Training in progress, step 20, 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 847599616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62ea57d95f189e14afef9705561103ee91a6e9ba01e5482796d19fc2459d71e2
|
| 3 |
size 847599616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 350603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b13a9b1047bb4b6f08d1c4cdf00a090f9620623f2a2765a0675250b296d77fd
|
| 3 |
size 350603
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -11,39 +11,39 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.025,
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
"learning_rate": 1.6e-06,
|
| 16 |
-
"loss": 21.
|
| 17 |
"step": 5
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.05,
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
"learning_rate": 3.6e-06,
|
| 23 |
-
"loss": 21.
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.075,
|
| 28 |
-
"grad_norm": 0.
|
| 29 |
"learning_rate": 3.995627254437549e-06,
|
| 30 |
-
"loss": 21.
|
| 31 |
"step": 15
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.1,
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
"learning_rate": 3.9778957412029366e-06,
|
| 37 |
-
"loss": 21.
|
| 38 |
"step": 20
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.1,
|
| 42 |
-
"eval_accuracy": 1.
|
| 43 |
-
"eval_loss": 10.
|
| 44 |
-
"eval_runtime": 91.
|
| 45 |
-
"eval_samples_per_second": 10.
|
| 46 |
-
"eval_steps_per_second": 1.
|
| 47 |
"step": 20
|
| 48 |
}
|
| 49 |
],
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.025,
|
| 14 |
+
"grad_norm": 0.1988103836774826,
|
| 15 |
"learning_rate": 1.6e-06,
|
| 16 |
+
"loss": 21.93964385986328,
|
| 17 |
"step": 5
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.05,
|
| 21 |
+
"grad_norm": 0.19530963897705078,
|
| 22 |
"learning_rate": 3.6e-06,
|
| 23 |
+
"loss": 21.938941955566406,
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.075,
|
| 28 |
+
"grad_norm": 0.20022961497306824,
|
| 29 |
"learning_rate": 3.995627254437549e-06,
|
| 30 |
+
"loss": 21.935198974609374,
|
| 31 |
"step": 15
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.1,
|
| 35 |
+
"grad_norm": 0.18916556239128113,
|
| 36 |
"learning_rate": 3.9778957412029366e-06,
|
| 37 |
+
"loss": 21.938011169433594,
|
| 38 |
"step": 20
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.1,
|
| 42 |
+
"eval_accuracy": 1.3431013431013431e-05,
|
| 43 |
+
"eval_loss": 10.96605396270752,
|
| 44 |
+
"eval_runtime": 91.788,
|
| 45 |
+
"eval_samples_per_second": 10.895,
|
| 46 |
+
"eval_steps_per_second": 1.819,
|
| 47 |
"step": 20
|
| 48 |
}
|
| 49 |
],
|