Training in progress, step 2000, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +711 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 791869518
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ddca534d10503034b593de387e205daca04a072f5ccbe17faac957a202b96d5
|
| 3 |
size 791869518
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2375752250
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cb1c32a2fe7e4ba0a80cc2f7de739a5497222f3987f01b2d711b570de3fbe5c
|
| 3 |
size 2375752250
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06fea830cf5ad73ec00d500ea6fb952740ac936f18e93fa2d32abde1ea3ead92
|
| 3 |
size 15984
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be561d1df19be227394d8ea607c54262a06c9bf880af0aa5e04a52596a2a6cb0
|
| 3 |
size 15984
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03f3e24417a59435f5a8450a4aeb0f09cc92734b5c3b45a0701b2c043c415c05
|
| 3 |
size 15984
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2bea02744c29f30024590ab1629a0e7b7dabbf1e8476456c2e7c5ce46dc35c28
|
| 3 |
size 15984
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:041be966454b60c86af576fc1eb7f34189114689abff8f9622b947110f7334c8
|
| 3 |
size 15984
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b85766f6596d15a810177d77dd259d9b50588cf100ec5f8ebff5fed881d57957
|
| 3 |
size 15984
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8be75d04b1ebe614241b88fd010a5dda1b7bf703c00c6ebe310ca07975830fe7
|
| 3 |
size 15984
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4699833a7ab4cb692996ef7567f934c0bac79d6a067963a873f89a38e412bd48
|
| 3 |
size 15984
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06d017eeef42e0127e56fc73579acf75949a31de7dbc0f95bf4428c95dd75f92
|
| 3 |
size 1000
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2131,6 +2131,714 @@
|
|
| 2131 |
"eval_samples_per_second": 609.887,
|
| 2132 |
"eval_steps_per_second": 38.118,
|
| 2133 |
"step": 1500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2134 |
}
|
| 2135 |
],
|
| 2136 |
"logging_steps": 5,
|
|
@@ -2150,7 +2858,7 @@
|
|
| 2150 |
"attributes": {}
|
| 2151 |
}
|
| 2152 |
},
|
| 2153 |
-
"total_flos":
|
| 2154 |
"train_batch_size": 4,
|
| 2155 |
"trial_name": null,
|
| 2156 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.4623509098921422,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2131 |
"eval_samples_per_second": 609.887,
|
| 2132 |
"eval_steps_per_second": 38.118,
|
| 2133 |
"step": 1500
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.347919059693837,
|
| 2137 |
+
"grad_norm": 167.875,
|
| 2138 |
+
"learning_rate": 7.245632065775952e-06,
|
| 2139 |
+
"loss": 71.7923,
|
| 2140 |
+
"step": 1505
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.34907493696856734,
|
| 2144 |
+
"grad_norm": 180.625,
|
| 2145 |
+
"learning_rate": 7.2327852004111e-06,
|
| 2146 |
+
"loss": 72.9641,
|
| 2147 |
+
"step": 1510
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.3502308142432977,
|
| 2151 |
+
"grad_norm": 161.875,
|
| 2152 |
+
"learning_rate": 7.21993833504625e-06,
|
| 2153 |
+
"loss": 71.7615,
|
| 2154 |
+
"step": 1515
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.3513866915180281,
|
| 2158 |
+
"grad_norm": 183.5,
|
| 2159 |
+
"learning_rate": 7.2070914696813985e-06,
|
| 2160 |
+
"loss": 71.8146,
|
| 2161 |
+
"step": 1520
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.3525425687927584,
|
| 2165 |
+
"grad_norm": 162.0,
|
| 2166 |
+
"learning_rate": 7.194244604316547e-06,
|
| 2167 |
+
"loss": 71.0128,
|
| 2168 |
+
"step": 1525
