End of training
Browse files- trainer_state.json +1618 -0
trainer_state.json
ADDED
@@ -0,0 +1,1618 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.09492826159492826,
|
3 |
+
"best_model_checkpoint": "fine-w2v2base-bs16-ep100-lr2e-05-linguistic-rmsnorm-focal_ctc_a0.99_g1.0-0.05_10_0.004_40/checkpoint-4450",
|
4 |
+
"epoch": 100.0,
|
5 |
+
"eval_steps": 50,
|
6 |
+
"global_step": 5300,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.94,
|
13 |
+
"learning_rate": 7.169811320754717e-07,
|
14 |
+
"loss": 2181.8592,
|
15 |
+
"step": 50
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.94,
|
19 |
+
"eval_loss": 1087.6209716796875,
|
20 |
+
"eval_runtime": 3.509,
|
21 |
+
"eval_samples_per_second": 204.334,
|
22 |
+
"eval_steps_per_second": 6.555,
|
23 |
+
"eval_wer": 15.942025358692025,
|
24 |
+
"step": 50
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 1.89,
|
28 |
+
"learning_rate": 2.5660377358490568e-06,
|
29 |
+
"loss": 1908.6856,
|
30 |
+
"step": 100
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 1.89,
|
34 |
+
"eval_loss": 809.7703247070312,
|
35 |
+
"eval_runtime": 2.575,
|
36 |
+
"eval_samples_per_second": 278.445,
|
37 |
+
"eval_steps_per_second": 8.932,
|
38 |
+
"eval_wer": 15.877293960627293,
|
39 |
+
"step": 100
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 2.83,
|
43 |
+
"learning_rate": 4.452830188679246e-06,
|
44 |
+
"loss": 838.4017,
|
45 |
+
"step": 150
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 2.83,
|
49 |
+
"eval_loss": 112.64669036865234,
|
50 |
+
"eval_runtime": 2.3442,
|
51 |
+
"eval_samples_per_second": 305.859,
|
52 |
+
"eval_steps_per_second": 9.811,
|
53 |
+
"eval_wer": 0.9996663329996663,
|
54 |
+
"step": 150
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 3.77,
|
58 |
+
"learning_rate": 6.339622641509434e-06,
|
59 |
+
"loss": 117.7945,
|
60 |
+
"step": 200
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"epoch": 3.77,
|
64 |
+
"eval_loss": 85.67916107177734,
|
65 |
+
"eval_runtime": 2.4505,
|
66 |
+
"eval_samples_per_second": 292.59,
|
67 |
+
"eval_steps_per_second": 9.386,
|
68 |
+
"eval_wer": 1.0,
|
69 |
+
"step": 200
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 4.72,
|
73 |
+
"learning_rate": 8.226415094339623e-06,
|
74 |
+
"loss": 109.9946,
|
75 |
+
"step": 250
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 4.72,
|
79 |
+
"eval_loss": 82.57705688476562,
|
80 |
+
"eval_runtime": 2.3658,
|
81 |
+
"eval_samples_per_second": 303.071,
|
82 |
+
"eval_steps_per_second": 9.722,
|
83 |
+
"eval_wer": 1.0,
|
84 |
+
"step": 250
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 5.66,
|
88 |
+
"learning_rate": 1.0113207547169812e-05,
|
89 |
+
"loss": 105.7306,
|
90 |
+
"step": 300
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 5.66,
|
94 |
+
"eval_loss": 79.65999603271484,
|
95 |
+
"eval_runtime": 2.4085,
|
96 |
+
"eval_samples_per_second": 297.69,
|
97 |
+
"eval_steps_per_second": 9.549,
|
98 |
+
"eval_wer": 1.0,
|
99 |
+
"step": 300
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 6.6,
|
103 |
+
"learning_rate": 1.2e-05,
|
104 |
+
"loss": 102.0127,
|
105 |
+
"step": 350
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 6.6,
|
109 |
+
"eval_loss": 77.22874450683594,
|
110 |
+
"eval_runtime": 2.2958,
|
111 |
+
"eval_samples_per_second": 312.304,
|
112 |
+
"eval_steps_per_second": 10.018,
|
113 |
+
"eval_wer": 1.0,
|
114 |
+
"step": 350
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 7.55,
|
118 |
+
"learning_rate": 1.3886792452830189e-05,
|
119 |
+
"loss": 97.9428,
|
120 |
+
"step": 400
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 7.55,
|
124 |
+
"eval_loss": 75.43338012695312,
|
125 |
+
"eval_runtime": 2.2511,
|
126 |
+
"eval_samples_per_second": 318.513,
|
127 |
+
"eval_steps_per_second": 10.217,
|
128 |
+
"eval_wer": 1.0,
|
129 |
+
"step": 400
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 8.49,
|
133 |
+
"learning_rate": 1.577358490566038e-05,
|
134 |
+
"loss": 96.0055,
|
135 |
+
"step": 450
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 8.49,
|
139 |
+
"eval_loss": 74.68699645996094,
|
140 |
+
"eval_runtime": 2.3966,
|
141 |
+
"eval_samples_per_second": 299.18,
|
142 |
+
"eval_steps_per_second": 9.597,
|
143 |
+
"eval_wer": 1.0,
|
144 |
+
"step": 450
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 9.43,
|
148 |
+
"learning_rate": 1.766037735849057e-05,
|
149 |
+
"loss": 96.9376,
|
150 |
+
"step": 500
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 9.43,
|
154 |
+
"eval_loss": 74.24928283691406,
|
155 |
+
"eval_runtime": 2.3235,
|
156 |
+
"eval_samples_per_second": 308.593,
|
157 |
+
"eval_steps_per_second": 9.899,
|
158 |
+
"eval_wer": 1.0,
|
159 |
+
"step": 500
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 10.38,
|
163 |
+
"learning_rate": 1.9547169811320757e-05,
|
164 |
+
"loss": 95.6634,
|
165 |
+
"step": 550
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 10.38,
|
169 |
+
"eval_loss": 74.1340560913086,
|
170 |
+
"eval_runtime": 2.2186,
|
171 |
+
"eval_samples_per_second": 323.18,
|
172 |
+
"eval_steps_per_second": 10.367,
|
173 |
+
"eval_wer": 1.0,
|
174 |
+
"step": 550
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 11.32,
|
178 |
+
"learning_rate": 1.9996868319012422e-05,
|
179 |
+
"loss": 96.1578,
|
180 |
+
"step": 600
|
181 |
+
},
|
182 |
+
{
|
183 |
+
"epoch": 11.32,
|
184 |
+
"eval_loss": 74.90034484863281,
|
185 |
+
"eval_runtime": 2.3308,
|
186 |
+
"eval_samples_per_second": 307.622,
|
187 |
+
"eval_steps_per_second": 9.868,
|
188 |
+
"eval_wer": 1.0,
|
189 |
+
"step": 600
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 12.26,
|
193 |
+
"learning_rate": 1.9983208992285993e-05,
|
194 |
+
"loss": 92.5678,
|
195 |
+
"step": 650
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 12.26,
|
199 |
+
"eval_loss": 75.66032409667969,
|
200 |
+
"eval_runtime": 2.35,
|
201 |
+
"eval_samples_per_second": 305.104,
|
202 |
+
"eval_steps_per_second": 9.787,
|
203 |
+
"eval_wer": 1.0598098098098099,
|
204 |
+
"step": 650
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 13.21,
