +3 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +585 -0
- training_args.bin +3 -0
head.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9de687367ff8fe45475065b8b8947b7b83c29e481111d51ee9fc2645a7b7e4a9
|
| 3 |
+
size 83334
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a49986c22a74efef068d6327f934dd304e03a7a2d7920eba0523f294f176e069
|
| 3 |
+
size 15177808620
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d041cb580f219a0ac2de0df9278c020a88ba447d7318ffeff8af669ce78372f5
|
| 3 |
+
size 336087960
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c01cde4f49c4bb442d7aa76ff25f0943caa915570f932e3c2e67d76d21d0fbac
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3035aa2833cf83634813d0f3fd0b37c38da05a9e61e51aaed87fb19461708f93
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,585 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 37500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04,
|
| 13 |
+
"grad_norm": 30.94066619873047,
|
| 14 |
+
"learning_rate": 6.666666666666667e-06,
|
| 15 |
+
"loss": 2.5555,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.08,
|
| 20 |
+
"grad_norm": 31.642507553100586,
|
| 21 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 22 |
+
"loss": 1.5341,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.12,
|
| 27 |
+
"grad_norm": 15.755248069763184,
|
| 28 |
+
"learning_rate": 2e-05,
|
| 29 |
+
"loss": 0.8868,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.16,
|
| 34 |
+
"grad_norm": 16.84243392944336,
|
| 35 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 36 |
+
"loss": 0.7249,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.2,
|
| 41 |
+
"grad_norm": 16.464881896972656,
|
| 42 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 43 |
+
"loss": 0.6703,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.24,
|
| 48 |
+
"grad_norm": 22.622730255126953,
|
| 49 |
+
"learning_rate": 4e-05,
|
| 50 |
+
"loss": 0.6409,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.28,
|
| 55 |
+
"grad_norm": 36.4688720703125,
|
| 56 |
+
"learning_rate": 4.666666666666667e-05,
|
| 57 |
+
"loss": 0.622,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.32,
|
| 62 |
+
"grad_norm": 26.15373420715332,
|
| 63 |
+
"learning_rate": 4.9993231029486544e-05,
|
| 64 |
+
"loss": 0.6081,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.36,
|
| 69 |
+
"grad_norm": 18.611602783203125,
|
| 70 |
+
"learning_rate": 4.993910125649561e-05,
|
| 71 |
+
"loss": 0.5778,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.4,
|
| 76 |
+
"grad_norm": 8.996405601501465,
|
| 77 |
+
"learning_rate": 4.983095894354858e-05,
|
| 78 |
+
"loss": 0.5566,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.44,
|
| 83 |
+
"grad_norm": 9.431114196777344,
|
| 84 |
+
"learning_rate": 4.966903830281449e-05,
|
| 85 |
+
"loss": 0.5342,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.48,
|
| 90 |
+
"grad_norm": 7.389405250549316,
|
| 91 |
+
"learning_rate": 4.9453690018345144e-05,
|
| 92 |
+
"loss": 0.5065,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.52,
|
| 97 |
+
"grad_norm": 31.81819725036621,
|
| 98 |
+
"learning_rate": 4.9185380486571595e-05,
|
| 99 |
+
"loss": 0.4933,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.56,
|
| 104 |
+
"grad_norm": 0.20836065709590912,
|
| 105 |
+
"learning_rate": 4.88646908061933e-05,
|
| 106 |
+
"loss": 0.4857,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.6,
|
| 111 |
+
"grad_norm": 17.081571578979492,
|
| 112 |
+
"learning_rate": 4.849231551964771e-05,
|
