Training in progress, step 15000
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 1001681601
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d4c97ffaaa05461792e0e2229949f1e81684656529b17decb46dece12e85e31e
|
3 |
size 1001681601
|
last-checkpoint/pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 551185105
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:54783e67cf70c65d09958c26d62295c22bd36422c10c78b83bcc280e9065e915
|
3 |
size 551185105
|
last-checkpoint/scaler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 559
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9471fc05f097534af9b9399d150c8c30e9b19c816a337e78607a714f9bc6cb02
|
3 |
size 559
|
last-checkpoint/scheduler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 623
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:453fb9c8b00cf9c9b376c4e51da5a3ca306c7b7e2684cc56e9a48732e0e65bf5
|
3 |
size 623
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 0.
|
5 |
-
"global_step":
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
@@ -180,11 +180,17 @@
|
|
180 |
"learning_rate": 3.3525e-05,
|
181 |
"loss": 0.0353,
|
182 |
"step": 14500
|
|
|
|
|
|
|
|
|
|
|
|
|
183 |
}
|
184 |
],
|
185 |
"max_steps": 44000,
|
186 |
"num_train_epochs": 1,
|
187 |
-
"total_flos": 1.
|
188 |
"trial_name": null,
|
189 |
"trial_params": null
|
190 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.20347811931956916,
|
5 |
+
"global_step": 15000,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
|
|
180 |
"learning_rate": 3.3525e-05,
|
181 |
"loss": 0.0353,
|
182 |
"step": 14500
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"epoch": 0.2,
|
186 |
+
"learning_rate": 3.295681818181818e-05,
|
187 |
+
"loss": 0.034,
|
188 |
+
"step": 15000
|
189 |
}
|
190 |
],
|
191 |
"max_steps": 44000,
|
192 |
"num_train_epochs": 1,
|
193 |
+
"total_flos": 1.5674515497418752e+16,
|
194 |
"trial_name": null,
|
195 |
"trial_params": null
|
196 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 551185105
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:54783e67cf70c65d09958c26d62295c22bd36422c10c78b83bcc280e9065e915
|
3 |
size 551185105
|
runs/Oct19_12-48-30_40567d3bb726/events.out.tfevents.1666183950.40567d3bb726.23.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7af85d0a7b114e3c7a8348556983d65f1e68d84fa74a3c0e1a9c84606b6b2d46
|
3 |
+
size 50356
|