Training in progress, step 2000
Browse files- model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- runs/Jun07_12-33-16_DESKTOP-69FPKCK/events.out.tfevents.1717788805.DESKTOP-69FPKCK +2 -2
- scheduler.pt +3 -0
- trainer_state.json +730 -0
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 44644496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc3ff2b0341cfe326f9f163502af875a0edd6f1e706c9f117b934b936037b886
|
| 3 |
size 44644496
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:610cf6c0652909f68d3dc7d2e9626ea7ef4255245b6d5052cbe5d2aa8ec6262b
|
| 3 |
+
size 11230198
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ff264f99d31b522cc7e2a4eac9d38606d0c58a34c0adc74d71e0ca8b371dc36
|
| 3 |
+
size 14244
|
runs/Jun07_12-33-16_DESKTOP-69FPKCK/events.out.tfevents.1717788805.DESKTOP-69FPKCK
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:af5ce506e31362699c9c7324fe55ddcf9140b636d93dc6ba1de2c8bf2e2dab8e
|
| 3 |
+
size 25646
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d36825260a27087b0c3610daac7ed09da2423bee63d54819c2e9215d0ba00718
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,730 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.00904899578769246,
|
| 5 |
+
"eval_steps": 2000,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 9.048995787692461e-05,
|
| 13 |
+
"grad_norm": 1.1874778270721436,
|
| 14 |
+
"learning_rate": 2.7146864537145957e-07,
|
| 15 |
+
"loss": 10.3312,
|
| 16 |
+
"step": 20
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.00018097991575384922,
|
| 20 |
+
"grad_norm": 1.3932149410247803,
|
| 21 |
+
"learning_rate": 5.429372907429191e-07,
|
| 22 |
+
"loss": 10.3266,
|
| 23 |
+
"step": 40
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00027146987363077383,
|
| 27 |
+
"grad_norm": 1.2732529640197754,
|
| 28 |
+
"learning_rate": 8.144059361143787e-07,
|
| 29 |
+
"loss": 10.3163,
|
| 30 |
+
"step": 60
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.00036195983150769844,
|
| 34 |
+
"grad_norm": 1.07429039478302,
|
| 35 |
+
"learning_rate": 1.0858745814858383e-06,
|
| 36 |
+
"loss": 10.3044,
|
| 37 |
+
"step": 80
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.00045244978938462305,
|
| 41 |
+
"grad_norm": 1.0309141874313354,
|
| 42 |
+
"learning_rate": 1.357343226857298e-06,
|
| 43 |
+
"loss": 10.2959,
|
| 44 |
+
"step": 100
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0005429397472615477,
|
| 48 |
+
"grad_norm": 0.9270058870315552,
|
| 49 |
+
"learning_rate": 1.6288118722287574e-06,
|
| 50 |
+
"loss": 10.2818,
|
| 51 |
+
"step": 120
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0006334297051384723,
|
| 55 |
+
"grad_norm": 0.8409116864204407,
|
| 56 |
+
"learning_rate": 1.900280517600217e-06,
|
| 57 |
+
"loss": 10.2724,
|
| 58 |
+
"step": 140
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0007239196630153969,
|
| 62 |
+
"grad_norm": 0.7587267160415649,
|
| 63 |
+
"learning_rate": 2.1717491629716765e-06,
|
| 64 |
+
"loss": 10.2662,
|
| 65 |
+
"step": 160
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0008144096208923215,
|
| 69 |
+
"grad_norm": 0.8605366945266724,
|
| 70 |
+
"learning_rate": 2.4432178083431364e-06,
|
| 71 |
+
"loss": 10.2567,
|
| 72 |
+
"step": 180
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0009048995787692461,
|
| 76 |
+
"grad_norm": 0.8124440908432007,
|
| 77 |
+
"learning_rate": 2.714686453714596e-06,
|
| 78 |
+
"loss": 10.2513,
|
| 79 |
+
"step": 200