| 113 |
+
"loss": 0.4848,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.64,
|
| 118 |
+
"grad_norm": 18.848472595214844,
|
| 119 |
+
"learning_rate": 4.806906110888606e-05,
|
| 120 |
+
"loss": 0.4648,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.68,
|
| 125 |
+
"grad_norm": 28.703367233276367,
|
| 126 |
+
"learning_rate": 4.759584424871302e-05,
|
| 127 |
+
"loss": 0.4274,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.72,
|
| 132 |
+
"grad_norm": 25.15323829650879,
|
| 133 |
+
"learning_rate": 4.707368982147318e-05,
|
| 134 |
+
"loss": 0.4581,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.76,
|
| 139 |
+
"grad_norm": 24.141220092773438,
|
| 140 |
+
"learning_rate": 4.650372869738414e-05,
|
| 141 |
+
"loss": 0.4331,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.8,
|
| 146 |
+
"grad_norm": 11.99329662322998,
|
| 147 |
+
"learning_rate": 4.588719528532342e-05,
|
| 148 |
+
"loss": 0.4188,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.84,
|
| 153 |
+
"grad_norm": 5.699385166168213,
|
| 154 |
+
"learning_rate": 4.522542485937369e-05,
|
| 155 |
+
"loss": 0.4158,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.88,
|
| 160 |
+
"grad_norm": 22.17208480834961,
|
| 161 |
+
"learning_rate": 4.4519850666916484e-05,
|
| 162 |
+
"loss": 0.4045,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.92,
|
| 167 |
+
"grad_norm": 0.43230658769607544,
|
| 168 |
+
"learning_rate": 4.377200082453749e-05,
|
| 169 |
+
"loss": 0.3884,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.96,
|
| 174 |
+
"grad_norm": 11.466931343078613,
|
| 175 |
+
"learning_rate": 4.2983495008466276e-05,
|
| 176 |
+
"loss": 0.3896,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.0,
|
| 181 |
+
"grad_norm": 15.537298202514648,
|
| 182 |
+
"learning_rate": 4.215604094671835e-05,
|
| 183 |
+
"loss": 0.3913,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.0,
|
| 188 |
+
"eval_accuracy": 0.887,
|
| 189 |
+
"eval_loss": 0.3431779742240906,
|
| 190 |
+
"eval_runtime": 577.7398,
|
| 191 |
+
"eval_samples_per_second": 17.309,
|
| 192 |
+
"eval_steps_per_second": 2.164,
|
| 193 |
+
"step": 12500
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.04,
|
| 197 |
+
"grad_norm": 23.692087173461914,
|
| 198 |
+
"learning_rate": 4.129143072053638e-05,
|
| 199 |
+
"loss": 0.3064,
|
| 200 |
+
"step": 13000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.08,
|
| 204 |
+
"grad_norm": 0.12528562545776367,
|
| 205 |
+
"learning_rate": 4.039153688314145e-05,
|
| 206 |
+
"loss": 0.301,
|
| 207 |
+
"step": 13500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.12,
|
| 211 |
+
"grad_norm": 42.4860725402832,
|
| 212 |
+
"learning_rate": 3.945830840419966e-05,
|
| 213 |
+
"loss": 0.3009,
|
| 214 |
+
"step": 14000
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.16,
|
| 218 |
+
"grad_norm": 15.097381591796875,
|
| 219 |
+
"learning_rate": 3.8493766448787825e-05,
|
| 220 |
+
"loss": 0.3001,
|
| 221 |
+
"step": 14500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.2,
|
| 225 |
+
"grad_norm": 16.759035110473633,
|
| 226 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 227 |
+
"loss": 0.2997,
|
| 228 |
+
"step": 15000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.24,
|
| 232 |
+
"grad_norm": 17.78322410583496,
|
| 233 |
+
"learning_rate": 3.6479161334675296e-05,
|
| 234 |
+
"loss": 0.3082,
|
| 235 |
+
"step": 15500
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.28,
|
| 239 |
+
"grad_norm": 20.591352462768555,
|
| 240 |
+
