Training in progress, step 4000, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1583480280
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d1de8fb6d5d71eb33c8845604be2dd7deaeeed74d9c9a4b7746d29050d18cd3
|
| 3 |
size 1583480280
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3166958572
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3648cb7e04dbe7d615d7389243cf296545176724dfd1acef9c321f68bbf4de41
|
| 3 |
size 3166958572
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e718af3156d6b40c17f5c33f24275fc2398dcbad98f0b2caad8b6817b49979a4
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1318375700a60b04c0a22b76fc70122f8e0729a2ae5ca4a0077237344a45104f
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43dd10e9f23213b9d3b365e2255a723d72c3f7686dc7616d93e93d05cf6d351e
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 1.
|
| 4 |
-
"best_model_checkpoint": "hieptt/vietnamese-correction-ft/checkpoint-
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2142,6 +2142,715 @@
|
|
| 2142 |
"eval_samples_per_second": 81.49,
|
| 2143 |
"eval_steps_per_second": 0.637,
|
| 2144 |
"step": 3000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2145 |
}
|
| 2146 |
],
|
| 2147 |
"logging_steps": 10,
|
|
@@ -2170,7 +2879,7 @@
|
|
| 2170 |
"attributes": {}
|
| 2171 |
}
|
| 2172 |
},
|
| 2173 |
-
"total_flos":
|
| 2174 |
"train_batch_size": 64,
|
| 2175 |
"trial_name": null,
|
| 2176 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 4000,
|
| 3 |
+
"best_metric": 1.4903924465179443,
|
| 4 |
+
"best_model_checkpoint": "hieptt/vietnamese-correction-ft/checkpoint-4000",
|
| 5 |
+
"epoch": 0.09674454602621778,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 4000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2142 |
"eval_samples_per_second": 81.49,
|
| 2143 |
"eval_steps_per_second": 0.637,
|
| 2144 |
"step": 3000
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.07280027088472887,
|
| 2148 |
+
"grad_norm": 0.33310985565185547,
|
| 2149 |
+
"learning_rate": 4.9792901502764075e-05,
|
| 2150 |
+
"loss": 1.4985,
|
| 2151 |
+
"step": 3010
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.07304213224979442,
|
| 2155 |
+
"grad_norm": 0.4175123870372772,
|
| 2156 |
+
"learning_rate": 4.979083755114006e-05,
|
| 2157 |
+
"loss": 1.496,
|
| 2158 |
+
"step": 3020
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.07328399361485996,
|
| 2162 |
+
"grad_norm": 0.3468244969844818,
|
| 2163 |
+
"learning_rate": 4.978876340893915e-05,
|
| 2164 |
+
"loss": 1.4988,
|
| 2165 |
+
"step": 3030
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.0735258549799255,
|
| 2169 |
+
"grad_norm": 0.353986531496048,
|
| 2170 |
+
"learning_rate": 4.978667907701396e-05,
|
| 2171 |
+
"loss": 1.5031,
|
| 2172 |
+
"step": 3040
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.07376771634499105,
|
| 2176 |
+
"grad_norm": 0.46692317724227905,
|
| 2177 |
+
"learning_rate": 4.9784584556221275e-05,
|
| 2178 |
+
"loss": 1.496,
|
| 2179 |
+
"step": 3050
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.07400957771005659,
|
| 2183 |
+
"grad_norm": 0.45075079798698425,
|
| 2184 |
+
"learning_rate": 4.978247984742208e-05,
|
| 2185 |
+
"loss": 1.5003,
|
| 2186 |
+
"step": 3060
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.07425143907512213,
|
| 2190 |
+
"grad_norm": 0.36158666014671326,
|
| 2191 |
+
"learning_rate": 4.9780364951481525e-05,
|
| 2192 |
+
"loss": 1.4975,
|
| 2193 |
+
"step": 3070
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.07449330044018769,
|
| 2197 |
+
"grad_norm": 0.38103577494621277,
|
| 2198 |
+