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.0009953895366461706,
|
| 83 |
+
"grad_norm": 0.8214222192764282,
|
| 84 |
+
"learning_rate": 2.9861550990860553e-06,
|
| 85 |
+
"loss": 10.2396,
|
| 86 |
+
"step": 220
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0010858794945230953,
|
| 90 |
+
"grad_norm": 0.7500312924385071,
|
| 91 |
+
"learning_rate": 3.2576237444575148e-06,
|
| 92 |
+
"loss": 10.2378,
|
| 93 |
+
"step": 240
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0011763694524000198,
|
| 97 |
+
"grad_norm": 0.7709519267082214,
|
| 98 |
+
"learning_rate": 3.529092389828975e-06,
|
| 99 |
+
"loss": 10.2287,
|
| 100 |
+
"step": 260
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0012668594102769445,
|
| 104 |
+
"grad_norm": 0.8319140672683716,
|
| 105 |
+
"learning_rate": 3.800561035200434e-06,
|
| 106 |
+
"loss": 10.2214,
|
| 107 |
+
"step": 280
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.001357349368153869,
|
| 111 |
+
"grad_norm": 0.8057898283004761,
|
| 112 |
+
"learning_rate": 4.072029680571894e-06,
|
| 113 |
+
"loss": 10.2072,
|
| 114 |
+
"step": 300
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0014478393260307938,
|
| 118 |
+
"grad_norm": 0.6834843754768372,
|
| 119 |
+
"learning_rate": 4.343498325943353e-06,
|
| 120 |
+
"loss": 10.1983,
|
| 121 |
+
"step": 320
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.0015383292839077183,
|
| 125 |
+
"grad_norm": 0.8223700523376465,
|
| 126 |
+
"learning_rate": 4.614966971314813e-06,
|
| 127 |
+
"loss": 10.1884,
|
| 128 |
+
"step": 340
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.001628819241784643,
|
| 132 |
+
"grad_norm": 0.8147690892219543,
|
| 133 |
+
"learning_rate": 4.886435616686273e-06,
|
| 134 |
+
"loss": 10.1814,
|
| 135 |
+
"step": 360
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.0017193091996615675,
|
| 139 |
+
"grad_norm": 0.8512526750564575,
|
| 140 |
+
"learning_rate": 5.157904262057733e-06,
|
| 141 |
+
"loss": 10.1713,
|
| 142 |
+
"step": 380
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.0018097991575384922,
|
| 146 |
+
"grad_norm": 0.8844230771064758,
|
| 147 |
+
"learning_rate": 5.429372907429192e-06,
|
| 148 |
+
"loss": 10.1572,
|
| 149 |
+
"step": 400
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0019002891154154167,
|
| 153 |
+
"grad_norm": 0.9605993628501892,
|
| 154 |
+
"learning_rate": 5.700841552800652e-06,
|
| 155 |
+
"loss": 10.1496,
|
| 156 |
+
"step": 420
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.001990779073292341,
|
| 160 |
+
"grad_norm": 1.2027961015701294,
|
| 161 |
+
"learning_rate": 5.972310198172111e-06,
|
| 162 |
+
"loss": 10.1298,
|
| 163 |
+
"step": 440
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.002081269031169266,
|
| 167 |
+
"grad_norm": 1.4069308042526245,
|
| 168 |
+
"learning_rate": 6.2437788435435705e-06,
|
| 169 |
+
"loss": 10.1092,
|
| 170 |
+
"step": 460
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.0021717589890461906,
|
| 174 |
+
"grad_norm": 1.7658456563949585,
|
| 175 |
+
"learning_rate": 6.5152474889150296e-06,
|
| 176 |
+
"loss": 10.0954,
|
| 177 |
+
"step": 480
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.002262248946923115,
|
| 181 |
+
"grad_norm": 1.6941689252853394,
|
| 182 |
+
"learning_rate": 6.7867161342864895e-06,
|
| 183 |
+
"loss": 10.0746,
|
| 184 |
+
"step": 500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.0023527389048000396,
|
| 188 |
+
"grad_norm": 2.362786293029785,
|
| 189 |
+
