Upload folder using huggingface_hub
Browse files- README.md +1 -1
- adapter_config.json +2 -4
- adapter_model.safetensors +1 -1
- checkpoint-14145/README.md +1 -1
- checkpoint-14145/adapter_config.json +2 -4
- checkpoint-14145/adapter_model.safetensors +1 -1
- checkpoint-14145/optimizer.pt +1 -1
- checkpoint-14145/scaler.pt +1 -1
- checkpoint-14145/scheduler.pt +1 -1
- checkpoint-14145/trainer_state.json +893 -878
- checkpoint-14145/training_args.bin +1 -1
README.md
CHANGED
|
@@ -199,4 +199,4 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
-
- PEFT 0.
|
|
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
+
- PEFT 0.14.0
|
adapter_config.json
CHANGED
|
@@ -3,19 +3,18 @@
|
|
| 3 |
"auto_mapping": null,
|
| 4 |
"base_model_name_or_path": "Salesforce/codet5-base",
|
| 5 |
"bias": "none",
|
| 6 |
-
"corda_config": null,
|
| 7 |
"eva_config": null,
|
| 8 |
"exclude_modules": null,
|
| 9 |
"fan_in_fan_out": false,
|
| 10 |
"inference_mode": true,
|
| 11 |
-
"init_lora_weights":
|
| 12 |
"layer_replication": null,
|
| 13 |
"layers_pattern": null,
|
| 14 |
"layers_to_transform": null,
|
| 15 |
"loftq_config": {},
|
| 16 |
"lora_alpha": 16,
|
| 17 |
"lora_bias": false,
|
| 18 |
-
"lora_dropout": 0.
|
| 19 |
"megatron_config": null,
|
| 20 |
"megatron_core": "megatron.core",
|
| 21 |
"modules_to_save": null,
|
|
@@ -28,7 +27,6 @@
|
|
| 28 |
"v"
|
| 29 |
],
|
| 30 |
"task_type": "SEQ_2_SEQ_LM",
|
| 31 |
-
"trainable_token_indices": null,
|
| 32 |
"use_dora": false,
|
| 33 |
"use_rslora": false
|
| 34 |
}
|
|
|
|
| 3 |
"auto_mapping": null,
|
| 4 |
"base_model_name_or_path": "Salesforce/codet5-base",
|
| 5 |
"bias": "none",
|
|
|
|
| 6 |
"eva_config": null,
|
| 7 |
"exclude_modules": null,
|
| 8 |
"fan_in_fan_out": false,
|
| 9 |
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": "gaussian",
|
| 11 |
"layer_replication": null,
|
| 12 |
"layers_pattern": null,
|
| 13 |
"layers_to_transform": null,
|
| 14 |
"loftq_config": {},
|
| 15 |
"lora_alpha": 16,
|
| 16 |
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.05,
|
| 18 |
"megatron_config": null,
|
| 19 |
"megatron_core": "megatron.core",
|
| 20 |
"modules_to_save": null,
|
|
|
|
| 27 |
"v"
|
| 28 |
],
|
| 29 |
"task_type": "SEQ_2_SEQ_LM",
|
|
|
|
| 30 |
"use_dora": false,
|
| 31 |
"use_rslora": false
|
| 32 |
}
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3558888
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1280f7f2b208c156d2a4a277a7a1a7316c93f88988b5c754c16059bf2e0f289c
|
| 3 |
size 3558888
|
checkpoint-14145/README.md
CHANGED
|
@@ -199,4 +199,4 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
-
- PEFT 0.
|
|
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
+
- PEFT 0.14.0
|
checkpoint-14145/adapter_config.json
CHANGED
|
@@ -3,19 +3,18 @@
|
|
| 3 |
"auto_mapping": null,
|
| 4 |
"base_model_name_or_path": "Salesforce/codet5-base",
|
| 5 |
"bias": "none",
|
| 6 |
-
"corda_config": null,
|
| 7 |
"eva_config": null,
|
| 8 |
"exclude_modules": null,
|
| 9 |
"fan_in_fan_out": false,
|
| 10 |
"inference_mode": true,
|
| 11 |
-
"init_lora_weights":
|
| 12 |
"layer_replication": null,
|
| 13 |
"layers_pattern": null,
|
| 14 |
"layers_to_transform": null,
|
| 15 |
"loftq_config": {},
|
| 16 |
"lora_alpha": 16,
|
| 17 |
"lora_bias": false,
|
| 18 |
-
"lora_dropout": 0.
|
| 19 |
"megatron_config": null,
|
| 20 |
"megatron_core": "megatron.core",
|
| 21 |
"modules_to_save": null,
|
|
@@ -28,7 +27,6 @@
|
|
| 28 |
"v"
|
| 29 |
],
|
| 30 |
"task_type": "SEQ_2_SEQ_LM",
|
| 31 |
-
"trainable_token_indices": null,
|
| 32 |
"use_dora": false,
|
| 33 |
"use_rslora": false
|
| 34 |
}
|
|
|
|
| 3 |
"auto_mapping": null,
|
| 4 |
"base_model_name_or_path": "Salesforce/codet5-base",
|
| 5 |
"bias": "none",
|
|
|
|
| 6 |
"eva_config": null,
|
| 7 |
"exclude_modules": null,
|
| 8 |
"fan_in_fan_out": false,
|
| 9 |
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": "gaussian",
|
| 11 |
"layer_replication": null,
|
| 12 |
"layers_pattern": null,
|
| 13 |
"layers_to_transform": null,
|
| 14 |
"loftq_config": {},
|
| 15 |
"lora_alpha": 16,
|
| 16 |
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.05,
|
| 18 |
"megatron_config": null,
|
| 19 |
"megatron_core": "megatron.core",
|
| 20 |
"modules_to_save": null,
|
|
|
|
| 27 |
"v"
|
| 28 |
],
|
| 29 |
"task_type": "SEQ_2_SEQ_LM",
|
|
|
|
| 30 |
"use_dora": false,
|
| 31 |
"use_rslora": false
|
| 32 |
}
|
checkpoint-14145/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3558888
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1280f7f2b208c156d2a4a277a7a1a7316c93f88988b5c754c16059bf2e0f289c
|
| 3 |
size 3558888
|
checkpoint-14145/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7198906
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03e73f5f4a50e8492ca23abad4e74c002d3e1fe0681df8ea443c93e46d7d8cd0
|
| 3 |
size 7198906
|
checkpoint-14145/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:551d95fa01a350d649318071ec6c30c9643b1415452525afe82187c8c159941c
|
| 3 |
size 988
|
checkpoint-14145/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:898d0d2ace2142f45ba707687d167a2ef3ebdcb03bdca2695993e6fe1235bc54
|
| 3 |
size 1064
|
checkpoint-14145/trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 14145,
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint": "
|
| 5 |
"epoch": 5.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 14145,
|
|
@@ -11,2026 +11,2041 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.017674089784376106,
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss":
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.03534817956875221,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss":
|
| 24 |
"step": 100
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.053022269353128315,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss":
|
| 31 |
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.07069635913750442,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 1.
|
| 38 |
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.08837044892188052,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.10604453870625663,
|
| 49 |
-
"grad_norm": 2.
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 1.
|
| 52 |
"step": 300
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.12371862849063273,
|
| 56 |
-
"grad_norm": 2.
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 1.
|
| 59 |
"step": 350
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.14139271827500885,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 1.
|
| 66 |
"step": 400
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.15906680805938495,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss":
|
| 73 |
"step": 450
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.17674089784376104,
|
| 77 |
-
"grad_norm": 2.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss":
|
| 80 |
"step": 500
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.19441498762813716,
|
| 84 |
-
"grad_norm": 2.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss":
|
| 87 |
"step": 550
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.21208907741251326,
|
| 91 |
-
"grad_norm": 2.
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss":
|
| 94 |
"step": 600
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.22976316719688936,
|
| 98 |
-
"grad_norm":
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss":
|
| 101 |
"step": 650
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.24743725698126545,
|
| 105 |
-
"grad_norm": 2.
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss":
|
| 108 |
"step": 700
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.2651113467656416,
|
| 112 |
-
"grad_norm":
|
| 113 |
-
"learning_rate":
|
| 114 |
-
"loss":
|
| 115 |
"step": 750
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.2827854365500177,
|
| 119 |
-
"grad_norm":
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss":
|
| 122 |
"step": 800
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.30045952633439377,
|
| 126 |
-
"grad_norm":
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss":
|
| 129 |
"step": 850
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.3181336161187699,
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"loss":
|
| 136 |
"step": 900
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.335807705903146,
|
| 140 |
-
"grad_norm":
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss":
|
| 143 |
"step": 950
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.3534817956875221,
|
| 147 |
-
"grad_norm": 2.
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss":
|
| 150 |
"step": 1000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.3711558854718982,
|
| 154 |
-
"grad_norm":
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss":
|
| 157 |
"step": 1050
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.38882997525627433,
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss":
|
| 164 |
"step": 1100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.4065040650406504,
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss":
|
| 171 |
"step": 1150
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.4241781548250265,
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 1200
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.4418522446094026,
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss":
|
| 185 |
"step": 1250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.4595263343937787,
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 1300
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.47720042417815484,
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 1350
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.4948745139625309,
|
| 203 |
-
"grad_norm":
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 1400
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.512548603746907,
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss":
|
| 213 |
"step": 1450
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.5302226935312832,
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 1500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.5478967833156593,
|
| 224 |
-
"grad_norm":
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 1550
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.5655708731000354,
|
| 231 |
-
"grad_norm":
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 1600
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.5832449628844114,
|
| 238 |
-
"grad_norm":
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 1650
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.6009190526687875,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 1700
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.6185931424531637,
|
| 252 |
-
"grad_norm":
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 1750
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.6362672322375398,
|
| 259 |
-
"grad_norm": 3.
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 1800
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.6539413220219159,
|
| 266 |
-
"grad_norm":
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 1850
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.671615411806292,
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 1900
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.689289501590668,
|
| 280 |
-
"grad_norm":
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 1950
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.7069635913750442,
|
| 287 |
-
"grad_norm":
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 2000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.7246376811594203,
|
| 294 |
-
"grad_norm":
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 2050
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.7423117709437964,
|
| 301 |
-
"grad_norm":
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss": 0.
|
| 304 |
"step": 2100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.7599858607281725,
|
| 308 |
-
"grad_norm":
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 2150
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.7776599505125487,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss": 0.
|
| 318 |
"step": 2200
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.7953340402969247,
|
| 322 |
-
"grad_norm":
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 2250
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.8130081300813008,
|
| 329 |
-
"grad_norm":
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 2300
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.8306822198656769,
|
| 336 |
-
"grad_norm":
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 2350
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.848356309650053,
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 2400
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.8660303994344292,
|
| 350 |
-
"grad_norm":
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 2450
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.8837044892188052,
|
| 357 |
-
"grad_norm":
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 2500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.9013785790031813,
|
| 364 |
-
"grad_norm":
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 2550
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.9190526687875574,
|
| 371 |
-
"grad_norm":
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 2600
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.9367267585719335,
|
| 378 |
-
"grad_norm":
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 2650
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.9544008483563097,
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 2700
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.9720749381406858,
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 2750
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.9897490279250618,
|
| 399 |
-
"grad_norm":
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 2800
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.0,
|
| 406 |
-
"eval_bertscore_f1": 0.
|
| 407 |
-
"eval_bleu": 0.
|
| 408 |
-
"eval_loss": 0.
|
| 409 |
-
"
|
| 410 |
-
"
|
| 411 |
-
"
|
|
|
|
|
|
|
|
|
|
| 412 |
"step": 2829
|
| 413 |
},
|
| 414 |
{
|
| 415 |
"epoch": 1.007423117709438,
|
| 416 |
-
"grad_norm":
|
| 417 |
-
"learning_rate":
|
| 418 |
-
"loss": 0.
|
| 419 |
"step": 2850
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 1.025097207493814,
|
| 423 |
-
"grad_norm":
|
| 424 |
-
"learning_rate":
|
| 425 |
-
"loss": 0.
|
| 426 |
"step": 2900
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 1.0427712972781902,
|
| 430 |
-
"grad_norm":
|
| 431 |
-
"learning_rate":
|
| 432 |
-
"loss": 0.
|
| 433 |
"step": 2950
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 1.0604453870625663,
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate":
|
| 439 |
-
"loss": 0.
|
| 440 |
"step": 3000
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 1.0781194768469424,
|
| 444 |
-
"grad_norm":
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss": 0.
|
| 447 |
"step": 3050
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 1.0957935666313185,
|
| 451 |
-
"grad_norm":
|
| 452 |
-
"learning_rate":
|
| 453 |
-
"loss": 0.
|
| 454 |
"step": 3100
|
| 455 |
},
|
| 456 |
{
|
| 457 |
"epoch": 1.1134676564156947,
|
| 458 |
-
"grad_norm":
|
| 459 |
-
"learning_rate":
|
| 460 |
-
"loss": 0.
|
| 461 |
"step": 3150
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 1.1311417462000706,
|
| 465 |
-
"grad_norm":
|
| 466 |
-
"learning_rate":
|
| 467 |
-
"loss": 0.
|
| 468 |
"step": 3200
|
| 469 |
},
|
| 470 |
{
|
| 471 |
"epoch": 1.148815835984447,
|
| 472 |
-
"grad_norm":
|
| 473 |
-
"learning_rate":
|
| 474 |
-
"loss": 0.
|
| 475 |
"step": 3250
|
| 476 |
},
|
| 477 |
{
|
| 478 |
"epoch": 1.1664899257688228,
|
| 479 |
-
"grad_norm": 2.
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 0.
|
| 482 |
"step": 3300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 1.184164015553199,
|
| 486 |
-
"grad_norm":
|
| 487 |
-
"learning_rate":
|
| 488 |
-
"loss": 0.
|
| 489 |
"step": 3350
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 1.201838105337575,
|
| 493 |
-
"grad_norm":
|
| 494 |
-
"learning_rate":
|
| 495 |
-
"loss": 0.
|
| 496 |
"step": 3400
|
| 497 |
},
|
| 498 |
{
|
| 499 |
"epoch": 1.2195121951219512,
|
| 500 |
-
"grad_norm":
|
| 501 |
-
"learning_rate":
|
| 502 |
-
"loss": 0.
|
| 503 |
"step": 3450
|
| 504 |
},
|
| 505 |
{
|
| 506 |
"epoch": 1.2371862849063273,
|
| 507 |
-
"grad_norm":
|
| 508 |
-
"learning_rate":
|
| 509 |
-
"loss": 0.
|
| 510 |
"step": 3500
|
| 511 |
},
|
| 512 |
{
|
| 513 |
"epoch": 1.2548603746907034,
|
| 514 |
-
"grad_norm":
|
| 515 |
-
"learning_rate":
|
| 516 |
-
"loss": 0.
|
| 517 |
"step": 3550
|
| 518 |
},
|
| 519 |
{
|
| 520 |
"epoch": 1.2725344644750796,
|
| 521 |
-
"grad_norm":
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss": 0.
|
| 524 |
"step": 3600
|
| 525 |
},
|
| 526 |
{
|
| 527 |
"epoch": 1.2902085542594557,
|
| 528 |
-
"grad_norm":
|
| 529 |
-
"learning_rate":
|
| 530 |
-
"loss": 0.
|
| 531 |
"step": 3650
|
| 532 |
},
|
| 533 |
{
|
| 534 |
"epoch": 1.3078826440438318,
|
| 535 |
-
"grad_norm":
|
| 536 |
-
"learning_rate":
|
| 537 |
-
"loss": 0.
|
| 538 |
"step": 3700
|
| 539 |
},
|
| 540 |
{
|
| 541 |
"epoch": 1.3255567338282077,
|
| 542 |
-
"grad_norm":
|
| 543 |
-
"learning_rate":
|
| 544 |
-
"loss": 0.
|
| 545 |
"step": 3750
|
| 546 |
},
|
| 547 |
{
|
| 548 |
"epoch": 1.343230823612584,
|
| 549 |
-
"grad_norm":
|
| 550 |
-
"learning_rate":
|
| 551 |
-
"loss": 0.
|
| 552 |
"step": 3800
|
| 553 |
},
|
| 554 |
{
|
| 555 |
"epoch": 1.36090491339696,
|
| 556 |
-
"grad_norm":
|
| 557 |
-
"learning_rate":
|
| 558 |
-
"loss": 0.
|
| 559 |
"step": 3850
|
| 560 |
},
|
| 561 |
{
|
| 562 |
"epoch": 1.378579003181336,
|
| 563 |
-
"grad_norm":
|
| 564 |
-
"learning_rate":
|
| 565 |
-
"loss": 0.
|
| 566 |
"step": 3900
|
| 567 |
},
|
| 568 |
{
|
| 569 |
"epoch": 1.3962530929657122,
|
| 570 |
-
"grad_norm":
|
| 571 |
-
"learning_rate":
|
| 572 |
-
"loss": 0.
|
| 573 |
"step": 3950
|
| 574 |
},
|
| 575 |
{
|
| 576 |
"epoch": 1.4139271827500883,
|
| 577 |
-
"grad_norm":
|
| 578 |
-
"learning_rate":
|
| 579 |
-
"loss": 0.
|
| 580 |
"step": 4000
|
| 581 |
},
|
| 582 |
{
|
| 583 |
"epoch": 1.4316012725344645,
|
| 584 |
-
"grad_norm":
|
| 585 |
-
"learning_rate":
|
| 586 |
-
"loss": 0.
|
| 587 |
"step": 4050
|
| 588 |
},
|
| 589 |
{
|
| 590 |
"epoch": 1.4492753623188406,
|
| 591 |
-
"grad_norm":
|
| 592 |
-
"learning_rate":
|
| 593 |
-
"loss": 0.
|
| 594 |
"step": 4100
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 1.4669494521032167,
|
| 598 |
-
"grad_norm":
|
| 599 |
-
"learning_rate":
|
| 600 |
-
"loss": 0.
|
| 601 |
"step": 4150
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"epoch": 1.4846235418875928,
|
| 605 |
-
"grad_norm":
|
| 606 |
-
"learning_rate":
|
| 607 |
-
"loss": 0.
|
| 608 |
"step": 4200
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 1.502297631671969,
|
| 612 |
-
"grad_norm":
|
| 613 |
-
"learning_rate":
|
| 614 |
-
"loss": 0.
