Training in progress, step 285, checkpoint
Browse files- checkpoint-285/global_step284/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/global_step284/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt +1 -1
- checkpoint-285/model-00001-of-00014.safetensors +1 -1
- checkpoint-285/model-00002-of-00014.safetensors +1 -1
- checkpoint-285/model-00003-of-00014.safetensors +1 -1
- checkpoint-285/model-00004-of-00014.safetensors +1 -1
- checkpoint-285/model-00005-of-00014.safetensors +1 -1
- checkpoint-285/model-00006-of-00014.safetensors +1 -1
- checkpoint-285/model-00007-of-00014.safetensors +1 -1
- checkpoint-285/model-00008-of-00014.safetensors +1 -1
- checkpoint-285/model-00009-of-00014.safetensors +1 -1
- checkpoint-285/model-00010-of-00014.safetensors +1 -1
- checkpoint-285/model-00011-of-00014.safetensors +1 -1
- checkpoint-285/model-00012-of-00014.safetensors +1 -1
- checkpoint-285/model-00013-of-00014.safetensors +1 -1
- checkpoint-285/model-00014-of-00014.safetensors +1 -1
- checkpoint-285/scheduler.pt +1 -1
- checkpoint-285/trainer_state.json +853 -853
- checkpoint-285/training_args.bin +1 -1
checkpoint-285/global_step284/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de31ab4cc55b5a62a02311287c1c4727c533b8f311e2a9d2dfc62cc9bc0e13d5
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a4b18d2a78e83956f01de73a12f6e4073ee73a7aecaade2daedc7a9846265f2c
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da7ffd025ed28a5619ead7166b1c8f0c355478aeeb1cd72606519f0bb0c9e8f9
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7743641d34754453647cb84c53bc0e4ac5dcbeb0d3bbd2e76863cc9f02a55cf0
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1501dda8d48b87a7768c14609361d9ef225fdf4e044114b6901eedcda389ba8d
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2fa2f137899bedcb1e4c119d894eb991a11d22e3083de538541d7d7ebbdae57
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6c13d22c1fd3ac104de66b63a8439eff7253e8c8638787dd9de04cb833fb693
|
| 3 |
size 24702834279
|
checkpoint-285/global_step284/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24702834279
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f42819fba647e363429895c0a5b1c2b06a9a9ec455d353be37c37a503a56ebe
|
| 3 |
size 24702834279
|
checkpoint-285/model-00001-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4891730992
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ea61edbcc041ff874873784b2ea0ed8c3a8e61039e048d21944f403711ba32d
|
| 3 |
size 4891730992
|
checkpoint-285/model-00002-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059352
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d6344c83acb079116781c108e03221a8f14aaa0c4cdfecb38d5a31a0fe1d46c
|
| 3 |
size 4876059352
|
checkpoint-285/model-00003-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059384
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd504b64c920ca644ef03455badf3bef40c8f07909e704f1749699769b132792
|
| 3 |
size 4876059384
|
checkpoint-285/model-00004-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4582cc3002f1f8ff564b58a8f2d0be4b54ec8aec6f712f8ef6fb7cee92f8349c
|
| 3 |
size 4876059416
|
checkpoint-285/model-00005-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6673e14622d95b673ff9ac0e8a50aaf111a50dd1a2e623cea6a27b4c0a3aeb5e
|
| 3 |
size 4876059416
|
checkpoint-285/model-00006-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3733f1990dd750cde99801b3977e22d405b70360761216f9026297e4439d75dd
|
| 3 |
size 4876059416
|
checkpoint-285/model-00007-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d6ff20d4c0007e797b00c8621f4a605c61019456d2f2b14642eb703fd478134
|
| 3 |
size 4876059416
|
checkpoint-285/model-00008-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e803fcc649cbd9981515476d92049bb420330da7f1ea2e75447e51bb05ec20c1
|
| 3 |
size 4876059416
|
checkpoint-285/model-00009-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5d06042deae94dbaf66a368317cbba0a0e94606aabdf427a36f4504e59eb808
|
| 3 |
size 4876059416
|
checkpoint-285/model-00010-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7f594cbd3d30f4012a27083a171fd5cf89d94a62324444378425bc97f777313
|
| 3 |
size 4876059416
|
checkpoint-285/model-00011-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e54aa15ff3f38d10e456a9cdb0e18de707067412dd846593f5e8340fb3ea19e7
|
| 3 |
size 4876059416
|
checkpoint-285/model-00012-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e368c270123c34cb6b7a0666d8bd590dc54d1ea4ccd1ed978816de306af6b4b
|
| 3 |
size 4876059416
|
checkpoint-285/model-00013-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4876059416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ec1c581c74495ee32834a958994e33ea0ded5add67ccc7dceb06eda98815f6e
|
| 3 |
size 4876059416
|
checkpoint-285/model-00014-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2123397800
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c128cac91acee62f711921c9bd02d95fc475f3bad50fbc3a7f854b1d3c2deb03
|
| 3 |
size 2123397800
|
checkpoint-285/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d2293e17d174bd961d6dabc3bd55c0c178258aa07f87daf3a13f6b70c674f144
|
| 3 |
size 1064
|
checkpoint-285/trainer_state.json
CHANGED
|
@@ -10,1997 +10,1997 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.005249343832020997,
|
| 13 |
-
"grad_norm": 1.
