Training in progress, step 1200, checkpoint
Browse files
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 160285149
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9a748e4fabb1185fcc97bf0591137d28fac894d68e354057f9137276730d7892
|
3 |
size 160285149
|
last-checkpoint/rng_state.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14575
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f6ce383084f6b106e43851b732368fa1bb43173d4b082c8c2358258ae5094627
|
3 |
size 14575
|
last-checkpoint/scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 627
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:03baae644d6a5ff8182893827142cd7da004e6d4f167798346eedf849e69c4fa
|
3 |
size 627
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 1.
|
5 |
"eval_steps": 100,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -150,13 +150,26 @@
|
|
150 |
"eval_samples_per_second": 1.903,
|
151 |
"eval_steps_per_second": 0.239,
|
152 |
"step": 1100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
}
|
154 |
],
|
155 |
"logging_steps": 100,
|
156 |
"max_steps": 2025,
|
157 |
"num_train_epochs": 3,
|
158 |
"save_steps": 100,
|
159 |
-
"total_flos": 2.
|
160 |
"trial_name": null,
|
161 |
"trial_params": null
|
162 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.7777777777777777,
|
5 |
"eval_steps": 100,
|
6 |
+
"global_step": 1200,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
150 |
"eval_samples_per_second": 1.903,
|
151 |
"eval_steps_per_second": 0.239,
|
152 |
"step": 1100
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 1.78,
|
156 |
+
"learning_rate": 0.0,
|
157 |
+
"loss": 8.781,
|
158 |
+
"step": 1200
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 1.78,
|
162 |
+
"eval_runtime": 788.0952,
|
163 |
+
"eval_samples_per_second": 1.903,
|
164 |
+
"eval_steps_per_second": 0.239,
|
165 |
+
"step": 1200
|
166 |
}
|
167 |
],
|
168 |
"logging_steps": 100,
|
169 |
"max_steps": 2025,
|
170 |
"num_train_epochs": 3,
|
171 |
"save_steps": 100,
|
172 |
+
"total_flos": 2.49888478593024e+17,
|
173 |
"trial_name": null,
|
174 |
"trial_params": null
|
175 |
}
|