|
| 615 |
"step": 4250
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 1.5199717214563448,
|
| 619 |
-
"grad_norm":
|
| 620 |
-
"learning_rate":
|
| 621 |
-
"loss": 0.
|
| 622 |
"step": 4300
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 1.5376458112407212,
|
| 626 |
-
"grad_norm":
|
| 627 |
-
"learning_rate":
|
| 628 |
-
"loss": 0.
|
| 629 |
"step": 4350
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 1.555319901025097,
|
| 633 |
-
"grad_norm":
|
| 634 |
-
"learning_rate":
|
| 635 |
-
"loss": 0.
|
| 636 |
"step": 4400
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 1.5729939908094734,
|
| 640 |
-
"grad_norm":
|
| 641 |
-
"learning_rate":
|
| 642 |
-
"loss": 0.
|
| 643 |
"step": 4450
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 1.5906680805938493,
|
| 647 |
-
"grad_norm":
|
| 648 |
-
"learning_rate":
|
| 649 |
-
"loss": 0.
|
| 650 |
"step": 4500
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 1.6083421703782255,
|
| 654 |
-
"grad_norm":
|
| 655 |
-
"learning_rate":
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 4550
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 1.6260162601626016,
|
| 661 |
-
"grad_norm":
|
| 662 |
-
"learning_rate":
|
| 663 |
-
"loss": 0.
|
| 664 |
"step": 4600
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 1.6436903499469777,
|
| 668 |
-
"grad_norm":
|
| 669 |
-
"learning_rate":
|
| 670 |
-
"loss": 0.
|
| 671 |
"step": 4650
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 1.6613644397313538,
|
| 675 |
-
"grad_norm":
|
| 676 |
-
"learning_rate":
|
| 677 |
-
"loss": 0.
|
| 678 |
"step": 4700
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 1.67903852951573,
|
| 682 |
-
"grad_norm":
|
| 683 |
-
"learning_rate":
|
| 684 |
-
"loss": 0.
|
| 685 |
"step": 4750
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 1.696712619300106,
|
| 689 |
-
"grad_norm":
|
| 690 |
-
"learning_rate":
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 4800
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 1.714386709084482,
|
| 696 |
-
"grad_norm":
|
| 697 |
-
"learning_rate":
|
| 698 |
-
"loss": 0.
|
| 699 |
"step": 4850
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 1.7320607988688583,
|
| 703 |
-
"grad_norm":
|
| 704 |
-
"learning_rate":
|
| 705 |
-
"loss": 0.
|
| 706 |
"step": 4900
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 1.7497348886532342,
|
| 710 |
-
"grad_norm":
|
| 711 |
-
"learning_rate":
|
| 712 |
-
"loss": 0.
|
| 713 |
"step": 4950
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 1.7674089784376106,
|
| 717 |
-
"grad_norm":
|
| 718 |
-
"learning_rate":
|
| 719 |
-
"loss": 0.
|
| 720 |
"step": 5000
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 1.7850830682219865,
|
| 724 |
-
"grad_norm":
|
| 725 |
-
"learning_rate":
|
| 726 |
-
"loss": 0.
|
| 727 |
"step": 5050
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 1.8027571580063628,
|
| 731 |
-
"grad_norm":
|
| 732 |
-
"learning_rate":
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 5100
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 1.8204312477907387,
|
| 738 |
-
"grad_norm":
|
| 739 |
-
"learning_rate":
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 5150
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 1.8381053375751149,
|
| 745 |
-
"grad_norm":
|
| 746 |
-
"learning_rate":
|
| 747 |
-
"loss": 0.
|
| 748 |
"step": 5200
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 1.855779427359491,
|
| 752 |
-
"grad_norm":
|
| 753 |
-
"learning_rate":
|
| 754 |
-
"loss": 0.
|
| 755 |
"step": 5250
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 1.873453517143867,
|
| 759 |
-
"grad_norm":
|
| 760 |
-
"learning_rate":
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 5300
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 1.8911276069282432,
|
| 766 |
-
"grad_norm":
|
| 767 |
-
"learning_rate":
|
| 768 |
-
"loss": 0.
|
| 769 |
"step": 5350
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 1.9088016967126193,
|
| 773 |
-
"grad_norm":
|
| 774 |
-
"learning_rate":
|
| 775 |
-
"loss": 0.
|
| 776 |
"step": 5400
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 1.9264757864969955,
|
| 780 |
-
"grad_norm":
|
| 781 |
-
"learning_rate":
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 5450
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 1.9441498762813714,
|
| 787 |
-
"grad_norm":
|
| 788 |
-
"learning_rate":
|
| 789 |
-
"loss": 0.
|
| 790 |
"step": 5500
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 1.9618239660657477,
|
| 794 |
-
"grad_norm":
|
| 795 |
-
"learning_rate":
|
| 796 |
-
"loss": 0.
|
| 797 |
"step": 5550
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 1.9794980558501236,
|
| 801 |
-
"grad_norm":
|
| 802 |
-
"learning_rate":
|
| 803 |
-
"loss": 0.
|
| 804 |
"step": 5600
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 1.9971721456345,
|
| 808 |
-
"grad_norm":
|
| 809 |
-
"learning_rate":
|
| 810 |
-
"loss": 0.
|
| 811 |
"step": 5650
|
| 812 |
},
|
| 813 |
{
|
| 814 |
"epoch": 2.0,
|
| 815 |
-
"eval_bertscore_f1": 0.
|
| 816 |
-
"eval_bleu": 0.
|
| 817 |
-
"eval_loss": 0.
|
| 818 |
-
"
|
| 819 |
-
"
|
| 820 |
-
"
|
|
|
|
|
|
|
|
|
|
| 821 |
"step": 5658
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 2.014846235418876,
|
| 825 |
-
"grad_norm":
|
| 826 |
-
"learning_rate":
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 5700
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 2.032520325203252,
|
| 832 |
-
"grad_norm":
|
| 833 |
-
"learning_rate":
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 5750
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 2.050194414987628,
|
| 839 |
-
"grad_norm":
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 5800
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 2.0678685047720045,
|
| 846 |
-
"grad_norm":
|
| 847 |
-
"learning_rate":
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 5850
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 2.0855425945563804,
|
| 853 |
-
"grad_norm":
|
| 854 |
-
"learning_rate":
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 5900
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 2.1032166843407563,
|
| 860 |
-
"grad_norm":
|
| 861 |
-
"learning_rate":
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 5950
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 2.1208907741251326,
|
| 867 |
-
"grad_norm":
|
| 868 |
-
"learning_rate":
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 6000
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 2.1385648639095085,
|
| 874 |
-
"grad_norm":
|
| 875 |
-
"learning_rate":
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 6050
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 2.156238953693885,
|
| 881 |
-
"grad_norm":
|
| 882 |
-
"learning_rate":
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 6100
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 2.1739130434782608,
|
| 888 |
-
"grad_norm":
|
| 889 |
-
"learning_rate":
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 6150
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 2.191587133262637,
|
| 895 |
-
"grad_norm":
|
| 896 |
-
"learning_rate":
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 6200
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 2.209261223047013,
|
| 902 |
-
"grad_norm":
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 6250
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 2.2269353128313893,
|
| 909 |
-
"grad_norm":
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 6300
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 2.2446094026157652,
|
| 916 |
-
"grad_norm":
|
| 917 |
-
"learning_rate":
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 6350
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 2.262283492400141,
|
| 923 |
-
"grad_norm":
|
| 924 |
-
"learning_rate":
|
| 925 |
-
"loss": 0.
|
| 926 |
"step": 6400
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 2.2799575821845175,
|
| 930 |
-
"grad_norm":
|
| 931 |
-
"learning_rate":
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 6450
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 2.297631671968894,
|
| 937 |
-
"grad_norm":
|
| 938 |
-
"learning_rate":
|
| 939 |
-
"loss": 0.
|
| 940 |
"step": 6500
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 2.3153057617532697,
|
| 944 |
-
"grad_norm":
|
| 945 |
-
"learning_rate":
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 6550
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 2.3329798515376456,
|
| 951 |
-
"grad_norm":
|
| 952 |
-
"learning_rate":
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 6600
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 2.350653941322022,
|
| 958 |
-
"grad_norm":
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 6650
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 2.368328031106398,
|
| 965 |
-
"grad_norm":
|
| 966 |
-
"learning_rate":
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 6700
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 2.3860021208907742,
|
| 972 |
-
"grad_norm":
|
| 973 |
-
"learning_rate":
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 6750
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 2.40367621067515,
|
| 979 |
-
"grad_norm":
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 6800
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 2.4213503004595265,
|
| 986 |
-
"grad_norm":
|
| 987 |
-
"learning_rate":
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 6850
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 2.4390243902439024,
|
| 993 |
-
"grad_norm":
|
| 994 |
-
"learning_rate":
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 6900
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 2.4566984800282787,
|
| 1000 |
-
"grad_norm":
|
| 1001 |
-
"learning_rate":
|
| 1002 |
-
"loss": 0.
|
| 1003 |
"step": 6950
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 2.4743725698126546,
|
| 1007 |
-
"grad_norm":
|
| 1008 |
-
"learning_rate":
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 7000
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 2.4920466595970305,
|
| 1014 |
-
"grad_norm":
|
| 1015 |
-
"learning_rate":
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 7050
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 2.509720749381407,
|
| 1021 |
-
"grad_norm":
|
| 1022 |
-
"learning_rate":
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 7100
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 2.5273948391657832,
|
| 1028 |
-
"grad_norm":
|
| 1029 |
-
"learning_rate":
|
| 1030 |
-
"loss": 0.
|
| 1031 |
"step": 7150
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 2.545068928950159,
|
| 1035 |
-
"grad_norm":
|
| 1036 |
-
"learning_rate":
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 7200
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 2.562743018734535,
|
| 1042 |
-
"grad_norm":
|
| 1043 |
-
"learning_rate":
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 7250
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 2.5804171085189114,
|
| 1049 |
-
"grad_norm":
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 7300
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 2.5980911983032873,
|
| 1056 |
-
"grad_norm":
|
| 1057 |
-
"learning_rate":
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 7350
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 2.6157652880876636,
|
| 1063 |
-
"grad_norm":
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 7400
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 2.6334393778720395,
|
| 1070 |
-
"grad_norm":
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 7450
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 2.6511134676564154,
|
| 1077 |
-
"grad_norm":
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 7500
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 2.6687875574407918,
|
| 1084 |
-
"grad_norm":
|
| 1085 |
-
"learning_rate":
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 7550
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 2.686461647225168,
|
| 1091 |
-
"grad_norm":
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 7600
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 2.704135737009544,
|
| 1098 |
-
"grad_norm":
|
| 1099 |
-
"learning_rate":
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 7650
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 2.72180982679392,
|
| 1105 |
-
"grad_norm":
|
| 1106 |
-
"learning_rate":
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 7700
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 2.7394839165782963,
|
| 1112 |
-
"grad_norm":
|
| 1113 |
-
"learning_rate":
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 7750
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 2.757158006362672,
|
| 1119 |
-
"grad_norm":
|
| 1120 |
-
"learning_rate":
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 7800
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 2.7748320961470485,
|
| 1126 |
-
"grad_norm":
|
| 1127 |
-
"learning_rate":
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 7850
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 2.7925061859314244,
|
| 1133 |
-
"grad_norm":
|
| 1134 |
-
"learning_rate":
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 7900
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 2.8101802757158008,
|
| 1140 |
-
"grad_norm":
|
| 1141 |
-
"learning_rate":
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 7950
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 2.8278543655001767,
|
| 1147 |
-
"grad_norm":
|
| 1148 |
-
"learning_rate":
|
| 1149 |
-
"loss": 0.
|
| 1150 |
"step": 8000
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 2.845528455284553,
|
| 1154 |
-
"grad_norm":
|
| 1155 |
-
"learning_rate":
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 8050
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 2.863202545068929,
|
| 1161 |
-
"grad_norm":
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 8100
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 2.880876634853305,
|
| 1168 |
-
"grad_norm":
|
| 1169 |
-
"learning_rate":
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 8150
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 2.898550724637681,
|
| 1175 |
-
"grad_norm":
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 8200
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 2.9162248144220575,
|
| 1182 |
-
"grad_norm":
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 8250
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 2.9338989042064334,
|
| 1189 |
-
"grad_norm":
|
| 1190 |
-
"learning_rate":
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 8300
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 2.9515729939908093,
|
| 1196 |
-
"grad_norm":
|
| 1197 |
-
"learning_rate":
|
| 1198 |
-
"loss": 0.
|
| 1199 |
"step": 8350
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 2.9692470837751856,
|
| 1203 |
-
"grad_norm":
|
| 1204 |
-
"learning_rate":
|
| 1205 |
-
"loss": 0.
|
| 1206 |
"step": 8400
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 2.9869211735595615,
|
| 1210 |
-
"grad_norm":
|
| 1211 |
-
"learning_rate":
|
| 1212 |
-
"loss": 0.
|
| 1213 |
"step": 8450
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 3.0,
|
| 1217 |
-
"eval_bertscore_f1": 0.
|
| 1218 |
-
"eval_bleu": 0.
|
| 1219 |
-
"eval_loss": 0.
|
| 1220 |
-
"
|
| 1221 |
-
"
|
| 1222 |
-
"
|
|
|
|
|
|
|
|
|
|
| 1223 |
"step": 8487
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
"epoch": 3.004595263343938,
|
| 1227 |
-
"grad_norm":
|
| 1228 |
-
"learning_rate":
|
| 1229 |
-
"loss": 0.
|
| 1230 |
"step": 8500
|
| 1231 |
},
|
| 1232 |
{
|
| 1233 |
"epoch": 3.022269353128314,
|
| 1234 |
-
"grad_norm":
|
| 1235 |
-
"learning_rate":
|
| 1236 |
-
"loss": 0.
|
| 1237 |
"step": 8550
|
| 1238 |
},
|
| 1239 |
{
|
| 1240 |
"epoch": 3.03994344291269,
|
| 1241 |
-
"grad_norm":
|
| 1242 |
-
"learning_rate":
|
| 1243 |
-
"loss": 0.
|
| 1244 |
"step": 8600
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
"epoch": 3.057617532697066,
|
| 1248 |
-
"grad_norm":
|
| 1249 |
-
"learning_rate":
|
| 1250 |
-
"loss": 0.
|
| 1251 |
"step": 8650
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
"epoch": 3.0752916224814424,
|
| 1255 |
-
"grad_norm":
|
| 1256 |
-
"learning_rate":
|
| 1257 |
-
"loss": 0.
|
| 1258 |
"step": 8700
|
| 1259 |
},
|
| 1260 |
{
|
| 1261 |
"epoch": 3.0929657122658183,
|
| 1262 |
-
"grad_norm":
|
| 1263 |
-
"learning_rate":
|
| 1264 |
-
"loss": 0.
|
| 1265 |
"step": 8750
|
| 1266 |
},
|
| 1267 |
{
|
| 1268 |
"epoch": 3.110639802050194,
|
| 1269 |
-
"grad_norm":
|
| 1270 |
-
"learning_rate":
|
| 1271 |
-
"loss": 0.
|
| 1272 |
"step": 8800
|
| 1273 |
},
|
| 1274 |
{
|
| 1275 |
"epoch": 3.1283138918345705,
|
| 1276 |
-
"grad_norm":
|
| 1277 |
-
"learning_rate":
|
| 1278 |
-
"loss": 0.
|
| 1279 |
"step": 8850
|
| 1280 |
},
|
| 1281 |
{
|
| 1282 |
"epoch": 3.1459879816189464,
|
| 1283 |
-
"grad_norm":
|
| 1284 |
-
"learning_rate":
|
| 1285 |
-
"loss": 0.
|
| 1286 |
"step": 8900
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
"epoch": 3.163662071403323,
|
| 1290 |
-
"grad_norm":
|
| 1291 |
-
"learning_rate":
|
| 1292 |
-
"loss": 0.
|
| 1293 |
"step": 8950
|
| 1294 |
},
|
| 1295 |
{
|
| 1296 |
"epoch": 3.1813361611876987,
|
| 1297 |
-
"grad_norm":
|
| 1298 |
-
"learning_rate":
|
| 1299 |
-
"loss": 0.
|
| 1300 |
"step": 9000
|
| 1301 |
},
|
| 1302 |
{
|
| 1303 |
"epoch": 3.199010250972075,
|
| 1304 |
-
"grad_norm":
|
| 1305 |
-
"learning_rate":
|
| 1306 |
-
"loss": 0.
|
| 1307 |
"step": 9050
|
| 1308 |
},
|
| 1309 |
{
|
| 1310 |
"epoch": 3.216684340756451,
|
| 1311 |
-
"grad_norm":
|
| 1312 |
-
"learning_rate":
|
| 1313 |
-
"loss": 0.
|
| 1314 |
"step": 9100
|
| 1315 |
},
|
| 1316 |
{
|
| 1317 |
"epoch": 3.2343584305408273,
|
| 1318 |
-
"grad_norm":
|
| 1319 |
-
"learning_rate":
|
| 1320 |
-
"loss": 0.
|
| 1321 |
"step": 9150
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"epoch": 3.252032520325203,
|
| 1325 |
-
"grad_norm":
|
| 1326 |
-
"learning_rate":
|
| 1327 |
-
"loss": 0.
|
| 1328 |
"step": 9200
|
| 1329 |
},
|
| 1330 |
{
|
| 1331 |
"epoch": 3.2697066101095795,
|
| 1332 |
-
"grad_norm":
|
| 1333 |
-
"learning_rate":
|
| 1334 |
-
"loss": 0.
|
| 1335 |
"step": 9250
|
| 1336 |
},
|
| 1337 |
{
|
| 1338 |
"epoch": 3.2873806998939554,
|
| 1339 |
-
"grad_norm":
|
| 1340 |
-
"learning_rate":
|
| 1341 |
-
"loss": 0.
|
| 1342 |
"step": 9300
|
| 1343 |
},
|
| 1344 |
{
|
| 1345 |
"epoch": 3.3050547896783318,
|
| 1346 |
-
"grad_norm":
|
| 1347 |
-
"learning_rate":
|
| 1348 |
-
"loss": 0.
|
| 1349 |
"step": 9350
|
| 1350 |
},
|
| 1351 |
{
|
| 1352 |
"epoch": 3.3227288794627077,
|
| 1353 |
-
"grad_norm":
|
| 1354 |
-
"learning_rate":
|
| 1355 |
-
"loss": 0.