|
| 14 |
-
"learning_rate": 1.
|
| 15 |
"loss": 1.1087,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.010498687664041995,
|
| 20 |
-
"grad_norm": 1.
|
| 21 |
-
"learning_rate":
|
| 22 |
"loss": 1.1356,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.015748031496062992,
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.02099737532808399,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.026246719160104987,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 5e-
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.031496062992125984,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03674540682414698,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04199475065616798,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.047244094488188976,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.05249343832020997,
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05774278215223097,
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
-
"learning_rate": 1.
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.06299212598425197,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
-
"learning_rate": 1.
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06824146981627296,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
-
"learning_rate": 1.
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.07349081364829396,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07874015748031496,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.08398950131233596,
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08923884514435695,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.09448818897637795,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09973753280839895,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.10498687664041995,
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.11023622047244094,
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.11548556430446194,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
-
"learning_rate":
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.12073490813648294,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.12598425196850394,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.13123359580052493,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate":
|
| 183 |
-
"loss":
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.13648293963254593,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate":
|
| 190 |
-
"loss": 1.
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.14173228346456693,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate":
|
| 197 |
-
"loss": 1.
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.14698162729658792,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
-
"learning_rate":
|
| 204 |
-
"loss":
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.15223097112860892,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate":
|
| 211 |
-
"loss":
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.15748031496062992,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.16272965879265092,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
-
"learning_rate":
|
| 225 |
-
"loss":
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.1679790026246719,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
-
"learning_rate":
|
| 232 |
-
"loss": 1.
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.1732283464566929,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate":
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.1784776902887139,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate":
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.1837270341207349,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
-
"learning_rate":
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.1889763779527559,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.1942257217847769,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
-
"learning_rate":
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1994750656167979,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate":
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2047244094488189,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
-
"learning_rate":
|
| 281 |
-
"loss":
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.2099737532808399,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
-
"learning_rate":
|
| 288 |
-
"loss":
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.2152230971128609,
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
-
"learning_rate":
|
| 295 |
-
"loss":
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.2204724409448819,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss":
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.22572178477690288,
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
-
"learning_rate":
|
| 309 |
-
"loss":
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.23097112860892388,
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
-
"learning_rate":
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.23622047244094488,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate":
|
| 323 |
-
"loss":
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.24146981627296588,
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
-
"learning_rate":
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.24671916010498687,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate":
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.25196850393700787,
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.2572178477690289,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
-
"learning_rate":
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.26246719160104987,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
-
"learning_rate":
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.2677165354330709,
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
-
"learning_rate":
|
| 365 |
-
"loss": 1.
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.27296587926509186,
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
-
"learning_rate":
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.2782152230971129,
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate":
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.28346456692913385,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss":
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.2887139107611549,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
-
"learning_rate":
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.29396325459317585,
|
| 398 |
-
"grad_norm":
|
| 399 |
-
"learning_rate":
|
| 400 |
-
"loss":
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2992125984251969,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate":
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.30446194225721784,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
-
"learning_rate":
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.30971128608923887,
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
-
"learning_rate":
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.31496062992125984,
|
| 426 |
-
"grad_norm":
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.32020997375328086,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate":
|
| 435 |
-
"loss": 1.
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.32545931758530183,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate":
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.33070866141732286,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
-
"learning_rate":
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.3359580052493438,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
-
"learning_rate":
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.34120734908136485,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.3464566929133858,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss":
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.35170603674540685,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
-
"learning_rate":
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.3569553805774278,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
-
"learning_rate":
|
| 484 |
-
"loss":
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.36220472440944884,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
-
"learning_rate":
|
| 491 |
-
"loss":
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.3674540682414698,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
-
"learning_rate":
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.37270341207349084,
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
-
"learning_rate":
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.3779527559055118,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.38320209973753283,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
-
"learning_rate":
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.3884514435695538,
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
-
"learning_rate":
|
| 526 |
-
"loss": 0.