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.3536984460674888,
|
| 2172 |
+
"grad_norm": 163.375,
|
| 2173 |
+
"learning_rate": 7.181397738951696e-06,
|
| 2174 |
+
"loss": 71.2744,
|
| 2175 |
+
"step": 1530
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.3548543233422191,
|
| 2179 |
+
"grad_norm": 173.75,
|
| 2180 |
+
"learning_rate": 7.168550873586846e-06,
|
| 2181 |
+
"loss": 72.7448,
|
| 2182 |
+
"step": 1535
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.3560102006169495,
|
| 2186 |
+
"grad_norm": 182.625,
|
| 2187 |
+
"learning_rate": 7.155704008221994e-06,
|
| 2188 |
+
"loss": 72.3184,
|
| 2189 |
+
"step": 1540
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.35716607789167987,
|
| 2193 |
+
"grad_norm": 169.625,
|
| 2194 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 2195 |
+
"loss": 73.2466,
|
| 2196 |
+
"step": 1545
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.3583219551664102,
|
| 2200 |
+
"grad_norm": 169.375,
|
| 2201 |
+
"learning_rate": 7.1300102774922916e-06,
|
| 2202 |
+
"loss": 72.9034,
|
| 2203 |
+
"step": 1550
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.35947783244114057,
|
| 2207 |
+
"grad_norm": 155.0,
|
| 2208 |
+
"learning_rate": 7.117163412127441e-06,
|
| 2209 |
+
"loss": 70.7591,
|
| 2210 |
+
"step": 1555
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.36063370971587094,
|
| 2214 |
+
"grad_norm": 192.25,
|
| 2215 |
+
"learning_rate": 7.104316546762591e-06,
|
| 2216 |
+
"loss": 72.9981,
|
| 2217 |
+
"step": 1560
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.36178958699060126,
|
| 2221 |
+
"grad_norm": 167.375,
|
| 2222 |
+
"learning_rate": 7.091469681397739e-06,
|
| 2223 |
+
"loss": 71.7857,
|
| 2224 |
+
"step": 1565
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.36294546426533164,
|
| 2228 |
+
"grad_norm": 160.5,
|
| 2229 |
+
"learning_rate": 7.078622816032889e-06,
|
| 2230 |
+
"loss": 71.0528,
|
| 2231 |
+
"step": 1570
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.36410134154006196,
|
| 2235 |
+
"grad_norm": 178.75,
|
| 2236 |
+
"learning_rate": 7.0657759506680375e-06,
|
| 2237 |
+
"loss": 72.1775,
|
| 2238 |
+
"step": 1575
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.36525721881479234,
|
| 2242 |
+
"grad_norm": 161.5,
|
| 2243 |
+
"learning_rate": 7.052929085303186e-06,
|
| 2244 |
+
"loss": 71.9378,
|
| 2245 |
+
"step": 1580
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.3664130960895227,
|
| 2249 |
+
"grad_norm": 158.25,
|
| 2250 |
+
"learning_rate": 7.040082219938336e-06,
|
| 2251 |
+
"loss": 71.7813,
|
| 2252 |
+
"step": 1585
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.36756897336425304,
|
| 2256 |
+
"grad_norm": 160.0,
|
| 2257 |
+
"learning_rate": 7.027235354573485e-06,
|
| 2258 |
+
"loss": 71.8137,
|
| 2259 |
+
"step": 1590
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.3687248506389834,
|
| 2263 |
+
"grad_norm": 165.0,
|
| 2264 |
+
"learning_rate": 7.014388489208634e-06,
|
| 2265 |
+
"loss": 73.0897,
|
| 2266 |
+
"step": 1595
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.36988072791371374,
|
| 2270 |
+
"grad_norm": 166.125,
|
| 2271 |
+
"learning_rate": 7.0015416238437826e-06,
|
| 2272 |
+
"loss": 71.867,
|
| 2273 |
+
"step": 1600
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.3710366051884441,
|
| 2277 |
+
"grad_norm": 152.375,
|
| 2278 |
+
"learning_rate": 6.988694758478932e-06,
|
| 2279 |
+
"loss": 72.0306,
|
| 2280 |
+
"step": 1605
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 0.3721924824631745,