|
| 1356 |
"step": 9400
|
| 1357 |
},
|
| 1358 |
{
|
| 1359 |
"epoch": 3.3404029692470836,
|
| 1360 |
-
"grad_norm":
|
| 1361 |
-
"learning_rate":
|
| 1362 |
-
"loss": 0.
|
| 1363 |
"step": 9450
|
| 1364 |
},
|
| 1365 |
{
|
| 1366 |
"epoch": 3.35807705903146,
|
| 1367 |
-
"grad_norm":
|
| 1368 |
-
"learning_rate":
|
| 1369 |
-
"loss": 0.
|
| 1370 |
"step": 9500
|
| 1371 |
},
|
| 1372 |
{
|
| 1373 |
"epoch": 3.375751148815836,
|
| 1374 |
-
"grad_norm":
|
| 1375 |
-
"learning_rate":
|
| 1376 |
-
"loss": 0.
|
| 1377 |
"step": 9550
|
| 1378 |
},
|
| 1379 |
{
|
| 1380 |
"epoch": 3.393425238600212,
|
| 1381 |
-
"grad_norm":
|
| 1382 |
-
"learning_rate":
|
| 1383 |
-
"loss": 0.
|
| 1384 |
"step": 9600
|
| 1385 |
},
|
| 1386 |
{
|
| 1387 |
"epoch": 3.411099328384588,
|
| 1388 |
-
"grad_norm":
|
| 1389 |
-
"learning_rate":
|
| 1390 |
-
"loss": 0.
|
| 1391 |
"step": 9650
|
| 1392 |
},
|
| 1393 |
{
|
| 1394 |
"epoch": 3.4287734181689644,
|
| 1395 |
-
"grad_norm":
|
| 1396 |
-
"learning_rate":
|
| 1397 |
-
"loss": 0.
|
| 1398 |
"step": 9700
|
| 1399 |
},
|
| 1400 |
{
|
| 1401 |
"epoch": 3.4464475079533403,
|
| 1402 |
-
"grad_norm":
|
| 1403 |
-
"learning_rate":
|
| 1404 |
-
"loss": 0.
|
| 1405 |
"step": 9750
|
| 1406 |
},
|
| 1407 |
{
|
| 1408 |
"epoch": 3.4641215977377167,
|
| 1409 |
-
"grad_norm":
|
| 1410 |
-
"learning_rate":
|
| 1411 |
-
"loss": 0.
|
| 1412 |
"step": 9800
|
| 1413 |
},
|
| 1414 |
{
|
| 1415 |
"epoch": 3.4817956875220926,
|
| 1416 |
-
"grad_norm":
|
| 1417 |
-
"learning_rate":
|
| 1418 |
-
"loss": 0.
|
| 1419 |
"step": 9850
|
| 1420 |
},
|
| 1421 |
{
|
| 1422 |
"epoch": 3.499469777306469,
|
| 1423 |
-
"grad_norm":
|
| 1424 |
-
"learning_rate":
|
| 1425 |
-
"loss": 0.
|
| 1426 |
"step": 9900
|
| 1427 |
},
|
| 1428 |
{
|
| 1429 |
"epoch": 3.517143867090845,
|
| 1430 |
-
"grad_norm":
|
| 1431 |
-
"learning_rate":
|
| 1432 |
-
"loss": 0.
|
| 1433 |
"step": 9950
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
"epoch": 3.534817956875221,
|
| 1437 |
-
"grad_norm":
|
| 1438 |
-
"learning_rate":
|
| 1439 |
-
"loss": 0.
|
| 1440 |
"step": 10000
|
| 1441 |
},
|
| 1442 |
{
|
| 1443 |
"epoch": 3.552492046659597,
|
| 1444 |
-
"grad_norm":
|
| 1445 |
-
"learning_rate":
|
| 1446 |
-
"loss": 0.
|
| 1447 |
"step": 10050
|
| 1448 |
},
|
| 1449 |
{
|
| 1450 |
"epoch": 3.570166136443973,
|
| 1451 |
-
"grad_norm":
|
| 1452 |
-
"learning_rate":
|
| 1453 |
-
"loss": 0.
|
| 1454 |
"step": 10100
|
| 1455 |
},
|
| 1456 |
{
|
| 1457 |
"epoch": 3.5878402262283493,
|
| 1458 |
-
"grad_norm":
|
| 1459 |
-
"learning_rate":
|
| 1460 |
-
"loss": 0.
|
| 1461 |
"step": 10150
|
| 1462 |
},
|
| 1463 |
{
|
| 1464 |
"epoch": 3.605514316012725,
|
| 1465 |
-
"grad_norm":
|
| 1466 |
-
"learning_rate":
|
| 1467 |
-
"loss": 0.
|
| 1468 |
"step": 10200
|
| 1469 |
},
|
| 1470 |
{
|
| 1471 |
"epoch": 3.6231884057971016,
|
| 1472 |
-
"grad_norm":
|
| 1473 |
-
"learning_rate":
|
| 1474 |
-
"loss": 0.
|
| 1475 |
"step": 10250
|
| 1476 |
},
|
| 1477 |
{
|
| 1478 |
"epoch": 3.6408624955814775,
|
| 1479 |
-
"grad_norm":
|
| 1480 |
-
"learning_rate":
|
| 1481 |
-
"loss": 0.
|
| 1482 |
"step": 10300
|
| 1483 |
},
|
| 1484 |
{
|
| 1485 |
"epoch": 3.658536585365854,
|
| 1486 |
-
"grad_norm":
|
| 1487 |
-
"learning_rate":
|
| 1488 |
-
"loss": 0.
|
| 1489 |
"step": 10350
|
| 1490 |
},
|
| 1491 |
{
|
| 1492 |
"epoch": 3.6762106751502297,
|
| 1493 |
-
"grad_norm":
|
| 1494 |
-
"learning_rate":
|
| 1495 |
-
"loss": 0.
|
| 1496 |
"step": 10400
|
| 1497 |
},
|
| 1498 |
{
|
| 1499 |
"epoch": 3.693884764934606,
|
| 1500 |
-
"grad_norm":
|
| 1501 |
-
"learning_rate":
|
| 1502 |
-
"loss": 0.
|
| 1503 |
"step": 10450
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 3.711558854718982,
|
| 1507 |
-
"grad_norm":
|
| 1508 |
-
"learning_rate":
|
| 1509 |
-
"loss": 0.
|
| 1510 |
"step": 10500
|
| 1511 |
},
|
| 1512 |
{
|
| 1513 |
"epoch": 3.729232944503358,
|
| 1514 |
-
"grad_norm":
|
| 1515 |
-
"learning_rate":
|
| 1516 |
-
"loss": 0.
|
| 1517 |
"step": 10550
|
| 1518 |
},
|
| 1519 |
{
|
| 1520 |
"epoch": 3.746907034287734,
|
| 1521 |
-
"grad_norm":
|
| 1522 |
-
"learning_rate":
|
| 1523 |
-
"loss": 0.
|
| 1524 |
"step": 10600
|
| 1525 |
},
|
| 1526 |
{
|
| 1527 |
"epoch": 3.7645811240721105,
|
| 1528 |
-
"grad_norm":
|
| 1529 |
-
"learning_rate":
|
| 1530 |
-
"loss": 0.
|
| 1531 |
"step": 10650
|
| 1532 |
},
|
| 1533 |
{
|
| 1534 |
"epoch": 3.7822552138564864,
|
| 1535 |
-
"grad_norm":
|
| 1536 |
-
"learning_rate":
|
| 1537 |
-
"loss": 0.
|
| 1538 |
"step": 10700
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
"epoch": 3.7999293036408623,
|
| 1542 |
-
"grad_norm":
|
| 1543 |
-
"learning_rate":
|
| 1544 |
-
"loss": 0.
|
| 1545 |
"step": 10750
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 3.8176033934252387,
|
| 1549 |
-
"grad_norm":
|
| 1550 |
-
"learning_rate":
|
| 1551 |
-
"loss": 0.
|
| 1552 |
"step": 10800
|
| 1553 |
},
|
| 1554 |
{
|
| 1555 |
"epoch": 3.8352774832096146,
|
| 1556 |
-
"grad_norm":
|
| 1557 |
-
"learning_rate":
|
| 1558 |
-
"loss": 0.
|
| 1559 |
"step": 10850
|
| 1560 |
},
|
| 1561 |
{
|
| 1562 |
"epoch": 3.852951572993991,
|
| 1563 |
-
"grad_norm":
|
| 1564 |
-
"learning_rate":
|
| 1565 |
-
"loss": 0.
|
| 1566 |
"step": 10900
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 3.870625662778367,
|
| 1570 |
-
"grad_norm":
|
| 1571 |
-
"learning_rate":
|
| 1572 |
-
"loss": 0.
|
| 1573 |
"step": 10950
|
| 1574 |
},
|
| 1575 |
{
|
| 1576 |
"epoch": 3.888299752562743,
|
| 1577 |
-
"grad_norm":
|
| 1578 |
-
"learning_rate":
|
| 1579 |
-
"loss": 0.
|
| 1580 |
"step": 11000
|
| 1581 |
},
|
| 1582 |
{
|
| 1583 |
"epoch": 3.905973842347119,
|
| 1584 |
-
"grad_norm":
|
| 1585 |
-
"learning_rate":
|
| 1586 |
-
"loss": 0.
|
| 1587 |
"step": 11050
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
"epoch": 3.9236479321314954,
|
| 1591 |
-
"grad_norm":
|
| 1592 |
-
"learning_rate":
|
| 1593 |
-
"loss": 0.
|
| 1594 |
"step": 11100
|
| 1595 |
},
|
| 1596 |
{
|
| 1597 |
"epoch": 3.9413220219158713,
|
| 1598 |
-
"grad_norm":
|
| 1599 |
-
"learning_rate":
|
| 1600 |
-
"loss": 0.
|
| 1601 |
"step": 11150
|
| 1602 |
},
|
| 1603 |
{
|
| 1604 |
"epoch": 3.9589961117002472,
|
| 1605 |
-
"grad_norm":
|
| 1606 |
-
"learning_rate":
|
| 1607 |
-
"loss": 0.
|
| 1608 |
"step": 11200
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
"epoch": 3.9766702014846236,
|
| 1612 |
-
"grad_norm":
|
| 1613 |
-
"learning_rate":
|
| 1614 |
-
"loss": 0.
|
| 1615 |
"step": 11250
|
| 1616 |
},
|
| 1617 |
{
|
| 1618 |
"epoch": 3.9943442912689995,
|
| 1619 |
-
"grad_norm":
|
| 1620 |
-
"learning_rate":
|
| 1621 |
-
"loss": 0.
|
| 1622 |
"step": 11300
|
| 1623 |
},
|
| 1624 |
{
|
| 1625 |
"epoch": 4.0,
|
| 1626 |
-
"eval_bertscore_f1": 0.
|
| 1627 |
-
"eval_bleu": 0.
|
| 1628 |
-
"eval_loss": 0.
|
| 1629 |
-
"
|
| 1630 |
-
"
|
| 1631 |
-
"
|
|
|
|
|
|
|
|
|
|
| 1632 |
"step": 11316
|
| 1633 |
},
|
| 1634 |
{
|
| 1635 |
"epoch": 4.012018381053376,
|
| 1636 |
-
"grad_norm":
|
| 1637 |
-
"learning_rate":
|
| 1638 |
-
"loss": 0.
|
| 1639 |
"step": 11350
|
| 1640 |
},
|
| 1641 |
{
|
| 1642 |
"epoch": 4.029692470837752,
|
| 1643 |
-
"grad_norm":
|
| 1644 |
-
"learning_rate":
|
| 1645 |
-
"loss": 0.
|
| 1646 |
"step": 11400
|
| 1647 |
},
|
| 1648 |
{
|
| 1649 |
"epoch": 4.047366560622128,
|
| 1650 |
-
"grad_norm":
|
| 1651 |
-
"learning_rate":
|
| 1652 |
-
"loss": 0.
|
| 1653 |
"step": 11450
|
| 1654 |
},
|
| 1655 |
{
|
| 1656 |
"epoch": 4.065040650406504,
|
| 1657 |
-
"grad_norm":
|
| 1658 |
-
"learning_rate":
|
| 1659 |
-
"loss": 0.
|
| 1660 |
"step": 11500
|
| 1661 |
},
|
| 1662 |
{
|
| 1663 |
"epoch": 4.08271474019088,
|
| 1664 |
-
"grad_norm":
|
| 1665 |
-
"learning_rate":
|
| 1666 |
-
"loss": 0.
|
| 1667 |
"step": 11550
|
| 1668 |
},
|
| 1669 |
{
|
| 1670 |
"epoch": 4.100388829975256,
|
| 1671 |
-
"grad_norm":
|
| 1672 |
-
"learning_rate":
|
| 1673 |
-
"loss": 0.
|
| 1674 |
"step": 11600
|
| 1675 |
},
|
| 1676 |
{
|
| 1677 |
"epoch": 4.118062919759632,
|
| 1678 |
-
"grad_norm":
|
| 1679 |
-
"learning_rate":
|
| 1680 |
-
"loss": 0.
|
| 1681 |
"step": 11650
|
| 1682 |
},
|
| 1683 |
{
|
| 1684 |
"epoch": 4.135737009544009,
|
| 1685 |
-
"grad_norm":
|
| 1686 |
-
"learning_rate":
|
| 1687 |
-
"loss": 0.
|
| 1688 |
"step": 11700
|
| 1689 |
},
|
| 1690 |
{
|
| 1691 |
"epoch": 4.153411099328385,
|
| 1692 |
-
"grad_norm":
|
| 1693 |
-
"learning_rate":
|
| 1694 |
-
"loss": 0.
|
| 1695 |
"step": 11750
|
| 1696 |
},
|
| 1697 |
{
|
| 1698 |
"epoch": 4.171085189112761,
|
| 1699 |
-
"grad_norm":
|
| 1700 |
-
"learning_rate":
|
| 1701 |
-
"loss": 0.
|
| 1702 |
"step": 11800
|
| 1703 |
},
|
| 1704 |
{
|
| 1705 |
"epoch": 4.188759278897137,
|
| 1706 |
-
"grad_norm":
|
| 1707 |
-
"learning_rate":
|
| 1708 |
-
"loss": 0.
|
| 1709 |
"step": 11850
|
| 1710 |
},
|
| 1711 |
{
|
| 1712 |
"epoch": 4.2064333686815125,
|
| 1713 |
-
"grad_norm":
|
| 1714 |
-
"learning_rate":
|
| 1715 |
-
"loss": 0.
|
| 1716 |
"step": 11900
|
| 1717 |
},
|
| 1718 |
{
|
| 1719 |
"epoch": 4.224107458465889,
|
| 1720 |
-
"grad_norm":
|
| 1721 |
-
"learning_rate":
|
| 1722 |
-
"loss": 0.
|
| 1723 |
"step": 11950
|
| 1724 |
},
|
| 1725 |
{
|
| 1726 |
"epoch": 4.241781548250265,
|
| 1727 |
-
"grad_norm":
|
| 1728 |
-
"learning_rate":
|
| 1729 |
-
"loss": 0.
|
| 1730 |
"step": 12000
|
| 1731 |
},
|
| 1732 |
{
|
| 1733 |
"epoch": 4.259455638034641,
|
| 1734 |
-
"grad_norm":
|
| 1735 |
-
"learning_rate":
|
| 1736 |
-
"loss": 0.
|
| 1737 |
"step": 12050
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 4.277129727819017,
|
| 1741 |
-
"grad_norm":
|
| 1742 |
-
"learning_rate":
|
| 1743 |
-
"loss": 0.
|
| 1744 |
"step": 12100
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 4.294803817603394,
|
| 1748 |
-
"grad_norm":
|
| 1749 |
-
"learning_rate":
|
| 1750 |
-
"loss": 0.
|
| 1751 |
"step": 12150
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 4.31247790738777,
|
| 1755 |
-
"grad_norm":
|
| 1756 |
-
"learning_rate":
|
| 1757 |
-
"loss": 0.
|
| 1758 |
"step": 12200
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 4.330151997172146,
|
| 1762 |
-
"grad_norm":
|
| 1763 |
-
"learning_rate":
|
| 1764 |
-
"loss": 0.
|
| 1765 |
"step": 12250
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 4.3478260869565215,
|
| 1769 |
-
"grad_norm":
|
| 1770 |
-
"learning_rate":
|
| 1771 |
-
"loss": 0.
|
| 1772 |
"step": 12300
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 4.365500176740898,
|
| 1776 |
-
"grad_norm":
|
| 1777 |
-
"learning_rate":
|
| 1778 |
-
"loss": 0.
|
| 1779 |
"step": 12350
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 4.383174266525274,
|
| 1783 |
-
"grad_norm":
|
| 1784 |
-
"learning_rate":
|
| 1785 |
-
"loss": 0.
|
| 1786 |
"step": 12400
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 4.40084835630965,
|
| 1790 |
-
"grad_norm":
|
| 1791 |
-
"learning_rate":
|
| 1792 |
-
"loss": 0.
|
| 1793 |
"step": 12450
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 4.418522446094026,
|
| 1797 |
-
"grad_norm":
|
| 1798 |
-
"learning_rate":
|
| 1799 |
-
"loss": 0.
|
| 1800 |
"step": 12500
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 4.436196535878402,
|
| 1804 |
-
"grad_norm":
|
| 1805 |
-
"learning_rate":
|
| 1806 |
-
"loss": 0.
|
| 1807 |
"step": 12550
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 4.453870625662779,
|
| 1811 |
-
"grad_norm":
|
| 1812 |
-
"learning_rate":
|
| 1813 |
-
"loss": 0.
|
| 1814 |
"step": 12600
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 4.471544715447155,
|
| 1818 |
-
"grad_norm":
|
| 1819 |
-
"learning_rate":
|
| 1820 |
-
"loss": 0.
|
| 1821 |
"step": 12650
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 4.4892188052315305,
|
| 1825 |
-
"grad_norm":
|
| 1826 |
-
"learning_rate":
|
| 1827 |
-
"loss": 0.
|
| 1828 |
"step": 12700
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 4.506892895015906,
|
| 1832 |
-
"grad_norm":
|
| 1833 |
-
"learning_rate":
|
| 1834 |
-
"loss": 0.
|
| 1835 |
"step": 12750
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 4.524566984800282,
|
| 1839 |
-
"grad_norm":
|
| 1840 |
-
"learning_rate":
|
| 1841 |
-
"loss": 0.