"learning_rate": 3.543346136204545e-05,
|
| 241 |
+
"loss": 0.2789,
|
| 242 |
+
"step": 16000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.32,
|
| 246 |
+
"grad_norm": 2.979491949081421,
|
| 247 |
+
"learning_rate": 3.436516483539781e-05,
|
| 248 |
+
"loss": 0.299,
|
| 249 |
+
"step": 16500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.3599999999999999,
|
| 253 |
+
"grad_norm": 21.6214542388916,
|
| 254 |
+
"learning_rate": 3.327658544712395e-05,
|
| 255 |
+
"loss": 0.2841,
|
| 256 |
+
"step": 17000
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.4,
|
| 260 |
+
"grad_norm": 12.33953857421875,
|
| 261 |
+
"learning_rate": 3.217008081777726e-05,
|
| 262 |
+
"loss": 0.2803,
|
| 263 |
+
"step": 17500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.44,
|
| 267 |
+
"grad_norm": 1.7421214580535889,
|
| 268 |
+
"learning_rate": 3.104804738999169e-05,
|
| 269 |
+
"loss": 0.2799,
|
| 270 |
+
"step": 18000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.48,
|
| 274 |
+
"grad_norm": 3.7507829666137695,
|
| 275 |
+
"learning_rate": 2.9912915238320754e-05,
|
| 276 |
+
"loss": 0.2969,
|
| 277 |
+
"step": 18500
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.52,
|
| 281 |
+
"grad_norm": 0.3172661364078522,
|
| 282 |
+
"learning_rate": 2.876714280623708e-05,
|
| 283 |
+
"loss": 0.2839,
|
| 284 |
+
"step": 19000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.56,
|
| 288 |
+
"grad_norm": 0.3152337968349457,
|
| 289 |
+
"learning_rate": 2.761321158169134e-05,
|
| 290 |
+
"loss": 0.2934,
|
| 291 |
+
"step": 19500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.6,
|
| 295 |
+
"grad_norm": 6.9163289070129395,
|
| 296 |
+
"learning_rate": 2.6453620722761896e-05,
|
| 297 |
+
"loss": 0.2889,
|
| 298 |
+
"step": 20000
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.6400000000000001,
|
| 302 |
+
"grad_norm": 28.51948356628418,
|
| 303 |
+
"learning_rate": 2.5290881645034932e-05,
|
| 304 |
+
"loss": 0.2709,
|
| 305 |
+
"step": 20500
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.6800000000000002,
|
| 309 |
+
"grad_norm": 6.519890308380127,
|
| 310 |
+
"learning_rate": 2.4127512582437485e-05,
|
| 311 |
+
"loss": 0.2621,
|
| 312 |
+
"step": 21000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.72,
|
| 316 |
+
"grad_norm": 0.005845916923135519,
|
| 317 |
+
"learning_rate": 2.2966033133303545e-05,
|
| 318 |
+
"loss": 0.2588,
|
| 319 |
+
"step": 21500
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.76,
|
| 323 |
+
"grad_norm": 0.0003709389711730182,
|
| 324 |
+
"learning_rate": 2.1808958803485136e-05,
|
| 325 |
+
"loss": 0.2677,
|
| 326 |
+
"step": 22000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.8,
|
| 330 |
+
"grad_norm": 1.661258339881897,
|
| 331 |
+
"learning_rate": 2.0658795558326743e-05,
|
| 332 |
+
"loss": 0.26,
|
| 333 |
+
"step": 22500
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.8399999999999999,
|
| 337 |
+
"grad_norm": 20.34541130065918,
|
| 338 |
+
"learning_rate": 1.9518034395302414e-05,
|
| 339 |
+
"loss": 0.2442,
|
| 340 |
+
"step": 23000
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.88,
|
| 344 |
+
"grad_norm": 0.12481857091188431,
|
| 345 |
+
"learning_rate": 1.838914594906995e-05,
|
| 346 |
+
"loss": 0.2599,
|
| 347 |
+
"step": 23500
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.92,
|
| 351 |
+
"grad_norm": 12.972709655761719,
|
| 352 |
+
"learning_rate": 1.7274575140626318e-05,
|
| 353 |
+
"loss": 0.2513,
|
| 354 |
+
"step": 24000
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.96,