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.3937007874015748,
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
-
"learning_rate":
|
| 533 |
-
"loss":
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3989501312335958,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
-
"learning_rate":
|
| 540 |
-
"loss": 1.
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.4041994750656168,
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.4094488188976378,
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
-
"learning_rate":
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.4146981627296588,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
-
"learning_rate":
|
| 561 |
-
"loss": 0.
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.4199475065616798,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
-
"learning_rate":
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.4251968503937008,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
-
"learning_rate":
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.4304461942257218,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
-
"learning_rate":
|
| 582 |
-
"loss":
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.4356955380577428,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.4409448818897638,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.4461942257217848,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
-
"learning_rate":
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.45144356955380577,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
-
"learning_rate":
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.4566929133858268,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.46194225721784776,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
-
"learning_rate":
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.4671916010498688,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.47244094488188976,
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.4776902887139108,
|
| 643 |
-
"grad_norm": 0.
|
| 644 |
-
"learning_rate":
|
| 645 |
-
"loss": 1.
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.48293963254593175,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
-
"learning_rate":
|
| 652 |
-
"loss":
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.4881889763779528,
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
-
"learning_rate":
|
| 659 |
-
"loss":
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.49343832020997375,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
-
"learning_rate":
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.49868766404199477,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
-
"learning_rate":
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.5039370078740157,
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
-
"learning_rate":
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.5091863517060368,
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
-
"learning_rate":
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.5144356955380578,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
-
"learning_rate":
|
| 694 |
-
"loss": 0.
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.5196850393700787,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.5249343832020997,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
-
"learning_rate":
|
| 708 |
-
"loss": 0.
|
| 709 |
"step": 100
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.5301837270341208,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
-
"learning_rate":
|
| 715 |
-
"loss":
|
| 716 |
"step": 101
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.5354330708661418,
|
| 720 |
-
"grad_norm": 0.
|
| 721 |
-
"learning_rate":
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 102
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.5406824146981627,
|
| 727 |
-
"grad_norm": 0.
|
| 728 |
-
"learning_rate":
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 103
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.5459317585301837,
|
| 734 |
-
"grad_norm": 0.
|
| 735 |
-
"learning_rate":
|
| 736 |
-
"loss": 0.
|
| 737 |
"step": 104
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.5511811023622047,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
-
"learning_rate":
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 105
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.5564304461942258,
|
| 748 |
-
"grad_norm":
|
| 749 |
-
"learning_rate":
|
| 750 |
-
"loss":
|
| 751 |
"step": 106
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.5616797900262467,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
-
"learning_rate":
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 107
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.5669291338582677,
|
| 762 |
-
"grad_norm":
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 108
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.5721784776902887,
|
| 769 |
-
"grad_norm": 0.
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 109
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.5774278215223098,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
-
"learning_rate":
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 110
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.5826771653543307,
|
| 783 |
-
"grad_norm":
|
| 784 |
-
"learning_rate":
|
| 785 |
-
"loss":
|
| 786 |
"step": 111
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.5879265091863517,
|
| 790 |
-
"grad_norm": 0.
|
| 791 |
-
"learning_rate":
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 112
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.5931758530183727,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
-
"learning_rate":
|
| 799 |
-
"loss": 0.
|
| 800 |
"step": 113
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.5984251968503937,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
-
"learning_rate":
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 114
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.6036745406824147,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
-
"learning_rate":
|
| 813 |
-
"loss": 0.
|
| 814 |
"step": 115
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.6089238845144357,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
-
"learning_rate":
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 116
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.6141732283464567,
|
| 825 |
-
"grad_norm": 0.
|
| 826 |
-
"learning_rate":
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 117
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.6194225721784777,
|
| 832 |
-
"grad_norm": 0.
|
| 833 |
-
"learning_rate":
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 118
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.6246719160104987,
|
| 839 |
-
"grad_norm": 0.
|
| 840 |
-
"learning_rate":
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 119
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.6299212598425197,
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
-
"learning_rate":
|
| 848 |
-
"loss":
|
| 849 |
"step": 120
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.6351706036745407,
|
| 853 |
-
"grad_norm": 0.