"learning_rate": 7.05818477965795e-06,
|
| 190 |
+
"loss": 10.0613,
|
| 191 |
+
"step": 520
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.0024432288626769646,
|
| 195 |
+
"grad_norm": 1.827091932296753,
|
| 196 |
+
"learning_rate": 7.329653425029408e-06,
|
| 197 |
+
"loss": 10.045,
|
| 198 |
+
"step": 540
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.002533718820553889,
|
| 202 |
+
"grad_norm": 2.136615753173828,
|
| 203 |
+
"learning_rate": 7.601122070400868e-06,
|
| 204 |
+
"loss": 10.0243,
|
| 205 |
+
"step": 560
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.0026242087784308136,
|
| 209 |
+
"grad_norm": 2.501790761947632,
|
| 210 |
+
"learning_rate": 7.872590715772328e-06,
|
| 211 |
+
"loss": 10.0091,
|
| 212 |
+
"step": 580
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.002714698736307738,
|
| 216 |
+
"grad_norm": 2.7978005409240723,
|
| 217 |
+
"learning_rate": 8.144059361143788e-06,
|
| 218 |
+
"loss": 9.9957,
|
| 219 |
+
"step": 600
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.002805188694184663,
|
| 223 |
+
"grad_norm": 3.0485517978668213,
|
| 224 |
+
"learning_rate": 8.415528006515246e-06,
|
| 225 |
+
"loss": 9.9819,
|
| 226 |
+
"step": 620
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.0028956786520615875,
|
| 230 |
+
"grad_norm": 2.761986255645752,
|
| 231 |
+
"learning_rate": 8.686996651886706e-06,
|
| 232 |
+
"loss": 9.9596,
|
| 233 |
+
"step": 640
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.002986168609938512,
|
| 237 |
+
"grad_norm": 3.0985260009765625,
|
| 238 |
+
"learning_rate": 8.958465297258166e-06,
|
| 239 |
+
"loss": 9.9436,
|
| 240 |
+
"step": 660
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.0030766585678154365,
|
| 244 |
+
"grad_norm": 2.40391206741333,
|
| 245 |
+
"learning_rate": 9.229933942629626e-06,
|
| 246 |
+
"loss": 9.9226,
|
| 247 |
+
"step": 680
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.0031671485256923614,
|
| 251 |
+
"grad_norm": 1.933740496635437,
|
| 252 |
+
"learning_rate": 9.501402588001086e-06,
|
| 253 |
+
"loss": 9.9069,
|
| 254 |
+
"step": 700
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.003257638483569286,
|
| 258 |
+
"grad_norm": 2.518874168395996,
|
| 259 |
+
"learning_rate": 9.772871233372546e-06,
|
| 260 |
+
"loss": 9.881,
|
| 261 |
+
"step": 720
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.0033481284414462104,
|
| 265 |
+
"grad_norm": 2.8025624752044678,
|
| 266 |
+
"learning_rate": 1.0044339878744006e-05,
|
| 267 |
+
"loss": 9.8686,
|
| 268 |
+
"step": 740
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.003438618399323135,
|
| 272 |
+
"grad_norm": 1.943656086921692,
|
| 273 |
+
"learning_rate": 1.0315808524115465e-05,
|
| 274 |
+
"loss": 9.8463,
|
| 275 |
+
"step": 760
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.00352910835720006,
|
| 279 |
+
"grad_norm": 1.753179907798767,
|
| 280 |
+
"learning_rate": 1.0587277169486925e-05,
|
| 281 |
+
"loss": 9.8344,
|
| 282 |
+
"step": 780
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.0036195983150769844,
|
| 286 |
+
"grad_norm": 1.9388506412506104,
|
| 287 |
+
"learning_rate": 1.0858745814858383e-05,
|
| 288 |
+
"loss": 9.8144,
|
| 289 |
+
"step": 800
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.003710088272953909,
|
| 293 |
+
"grad_norm": 2.6278536319732666,
|
| 294 |
+
"learning_rate": 1.1130214460229843e-05,
|
| 295 |
+
"loss": 9.8008,
|
| 296 |
+
"step": 820