|
| 2284 |
+
"grad_norm": 153.625,
|
| 2285 |
+
"learning_rate": 6.97584789311408e-06,
|
| 2286 |
+
"loss": 71.2797,
|
| 2287 |
+
"step": 1610
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.3733483597379048,
|
| 2291 |
+
"grad_norm": 157.75,
|
| 2292 |
+
"learning_rate": 6.96300102774923e-06,
|
| 2293 |
+
"loss": 70.6556,
|
| 2294 |
+
"step": 1615
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 0.3745042370126352,
|
| 2298 |
+
"grad_norm": 206.25,
|
| 2299 |
+
"learning_rate": 6.950154162384378e-06,
|
| 2300 |
+
"loss": 70.427,
|
| 2301 |
+
"step": 1620
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.37566011428736557,
|
| 2305 |
+
"grad_norm": 198.0,
|
| 2306 |
+
"learning_rate": 6.937307297019528e-06,
|
| 2307 |
+
"loss": 70.8553,
|
| 2308 |
+
"step": 1625
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 0.3768159915620959,
|
| 2312 |
+
"grad_norm": 172.375,
|
| 2313 |
+
"learning_rate": 6.924460431654677e-06,
|
| 2314 |
+
"loss": 72.5536,
|
| 2315 |
+
"step": 1630
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.37797186883682626,
|
| 2319 |
+
"grad_norm": 162.25,
|
| 2320 |
+
"learning_rate": 6.911613566289825e-06,
|
| 2321 |
+
"loss": 71.3941,
|
| 2322 |
+
"step": 1635
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.3791277461115566,
|
| 2326 |
+
"grad_norm": 154.75,
|
| 2327 |
+
"learning_rate": 6.898766700924975e-06,
|
| 2328 |
+
"loss": 70.9165,
|
| 2329 |
+
"step": 1640
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.38028362338628696,
|
| 2333 |
+
"grad_norm": 181.75,
|
| 2334 |
+
"learning_rate": 6.885919835560124e-06,
|
| 2335 |
+
"loss": 72.0093,
|
| 2336 |
+
"step": 1645
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.38143950066101734,
|
| 2340 |
+
"grad_norm": 150.0,
|
| 2341 |
+
"learning_rate": 6.873072970195273e-06,
|
| 2342 |
+
"loss": 71.9246,
|
| 2343 |
+
"step": 1650
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.38259537793574766,
|
| 2347 |
+
"grad_norm": 164.875,
|
| 2348 |
+
"learning_rate": 6.860226104830422e-06,
|
| 2349 |
+
"loss": 71.8933,
|
| 2350 |
+
"step": 1655
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.38375125521047804,
|
| 2354 |
+
"grad_norm": 157.0,
|
| 2355 |
+
"learning_rate": 6.847379239465571e-06,
|
| 2356 |
+
"loss": 72.0258,
|
| 2357 |
+
"step": 1660
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.38490713248520836,
|
| 2361 |
+
"grad_norm": 177.375,
|
| 2362 |
+
"learning_rate": 6.834532374100719e-06,
|
| 2363 |
+
"loss": 72.3605,
|
| 2364 |
+
"step": 1665
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.38606300975993874,
|
| 2368 |
+
"grad_norm": 175.625,
|
| 2369 |
+
"learning_rate": 6.821685508735869e-06,
|
| 2370 |
+
"loss": 71.7816,
|
| 2371 |
+
"step": 1670
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.3872188870346691,
|
| 2375 |
+
"grad_norm": 178.375,
|
| 2376 |
+
"learning_rate": 6.808838643371019e-06,
|
| 2377 |
+
"loss": 71.8166,
|
| 2378 |
+
"step": 1675
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.38837476430939943,
|
| 2382 |
+
"grad_norm": 164.0,
|
| 2383 |
+
"learning_rate": 6.795991778006167e-06,
|
| 2384 |
+
"loss": 72.0027,
|
| 2385 |
+
"step": 1680
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.3895306415841298,
|
| 2389 |
+
"grad_norm": 168.0,
|
| 2390 |
+
"learning_rate": 6.783144912641316e-06,
|
| 2391 |
+
"loss": 71.8431,
|
| 2392 |
+
"step": 1685
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.3906865188588602,
|
| 2396 |
+
"grad_norm": 165.375,
|
| 2397 |
