Upload folder using huggingface_hub
Browse files- checkpoints/checkpoint-10240/ema.safetensors +3 -0
- checkpoints/checkpoint-10240/model.safetensors +3 -0
- checkpoints/checkpoint-10240/optimizer.pt +3 -0
- checkpoints/checkpoint-10240/rng_state.pth +3 -0
- checkpoints/checkpoint-10240/scaler.pt +3 -0
- checkpoints/checkpoint-10240/scheduler.pt +3 -0
- checkpoints/checkpoint-10240/trainer_state.json +464 -0
- checkpoints/checkpoint-10240/training_args.bin +3 -0
checkpoints/checkpoint-10240/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7f0949c4744e72fe73db782124fc4f290b639828c0631c12aa3d5f212ca9978
|
| 3 |
+
size 634904
|
checkpoints/checkpoint-10240/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e05f24acc8e4f4d75f94ac6f49a49b505fdda3c67b618e6220ee132cd57534c
|
| 3 |
+
size 19272752
|
checkpoints/checkpoint-10240/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0421234bf12d55e2308b14119e0395687cb7ffe91ba2f48abbe1d75de8ff6d62
|
| 3 |
+
size 1337227
|
checkpoints/checkpoint-10240/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2c04be13edf4b232ba08eac3996218459dd3d050909829aaa71859abddd8737
|
| 3 |
+
size 14645
|
checkpoints/checkpoint-10240/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94ddcf074384d3a34ade0cdda457b71df0096ecbfea07f5653bd35b87f62e3e9
|
| 3 |
+
size 1383
|
checkpoints/checkpoint-10240/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5eab6a06d702c463fa681f5c59aa4c76b4a7cb4dc7af81563f2aa2d201d5e87
|
| 3 |
+
size 1465
|
checkpoints/checkpoint-10240/trainer_state.json
ADDED
|
@@ -0,0 +1,464 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.472957369174634,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 10240,
|
| 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": 0.11696221679449081,
|
| 15 |
+
"learning_rate": 9.418282548476454e-05,
|
| 16 |
+
"loss": 1.568,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.14284025132656097,
|
| 22 |
+
"learning_rate": 0.00018873499538319484,
|
| 23 |
+
"loss": 1.4627,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.14623360335826874,
|
| 29 |
+
"learning_rate": 0.00028328716528162515,
|
| 30 |
+
"loss": 1.3763,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.1999262273311615,
|
| 36 |
+
"learning_rate": 0.0003778393351800554,
|
| 37 |
+
"loss": 1.325,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_loss": 1.2694775628172643,
|
| 43 |
+
"eval_mse_loss": 1.2694775628172643,
|
| 44 |
+
"step": 1024
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.047295736917463395,
|
| 48 |
+
"eval_loss": 1.2694775628172643,
|
| 49 |
+
"eval_mse_loss": 1.2694775628172643,
|
| 50 |
+
"eval_runtime": 99.7615,
|
| 51 |
+
"eval_samples_per_second": 280.599,
|
| 52 |
+
"eval_steps_per_second": 4.39,
|
| 53 |
+
"step": 1024
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.05911967114682925,
|
| 57 |
+
"grad_norm": 0.17195427417755127,
|
| 58 |
+
"learning_rate": 0.000399910381984159,
|
| 59 |
+
"loss": 1.2775,
|
| 60 |
+
"step": 1280
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.0709436053761951,
|
| 64 |
+
"grad_norm": 0.24726267158985138,
|
| 65 |
+
"learning_rate": 0.0003995235470960099,
|
| 66 |
+
"loss": 1.2335,
|
| 67 |
+
"step": 1536
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.08276753960556095,