"learning_rate": 4.977823986926898e-05,
|
| 2199 |
+
"loss": 1.5029,
|
| 2200 |
+
"step": 3080
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.07473516180525323,
|
| 2204 |
+
"grad_norm": 0.6088710427284241,
|
| 2205 |
+
"learning_rate": 4.977610460165799e-05,
|
| 2206 |
+
"loss": 1.4975,
|
| 2207 |
+
"step": 3090
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.07497702317031878,
|
| 2211 |
+
"grad_norm": 0.3984760642051697,
|
| 2212 |
+
"learning_rate": 4.977395914952627e-05,
|
| 2213 |
+
"loss": 1.4991,
|
| 2214 |
+
"step": 3100
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.07521888453538432,
|
| 2218 |
+
"grad_norm": 0.581408679485321,
|
| 2219 |
+
"learning_rate": 4.977180351375573e-05,
|
| 2220 |
+
"loss": 1.4974,
|
| 2221 |
+
"step": 3110
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.07546074590044986,
|
| 2225 |
+
"grad_norm": 0.4531524181365967,
|
| 2226 |
+
"learning_rate": 4.9769637695232486e-05,
|
| 2227 |
+
"loss": 1.4981,
|
| 2228 |
+
"step": 3120
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.0757026072655154,
|
| 2232 |
+
"grad_norm": 0.3169078826904297,
|
| 2233 |
+
"learning_rate": 4.976746169484681e-05,
|
| 2234 |
+
"loss": 1.4993,
|
| 2235 |
+
"step": 3130
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.07594446863058095,
|
| 2239 |
+
"grad_norm": 0.3769000172615051,
|
| 2240 |
+
"learning_rate": 4.976527551349318e-05,
|
| 2241 |
+
"loss": 1.4966,
|
| 2242 |
+
"step": 3140
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.07618632999564649,
|
| 2246 |
+
"grad_norm": 0.34887903928756714,
|
| 2247 |
+
"learning_rate": 4.976307915207026e-05,
|
| 2248 |
+
"loss": 1.4998,
|
| 2249 |
+
"step": 3150
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.07642819136071204,
|
| 2253 |
+
"grad_norm": 0.3374789357185364,
|
| 2254 |
+
"learning_rate": 4.976087261148088e-05,
|
| 2255 |
+
"loss": 1.4968,
|
| 2256 |
+
"step": 3160
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.07667005272577758,
|
| 2260 |
+
"grad_norm": 0.3697901964187622,
|
| 2261 |
+
"learning_rate": 4.9758655892632063e-05,
|
| 2262 |
+
"loss": 1.5013,
|
| 2263 |
+
"step": 3170
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.07691191409084312,
|
| 2267 |
+
"grad_norm": 0.4903365671634674,
|
| 2268 |
+
"learning_rate": 4.975642899643503e-05,
|
| 2269 |
+
"loss": 1.493,
|
| 2270 |
+
"step": 3180
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.07715377545590868,
|
| 2274 |
+
"grad_norm": 0.3290081024169922,
|
| 2275 |
+
"learning_rate": 4.975419192380517e-05,
|
| 2276 |
+
"loss": 1.5078,
|
| 2277 |
+
"step": 3190
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.07739563682097422,
|
| 2281 |
+
"grad_norm": 0.33302998542785645,
|
| 2282 |
+
"learning_rate": 4.975194467566206e-05,
|
| 2283 |
+
"loss": 1.4965,
|
| 2284 |
+
"step": 3200
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.07763749818603977,
|
| 2288 |
+
"grad_norm": 0.29818981885910034,
|
| 2289 |
+
"learning_rate": 4.974968725292945e-05,
|
| 2290 |
+
"loss": 1.4978,
|
| 2291 |
+
"step": 3210
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.07787935955110531,
|
| 2295 |
+
"grad_norm": 0.5125884413719177,
|
| 2296 |
+
"learning_rate": 4.97474196565353e-05,
|
| 2297 |
+
"loss": 1.5056,
|
| 2298 |
+
"step": 3220
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.07812122091617085,
|
| 2302 |
+
"grad_norm": 0.39383506774902344,
|
| 2303 |
+
"learning_rate": 4.9745141887411704e-05,
|
| 2304 |
+
"loss": 1.5096,
|
| 2305 |
+
"step": 3230
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.0783630822812364,
|
| 2309 |
+
"grad_norm": 0.3851451575756073,