+
"learning_rate": 6.770298047276464e-06,
|
| 2398 |
+
"loss": 71.9873,
|
| 2399 |
+
"step": 1690
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.3918423961335905,
|
| 2403 |
+
"grad_norm": 161.25,
|
| 2404 |
+
"learning_rate": 6.757451181911614e-06,
|
| 2405 |
+
"loss": 72.0379,
|
| 2406 |
+
"step": 1695
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.3929982734083209,
|
| 2410 |
+
"grad_norm": 175.125,
|
| 2411 |
+
"learning_rate": 6.744604316546764e-06,
|
| 2412 |
+
"loss": 72.1231,
|
| 2413 |
+
"step": 1700
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.3941541506830512,
|
| 2417 |
+
"grad_norm": 177.375,
|
| 2418 |
+
"learning_rate": 6.731757451181912e-06,
|
| 2419 |
+
"loss": 70.1524,
|
| 2420 |
+
"step": 1705
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.3953100279577816,
|
| 2424 |
+
"grad_norm": 167.75,
|
| 2425 |
+
"learning_rate": 6.7189105858170614e-06,
|
| 2426 |
+
"loss": 71.1926,
|
| 2427 |
+
"step": 1710
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.39646590523251196,
|
| 2431 |
+
"grad_norm": 148.125,
|
| 2432 |
+
"learning_rate": 6.70606372045221e-06,
|
| 2433 |
+
"loss": 72.7071,
|
| 2434 |
+
"step": 1715
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.3976217825072423,
|
| 2438 |
+
"grad_norm": 162.875,
|
| 2439 |
+
"learning_rate": 6.693216855087359e-06,
|
| 2440 |
+
"loss": 71.857,
|
| 2441 |
+
"step": 1720
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.39877765978197266,
|
| 2445 |
+
"grad_norm": 198.625,
|
| 2446 |
+
"learning_rate": 6.680369989722508e-06,
|
| 2447 |
+
"loss": 72.0683,
|
| 2448 |
+
"step": 1725
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.399933537056703,
|
| 2452 |
+
"grad_norm": 155.625,
|
| 2453 |
+
"learning_rate": 6.667523124357658e-06,
|
| 2454 |
+
"loss": 70.8161,
|
| 2455 |
+
"step": 1730
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.40108941433143336,
|
| 2459 |
+
"grad_norm": 164.0,
|
| 2460 |
+
"learning_rate": 6.654676258992806e-06,
|
| 2461 |
+
"loss": 70.9434,
|
| 2462 |
+
"step": 1735
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.40224529160616374,
|
| 2466 |
+
"grad_norm": 159.375,
|
| 2467 |
+
"learning_rate": 6.641829393627955e-06,
|
| 2468 |
+
"loss": 71.407,
|
| 2469 |
+
"step": 1740
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.40340116888089406,
|
| 2473 |
+
"grad_norm": 161.875,
|
| 2474 |
+
"learning_rate": 6.628982528263105e-06,
|
| 2475 |
+
"loss": 71.7674,
|
| 2476 |
+
"step": 1745
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.40455704615562443,
|
| 2480 |
+
"grad_norm": 162.625,
|
| 2481 |
+
"learning_rate": 6.616135662898253e-06,
|
| 2482 |
+
"loss": 72.8434,
|
| 2483 |
+
"step": 1750
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.40571292343035475,
|
| 2487 |
+
"grad_norm": 205.75,
|
| 2488 |
+
"learning_rate": 6.603288797533403e-06,
|
| 2489 |
+
"loss": 71.8435,
|
| 2490 |
+
"step": 1755
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.40686880070508513,
|
| 2494 |
+
"grad_norm": 149.625,
|
| 2495 |
+
"learning_rate": 6.590441932168551e-06,
|
| 2496 |
+
"loss": 71.5685,
|
| 2497 |
+
"step": 1760
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.4080246779798155,
|
| 2501 |
+
"grad_norm": 144.625,
|
| 2502 |
+
"learning_rate": 6.5775950668037005e-06,
|
| 2503 |
+
"loss": 71.5258,
|
| 2504 |
+
"step": 1765
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.40918055525454583,
|
| 2508 |
+
"grad_norm": 171.5,
|
| 2509 |
+
"learning_rate": 6.564748201438849e-06,
|
| 2510 |
+
