Upload 8 files
Browse files- model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +342 -0
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2422974460
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7236b7d224d5aa3553d0c13a8e81d91b7a8d56ec4d855f3268926f531b43b991
|
| 3 |
size 2422974460
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8fb614f6258219a1a8afa5beb50209824c4ccfbb81d01c3e5f92fc1199c026d8
|
| 3 |
+
size 4846409706
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f7c0e26e544562a785563534b37a6a9d9cad5b67dcb89ac3dae1cdb8bfa9a9b
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:665c895d17e43975c6948ab4f9c42311a2215bbd4abf94b595fda0aac39265bd
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,342 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.04373383894562721,
|
| 3 |
+
"best_model_checkpoint": "w2v-bert-final-v2/checkpoint-10000",
|
| 4 |
+
"epoch": 7.716049382716049,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.23148148148148148,
|
| 13 |
+
"grad_norm": 7.229776859283447,
|
| 14 |
+
"learning_rate": 7.425e-06,
|
| 15 |
+
"loss": 3.4047,
|
| 16 |
+
"step": 300
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.46296296296296297,
|
| 20 |
+
"grad_norm": 11.000567436218262,
|
| 21 |
+
"learning_rate": 1.4925e-05,
|
| 22 |
+
"loss": 0.8648,
|
| 23 |
+
"step": 600
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.6944444444444444,
|
| 27 |
+
"grad_norm": 11.643492698669434,
|
| 28 |
+
"learning_rate": 2.2400000000000002e-05,
|
| 29 |
+
"loss": 0.6263,
|
| 30 |
+
"step": 900
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.7716049382716049,
|
| 34 |
+
"eval_loss": 0.5938597321510315,
|
| 35 |
+
"eval_runtime": 49.2872,
|
| 36 |
+
"eval_samples_per_second": 38.408,
|
| 37 |
+
"eval_steps_per_second": 4.809,
|
| 38 |
+
"eval_wer": 0.509299543180335,
|
| 39 |
+
"step": 1000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.9259259259259259,
|
| 43 |
+
"grad_norm": 16.209674835205078,
|
| 44 |
+
"learning_rate": 2.9900000000000002e-05,
|
| 45 |
+
"loss": 0.5512,
|
| 46 |
+
"step": 1200
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.1574074074074074,
|
| 50 |
+
"grad_norm": 2.7776882648468018,
|
| 51 |
+
"learning_rate": 3.74e-05,
|
| 52 |
+
"loss": 0.4844,
|
| 53 |
+
"step": 1500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.3888888888888888,
|
| 57 |
+
"grad_norm": 5.179060459136963,
|
| 58 |
+
"learning_rate": 4.4875e-05,
|
| 59 |
+
"loss": 0.4641,
|
| 60 |
+
"step": 1800
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.5432098765432098,
|
| 64 |
+
"eval_loss": 0.5058629512786865,
|
| 65 |
+
"eval_runtime": 49.9367,
|
| 66 |
+
"eval_samples_per_second": 37.908,
|
| 67 |
+
"eval_steps_per_second": 4.746,
|
| 68 |
+
"eval_wer": 0.45061996954535566,
|
| 69 |
+
"step": 2000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.6203703703703702,
|
| 73 |
+
"grad_norm": 2.7534990310668945,
|
| 74 |
+
"learning_rate": 4.943236137667304e-05,
|
| 75 |
+
"loss": 0.4629,
|
| 76 |
+
"step": 2100
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.8518518518518519,
|
| 80 |
+
"grad_norm": 3.1153225898742676,
|
| 81 |
+
"learning_rate": 4.763981835564054e-05,
|
| 82 |
+
"loss": 0.4393,
|
| 83 |
+
"step": 2400
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 2.0833333333333335,
|
| 87 |
+
"grad_norm": 2.560495376586914,
|
| 88 |
+
"learning_rate": 4.584727533460803e-05,
|
| 89 |
+
"loss": 0.3782,
|
| 90 |
+
"step": 2700
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 2.314814814814815,
|
| 94 |
+
"grad_norm": 2.7176785469055176,
|
| 95 |
+
"learning_rate": 4.4054732313575525e-05,
|
| 96 |
+
"loss": 0.3054,
|
| 97 |
+
"step": 3000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 2.314814814814815,
|
| 101 |
+
"eval_loss": 0.34695935249328613,
|
| 102 |
+
"eval_runtime": 49.4937,
|
| 103 |
+
"eval_samples_per_second": 38.247,
|
| 104 |
+