|
| 2310 |
+
"learning_rate": 4.974285394649499e-05,
|
| 2311 |
+
"loss": 1.501,
|
| 2312 |
+
"step": 3240
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.07860494364630194,
|
| 2316 |
+
"grad_norm": 0.37914183735847473,
|
| 2317 |
+
"learning_rate": 4.9740555834725646e-05,
|
| 2318 |
+
"loss": 1.4854,
|
| 2319 |
+
"step": 3250
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.07884680501136748,
|
| 2323 |
+
"grad_norm": 0.3494224548339844,
|
| 2324 |
+
"learning_rate": 4.973824755304832e-05,
|
| 2325 |
+
"loss": 1.5,
|
| 2326 |
+
"step": 3260
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.07908866637643303,
|
| 2330 |
+
"grad_norm": 0.41473448276519775,
|
| 2331 |
+
"learning_rate": 4.973592910241188e-05,
|
| 2332 |
+
"loss": 1.5026,
|
| 2333 |
+
"step": 3270
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.07933052774149857,
|
| 2337 |
+
"grad_norm": 0.5182355046272278,
|
| 2338 |
+
"learning_rate": 4.973360048376934e-05,
|
| 2339 |
+
"loss": 1.4998,
|
| 2340 |
+
"step": 3280
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.07957238910656411,
|
| 2344 |
+
"grad_norm": 0.3290833532810211,
|
| 2345 |
+
"learning_rate": 4.973126169807791e-05,
|
| 2346 |
+
"loss": 1.498,
|
| 2347 |
+
"step": 3290
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.07981425047162966,
|
| 2351 |
+
"grad_norm": 0.3534269332885742,
|
| 2352 |
+
"learning_rate": 4.9728912746298975e-05,
|
| 2353 |
+
"loss": 1.4917,
|
| 2354 |
+
"step": 3300
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.08005611183669521,
|
| 2358 |
+
"grad_norm": 0.32593613862991333,
|
| 2359 |
+
"learning_rate": 4.97265536293981e-05,
|
| 2360 |
+
"loss": 1.4909,
|
| 2361 |
+
"step": 3310
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.08029797320176076,
|
| 2365 |
+
"grad_norm": 0.38789451122283936,
|
| 2366 |
+
"learning_rate": 4.972418434834504e-05,
|
| 2367 |
+
"loss": 1.5006,
|
| 2368 |
+
"step": 3320
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.0805398345668263,
|
| 2372 |
+
"grad_norm": 0.46936559677124023,
|
| 2373 |
+
"learning_rate": 4.9721804904113705e-05,
|
| 2374 |
+
"loss": 1.5049,
|
| 2375 |
+
"step": 3330
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.08078169593189184,
|
| 2379 |
+
"grad_norm": 0.3091851472854614,
|
| 2380 |
+
"learning_rate": 4.97194152976822e-05,
|
| 2381 |
+
"loss": 1.4966,
|
| 2382 |
+
"step": 3340
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.08102355729695739,
|
| 2386 |
+
"grad_norm": 0.45702749490737915,
|
| 2387 |
+
"learning_rate": 4.97170155300328e-05,
|
| 2388 |
+
"loss": 1.5079,
|
| 2389 |
+
"step": 3350
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.08126541866202293,
|
| 2393 |
+
"grad_norm": 0.40697166323661804,
|
| 2394 |
+
"learning_rate": 4.9714605602151955e-05,
|
| 2395 |
+
"loss": 1.5018,
|
| 2396 |
+
"step": 3360
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.08150728002708847,
|
| 2400 |
+
"grad_norm": 0.532478392124176,
|
| 2401 |
+
"learning_rate": 4.971218551503029e-05,
|
| 2402 |
+
"loss": 1.4866,
|
| 2403 |
+
"step": 3370
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.08174914139215401,
|
| 2407 |
+
"grad_norm": 0.4273573160171509,
|
| 2408 |
+
"learning_rate": 4.970975526966264e-05,
|
| 2409 |
+
"loss": 1.495,
|
| 2410 |
+
"step": 3380
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.08199100275721956,
|
| 2414 |
+
"grad_norm": 0.4085642397403717,
|
| 2415 |
+
"learning_rate": 4.970731486704796e-05,
|
| 2416 |
+
"loss": 1.4925,
|
| 2417 |
+
"step": 3390
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.0822328641222851,