"loss": 70.664,
|
| 2511 |
+
"step": 1770
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.4103364325292762,
|
| 2515 |
+
"grad_norm": 177.875,
|
| 2516 |
+
"learning_rate": 6.551901336073998e-06,
|
| 2517 |
+
"loss": 71.853,
|
| 2518 |
+
"step": 1775
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.4114923098040066,
|
| 2522 |
+
"grad_norm": 168.5,
|
| 2523 |
+
"learning_rate": 6.539054470709148e-06,
|
| 2524 |
+
"loss": 71.6207,
|
| 2525 |
+
"step": 1780
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.4126481870787369,
|
| 2529 |
+
"grad_norm": 166.375,
|
| 2530 |
+
"learning_rate": 6.526207605344297e-06,
|
| 2531 |
+
"loss": 71.5577,
|
| 2532 |
+
"step": 1785
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.4138040643534673,
|
| 2536 |
+
"grad_norm": 187.375,
|
| 2537 |
+
"learning_rate": 6.5133607399794455e-06,
|
| 2538 |
+
"loss": 71.5951,
|
| 2539 |
+
"step": 1790
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.4149599416281976,
|
| 2543 |
+
"grad_norm": 167.625,
|
| 2544 |
+
"learning_rate": 6.500513874614594e-06,
|
| 2545 |
+
"loss": 72.7333,
|
| 2546 |
+
"step": 1795
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.416115818902928,
|
| 2550 |
+
"grad_norm": 177.375,
|
| 2551 |
+
"learning_rate": 6.487667009249744e-06,
|
| 2552 |
+
"loss": 71.4623,
|
| 2553 |
+
"step": 1800
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.41727169617765836,
|
| 2557 |
+
"grad_norm": 176.25,
|
| 2558 |
+
"learning_rate": 6.474820143884892e-06,
|
| 2559 |
+
"loss": 70.88,
|
| 2560 |
+
"step": 1805
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.4184275734523887,
|
| 2564 |
+
"grad_norm": 178.25,
|
| 2565 |
+
"learning_rate": 6.461973278520042e-06,
|
| 2566 |
+
"loss": 70.7191,
|
| 2567 |
+
"step": 1810
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.41958345072711906,
|
| 2571 |
+
"grad_norm": 190.75,
|
| 2572 |
+
"learning_rate": 6.4491264131551915e-06,
|
| 2573 |
+
"loss": 71.4613,
|
| 2574 |
+
"step": 1815
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.4207393280018494,
|
| 2578 |
+
"grad_norm": 175.75,
|
| 2579 |
+
"learning_rate": 6.4362795477903395e-06,
|
| 2580 |
+
"loss": 71.9579,
|
| 2581 |
+
"step": 1820
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.42189520527657975,
|
| 2585 |
+
"grad_norm": 159.625,
|
| 2586 |
+
"learning_rate": 6.423432682425489e-06,
|
| 2587 |
+
"loss": 70.9512,
|
| 2588 |
+
"step": 1825
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.42305108255131013,
|
| 2592 |
+
"grad_norm": 171.0,
|
| 2593 |
+
"learning_rate": 6.410585817060637e-06,
|
| 2594 |
+
"loss": 71.0018,
|
| 2595 |
+
"step": 1830
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.42420695982604045,
|
| 2599 |
+
"grad_norm": 168.625,
|
| 2600 |
+
"learning_rate": 6.397738951695787e-06,
|
| 2601 |
+
"loss": 69.9553,
|
| 2602 |
+
"step": 1835
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.42536283710077083,
|
| 2606 |
+
"grad_norm": 161.875,
|
| 2607 |
+
"learning_rate": 6.384892086330936e-06,
|
| 2608 |
+
"loss": 71.8582,
|
| 2609 |
+
"step": 1840
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.4265187143755012,
|
| 2613 |
+
"grad_norm": 159.375,
|
| 2614 |
+
"learning_rate": 6.3720452209660846e-06,
|
| 2615 |
+
"loss": 71.7597,
|
| 2616 |
+
"step": 1845
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.42767459165023153,
|
| 2620 |
+
"grad_norm": 168.75,
|
| 2621 |
+
"learning_rate": 6.359198355601233e-06,
|
| 2622 |
+
"loss": 72.3329,
|
| 2623 |
+
"step": 1850
