Upload folder using huggingface_hub
Browse files- checkpoints/checkpoint-1010/model.safetensors +3 -0
- checkpoints/checkpoint-1010/optimizer.pt +3 -0
- checkpoints/checkpoint-1010/rng_state.pth +3 -0
- checkpoints/checkpoint-1010/scheduler.pt +3 -0
- checkpoints/checkpoint-1010/trainer_state.json +569 -0
- checkpoints/checkpoint-1010/training_args.bin +3 -0
checkpoints/checkpoint-1010/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a0b276628fe265b1da89742723d1e531e444e357b59e2e04a3ca1ba1ad1f0f4
|
| 3 |
+
size 6020469052
|
checkpoints/checkpoint-1010/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8057b78d7969d06813f11870f3f638ce999f3cfbae9c8991184580d0c545398c
|
| 3 |
+
size 204549186
|
checkpoints/checkpoint-1010/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb9d3a5dcef41265354a635ed4d257650889c2c54bf71381fb8a48f7bea6654e
|
| 3 |
+
size 14180
|
checkpoints/checkpoint-1010/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:babddefb328e67e86fb7e81945db16146b67f693469048a8e2b8bb25bdeca0d9
|
| 3 |
+
size 1064
|
checkpoints/checkpoint-1010/trainer_state.json
ADDED
|
@@ -0,0 +1,569 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.987654320987654,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1010,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.07901234567901234,
|
| 13 |
+
"grad_norm": 24.433338165283203,
|
| 14 |
+
"learning_rate": 3.1683168316831685e-06,
|
| 15 |
+
"loss": 10.4193,
|
| 16 |
+
"step": 16
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1580246913580247,
|
| 20 |
+
"grad_norm": 12.38017749786377,
|
| 21 |
+
"learning_rate": 6.336633663366337e-06,
|
| 22 |
+
"loss": 9.306,
|
| 23 |
+
"step": 32
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.23703703703703705,
|
| 27 |
+
"grad_norm": 9.08733081817627,
|
| 28 |
+
"learning_rate": 9.504950495049505e-06,
|
| 29 |
+
"loss": 8.5947,
|
| 30 |
+
"step": 48
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.3160493827160494,
|
| 34 |
+
"grad_norm": 5.627976894378662,
|
| 35 |
+
"learning_rate": 1.2673267326732674e-05,
|
| 36 |
+
"loss": 8.3942,
|
| 37 |
+
"step": 64
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.3950617283950617,
|
| 41 |
+
"grad_norm": 4.160132884979248,
|
| 42 |
+
"learning_rate": 1.5841584158415843e-05,
|
| 43 |
+
"loss": 8.3317,
|
| 44 |
+
"step": 80
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.4740740740740741,
|
| 48 |
+
"grad_norm": 3.8646466732025146,
|
| 49 |
+
"learning_rate": 1.900990099009901e-05,
|
| 50 |
+
"loss": 8.2782,
|
| 51 |
+
"step": 96
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.5530864197530864,
|
| 55 |
+
"grad_norm": 3.153797149658203,
|
| 56 |
+
"learning_rate": 1.999277438119978e-05,
|
| 57 |
+
"loss": 8.2797,
|
| 58 |
+
"step": 112
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.6320987654320988,
|
| 62 |
+
"grad_norm": 3.0883467197418213,
|
| 63 |
+
"learning_rate": 1.995649347969019e-05,
|
| 64 |
+
"loss": 8.2899,
|
| 65 |
+
"step": 128
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.7111111111111111,
|
| 69 |
+
"grad_norm": 2.2813191413879395,
|
| 70 |
+
"learning_rate": 1.9889775168565942e-05,
|
| 71 |
+
"loss": 8.2603,
|
| 72 |
+
"step": 144
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.7901234567901234,