|
| 2421 |
+
"grad_norm": 0.28924253582954407,
|
| 2422 |
+
"learning_rate": 4.9704864308189415e-05,
|
| 2423 |
+
"loss": 1.4942,
|
| 2424 |
+
"step": 3400
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.08247472548735064,
|
| 2428 |
+
"grad_norm": 0.47220590710639954,
|
| 2429 |
+
"learning_rate": 4.9702403594094335e-05,
|
| 2430 |
+
"loss": 1.4882,
|
| 2431 |
+
"step": 3410
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.0827165868524162,
|
| 2435 |
+
"grad_norm": 0.35851794481277466,
|
| 2436 |
+
"learning_rate": 4.9699932725774234e-05,
|
| 2437 |
+
"loss": 1.4992,
|
| 2438 |
+
"step": 3420
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.08295844821748175,
|
| 2442 |
+
"grad_norm": 0.40433382987976074,
|
| 2443 |
+
"learning_rate": 4.969745170424478e-05,
|
| 2444 |
+
"loss": 1.5014,
|
| 2445 |
+
"step": 3430
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.08320030958254729,
|
| 2449 |
+
"grad_norm": 0.40462055802345276,
|
| 2450 |
+
"learning_rate": 4.969496053052584e-05,
|
| 2451 |
+
"loss": 1.4984,
|
| 2452 |
+
"step": 3440
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.08344217094761283,
|
| 2456 |
+
"grad_norm": 0.39987021684646606,
|
| 2457 |
+
"learning_rate": 4.969245920564144e-05,
|
| 2458 |
+
"loss": 1.4943,
|
| 2459 |
+
"step": 3450
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.08368403231267837,
|
| 2463 |
+
"grad_norm": 0.4980989992618561,
|
| 2464 |
+
"learning_rate": 4.968994773061977e-05,
|
| 2465 |
+
"loss": 1.4958,
|
| 2466 |
+
"step": 3460
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.08392589367774392,
|
| 2470 |
+
"grad_norm": 0.32841891050338745,
|
| 2471 |
+
"learning_rate": 4.968742610649321e-05,
|
| 2472 |
+
"loss": 1.4957,
|
| 2473 |
+
"step": 3470
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.08416775504280946,
|
| 2477 |
+
"grad_norm": 0.36307471990585327,
|
| 2478 |
+
"learning_rate": 4.96848943342983e-05,
|
| 2479 |
+
"loss": 1.5073,
|
| 2480 |
+
"step": 3480
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.084409616407875,
|
| 2484 |
+
"grad_norm": 0.5699554681777954,
|
| 2485 |
+
"learning_rate": 4.968235241507576e-05,
|
| 2486 |
+
"loss": 1.4998,
|
| 2487 |
+
"step": 3490
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.08465147777294055,
|
| 2491 |
+
"grad_norm": 0.36965617537498474,
|
| 2492 |
+
"learning_rate": 4.967980034987048e-05,
|
| 2493 |
+
"loss": 1.4925,
|
| 2494 |
+
"step": 3500
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.08489333913800609,
|
| 2498 |
+
"grad_norm": 0.3929795026779175,
|
| 2499 |
+
"learning_rate": 4.967723813973151e-05,
|
| 2500 |
+
"loss": 1.4951,
|
| 2501 |
+
"step": 3510
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.08513520050307163,
|
| 2505 |
+
"grad_norm": 0.35785186290740967,
|
| 2506 |
+
"learning_rate": 4.967466578571209e-05,
|
| 2507 |
+
"loss": 1.4903,
|
| 2508 |
+
"step": 3520
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.08537706186813718,
|
| 2512 |
+
"grad_norm": 0.3917689025402069,
|
| 2513 |
+
"learning_rate": 4.96720832888696e-05,
|
| 2514 |
+
"loss": 1.496,
|
| 2515 |
+
"step": 3530
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.08561892323320273,
|
| 2519 |
+
"grad_norm": 0.37746888399124146,
|
| 2520 |
+
"learning_rate": 4.9669490650265624e-05,
|
| 2521 |
+
"loss": 1.488,
|
| 2522 |
+
"step": 3540
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.08586078459826828,
|
| 2526 |
+
"grad_norm": 0.4309702217578888,
|
| 2527 |
+
"learning_rate": 4.9666887870965885e-05,
|
| 2528 |
+
"loss": 1.4991,
|
| 2529 |
+
"step": 3550