"eval_steps_per_second": 4.788,
|
| 105 |
+
"eval_wer": 0.3308135740700457,
|
| 106 |
+
"step": 3000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 2.5462962962962963,
|
| 110 |
+
"grad_norm": 2.319594383239746,
|
| 111 |
+
"learning_rate": 4.226816443594647e-05,
|
| 112 |
+
"loss": 0.3079,
|
| 113 |
+
"step": 3300
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 2.7777777777777777,
|
| 117 |
+
"grad_norm": 2.165224552154541,
|
| 118 |
+
"learning_rate": 4.047562141491396e-05,
|
| 119 |
+
"loss": 0.301,
|
| 120 |
+
"step": 3600
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 3.009259259259259,
|
| 124 |
+
"grad_norm": 3.609168529510498,
|
| 125 |
+
"learning_rate": 3.8683078393881456e-05,
|
| 126 |
+
"loss": 0.2837,
|
| 127 |
+
"step": 3900
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 3.0864197530864197,
|
| 131 |
+
"eval_loss": 0.2659013867378235,
|
| 132 |
+
"eval_runtime": 50.1986,
|
| 133 |
+
"eval_samples_per_second": 37.71,
|
| 134 |
+
"eval_steps_per_second": 4.721,
|
| 135 |
+
"eval_wer": 0.2583206438981945,
|
| 136 |
+
"step": 4000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 3.240740740740741,
|
| 140 |
+
"grad_norm": 1.6436126232147217,
|
| 141 |
+
"learning_rate": 3.689053537284895e-05,
|
| 142 |
+
"loss": 0.2191,
|
| 143 |
+
"step": 4200
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 3.4722222222222223,
|
| 147 |
+
"grad_norm": 1.704137921333313,
|
| 148 |
+
"learning_rate": 3.5103967495219884e-05,
|
| 149 |
+
"loss": 0.2103,
|
| 150 |
+
"step": 4500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 3.7037037037037037,
|
| 154 |
+
"grad_norm": 6.373330593109131,
|
| 155 |
+
"learning_rate": 3.331142447418738e-05,
|
| 156 |
+
"loss": 0.2174,
|
| 157 |
+
"step": 4800
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 3.8580246913580245,
|
| 161 |
+
"eval_loss": 0.19009661674499512,
|
| 162 |
+
"eval_runtime": 49.6613,
|
| 163 |
+
"eval_samples_per_second": 38.118,
|
| 164 |
+
"eval_steps_per_second": 4.772,
|
| 165 |
+
"eval_wer": 0.195072873613226,
|
| 166 |
+
"step": 5000
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 3.935185185185185,
|
| 170 |
+
"grad_norm": 2.4362969398498535,
|
| 171 |
+
"learning_rate": 3.151888145315488e-05,
|
| 172 |
+
"loss": 0.1939,
|
| 173 |
+
"step": 5100
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 4.166666666666667,
|
| 177 |
+
"grad_norm": 1.4299020767211914,
|
| 178 |
+
"learning_rate": 2.9726338432122373e-05,
|
| 179 |
+
"loss": 0.1666,
|
| 180 |
+
"step": 5400
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.398148148148148,
|
| 184 |
+
"grad_norm": 1.3548126220703125,
|
| 185 |
+
"learning_rate": 2.793379541108987e-05,
|
| 186 |
+
"loss": 0.1569,
|
| 187 |
+
"step": 5700
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 4.62962962962963,
|
| 191 |
+
"grad_norm": 1.5247106552124023,
|
| 192 |
+
"learning_rate": 2.6141252390057363e-05,
|
| 193 |
+
"loss": 0.152,
|
| 194 |
+
"step": 6000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.62962962962963,
|
| 198 |
+
"eval_loss": 0.15165844559669495,
|
| 199 |
+
"eval_runtime": 49.8658,
|
| 200 |
+
"eval_samples_per_second": 37.962,
|
| 201 |
+
"eval_steps_per_second": 4.753,
|
| 202 |
+
"eval_wer": 0.1540678703502284,
|
| 203 |
+
"step": 6000
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 4.861111111111111,
|
| 207 |
+
"grad_norm": 3.228041410446167,
|
| 208 |
+
"learning_rate": 2.434870936902486e-05,
|
| 209 |
+
"loss": 0.1481,
|
| 210 |
+
"step": 6300
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 5.092592592592593,
|
| 214 |
+
"grad_norm": 1.5059906244277954,
|
| 215 |
+
"learning_rate": 2.2562141491395794e-05,
|
| 216 |
+
"loss": 0.1256,
|
| 217 |
+
"step": 6600
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 5.324074074074074,
|
| 221 |
+
"grad_norm": 3.538102626800537,
|
| 222 |
+
"learning_rate": 2.0769598470363287e-05,
|