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.4288304689249619,
|
| 2627 |
+
"grad_norm": 227.625,
|
| 2628 |
+
"learning_rate": 6.346351490236383e-06,
|
| 2629 |
+
"loss": 70.8971,
|
| 2630 |
+
"step": 1855
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 0.4299863461996922,
|
| 2634 |
+
"grad_norm": 162.125,
|
| 2635 |
+
"learning_rate": 6.333504624871531e-06,
|
| 2636 |
+
"loss": 72.014,
|
| 2637 |
+
"step": 1860
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.4311422234744226,
|
| 2641 |
+
"grad_norm": 167.375,
|
| 2642 |
+
"learning_rate": 6.320657759506681e-06,
|
| 2643 |
+
"loss": 70.4521,
|
| 2644 |
+
"step": 1865
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 0.432298100749153,
|
| 2648 |
+
"grad_norm": 176.625,
|
| 2649 |
+
"learning_rate": 6.3078108941418305e-06,
|
| 2650 |
+
"loss": 71.6845,
|
| 2651 |
+
"step": 1870
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 0.4334539780238833,
|
| 2655 |
+
"grad_norm": 160.5,
|
| 2656 |
+
"learning_rate": 6.2949640287769785e-06,
|
| 2657 |
+
"loss": 70.4146,
|
| 2658 |
+
"step": 1875
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 0.4346098552986137,
|
| 2662 |
+
"grad_norm": 169.0,
|
| 2663 |
+
"learning_rate": 6.282117163412128e-06,
|
| 2664 |
+
"loss": 70.7766,
|
| 2665 |
+
"step": 1880
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 0.435765732573344,
|
| 2669 |
+
"grad_norm": 166.125,
|
| 2670 |
+
"learning_rate": 6.269270298047278e-06,
|
| 2671 |
+
"loss": 70.6441,
|
| 2672 |
+
"step": 1885
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 0.4369216098480744,
|
| 2676 |
+
"grad_norm": 169.875,
|
| 2677 |
+
"learning_rate": 6.256423432682426e-06,
|
| 2678 |
+
"loss": 70.6906,
|
| 2679 |
+
"step": 1890
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 0.43807748712280475,
|
| 2683 |
+
"grad_norm": 189.0,
|
| 2684 |
+
"learning_rate": 6.2435765673175756e-06,
|
| 2685 |
+
"loss": 69.9695,
|
| 2686 |
+
"step": 1895
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 0.4392333643975351,
|
| 2690 |
+
"grad_norm": 163.5,
|
| 2691 |
+
"learning_rate": 6.2307297019527236e-06,
|
| 2692 |
+
"loss": 72.2346,
|
| 2693 |
+
"step": 1900
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 0.44038924167226545,
|
| 2697 |
+
"grad_norm": 168.25,
|
| 2698 |
+
"learning_rate": 6.217882836587873e-06,
|
| 2699 |
+
"loss": 70.2958,
|
| 2700 |
+
"step": 1905
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.44154511894699583,
|
| 2704 |
+
"grad_norm": 187.875,
|
| 2705 |
+
"learning_rate": 6.205035971223022e-06,
|
| 2706 |
+
"loss": 69.8792,
|
| 2707 |
+
"step": 1910
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 0.44270099622172615,
|
| 2711 |
+
"grad_norm": 149.625,
|
| 2712 |
+
"learning_rate": 6.192189105858171e-06,
|
| 2713 |
+
"loss": 72.0271,
|
| 2714 |
+
"step": 1915
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 0.4438568734964565,
|
| 2718 |
+
"grad_norm": 160.625,
|
| 2719 |
+
"learning_rate": 6.17934224049332e-06,
|
| 2720 |
+
"loss": 71.8988,
|
| 2721 |
+
"step": 1920
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 0.44501275077118685,
|
| 2725 |
+
"grad_norm": 186.875,
|
| 2726 |
+
"learning_rate": 6.1664953751284695e-06,
|
| 2727 |
+
"loss": 71.6762,
|
| 2728 |
+
"step": 1925
|
| 2729 |
+
},
|
| 2730 |
+
{
|
| 2731 |
+
"epoch": 0.4461686280459172,
|
| 2732 |
+
"grad_norm": 166.125,
|
| 2733 |
+
"learning_rate": 6.1536485097636175e-06,
|
| 2734 |
+
"loss": 70.4196,
|
| 2735 |
+
"step": 1930
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 0.4473245053206476,