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.0038005782308308334,
|
| 300 |
+
"grad_norm": 1.8270655870437622,
|
| 301 |
+
"learning_rate": 1.1401683105601303e-05,
|
| 302 |
+
"loss": 9.7791,
|
| 303 |
+
"step": 840
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.0038910681887077583,
|
| 307 |
+
"grad_norm": 1.656563639640808,
|
| 308 |
+
"learning_rate": 1.1673151750972763e-05,
|
| 309 |
+
"loss": 9.7677,
|
| 310 |
+
"step": 860
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.003981558146584682,
|
| 314 |
+
"grad_norm": 1.6003910303115845,
|
| 315 |
+
"learning_rate": 1.1944620396344221e-05,
|
| 316 |
+
"loss": 9.7465,
|
| 317 |
+
"step": 880
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.004072048104461608,
|
| 321 |
+
"grad_norm": 1.5632762908935547,
|
| 322 |
+
"learning_rate": 1.2216089041715681e-05,
|
| 323 |
+
"loss": 9.73,
|
| 324 |
+
"step": 900
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.004162538062338532,
|
| 328 |
+
"grad_norm": 1.4974184036254883,
|
| 329 |
+
"learning_rate": 1.2487557687087141e-05,
|
| 330 |
+
"loss": 9.7067,
|
| 331 |
+
"step": 920
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.004253028020215457,
|
| 335 |
+
"grad_norm": 1.811112880706787,
|
| 336 |
+
"learning_rate": 1.2759026332458601e-05,
|
| 337 |
+
"loss": 9.6956,
|
| 338 |
+
"step": 940
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.004343517978092381,
|
| 342 |
+
"grad_norm": 1.505334734916687,
|
| 343 |
+
"learning_rate": 1.3030494977830059e-05,
|
| 344 |
+
"loss": 9.6667,
|
| 345 |
+
"step": 960
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.004434007935969306,
|
| 349 |
+
"grad_norm": 1.6951265335083008,
|
| 350 |
+
"learning_rate": 1.3301963623201519e-05,
|
| 351 |
+
"loss": 9.6505,
|
| 352 |
+
"step": 980
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.00452449789384623,
|
| 356 |
+
"grad_norm": 1.6119604110717773,
|
| 357 |
+
"learning_rate": 1.3573432268572979e-05,
|
| 358 |
+
"loss": 9.6381,
|
| 359 |
+
"step": 1000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.004614987851723155,
|
| 363 |
+
"grad_norm": 1.1929903030395508,
|
| 364 |
+
"learning_rate": 1.3844900913944439e-05,
|
| 365 |
+
"loss": 9.6209,
|
| 366 |
+
"step": 1020
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.004705477809600079,
|
| 370 |
+
"grad_norm": 1.5701353549957275,
|
| 371 |
+
"learning_rate": 1.41163695593159e-05,
|
| 372 |
+
"loss": 9.5956,
|
| 373 |
+
"step": 1040
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.004795967767477005,
|
| 377 |
+
"grad_norm": 1.32628333568573,
|
| 378 |
+
"learning_rate": 1.4387838204687359e-05,
|
| 379 |
+
"loss": 9.5899,
|
| 380 |
+
"step": 1060
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.004886457725353929,
|
| 384 |
+
"grad_norm": 1.5850657224655151,
|
| 385 |
+
"learning_rate": 1.4659306850058817e-05,
|
| 386 |
+
"loss": 9.5779,
|
| 387 |
+
"step": 1080
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.004976947683230854,
|
| 391 |
+
"grad_norm": 1.3933109045028687,
|
| 392 |
+
"learning_rate": 1.4930775495430278e-05,
|
| 393 |
+
"loss": 9.5701,
|
| 394 |
+
"step": 1100
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.005067437641107778,
|
| 398 |
+
"grad_norm": 1.258367657661438,
|
| 399 |
+
"learning_rate": 1.5202244140801737e-05,
|
| 400 |
+
"loss": 9.5468,
|
| 401 |
+
"step": 1120
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.005157927598984703,
|
| 405 |
+
"grad_norm": 1.3926512002944946,
|
| 406 |
+