|
| 76 |
+
"grad_norm": 1.9534201622009277,
|
| 77 |
+
"learning_rate": 1.9792823408445173e-05,
|
| 78 |
+
"loss": 8.2206,
|
| 79 |
+
"step": 160
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.8691358024691358,
|
| 83 |
+
"grad_norm": 2.1486434936523438,
|
| 84 |
+
"learning_rate": 1.966593458484168e-05,
|
| 85 |
+
"loss": 8.2437,
|
| 86 |
+
"step": 176
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.9481481481481482,
|
| 90 |
+
"grad_norm": 1.717513918876648,
|
| 91 |
+
"learning_rate": 1.9509496602102253e-05,
|
| 92 |
+
"loss": 8.2006,
|
| 93 |
+
"step": 192
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.9975308641975309,
|
| 97 |
+
"eval_bleu": 0.04760109116359015,
|
| 98 |
+
"eval_cap_loss": 2.677484789315392,
|
| 99 |
+
"eval_con_loss": 2.058648048662672,
|
| 100 |
+
"eval_loss": 6.794780893653047,
|
| 101 |
+
"step": 202
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.9975308641975309,
|
| 105 |
+
"eval_bleu": 0.04760109116359015,
|
| 106 |
+
"eval_cap_loss": 2.677484789315392,
|
| 107 |
+
"eval_con_loss": 2.058648048662672,
|
| 108 |
+
"eval_loss": 6.794780893653047,
|
| 109 |
+
"eval_runtime": 78.7662,
|
| 110 |
+
"eval_samples_per_second": 10.271,
|
| 111 |
+
"eval_steps_per_second": 1.295,
|
| 112 |
+
"step": 202
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.0271604938271606,
|
| 116 |
+
"grad_norm": 1.9594453573226929,
|
| 117 |
+
"learning_rate": 1.932398769756714e-05,
|
| 118 |
+
"loss": 8.1451,
|
| 119 |
+
"step": 208
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.106172839506173,
|
| 123 |
+
"grad_norm": 1.848238229751587,
|
| 124 |
+
"learning_rate": 1.9109974979578852e-05,
|
| 125 |
+
"loss": 8.1435,
|
| 126 |
+
"step": 224
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1851851851851851,
|
| 130 |
+
"grad_norm": 1.7789127826690674,
|
| 131 |
+
"learning_rate": 1.8868112693808664e-05,
|
| 132 |
+
"loss": 8.171,
|
| 133 |
+
"step": 240
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.2641975308641975,
|
| 137 |
+
"grad_norm": 1.5801581144332886,
|
| 138 |
+
"learning_rate": 1.8599140223200716e-05,
|
| 139 |
+
"loss": 8.1993,
|
| 140 |
+
"step": 256
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.34320987654321,
|
| 144 |
+
"grad_norm": 1.7825208902359009,
|
| 145 |
+
"learning_rate": 1.8303879827647977e-05,
|
| 146 |
+
"loss": 8.1742,
|
| 147 |
+
"step": 272
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.4222222222222223,
|
| 151 |
+
"grad_norm": 1.8662821054458618,
|
| 152 |
+
"learning_rate": 1.798323413030997e-05,
|
| 153 |
+
"loss": 8.152,
|
| 154 |
+
"step": 288
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.5012345679012347,
|
| 158 |
+
"grad_norm": 1.6635081768035889,
|
| 159 |
+
"learning_rate": 1.76381833582567e-05,
|
| 160 |
+
"loss": 8.1432,
|
| 161 |
+
"step": 304
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.5802469135802468,
|
| 165 |
+
"grad_norm": 1.480753779411316,
|
| 166 |
+
"learning_rate": 1.7269782345874204e-05,
|
| 167 |
+
"loss": 8.123,
|
| 168 |
+
"step": 320
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.6592592592592592,
|
| 172 |
+
"grad_norm": 1.7623605728149414,
|
| 173 |
+
"learning_rate": 1.6879157310192537e-05,
|
| 174 |