|
208 |
+
"learning_rate": 1.9958724515842856e-05,
|
209 |
+
"loss": 90.5927,
|
210 |
+
"step": 700
|
211 |
+
},
|
212 |
+
{
|
213 |
+
"epoch": 13.21,
|
214 |
+
"eval_loss": 73.4554672241211,
|
215 |
+
"eval_runtime": 2.3862,
|
216 |
+
"eval_samples_per_second": 300.482,
|
217 |
+
"eval_steps_per_second": 9.639,
|
218 |
+
"eval_wer": 1.0538872205538872,
|
219 |
+
"step": 700
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 14.15,
|
223 |
+
"learning_rate": 1.9923441439074434e-05,
|
224 |
+
"loss": 87.8965,
|
225 |
+
"step": 750
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 14.15,
|
229 |
+
"eval_loss": 72.41024017333984,
|
230 |
+
"eval_runtime": 2.3682,
|
231 |
+
"eval_samples_per_second": 302.757,
|
232 |
+
"eval_steps_per_second": 9.712,
|
233 |
+
"eval_wer": 0.9986653319986654,
|
234 |
+
"step": 750
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 15.09,
|
238 |
+
"learning_rate": 1.9877398020679796e-05,
|
239 |
+
"loss": 86.8467,
|
240 |
+
"step": 800
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 15.09,
|
244 |
+
"eval_loss": 69.77374267578125,
|
245 |
+
"eval_runtime": 2.3633,
|
246 |
+
"eval_samples_per_second": 303.392,
|
247 |
+
"eval_steps_per_second": 9.732,
|
248 |
+
"eval_wer": 0.9984150817484151,
|
249 |
+
"step": 800
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 16.04,
|
253 |
+
"learning_rate": 1.9820644187180354e-05,
|
254 |
+
"loss": 85.3381,
|
255 |
+
"step": 850
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 16.04,
|
259 |
+
"eval_loss": 67.84330749511719,
|
260 |
+
"eval_runtime": 2.282,
|
261 |
+
"eval_samples_per_second": 314.204,
|
262 |
+
"eval_steps_per_second": 10.079,
|
263 |
+
"eval_wer": 0.9717217217217218,
|
264 |
+
"step": 850
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 16.98,
|
268 |
+
"learning_rate": 1.975324147878278e-05,
|
269 |
+
"loss": 80.3298,
|
270 |
+
"step": 900
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 16.98,
|
274 |
+
"eval_loss": 52.40813064575195,
|
275 |
+
"eval_runtime": 2.2309,
|
276 |
+
"eval_samples_per_second": 321.398,
|
277 |
+
"eval_steps_per_second": 10.31,
|
278 |
+
"eval_wer": 0.8594427761094428,
|
279 |
+
"step": 900
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 17.92,
|
283 |
+
"learning_rate": 1.9675262982648757e-05,
|
284 |
+
"loss": 56.9494,
|
285 |
+
"step": 950
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 17.92,
|
289 |
+
"eval_loss": 25.267763137817383,
|
290 |
+
"eval_runtime": 2.3806,
|
291 |
+
"eval_samples_per_second": 301.181,
|
292 |
+
"eval_steps_per_second": 9.661,
|
293 |
+
"eval_wer": 0.3554387721054388,
|
294 |
+
"step": 950
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"epoch": 18.87,
|
298 |
+
"learning_rate": 1.958679325364396e-05,
|
299 |
+
"loss": 32.292,
|
300 |
+
"step": 1000
|
301 |
+
},
|
302 |
+
{
|
303 |
+
"epoch": 18.87,
|
304 |
+
"eval_loss": 14.86340618133545,
|
305 |
+
"eval_runtime": 2.2474,
|
306 |
+
"eval_samples_per_second": 319.039,
|
307 |
+
"eval_steps_per_second": 10.234,
|
308 |
+
"eval_wer": 0.21896896896896897,
|
309 |
+
"step": 1000
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 19.81,
|
313 |
+
"learning_rate": 1.9487928222652195e-05,
|
314 |
+
"loss": 22.3255,
|
315 |
+
"step": 1050
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 19.81,
|
319 |
+
"eval_loss": 11.289799690246582,
|
320 |
+
"eval_runtime": 2.4098,
|
321 |
+
"eval_samples_per_second": 297.529,
|
322 |
+
"eval_steps_per_second": 9.544,
|
323 |
+
"eval_wer": 0.182349015682349,
|
324 |
+
"step": 1050
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 20.75,
|
328 |
+
"learning_rate": 1.9378775092554124e-05,
|
329 |
+
"loss": 17.6187,
|
330 |
+
"step": 1100
|
331 |
+
},
|
332 |
+
{
|
333 |
+
"epoch": 20.75,
|
334 |
+
"eval_loss": 9.138720512390137,
|
335 |
+
"eval_runtime": 2.3047,
|
336 |
+
"eval_samples_per_second": 311.103,
|
337 |
+
"eval_steps_per_second": 9.98,
|
338 |
+
"eval_wer": 0.15340340340340342,
|
339 |
+
"step": 1100
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 21.7,
|
343 |
+
"learning_rate": 1.925945222198336e-05,
|
344 |
+
"loss": 15.1531,
|
345 |
+
"step": 1150
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 21.7,
|
349 |
+
"eval_loss": 7.663585662841797,
|
350 |
+
"eval_runtime": 2.3252,
|
351 |
+
"eval_samples_per_second": 308.358,
|
352 |
+
"eval_steps_per_second": 9.892,
|
353 |
+
"eval_wer": 0.13680347013680347,
|
354 |
+
"step": 1150
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 22.64,
|
358 |
+
"learning_rate": 1.9130088996985967e-05,
|
359 |
+
"loss": 13.1696,
|
360 |
+
"step": 1200
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 22.64,
|
364 |
+
"eval_loss": 7.029110908508301,
|
365 |
+
"eval_runtime": 2.4316,
|
366 |
+
"eval_samples_per_second": 294.871,
|
367 |
+
"eval_steps_per_second": 9.459,
|
368 |
+
"eval_wer": 0.14339339339339338,
|
369 |
+
"step": 1200
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 23.58,
|
373 |
+
"learning_rate": 1.8990825690722557e-05,
|
374 |
+
"loss": 11.9792,
|
375 |
+
"step": 1250
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 23.58,
|
379 |
+
"eval_loss": 6.686671257019043,
|
380 |
+
"eval_runtime": 2.3989,
|
381 |
+
"eval_samples_per_second": 298.889,
|
382 |
+
"eval_steps_per_second": 9.588,
|
383 |
+
"eval_wer": 0.1324657991324658,
|
384 |
+
"step": 1250
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"epoch": 24.53,
|
388 |
+
"learning_rate": 1.8841813311365105e-05,
|
389 |
+
"loss": 11.2404,
|
390 |
+
"step": 1300
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 24.53,
|
394 |
+
"eval_loss": 6.294769287109375,
|
395 |
+
"eval_runtime": 2.3601,
|
396 |
+
"eval_samples_per_second": 303.796,
|
397 |
+
"eval_steps_per_second": 9.745,
|
398 |
+
"eval_wer": 0.12128795462128796,
|
399 |
+
"step": 1300
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 25.47,
|
403 |
+
"learning_rate": 1.868321343835339e-05,
|
404 |
+
"loss": 10.6256,
|
405 |
+
"step": 1350
|
406 |
+
},
|
407 |
+
{
|
408 |
+
"epoch": 25.47,
|
409 |
+
"eval_loss": 5.715055465698242,
|