| 223 |
+
"loss": 0.1089,
|
| 224 |
+
"step": 6900
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.401234567901234,
|
| 228 |
+
"eval_loss": 0.11081259697675705,
|
| 229 |
+
"eval_runtime": 49.8031,
|
| 230 |
+
"eval_samples_per_second": 38.01,
|
| 231 |
+
"eval_steps_per_second": 4.759,
|
| 232 |
+
"eval_wer": 0.11404176636937133,
|
| 233 |
+
"step": 7000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 5.555555555555555,
|
| 237 |
+
"grad_norm": 0.6531468033790588,
|
| 238 |
+
"learning_rate": 1.8977055449330787e-05,
|
| 239 |
+
"loss": 0.1022,
|
| 240 |
+
"step": 7200
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.787037037037037,
|
| 244 |
+
"grad_norm": 1.7070139646530151,
|
| 245 |
+
"learning_rate": 1.718451242829828e-05,
|
| 246 |
+
"loss": 0.1022,
|
| 247 |
+
"step": 7500
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 6.018518518518518,
|
| 251 |
+
"grad_norm": 1.612425446510315,
|
| 252 |
+
"learning_rate": 1.5397944550669215e-05,
|
| 253 |
+
"loss": 0.0986,
|
| 254 |
+
"step": 7800
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.172839506172839,
|
| 258 |
+
"eval_loss": 0.08366883546113968,
|
| 259 |
+
"eval_runtime": 51.7773,
|
| 260 |
+
"eval_samples_per_second": 36.56,
|
| 261 |
+
"eval_steps_per_second": 4.577,
|
| 262 |
+
"eval_wer": 0.08902545138133565,
|
| 263 |
+
"step": 8000
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 6.25,
|
| 267 |
+
"grad_norm": 0.7614215612411499,
|
| 268 |
+
"learning_rate": 1.3605401529636713e-05,
|
| 269 |
+
"loss": 0.0766,
|
| 270 |
+
"step": 8100
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 6.481481481481482,
|
| 274 |
+
"grad_norm": 2.542365550994873,
|
| 275 |
+
"learning_rate": 1.181883365200765e-05,
|
| 276 |
+
"loss": 0.0758,
|
| 277 |
+
"step": 8400
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.712962962962963,
|
| 281 |
+
"grad_norm": 1.319675326347351,
|
| 282 |
+
"learning_rate": 1.0026290630975144e-05,
|
| 283 |
+
"loss": 0.0726,
|
| 284 |
+
"step": 8700
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.944444444444445,
|
| 288 |
+
"grad_norm": 0.8109455704689026,
|
| 289 |
+
"learning_rate": 8.233747609942639e-06,
|
| 290 |
+
"loss": 0.0648,
|
| 291 |
+
"step": 9000
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 6.944444444444445,
|
| 295 |
+
"eval_loss": 0.05808680132031441,
|
| 296 |
+
"eval_runtime": 51.2984,
|
| 297 |
+
"eval_samples_per_second": 36.902,
|
| 298 |
+
"eval_steps_per_second": 4.62,
|
| 299 |
+
"eval_wer": 0.06031107243854688,
|
| 300 |
+
"step": 9000
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 7.175925925925926,
|
| 304 |
+
"grad_norm": 1.2560392618179321,
|
| 305 |
+
"learning_rate": 6.441204588910134e-06,
|
| 306 |
+
"loss": 0.051,
|
| 307 |
+
"step": 9300
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 7.407407407407407,
|
| 311 |
+
"grad_norm": 3.7403156757354736,
|
| 312 |
+
"learning_rate": 4.654636711281071e-06,
|
| 313 |
+
"loss": 0.0473,
|
| 314 |
+
"step": 9600
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 7.638888888888889,
|
| 318 |
+
"grad_norm": 0.7289965748786926,
|
| 319 |
+
"learning_rate": 2.862093690248566e-06,
|
| 320 |
+
"loss": 0.0499,
|
| 321 |
+
"step": 9900
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 7.716049382716049,
|
| 325 |
+
"eval_loss": 0.04373383894562721,
|
| 326 |
+
"eval_runtime": 50.2864,
|
| 327 |
+
"eval_samples_per_second": 37.644,
|
| 328 |
+
"eval_steps_per_second": 4.713,
|
| 329 |
+
"eval_wer": 0.04606264955405699,
|
| 330 |
+
"step": 10000
|
| 331 |
+
}
|
| 332 |
+
],
|
| 333 |
+
"logging_steps": 300,
|
| 334 |
+
"max_steps": 10368,
|
| 335 |
+
"num_input_tokens_seen": 0,
|
| 336 |
+
"num_train_epochs": 8,
|
| 337 |
+
"save_steps": 1000,
|
| 338 |
+
"total_flos": 2.075992650724199e+19,
|
| 339 |
+
"train_batch_size": 8,
|
| 340 |
+
"trial_name": null,
|
| 341 |
+
"trial_params": null
|
| 342 |
+
}
|