+
"loss": 8.1464,
|
| 175 |
+
"step": 336
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.7382716049382716,
|
| 179 |
+
"grad_norm": 1.7219517230987549,
|
| 180 |
+
"learning_rate": 1.6467502407993995e-05,
|
| 181 |
+
"loss": 8.1435,
|
| 182 |
+
"step": 352
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.817283950617284,
|
| 186 |
+
"grad_norm": 1.6134917736053467,
|
| 187 |
+
"learning_rate": 1.6036076085226813e-05,
|
| 188 |
+
"loss": 8.154,
|
| 189 |
+
"step": 368
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.8962962962962964,
|
| 193 |
+
"grad_norm": 1.7360373735427856,
|
| 194 |
+
"learning_rate": 1.5586197229884185e-05,
|
| 195 |
+
"loss": 8.1822,
|
| 196 |
+
"step": 384
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.9753086419753085,
|
| 200 |
+
"grad_norm": 1.8116141557693481,
|
| 201 |
+
"learning_rate": 1.5119241140109466e-05,
|
| 202 |
+
"loss": 8.1585,
|
| 203 |
+
"step": 400
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 2.0,
|
| 207 |
+
"eval_bleu": 0.048510890987172074,
|
| 208 |
+
"eval_cap_loss": 2.6132850997588215,
|
| 209 |
+
"eval_con_loss": 2.058407783508301,
|
| 210 |
+
"eval_loss": 6.730100673787734,
|
| 211 |
+
"step": 405
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 2.0,
|
| 215 |
+
"eval_bleu": 0.048510890987172074,
|
| 216 |
+
"eval_cap_loss": 2.6132850997588215,
|
| 217 |
+
"eval_con_loss": 2.058407783508301,
|
| 218 |
+
"eval_loss": 6.730100673787734,
|
| 219 |
+
"eval_runtime": 82.3042,
|
| 220 |
+
"eval_samples_per_second": 9.829,
|
| 221 |
+
"eval_steps_per_second": 1.239,
|
| 222 |
+
"step": 405
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 2.054320987654321,
|
| 226 |
+
"grad_norm": 1.8309221267700195,
|
| 227 |
+
"learning_rate": 1.4636635319853274e-05,
|
| 228 |
+
"loss": 8.1317,
|
| 229 |
+
"step": 416
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 2.1333333333333333,
|
| 233 |
+
"grad_norm": 1.6076328754425049,
|
| 234 |
+
"learning_rate": 1.4139855114935253e-05,
|
| 235 |
+
"loss": 8.1451,
|
| 236 |
+
"step": 432
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.212345679012346,
|
| 240 |
+
"grad_norm": 1.7202519178390503,
|
| 241 |
+
"learning_rate": 1.3630419202851287e-05,
|
| 242 |
+
"loss": 8.1401,
|
| 243 |
+
"step": 448
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 2.291358024691358,
|
| 247 |
+
"grad_norm": 1.7173830270767212,
|
| 248 |
+
"learning_rate": 1.3109884950114007e-05,
|
| 249 |
+
"loss": 8.1416,
|
| 250 |
+
"step": 464
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 2.3703703703703702,
|
| 254 |
+
"grad_norm": 1.7003859281539917,
|
| 255 |
+
"learning_rate": 1.2579843651319382e-05,
|
| 256 |
+
"loss": 8.1209,
|
| 257 |
+
"step": 480
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 2.449382716049383,
|
| 261 |
+
"grad_norm": 1.6290701627731323,
|
| 262 |
+
"learning_rate": 1.2041915664493763e-05,
|
| 263 |
+
"loss": 8.1018,
|
| 264 |
+
"step": 496
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 2.528395061728395,
|
| 268 |
+
"grad_norm": 1.78144371509552,
|
| 269 |
+
"learning_rate": 1.1497745457592817e-05,
|
| 270 |
+
"loss": 8.1039,
|
| 271 |
+
"step": 512
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 2.6074074074074076,
|
| 275 |