|
| 854 |
-
"learning_rate":
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 121
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.6404199475065617,
|
| 860 |
-
"grad_norm": 0.
|
| 861 |
-
"learning_rate":
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 122
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.6456692913385826,
|
| 867 |
-
"grad_norm": 0.
|
| 868 |
-
"learning_rate":
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 123
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.6509186351706037,
|
| 874 |
-
"grad_norm": 0.
|
| 875 |
-
"learning_rate":
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 124
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.6561679790026247,
|
| 881 |
-
"grad_norm": 0.
|
| 882 |
-
"learning_rate":
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 125
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.6614173228346457,
|
| 888 |
-
"grad_norm": 0.
|
| 889 |
-
"learning_rate":
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 126
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.6666666666666666,
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
-
"learning_rate":
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 127
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.6719160104986877,
|
| 902 |
-
"grad_norm": 0.
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 128
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.6771653543307087,
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 129
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.6824146981627297,
|
| 916 |
-
"grad_norm": 0.
|
| 917 |
-
"learning_rate":
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 130
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.6876640419947506,
|
| 923 |
-
"grad_norm": 0.
|
| 924 |
-
"learning_rate":
|
| 925 |
-
"loss": 0.
|
| 926 |
"step": 131
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.6929133858267716,
|
| 930 |
-
"grad_norm": 0.
|
| 931 |
-
"learning_rate":
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 132
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.6981627296587927,
|
| 937 |
-
"grad_norm": 0.
|
| 938 |
-
"learning_rate":
|
| 939 |
-
"loss": 0.
|
| 940 |
"step": 133
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.7034120734908137,
|
| 944 |
-
"grad_norm": 0.
|
| 945 |
-
"learning_rate":
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 134
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.7086614173228346,
|
| 951 |
-
"grad_norm": 0.
|
| 952 |
-
"learning_rate":
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 135
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.7139107611548556,
|
| 958 |
-
"grad_norm": 0.
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 136
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.7191601049868767,
|
| 965 |
-
"grad_norm": 0.
|
| 966 |
-
"learning_rate":
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 137
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.7244094488188977,
|
| 972 |
-
"grad_norm": 0.
|
| 973 |
-
"learning_rate":
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 138
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.7296587926509186,
|
| 979 |
-
"grad_norm": 0.
|
| 980 |
-
"learning_rate":
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 139
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.7349081364829396,
|
| 986 |
-
"grad_norm": 0.
|
| 987 |
-
"learning_rate":
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 140
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.7401574803149606,
|
| 993 |
-
"grad_norm": 0.
|
| 994 |
-
"learning_rate":
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 141
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.7454068241469817,
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
-
"learning_rate":
|
| 1002 |
-
"loss": 0.
|
| 1003 |
"step": 142
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.7506561679790026,
|
| 1007 |
-
"grad_norm": 0.
|
| 1008 |
-
"learning_rate":
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 143
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.7559055118110236,
|
| 1014 |
-
"grad_norm": 0.
|
| 1015 |
-
"learning_rate":
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 144
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.7611548556430446,
|
| 1021 |
-
"grad_norm": 0.
|
| 1022 |
-
"learning_rate":
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 145
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.7664041994750657,
|
| 1028 |
-
"grad_norm": 0.
|
| 1029 |
-
"learning_rate":
|
| 1030 |
-
"loss": 0.
|
| 1031 |
"step": 146
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.7716535433070866,
|
| 1035 |
-
"grad_norm": 0.
|
| 1036 |
-
"learning_rate":
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 147
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.7769028871391076,
|
| 1042 |
-
"grad_norm": 0.
|
| 1043 |
-
"learning_rate":
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 148
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.7821522309711286,
|
| 1049 |
-
"grad_norm": 0.
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 149
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.7874015748031497,
|
| 1056 |
-
"grad_norm":
|
| 1057 |
-
"learning_rate":
|
| 1058 |
-
"loss": 1.
|
| 1059 |
"step": 150
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.7926509186351706,
|
| 1063 |
-
"grad_norm": 0.
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 151
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.7979002624671916,
|
| 1070 |
-
"grad_norm": 0.
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 152
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.8031496062992126,
|
| 1077 |
-
"grad_norm": 0.
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 153
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.8083989501312336,
|
| 1084 |
-
"grad_norm": 0.
|
| 1085 |
-
"learning_rate":
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 154
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.8136482939632546,
|
| 1091 |
-
"grad_norm": 0.