"learning_rate": 1.5473712786173196e-05,
|
| 407 |
+
"loss": 9.5392,
|
| 408 |
+
"step": 1140
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.005248417556861627,
|
| 412 |
+
"grad_norm": 1.1674704551696777,
|
| 413 |
+
"learning_rate": 1.5745181431544656e-05,
|
| 414 |
+
"loss": 9.5291,
|
| 415 |
+
"step": 1160
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.005338907514738552,
|
| 419 |
+
"grad_norm": 1.4704829454421997,
|
| 420 |
+
"learning_rate": 1.6016650076916116e-05,
|
| 421 |
+
"loss": 9.5219,
|
| 422 |
+
"step": 1180
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.005429397472615476,
|
| 426 |
+
"grad_norm": 1.6223082542419434,
|
| 427 |
+
"learning_rate": 1.6288118722287576e-05,
|
| 428 |
+
"loss": 9.4918,
|
| 429 |
+
"step": 1200
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.0055198874304924015,
|
| 433 |
+
"grad_norm": 1.8586570024490356,
|
| 434 |
+
"learning_rate": 1.6559587367659036e-05,
|
| 435 |
+
"loss": 9.4895,
|
| 436 |
+
"step": 1220
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.005610377388369326,
|
| 440 |
+
"grad_norm": 1.4105405807495117,
|
| 441 |
+
"learning_rate": 1.6831056013030492e-05,
|
| 442 |
+
"loss": 9.4886,
|
| 443 |
+
"step": 1240
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.0057008673462462505,
|
| 447 |
+
"grad_norm": 1.4756163358688354,
|
| 448 |
+
"learning_rate": 1.7102524658401956e-05,
|
| 449 |
+
"loss": 9.4702,
|
| 450 |
+
"step": 1260
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.005791357304123175,
|
| 454 |
+
"grad_norm": 1.3847874402999878,
|
| 455 |
+
"learning_rate": 1.7373993303773412e-05,
|
| 456 |
+
"loss": 9.4638,
|
| 457 |
+
"step": 1280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.0058818472620000995,
|
| 461 |
+
"grad_norm": 1.5135865211486816,
|
| 462 |
+
"learning_rate": 1.7645461949144875e-05,
|
| 463 |
+
"loss": 9.4583,
|
| 464 |
+
"step": 1300
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.005972337219877024,
|
| 468 |
+
"grad_norm": 1.462760329246521,
|
| 469 |
+
"learning_rate": 1.7916930594516332e-05,
|
| 470 |
+
"loss": 9.4292,
|
| 471 |
+
"step": 1320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.0060628271777539485,
|
| 475 |
+
"grad_norm": 1.646760106086731,
|
| 476 |
+
"learning_rate": 1.8188399239887792e-05,
|
| 477 |
+
"loss": 9.4419,
|
| 478 |
+
"step": 1340
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.006153317135630873,
|
| 482 |
+
"grad_norm": 1.3564046621322632,
|
| 483 |
+
"learning_rate": 1.8459867885259252e-05,
|
| 484 |
+
"loss": 9.4283,
|
| 485 |
+
"step": 1360
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.006243807093507798,
|
| 489 |
+
"grad_norm": 1.4385489225387573,
|
| 490 |
+
"learning_rate": 1.873133653063071e-05,
|
| 491 |
+
"loss": 9.4208,
|
| 492 |
+
"step": 1380
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.006334297051384723,
|
| 496 |
+
"grad_norm": 1.3975261449813843,
|
| 497 |
+
"learning_rate": 1.900280517600217e-05,
|
| 498 |
+
"loss": 9.4015,
|
| 499 |
+
"step": 1400
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.006424787009261647,
|
| 503 |
+
"grad_norm": 1.4809174537658691,
|
| 504 |
+
"learning_rate": 1.927427382137363e-05,
|
| 505 |
+
"loss": 9.4009,
|
| 506 |
+
"step": 1420
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.006515276967138572,
|
| 510 |
+
"grad_norm": 1.5181605815887451,
|
| 511 |
+
"learning_rate": 1.954574246674509e-05,
|
| 512 |
+
"loss": 9.3969,
|
| 513 |
+
"step": 1440