|
| 358 |
+
"grad_norm": 15.054149627685547,
|
| 359 |
+
"learning_rate": 1.617673588215328e-05,
|
| 360 |
+
"loss": 0.2497,
|
| 361 |
+
"step": 24500
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.0,
|
| 365 |
+
"grad_norm": 0.9500556588172913,
|
| 366 |
+
"learning_rate": 1.509800584902108e-05,
|
| 367 |
+
"loss": 0.2686,
|
| 368 |
+
"step": 25000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.0,
|
| 372 |
+
"eval_accuracy": 0.9154,
|
| 373 |
+
"eval_loss": 0.3040485382080078,
|
| 374 |
+
"eval_runtime": 578.6501,
|
| 375 |
+
"eval_samples_per_second": 17.282,
|
| 376 |
+
"eval_steps_per_second": 2.16,
|
| 377 |
+
"step": 25000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.04,
|
| 381 |
+
"grad_norm": 3.1150968074798584,
|
| 382 |
+
"learning_rate": 1.4040721330273062e-05,
|
| 383 |
+
"loss": 0.1121,
|
| 384 |
+
"step": 25500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.08,
|
| 388 |
+
"grad_norm": 1.2094102203263901e-05,
|
| 389 |
+
"learning_rate": 1.3007172168743854e-05,
|
| 390 |
+
"loss": 0.1066,
|
| 391 |
+
"step": 26000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.12,
|
| 395 |
+
"grad_norm": 0.005309335887432098,
|
| 396 |
+
"learning_rate": 1.1999596801769616e-05,
|
| 397 |
+
"loss": 0.104,
|
| 398 |
+
"step": 26500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.16,
|
| 402 |
+
"grad_norm": 0.26751643419265747,
|
| 403 |
+
"learning_rate": 1.1020177413231334e-05,
|
| 404 |
+
"loss": 0.1105,
|
| 405 |
+
"step": 27000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.2,
|
| 409 |
+
"grad_norm": 20.066680908203125,
|
| 410 |
+
"learning_rate": 1.0071035207430352e-05,
|
| 411 |
+
"loss": 0.1103,
|
| 412 |
+
"step": 27500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.24,
|
| 416 |
+
"grad_norm": 9.758234977722168,
|
| 417 |
+
"learning_rate": 9.154225815032242e-06,
|
| 418 |
+
"loss": 0.1216,
|
| 419 |
+
"step": 28000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.2800000000000002,
|
| 423 |
+
"grad_norm": 39.75688934326172,
|
| 424 |
+
"learning_rate": 8.271734841028553e-06,
|
| 425 |
+
"loss": 0.121,
|
| 426 |
+
"step": 28500
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.32,
|
| 430 |
+
"grad_norm": 8.768061637878418,
|
| 431 |
+
"learning_rate": 7.4254735643584564e-06,
|
| 432 |
+
"loss": 0.0926,
|
| 433 |
+
"step": 29000
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.36,
|
| 437 |
+
"grad_norm": 1.5957878828048706,
|
| 438 |
+
"learning_rate": 6.617274798504286e-06,
|
| 439 |
+
"loss": 0.1115,
|
| 440 |
+
"step": 29500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.4,
|
| 444 |
+
"grad_norm": 0.007041580975055695,
|
| 445 |
+
"learning_rate": 5.848888922025553e-06,
|
| 446 |
+
"loss": 0.113,
|
| 447 |
+
"step": 30000
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.44,
|
| 451 |
+
"grad_norm": 0.0010586841963231564,
|
| 452 |
+
"learning_rate": 5.121980087628803e-06,
|
| 453 |
+
"loss": 0.1154,
|
| 454 |
+
"step": 30500
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.48,
|
| 458 |
+
"grad_norm": 10.376145362854004,
|
| 459 |
+
"learning_rate": 4.438122617983443e-06,
|
| 460 |
+
"loss": 0.117,
|
| 461 |
+
"step": 31000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.52,
|
| 465 |
+
"grad_norm": 5.163309992894938e-07,
|
| 466 |
+
"learning_rate": 3.798797596089351e-06,
|
| 467 |
+
"loss": 0.0858,
|
| 468 |
+
"step": 31500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.56,
|
| 472 |
+
"grad_norm": 8.937009260989726e-05,
|
| 473 |
+
