Upload 12 files
Browse files- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +2 -2
- scaler.pt +1 -1
- scheduler.pt +1 -1
- trainer_state.json +38 -35
- training_args.bin +1 -1
optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 2490594117
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:70209bf90ff6329bc9d3951ed7f6f3337707f480a0a4685bc64b6d32752ce81a
|
3 |
size 2490594117
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1262168365
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3d2724b1a2e8ade88b4044a5113a5fb703997cf65bb5f5db0d9613cf94c44859
|
3 |
size 1262168365
|
rng_state.pth
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 sha256:29508f8f7db7db5f8e88e735adfff7e49ffff6a6b57ff18d5d62dda777ff736d
|
3 |
+
size 14575
|
scaler.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 557
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:68cff80b680ddf6e7abbef98b5f336b97f9b5963e2209307f639383870e8cc71
|
3 |
size 557
|
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 sha256:95abbc43524e0f4952e011c561442dae9f3d8b40eb09aba46d0b62f4330ba829
|
3 |
size 627
|
trainer_state.json
CHANGED
@@ -1,67 +1,70 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch":
|
5 |
-
"global_step":
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
-
"epoch":
|
12 |
-
"learning_rate": 4.
|
13 |
-
"loss": 0.
|
14 |
"step": 150
|
15 |
},
|
16 |
{
|
17 |
-
"epoch":
|
18 |
-
"learning_rate": 4.
|
19 |
-
"loss": 0.
|
20 |
"step": 300
|
21 |
},
|
22 |
{
|
23 |
-
"epoch":
|
24 |
-
"learning_rate": 4.
|
25 |
-
"loss": 0.
|
26 |
"step": 450
|
27 |
},
|
28 |
{
|
29 |
-
"epoch": 6.
|
30 |
-
"
|
31 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
"step": 600
|
33 |
},
|
34 |
{
|
35 |
-
"epoch":
|
36 |
-
"learning_rate": 4.
|
37 |
-
"loss": 0.
|
38 |
"step": 750
|
39 |
},
|
40 |
{
|
41 |
-
"epoch":
|
42 |
-
"learning_rate": 4.
|
43 |
-
"loss": 0.
|
44 |
"step": 900
|
45 |
},
|
46 |
{
|
47 |
-
"epoch":
|
48 |
-
"
|
49 |
-
"
|
50 |
-
"
|
51 |
-
|
52 |
-
|
53 |
-
"
|
54 |
-
"eval_loss": 0.23020002245903015,
|
55 |
-
"eval_runtime": 433.719,
|
56 |
-
"eval_samples_per_second": 9.407,
|
57 |
-
"eval_steps_per_second": 1.176,
|
58 |
-
"eval_wer": 0.16092909535452324,
|
59 |
-
"step": 1100
|
60 |
}
|
61 |
],
|
62 |
-
"max_steps":
|
63 |
"num_train_epochs": 10000,
|
64 |
-
"total_flos":
|
65 |
"trial_name": null,
|
66 |
"trial_params": null
|
67 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 13.693877551020408,
|
5 |
+
"global_step": 1000,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
+
"epoch": 2.05,
|
12 |
+
"learning_rate": 4.999657487327031e-06,
|
13 |
+
"loss": 0.0321,
|
14 |
"step": 150
|
15 |
},
|
16 |
{
|
17 |
+
"epoch": 4.11,
|
18 |
+
"learning_rate": 4.998629949308124e-06,
|
19 |
+
"loss": 0.0265,
|
20 |
"step": 300
|
21 |
},
|
22 |
{
|
23 |
+
"epoch": 6.16,
|
24 |
+
"learning_rate": 4.997602411289218e-06,
|
25 |
+
"loss": 0.0244,
|
26 |
"step": 450
|
27 |
},
|
28 |
{
|
29 |
+
"epoch": 6.84,
|
30 |
+
"eval_loss": 0.19918830692768097,
|
31 |
+
"eval_runtime": 422.7179,
|
32 |
+
"eval_samples_per_second": 9.924,
|
33 |
+
"eval_steps_per_second": 1.242,
|
34 |
+
"eval_wer": 0.16014906363507714,
|
35 |
+
"step": 500
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"epoch": 8.22,
|
39 |
+
"learning_rate": 4.9965748732703115e-06,
|
40 |
+
"loss": 0.0242,
|
41 |
"step": 600
|
42 |
},
|
43 |
{
|
44 |
+
"epoch": 10.27,
|
45 |
+
"learning_rate": 4.995547335251405e-06,
|
46 |
+
"loss": 0.0228,
|
47 |
"step": 750
|
48 |
},
|
49 |
{
|
50 |
+
"epoch": 12.33,
|
51 |
+
"learning_rate": 4.994519797232499e-06,
|
52 |
+
"loss": 0.0228,
|
53 |
"step": 900
|
54 |
},
|
55 |
{
|
56 |
+
"epoch": 13.69,
|
57 |
+
"eval_loss": 0.2058412730693817,
|
58 |
+
"eval_runtime": 245.0508,
|
59 |
+
"eval_samples_per_second": 17.119,
|
60 |
+
"eval_steps_per_second": 2.142,
|
61 |
+
"eval_wer": 0.15878107457898957,
|
62 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
}
|
64 |
],
|
65 |
+
"max_steps": 730000,
|
66 |
"num_train_epochs": 10000,
|
67 |
+
"total_flos": 4.5875321808669204e+19,
|
68 |
"trial_name": null,
|
69 |
"trial_params": null
|
70 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3323
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:da61f32f17ff4a89629ecab29356f718e95133ce74a533e043ffb4e6d39fcc9a
|
3 |
size 3323
|