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.08610264596333382,
|
| 2533 |
+
"grad_norm": 0.4067014753818512,
|
| 2534 |
+
"learning_rate": 4.966427495204029e-05,
|
| 2535 |
+
"loss": 1.495,
|
| 2536 |
+
"step": 3560
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.08634450732839936,
|
| 2540 |
+
"grad_norm": 0.29803594946861267,
|
| 2541 |
+
"learning_rate": 4.966165189456291e-05,
|
| 2542 |
+
"loss": 1.4948,
|
| 2543 |
+
"step": 3570
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.08658636869346491,
|
| 2547 |
+
"grad_norm": 0.32155147194862366,
|
| 2548 |
+
"learning_rate": 4.9659018699611995e-05,
|
| 2549 |
+
"loss": 1.4942,
|
| 2550 |
+
"step": 3580
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.08682823005853045,
|
| 2554 |
+
"grad_norm": 0.30671581625938416,
|
| 2555 |
+
"learning_rate": 4.965637536826995e-05,
|
| 2556 |
+
"loss": 1.4941,
|
| 2557 |
+
"step": 3590
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.087070091423596,
|
| 2561 |
+
"grad_norm": 0.3266153931617737,
|
| 2562 |
+
"learning_rate": 4.965372190162333e-05,
|
| 2563 |
+
"loss": 1.5009,
|
| 2564 |
+
"step": 3600
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.08731195278866154,
|
| 2568 |
+
"grad_norm": 0.2917800545692444,
|
| 2569 |
+
"learning_rate": 4.9651058300762895e-05,
|
| 2570 |
+
"loss": 1.4879,
|
| 2571 |
+
"step": 3610
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.08755381415372708,
|
| 2575 |
+
"grad_norm": 0.3614349067211151,
|
| 2576 |
+
"learning_rate": 4.9648384566783545e-05,
|
| 2577 |
+
"loss": 1.4868,
|
| 2578 |
+
"step": 3620
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.08779567551879262,
|
| 2582 |
+
"grad_norm": 0.3470216691493988,
|
| 2583 |
+
"learning_rate": 4.964570070078435e-05,
|
| 2584 |
+
"loss": 1.4854,
|
| 2585 |
+
"step": 3630
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.08803753688385817,
|
| 2589 |
+
"grad_norm": 0.5655723214149475,
|
| 2590 |
+
"learning_rate": 4.964300670386856e-05,
|
| 2591 |
+
"loss": 1.4926,
|
| 2592 |
+
"step": 3640
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.08827939824892371,
|
| 2596 |
+
"grad_norm": 1.6314070224761963,
|
| 2597 |
+
"learning_rate": 4.9640302577143536e-05,
|
| 2598 |
+
"loss": 1.5252,
|
| 2599 |
+
"step": 3650
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.08852125961398927,
|
| 2603 |
+
"grad_norm": 34.27970886230469,
|
| 2604 |
+
"learning_rate": 4.963758832172088e-05,
|
| 2605 |
+
"loss": 4.7597,
|
| 2606 |
+
"step": 3660
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.08876312097905481,
|
| 2610 |
+
"grad_norm": 9.672608375549316,
|
| 2611 |
+
"learning_rate": 4.9634863938716315e-05,
|
| 2612 |
+
"loss": 9.2436,
|
| 2613 |
+
"step": 3670
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.08900498234412035,
|
| 2617 |
+
"grad_norm": 18.468290328979492,
|
| 2618 |
+
"learning_rate": 4.9632129429249716e-05,
|
| 2619 |
+
"loss": 8.3476,
|
| 2620 |
+
"step": 3680
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.0892468437091859,
|
| 2624 |
+
"grad_norm": 4.984256744384766,
|
| 2625 |
+
"learning_rate": 4.9629384794445157e-05,
|
| 2626 |
+
"loss": 8.0705,
|
| 2627 |
+
"step": 3690
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.08948870507425144,
|
| 2631 |
+
"grad_norm": 3.9478564262390137,
|
| 2632 |
+
"learning_rate": 4.962663003543083e-05,
|
| 2633 |
+
"loss": 7.7813,
|
| 2634 |
+
"step": 3700
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.08973056643931698,
|
| 2638 |
+
"grad_norm": 3.284527063369751,
|
| 2639 |
+
"learning_rate": 4.962386515333913e-05,
|
| 2640 |
+
"loss": 7.5749,
|
| 2641 |
+
"step": 3710