"learning_rate": 3.205389657580943e-06,
|
| 474 |
+
"loss": 0.0908,
|
| 475 |
+
"step": 32000
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.6,
|
| 479 |
+
"grad_norm": 65.54885864257812,
|
| 480 |
+
"learning_rate": 2.659183991914696e-06,
|
| 481 |
+
"loss": 0.1016,
|
| 482 |
+
"step": 32500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.64,
|
| 486 |
+
"grad_norm": 9.665397556091193e-06,
|
| 487 |
+
"learning_rate": 2.1613635589349756e-06,
|
| 488 |
+
"loss": 0.081,
|
| 489 |
+
"step": 33000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.68,
|
| 493 |
+
"grad_norm": 0.019450731575489044,
|
| 494 |
+
"learning_rate": 1.713006526846439e-06,
|
| 495 |
+
"loss": 0.1056,
|
| 496 |
+
"step": 33500
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.7199999999999998,
|
| 500 |
+
"grad_norm": 0.04092613607645035,
|
| 501 |
+
"learning_rate": 1.31508393714177e-06,
|
| 502 |
+
"loss": 0.1185,
|
| 503 |
+
"step": 34000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.76,
|
| 507 |
+
"grad_norm": 0.00013055396266281605,
|
| 508 |
+
"learning_rate": 9.684576015420278e-07,
|
| 509 |
+
"loss": 0.1084,
|
| 510 |
+
"step": 34500
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.8,
|
| 514 |
+
"grad_norm": 0.004797887057065964,
|
| 515 |
+
"learning_rate": 6.738782355044049e-07,
|
| 516 |
+
"loss": 0.0973,
|
| 517 |
+
"step": 35000
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.84,
|
| 521 |
+
"grad_norm": 0.0011987409088760614,
|
| 522 |
+
"learning_rate": 4.319838323396691e-07,
|
| 523 |
+
"loss": 0.0883,
|
| 524 |
+
"step": 35500
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.88,
|
| 528 |
+
"grad_norm": 4.8637772124493495e-05,
|
| 529 |
+
"learning_rate": 2.4329828146074095e-07,
|
| 530 |
+
"loss": 0.0966,
|
| 531 |
+
"step": 36000
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.92,
|
| 535 |
+
"grad_norm": 69.71770477294922,
|
| 536 |
+
"learning_rate": 1.0823023375489127e-07,
|
| 537 |
+
"loss": 0.0953,
|
| 538 |
+
"step": 36500
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.96,
|
| 542 |
+
"grad_norm": 3.9461236000061035,
|
| 543 |
+
"learning_rate": 2.7072216536885853e-08,
|
| 544 |
+
"loss": 0.0754,
|
| 545 |
+
"step": 37000
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 3.0,
|
| 549 |
+
"grad_norm": 0.002531010191887617,
|
| 550 |
+
"learning_rate": 0.0,
|
| 551 |
+
"loss": 0.0937,
|
| 552 |
+
"step": 37500
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 3.0,
|
| 556 |
+
"eval_accuracy": 0.9197,
|
| 557 |
+
"eval_loss": 0.49928906559944153,
|
| 558 |
+
"eval_runtime": 577.5752,
|
| 559 |
+
"eval_samples_per_second": 17.314,
|
| 560 |
+
"eval_steps_per_second": 2.164,
|
| 561 |
+
"step": 37500
|
| 562 |
+
}
|
| 563 |
+
],
|
| 564 |
+
"logging_steps": 500,
|
| 565 |
+
"max_steps": 37500,
|
| 566 |
+
"num_input_tokens_seen": 0,
|
| 567 |
+
"num_train_epochs": 3,
|
| 568 |
+
"save_steps": 500,
|
| 569 |
+
"stateful_callbacks": {
|
| 570 |
+
"TrainerControl": {
|
| 571 |
+
"args": {
|
| 572 |
+
"should_epoch_stop": false,
|
| 573 |
+
"should_evaluate": false,
|
| 574 |
+
"should_log": false,
|
| 575 |
+
"should_save": true,
|
| 576 |
+
"should_training_stop": true
|
| 577 |
+
},
|
| 578 |
+
"attributes": {}
|
| 579 |
+
}
|
| 580 |
+
},
|
| 581 |
+
"total_flos": 0.0,
|
| 582 |
+
"train_batch_size": 8,
|
| 583 |
+
"trial_name": null,
|
| 584 |
+
"trial_params": null
|
| 585 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ef1085d1d8d1536642d545d703923c054f76f6e4478afcf8e52afd26e933b80
|
| 3 |
+
size 5496
|