|
| 1842 |
"step": 12800
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 4.542241074584659,
|
| 1846 |
-
"grad_norm":
|
| 1847 |
-
"learning_rate":
|
| 1848 |
-
"loss": 0.
|
| 1849 |
"step": 12850
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 4.559915164369035,
|
| 1853 |
-
"grad_norm":
|
| 1854 |
-
"learning_rate":
|
| 1855 |
-
"loss": 0.
|
| 1856 |
"step": 12900
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 4.577589254153411,
|
| 1860 |
-
"grad_norm":
|
| 1861 |
-
"learning_rate":
|
| 1862 |
-
"loss": 0.
|
| 1863 |
"step": 12950
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 4.595263343937788,
|
| 1867 |
-
"grad_norm":
|
| 1868 |
-
"learning_rate":
|
| 1869 |
-
"loss": 0.
|
| 1870 |
"step": 13000
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 4.612937433722164,
|
| 1874 |
-
"grad_norm":
|
| 1875 |
-
"learning_rate":
|
| 1876 |
-
"loss": 0.
|
| 1877 |
"step": 13050
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 4.6306115235065395,
|
| 1881 |
-
"grad_norm":
|
| 1882 |
-
"learning_rate":
|
| 1883 |
-
"loss": 0.
|
| 1884 |
"step": 13100
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 4.648285613290915,
|
| 1888 |
-
"grad_norm":
|
| 1889 |
-
"learning_rate":
|
| 1890 |
-
"loss": 0.
|
| 1891 |
"step": 13150
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 4.665959703075291,
|
| 1895 |
-
"grad_norm":
|
| 1896 |
-
"learning_rate":
|
| 1897 |
-
"loss": 0.
|
| 1898 |
"step": 13200
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 4.683633792859668,
|
| 1902 |
-
"grad_norm":
|
| 1903 |
-
"learning_rate":
|
| 1904 |
-
"loss": 0.
|
| 1905 |
"step": 13250
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 4.701307882644044,
|
| 1909 |
-
"grad_norm":
|
| 1910 |
-
"learning_rate":
|
| 1911 |
-
"loss": 0.
|
| 1912 |
"step": 13300
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 4.71898197242842,
|
| 1916 |
-
"grad_norm":
|
| 1917 |
-
"learning_rate":
|
| 1918 |
-
"loss": 0.
|
| 1919 |
"step": 13350
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 4.736656062212796,
|
| 1923 |
-
"grad_norm":
|
| 1924 |
-
"learning_rate":
|
| 1925 |
-
"loss": 0.
|
| 1926 |
"step": 13400
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 4.754330151997172,
|
| 1930 |
-
"grad_norm":
|
| 1931 |
-
"learning_rate":
|
| 1932 |
-
"loss": 0.
|
| 1933 |
"step": 13450
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 4.7720042417815485,
|
| 1937 |
-
"grad_norm":
|
| 1938 |
-
"learning_rate":
|
| 1939 |
-
"loss": 0.
|
| 1940 |
"step": 13500
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 4.789678331565924,
|
| 1944 |
-
"grad_norm":
|
| 1945 |
-
"learning_rate":
|
| 1946 |
-
"loss": 0.
|
| 1947 |
"step": 13550
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 4.8073524213503,
|
| 1951 |
-
"grad_norm":
|
| 1952 |
-
"learning_rate": 1.
|
| 1953 |
-
"loss": 0.
|
| 1954 |
"step": 13600
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 4.825026511134676,
|
| 1958 |
-
"grad_norm":
|
| 1959 |
-
"learning_rate": 1.
|
| 1960 |
-
"loss": 0.
|
| 1961 |
"step": 13650
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 4.842700600919053,
|
| 1965 |
-
"grad_norm":
|
| 1966 |
-
"learning_rate":
|
| 1967 |
-
"loss": 0.
|
| 1968 |
"step": 13700
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 4.860374690703429,
|
| 1972 |
-
"grad_norm":
|
| 1973 |
-
"learning_rate":
|
| 1974 |
-
"loss": 0.
|
| 1975 |
"step": 13750
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 4.878048780487805,
|
| 1979 |
-
"grad_norm":
|
| 1980 |
-
"learning_rate":
|
| 1981 |
-
"loss": 0.
|
| 1982 |
"step": 13800
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 4.895722870272181,
|
| 1986 |
-
"grad_norm":
|
| 1987 |
-
"learning_rate":
|
| 1988 |
-
"loss": 0.
|
| 1989 |
"step": 13850
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 4.9133969600565575,
|
| 1993 |
-
"grad_norm":
|
| 1994 |
-
"learning_rate":
|
| 1995 |
-
"loss": 0.
|
| 1996 |
"step": 13900
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 4.931071049840933,
|
| 2000 |
-
"grad_norm":
|
| 2001 |
-
"learning_rate":
|
| 2002 |
-
"loss": 0.
|
| 2003 |
"step": 13950
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 4.948745139625309,
|
| 2007 |
-
"grad_norm":
|
| 2008 |
-
"learning_rate":
|
| 2009 |
-
"loss": 0.
|
| 2010 |
"step": 14000
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 4.966419229409685,
|
| 2014 |
-
"grad_norm":
|
| 2015 |
-
"learning_rate":
|
| 2016 |
-
"loss": 0.
|
| 2017 |
"step": 14050
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 4.984093319194061,
|
| 2021 |
-
"grad_norm":
|
| 2022 |
-
"learning_rate": 1.
|
| 2023 |
-
"loss": 0.
|
| 2024 |
"step": 14100
|
| 2025 |
},
|
| 2026 |
{
|
| 2027 |
"epoch": 5.0,
|
| 2028 |
-
"eval_bertscore_f1": 0.
|
| 2029 |
-
"eval_bleu": 0.
|
| 2030 |
-
"eval_loss": 0.
|
| 2031 |
-
"
|
| 2032 |
-
"
|
| 2033 |
-
"
|
|
|
|
|
|
|
|
|
|
| 2034 |
"step": 14145
|
| 2035 |
}
|
| 2036 |
],
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 14145,
|
| 3 |
+
"best_metric": 0.9658045416033947,
|
| 4 |
+
"best_model_checkpoint": "/kaggle/working/codet5-k8s-lora/checkpoint-14145",
|
| 5 |
"epoch": 5.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 14145,
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.017674089784376106,
|
| 14 |
+
"grad_norm": 2.7520432472229004,
|
| 15 |
+
"learning_rate": 4.983739837398374e-05,
|
| 16 |
+
"loss": 4.115,
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.03534817956875221,
|
| 21 |
+
"grad_norm": 17.1498966217041,
|
| 22 |
+
"learning_rate": 4.966065747613998e-05,
|
| 23 |
+
"loss": 3.2283,
|
| 24 |
"step": 100
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.053022269353128315,
|
| 28 |
+
"grad_norm": 2.4044106006622314,
|
| 29 |
+
"learning_rate": 4.948391657829622e-05,
|
| 30 |
+
"loss": 2.2704,
|
| 31 |
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.07069635913750442,
|
| 35 |
+
"grad_norm": 2.3264973163604736,
|
| 36 |
+
"learning_rate": 4.930717568045246e-05,
|
| 37 |
+
"loss": 1.963,
|
| 38 |
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.08837044892188052,
|
| 42 |
+
"grad_norm": 2.146696090698242,
|
| 43 |
+
"learning_rate": 4.91304347826087e-05,
|
| 44 |
+
"loss": 1.8358,
|
| 45 |
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.10604453870625663,
|
| 49 |
+
"grad_norm": 2.6933417320251465,
|
| 50 |
+
"learning_rate": 4.895369388476494e-05,
|
| 51 |
+
"loss": 1.7542,
|
| 52 |
"step": 300
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.12371862849063273,
|
| 56 |
+
"grad_norm": 2.71453857421875,
|
| 57 |
+
"learning_rate": 4.8776952986921177e-05,
|
| 58 |
+
"loss": 1.6651,
|
| 59 |
"step": 350
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.14139271827500885,
|
| 63 |
+
"grad_norm": 2.239396095275879,
|
| 64 |
+
"learning_rate": 4.8600212089077416e-05,
|
| 65 |
+
"loss": 1.5549,
|
| 66 |
"step": 400
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.15906680805938495,
|
| 70 |
+
"grad_norm": 3.038501262664795,
|
| 71 |
+
"learning_rate": 4.8423471191233655e-05,
|
| 72 |
+
"loss": 1.4887,
|
| 73 |
"step": 450
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.17674089784376104,
|
| 77 |
+
"grad_norm": 2.988568067550659,
|
| 78 |
+
"learning_rate": 4.825026511134676e-05,
|
| 79 |
+
"loss": 1.4896,
|
| 80 |
"step": 500
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.19441498762813716,
|
| 84 |
+
"grad_norm": 2.7871344089508057,
|
| 85 |
+
"learning_rate": 4.807352421350301e-05,
|
| 86 |
+
"loss": 1.3147,
|
| 87 |
"step": 550
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.21208907741251326,
|
| 91 |
+
"grad_norm": 2.9361987113952637,
|
| 92 |
+
"learning_rate": 4.789678331565924e-05,
|
| 93 |
+
"loss": 1.4059,
|
| 94 |
"step": 600
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.22976316719688936,
|
| 98 |
+
"grad_norm": 3.7504327297210693,
|
| 99 |
+
"learning_rate": 4.7720042417815487e-05,
|
| 100 |
+
"loss": 1.3733,
|
| 101 |
"step": 650
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.24743725698126545,
|
| 105 |
+
"grad_norm": 2.8540990352630615,
|
| 106 |
+
"learning_rate": 4.7543301519971726e-05,
|
| 107 |
+
"loss": 1.2048,
|
| 108 |
"step": 700
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.2651113467656416,
|
| 112 |
+
"grad_norm": 3.3986833095550537,
|
| 113 |
+
"learning_rate": 4.7366560622127965e-05,
|
| 114 |
+
"loss": 1.3204,
|
| 115 |
"step": 750
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.2827854365500177,
|
| 119 |
+
"grad_norm": 3.490173816680908,
|
| 120 |
+
"learning_rate": 4.7189819724284204e-05,
|
| 121 |
+
"loss": 1.2636,
|
| 122 |
"step": 800
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.30045952633439377,
|
| 126 |
+
"grad_norm": 3.0693671703338623,
|
| 127 |
+
"learning_rate": 4.701307882644044e-05,
|
| 128 |
+
"loss": 1.1539,
|
| 129 |
"step": 850
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.3181336161187699,
|
| 133 |
+
"grad_norm": 6.666368007659912,
|
| 134 |
+
"learning_rate": 4.683633792859668e-05,
|
| 135 |
+
"loss": 1.1835,
|
| 136 |
"step": 900
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.335807705903146,
|
| 140 |
+
"grad_norm": 3.3494420051574707,
|
| 141 |
+
"learning_rate": 4.6659597030752915e-05,
|
| 142 |
+
"loss": 1.0956,
|
| 143 |
"step": 950
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.3534817956875221,
|
| 147 |
+
"grad_norm": 2.6836087703704834,
|
| 148 |
+
"learning_rate": 4.648285613290916e-05,
|
| 149 |
+
"loss": 1.1311,
|
| 150 |
"step": 1000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.3711558854718982,
|
| 154 |
+
"grad_norm": 2.8905951976776123,
|
| 155 |
+
"learning_rate": 4.630611523506539e-05,
|
| 156 |
+
"loss": 1.0765,
|
| 157 |
"step": 1050
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.38882997525627433,
|
| 161 |
+
"grad_norm": 3.939500093460083,
|
| 162 |
+
"learning_rate": 4.612937433722164e-05,
|
| 163 |
+
"loss": 1.0456,
|
| 164 |
"step": 1100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.4065040650406504,
|
| 168 |
+
"grad_norm": 3.1198384761810303,
|
| 169 |
+
"learning_rate": 4.595263343937787e-05,
|
| 170 |
+
"loss": 1.0448,
|
| 171 |
"step": 1150
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.4241781548250265,
|
| 175 |
+
"grad_norm": 3.0342133045196533,
|
| 176 |
+
"learning_rate": 4.577589254153412e-05,
|
| 177 |
+
"loss": 0.9943,
|
| 178 |
"step": 1200
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.4418522446094026,
|
| 182 |
+
"grad_norm": 3.555405616760254,
|
| 183 |
+
"learning_rate": 4.559915164369035e-05,
|
| 184 |
+
"loss": 1.045,
|
| 185 |
"step": 1250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.4595263343937787,
|
| 189 |
+
"grad_norm": 3.0326290130615234,
|
| 190 |
+
"learning_rate": 4.542241074584659e-05,
|
| 191 |
+
"loss": 0.9807,
|
| 192 |
"step": 1300
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.47720042417815484,
|
| 196 |
+
"grad_norm": 3.1818552017211914,
|
| 197 |
+
"learning_rate": 4.524566984800283e-05,
|
| 198 |
+
"loss": 0.9889,
|
| 199 |
"step": 1350
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.4948745139625309,
|
| 203 |
+
"grad_norm": 2.933518171310425,
|
| 204 |
+
"learning_rate": 4.506892895015907e-05,
|
| 205 |
+
"loss": 0.9373,
|
| 206 |
"step": 1400
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.512548603746907,
|
| 210 |
+
"grad_norm": 4.3423566818237305,
|
| 211 |
+
"learning_rate": 4.489218805231531e-05,
|
| 212 |
+
"loss": 1.0253,
|
| 213 |
"step": 1450
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.5302226935312832,
|
| 217 |
+
"grad_norm": 3.6613636016845703,
|
| 218 |
+
"learning_rate": 4.4715447154471546e-05,
|
| 219 |
+
"loss": 0.9458,
|
| 220 |
"step": 1500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.5478967833156593,
|
| 224 |
+
"grad_norm": 2.736353635787964,
|
| 225 |
+
"learning_rate": 4.454224107458466e-05,
|
| 226 |
+
"loss": 0.9205,
|
| 227 |
"step": 1550
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.5655708731000354,
|
| 231 |
+
"grad_norm": 3.9614179134368896,
|
| 232 |
+
"learning_rate": 4.43655001767409e-05,
|
| 233 |
+
"loss": 0.8896,
|
| 234 |
"step": 1600
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.5832449628844114,
|
| 238 |
+
"grad_norm": 3.397909164428711,
|
| 239 |
+
"learning_rate": 4.418875927889714e-05,
|
| 240 |
+
"loss": 0.9247,
|
| 241 |
"step": 1650
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.6009190526687875,
|
| 245 |
+
"grad_norm": 3.26153564453125,
|
| 246 |
+
"learning_rate": 4.401201838105338e-05,
|
| 247 |
+
"loss": 0.8023,
|
| 248 |
"step": 1700
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.6185931424531637,
|
| 252 |
+
"grad_norm": 4.067619800567627,
|
| 253 |
+
"learning_rate": 4.383527748320962e-05,
|
| 254 |
+
"loss": 0.8967,
|
| 255 |
"step": 1750
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.6362672322375398,
|
| 259 |
+
"grad_norm": 3.9123120307922363,
|
| 260 |
+
"learning_rate": 4.3658536585365856e-05,
|
| 261 |
+
"loss": 0.8668,
|
| 262 |
"step": 1800
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.6539413220219159,
|
| 266 |
+
"grad_norm": 3.9695286750793457,
|
| 267 |
+
"learning_rate": 4.3481795687522095e-05,
|
| 268 |
+
"loss": 0.8688,
|
| 269 |
"step": 1850
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.671615411806292,
|
| 273 |
+
"grad_norm": 3.176691770553589,
|
| 274 |
+
"learning_rate": 4.3305054789678334e-05,
|
| 275 |
+
"loss": 0.8475,
|
| 276 |
"step": 1900
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.689289501590668,
|
| 280 |
+
"grad_norm": 4.363101005554199,
|
| 281 |
+
"learning_rate": 4.3128313891834574e-05,
|
| 282 |
+
"loss": 0.8343,
|
| 283 |
"step": 1950
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.7069635913750442,
|
| 287 |
+
"grad_norm": 4.429725170135498,
|
| 288 |
+
"learning_rate": 4.295157299399081e-05,
|
| 289 |
+
"loss": 0.8001,
|
| 290 |
"step": 2000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.7246376811594203,
|
| 294 |
+
"grad_norm": 3.035944700241089,
|
| 295 |
+
"learning_rate": 4.277483209614705e-05,
|
| 296 |
+