410 |
+
"eval_runtime": 2.2982,
|
411 |
+
"eval_samples_per_second": 311.988,
|
412 |
+
"eval_steps_per_second": 10.008,
|
413 |
+
"eval_wer": 0.1180347013680347,
|
414 |
+
"step": 1350
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 26.42,
|
418 |
+
"learning_rate": 1.8515198047188652e-05,
|
419 |
+
"loss": 9.452,
|
420 |
+
"step": 1400
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"epoch": 26.42,
|
424 |
+
"eval_loss": 5.419599533081055,
|
425 |
+
"eval_runtime": 2.4076,
|
426 |
+
"eval_samples_per_second": 297.802,
|
427 |
+
"eval_steps_per_second": 9.553,
|
428 |
+
"eval_wer": 0.1175342008675342,
|
429 |
+
"step": 1400
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 27.36,
|
433 |
+
"learning_rate": 1.833794932295441e-05,
|
434 |
+
"loss": 9.3087,
|
435 |
+
"step": 1450
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 27.36,
|
439 |
+
"eval_loss": 5.292884826660156,
|
440 |
+
"eval_runtime": 2.3017,
|
441 |
+
"eval_samples_per_second": 311.506,
|
442 |
+
"eval_steps_per_second": 9.993,
|
443 |
+
"eval_wer": 0.11244577911244578,
|
444 |
+
"step": 1450
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 28.3,
|
448 |
+
"learning_rate": 1.8151659462766685e-05,
|
449 |
+
"loss": 8.5149,
|
450 |
+
"step": 1500
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 28.3,
|
454 |
+
"eval_loss": 5.13940954208374,
|
455 |
+
"eval_runtime": 2.4233,
|
456 |
+
"eval_samples_per_second": 295.882,
|
457 |
+
"eval_steps_per_second": 9.491,
|
458 |
+
"eval_wer": 0.11628294961628295,
|
459 |
+
"step": 1500
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 29.25,
|
463 |
+
"learning_rate": 1.7956530467367805e-05,
|
464 |
+
"loss": 8.3662,
|
465 |
+
"step": 1550
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 29.25,
|
469 |
+
"eval_loss": 5.127543926239014,
|
470 |
+
"eval_runtime": 2.3508,
|
471 |
+
"eval_samples_per_second": 304.996,
|
472 |
+
"eval_steps_per_second": 9.784,
|
473 |
+
"eval_wer": 0.12128795462128796,
|
474 |
+
"step": 1550
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 30.19,
|
478 |
+
"learning_rate": 1.7752773922089784e-05,
|
479 |
+
"loss": 7.8852,
|
480 |
+
"step": 1600
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"epoch": 30.19,
|
484 |
+
"eval_loss": 4.903261184692383,
|
485 |
+
"eval_runtime": 2.4091,
|
486 |
+
"eval_samples_per_second": 297.622,
|
487 |
+
"eval_steps_per_second": 9.547,
|
488 |
+
"eval_wer": 0.10927594260927594,
|
489 |
+
"step": 1600
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 31.13,
|
493 |
+
"learning_rate": 1.7540610767424813e-05,
|
494 |
+
"loss": 7.5135,
|
495 |
+
"step": 1650
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 31.13,
|
499 |
+
"eval_loss": 4.957184314727783,
|
500 |
+
"eval_runtime": 2.4012,
|
501 |
+
"eval_samples_per_second": 298.606,
|
502 |
+
"eval_steps_per_second": 9.579,
|
503 |
+
"eval_wer": 0.10969302635969302,
|
504 |
+
"step": 1650
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 32.08,
|
508 |
+
"learning_rate": 1.7320271059451597e-05,
|
509 |
+
"loss": 7.5374,
|
510 |
+
"step": 1700
|
511 |
+
},
|
512 |
+
{
|
513 |
+
"epoch": 32.08,
|
514 |
+
"eval_loss": 4.758788108825684,
|
515 |
+
"eval_runtime": 2.4229,
|
516 |
+
"eval_samples_per_second": 295.923,
|
517 |
+
"eval_steps_per_second": 9.493,
|
518 |
+
"eval_wer": 0.1016016016016016,
|
519 |
+
"step": 1700
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 33.02,
|
523 |
+
"learning_rate": 1.7091993720377336e-05,
|
524 |
+
"loss": 7.2968,
|
525 |
+
"step": 1750
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 33.02,
|
529 |
+
"eval_loss": 4.7317328453063965,
|
530 |
+
"eval_runtime": 2.2458,
|
531 |
+
"eval_samples_per_second": 319.258,
|
532 |
+
"eval_steps_per_second": 10.241,
|
533 |
+
"eval_wer": 0.10326993660326994,
|
534 |
+
"step": 1750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 33.96,
|
538 |
+
"learning_rate": 1.685602627946584e-05,
|
539 |
+
"loss": 7.0861,
|
540 |
+
"step": 1800
|
541 |
+
},
|
542 |
+
{
|
543 |
+
"epoch": 33.96,
|
544 |
+
"eval_loss": 4.791558265686035,
|
545 |
+
"eval_runtime": 2.2472,
|
546 |
+
"eval_samples_per_second": 319.068,
|
547 |
+
"eval_steps_per_second": 10.235,
|
548 |
+
"eval_wer": 0.10869202535869203,
|
549 |
+
"step": 1800
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 34.91,
|
553 |
+
"learning_rate": 1.661262460463274e-05,
|
554 |
+
"loss": 6.6371,
|
555 |
+
"step": 1850
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 34.91,
|
559 |
+
"eval_loss": 4.7941060066223145,
|
560 |
+
"eval_runtime": 2.2988,
|
561 |
+
"eval_samples_per_second": 311.908,
|
562 |
+
"eval_steps_per_second": 10.005,
|
563 |
+
"eval_wer": 0.11319652986319653,
|
564 |
+
"step": 1850
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 35.85,
|
568 |
+
"learning_rate": 1.6362052624998767e-05,
|
569 |
+
"loss": 6.6186,
|
570 |
+
"step": 1900
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 35.85,
|
574 |
+
"eval_loss": 4.660822868347168,
|
575 |
+
"eval_runtime": 2.4244,
|
576 |
+
"eval_samples_per_second": 295.746,
|
577 |
+
"eval_steps_per_second": 9.487,
|
578 |
+
"eval_wer": 0.1036036036036036,
|
579 |
+
"step": 1900
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 36.79,
|
583 |
+
"learning_rate": 1.6104582044701983e-05,
|
584 |
+
"loss": 6.6288,
|
585 |
+
"step": 1950
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 36.79,
|
589 |
+
"eval_loss": 4.679046154022217,
|
590 |
+
"eval_runtime": 2.2429,
|
591 |
+
"eval_samples_per_second": 319.68,
|
592 |
+
"eval_steps_per_second": 10.255,
|
593 |
+
"eval_wer": 0.10744077410744077,
|
594 |
+
"step": 1950
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"epoch": 37.74,
|
598 |
+
"learning_rate": 1.584049204827929e-05,
|
599 |
+
"loss": 6.2433,
|
600 |
+
"step": 2000
|
601 |
+
},
|
602 |
+
{
|
603 |
+
"epoch": 37.74,
|
604 |
+
"eval_loss": 4.77145528793335,
|
605 |
+
"eval_runtime": 2.2791,
|
606 |
+
"eval_samples_per_second": 314.598,
|
607 |
+
"eval_steps_per_second": 10.092,
|
608 |
+
"eval_wer": 0.11211211211211211,
|
609 |
+
"step": 2000