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.08997242780438253,
|
| 2645 |
+
"grad_norm": 4.682032585144043,
|
| 2646 |
+
"learning_rate": 4.962109014930659e-05,
|
| 2647 |
+
"loss": 7.3946,
|
| 2648 |
+
"step": 3720
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.09021428916944807,
|
| 2652 |
+
"grad_norm": 3.8620240688323975,
|
| 2653 |
+
"learning_rate": 4.9618305024473905e-05,
|
| 2654 |
+
"loss": 5.399,
|
| 2655 |
+
"step": 3730
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.09045615053451361,
|
| 2659 |
+
"grad_norm": 1.1052254438400269,
|
| 2660 |
+
"learning_rate": 4.961550977998594e-05,
|
| 2661 |
+
"loss": 1.7208,
|
| 2662 |
+
"step": 3740
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.09069801189957916,
|
| 2666 |
+
"grad_norm": 45.039066314697266,
|
| 2667 |
+
"learning_rate": 4.9612704416991716e-05,
|
| 2668 |
+
"loss": 1.6211,
|
| 2669 |
+
"step": 3750
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.0909398732646447,
|
| 2673 |
+
"grad_norm": 0.7098454236984253,
|
| 2674 |
+
"learning_rate": 4.960988893664441e-05,
|
| 2675 |
+
"loss": 1.6957,
|
| 2676 |
+
"step": 3760
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.09118173462971026,
|
| 2680 |
+
"grad_norm": 2.7060585021972656,
|
| 2681 |
+
"learning_rate": 4.9607063340101354e-05,
|
| 2682 |
+
"loss": 1.6564,
|
| 2683 |
+
"step": 3770
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.0914235959947758,
|
| 2687 |
+
"grad_norm": 2.3458268642425537,
|
| 2688 |
+
"learning_rate": 4.960422762852406e-05,
|
| 2689 |
+
"loss": 2.4888,
|
| 2690 |
+
"step": 3780
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.09166545735984134,
|
| 2694 |
+
"grad_norm": 12.697403907775879,
|
| 2695 |
+
"learning_rate": 4.960138180307816e-05,
|
| 2696 |
+
"loss": 2.405,
|
| 2697 |
+
"step": 3790
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.09190731872490689,
|
| 2701 |
+
"grad_norm": 7.4579854011535645,
|
| 2702 |
+
"learning_rate": 4.959852586493349e-05,
|
| 2703 |
+
"loss": 3.2748,
|
| 2704 |
+
"step": 3800
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.09214918008997243,
|
| 2708 |
+
"grad_norm": 0.6929895877838135,
|
| 2709 |
+
"learning_rate": 4.9595659815263985e-05,
|
| 2710 |
+
"loss": 2.112,
|
| 2711 |
+
"step": 3810
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.09239104145503797,
|
| 2715 |
+
"grad_norm": 0.563450276851654,
|
| 2716 |
+
"learning_rate": 4.95927836552478e-05,
|
| 2717 |
+
"loss": 1.6898,
|
| 2718 |
+
"step": 3820
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.09263290282010352,
|
| 2722 |
+
"grad_norm": 0.4547874629497528,
|
| 2723 |
+
"learning_rate": 4.9589897386067205e-05,
|
| 2724 |
+
"loss": 1.5361,
|
| 2725 |
+
"step": 3830
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.09287476418516906,
|
| 2729 |
+
"grad_norm": 0.525037407875061,
|
| 2730 |
+
"learning_rate": 4.958700100890863e-05,
|
| 2731 |
+
"loss": 1.5204,
|
| 2732 |
+
"step": 3840
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.0931166255502346,
|
| 2736 |
+
"grad_norm": 0.826223611831665,
|
| 2737 |
+
"learning_rate": 4.9584094524962684e-05,
|
| 2738 |
+
"loss": 1.5188,
|
| 2739 |
+
"step": 3850
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.09335848691530015,
|
| 2743 |
+
"grad_norm": 0.45551028847694397,
|
| 2744 |
+
"learning_rate": 4.95811779354241e-05,
|
| 2745 |
+
"loss": 1.5779,
|
| 2746 |
+
"step": 3860
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.09360034828036569,
|
| 2750 |
+
"grad_norm": 0.4146782159805298,
|
| 2751 |
+
"learning_rate": 4.957825124149177e-05,
|
| 2752 |
+