"loss": 0.7387,
|
| 297 |
"step": 2050
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.7423117709437964,
|
| 301 |
+
"grad_norm": 3.1769495010375977,
|
| 302 |
+
"learning_rate": 4.259809119830329e-05,
|
| 303 |
+
"loss": 0.7837,
|
| 304 |
"step": 2100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.7599858607281725,
|
| 308 |
+
"grad_norm": 3.7096972465515137,
|
| 309 |
+
"learning_rate": 4.242135030045953e-05,
|
| 310 |
+
"loss": 0.8015,
|
| 311 |
"step": 2150
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.7776599505125487,
|
| 315 |
+
"grad_norm": 4.8830718994140625,
|
| 316 |
+
"learning_rate": 4.224460940261576e-05,
|
| 317 |
+
"loss": 0.8145,
|
| 318 |
"step": 2200
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.7953340402969247,
|
| 322 |
+
"grad_norm": 4.858372688293457,
|
| 323 |
+
"learning_rate": 4.206786850477201e-05,
|
| 324 |
+
"loss": 0.7707,
|
| 325 |
"step": 2250
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.8130081300813008,
|
| 329 |
+
"grad_norm": 4.0257697105407715,
|
| 330 |
+
"learning_rate": 4.189112760692824e-05,
|
| 331 |
+
"loss": 0.7365,
|
| 332 |
"step": 2300
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.8306822198656769,
|
| 336 |
+
"grad_norm": 3.4525439739227295,
|
| 337 |
+
"learning_rate": 4.171438670908449e-05,
|
| 338 |
+
"loss": 0.6824,
|
| 339 |
"step": 2350
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.848356309650053,
|
| 343 |
+
"grad_norm": 3.3290514945983887,
|
| 344 |
+
"learning_rate": 4.153764581124072e-05,
|
| 345 |
+
"loss": 0.696,
|
| 346 |
"step": 2400
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.8660303994344292,
|
| 350 |
+
"grad_norm": 3.284925937652588,
|
| 351 |
+
"learning_rate": 4.1360904913396966e-05,
|
| 352 |
+
"loss": 0.7506,
|
| 353 |
"step": 2450
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.8837044892188052,
|
| 357 |
+
"grad_norm": 7.381216526031494,
|
| 358 |
+
"learning_rate": 4.11841640155532e-05,
|
| 359 |
+
"loss": 0.7282,
|
| 360 |
"step": 2500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.9013785790031813,
|
| 364 |
+
"grad_norm": 3.3384296894073486,
|
| 365 |
+
"learning_rate": 4.1007423117709444e-05,
|
| 366 |
+
"loss": 0.7632,
|
| 367 |
"step": 2550
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.9190526687875574,
|
| 371 |
+
"grad_norm": 3.1576461791992188,
|
| 372 |
+
"learning_rate": 4.0830682219865676e-05,
|
| 373 |
+
"loss": 0.739,
|
| 374 |
"step": 2600
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.9367267585719335,
|
| 378 |
+
"grad_norm": 3.626567840576172,
|
| 379 |
+
"learning_rate": 4.0653941322021916e-05,
|
| 380 |
+
"loss": 0.6791,
|
| 381 |
"step": 2650
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.9544008483563097,
|
| 385 |
+
"grad_norm": 4.196743011474609,
|
| 386 |
+
"learning_rate": 4.0477200424178155e-05,
|
| 387 |
+
"loss": 0.7332,
|
| 388 |
"step": 2700
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.9720749381406858,
|
| 392 |
+
"grad_norm": 3.1296284198760986,
|
| 393 |
+
"learning_rate": 4.0300459526334394e-05,
|
| 394 |
+
"loss": 0.6539,
|
| 395 |
"step": 2750
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.9897490279250618,
|
| 399 |
+
"grad_norm": 5.278796672821045,
|
| 400 |
+
"learning_rate": 4.012371862849063e-05,
|
| 401 |
+
"loss": 0.737,
|
| 402 |
"step": 2800
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.0,
|
| 406 |
+
"eval_bertscore_f1": 0.9560778172188761,
|
| 407 |
+
"eval_bleu": 0.4824577747321971,
|
| 408 |
+
"eval_loss": 0.5060375928878784,
|
| 409 |
+
"eval_meteor": 0.6530790735036477,
|
| 410 |
+
"eval_rouge1": 0.7971610347699927,
|
| 411 |
+
"eval_rouge2": 0.6895819892909911,
|
| 412 |
+
"eval_runtime": 1389.2459,
|
| 413 |
+
"eval_samples_per_second": 4.653,
|
| 414 |
+
"eval_steps_per_second": 0.582,
|
| 415 |
"step": 2829
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 1.007423117709438,
|
| 419 |
+
"grad_norm": 3.9626195430755615,
|
| 420 |
+
"learning_rate": 3.994697773064687e-05,
|
| 421 |
+
"loss": 0.7198,
|
| 422 |
"step": 2850
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 1.025097207493814,
|
| 426 |
+
"grad_norm": 4.509051322937012,
|
| 427 |
+
"learning_rate": 3.977023683280312e-05,
|
| 428 |
+
"loss": 0.6777,
|
| 429 |
"step": 2900
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 1.0427712972781902,
|
| 433 |
+
"grad_norm": 3.044351816177368,
|
| 434 |
+
"learning_rate": 3.959349593495935e-05,
|
| 435 |
+
"loss": 0.6675,
|
| 436 |
"step": 2950
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 1.0604453870625663,
|
| 440 |
+
"grad_norm": 3.598339319229126,
|
| 441 |
+
"learning_rate": 3.941675503711559e-05,
|
| 442 |
+
"loss": 0.7256,
|
| 443 |
"step": 3000
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 1.0781194768469424,
|
| 447 |
+
"grad_norm": 5.248291015625,
|
| 448 |
+
"learning_rate": 3.924001413927183e-05,
|
| 449 |
+
"loss": 0.5862,
|
| 450 |
"step": 3050
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 1.0957935666313185,
|
| 454 |
+
"grad_norm": 3.2210874557495117,
|
| 455 |
+
"learning_rate": 3.906327324142807e-05,
|
| 456 |
+
"loss": 0.7295,
|
| 457 |
"step": 3100
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 1.1134676564156947,
|
| 461 |
+
"grad_norm": 3.6727230548858643,
|
| 462 |
+
"learning_rate": 3.888653234358431e-05,
|
| 463 |
+
"loss": 0.6206,
|
| 464 |
"step": 3150
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 1.1311417462000706,
|
| 468 |
+
"grad_norm": 4.341032981872559,
|
| 469 |
+
"learning_rate": 3.870979144574055e-05,
|
| 470 |
+
"loss": 0.676,
|
| 471 |
"step": 3200
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 1.148815835984447,
|
| 475 |
+
"grad_norm": 3.0379395484924316,
|
| 476 |
+
"learning_rate": 3.8533050547896786e-05,
|
| 477 |
+
"loss": 0.6372,
|
| 478 |
"step": 3250
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 1.1664899257688228,
|
| 482 |
+
"grad_norm": 2.846511125564575,
|
| 483 |
+
"learning_rate": 3.8356309650053025e-05,
|
| 484 |
+
"loss": 0.6758,
|
| 485 |
"step": 3300
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 1.184164015553199,
|
| 489 |
+
"grad_norm": 3.267794132232666,
|
| 490 |
+
"learning_rate": 3.817956875220926e-05,
|
| 491 |
+
"loss": 0.6035,
|
| 492 |
"step": 3350
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 1.201838105337575,
|
| 496 |
+
"grad_norm": 5.214766025543213,
|
| 497 |
+
"learning_rate": 3.8002827854365503e-05,
|
| 498 |
+
"loss": 0.6907,
|
| 499 |
"step": 3400
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 1.2195121951219512,
|
| 503 |
+
"grad_norm": 5.761065483093262,
|
| 504 |
+
"learning_rate": 3.7826086956521736e-05,
|
| 505 |
+
"loss": 0.6092,
|
| 506 |
"step": 3450
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 1.2371862849063273,
|
| 510 |
+
"grad_norm": 4.127236843109131,
|
| 511 |
+
"learning_rate": 3.764934605867798e-05,
|
| 512 |
+
"loss": 0.6127,
|
| 513 |
"step": 3500
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 1.2548603746907034,
|
| 517 |
+
"grad_norm": 2.9047141075134277,
|
| 518 |
+
"learning_rate": 3.747260516083422e-05,
|
| 519 |
+
"loss": 0.6486,
|
| 520 |
"step": 3550
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 1.2725344644750796,
|
| 524 |
+
"grad_norm": 3.972148895263672,
|
| 525 |
+
"learning_rate": 3.729586426299046e-05,
|
| 526 |
+
"loss": 0.6933,
|
| 527 |
"step": 3600
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 1.2902085542594557,
|
| 531 |
+
"grad_norm": 3.2735204696655273,
|
| 532 |
+
"learning_rate": 3.71191233651467e-05,
|
| 533 |
+
"loss": 0.6808,
|
| 534 |
"step": 3650
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 1.3078826440438318,
|
| 538 |
+
"grad_norm": 5.26752233505249,
|
| 539 |
+
"learning_rate": 3.694238246730294e-05,
|
| 540 |
+
"loss": 0.6672,
|
| 541 |
"step": 3700
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 1.3255567338282077,
|
| 545 |
+
"grad_norm": 3.852576971054077,
|
| 546 |
+
"learning_rate": 3.676564156945918e-05,
|
| 547 |
+
"loss": 0.6458,
|
| 548 |
"step": 3750
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 1.343230823612584,
|
| 552 |
+
"grad_norm": 5.333845138549805,
|
| 553 |
+
"learning_rate": 3.658890067161541e-05,
|
| 554 |
+
"loss": 0.5771,
|
| 555 |
"step": 3800
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 1.36090491339696,
|
| 559 |
+
"grad_norm": 3.3407108783721924,
|
| 560 |
+
"learning_rate": 3.6412159773771656e-05,
|
| 561 |
+
"loss": 0.6281,
|
| 562 |
"step": 3850
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 1.378579003181336,
|
| 566 |
+
"grad_norm": 3.874502658843994,
|
| 567 |
+
"learning_rate": 3.623541887592789e-05,
|
| 568 |
+
"loss": 0.6716,
|
| 569 |
"step": 3900
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 1.3962530929657122,
|
| 573 |
+
"grad_norm": 23.51529884338379,
|
| 574 |
+
"learning_rate": 3.6058677978084134e-05,
|
| 575 |
+
"loss": 0.5915,
|
| 576 |
"step": 3950
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 1.4139271827500883,
|
| 580 |
+
"grad_norm": 4.40012788772583,
|
| 581 |
+
"learning_rate": 3.588193708024037e-05,
|
| 582 |
+
"loss": 0.6098,
|
| 583 |
"step": 4000
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 1.4316012725344645,
|
| 587 |
+
"grad_norm": 2.98525071144104,
|
| 588 |
+
"learning_rate": 3.570519618239661e-05,
|
| 589 |
+
"loss": 0.5767,
|
| 590 |
"step": 4050
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 1.4492753623188406,
|
| 594 |
+
"grad_norm": 3.8279759883880615,
|
| 595 |
+
"learning_rate": 3.5528455284552845e-05,
|
| 596 |
+
"loss": 0.5843,
|
| 597 |
"step": 4100
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 1.4669494521032167,
|
| 601 |
+
"grad_norm": 2.8104283809661865,
|
| 602 |
+
"learning_rate": 3.5351714386709084e-05,
|
| 603 |
+
"loss": 0.5314,
|
| 604 |
"step": 4150
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 1.4846235418875928,
|
| 608 |
+
"grad_norm": 3.081321954727173,
|
| 609 |
+
"learning_rate": 3.5174973488865324e-05,
|
| 610 |
+
"loss": 0.5187,
|
| 611 |
"step": 4200
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 1.502297631671969,
|
| 615 |
+
"grad_norm": 2.8093416690826416,
|
| 616 |
+
"learning_rate": 3.499823259102156e-05,
|
| 617 |
+
"loss": 0.5663,
|
| 618 |
"step": 4250
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 1.5199717214563448,
|
| 622 |
+
"grad_norm": 3.7971787452697754,
|
| 623 |
+
"learning_rate": 3.48214916931778e-05,
|
| 624 |
+
"loss": 0.5069,
|
| 625 |
"step": 4300
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 1.5376458112407212,
|
| 629 |
+
"grad_norm": 3.116645336151123,
|
| 630 |
+
"learning_rate": 3.464475079533404e-05,
|
| 631 |
+
"loss": 0.4945,
|
| 632 |
"step": 4350
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 1.555319901025097,
|
| 636 |
+
"grad_norm": 2.9984517097473145,
|
| 637 |
+
"learning_rate": 3.446800989749028e-05,
|
| 638 |
+
"loss": 0.5399,
|
| 639 |
"step": 4400
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 1.5729939908094734,
|
| 643 |
+
"grad_norm": 3.3107683658599854,
|
| 644 |
+
"learning_rate": 3.429126899964652e-05,
|
| 645 |
+
"loss": 0.5507,
|
| 646 |
"step": 4450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 1.5906680805938493,
|
| 650 |
+
"grad_norm": 4.328573226928711,
|
| 651 |
+
"learning_rate": 3.411452810180276e-05,
|
| 652 |
+
"loss": 0.5695,
|
| 653 |
"step": 4500
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 1.6083421703782255,
|
| 657 |
+
"grad_norm": 4.086219787597656,
|
| 658 |
+
"learning_rate": 3.3937787203959e-05,
|
| 659 |
+
"loss": 0.5234,
|
| 660 |
"step": 4550
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 1.6260162601626016,
|
| 664 |
+
"grad_norm": 3.971590280532837,
|
| 665 |
+
"learning_rate": 3.376104630611524e-05,
|
| 666 |
+
"loss": 0.6107,
|
| 667 |
"step": 4600
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 1.6436903499469777,
|
| 671 |
+
"grad_norm": 3.0131218433380127,
|
| 672 |
+
"learning_rate": 3.3584305408271476e-05,
|
| 673 |
+
"loss": 0.5763,
|
| 674 |
"step": 4650
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 1.6613644397313538,
|
| 678 |
+
"grad_norm": 3.862619161605835,
|
| 679 |
+
"learning_rate": 3.3407564510427716e-05,
|
| 680 |
+
"loss": 0.5422,
|
| 681 |
"step": 4700
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 1.67903852951573,
|
| 685 |
+
"grad_norm": 4.276158332824707,
|
| 686 |
+
"learning_rate": 3.3230823612583955e-05,
|
| 687 |
+
"loss": 0.6326,
|
| 688 |
"step": 4750
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 1.696712619300106,
|
| 692 |
+
"grad_norm": 4.451511383056641,
|
| 693 |
+
"learning_rate": 3.3054082714740194e-05,
|
| 694 |
+
"loss": 0.5883,
|
| 695 |
"step": 4800
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 1.714386709084482,
|
| 699 |
+
"grad_norm": 3.8793303966522217,
|
| 700 |
+
"learning_rate": 3.2877341816896426e-05,
|
| 701 |
+
"loss": 0.4958,
|
| 702 |
"step": 4850
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 1.7320607988688583,
|
| 706 |
+
"grad_norm": 3.0720949172973633,
|
| 707 |
+
"learning_rate": 3.270060091905267e-05,
|
| 708 |
+
"loss": 0.5962,
|
| 709 |
"step": 4900
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 1.7497348886532342,
|
| 713 |
+
"grad_norm": 3.674368143081665,
|
| 714 |
+
"learning_rate": 3.2523860021208905e-05,
|
| 715 |
+
"loss": 0.5179,
|
| 716 |
"step": 4950
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 1.7674089784376106,
|
| 720 |
+
"grad_norm": 4.063836574554443,
|
| 721 |
+
"learning_rate": 3.234711912336515e-05,
|
| 722 |
+
"loss": 0.5588,
|
| 723 |
"step": 5000
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 1.7850830682219865,
|
| 727 |
+
"grad_norm": 5.352965354919434,
|
| 728 |
+
"learning_rate": 3.217037822552138e-05,
|
| 729 |
+
"loss": 0.5762,
|
| 730 |
"step": 5050
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 1.8027571580063628,
|
| 734 |
+
"grad_norm": 4.00150203704834,
|
| 735 |
+
"learning_rate": 3.199363732767763e-05,
|
| 736 |
+
"loss": 0.5546,
|
| 737 |
"step": 5100
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 1.8204312477907387,