|
| 71 |
+
"grad_norm": 0.29991263151168823,
|
| 72 |
+
"learning_rate": 0.000398831687956263,
|
| 73 |
+
"loss": 1.2043,
|
| 74 |
+
"step": 1792
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.09459147383492679,
|
| 78 |
+
"grad_norm": 0.2526465356349945,
|
| 79 |
+
"learning_rate": 0.00039783586225369426,
|
| 80 |
+
"loss": 1.1758,
|
| 81 |
+
"step": 2048
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.09459147383492679,
|
| 85 |
+
"eval_loss": 1.1193939001048536,
|
| 86 |
+
"eval_mse_loss": 1.1193939001048536,
|
| 87 |
+
"step": 2048
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09459147383492679,
|
| 91 |
+
"eval_loss": 1.1193939001048536,
|
| 92 |
+
"eval_mse_loss": 1.1193939001048536,
|
| 93 |
+
"eval_runtime": 99.592,
|
| 94 |
+
"eval_samples_per_second": 281.077,
|
| 95 |
+
"eval_steps_per_second": 4.398,
|
| 96 |
+
"step": 2048
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.10641540806429264,
|
| 100 |
+
"grad_norm": 0.23426194489002228,
|
| 101 |
+
"learning_rate": 0.0003965375923699678,
|
| 102 |
+
"loss": 1.1514,
|
| 103 |
+
"step": 2304
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.1182393422936585,
|
| 107 |
+
"grad_norm": 0.2490096390247345,
|
| 108 |
+
"learning_rate": 0.0003949388630522747,
|
| 109 |
+
"loss": 1.1287,
|
| 110 |
+
"step": 2560
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.13006327652302435,
|
| 114 |
+
"grad_norm": 0.3001377582550049,
|
| 115 |
+
"learning_rate": 0.0003930421183791246,
|
| 116 |
+
"loss": 1.1064,
|
| 117 |
+
"step": 2816
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.1418872107523902,
|
| 121 |
+
"grad_norm": 0.31271305680274963,
|
| 122 |
+
"learning_rate": 0.00039085025802392876,
|
| 123 |
+
"loss": 1.0907,
|
| 124 |
+
"step": 3072
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.1418872107523902,
|
| 128 |
+
"eval_loss": 1.0367718265481192,
|
| 129 |
+
"eval_mse_loss": 1.0367718265481192,
|
| 130 |
+
"step": 3072
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.1418872107523902,
|
| 134 |
+
"eval_loss": 1.0367718265481192,
|
| 135 |
+
"eval_mse_loss": 1.0367718265481192,
|
| 136 |
+
"eval_runtime": 99.0318,
|
| 137 |
+
"eval_samples_per_second": 282.667,
|
| 138 |
+
"eval_steps_per_second": 4.423,
|
| 139 |
+
"step": 3072
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.15371114498175603,
|
| 143 |
+
"grad_norm": 0.31617286801338196,
|
| 144 |
+
"learning_rate": 0.0003883666328220855,
|
| 145 |
+
"loss": 1.0795,
|
| 146 |
+
"step": 3328
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.1655350792111219,
|
| 150 |
+
"grad_norm": 0.3006609082221985,
|
| 151 |
+
"learning_rate": 0.00038559503964834713,
|
| 152 |
+
"loss": 1.072,
|
| 153 |
+
"step": 3584
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.17735901344048774,
|
| 157 |
+
"grad_norm": 0.3017251491546631,
|
| 158 |
+
"learning_rate": 0.00038253971561229733,
|
| 159 |
+
"loss": 1.0632,
|
| 160 |
+
"step": 3840
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.18918294766985358,
|
| 164 |
+
"grad_norm": 0.29505959153175354,
|
| 165 |
+
"learning_rate": 0.00037920533158081383,
|
| 166 |
+
"loss": 1.0583,
|
| 167 |
+
"step": 4096
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.18918294766985358,