+
"grad_norm": 1.551781415939331,
|
| 276 |
+
"learning_rate": 1.0948996581295437e-05,
|
| 277 |
+
"loss": 8.1037,
|
| 278 |
+
"step": 528
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.68641975308642,
|
| 282 |
+
"grad_norm": 2.156895637512207,
|
| 283 |
+
"learning_rate": 1.0397346583460972e-05,
|
| 284 |
+
"loss": 8.1212,
|
| 285 |
+
"step": 544
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 2.765432098765432,
|
| 289 |
+
"grad_norm": 1.6621625423431396,
|
| 290 |
+
"learning_rate": 9.844481880796492e-06,
|
| 291 |
+
"loss": 8.1438,
|
| 292 |
+
"step": 560
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 2.8444444444444446,
|
| 296 |
+
"grad_norm": 1.603384017944336,
|
| 297 |
+
"learning_rate": 9.292092603411642e-06,
|
| 298 |
+
"loss": 8.0666,
|
| 299 |
+
"step": 576
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 2.9234567901234567,
|
| 303 |
+
"grad_norm": 1.8245571851730347,
|
| 304 |
+
"learning_rate": 8.741867428021447e-06,
|
| 305 |
+
"loss": 8.1017,
|
| 306 |
+
"step": 592
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 2.9975308641975307,
|
| 310 |
+
"eval_bleu": 0.05650729222483089,
|
| 311 |
+
"eval_cap_loss": 2.583449826521032,
|
| 312 |
+
"eval_con_loss": 2.0582562147402297,
|
| 313 |
+
"eval_loss": 6.699962258338928,
|
| 314 |
+
"step": 607
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.9975308641975307,
|
| 318 |
+
"eval_bleu": 0.05650729222483089,
|
| 319 |
+
"eval_cap_loss": 2.583449826521032,
|
| 320 |
+
"eval_con_loss": 2.0582562147402297,
|
| 321 |
+
"eval_loss": 6.699962258338928,
|
| 322 |
+
"eval_runtime": 81.7977,
|
| 323 |
+
"eval_samples_per_second": 9.89,
|
| 324 |
+
"eval_steps_per_second": 1.247,
|
| 325 |
+
"step": 607
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 3.0024691358024693,
|
| 329 |
+
"grad_norm": 2.2745141983032227,
|
| 330 |
+
"learning_rate": 8.195488415592238e-06,
|
| 331 |
+
"loss": 8.0643,
|
| 332 |
+
"step": 608
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 3.0814814814814815,
|
| 336 |
+
"grad_norm": 1.972432017326355,
|
| 337 |
+
"learning_rate": 7.654625869212147e-06,
|
| 338 |
+
"loss": 8.09,
|
| 339 |
+
"step": 624
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 3.1604938271604937,
|
| 343 |
+
"grad_norm": 1.9584094285964966,
|
| 344 |
+
"learning_rate": 7.120933227905971e-06,
|
| 345 |
+
"loss": 8.1158,
|
| 346 |
+
"step": 640
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 3.2395061728395063,
|
| 350 |
+
"grad_norm": 1.9004019498825073,
|
| 351 |
+
"learning_rate": 6.59604201200412e-06,
|
| 352 |
+
"loss": 8.1292,
|
| 353 |
+
"step": 656
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 3.3185185185185184,
|
| 357 |
+
"grad_norm": 1.9189977645874023,
|
| 358 |
+
"learning_rate": 6.081556835517955e-06,
|
| 359 |
+
"loss": 8.1141,
|
| 360 |
+
"step": 672
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 3.397530864197531,
|
| 364 |
+
"grad_norm": 2.1858201026916504,
|
| 365 |
+
"learning_rate": 5.579050500768837e-06,
|
| 366 |
+
"loss": 8.0591,
|
| 367 |
+
"step": 688
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 3.476543209876543,
|
| 371 |
+
"grad_norm": 1.8498514890670776,
|
| 372 |
+
"learning_rate": 5.090059190266779e-06,
|
| 373 |
+