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 38.68,
|
613 |
+
"learning_rate": 1.5570068997936686e-05,
|
614 |
+
"loss": 6.2362,
|
615 |
+
"step": 2050
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 38.68,
|
619 |
+
"eval_loss": 4.6420440673828125,
|
620 |
+
"eval_runtime": 2.4168,
|
621 |
+
"eval_samples_per_second": 296.676,
|
622 |
+
"eval_steps_per_second": 9.517,
|
623 |
+
"eval_wer": 0.10343677010343677,
|
624 |
+
"step": 2050
|
625 |
+
},
|
626 |
+
{
|
627 |
+
"epoch": 39.62,
|
628 |
+
"learning_rate": 1.5293606123036508e-05,
|
629 |
+
"loss": 5.957,
|
630 |
+
"step": 2100
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"epoch": 39.62,
|
634 |
+
"eval_loss": 4.575562953948975,
|
635 |
+
"eval_runtime": 2.4332,
|
636 |
+
"eval_samples_per_second": 294.67,
|
637 |
+
"eval_steps_per_second": 9.452,
|
638 |
+
"eval_wer": 0.10702369035702369,
|
639 |
+
"step": 2100
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 40.57,
|
643 |
+
"learning_rate": 1.5011403202138346e-05,
|
644 |
+
"loss": 5.8034,
|
645 |
+
"step": 2150
|
646 |
+
},
|
647 |
+
{
|
648 |
+
"epoch": 40.57,
|
649 |
+
"eval_loss": 4.411165714263916,
|
650 |
+
"eval_runtime": 2.3583,
|
651 |
+
"eval_samples_per_second": 304.027,
|
652 |
+
"eval_steps_per_second": 9.753,
|
653 |
+
"eval_wer": 0.10602268935602269,
|
654 |
+
"step": 2150
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 41.51,
|
658 |
+
"learning_rate": 1.4723766237938495e-05,
|
659 |
+
"loss": 5.4943,
|
660 |
+
"step": 2200
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 41.51,
|
664 |
+
"eval_loss": 4.563217639923096,
|
665 |
+
"eval_runtime": 2.4228,
|
666 |
+
"eval_samples_per_second": 295.938,
|
667 |
+
"eval_steps_per_second": 9.493,
|
668 |
+
"eval_wer": 0.10335335335335336,
|
669 |
+
"step": 2200
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 42.45,
|
673 |
+
"learning_rate": 1.4431007125460274e-05,
|
674 |
+
"loss": 5.5593,
|
675 |
+
"step": 2250
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 42.45,
|
679 |
+
"eval_loss": 4.537557601928711,
|
680 |
+
"eval_runtime": 2.4026,
|
681 |
+
"eval_samples_per_second": 298.421,
|
682 |
+
"eval_steps_per_second": 9.573,
|
683 |
+
"eval_wer": 0.11052719386052719,
|
684 |
+
"step": 2250
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"epoch": 43.4,
|
688 |
+
"learning_rate": 1.4133443313855155e-05,
|
689 |
+
"loss": 5.3447,
|
690 |
+
"step": 2300
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 43.4,
|
694 |
+
"eval_loss": 4.54231071472168,
|
695 |
+
"eval_runtime": 2.2971,
|
696 |
+
"eval_samples_per_second": 312.127,
|
697 |
+
"eval_steps_per_second": 10.012,
|
698 |
+
"eval_wer": 0.1006006006006006,
|
699 |
+
"step": 2300
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 44.34,
|
703 |
+
"learning_rate": 1.3831397462181298e-05,
|
704 |
+
"loss": 5.4181,
|
705 |
+
"step": 2350
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"epoch": 44.34,
|
709 |
+
"eval_loss": 4.378854274749756,
|
710 |
+
"eval_runtime": 2.3724,
|
711 |
+
"eval_samples_per_second": 302.223,
|
712 |
+
"eval_steps_per_second": 9.695,
|
713 |
+
"eval_wer": 0.09926593259926593,
|
714 |
+
"step": 2350
|
715 |
+
},
|
716 |
+
{
|
717 |
+
"epoch": 45.28,
|
718 |
+
"learning_rate": 1.3525197089532833e-05,
|
719 |
+
"loss": 5.222,
|
720 |
+
"step": 2400
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 45.28,
|
724 |
+
"eval_loss": 4.369490146636963,
|
725 |
+
"eval_runtime": 2.2903,
|
726 |
+
"eval_samples_per_second": 313.058,
|
727 |
+
"eval_steps_per_second": 10.042,
|
728 |
+
"eval_wer": 0.1031031031031031,
|
729 |
+
"step": 2400
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 46.23,
|
733 |
+
"learning_rate": 1.3215174219899224e-05,
|
734 |
+
"loss": 5.1146,
|
735 |
+
"step": 2450
|
736 |
+
},
|
737 |
+
{
|
738 |
+
"epoch": 46.23,
|
739 |
+
"eval_loss": 4.410806179046631,
|
740 |
+
"eval_runtime": 2.3034,
|
741 |
+
"eval_samples_per_second": 311.283,
|
742 |
+
"eval_steps_per_second": 9.985,
|
743 |
+
"eval_wer": 0.10844177510844177,
|
744 |
+
"step": 2450
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 47.17,
|
748 |
+
"learning_rate": 1.2901665022139796e-05,
|
749 |
+
"loss": 5.0952,
|
750 |
+
"step": 2500
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"epoch": 47.17,
|
754 |
+
"eval_loss": 4.295694828033447,
|
755 |
+
"eval_runtime": 2.2425,
|
756 |
+
"eval_samples_per_second": 319.732,
|
757 |
+
"eval_steps_per_second": 10.256,
|
758 |
+
"eval_wer": 0.1016016016016016,
|
759 |
+
"step": 2500
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 48.11,
|
763 |
+
"learning_rate": 1.2585009445463867e-05,
|
764 |
+
"loss": 4.9023,
|
765 |
+
"step": 2550
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 48.11,
|
769 |
+
"eval_loss": 4.3768768310546875,
|
770 |
+
"eval_runtime": 2.4393,
|
771 |
+
"eval_samples_per_second": 293.942,
|
772 |
+
"eval_steps_per_second": 9.429,
|
773 |
+
"eval_wer": 0.1021021021021021,
|
774 |
+
"step": 2550
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 49.06,
|
778 |
+
"learning_rate": 1.2265550850811663e-05,
|
779 |
+
"loss": 5.1633,
|
780 |
+
"step": 2600
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 49.06,
|
784 |
+
"eval_loss": 4.36325216293335,
|
785 |
+
"eval_runtime": 2.3579,
|
786 |
+
"eval_samples_per_second": 304.084,
|
787 |
+
"eval_steps_per_second": 9.754,
|
788 |
+
"eval_wer": 0.10627293960627295,
|
789 |
+
"step": 2600
|
790 |
+
},
|
791 |
+
{
|
792 |
+
"epoch": 50.0,
|
793 |
+
"learning_rate": 1.1943635638535827e-05,
|
794 |
+
"loss": 4.9489,
|
795 |
+
"step": 2650
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 50.0,
|
799 |
+
"eval_loss": 4.3422441482543945,
|
800 |
+
"eval_runtime": 2.2967,
|
801 |
+
"eval_samples_per_second": 312.186,
|
802 |
+
"eval_steps_per_second": 10.014,
|
803 |
+
"eval_wer": 0.10452118785452119,
|
804 |
+
"step": 2650
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 50.94,
|
808 |
+
"learning_rate": 1.1619612872787144e-05,
|
809 |
+
"loss": 4.7391,
|
810 |
+
"step": 2700
|
811 |
+
},
|
812 |
+
{
|
813 |
+
"epoch": 50.94,
|