|
| 2739 |
+
"grad_norm": 169.625,
|
| 2740 |
+
"learning_rate": 6.140801644398767e-06,
|
| 2741 |
+
"loss": 70.9007,
|
| 2742 |
+
"step": 1935
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 0.4484803825953779,
|
| 2746 |
+
"grad_norm": 166.25,
|
| 2747 |
+
"learning_rate": 6.127954779033917e-06,
|
| 2748 |
+
"loss": 69.9268,
|
| 2749 |
+
"step": 1940
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 0.4496362598701083,
|
| 2753 |
+
"grad_norm": 151.625,
|
| 2754 |
+
"learning_rate": 6.115107913669065e-06,
|
| 2755 |
+
"loss": 71.6101,
|
| 2756 |
+
"step": 1945
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 0.4507921371448386,
|
| 2760 |
+
"grad_norm": 169.125,
|
| 2761 |
+
"learning_rate": 6.102261048304215e-06,
|
| 2762 |
+
"loss": 72.2735,
|
| 2763 |
+
"step": 1950
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 0.451948014419569,
|
| 2767 |
+
"grad_norm": 161.375,
|
| 2768 |
+
"learning_rate": 6.0894141829393634e-06,
|
| 2769 |
+
"loss": 71.2896,
|
| 2770 |
+
"step": 1955
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 0.4531038916942994,
|
| 2774 |
+
"grad_norm": 174.875,
|
| 2775 |
+
"learning_rate": 6.076567317574512e-06,
|
| 2776 |
+
"loss": 71.3177,
|
| 2777 |
+
"step": 1960
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 0.4542597689690297,
|
| 2781 |
+
"grad_norm": 170.5,
|
| 2782 |
+
"learning_rate": 6.063720452209661e-06,
|
| 2783 |
+
"loss": 70.5251,
|
| 2784 |
+
"step": 1965
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 0.4554156462437601,
|
| 2788 |
+
"grad_norm": 156.875,
|
| 2789 |
+
"learning_rate": 6.05087358684481e-06,
|
| 2790 |
+
"loss": 70.9012,
|
| 2791 |
+
"step": 1970
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 0.45657152351849045,
|
| 2795 |
+
"grad_norm": 156.0,
|
| 2796 |
+
"learning_rate": 6.03802672147996e-06,
|
| 2797 |
+
"loss": 70.7252,
|
| 2798 |
+
"step": 1975
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 0.4577274007932208,
|
| 2802 |
+
"grad_norm": 152.75,
|
| 2803 |
+
"learning_rate": 6.0251798561151085e-06,
|
| 2804 |
+
"loss": 70.9783,
|
| 2805 |
+
"step": 1980
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 0.45888327806795115,
|
| 2809 |
+
"grad_norm": 159.5,
|
| 2810 |
+
"learning_rate": 6.012332990750257e-06,
|
| 2811 |
+
"loss": 70.2681,
|
| 2812 |
+
"step": 1985
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 0.46003915534268147,
|
| 2816 |
+
"grad_norm": 178.0,
|
| 2817 |
+
"learning_rate": 5.999486125385406e-06,
|
| 2818 |
+
"loss": 70.291,
|
| 2819 |
+
"step": 1990
|
| 2820 |
+
},
|
| 2821 |
+
{
|
| 2822 |
+
"epoch": 0.46119503261741185,
|
| 2823 |
+
"grad_norm": 176.875,
|
| 2824 |
+
"learning_rate": 5.986639260020556e-06,
|
| 2825 |
+
"loss": 70.9205,
|
| 2826 |
+
"step": 1995
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 0.4623509098921422,
|
| 2830 |
+
"grad_norm": 188.75,
|
| 2831 |
+
"learning_rate": 5.973792394655704e-06,
|
| 2832 |
+
"loss": 72.2977,
|
| 2833 |
+
"step": 2000
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"epoch": 0.4623509098921422,
|
| 2837 |
+
"eval_loss": NaN,
|
| 2838 |
+
"eval_runtime": 382.3957,
|
| 2839 |
+
"eval_samples_per_second": 609.667,
|
| 2840 |
+
"eval_steps_per_second": 38.105,
|
| 2841 |
+
"step": 2000
|
| 2842 |
}
|
| 2843 |
],
|
| 2844 |
"logging_steps": 5,
|
|
|
|
| 2858 |
"attributes": {}
|
| 2859 |
}
|
| 2860 |
},
|
| 2861 |
+
"total_flos": 8.664715985577574e+18,
|
| 2862 |
"train_batch_size": 4,
|
| 2863 |
"trial_name": null,
|
| 2864 |
"trial_params": null
|