|
| 171 |
+
"eval_loss": 1.0053560841301261,
|
| 172 |
+
"eval_mse_loss": 1.0053560841301261,
|
| 173 |
+
"step": 4096
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.18918294766985358,
|
| 177 |
+
"eval_loss": 1.0053560841301261,
|
| 178 |
+
"eval_mse_loss": 1.0053560841301261,
|
| 179 |
+
"eval_runtime": 99.7606,
|
| 180 |
+
"eval_samples_per_second": 280.602,
|
| 181 |
+
"eval_steps_per_second": 4.391,
|
| 182 |
+
"step": 4096
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.20100688189921945,
|
| 186 |
+
"grad_norm": 0.327700138092041,
|
| 187 |
+
"learning_rate": 0.00037559698503741954,
|
| 188 |
+
"loss": 1.0539,
|
| 189 |
+
"step": 4352
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.2128308161285853,
|
| 193 |
+
"grad_norm": 0.3910951316356659,
|
| 194 |
+
"learning_rate": 0.0003717201922894372,
|
| 195 |
+
"loss": 1.0525,
|
| 196 |
+
"step": 4608
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.22465475035795113,
|
| 200 |
+
"grad_norm": 0.3237537443637848,
|
| 201 |
+
"learning_rate": 0.0003675808800348619,
|
| 202 |
+
"loss": 1.0461,
|
| 203 |
+
"step": 4864
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.236478684587317,
|
| 207 |
+
"grad_norm": 0.4316788911819458,
|
| 208 |
+
"learning_rate": 0.0003631853763018435,
|
| 209 |
+
"loss": 1.042,
|
| 210 |
+
"step": 5120
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.236478684587317,
|
| 214 |
+
"eval_loss": 0.9950277422933274,
|
| 215 |
+
"eval_mse_loss": 0.9950277422933274,
|
| 216 |
+
"step": 5120
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.236478684587317,
|
| 220 |
+
"eval_loss": 0.9950277422933274,
|
| 221 |
+
"eval_mse_loss": 0.9950277422933274,
|
| 222 |
+
"eval_runtime": 99.7527,
|
| 223 |
+
"eval_samples_per_second": 280.624,
|
| 224 |
+
"eval_steps_per_second": 4.391,
|
| 225 |
+
"step": 5120
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.24830261881668284,
|
| 229 |
+
"grad_norm": 0.32822364568710327,
|
| 230 |
+
"learning_rate": 0.0003585404007746302,
|
| 231 |
+
"loss": 1.0385,
|
| 232 |
+
"step": 5376
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.2601265530460487,
|
| 236 |
+
"grad_norm": 0.40738987922668457,
|
| 237 |
+
"learning_rate": 0.0003536530545207627,
|
| 238 |
+
"loss": 1.0378,
|
| 239 |
+
"step": 5632
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.27195048727541454,
|
| 243 |
+
"grad_norm": 0.318866103887558,
|
| 244 |
+
"learning_rate": 0.0003485308091352234,
|
| 245 |
+
"loss": 1.0375,
|
| 246 |
+
"step": 5888
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.2837744215047804,
|
| 250 |
+
"grad_norm": 0.34148862957954407,
|
| 251 |
+
"learning_rate": 0.0003431814953181376,
|
| 252 |
+
"loss": 1.0351,
|
| 253 |
+
"step": 6144
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.2837744215047804,
|
| 257 |
+
"eval_loss": 0.9900110875634842,
|
| 258 |
+
"eval_mse_loss": 0.9900110875634842,
|
| 259 |
+
"step": 6144
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.2837744215047804,
|
| 263 |
+
"eval_loss": 0.9900110875634842,
|
| 264 |
+
"eval_mse_loss": 0.9900110875634842,
|
| 265 |
+
"eval_runtime": 99.6906,
|
| 266 |
+
"eval_samples_per_second": 280.799,
|
| 267 |
+
"eval_steps_per_second": 4.394,
|