814 |
+
"eval_loss": 4.251036643981934,
|
815 |
+
"eval_runtime": 2.3762,
|
816 |
+
"eval_samples_per_second": 301.737,
|
817 |
+
"eval_steps_per_second": 9.679,
|
818 |
+
"eval_wer": 0.10293626960293627,
|
819 |
+
"step": 2700
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 51.89,
|
823 |
+
"learning_rate": 1.1293833903011819e-05,
|
824 |
+
"loss": 4.7996,
|
825 |
+
"step": 2750
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 51.89,
|
829 |
+
"eval_loss": 4.3253912925720215,
|
830 |
+
"eval_runtime": 2.3965,
|
831 |
+
"eval_samples_per_second": 299.183,
|
832 |
+
"eval_steps_per_second": 9.597,
|
833 |
+
"eval_wer": 0.10118451785118451,
|
834 |
+
"step": 2750
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"epoch": 52.83,
|
838 |
+
"learning_rate": 1.0966651982970757e-05,
|
839 |
+
"loss": 4.244,
|
840 |
+
"step": 2800
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"epoch": 52.83,
|
844 |
+
"eval_loss": 4.41210412979126,
|
845 |
+
"eval_runtime": 2.392,
|
846 |
+
"eval_samples_per_second": 299.749,
|
847 |
+
"eval_steps_per_second": 9.615,
|
848 |
+
"eval_wer": 0.10352018685352019,
|
849 |
+
"step": 2800
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 53.77,
|
853 |
+
"learning_rate": 1.0638421887693887e-05,
|
854 |
+
"loss": 4.5831,
|
855 |
+
"step": 2850
|
856 |
+
},
|
857 |
+
{
|
858 |
+
"epoch": 53.77,
|
859 |
+
"eval_loss": 4.405577182769775,
|
860 |
+
"eval_runtime": 2.3528,
|
861 |
+
"eval_samples_per_second": 304.748,
|
862 |
+
"eval_steps_per_second": 9.776,
|
863 |
+
"eval_wer": 0.10443777110443778,
|
864 |
+
"step": 2850
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 54.72,
|
868 |
+
"learning_rate": 1.0309499528784948e-05,
|
869 |
+
"loss": 4.5198,
|
870 |
+
"step": 2900
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 54.72,
|
874 |
+
"eval_loss": 4.363803386688232,
|
875 |
+
"eval_runtime": 2.3305,
|
876 |
+
"eval_samples_per_second": 307.658,
|
877 |
+
"eval_steps_per_second": 9.869,
|
878 |
+
"eval_wer": 0.1050216883550217,
|
879 |
+
"step": 2900
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 55.66,
|
883 |
+
"learning_rate": 9.980241568493834e-06,
|
884 |
+
"loss": 4.1964,
|
885 |
+
"step": 2950
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 55.66,
|
889 |
+
"eval_loss": 4.339745044708252,
|
890 |
+
"eval_runtime": 2.3557,
|
891 |
+
"eval_samples_per_second": 304.369,
|
892 |
+
"eval_steps_per_second": 9.764,
|
893 |
+
"eval_wer": 0.10710710710710711,
|
894 |
+
"step": 2950
|
895 |
+
},
|
896 |
+
{
|
897 |
+
"epoch": 56.6,
|
898 |
+
"learning_rate": 9.651005032974994e-06,
|
899 |
+
"loss": 4.0544,
|
900 |
+
"step": 3000
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 56.6,
|
904 |
+
"eval_loss": 4.349318981170654,
|
905 |
+
"eval_runtime": 2.4132,
|
906 |
+
"eval_samples_per_second": 297.115,
|
907 |
+
"eval_steps_per_second": 9.531,
|
908 |
+
"eval_wer": 0.1031031031031031,
|
909 |
+
"step": 3000
|
910 |
+
},
|
911 |
+
{
|
912 |
+
"epoch": 57.55,
|
913 |
+
"learning_rate": 9.322146925151226e-06,
|
914 |
+
"loss": 4.3568,
|
915 |
+
"step": 3050
|
916 |
+
},
|
917 |
+
{
|
918 |
+
"epoch": 57.55,
|
919 |
+
"eval_loss": 4.472135066986084,
|
920 |
+
"eval_runtime": 2.3982,
|
921 |
+
"eval_samples_per_second": 298.97,
|
922 |
+
"eval_steps_per_second": 9.59,
|
923 |
+
"eval_wer": 0.10593927260593927,
|
924 |
+
"step": 3050
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 58.49,
|
928 |
+
"learning_rate": 8.994023837602694e-06,
|
929 |
+
"loss": 4.2692,
|
930 |
+
"step": 3100
|
931 |
+
},
|
932 |
+
{
|
933 |
+
"epoch": 58.49,
|
934 |
+
"eval_loss": 4.427754878997803,
|
935 |
+
"eval_runtime": 2.4471,
|
936 |
+
"eval_samples_per_second": 292.994,
|
937 |
+
"eval_steps_per_second": 9.399,
|
938 |
+
"eval_wer": 0.11169502836169502,
|
939 |
+
"step": 3100
|
940 |
+
},
|
941 |
+
{
|
942 |
+
"epoch": 59.43,
|
943 |
+
"learning_rate": 8.666991565900827e-06,
|
944 |
+
"loss": 4.1226,
|
945 |
+
"step": 3150
|
946 |
+
},
|
947 |
+
{
|
948 |
+
"epoch": 59.43,
|
949 |
+
"eval_loss": 4.308145523071289,
|
950 |
+
"eval_runtime": 2.3553,
|
951 |
+
"eval_samples_per_second": 304.422,
|
952 |
+
"eval_steps_per_second": 9.765,
|
953 |
+
"eval_wer": 0.10035035035035035,
|
954 |
+
"step": 3150
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 60.38,
|
958 |
+
"learning_rate": 8.341404722806525e-06,
|
959 |
+
"loss": 4.2681,
|
960 |
+
"step": 3200
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 60.38,
|
964 |
+
"eval_loss": 4.417555332183838,
|
965 |
+
"eval_runtime": 2.2916,
|
966 |
+
"eval_samples_per_second": 312.876,
|
967 |
+
"eval_steps_per_second": 10.036,
|
968 |
+
"eval_wer": 0.10585585585585586,
|
969 |
+
"step": 3200
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 61.32,
|
973 |
+
"learning_rate": 8.017616353750874e-06,
|
974 |
+
"loss": 3.8412,
|
975 |
+
"step": 3250
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 61.32,
|
979 |
+
"eval_loss": 4.321342945098877,
|
980 |
+
"eval_runtime": 2.3267,
|
981 |
+
"eval_samples_per_second": 308.163,
|
982 |
+
"eval_steps_per_second": 9.885,
|
983 |
+
"eval_wer": 0.10276943610276944,
|
984 |
+
"step": 3250
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 62.26,
|
988 |
+
"learning_rate": 7.695977554015387e-06,
|
989 |
+
"loss": 4.1387,
|
990 |
+
"step": 3300
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 62.26,
|
994 |
+
"eval_loss": 4.341909408569336,
|
995 |
+
"eval_runtime": 2.2567,
|
996 |
+
"eval_samples_per_second": 317.721,
|
997 |
+
"eval_steps_per_second": 10.192,
|
998 |
+
"eval_wer": 0.10560560560560561,
|
999 |
+
"step": 3300
|
1000 |
+
},
|
1001 |
+
{
|
1002 |
+
"epoch": 63.21,
|
1003 |
+
"learning_rate": 7.376837088026863e-06,
|
1004 |
+
"loss": 3.6847,
|
1005 |
+
"step": 3350
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 63.21,
|
1009 |
+
"eval_loss": 4.249768257141113,
|
1010 |
+
"eval_runtime": 2.4172,
|
1011 |
+
"eval_samples_per_second": 296.628,
|
1012 |
+
"eval_steps_per_second": 9.515,
|
1013 |
+
"eval_wer": 0.10652318985652319,
|
1014 |
+
"step": 3350
|
1015 |
+
},
|
1016 |