"loss": 1.5293,
|
| 2753 |
+
"step": 3870
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.09384220964543123,
|
| 2757 |
+
"grad_norm": 0.40724071860313416,
|
| 2758 |
+
"learning_rate": 4.9575314444368756e-05,
|
| 2759 |
+
"loss": 1.5346,
|
| 2760 |
+
"step": 3880
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.09408407101049679,
|
| 2764 |
+
"grad_norm": 0.47108572721481323,
|
| 2765 |
+
"learning_rate": 4.957236754526227e-05,
|
| 2766 |
+
"loss": 1.5205,
|
| 2767 |
+
"step": 3890
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.09432593237556233,
|
| 2771 |
+
"grad_norm": 0.45449286699295044,
|
| 2772 |
+
"learning_rate": 4.9569410545383665e-05,
|
| 2773 |
+
"loss": 1.5149,
|
| 2774 |
+
"step": 3900
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.09456779374062788,
|
| 2778 |
+
"grad_norm": 0.5905476808547974,
|
| 2779 |
+
"learning_rate": 4.956644344594845e-05,
|
| 2780 |
+
"loss": 1.5242,
|
| 2781 |
+
"step": 3910
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.09480965510569342,
|
| 2785 |
+
"grad_norm": 0.4378545582294464,
|
| 2786 |
+
"learning_rate": 4.956346624817629e-05,
|
| 2787 |
+
"loss": 1.495,
|
| 2788 |
+
"step": 3920
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.09505151647075896,
|
| 2792 |
+
"grad_norm": 0.3891792893409729,
|
| 2793 |
+
"learning_rate": 4.956047895329101e-05,
|
| 2794 |
+
"loss": 1.5011,
|
| 2795 |
+
"step": 3930
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.0952933778358245,
|
| 2799 |
+
"grad_norm": 0.4357585310935974,
|
| 2800 |
+
"learning_rate": 4.9557481562520546e-05,
|
| 2801 |
+
"loss": 1.5061,
|
| 2802 |
+
"step": 3940
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.09553523920089005,
|
| 2806 |
+
"grad_norm": 0.36777547001838684,
|
| 2807 |
+
"learning_rate": 4.955447407709704e-05,
|
| 2808 |
+
"loss": 1.4988,
|
| 2809 |
+
"step": 3950
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.09577710056595559,
|
| 2813 |
+
"grad_norm": 0.42386820912361145,
|
| 2814 |
+
"learning_rate": 4.9551456498256744e-05,
|
| 2815 |
+
"loss": 1.5029,
|
| 2816 |
+
"step": 3960
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 0.09601896193102114,
|
| 2820 |
+
"grad_norm": 0.5050960183143616,
|
| 2821 |
+
"learning_rate": 4.954842882724007e-05,
|
| 2822 |
+
"loss": 1.5055,
|
| 2823 |
+
"step": 3970
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 0.09626082329608668,
|
| 2827 |
+
"grad_norm": 0.32713472843170166,
|
| 2828 |
+
"learning_rate": 4.954539106529157e-05,
|
| 2829 |
+
"loss": 1.5023,
|
| 2830 |
+
"step": 3980
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 0.09650268466115222,
|
| 2834 |
+
"grad_norm": 0.2444874346256256,
|
| 2835 |
+
"learning_rate": 4.9542343213659974e-05,
|
| 2836 |
+
"loss": 1.4971,
|
| 2837 |
+
"step": 3990
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 0.09674454602621778,
|
| 2841 |
+
"grad_norm": 0.4041503965854645,
|
| 2842 |
+
"learning_rate": 4.953928527359812e-05,
|
| 2843 |
+
"loss": 1.503,
|
| 2844 |
+
"step": 4000
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.09674454602621778,
|
| 2848 |
+
"eval_loss": 1.4903924465179443,
|
| 2849 |
+
"eval_runtime": 1236.6443,
|
| 2850 |
+
"eval_sacrebleu": 96.6133417933895,
|
| 2851 |
+
"eval_samples_per_second": 81.71,
|
| 2852 |
+
"eval_steps_per_second": 0.639,
|
| 2853 |
+
"step": 4000
|
| 2854 |
}
|
| 2855 |
],
|
| 2856 |
"logging_steps": 10,
|
|
|
|
| 2879 |
"attributes": {}
|
| 2880 |
}
|
| 2881 |
},
|
| 2882 |
+
"total_flos": 8.137680635088077e+16,
|
| 2883 |
"train_batch_size": 64,
|
| 2884 |
"trial_name": null,
|
| 2885 |
"trial_params": null
|