|
| 741 |
+
"grad_norm": 4.444275856018066,
|
| 742 |
+
"learning_rate": 3.181689642983386e-05,
|
| 743 |
+
"loss": 0.5321,
|
| 744 |
"step": 5150
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 1.8381053375751149,
|
| 748 |
+
"grad_norm": 2.783729076385498,
|
| 749 |
+
"learning_rate": 3.164015553199011e-05,
|
| 750 |
+
"loss": 0.5859,
|
| 751 |
"step": 5200
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 1.855779427359491,
|
| 755 |
+
"grad_norm": 4.168649673461914,
|
| 756 |
+
"learning_rate": 3.146341463414634e-05,
|
| 757 |
+
"loss": 0.5353,
|
| 758 |
"step": 5250
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 1.873453517143867,
|
| 762 |
+
"grad_norm": 4.757116794586182,
|
| 763 |
+
"learning_rate": 3.128667373630258e-05,
|
| 764 |
+
"loss": 0.4766,
|
| 765 |
"step": 5300
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 1.8911276069282432,
|
| 769 |
+
"grad_norm": 2.5472869873046875,
|
| 770 |
+
"learning_rate": 3.1109932838458825e-05,
|
| 771 |
+
"loss": 0.4687,
|
| 772 |
"step": 5350
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 1.9088016967126193,
|
| 776 |
+
"grad_norm": 2.9616148471832275,
|
| 777 |
+
"learning_rate": 3.093319194061506e-05,
|
| 778 |
+
"loss": 0.5018,
|
| 779 |
"step": 5400
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 1.9264757864969955,
|
| 783 |
+
"grad_norm": 3.773808240890503,
|
| 784 |
+
"learning_rate": 3.0756451042771303e-05,
|
| 785 |
+
"loss": 0.5798,
|
| 786 |
"step": 5450
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 1.9441498762813714,
|
| 790 |
+
"grad_norm": 3.725301742553711,
|
| 791 |
+
"learning_rate": 3.0579710144927536e-05,
|
| 792 |
+
"loss": 0.5041,
|
| 793 |
"step": 5500
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 1.9618239660657477,
|
| 797 |
+
"grad_norm": 3.422393560409546,
|
| 798 |
+
"learning_rate": 3.040296924708378e-05,
|
| 799 |
+
"loss": 0.5053,
|
| 800 |
"step": 5550
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 1.9794980558501236,
|
| 804 |
+
"grad_norm": 3.7830734252929688,
|
| 805 |
+
"learning_rate": 3.0226228349240014e-05,
|
| 806 |
+
"loss": 0.5573,
|
| 807 |
"step": 5600
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 1.9971721456345,
|
| 811 |
+
"grad_norm": 2.845203161239624,
|
| 812 |
+
"learning_rate": 3.0049487451396253e-05,
|
| 813 |
+
"loss": 0.4995,
|
| 814 |
"step": 5650
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 2.0,
|
| 818 |
+
"eval_bertscore_f1": 0.9619662560301252,
|
| 819 |
+
"eval_bleu": 0.5490499087973257,
|
| 820 |
+
"eval_loss": 0.3714849352836609,
|
| 821 |
+
"eval_meteor": 0.7078770382751671,
|
| 822 |
+
"eval_rouge1": 0.8227979006513153,
|
| 823 |
+
"eval_rouge2": 0.7467554150541201,
|
| 824 |
+
"eval_runtime": 1347.4627,
|
| 825 |
+
"eval_samples_per_second": 4.797,
|
| 826 |
+
"eval_steps_per_second": 0.6,
|
| 827 |
"step": 5658
|
| 828 |
},
|
| 829 |
{
|
| 830 |
"epoch": 2.014846235418876,
|
| 831 |
+
"grad_norm": 3.069786787033081,
|
| 832 |
+
"learning_rate": 2.9872746553552493e-05,
|
| 833 |
+
"loss": 0.4827,
|
| 834 |
"step": 5700
|
| 835 |
},
|
| 836 |
{
|
| 837 |
"epoch": 2.032520325203252,
|
| 838 |
+
"grad_norm": 3.7215096950531006,
|
| 839 |
+
"learning_rate": 2.9696005655708732e-05,
|
| 840 |
+
"loss": 0.5742,
|
| 841 |
"step": 5750
|
| 842 |
},
|
| 843 |
{
|
| 844 |
"epoch": 2.050194414987628,
|
| 845 |
+
"grad_norm": 4.543232440948486,
|
| 846 |
+
"learning_rate": 2.9519264757864974e-05,
|
| 847 |
+
"loss": 0.478,
|
| 848 |
"step": 5800
|
| 849 |
},
|
| 850 |
{
|
| 851 |
"epoch": 2.0678685047720045,
|
| 852 |
+
"grad_norm": 3.076716423034668,
|
| 853 |
+
"learning_rate": 2.934252386002121e-05,
|
| 854 |
+
"loss": 0.4619,
|
| 855 |
"step": 5850
|
| 856 |
},
|
| 857 |
{
|
| 858 |
"epoch": 2.0855425945563804,
|
| 859 |
+
"grad_norm": 3.951244592666626,
|
| 860 |
+
"learning_rate": 2.9165782962177453e-05,
|
| 861 |
+
"loss": 0.513,
|
| 862 |
"step": 5900
|
| 863 |
},
|
| 864 |
{
|
| 865 |
"epoch": 2.1032166843407563,
|
| 866 |
+
"grad_norm": 2.552813768386841,
|
| 867 |
+
"learning_rate": 2.898904206433369e-05,
|
| 868 |
+
"loss": 0.491,
|
| 869 |
"step": 5950
|
| 870 |
},
|
| 871 |
{
|
| 872 |
"epoch": 2.1208907741251326,
|
| 873 |
+
"grad_norm": 2.2147703170776367,
|
| 874 |
+
"learning_rate": 2.8812301166489924e-05,
|
| 875 |
+
"loss": 0.4333,
|
| 876 |
"step": 6000
|
| 877 |
},
|
| 878 |
{
|
| 879 |
"epoch": 2.1385648639095085,
|
| 880 |
+
"grad_norm": 2.9662623405456543,
|
| 881 |
+
"learning_rate": 2.8635560268646167e-05,
|
| 882 |
+
"loss": 0.4287,
|
| 883 |
"step": 6050
|
| 884 |
},
|
| 885 |
{
|
| 886 |
"epoch": 2.156238953693885,
|
| 887 |
+
"grad_norm": 1.9441404342651367,
|
| 888 |
+
"learning_rate": 2.8458819370802403e-05,
|
| 889 |
+
"loss": 0.5373,
|
| 890 |
"step": 6100
|
| 891 |
},
|
| 892 |
{
|
| 893 |
"epoch": 2.1739130434782608,
|
| 894 |
+
"grad_norm": 4.967250823974609,
|
| 895 |
+
"learning_rate": 2.8282078472958645e-05,
|
| 896 |
+
"loss": 0.4971,
|
| 897 |
"step": 6150
|
| 898 |
},
|
| 899 |
{
|
| 900 |
"epoch": 2.191587133262637,
|
| 901 |
+
"grad_norm": 3.1946051120758057,
|
| 902 |
+
"learning_rate": 2.810533757511488e-05,
|
| 903 |
+
"loss": 0.5078,
|
| 904 |
"step": 6200
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 2.209261223047013,
|
| 908 |
+
"grad_norm": 2.941650152206421,
|
| 909 |
+
"learning_rate": 2.7928596677271124e-05,
|
| 910 |
+
"loss": 0.5376,
|
| 911 |
"step": 6250
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 2.2269353128313893,
|
| 915 |
+
"grad_norm": 4.430084705352783,
|
| 916 |
+
"learning_rate": 2.775185577942736e-05,
|
| 917 |
+
"loss": 0.5443,
|
| 918 |
"step": 6300
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 2.2446094026157652,
|
| 922 |
+
"grad_norm": 4.810621738433838,
|
| 923 |
+
"learning_rate": 2.7575114881583602e-05,
|
| 924 |
+
"loss": 0.5101,
|
| 925 |
"step": 6350
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 2.262283492400141,
|
| 929 |
+
"grad_norm": 4.557118892669678,
|
| 930 |
+
"learning_rate": 2.7401908801696713e-05,
|
| 931 |
+
"loss": 0.5837,
|
| 932 |
"step": 6400
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 2.2799575821845175,
|
| 936 |
+
"grad_norm": 3.7677035331726074,
|
| 937 |
+
"learning_rate": 2.7225167903852955e-05,
|
| 938 |
+
"loss": 0.4693,
|
| 939 |
"step": 6450
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 2.297631671968894,
|
| 943 |
+
"grad_norm": 2.5755605697631836,
|
| 944 |
+
"learning_rate": 2.704842700600919e-05,
|
| 945 |
+
"loss": 0.4601,
|
| 946 |
"step": 6500
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 2.3153057617532697,
|
| 950 |
+
"grad_norm": 4.114721775054932,
|
| 951 |
+
"learning_rate": 2.6871686108165427e-05,
|
| 952 |
+
"loss": 0.5209,
|
| 953 |
"step": 6550
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 2.3329798515376456,
|
| 957 |
+
"grad_norm": 4.422333717346191,
|
| 958 |
+
"learning_rate": 2.669494521032167e-05,
|
| 959 |
+
"loss": 0.5097,
|
| 960 |
"step": 6600
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 2.350653941322022,
|
| 964 |
+
"grad_norm": 2.2840325832366943,
|
| 965 |
+
"learning_rate": 2.6518204312477905e-05,
|
| 966 |
+
"loss": 0.4896,
|
| 967 |
"step": 6650
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 2.368328031106398,
|
| 971 |
+
"grad_norm": 4.739809036254883,
|
| 972 |
+
"learning_rate": 2.6341463414634148e-05,
|
| 973 |
+
"loss": 0.5377,
|
| 974 |
"step": 6700
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 2.3860021208907742,
|
| 978 |
+
"grad_norm": 4.174150466918945,
|
| 979 |
+
"learning_rate": 2.6164722516790384e-05,
|
| 980 |
+
"loss": 0.4754,
|
| 981 |
"step": 6750
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 2.40367621067515,
|
| 985 |
+
"grad_norm": 3.5064797401428223,
|
| 986 |
+
"learning_rate": 2.5987981618946626e-05,
|
| 987 |
+
"loss": 0.4375,
|
| 988 |
"step": 6800
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 2.4213503004595265,
|
| 992 |
+
"grad_norm": 5.094990253448486,
|
| 993 |
+
"learning_rate": 2.5811240721102865e-05,
|
| 994 |
+
"loss": 0.5027,
|
| 995 |
"step": 6850
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 2.4390243902439024,
|
| 999 |
+
"grad_norm": 6.338164329528809,
|
| 1000 |
+
"learning_rate": 2.5634499823259105e-05,
|
| 1001 |
+
"loss": 0.5016,
|
| 1002 |
"step": 6900
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 2.4566984800282787,
|
| 1006 |
+
"grad_norm": 3.988973379135132,
|
| 1007 |
+
"learning_rate": 2.5457758925415344e-05,
|
| 1008 |
+
"loss": 0.4173,
|
| 1009 |
"step": 6950
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 2.4743725698126546,
|
| 1013 |
+
"grad_norm": 2.3877015113830566,
|
| 1014 |
+
"learning_rate": 2.528101802757158e-05,
|
| 1015 |
+
"loss": 0.4857,
|
| 1016 |
"step": 7000
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 2.4920466595970305,
|
| 1020 |
+
"grad_norm": 4.041755199432373,
|
| 1021 |
+
"learning_rate": 2.5104277129727822e-05,
|
| 1022 |
+
"loss": 0.4463,
|
| 1023 |
"step": 7050
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 2.509720749381407,
|
| 1027 |
+
"grad_norm": 3.5311896800994873,
|
| 1028 |
+
"learning_rate": 2.492753623188406e-05,
|
| 1029 |
+
"loss": 0.4899,
|
| 1030 |
"step": 7100
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 2.5273948391657832,
|
| 1034 |
+
"grad_norm": 2.9291367530822754,
|
| 1035 |
+
"learning_rate": 2.4750795334040297e-05,
|
| 1036 |
+
"loss": 0.5041,
|
| 1037 |
"step": 7150
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 2.545068928950159,
|
| 1041 |
+
"grad_norm": 3.6040406227111816,
|
| 1042 |
+
"learning_rate": 2.4574054436196536e-05,
|
| 1043 |
+
"loss": 0.4374,
|
| 1044 |
"step": 7200
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 2.562743018734535,
|
| 1048 |
+
"grad_norm": 4.614346981048584,
|
| 1049 |
+
"learning_rate": 2.4397313538352776e-05,
|
| 1050 |
+
"loss": 0.4659,
|
| 1051 |
"step": 7250
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
"epoch": 2.5804171085189114,
|
| 1055 |
+
"grad_norm": 5.035871505737305,
|
| 1056 |
+
"learning_rate": 2.4220572640509015e-05,
|
| 1057 |
+
"loss": 0.4419,
|
| 1058 |
"step": 7300
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
"epoch": 2.5980911983032873,
|
| 1062 |
+
"grad_norm": 3.1209981441497803,
|
| 1063 |
+
"learning_rate": 2.4043831742665254e-05,
|
| 1064 |
+
"loss": 0.4863,
|
| 1065 |
"step": 7350
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 2.6157652880876636,
|
| 1069 |
+
"grad_norm": 2.7074010372161865,
|
| 1070 |
+
"learning_rate": 2.3867090844821493e-05,
|
| 1071 |
+
"loss": 0.4899,
|
| 1072 |
"step": 7400
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
"epoch": 2.6334393778720395,
|
| 1076 |
+
"grad_norm": 4.56402587890625,
|
| 1077 |
+
"learning_rate": 2.3690349946977732e-05,
|
| 1078 |
+
"loss": 0.5104,
|
| 1079 |
"step": 7450
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
"epoch": 2.6511134676564154,
|
| 1083 |
+
"grad_norm": 3.637251377105713,
|
| 1084 |
+
"learning_rate": 2.3513609049133968e-05,
|
| 1085 |
+
"loss": 0.4638,
|
| 1086 |
"step": 7500
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 2.6687875574407918,
|
| 1090 |
+
"grad_norm": 2.912982702255249,
|
| 1091 |
+
"learning_rate": 2.3336868151290207e-05,
|
| 1092 |
+
"loss": 0.4572,
|
| 1093 |
"step": 7550
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
"epoch": 2.686461647225168,
|
| 1097 |
+
"grad_norm": 2.9806952476501465,
|
| 1098 |
+
"learning_rate": 2.3160127253446447e-05,
|
| 1099 |
+
"loss": 0.4544,
|
| 1100 |
"step": 7600
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
"epoch": 2.704135737009544,
|
| 1104 |
+
"grad_norm": 3.6673879623413086,
|
| 1105 |
+
"learning_rate": 2.2983386355602686e-05,
|
| 1106 |
+
"loss": 0.4428,
|
| 1107 |
"step": 7650
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
"epoch": 2.72180982679392,
|
| 1111 |
+
"grad_norm": 7.318435192108154,
|
| 1112 |
+
"learning_rate": 2.280664545775893e-05,
|
| 1113 |
+
"loss": 0.4137,
|
| 1114 |
"step": 7700
|
| 1115 |
},
|
| 1116 |
{
|
| 1117 |
"epoch": 2.7394839165782963,
|
| 1118 |
+
"grad_norm": 7.58805513381958,
|
| 1119 |
+
"learning_rate": 2.2629904559915168e-05,
|
| 1120 |
+
"loss": 0.4501,
|
| 1121 |
"step": 7750
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
"epoch": 2.757158006362672,
|
| 1125 |
+
"grad_norm": 3.000596046447754,
|
| 1126 |
+
"learning_rate": 2.2453163662071407e-05,
|
| 1127 |
+
"loss": 0.4104,
|
| 1128 |
"step": 7800
|
| 1129 |
},
|
| 1130 |
{
|
| 1131 |
"epoch": 2.7748320961470485,
|
| 1132 |
+
"grad_norm": 6.188124656677246,
|
| 1133 |
+
"learning_rate": 2.2276422764227646e-05,
|
| 1134 |
+
"loss": 0.4833,
|
| 1135 |
"step": 7850
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
"epoch": 2.7925061859314244,
|
| 1139 |
+
"grad_norm": 2.526710033416748,
|
| 1140 |
+
"learning_rate": 2.209968186638388e-05,
|
| 1141 |
+
"loss": 0.4377,
|
| 1142 |
"step": 7900
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
"epoch": 2.8101802757158008,
|
| 1146 |
+
"grad_norm": 3.080709934234619,
|
| 1147 |
+
"learning_rate": 2.192294096854012e-05,
|
| 1148 |
+
"loss": 0.4289,
|
| 1149 |
"step": 7950
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
"epoch": 2.8278543655001767,
|
| 1153 |
+
"grad_norm": 2.859811544418335,
|
| 1154 |
+
"learning_rate": 2.174620007069636e-05,
|
| 1155 |
+
"loss": 0.4856,
|
| 1156 |
"step": 8000
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
"epoch": 2.845528455284553,
|
| 1160 |
+
"grad_norm": 4.68251371383667,
|
| 1161 |
+
"learning_rate": 2.15694591728526e-05,
|
| 1162 |
+
"loss": 0.4901,
|
| 1163 |
"step": 8050
|
| 1164 |
},
|
| 1165 |
{
|
| 1166 |
"epoch": 2.863202545068929,
|
| 1167 |
+
"grad_norm": 4.8537211418151855,
|
| 1168 |
+
"learning_rate": 2.139271827500884e-05,
|
| 1169 |
+
"loss": 0.3815,
|
| 1170 |
"step": 8100
|
| 1171 |
},
|
| 1172 |
{
|
| 1173 |
"epoch": 2.880876634853305,
|
| 1174 |
+
"grad_norm": 3.1829328536987305,
|
| 1175 |
+
"learning_rate": 2.1215977377165078e-05,
|
| 1176 |
+
"loss": 0.4713,
|
| 1177 |
"step": 8150
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"epoch": 2.898550724637681,