+
{
|
1017 |
+
"epoch": 64.15,
|
1018 |
+
"learning_rate": 7.0605410111796855e-06,
|
1019 |
+
"loss": 3.8768,
|
1020 |
+
"step": 3400
|
1021 |
+
},
|
1022 |
+
{
|
1023 |
+
"epoch": 64.15,
|
1024 |
+
"eval_loss": 4.277639389038086,
|
1025 |
+
"eval_runtime": 2.4245,
|
1026 |
+
"eval_samples_per_second": 295.733,
|
1027 |
+
"eval_steps_per_second": 9.487,
|
1028 |
+
"eval_wer": 0.10276943610276944,
|
1029 |
+
"step": 3400
|
1030 |
+
},
|
1031 |
+
{
|
1032 |
+
"epoch": 65.09,
|
1033 |
+
"learning_rate": 6.7536610307503735e-06,
|
1034 |
+
"loss": 3.659,
|
1035 |
+
"step": 3450
|
1036 |
+
},
|
1037 |
+
{
|
1038 |
+
"epoch": 65.09,
|
1039 |
+
"eval_loss": 4.298828125,
|
1040 |
+
"eval_runtime": 2.4088,
|
1041 |
+
"eval_samples_per_second": 297.656,
|
1042 |
+
"eval_steps_per_second": 9.548,
|
1043 |
+
"eval_wer": 0.10076743410076744,
|
1044 |
+
"step": 3450
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 66.04,
|
1048 |
+
"learning_rate": 6.444005348764207e-06,
|
1049 |
+
"loss": 3.809,
|
1050 |
+
"step": 3500
|
1051 |
+
},
|
1052 |
+
{
|
1053 |
+
"epoch": 66.04,
|
1054 |
+
"eval_loss": 4.304114818572998,
|
1055 |
+
"eval_runtime": 2.3018,
|
1056 |
+
"eval_samples_per_second": 311.5,
|
1057 |
+
"eval_steps_per_second": 9.992,
|
1058 |
+
"eval_wer": 0.10343677010343677,
|
1059 |
+
"step": 3500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 66.98,
|
1063 |
+
"learning_rate": 6.138205558658212e-06,
|
1064 |
+
"loss": 3.7459,
|
1065 |
+
"step": 3550
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 66.98,
|
1069 |
+
"eval_loss": 4.295498847961426,
|
1070 |
+
"eval_runtime": 2.3818,
|
1071 |
+
"eval_samples_per_second": 301.033,
|
1072 |
+
"eval_steps_per_second": 9.657,
|
1073 |
+
"eval_wer": 0.09951618284951619,
|
1074 |
+
"step": 3550
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 67.92,
|
1078 |
+
"learning_rate": 5.83659325005591e-06,
|
1079 |
+
"loss": 3.7996,
|
1080 |
+
"step": 3600
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 67.92,
|
1084 |
+
"eval_loss": 4.284261703491211,
|
1085 |
+
"eval_runtime": 2.241,
|
1086 |
+
"eval_samples_per_second": 319.944,
|
1087 |
+
"eval_steps_per_second": 10.263,
|
1088 |
+
"eval_wer": 0.09926593259926593,
|
1089 |
+
"step": 3600
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 68.87,
|
1093 |
+
"learning_rate": 5.53949547194521e-06,
|
1094 |
+
"loss": 3.6773,
|
1095 |
+
"step": 3650
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 68.87,
|
1099 |
+
"eval_loss": 4.239638328552246,
|
1100 |
+
"eval_runtime": 2.3554,
|
1101 |
+
"eval_samples_per_second": 304.413,
|
1102 |
+
"eval_steps_per_second": 9.765,
|
1103 |
+
"eval_wer": 0.09876543209876543,
|
1104 |
+
"step": 3650
|
1105 |
+
},
|
1106 |
+
{
|
1107 |
+
"epoch": 69.81,
|
1108 |
+
"learning_rate": 5.247234378047524e-06,
|
1109 |
+
"loss": 3.6364,
|
1110 |
+
"step": 3700
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"epoch": 69.81,
|
1114 |
+
"eval_loss": 4.220588207244873,
|
1115 |
+
"eval_runtime": 2.4135,
|
1116 |
+
"eval_samples_per_second": 297.083,
|
1117 |
+
"eval_steps_per_second": 9.53,
|
1118 |
+
"eval_wer": 0.09634634634634634,
|
1119 |
+
"step": 3700
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 70.75,
|
1123 |
+
"learning_rate": 4.960126877495005e-06,
|
1124 |
+
"loss": 3.6342,
|
1125 |
+
"step": 3750
|
1126 |
+
},
|
1127 |
+
{
|
1128 |
+
"epoch": 70.75,
|
1129 |
+
"eval_loss": 4.290452480316162,
|
1130 |
+
"eval_runtime": 2.3553,
|
1131 |
+
"eval_samples_per_second": 304.417,
|
1132 |
+
"eval_steps_per_second": 9.765,
|
1133 |
+
"eval_wer": 0.10176843510176843,
|
1134 |
+
"step": 3750
|
1135 |
+
},
|
1136 |
+
{
|
1137 |
+
"epoch": 71.7,
|
1138 |
+
"learning_rate": 4.67848429119466e-06,
|
1139 |
+
"loss": 3.7012,
|
1140 |
+
"step": 3800
|
1141 |
+
},
|
1142 |
+
{
|
1143 |
+
"epoch": 71.7,
|
1144 |
+
"eval_loss": 4.308350086212158,
|
1145 |
+
"eval_runtime": 2.3492,
|
1146 |
+
"eval_samples_per_second": 305.208,
|
1147 |
+
"eval_steps_per_second": 9.79,
|
1148 |
+
"eval_wer": 0.09943276609943276,
|
1149 |
+
"step": 3800
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 72.64,
|
1153 |
+
"learning_rate": 4.402612014251967e-06,
|
1154 |
+
"loss": 3.4846,
|
1155 |
+
"step": 3850
|
1156 |
+
},
|
1157 |
+
{
|
1158 |
+
"epoch": 72.64,
|
1159 |
+
"eval_loss": 4.287242412567139,
|
1160 |
+
"eval_runtime": 2.3668,
|
1161 |
+
"eval_samples_per_second": 302.942,
|
1162 |
+
"eval_steps_per_second": 9.718,
|
1163 |
+
"eval_wer": 0.09759759759759759,
|
1164 |
+
"step": 3850
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 73.58,
|
1168 |
+
"learning_rate": 4.132809184820095e-06,
|
1169 |
+
"loss": 3.4814,
|
1170 |
+
"step": 3900
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"epoch": 73.58,
|
1174 |
+
"eval_loss": 4.259552955627441,
|
1175 |
+
"eval_runtime": 2.2432,
|
1176 |
+
"eval_samples_per_second": 319.627,
|
1177 |
+
"eval_steps_per_second": 10.253,
|
1178 |
+
"eval_wer": 0.10026693360026694,
|
1179 |
+
"step": 3900
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 74.53,
|
1183 |
+
"learning_rate": 3.869368359733711e-06,
|
1184 |
+
"loss": 3.3212,
|
1185 |
+
"step": 3950
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 74.53,
|
1189 |
+
"eval_loss": 4.226958274841309,
|
1190 |
+
"eval_runtime": 2.3893,
|
1191 |
+
"eval_samples_per_second": 300.093,
|
1192 |
+
"eval_steps_per_second": 9.626,
|
1193 |
+
"eval_wer": 0.09642976309642977,
|
1194 |
+
"step": 3950
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"epoch": 75.47,
|
1198 |
+
"learning_rate": 3.6125751972791635e-06,
|
1199 |
+
"loss": 3.6578,
|
1200 |
+
"step": 4000
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 75.47,
|
1204 |
+
"eval_loss": 4.247718811035156,
|
1205 |
+
"eval_runtime": 2.4075,
|
1206 |
+
"eval_samples_per_second": 297.814,
|
1207 |
+
"eval_steps_per_second": 9.553,
|
1208 |
+
"eval_wer": 0.09776443109776443,
|
1209 |
+
"step": 4000
|
1210 |
+
},
|
1211 |
+
{
|
1212 |
+
"epoch": 76.42,
|
1213 |
+
"learning_rate": 3.3627081474450273e-06,
|
1214 |
+
"loss": 3.4573,
|
1215 |
+
"step": 4050