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.006605766925015496,
|
| 517 |
+
"grad_norm": 1.4760838747024536,
|
| 518 |
+
"learning_rate": 1.981721111211655e-05,
|
| 519 |
+
"loss": 9.395,
|
| 520 |
+
"step": 1460
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.006696256882892421,
|
| 524 |
+
"grad_norm": 1.6140539646148682,
|
| 525 |
+
"learning_rate": 2.008867975748801e-05,
|
| 526 |
+
"loss": 9.3868,
|
| 527 |
+
"step": 1480
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.006786746840769345,
|
| 531 |
+
"grad_norm": 1.469307541847229,
|
| 532 |
+
"learning_rate": 2.0360148402859468e-05,
|
| 533 |
+
"loss": 9.3766,
|
| 534 |
+
"step": 1500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.00687723679864627,
|
| 538 |
+
"grad_norm": 1.8742159605026245,
|
| 539 |
+
"learning_rate": 2.063161704823093e-05,
|
| 540 |
+
"loss": 9.3715,
|
| 541 |
+
"step": 1520
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.006967726756523195,
|
| 545 |
+
"grad_norm": 1.5996043682098389,
|
| 546 |
+
"learning_rate": 2.0903085693602387e-05,
|
| 547 |
+
"loss": 9.3622,
|
| 548 |
+
"step": 1540
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.00705821671440012,
|
| 552 |
+
"grad_norm": 1.867632508277893,
|
| 553 |
+
"learning_rate": 2.117455433897385e-05,
|
| 554 |
+
"loss": 9.3704,
|
| 555 |
+
"step": 1560
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.007148706672277044,
|
| 559 |
+
"grad_norm": 1.4762872457504272,
|
| 560 |
+
"learning_rate": 2.1446022984345307e-05,
|
| 561 |
+
"loss": 9.3741,
|
| 562 |
+
"step": 1580
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.007239196630153969,
|
| 566 |
+
"grad_norm": 1.5752198696136475,
|
| 567 |
+
"learning_rate": 2.1717491629716767e-05,
|
| 568 |
+
"loss": 9.3561,
|
| 569 |
+
"step": 1600
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.007329686588030893,
|
| 573 |
+
"grad_norm": 1.637786865234375,
|
| 574 |
+
"learning_rate": 2.1988960275088227e-05,
|
| 575 |
+
"loss": 9.3535,
|
| 576 |
+
"step": 1620
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.007420176545907818,
|
| 580 |
+
"grad_norm": 2.6087028980255127,
|
| 581 |
+
"learning_rate": 2.2260428920459687e-05,
|
| 582 |
+
"loss": 9.3541,
|
| 583 |
+
"step": 1640
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.007510666503784742,
|
| 587 |
+
"grad_norm": 1.977252721786499,
|
| 588 |
+
"learning_rate": 2.2531897565831143e-05,
|
| 589 |
+
"loss": 9.3341,
|
| 590 |
+
"step": 1660
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.007601156461661667,
|
| 594 |
+
"grad_norm": 1.9511388540267944,
|
| 595 |
+
"learning_rate": 2.2803366211202606e-05,
|
| 596 |
+
"loss": 9.339,
|
| 597 |
+
"step": 1680
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.007691646419538592,
|
| 601 |
+
"grad_norm": 1.8821523189544678,
|
| 602 |
+
"learning_rate": 2.3074834856574063e-05,
|
| 603 |
+
"loss": 9.3234,
|
| 604 |
+
"step": 1700
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.007782136377415517,
|
| 608 |
+
"grad_norm": 1.5517367124557495,
|
| 609 |
+
"learning_rate": 2.3346303501945526e-05,
|
| 610 |
+
"loss": 9.3367,
|
| 611 |
+
"step": 1720
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.00787262633529244,
|
| 615 |
+
"grad_norm": 2.164625883102417,
|
| 616 |
+
"learning_rate": 2.3617772147316983e-05,
|
| 617 |
+
"loss": 9.3366,
|
| 618 |
+
"step": 1740
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.007963116293169365,
|
| 622 |
+
"grad_norm": 2.4158406257629395,