|
| 1181 |
+
"grad_norm": 4.081786155700684,
|
| 1182 |
+
"learning_rate": 2.1039236479321317e-05,
|
| 1183 |
+
"loss": 0.4494,
|
| 1184 |
"step": 8200
|
| 1185 |
},
|
| 1186 |
{
|
| 1187 |
"epoch": 2.9162248144220575,
|
| 1188 |
+
"grad_norm": 4.547771453857422,
|
| 1189 |
+
"learning_rate": 2.0862495581477553e-05,
|
| 1190 |
+
"loss": 0.4306,
|
| 1191 |
"step": 8250
|
| 1192 |
},
|
| 1193 |
{
|
| 1194 |
"epoch": 2.9338989042064334,
|
| 1195 |
+
"grad_norm": 2.5716054439544678,
|
| 1196 |
+
"learning_rate": 2.0685754683633792e-05,
|
| 1197 |
+
"loss": 0.3865,
|
| 1198 |
"step": 8300
|
| 1199 |
},
|
| 1200 |
{
|
| 1201 |
"epoch": 2.9515729939908093,
|
| 1202 |
+
"grad_norm": 3.1603822708129883,
|
| 1203 |
+
"learning_rate": 2.050901378579003e-05,
|
| 1204 |
+
"loss": 0.4292,
|
| 1205 |
"step": 8350
|
| 1206 |
},
|
| 1207 |
{
|
| 1208 |
"epoch": 2.9692470837751856,
|
| 1209 |
+
"grad_norm": 3.2110049724578857,
|
| 1210 |
+
"learning_rate": 2.033227288794627e-05,
|
| 1211 |
+
"loss": 0.4612,
|
| 1212 |
"step": 8400
|
| 1213 |
},
|
| 1214 |
{
|
| 1215 |
"epoch": 2.9869211735595615,
|
| 1216 |
+
"grad_norm": 3.1889193058013916,
|
| 1217 |
+
"learning_rate": 2.015553199010251e-05,
|
| 1218 |
+
"loss": 0.4432,
|
| 1219 |
"step": 8450
|
| 1220 |
},
|
| 1221 |
{
|
| 1222 |
"epoch": 3.0,
|
| 1223 |
+
"eval_bertscore_f1": 0.9641102957124315,
|
| 1224 |
+
"eval_bleu": 0.5720155666086876,
|
| 1225 |
+
"eval_loss": 0.32330864667892456,
|
| 1226 |
+
"eval_meteor": 0.7255114484352385,
|
| 1227 |
+
"eval_rouge1": 0.8322775652472418,
|
| 1228 |
+
"eval_rouge2": 0.7664804459108749,
|
| 1229 |
+
"eval_runtime": 1341.3557,
|
| 1230 |
+
"eval_samples_per_second": 4.819,
|
| 1231 |
+
"eval_steps_per_second": 0.602,
|
| 1232 |
"step": 8487
|
| 1233 |
},
|
| 1234 |
{
|
| 1235 |
"epoch": 3.004595263343938,
|
| 1236 |
+
"grad_norm": 3.6943013668060303,
|
| 1237 |
+
"learning_rate": 1.997879109225875e-05,
|
| 1238 |
+
"loss": 0.4008,
|
| 1239 |
"step": 8500
|
| 1240 |
},
|
| 1241 |
{
|
| 1242 |
"epoch": 3.022269353128314,
|
| 1243 |
+
"grad_norm": 3.6444859504699707,
|
| 1244 |
+
"learning_rate": 1.9802050194414988e-05,
|
| 1245 |
+
"loss": 0.4435,
|
| 1246 |
"step": 8550
|
| 1247 |
},
|
| 1248 |
{
|
| 1249 |
"epoch": 3.03994344291269,
|
| 1250 |
+
"grad_norm": 2.0229413509368896,
|
| 1251 |
+
"learning_rate": 1.962530929657123e-05,
|
| 1252 |
+
"loss": 0.4479,
|
| 1253 |
"step": 8600
|
| 1254 |
},
|
| 1255 |
{
|
| 1256 |
"epoch": 3.057617532697066,
|
| 1257 |
+
"grad_norm": 2.651965618133545,
|
| 1258 |
+
"learning_rate": 1.9448568398727466e-05,
|
| 1259 |
+
"loss": 0.4386,
|
| 1260 |
"step": 8650
|
| 1261 |
},
|
| 1262 |
{
|
| 1263 |
"epoch": 3.0752916224814424,
|
| 1264 |
+
"grad_norm": 3.405768394470215,
|
| 1265 |
+
"learning_rate": 1.9271827500883705e-05,
|
| 1266 |
+
"loss": 0.3994,
|
| 1267 |
"step": 8700
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
"epoch": 3.0929657122658183,
|
| 1271 |
+
"grad_norm": 4.544278144836426,
|
| 1272 |
+
"learning_rate": 1.9095086603039945e-05,
|
| 1273 |
+
"loss": 0.4002,
|
| 1274 |
"step": 8750
|
| 1275 |
},
|
| 1276 |
{
|
| 1277 |
"epoch": 3.110639802050194,
|
| 1278 |
+
"grad_norm": 4.592613220214844,
|
| 1279 |
+
"learning_rate": 1.8918345705196184e-05,
|
| 1280 |
+
"loss": 0.4925,
|
| 1281 |
"step": 8800
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
"epoch": 3.1283138918345705,
|
| 1285 |
+
"grad_norm": 4.989655017852783,
|
| 1286 |
+
"learning_rate": 1.8741604807352423e-05,
|
| 1287 |
+
"loss": 0.411,
|
| 1288 |
"step": 8850
|
| 1289 |
},
|
| 1290 |
{
|
| 1291 |
"epoch": 3.1459879816189464,
|
| 1292 |
+
"grad_norm": 5.274810791015625,
|
| 1293 |
+
"learning_rate": 1.8568398727465537e-05,
|
| 1294 |
+
"loss": 0.4002,
|
| 1295 |
"step": 8900
|
| 1296 |
},
|
| 1297 |
{
|
| 1298 |
"epoch": 3.163662071403323,
|
| 1299 |
+
"grad_norm": 3.1220662593841553,
|
| 1300 |
+
"learning_rate": 1.8391657829621776e-05,
|
| 1301 |
+
"loss": 0.4138,
|
| 1302 |
"step": 8950
|
| 1303 |
},
|
| 1304 |
{
|
| 1305 |
"epoch": 3.1813361611876987,
|
| 1306 |
+
"grad_norm": 3.1085612773895264,
|
| 1307 |
+
"learning_rate": 1.8214916931778015e-05,
|
| 1308 |
+
"loss": 0.4498,
|
| 1309 |
"step": 9000
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
"epoch": 3.199010250972075,
|
| 1313 |
+
"grad_norm": 2.929586410522461,
|
| 1314 |
+
"learning_rate": 1.8038176033934255e-05,
|
| 1315 |
+
"loss": 0.4271,
|
| 1316 |
"step": 9050
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
"epoch": 3.216684340756451,
|
| 1320 |
+
"grad_norm": 3.6650078296661377,
|
| 1321 |
+
"learning_rate": 1.7861435136090494e-05,
|
| 1322 |
+
"loss": 0.4642,
|
| 1323 |
"step": 9100
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
"epoch": 3.2343584305408273,
|
| 1327 |
+
"grad_norm": 4.111539363861084,
|
| 1328 |
+
"learning_rate": 1.7684694238246733e-05,
|
| 1329 |
+
"loss": 0.3974,
|
| 1330 |
"step": 9150
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 3.252032520325203,
|
| 1334 |
+
"grad_norm": 3.7882914543151855,
|
| 1335 |
+
"learning_rate": 1.750795334040297e-05,
|
| 1336 |
+
"loss": 0.4961,
|
| 1337 |
"step": 9200
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"epoch": 3.2697066101095795,
|
| 1341 |
+
"grad_norm": 3.846184253692627,
|
| 1342 |
+
"learning_rate": 1.7331212442559208e-05,
|
| 1343 |
+
"loss": 0.4681,
|
| 1344 |
"step": 9250
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
"epoch": 3.2873806998939554,
|
| 1348 |
+
"grad_norm": 2.3030924797058105,
|
| 1349 |
+
"learning_rate": 1.7154471544715447e-05,
|
| 1350 |
+
"loss": 0.3746,
|
| 1351 |
"step": 9300
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
"epoch": 3.3050547896783318,
|
| 1355 |
+
"grad_norm": 5.14872407913208,
|
| 1356 |
+
"learning_rate": 1.6977730646871686e-05,
|
| 1357 |
+
"loss": 0.3876,
|
| 1358 |
"step": 9350
|
| 1359 |
},
|
| 1360 |
{
|
| 1361 |
"epoch": 3.3227288794627077,
|
| 1362 |
+
"grad_norm": 3.8709867000579834,
|
| 1363 |
+
"learning_rate": 1.6800989749027926e-05,
|
| 1364 |
+
"loss": 0.4326,
|
| 1365 |
"step": 9400
|
| 1366 |
},
|
| 1367 |
{
|
| 1368 |
"epoch": 3.3404029692470836,
|
| 1369 |
+
"grad_norm": 2.4771230220794678,
|
| 1370 |
+
"learning_rate": 1.6624248851184165e-05,
|
| 1371 |
+
"loss": 0.4111,
|
| 1372 |
"step": 9450
|
| 1373 |
},
|
| 1374 |
{
|
| 1375 |
"epoch": 3.35807705903146,
|
| 1376 |
+
"grad_norm": 4.154597282409668,
|
| 1377 |
+
"learning_rate": 1.6447507953340404e-05,
|
| 1378 |
+
"loss": 0.514,
|
| 1379 |
"step": 9500
|
| 1380 |
},
|
| 1381 |
{
|
| 1382 |
"epoch": 3.375751148815836,
|
| 1383 |
+
"grad_norm": 3.861116409301758,
|
| 1384 |
+
"learning_rate": 1.6270767055496643e-05,
|
| 1385 |
+
"loss": 0.4171,
|
| 1386 |
"step": 9550
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
"epoch": 3.393425238600212,
|
| 1390 |
+
"grad_norm": 3.292591094970703,
|
| 1391 |
+
"learning_rate": 1.609402615765288e-05,
|
| 1392 |
+
"loss": 0.4192,
|
| 1393 |
"step": 9600
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
"epoch": 3.411099328384588,
|
| 1397 |
+
"grad_norm": 3.9203121662139893,
|
| 1398 |
+
"learning_rate": 1.591728525980912e-05,
|
| 1399 |
+
"loss": 0.4036,
|
| 1400 |
"step": 9650
|
| 1401 |
},
|
| 1402 |
{
|
| 1403 |
"epoch": 3.4287734181689644,
|
| 1404 |
+
"grad_norm": 3.337324857711792,
|
| 1405 |
+
"learning_rate": 1.574054436196536e-05,
|
| 1406 |
+
"loss": 0.3511,
|
| 1407 |
"step": 9700
|
| 1408 |
},
|
| 1409 |
{
|
| 1410 |
"epoch": 3.4464475079533403,
|
| 1411 |
+
"grad_norm": 2.305972099304199,
|
| 1412 |
+
"learning_rate": 1.55638034641216e-05,
|
| 1413 |
+
"loss": 0.4891,
|
| 1414 |
"step": 9750
|
| 1415 |
},
|
| 1416 |
{
|
| 1417 |
"epoch": 3.4641215977377167,
|
| 1418 |
+
"grad_norm": 4.972841739654541,
|
| 1419 |
+
"learning_rate": 1.538706256627784e-05,
|
| 1420 |
+
"loss": 0.4172,
|
| 1421 |
"step": 9800
|
| 1422 |
},
|
| 1423 |
{
|
| 1424 |
"epoch": 3.4817956875220926,
|
| 1425 |
+
"grad_norm": 3.111032009124756,
|
| 1426 |
+
"learning_rate": 1.5210321668434077e-05,
|
| 1427 |
+
"loss": 0.4401,
|
| 1428 |
"step": 9850
|
| 1429 |
},
|
| 1430 |
{
|
| 1431 |
"epoch": 3.499469777306469,
|
| 1432 |
+
"grad_norm": 1.9676620960235596,
|
| 1433 |
+
"learning_rate": 1.5033580770590316e-05,
|
| 1434 |
+
"loss": 0.4398,
|
| 1435 |
"step": 9900
|
| 1436 |
},
|
| 1437 |
{
|
| 1438 |
"epoch": 3.517143867090845,
|
| 1439 |
+
"grad_norm": 3.4216668605804443,
|
| 1440 |
+
"learning_rate": 1.4856839872746553e-05,
|
| 1441 |
+
"loss": 0.4015,
|
| 1442 |
"step": 9950
|
| 1443 |
},
|
| 1444 |
{
|
| 1445 |
"epoch": 3.534817956875221,
|
| 1446 |
+
"grad_norm": 3.161693811416626,
|
| 1447 |
+
"learning_rate": 1.4680098974902792e-05,
|
| 1448 |
+
"loss": 0.4216,
|
| 1449 |
"step": 10000
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 3.552492046659597,
|
| 1453 |
+
"grad_norm": 5.103592395782471,
|
| 1454 |
+
"learning_rate": 1.4503358077059032e-05,
|
| 1455 |
+
"loss": 0.4489,
|
| 1456 |
"step": 10050
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 3.570166136443973,
|
| 1460 |
+
"grad_norm": 3.90478777885437,
|
| 1461 |
+
"learning_rate": 1.432661717921527e-05,
|
| 1462 |
+
"loss": 0.4565,
|
| 1463 |
"step": 10100
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 3.5878402262283493,
|
| 1467 |
+
"grad_norm": 2.7845191955566406,
|
| 1468 |
+
"learning_rate": 1.414987628137151e-05,
|
| 1469 |
+
"loss": 0.4312,
|
| 1470 |
"step": 10150
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 3.605514316012725,
|
| 1474 |
+
"grad_norm": 4.3978729248046875,
|
| 1475 |
+
"learning_rate": 1.397313538352775e-05,
|
| 1476 |
+
"loss": 0.5477,
|
| 1477 |
"step": 10200
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 3.6231884057971016,
|
| 1481 |
+
"grad_norm": 3.3587982654571533,
|
| 1482 |
+
"learning_rate": 1.3796394485683988e-05,
|
| 1483 |
+
"loss": 0.4373,
|
| 1484 |
"step": 10250
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 3.6408624955814775,
|
| 1488 |
+
"grad_norm": 2.463456392288208,
|
| 1489 |
+
"learning_rate": 1.3619653587840228e-05,
|
| 1490 |
+
"loss": 0.4234,
|
| 1491 |
"step": 10300
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 3.658536585365854,
|
| 1495 |
+
"grad_norm": 3.532365560531616,
|
| 1496 |
+
"learning_rate": 1.3442912689996465e-05,
|
| 1497 |
+
"loss": 0.4115,
|
| 1498 |
"step": 10350
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 3.6762106751502297,
|
| 1502 |
+
"grad_norm": 3.4025349617004395,
|
| 1503 |
+
"learning_rate": 1.3266171792152704e-05,
|
| 1504 |
+
"loss": 0.4462,
|
| 1505 |
"step": 10400
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 3.693884764934606,
|
| 1509 |
+
"grad_norm": 1.0425785779953003,
|
| 1510 |
+
"learning_rate": 1.3089430894308943e-05,
|
| 1511 |
+
"loss": 0.3494,
|
| 1512 |
"step": 10450
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 3.711558854718982,
|
| 1516 |
+
"grad_norm": 4.738943099975586,
|
| 1517 |
+
"learning_rate": 1.2912689996465183e-05,
|
| 1518 |
+
"loss": 0.4316,
|
| 1519 |
"step": 10500
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 3.729232944503358,
|
| 1523 |
+
"grad_norm": 2.0041253566741943,
|
| 1524 |
+
"learning_rate": 1.2735949098621422e-05,
|
| 1525 |
+
"loss": 0.4596,
|
| 1526 |
"step": 10550
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 3.746907034287734,
|
| 1530 |
+
"grad_norm": 4.681216239929199,
|
| 1531 |
+
"learning_rate": 1.2559208200777661e-05,
|
| 1532 |
+
"loss": 0.4013,
|
| 1533 |
"step": 10600
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 3.7645811240721105,
|
| 1537 |
+
"grad_norm": 4.387250900268555,
|
| 1538 |
+
"learning_rate": 1.2382467302933899e-05,
|
| 1539 |
+
"loss": 0.387,
|
| 1540 |
"step": 10650
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 3.7822552138564864,
|
| 1544 |
+
"grad_norm": 2.4324512481689453,
|
| 1545 |
+
"learning_rate": 1.2205726405090138e-05,
|
| 1546 |
+
"loss": 0.4383,
|
| 1547 |
"step": 10700
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 3.7999293036408623,
|
| 1551 |
+
"grad_norm": 4.895308494567871,
|
| 1552 |
+
"learning_rate": 1.2028985507246379e-05,
|
| 1553 |
+
"loss": 0.4567,
|
| 1554 |
"step": 10750
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 3.8176033934252387,
|
| 1558 |
+
"grad_norm": 3.3893020153045654,
|
| 1559 |
+
"learning_rate": 1.1852244609402616e-05,
|
| 1560 |
+
"loss": 0.4316,
|
| 1561 |
"step": 10800
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 3.8352774832096146,
|
| 1565 |
+
"grad_norm": 2.8720388412475586,
|
| 1566 |
+
"learning_rate": 1.1675503711558855e-05,
|
| 1567 |
+
"loss": 0.3712,
|
| 1568 |
"step": 10850
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 3.852951572993991,
|
| 1572 |
+
"grad_norm": 2.9342293739318848,
|
| 1573 |
+
"learning_rate": 1.1498762813715094e-05,
|
| 1574 |
+
"loss": 0.3595,
|
| 1575 |
"step": 10900
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 3.870625662778367,
|
| 1579 |
+
"grad_norm": 5.391896724700928,
|
| 1580 |
+
"learning_rate": 1.1322021915871334e-05,
|
| 1581 |
+
"loss": 0.3945,
|
| 1582 |
"step": 10950
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 3.888299752562743,
|
| 1586 |
+
"grad_norm": 2.1623027324676514,
|
| 1587 |
+
"learning_rate": 1.1145281018027571e-05,
|
| 1588 |
+
"loss": 0.3849,
|
| 1589 |
"step": 11000
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 3.905973842347119,
|
| 1593 |
+
"grad_norm": 2.6391725540161133,
|
| 1594 |
+
"learning_rate": 1.096854012018381e-05,
|
| 1595 |
+
"loss": 0.3837,
|
| 1596 |
"step": 11050
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 3.9236479321314954,
|
| 1600 |
+
"grad_norm": 2.8502111434936523,
|
| 1601 |
+
"learning_rate": 1.079179922234005e-05,
|
| 1602 |
+
"loss": 0.4109,
|
| 1603 |
"step": 11100
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 3.9413220219158713,
|
| 1607 |
+
"grad_norm": 3.7358663082122803,
|
| 1608 |
+
"learning_rate": 1.0615058324496289e-05,
|
| 1609 |
+
"loss": 0.4113,
|
| 1610 |
"step": 11150
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 3.9589961117002472,
|
| 1614 |
+
"grad_norm": 3.771918535232544,
|