| 268 |
+
"step": 6144
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2955983557341462,
|
| 272 |
+
"grad_norm": 0.2927286624908447,
|
| 273 |
+
"learning_rate": 0.0003376132909034871,
|
| 274 |
+
"loss": 1.0342,
|
| 275 |
+
"step": 6400
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.30742228996351206,
|
| 279 |
+
"grad_norm": 0.3355765640735626,
|
| 280 |
+
"learning_rate": 0.000331834708357139,
|
| 281 |
+
"loss": 1.0334,
|
| 282 |
+
"step": 6656
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.3192462241928779,
|
| 286 |
+
"grad_norm": 0.3341493308544159,
|
| 287 |
+
"learning_rate": 0.00032585458176330104,
|
| 288 |
+
"loss": 1.0315,
|
| 289 |
+
"step": 6912
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3310701584222438,
|
| 293 |
+
"grad_norm": 0.3863889276981354,
|
| 294 |
+
"learning_rate": 0.0003196820533192994,
|
| 295 |
+
"loss": 1.0316,
|
| 296 |
+
"step": 7168
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3310701584222438,
|
| 300 |
+
"eval_loss": 0.9869283731669596,
|
| 301 |
+
"eval_mse_loss": 0.9869283731669596,
|
| 302 |
+
"step": 7168
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.3310701584222438,
|
| 306 |
+
"eval_loss": 0.9869283731669596,
|
| 307 |
+
"eval_mse_loss": 0.9869283731669596,
|
| 308 |
+
"eval_runtime": 99.9137,
|
| 309 |
+
"eval_samples_per_second": 280.172,
|
| 310 |
+
"eval_steps_per_second": 4.384,
|
| 311 |
+
"step": 7168
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.34289409265160964,
|
| 315 |
+
"grad_norm": 0.37314340472221375,
|
| 316 |
+
"learning_rate": 0.00031332655935932356,
|
| 317 |
+
"loss": 1.0304,
|
| 318 |
+
"step": 7424
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3547180268809755,
|
| 322 |
+
"grad_norm": 0.31663990020751953,
|
| 323 |
+
"learning_rate": 0.0003067978159285067,
|
| 324 |
+
"loss": 1.0291,
|
| 325 |
+
"step": 7680
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3665419611103413,
|
| 329 |
+
"grad_norm": 0.3562430143356323,
|
| 330 |
+
"learning_rate": 0.0003001058039293929,
|
| 331 |
+
"loss": 1.0282,
|
| 332 |
+
"step": 7936
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.37836589533970716,
|
| 336 |
+
"grad_norm": 0.32352393865585327,
|
| 337 |
+
"learning_rate": 0.0002932607538635016,
|
| 338 |
+
"loss": 1.0289,
|
| 339 |
+
"step": 8192
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.37836589533970716,
|
| 343 |
+
"eval_loss": 0.9830868584380302,
|
| 344 |
+
"eval_mse_loss": 0.9830868584380302,
|
| 345 |
+
"step": 8192
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.37836589533970716,
|
| 349 |
+
"eval_loss": 0.9830868584380302,
|
| 350 |
+
"eval_mse_loss": 0.9830868584380302,
|
| 351 |
+
"eval_runtime": 100.3923,
|
| 352 |
+
"eval_samples_per_second": 278.836,
|
| 353 |
+
"eval_steps_per_second": 4.363,
|
| 354 |
+
"step": 8192
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.390189829569073,
|
| 358 |
+
"grad_norm": 0.3661426603794098,
|
| 359 |
+
"learning_rate": 0.0002862731301913133,
|
| 360 |
+
"loss": 1.0279,
|
| 361 |
+
"step": 8448
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.4020137637984389,
|
| 365 |
+
"grad_norm": 0.3046676218509674,
|
| 366 |
+
"learning_rate": 0.0002791536153345875,
|
| 367 |
+
"loss": 1.0275,
|