"loss": 8.0868,
|
| 374 |
+
"step": 704
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.5555555555555554,
|
| 378 |
+
"grad_norm": 1.8132171630859375,
|
| 379 |
+
"learning_rate": 4.616077770537453e-06,
|
| 380 |
+
"loss": 8.1157,
|
| 381 |
+
"step": 720
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 3.634567901234568,
|
| 385 |
+
"grad_norm": 1.8233318328857422,
|
| 386 |
+
"learning_rate": 4.158555222253772e-06,
|
| 387 |
+
"loss": 8.0593,
|
| 388 |
+
"step": 736
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 3.71358024691358,
|
| 392 |
+
"grad_norm": 2.037637710571289,
|
| 393 |
+
"learning_rate": 3.718890210642442e-06,
|
| 394 |
+
"loss": 8.0928,
|
| 395 |
+
"step": 752
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.7925925925925927,
|
| 399 |
+
"grad_norm": 2.0393972396850586,
|
| 400 |
+
"learning_rate": 3.2984268097069284e-06,
|
| 401 |
+
"loss": 8.1042,
|
| 402 |
+
"step": 768
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.871604938271605,
|
| 406 |
+
"grad_norm": 2.3308756351470947,
|
| 407 |
+
"learning_rate": 2.898450393337977e-06,
|
| 408 |
+
"loss": 8.0735,
|
| 409 |
+
"step": 784
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.950617283950617,
|
| 413 |
+
"grad_norm": 2.4082491397857666,
|
| 414 |
+
"learning_rate": 2.5201837058728506e-06,
|
| 415 |
+
"loss": 8.0676,
|
| 416 |
+
"step": 800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.0,
|
| 420 |
+
"eval_bleu": 0.05697650729846282,
|
| 421 |
+
"eval_cap_loss": 2.558951176849066,
|
| 422 |
+
"eval_con_loss": 2.0581041083616367,
|
| 423 |
+
"eval_loss": 6.675159391234903,
|
| 424 |
+
"step": 810
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 4.0,
|
| 428 |
+
"eval_bleu": 0.05697650729846282,
|
| 429 |
+
"eval_cap_loss": 2.558951176849066,
|
| 430 |
+
"eval_con_loss": 2.0581041083616367,
|
| 431 |
+
"eval_loss": 6.675159391234903,
|
| 432 |
+
"eval_runtime": 82.0084,
|
| 433 |
+
"eval_samples_per_second": 9.865,
|
| 434 |
+
"eval_steps_per_second": 1.244,
|
| 435 |
+
"step": 810
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 4.029629629629629,
|
| 439 |
+
"grad_norm": 1.9233280420303345,
|
| 440 |
+
"learning_rate": 2.1647831241156304e-06,
|
| 441 |
+
"loss": 8.036,
|
| 442 |
+
"step": 816
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.108641975308642,
|
| 446 |
+
"grad_norm": 2.235079288482666,
|
| 447 |
+
"learning_rate": 1.8333351222458407e-06,
|
| 448 |
+
"loss": 8.118,
|
| 449 |
+
"step": 832
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 4.187654320987654,
|
| 453 |
+
"grad_norm": 1.9698057174682617,
|
| 454 |
+
"learning_rate": 1.5268529504222262e-06,
|
| 455 |
+
"loss": 8.0767,
|
| 456 |
+
"step": 848
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 4.266666666666667,
|
| 460 |
+
"grad_norm": 2.0586202144622803,
|
| 461 |
+
"learning_rate": 1.2462735372353996e-06,
|
| 462 |
+
"loss": 8.0563,
|
| 463 |
+
"step": 864
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 4.345679012345679,
|
| 467 |
+
"grad_norm": 2.1375420093536377,
|
| 468 |
+
"learning_rate": 9.924546254786493e-07,
|
| 469 |
+
"loss": 8.0677,
|
| 470 |
+
"step": 880
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 4.424691358024692,
|
| 474 |
+