| 1615 |
+
"learning_rate": 1.0438317426652528e-05,
|
| 1616 |
+
"loss": 0.3664,
|
| 1617 |
"step": 11200
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 3.9766702014846236,
|
| 1621 |
+
"grad_norm": 2.2865407466888428,
|
| 1622 |
+
"learning_rate": 1.0261576528808767e-05,
|
| 1623 |
+
"loss": 0.427,
|
| 1624 |
"step": 11250
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 3.9943442912689995,
|
| 1628 |
+
"grad_norm": 3.386460781097412,
|
| 1629 |
+
"learning_rate": 1.0084835630965006e-05,
|
| 1630 |
+
"loss": 0.4252,
|
| 1631 |
"step": 11300
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 4.0,
|
| 1635 |
+
"eval_bertscore_f1": 0.965401310132503,
|
| 1636 |
+
"eval_bleu": 0.5825957176218498,
|
| 1637 |
+
"eval_loss": 0.29947343468666077,
|
| 1638 |
+
"eval_meteor": 0.7356418884759407,
|
| 1639 |
+
"eval_rouge1": 0.8377531126193314,
|
| 1640 |
+
"eval_rouge2": 0.7761899021152611,
|
| 1641 |
+
"eval_runtime": 1380.5465,
|
| 1642 |
+
"eval_samples_per_second": 4.682,
|
| 1643 |
+
"eval_steps_per_second": 0.585,
|
| 1644 |
"step": 11316
|
| 1645 |
},
|
| 1646 |
{
|
| 1647 |
"epoch": 4.012018381053376,
|
| 1648 |
+
"grad_norm": 4.861545085906982,
|
| 1649 |
+
"learning_rate": 9.908094733121245e-06,
|
| 1650 |
+
"loss": 0.4298,
|
| 1651 |
"step": 11350
|
| 1652 |
},
|
| 1653 |
{
|
| 1654 |
"epoch": 4.029692470837752,
|
| 1655 |
+
"grad_norm": 6.035098075866699,
|
| 1656 |
+
"learning_rate": 9.731353835277483e-06,
|
| 1657 |
+
"loss": 0.3755,
|
| 1658 |
"step": 11400
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
"epoch": 4.047366560622128,
|
| 1662 |
+
"grad_norm": 2.295048952102661,
|
| 1663 |
+
"learning_rate": 9.554612937433722e-06,
|
| 1664 |
+
"loss": 0.3772,
|
| 1665 |
"step": 11450
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 4.065040650406504,
|
| 1669 |
+
"grad_norm": 3.09201979637146,
|
| 1670 |
+
"learning_rate": 9.377872039589961e-06,
|
| 1671 |
+
"loss": 0.4315,
|
| 1672 |
"step": 11500
|
| 1673 |
},
|
| 1674 |
{
|
| 1675 |
"epoch": 4.08271474019088,
|
| 1676 |
+
"grad_norm": 7.435740947723389,
|
| 1677 |
+
"learning_rate": 9.2011311417462e-06,
|
| 1678 |
+
"loss": 0.4276,
|
| 1679 |
"step": 11550
|
| 1680 |
},
|
| 1681 |
{
|
| 1682 |
"epoch": 4.100388829975256,
|
| 1683 |
+
"grad_norm": 2.680793046951294,
|
| 1684 |
+
"learning_rate": 9.02439024390244e-06,
|
| 1685 |
+
"loss": 0.3664,
|
| 1686 |
"step": 11600
|
| 1687 |
},
|
| 1688 |
{
|
| 1689 |
"epoch": 4.118062919759632,
|
| 1690 |
+
"grad_norm": 2.200497627258301,
|
| 1691 |
+
"learning_rate": 8.847649346058679e-06,
|
| 1692 |
+
"loss": 0.4054,
|
| 1693 |
"step": 11650
|
| 1694 |
},
|
| 1695 |
{
|
| 1696 |
"epoch": 4.135737009544009,
|
| 1697 |
+
"grad_norm": 3.864414930343628,
|
| 1698 |
+
"learning_rate": 8.670908448214918e-06,
|
| 1699 |
+
"loss": 0.4256,
|
| 1700 |
"step": 11700
|
| 1701 |
},
|
| 1702 |
{
|
| 1703 |
"epoch": 4.153411099328385,
|
| 1704 |
+
"grad_norm": 4.263733863830566,
|
| 1705 |
+
"learning_rate": 8.494167550371156e-06,
|
| 1706 |
+
"loss": 0.4066,
|
| 1707 |
"step": 11750
|
| 1708 |
},
|
| 1709 |
{
|
| 1710 |
"epoch": 4.171085189112761,
|
| 1711 |
+
"grad_norm": 4.246395587921143,
|
| 1712 |
+
"learning_rate": 8.317426652527395e-06,
|
| 1713 |
+
"loss": 0.3858,
|
| 1714 |
"step": 11800
|
| 1715 |
},
|
| 1716 |
{
|
| 1717 |
"epoch": 4.188759278897137,
|
| 1718 |
+
"grad_norm": 4.842310905456543,
|
| 1719 |
+
"learning_rate": 8.140685754683634e-06,
|
| 1720 |
+
"loss": 0.4026,
|
| 1721 |
"step": 11850
|
| 1722 |
},
|
| 1723 |
{
|
| 1724 |
"epoch": 4.2064333686815125,
|
| 1725 |
+
"grad_norm": 2.5259275436401367,
|
| 1726 |
+
"learning_rate": 7.963944856839873e-06,
|
| 1727 |
+
"loss": 0.3624,
|
| 1728 |
"step": 11900
|
| 1729 |
},
|
| 1730 |
{
|
| 1731 |
"epoch": 4.224107458465889,
|
| 1732 |
+
"grad_norm": 2.959528923034668,
|
| 1733 |
+
"learning_rate": 7.787203958996112e-06,
|
| 1734 |
+
"loss": 0.3971,
|
| 1735 |
"step": 11950
|
| 1736 |
},
|
| 1737 |
{
|
| 1738 |
"epoch": 4.241781548250265,
|
| 1739 |
+
"grad_norm": 2.9087865352630615,
|
| 1740 |
+
"learning_rate": 7.610463061152351e-06,
|
| 1741 |
+
"loss": 0.4095,
|
| 1742 |
"step": 12000
|
| 1743 |
},
|
| 1744 |
{
|
| 1745 |
"epoch": 4.259455638034641,
|
| 1746 |
+
"grad_norm": 3.8024725914001465,
|
| 1747 |
+
"learning_rate": 7.43372216330859e-06,
|
| 1748 |
+
"loss": 0.3815,
|
| 1749 |
"step": 12050
|
| 1750 |
},
|
| 1751 |
{
|
| 1752 |
"epoch": 4.277129727819017,
|
| 1753 |
+
"grad_norm": 3.2564175128936768,
|
| 1754 |
+
"learning_rate": 7.256981265464829e-06,
|
| 1755 |
+
"loss": 0.3752,
|
| 1756 |
"step": 12100
|
| 1757 |
},
|
| 1758 |
{
|
| 1759 |
"epoch": 4.294803817603394,
|
| 1760 |
+
"grad_norm": 1.157267689704895,
|
| 1761 |
+
"learning_rate": 7.080240367621067e-06,
|
| 1762 |
+
"loss": 0.3769,
|
| 1763 |
"step": 12150
|
| 1764 |
},
|
| 1765 |
{
|
| 1766 |
"epoch": 4.31247790738777,
|
| 1767 |
+
"grad_norm": 11.618633270263672,
|
| 1768 |
+
"learning_rate": 6.903499469777307e-06,
|
| 1769 |
+
"loss": 0.4042,
|
| 1770 |
"step": 12200
|
| 1771 |
},
|
| 1772 |
{
|
| 1773 |
"epoch": 4.330151997172146,
|
| 1774 |
+
"grad_norm": 4.113063812255859,
|
| 1775 |
+
"learning_rate": 6.726758571933546e-06,
|
| 1776 |
+
"loss": 0.4441,
|
| 1777 |
"step": 12250
|
| 1778 |
},
|
| 1779 |
{
|
| 1780 |
"epoch": 4.3478260869565215,
|
| 1781 |
+
"grad_norm": 4.125561237335205,
|
| 1782 |
+
"learning_rate": 6.550017674089785e-06,
|
| 1783 |
+
"loss": 0.421,
|
| 1784 |
"step": 12300
|
| 1785 |
},
|
| 1786 |
{
|
| 1787 |
"epoch": 4.365500176740898,
|
| 1788 |
+
"grad_norm": 3.2781832218170166,
|
| 1789 |
+
"learning_rate": 6.373276776246023e-06,
|
| 1790 |
+
"loss": 0.46,
|
| 1791 |
"step": 12350
|
| 1792 |
},
|
| 1793 |
{
|
| 1794 |
"epoch": 4.383174266525274,
|
| 1795 |
+
"grad_norm": 8.616064071655273,
|
| 1796 |
+
"learning_rate": 6.1965358784022625e-06,
|
| 1797 |
+
"loss": 0.3905,
|
| 1798 |
"step": 12400
|
| 1799 |
},
|
| 1800 |
{
|
| 1801 |
"epoch": 4.40084835630965,
|
| 1802 |
+
"grad_norm": 3.173698902130127,
|
| 1803 |
+
"learning_rate": 6.019794980558501e-06,
|
| 1804 |
+
"loss": 0.3529,
|
| 1805 |
"step": 12450
|
| 1806 |
},
|
| 1807 |
{
|
| 1808 |
"epoch": 4.418522446094026,
|
| 1809 |
+
"grad_norm": 2.7585160732269287,
|
| 1810 |
+
"learning_rate": 5.843054082714741e-06,
|
| 1811 |
+
"loss": 0.4236,
|
| 1812 |
"step": 12500
|
| 1813 |
},
|
| 1814 |
{
|
| 1815 |
"epoch": 4.436196535878402,
|
| 1816 |
+
"grad_norm": 1.5063729286193848,
|
| 1817 |
+
"learning_rate": 5.666313184870979e-06,
|
| 1818 |
+
"loss": 0.4182,
|
| 1819 |
"step": 12550
|
| 1820 |
},
|
| 1821 |
{
|
| 1822 |
"epoch": 4.453870625662779,
|
| 1823 |
+
"grad_norm": 5.755438327789307,
|
| 1824 |
+
"learning_rate": 5.4895722870272184e-06,
|
| 1825 |
+
"loss": 0.3909,
|
| 1826 |
"step": 12600
|
| 1827 |
},
|
| 1828 |
{
|
| 1829 |
"epoch": 4.471544715447155,
|
| 1830 |
+
"grad_norm": 2.465946674346924,
|
| 1831 |
+
"learning_rate": 5.312831389183457e-06,
|
| 1832 |
+
"loss": 0.4078,
|
| 1833 |
"step": 12650
|
| 1834 |
},
|
| 1835 |
{
|
| 1836 |
"epoch": 4.4892188052315305,
|
| 1837 |
+
"grad_norm": 2.642314910888672,
|
| 1838 |
+
"learning_rate": 5.136090491339696e-06,
|
| 1839 |
+
"loss": 0.4581,
|
| 1840 |
"step": 12700
|
| 1841 |
},
|
| 1842 |
{
|
| 1843 |
"epoch": 4.506892895015906,
|
| 1844 |
+
"grad_norm": 3.11537766456604,
|
| 1845 |
+
"learning_rate": 4.959349593495935e-06,
|
| 1846 |
+
"loss": 0.3574,
|
| 1847 |
"step": 12750
|
| 1848 |
},
|
| 1849 |
{
|
| 1850 |
"epoch": 4.524566984800282,
|
| 1851 |
+
"grad_norm": 5.104282855987549,
|
| 1852 |
+
"learning_rate": 4.782608695652174e-06,
|
| 1853 |
+
"loss": 0.3889,
|
| 1854 |
"step": 12800
|
| 1855 |
},
|
| 1856 |
{
|
| 1857 |
"epoch": 4.542241074584659,
|
| 1858 |
+
"grad_norm": 3.2097325325012207,
|
| 1859 |
+
"learning_rate": 4.605867797808413e-06,
|
| 1860 |
+
"loss": 0.3723,
|
| 1861 |
"step": 12850
|
| 1862 |
},
|
| 1863 |
{
|
| 1864 |
"epoch": 4.559915164369035,
|
| 1865 |
+
"grad_norm": 2.8303864002227783,
|
| 1866 |
+
"learning_rate": 4.429126899964652e-06,
|
| 1867 |
+
"loss": 0.4222,
|
| 1868 |
"step": 12900
|
| 1869 |
},
|
| 1870 |
{
|
| 1871 |
"epoch": 4.577589254153411,
|
| 1872 |
+
"grad_norm": 3.508904457092285,
|
| 1873 |
+
"learning_rate": 4.252386002120891e-06,
|
| 1874 |
+
"loss": 0.4109,
|
| 1875 |
"step": 12950
|
| 1876 |
},
|
| 1877 |
{
|
| 1878 |
"epoch": 4.595263343937788,
|
| 1879 |
+
"grad_norm": 3.8901443481445312,
|
| 1880 |
+
"learning_rate": 4.07564510427713e-06,
|
| 1881 |
+
"loss": 0.4248,
|
| 1882 |
"step": 13000
|
| 1883 |
},
|
| 1884 |
{
|
| 1885 |
"epoch": 4.612937433722164,
|
| 1886 |
+
"grad_norm": 4.3722920417785645,
|
| 1887 |
+
"learning_rate": 3.898904206433369e-06,
|
| 1888 |
+
"loss": 0.4139,
|
| 1889 |
"step": 13050
|
| 1890 |
},
|
| 1891 |
{
|
| 1892 |
"epoch": 4.6306115235065395,
|
| 1893 |
+
"grad_norm": 3.87107515335083,
|
| 1894 |
+
"learning_rate": 3.722163308589608e-06,
|
| 1895 |
+
"loss": 0.3862,
|
| 1896 |
"step": 13100
|
| 1897 |
},
|
| 1898 |
{
|
| 1899 |
"epoch": 4.648285613290915,
|
| 1900 |
+
"grad_norm": 4.208980560302734,
|
| 1901 |
+
"learning_rate": 3.5454224107458466e-06,
|
| 1902 |
+
"loss": 0.3971,
|
| 1903 |
"step": 13150
|
| 1904 |
},
|
| 1905 |
{
|
| 1906 |
"epoch": 4.665959703075291,
|
| 1907 |
+
"grad_norm": 3.0796680450439453,
|
| 1908 |
+
"learning_rate": 3.368681512902086e-06,
|
| 1909 |
+
"loss": 0.4466,
|
| 1910 |
"step": 13200
|
| 1911 |
},
|
| 1912 |
{
|
| 1913 |
"epoch": 4.683633792859668,
|
| 1914 |
+
"grad_norm": 3.2411413192749023,
|
| 1915 |
+
"learning_rate": 3.1919406150583245e-06,
|
| 1916 |
+
"loss": 0.3462,
|
| 1917 |
"step": 13250
|
| 1918 |
},
|
| 1919 |
{
|
| 1920 |
"epoch": 4.701307882644044,
|
| 1921 |
+
"grad_norm": 3.422546625137329,
|
| 1922 |
+
"learning_rate": 3.0151997172145637e-06,
|
| 1923 |
+
"loss": 0.4942,
|
| 1924 |
"step": 13300
|
| 1925 |
},
|
| 1926 |
{
|
| 1927 |
"epoch": 4.71898197242842,
|
| 1928 |
+
"grad_norm": 5.258462905883789,
|
| 1929 |
+
"learning_rate": 2.8384588193708025e-06,
|
| 1930 |
+
"loss": 0.3842,
|
| 1931 |
"step": 13350
|
| 1932 |
},
|
| 1933 |
{
|
| 1934 |
"epoch": 4.736656062212796,
|
| 1935 |
+
"grad_norm": 3.634772300720215,
|
| 1936 |
+
"learning_rate": 2.6617179215270417e-06,
|
| 1937 |
+
"loss": 0.386,
|
| 1938 |
"step": 13400
|
| 1939 |
},
|
| 1940 |
{
|
| 1941 |
"epoch": 4.754330151997172,
|
| 1942 |
+
"grad_norm": 2.493283987045288,
|
| 1943 |
+
"learning_rate": 2.4849770236832804e-06,
|
| 1944 |
+
"loss": 0.3646,
|
| 1945 |
"step": 13450
|
| 1946 |
},
|
| 1947 |
{
|
| 1948 |
"epoch": 4.7720042417815485,
|
| 1949 |
+
"grad_norm": 3.546058416366577,
|
| 1950 |
+
"learning_rate": 2.3082361258395196e-06,
|
| 1951 |
+
"loss": 0.4085,
|
| 1952 |
"step": 13500
|
| 1953 |
},
|
| 1954 |
{
|
| 1955 |
"epoch": 4.789678331565924,
|
| 1956 |
+
"grad_norm": 2.0962002277374268,
|
| 1957 |
+
"learning_rate": 2.1314952279957584e-06,
|
| 1958 |
+
"loss": 0.3869,
|
| 1959 |
"step": 13550
|
| 1960 |
},
|
| 1961 |
{
|
| 1962 |
"epoch": 4.8073524213503,
|
| 1963 |
+
"grad_norm": 2.2293384075164795,
|
| 1964 |
+
"learning_rate": 1.9547543301519976e-06,
|
| 1965 |
+
"loss": 0.4841,
|
| 1966 |
"step": 13600
|
| 1967 |
},
|
| 1968 |
{
|
| 1969 |
"epoch": 4.825026511134676,
|
| 1970 |
+
"grad_norm": 3.2926249504089355,
|
| 1971 |
+
"learning_rate": 1.7780134323082363e-06,
|
| 1972 |
+
"loss": 0.4599,
|
| 1973 |
"step": 13650
|
| 1974 |
},
|
| 1975 |
{
|
| 1976 |
"epoch": 4.842700600919053,
|
| 1977 |
+
"grad_norm": 5.047961235046387,
|
| 1978 |
+
"learning_rate": 1.6012725344644753e-06,
|
| 1979 |
+
"loss": 0.3796,
|
| 1980 |
"step": 13700
|
| 1981 |
},
|
| 1982 |
{
|
| 1983 |
"epoch": 4.860374690703429,
|
| 1984 |
+
"grad_norm": 3.179448366165161,
|
| 1985 |
+
"learning_rate": 1.424531636620714e-06,
|
| 1986 |
+
"loss": 0.3898,
|
| 1987 |
"step": 13750
|
| 1988 |
},
|
| 1989 |
{
|
| 1990 |
"epoch": 4.878048780487805,
|
| 1991 |
+
"grad_norm": 5.14663028717041,
|
| 1992 |
+
"learning_rate": 1.247790738776953e-06,
|
| 1993 |
+
"loss": 0.383,
|
| 1994 |
"step": 13800
|
| 1995 |
},
|
| 1996 |
{
|
| 1997 |
"epoch": 4.895722870272181,
|
| 1998 |
+
"grad_norm": 2.7722623348236084,
|
| 1999 |
+
"learning_rate": 1.071049840933192e-06,
|
| 2000 |
+
"loss": 0.3923,
|
| 2001 |
"step": 13850
|
| 2002 |
},
|
| 2003 |
{
|
| 2004 |
"epoch": 4.9133969600565575,
|
| 2005 |
+
"grad_norm": 4.3328447341918945,
|
| 2006 |
+
"learning_rate": 8.94308943089431e-07,
|
| 2007 |
+
"loss": 0.3859,
|
| 2008 |
"step": 13900
|
| 2009 |
},
|
| 2010 |
{
|
| 2011 |
"epoch": 4.931071049840933,
|
| 2012 |
+
"grad_norm": 3.5014865398406982,
|
| 2013 |
+
"learning_rate": 7.175680452456699e-07,
|
| 2014 |
+
"loss": 0.3909,
|
| 2015 |
"step": 13950
|
| 2016 |
},
|
| 2017 |
{
|
| 2018 |
"epoch": 4.948745139625309,
|
| 2019 |
+
"grad_norm": 4.449154376983643,
|
| 2020 |
+
"learning_rate": 5.408271474019089e-07,
|
| 2021 |
+
"loss": 0.4711,
|
| 2022 |
"step": 14000
|
| 2023 |
},
|
| 2024 |
{
|
| 2025 |
"epoch": 4.966419229409685,
|
| 2026 |
+
"grad_norm": 2.2578201293945312,
|
| 2027 |
+
"learning_rate": 3.640862495581478e-07,
|
| 2028 |
+
"loss": 0.3719,
|
| 2029 |
"step": 14050
|
| 2030 |
},
|
| 2031 |
{
|
| 2032 |
"epoch": 4.984093319194061,
|
| 2033 |
+
"grad_norm": 1.688942313194275,
|
| 2034 |
+
"learning_rate": 1.8734535171438673e-07,
|
| 2035 |
+
"loss": 0.4053,
|
| 2036 |
"step": 14100
|
| 2037 |
},
|
| 2038 |
{
|
| 2039 |
"epoch": 5.0,
|
| 2040 |
+
"eval_bertscore_f1": 0.9658045416033947,
|
| 2041 |
+
"eval_bleu": 0.5865134487850142,
|
| 2042 |
+
"eval_loss": 0.2939385771751404,
|
| 2043 |
+
"eval_meteor": 0.738667698887171,
|
| 2044 |
+
"eval_rouge1": 0.8397011041728719,
|
| 2045 |
+
"eval_rouge2": 0.7793367916496452,
|
| 2046 |
+
"eval_runtime": 1419.0322,
|
| 2047 |
+
"eval_samples_per_second": 4.555,
|
| 2048 |
+
"eval_steps_per_second": 0.569,
|
| 2049 |
"step": 14145
|
| 2050 |
}
|
| 2051 |
],
|
checkpoint-14145/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5432
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84eb264dc96b91d8b1c338a7f669bf17dafa7f32a2801b1215e7c0b8df1ea575
|
| 3 |
size 5432
|