| 368 |
+
"step": 8704
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.41383769802780473,
|
| 372 |
+
"grad_norm": 0.3376181721687317,
|
| 373 |
+
"learning_rate": 0.0002719130933454701,
|
| 374 |
+
"loss": 1.0282,
|
| 375 |
+
"step": 8960
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.4256616322571706,
|
| 379 |
+
"grad_norm": 0.3915064334869385,
|
| 380 |
+
"learning_rate": 0.00026456263326735493,
|
| 381 |
+
"loss": 1.0307,
|
| 382 |
+
"step": 9216
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.4256616322571706,
|
| 386 |
+
"eval_loss": 0.984883288406346,
|
| 387 |
+
"eval_mse_loss": 0.984883288406346,
|
| 388 |
+
"step": 9216
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.4256616322571706,
|
| 392 |
+
"eval_loss": 0.984883288406346,
|
| 393 |
+
"eval_mse_loss": 0.984883288406346,
|
| 394 |
+
"eval_runtime": 100.2629,
|
| 395 |
+
"eval_samples_per_second": 279.196,
|
| 396 |
+
"eval_steps_per_second": 4.369,
|
| 397 |
+
"step": 9216
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.4374855664865364,
|
| 401 |
+
"grad_norm": 0.33798637986183167,
|
| 402 |
+
"learning_rate": 0.0002571134722129381,
|
| 403 |
+
"loss": 1.0297,
|
| 404 |
+
"step": 9472
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.44930950071590225,
|
| 408 |
+
"grad_norm": 0.2875025272369385,
|
| 409 |
+
"learning_rate": 0.0002495769981853336,
|
| 410 |
+
"loss": 1.0315,
|
| 411 |
+
"step": 9728
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.4611334349452681,
|
| 415 |
+
"grad_norm": 0.33755892515182495,
|
| 416 |
+
"learning_rate": 0.0002419647326685136,
|
| 417 |
+
"loss": 1.031,
|
| 418 |
+
"step": 9984
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.472957369174634,
|
| 422 |
+
"grad_norm": 0.2760786712169647,
|
| 423 |
+
"learning_rate": 0.00023431840867685062,
|
| 424 |
+
"loss": 1.0319,
|
| 425 |
+
"step": 10240
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.472957369174634,
|
| 429 |
+
"eval_loss": 0.9867455217664101,
|
| 430 |
+
"eval_mse_loss": 0.9867455217664101,
|
| 431 |
+
"step": 10240
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.472957369174634,
|
| 435 |
+
"eval_loss": 0.9867455217664101,
|
| 436 |
+
"eval_mse_loss": 0.9867455217664101,
|
| 437 |
+
"eval_runtime": 100.0615,
|
| 438 |
+
"eval_samples_per_second": 279.758,
|
| 439 |
+
"eval_steps_per_second": 4.377,
|
| 440 |
+
"step": 10240
|
| 441 |
+
}
|
| 442 |
+
],
|
| 443 |
+
"logging_steps": 256,
|
| 444 |
+
"max_steps": 21651,
|
| 445 |
+
"num_input_tokens_seen": 0,
|
| 446 |
+
"num_train_epochs": 1,
|
| 447 |
+
"save_steps": 1024,
|
| 448 |
+
"stateful_callbacks": {
|
| 449 |
+
"TrainerControl": {
|
| 450 |
+
"args": {
|
| 451 |
+
"should_epoch_stop": false,
|
| 452 |
+
"should_evaluate": false,
|
| 453 |
+
"should_log": false,
|
| 454 |
+
"should_save": true,
|
| 455 |
+
"should_training_stop": false
|
| 456 |
+
},
|
| 457 |
+
"attributes": {}
|
| 458 |
+
}
|
| 459 |
+
},
|
| 460 |
+
"total_flos": 0.0,
|
| 461 |
+
"train_batch_size": 64,
|
| 462 |
+
"trial_name": null,
|
| 463 |
+
"trial_params": null
|
| 464 |
+
}
|
checkpoints/checkpoint-10240/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7cb9674560e8f032729757b059398ec449fd02c0ccb60413a5d94de5b6442629
|
| 3 |
+
size 5777
|