"grad_norm": 2.169581651687622,
|
| 475 |
+
"learning_rate": 7.661721499929753e-07,
|
| 476 |
+
"loss": 8.0669,
|
| 477 |
+
"step": 896
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 4.503703703703704,
|
| 481 |
+
"grad_norm": 2.463855743408203,
|
| 482 |
+
"learning_rate": 5.681178656024055e-07,
|
| 483 |
+
"loss": 8.084,
|
| 484 |
+
"step": 912
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 4.582716049382716,
|
| 488 |
+
"grad_norm": 1.940145492553711,
|
| 489 |
+
"learning_rate": 3.988972323910778e-07,
|
| 490 |
+
"loss": 8.0663,
|
| 491 |
+
"step": 928
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 4.661728395061728,
|
| 495 |
+
"grad_norm": 1.9443929195404053,
|
| 496 |
+
"learning_rate": 2.5902756478688674e-07,
|
| 497 |
+
"loss": 8.0821,
|
| 498 |
+
"step": 944
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 4.7407407407407405,
|
| 502 |
+
"grad_norm": 2.3160784244537354,
|
| 503 |
+
"learning_rate": 1.489364501100332e-07,
|
| 504 |
+
"loss": 8.0705,
|
| 505 |
+
"step": 960
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 4.8197530864197535,
|
| 509 |
+
"grad_norm": 2.41857647895813,
|
| 510 |
+
"learning_rate": 6.896044142100433e-08,
|
| 511 |
+
"loss": 8.0233,
|
| 512 |
+
"step": 976
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 4.898765432098766,
|
| 516 |
+
"grad_norm": 2.4384217262268066,
|
| 517 |
+
"learning_rate": 1.9344028664056715e-08,
|
| 518 |
+
"loss": 8.0523,
|
| 519 |
+
"step": 992
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 4.977777777777778,
|
| 523 |
+
"grad_norm": 2.1525607109069824,
|
| 524 |
+
"learning_rate": 2.388912514017516e-10,
|
| 525 |
+
"loss": 8.0613,
|
| 526 |
+
"step": 1008
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.987654320987654,
|
| 530 |
+
"eval_bleu": 0.058703717215517995,
|
| 531 |
+
"eval_cap_loss": 2.577115478468876,
|
| 532 |
+
"eval_con_loss": 2.058133321649888,
|
| 533 |
+
"eval_loss": 6.693382112418904,
|
| 534 |
+
"step": 1010
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 4.987654320987654,
|
| 538 |
+
"eval_bleu": 0.058703717215517995,
|
| 539 |
+
"eval_cap_loss": 2.577115478468876,
|
| 540 |
+
"eval_con_loss": 2.058133321649888,
|
| 541 |
+
"eval_loss": 6.693382112418904,
|
| 542 |
+
"eval_runtime": 79.8358,
|
| 543 |
+
"eval_samples_per_second": 10.133,
|
| 544 |
+
"eval_steps_per_second": 1.278,
|
| 545 |
+
"step": 1010
|
| 546 |
+
}
|
| 547 |
+
],
|
| 548 |
+
"logging_steps": 16,
|
| 549 |
+
"max_steps": 1010,
|
| 550 |
+
"num_input_tokens_seen": 0,
|
| 551 |
+
"num_train_epochs": 5,
|
| 552 |
+
"save_steps": 500,
|
| 553 |
+
"stateful_callbacks": {
|
| 554 |
+
"TrainerControl": {
|
| 555 |
+
"args": {
|
| 556 |
+
"should_epoch_stop": false,
|
| 557 |
+
"should_evaluate": false,
|
| 558 |
+
"should_log": false,
|
| 559 |
+
"should_save": true,
|
| 560 |
+
"should_training_stop": true
|
| 561 |
+
},
|
| 562 |
+
"attributes": {}
|
| 563 |
+
}
|
| 564 |
+
},
|
| 565 |
+
"total_flos": 0.0,
|
| 566 |
+
"train_batch_size": 16,
|
| 567 |
+
"trial_name": null,
|
| 568 |
+
"trial_params": null
|
| 569 |
+
}
|
checkpoints/checkpoint-1010/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee2a8b576b9d39a0aa2b1efffaf0e18c5b6404521c5e4285d4b45363bd019569
|
| 3 |
+
size 5176
|