|
| 623 |
+
"learning_rate": 2.3889240792688443e-05,
|
| 624 |
+
"loss": 9.3221,
|
| 625 |
+
"step": 1760
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.00805360625104629,
|
| 629 |
+
"grad_norm": 1.8652360439300537,
|
| 630 |
+
"learning_rate": 2.4160709438059906e-05,
|
| 631 |
+
"loss": 9.3098,
|
| 632 |
+
"step": 1780
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.008144096208923216,
|
| 636 |
+
"grad_norm": 1.8249917030334473,
|
| 637 |
+
"learning_rate": 2.4432178083431362e-05,
|
| 638 |
+
"loss": 9.3094,
|
| 639 |
+
"step": 1800
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.00823458616680014,
|
| 643 |
+
"grad_norm": 2.06990647315979,
|
| 644 |
+
"learning_rate": 2.4703646728802822e-05,
|
| 645 |
+
"loss": 9.2994,
|
| 646 |
+
"step": 1820
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.008325076124677065,
|
| 650 |
+
"grad_norm": 2.461805582046509,
|
| 651 |
+
"learning_rate": 2.4975115374174282e-05,
|
| 652 |
+
"loss": 9.3157,
|
| 653 |
+
"step": 1840
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.008415566082553989,
|
| 657 |
+
"grad_norm": 2.1320767402648926,
|
| 658 |
+
"learning_rate": 2.5246584019545742e-05,
|
| 659 |
+
"loss": 9.281,
|
| 660 |
+
"step": 1860
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.008506056040430914,
|
| 664 |
+
"grad_norm": 2.6872756481170654,
|
| 665 |
+
"learning_rate": 2.5518052664917202e-05,
|
| 666 |
+
"loss": 9.2917,
|
| 667 |
+
"step": 1880
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.008596545998307838,
|
| 671 |
+
"grad_norm": 2.4759294986724854,
|
| 672 |
+
"learning_rate": 2.5789521310288662e-05,
|
| 673 |
+
"loss": 9.2941,
|
| 674 |
+
"step": 1900
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.008687035956184763,
|
| 678 |
+
"grad_norm": 1.8129667043685913,
|
| 679 |
+
"learning_rate": 2.6060989955660118e-05,
|
| 680 |
+
"loss": 9.2815,
|
| 681 |
+
"step": 1920
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.008777525914061687,
|
| 685 |
+
"grad_norm": 2.9053220748901367,
|
| 686 |
+
"learning_rate": 2.633245860103158e-05,
|
| 687 |
+
"loss": 9.2801,
|
| 688 |
+
"step": 1940
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.008868015871938612,
|
| 692 |
+
"grad_norm": 2.412623167037964,
|
| 693 |
+
"learning_rate": 2.6603927246403038e-05,
|
| 694 |
+
"loss": 9.2719,
|
| 695 |
+
"step": 1960
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.008958505829815536,
|
| 699 |
+
"grad_norm": 1.972790002822876,
|
| 700 |
+
"learning_rate": 2.6875395891774498e-05,
|
| 701 |
+
"loss": 9.2729,
|
| 702 |
+
"step": 1980
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.00904899578769246,
|
| 706 |
+
"grad_norm": 3.04768705368042,
|
| 707 |
+
"learning_rate": 2.7146864537145958e-05,
|
| 708 |
+
"loss": 9.2653,
|
| 709 |
+
"step": 2000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.00904899578769246,
|
| 713 |
+
"eval_accuracy": 0.10545615706904701,
|
| 714 |
+
"eval_loss": 9.261013984680176,
|
| 715 |
+
"eval_runtime": 215.2628,
|
| 716 |
+
"eval_samples_per_second": 2823.711,
|
| 717 |
+
"eval_steps_per_second": 11.033,
|
| 718 |
+
"step": 2000
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 20,
|
| 722 |
+
"max_steps": 663057,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 3,
|
| 725 |
+
"save_steps": 100,
|
| 726 |
+
"total_flos": 719181053952000.0,
|
| 727 |
+
"train_batch_size": 256,
|
| 728 |
+
"trial_name": null,
|
| 729 |
+
"trial_params": null
|
| 730 |
+
}
|