|
1216 |
+
},
|
1217 |
+
{
|
1218 |
+
"epoch": 76.42,
|
1219 |
+
"eval_loss": 4.238850116729736,
|
1220 |
+
"eval_runtime": 2.3207,
|
1221 |
+
"eval_samples_per_second": 308.956,
|
1222 |
+
"eval_steps_per_second": 9.911,
|
1223 |
+
"eval_wer": 0.09726393059726393,
|
1224 |
+
"step": 4050
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 77.36,
|
1228 |
+
"learning_rate": 3.120038149988832e-06,
|
1229 |
+
"loss": 3.5776,
|
1230 |
+
"step": 4100
|
1231 |
+
},
|
1232 |
+
{
|
1233 |
+
"epoch": 77.36,
|
1234 |
+
"eval_loss": 4.282679557800293,
|
1235 |
+
"eval_runtime": 2.3987,
|
1236 |
+
"eval_samples_per_second": 298.911,
|
1237 |
+
"eval_steps_per_second": 9.589,
|
1238 |
+
"eval_wer": 0.09893226559893227,
|
1239 |
+
"step": 4100
|
1240 |
+
},
|
1241 |
+
{
|
1242 |
+
"epoch": 78.3,
|
1243 |
+
"learning_rate": 2.884828340647414e-06,
|
1244 |
+
"loss": 3.5116,
|
1245 |
+
"step": 4150
|
1246 |
+
},
|
1247 |
+
{
|
1248 |
+
"epoch": 78.3,
|
1249 |
+
"eval_loss": 4.324526786804199,
|
1250 |
+
"eval_runtime": 2.3498,
|
1251 |
+
"eval_samples_per_second": 305.135,
|
1252 |
+
"eval_steps_per_second": 9.788,
|
1253 |
+
"eval_wer": 0.10018351685018352,
|
1254 |
+
"step": 4150
|
1255 |
+
},
|
1256 |
+
{
|
1257 |
+
"epoch": 79.25,
|
1258 |
+
"learning_rate": 2.657333765809459e-06,
|
1259 |
+
"loss": 3.3334,
|
1260 |
+
"step": 4200
|
1261 |
+
},
|
1262 |
+
{
|
1263 |
+
"epoch": 79.25,
|
1264 |
+
"eval_loss": 4.270716667175293,
|
1265 |
+
"eval_runtime": 2.4082,
|
1266 |
+
"eval_samples_per_second": 297.731,
|
1267 |
+
"eval_steps_per_second": 9.551,
|
1268 |
+
"eval_wer": 0.0995995995995996,
|
1269 |
+
"step": 4200
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 80.19,
|
1273 |
+
"learning_rate": 2.437801105959594e-06,
|
1274 |
+
"loss": 3.4829,
|
1275 |
+
"step": 4250
|
1276 |
+
},
|
1277 |
+
{
|
1278 |
+
"epoch": 80.19,
|
1279 |
+
"eval_loss": 4.245628833770752,
|
1280 |
+
"eval_runtime": 2.3251,
|
1281 |
+
"eval_samples_per_second": 308.368,
|
1282 |
+
"eval_steps_per_second": 9.892,
|
1283 |
+
"eval_wer": 0.09818151484818151,
|
1284 |
+
"step": 4250
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 81.13,
|
1288 |
+
"learning_rate": 2.2264684081939447e-06,
|
1289 |
+
"loss": 3.44,
|
1290 |
+
"step": 4300
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 81.13,
|
1294 |
+
"eval_loss": 4.284632682800293,
|
1295 |
+
"eval_runtime": 2.3871,
|
1296 |
+
"eval_samples_per_second": 300.371,
|
1297 |
+
"eval_steps_per_second": 9.635,
|
1298 |
+
"eval_wer": 0.10026693360026694,
|
1299 |
+
"step": 4300
|
1300 |
+
},
|
1301 |
+
{
|
1302 |
+
"epoch": 82.08,
|
1303 |
+
"learning_rate": 2.023564828097159e-06,
|
1304 |
+
"loss": 3.4112,
|
1305 |
+
"step": 4350
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 82.08,
|
1309 |
+
"eval_loss": 4.280004501342773,
|
1310 |
+
"eval_runtime": 2.3804,
|
1311 |
+
"eval_samples_per_second": 301.204,
|
1312 |
+
"eval_steps_per_second": 9.662,
|
1313 |
+
"eval_wer": 0.09768101434768102,
|
1314 |
+
"step": 4350
|
1315 |
+
},
|
1316 |
+
{
|
1317 |
+
"epoch": 83.02,
|
1318 |
+
"learning_rate": 1.829310381260848e-06,
|
1319 |
+
"loss": 3.3825,
|
1320 |
+
"step": 4400
|
1321 |
+
},
|
1322 |
+
{
|
1323 |
+
"epoch": 83.02,
|
1324 |
+
"eval_loss": 4.256912708282471,
|
1325 |
+
"eval_runtime": 2.3009,
|
1326 |
+
"eval_samples_per_second": 311.615,
|
1327 |
+
"eval_steps_per_second": 9.996,
|
1328 |
+
"eval_wer": 0.09759759759759759,
|
1329 |
+
"step": 4400
|
1330 |
+
},
|
1331 |
+
{
|
1332 |
+
"epoch": 83.96,
|
1333 |
+
"learning_rate": 1.647535456169591e-06,
|
1334 |
+
"loss": 3.3444,
|
1335 |
+
"step": 4450
|
1336 |
+
},
|
1337 |
+
{
|
1338 |
+
"epoch": 83.96,
|
1339 |
+
"eval_loss": 4.2334303855896,
|
1340 |
+
"eval_runtime": 2.4524,
|
1341 |
+
"eval_samples_per_second": 292.361,
|
1342 |
+
"eval_steps_per_second": 9.378,
|
1343 |
+
"eval_wer": 0.09492826159492826,
|
1344 |
+
"step": 4450
|
1345 |
+
},
|
1346 |
+
{
|
1347 |
+
"epoch": 84.91,
|
1348 |
+
"learning_rate": 1.4710184570696184e-06,
|
1349 |
+
"loss": 3.5125,
|
1350 |
+
"step": 4500
|
1351 |
+
},
|
1352 |
+
{
|
1353 |
+
"epoch": 84.91,
|
1354 |
+
"eval_loss": 4.26320743560791,
|
1355 |
+
"eval_runtime": 2.3986,
|
1356 |
+
"eval_samples_per_second": 298.918,
|
1357 |
+
"eval_steps_per_second": 9.589,
|
1358 |
+
"eval_wer": 0.09784784784784785,
|
1359 |
+
"step": 4500
|
1360 |
+
},
|
1361 |
+
{
|
1362 |
+
"epoch": 85.85,
|
1363 |
+
"learning_rate": 1.3037497369753871e-06,
|
1364 |
+
"loss": 3.3393,
|
1365 |
+
"step": 4550
|
1366 |
+
},
|
1367 |
+
{
|
1368 |
+
"epoch": 85.85,
|
1369 |
+
"eval_loss": 4.25075626373291,
|
1370 |
+
"eval_runtime": 2.4865,
|
1371 |
+
"eval_samples_per_second": 288.354,
|
1372 |
+
"eval_steps_per_second": 9.25,
|
1373 |
+
"eval_wer": 0.09793126459793126,
|
1374 |
+
"step": 4550
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 86.79,
|
1378 |
+
"learning_rate": 1.1459106713283286e-06,
|
1379 |
+
"loss": 3.4698,
|
1380 |
+
"step": 4600
|
1381 |
+
},
|
1382 |
+
{
|
1383 |
+
"epoch": 86.79,
|
1384 |
+
"eval_loss": 4.248310089111328,
|
1385 |
+
"eval_runtime": 2.3393,
|
1386 |
+
"eval_samples_per_second": 306.501,
|
1387 |
+
"eval_steps_per_second": 9.832,
|
1388 |
+
"eval_wer": 0.10001668335001668,
|
1389 |
+
"step": 4600
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 87.74,
|
1393 |
+
"learning_rate": 9.976724106591128e-07,
|
1394 |
+
"loss": 3.3466,
|
1395 |
+
"step": 4650
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 87.74,
|
1399 |
+
"eval_loss": 4.256034851074219,
|
1400 |
+
"eval_runtime": 2.41,
|
1401 |
+
"eval_samples_per_second": 297.505,
|
1402 |
+
"eval_steps_per_second": 9.543,
|
1403 |
+
"eval_wer": 0.09851518184851518,
|
1404 |
+
"step": 4650
|
1405 |
+
},
|
1406 |
+
{
|
1407 |
+
"epoch": 88.68,
|
1408 |
+
"learning_rate": 8.591956950030067e-07,
|
1409 |
+
"loss": 3.3808,
|
1410 |
+
"step": 4700
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 88.68,
|
1414 |
+
"eval_loss": 4.255034446716309,
|
1415 |
+
"eval_runtime": 2.293,
|
1416 |
+
"eval_samples_per_second": 312.694,
|
1417 |
+
"eval_steps_per_second": 10.031,
|
1418 |
+
"eval_wer": 0.09734734734734735,
|
1419 |
+
"step": 4700
|
1420 |
+
},
|
1421 |
+
{
|
1422 |
+
"epoch": 89.62,
|
1423 |
+
"learning_rate": 7.306306796037188e-07,
|
1424 |
+
"loss": 3.3442,
|
1425 |
+
"step": 4750
|
1426 |
+
},
|
1427 |
+
{
|
1428 |
+
"epoch": 89.62,
|
1429 |
+
"eval_loss": 4.2573628425598145,
|
1430 |
+
"eval_runtime": 2.3853,
|
1431 |
+
"eval_samples_per_second": 300.593,
|
1432 |
+
"eval_steps_per_second": 9.642,
|
1433 |
+
"eval_wer": 0.09818151484818151,
|
1434 |
+
"step": 4750
|
1435 |
+
},
|
1436 |
+
{
|
1437 |
+
"epoch": 90.57,
|
1438 |
+
"learning_rate": 6.121167720947174e-07,
|
1439 |
+
"loss": 3.0359,
|
1440 |
+
"step": 4800
|
1441 |
+
},
|
1442 |
+
{
|
1443 |
+
"epoch": 90.57,
|
1444 |
+
"eval_loss": 4.257233142852783,
|
1445 |
+
"eval_runtime": 2.3441,
|
1446 |
+
"eval_samples_per_second": 305.87,
|
1447 |
+
"eval_steps_per_second": 9.812,
|
1448 |
+
"eval_wer": 0.09934934934934934,
|
1449 |
+
"step": 4800
|
1450 |
+
},
|
1451 |
+
{
|
1452 |
+
"epoch": 91.51,
|
1453 |
+
"learning_rate": 5.037824813345571e-07,
|
1454 |
+
"loss": 3.5286,
|
1455 |
+
"step": 4850
|
1456 |
+
},
|
1457 |
+
{
|
1458 |
+
"epoch": 91.51,
|
1459 |
+
"eval_loss": 4.250905513763428,
|
1460 |
+
"eval_runtime": 2.4654,
|
1461 |
+
"eval_samples_per_second": 290.825,
|
1462 |
+
"eval_steps_per_second": 9.329,
|
1463 |
+
"eval_wer": 0.09934934934934934,
|
1464 |
+
"step": 4850
|
1465 |
+
},
|
1466 |
+
{
|
1467 |
+
"epoch": 92.45,
|
1468 |
+
"learning_rate": 4.057452780601334e-07,
|
1469 |
+
"loss": 3.0826,
|
1470 |
+
"step": 4900
|
1471 |
+
},
|
1472 |
+
{
|
1473 |
+
"epoch": 92.45,
|
1474 |
+
"eval_loss": 4.240777492523193,
|
1475 |
+
"eval_runtime": 2.4178,
|
1476 |
+
"eval_samples_per_second": 296.545,
|
1477 |
+
"eval_steps_per_second": 9.513,
|
1478 |
+
"eval_wer": 0.09768101434768102,
|
1479 |
+
"step": 4900
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 93.4,
|
1483 |
+
"learning_rate": 3.1811146750898025e-07,
|
1484 |
+
"loss": 3.513,
|
1485 |
+
"step": 4950
|
1486 |
+
},
|
1487 |
+
{
|
1488 |
+
"epoch": 93.4,
|
1489 |
+
"eval_loss": 4.253066539764404,
|
1490 |
+
"eval_runtime": 2.2776,
|
1491 |
+
"eval_samples_per_second": 314.806,
|
1492 |
+
"eval_steps_per_second": 10.098,
|
1493 |
+
"eval_wer": 0.09901568234901569,
|
1494 |
+
"step": 4950
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 94.34,
|
1498 |
+
"learning_rate": 2.4097607414869995e-07,
|
1499 |
+
"loss": 3.272,
|
1500 |
+
"step": 5000
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 94.34,
|
1504 |
+
"eval_loss": 4.255825996398926,
|
1505 |
+
"eval_runtime": 2.3617,
|
1506 |
+
"eval_samples_per_second": 303.593,
|
1507 |
+
"eval_steps_per_second": 9.739,
|
1508 |
+
"eval_wer": 0.09951618284951619,
|
1509 |
+
"step": 5000
|
1510 |
+
},
|
1511 |
+
{
|
1512 |
+
"epoch": 95.28,
|
1513 |
+
"learning_rate": 1.7442273863854553e-07,
|
1514 |
+
"loss": 3.2433,
|
1515 |
+
"step": 5050
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 95.28,
|
1519 |
+
"eval_loss": 4.2515153884887695,
|
1520 |
+
"eval_runtime": 2.4174,
|
1521 |
+
"eval_samples_per_second": 296.604,
|
1522 |
+
"eval_steps_per_second": 9.515,
|
1523 |
+
"eval_wer": 0.09918251584918251,
|
1524 |
+
"step": 5050
|
1525 |
+
},
|
1526 |
+
{
|
1527 |
+
"epoch": 96.23,
|
1528 |
+
"learning_rate": 1.185236271348722e-07,
|
1529 |
+
"loss": 3.3373,
|
1530 |
+
"step": 5100
|
1531 |
+
},
|
1532 |
+
{
|
1533 |
+
"epoch": 96.23,
|
1534 |
+
"eval_loss": 4.252400875091553,
|
1535 |
+
"eval_runtime": 2.3337,
|
1536 |
+
"eval_samples_per_second": 307.237,
|
1537 |
+
"eval_steps_per_second": 9.856,
|
1538 |
+
"eval_wer": 0.1001001001001001,
|
1539 |
+
"step": 5100
|
1540 |
+
},
|
1541 |
+
{
|
1542 |
+
"epoch": 97.17,
|
1543 |
+
"learning_rate": 7.33393530387927e-08,
|
1544 |
+
"loss": 3.2239,
|
1545 |
+
"step": 5150
|
1546 |
+
},
|
1547 |
+
{
|
1548 |
+
"epoch": 97.17,
|
1549 |
+
"eval_loss": 4.253963947296143,
|
1550 |
+
"eval_runtime": 2.307,
|
1551 |
+
"eval_samples_per_second": 310.799,
|
1552 |
+
"eval_steps_per_second": 9.97,
|
1553 |
+
"eval_wer": 0.09951618284951619,
|
1554 |
+
"step": 5150
|
1555 |
+
},
|
1556 |
+
{
|
1557 |
+
"epoch": 98.11,
|
1558 |
+
"learning_rate": 3.8918911270908745e-08,
|
1559 |
+
"loss": 3.4072,
|
1560 |
+
"step": 5200
|
1561 |
+
},
|
1562 |
+
{
|
1563 |
+
"epoch": 98.11,
|
1564 |
+
"eval_loss": 4.248571872711182,
|
1565 |
+
"eval_runtime": 2.3602,
|
1566 |
+
"eval_samples_per_second": 303.785,
|
1567 |
+
"eval_steps_per_second": 9.745,
|
1568 |
+
"eval_wer": 0.09934934934934934,
|
1569 |
+
"step": 5200
|
1570 |
+
},
|
1571 |
+
{
|
1572 |
+
"epoch": 99.06,
|
1573 |
+
"learning_rate": 1.5665974539441632e-08,
|
1574 |
+
"loss": 3.3015,
|
1575 |
+
"step": 5250
|
1576 |
+
},
|
1577 |
+
{
|
1578 |
+
"epoch": 99.06,
|
1579 |
+
"eval_loss": 4.249679088592529,
|
1580 |
+
"eval_runtime": 2.4076,
|
1581 |
+
"eval_samples_per_second": 297.807,
|
1582 |
+
"eval_steps_per_second": 9.553,
|
1583 |
+
"eval_wer": 0.09876543209876543,
|
1584 |
+
"step": 5250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 100.0,
|
1588 |
+
"learning_rate": 2.6567448613734612e-09,
|
1589 |
+
"loss": 3.329,
|
1590 |
+
"step": 5300
|
1591 |
+
},
|
1592 |
+
{
|
1593 |
+
"epoch": 100.0,
|
1594 |
+
"eval_loss": 4.2488484382629395,
|
1595 |
+
"eval_runtime": 2.329,
|
1596 |
+
"eval_samples_per_second": 307.858,
|
1597 |
+
"eval_steps_per_second": 9.876,
|
1598 |
+
"eval_wer": 0.09901568234901569,
|
1599 |
+
"step": 5300
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 100.0,
|
1603 |
+
"step": 5300,
|
1604 |
+
"total_flos": 8.034472481207091e+16,
|
1605 |
+
"train_loss": 65.34310805986513,
|
1606 |
+
"train_runtime": 4047.7439,
|
1607 |
+
"train_samples_per_second": 83.602,
|
1608 |
+
"train_steps_per_second": 1.309
|
1609 |
+
}
|
1610 |
+
],
|
1611 |
+
"logging_steps": 50,
|
1612 |
+
"max_steps": 5300,
|
1613 |
+
"num_train_epochs": 100,
|
1614 |
+
"save_steps": 50,
|
1615 |
+
"total_flos": 8.034472481207091e+16,
|
1616 |
+
"trial_name": null,
|
1617 |
+
"trial_params": null
|
1618 |
+
}
|