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 155
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.8188976377952756,
|
| 1098 |
-
"grad_norm": 0.
|
| 1099 |
-
"learning_rate":
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 156
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.8241469816272966,
|
| 1105 |
-
"grad_norm": 0.
|
| 1106 |
-
"learning_rate":
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 157
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.8293963254593176,
|
| 1112 |
-
"grad_norm": 0.
|
| 1113 |
-
"learning_rate":
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 158
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.8346456692913385,
|
| 1119 |
-
"grad_norm": 0.
|
| 1120 |
-
"learning_rate":
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 159
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.8398950131233596,
|
| 1126 |
-
"grad_norm": 0.
|
| 1127 |
-
"learning_rate":
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 160
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.8451443569553806,
|
| 1133 |
-
"grad_norm": 0.
|
| 1134 |
-
"learning_rate":
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 161
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.8503937007874016,
|
| 1140 |
-
"grad_norm": 0.
|
| 1141 |
-
"learning_rate":
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 162
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.8556430446194225,
|
| 1147 |
-
"grad_norm": 0.
|
| 1148 |
-
"learning_rate":
|
| 1149 |
-
"loss": 1.
|
| 1150 |
"step": 163
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.8608923884514436,
|
| 1154 |
-
"grad_norm": 0.
|
| 1155 |
-
"learning_rate":
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 164
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.8661417322834646,
|
| 1161 |
-
"grad_norm": 0.
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 165
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.8713910761154856,
|
| 1168 |
-
"grad_norm": 0.
|
| 1169 |
-
"learning_rate":
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 166
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.8766404199475065,
|
| 1175 |
-
"grad_norm": 0.
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 167
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.8818897637795275,
|
| 1182 |
-
"grad_norm": 0.
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 168
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.8871391076115486,
|
| 1189 |
-
"grad_norm": 0.
|
| 1190 |
-
"learning_rate":
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 169
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.8923884514435696,
|
| 1196 |
-
"grad_norm": 0.
|
| 1197 |
-
"learning_rate":
|
| 1198 |
-
"loss": 0.
|
| 1199 |
"step": 170
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.8976377952755905,
|
| 1203 |
-
"grad_norm": 0.
|
| 1204 |
-
"learning_rate":
|
| 1205 |
-
"loss":
|
| 1206 |
"step": 171
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.9028871391076115,
|
| 1210 |
-
"grad_norm": 0.
|
| 1211 |
-
"learning_rate":
|
| 1212 |
-
"loss": 0.
|
| 1213 |
"step": 172
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.9081364829396326,
|
| 1217 |
-
"grad_norm": 0.
|
| 1218 |
-
"learning_rate":
|
| 1219 |
-
"loss": 0.
|
| 1220 |
"step": 173
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
"epoch": 0.9133858267716536,
|
| 1224 |
-
"grad_norm": 0.
|
| 1225 |
-
"learning_rate":
|
| 1226 |
-
"loss": 0.
|
| 1227 |
"step": 174
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.9186351706036745,
|
| 1231 |
-
"grad_norm": 0.
|
| 1232 |
-
"learning_rate":
|
| 1233 |
-
"loss": 0.
|
| 1234 |
"step": 175
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
"epoch": 0.9238845144356955,
|
| 1238 |
-
"grad_norm": 0.
|
| 1239 |
-
"learning_rate":
|
| 1240 |
-
"loss": 0.
|
| 1241 |
"step": 176
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.9291338582677166,
|
| 1245 |
-
"grad_norm": 0.
|
| 1246 |
-
"learning_rate":
|
| 1247 |
-
"loss": 0.
|
| 1248 |
"step": 177
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
"epoch": 0.9343832020997376,
|
| 1252 |
-
"grad_norm": 0.
|
| 1253 |
-
"learning_rate":
|
| 1254 |
-
"loss": 0.
|
| 1255 |
"step": 178
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.9396325459317585,
|
| 1259 |
-
"grad_norm": 0.
|
| 1260 |
-
"learning_rate":
|
| 1261 |
-
"loss": 0.
|
| 1262 |
"step": 179
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
"epoch": 0.9448818897637795,
|
| 1266 |
-
"grad_norm": 0.
|
| 1267 |
-
"learning_rate":
|
| 1268 |
-
"loss": 0.
|
| 1269 |
"step": 180
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.9501312335958005,
|
| 1273 |
-
"grad_norm": 0.
|
| 1274 |
-
"learning_rate":
|
| 1275 |
-
"loss": 0.
|
| 1276 |
"step": 181
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
"epoch": 0.9553805774278216,
|
| 1280 |
-
"grad_norm": 0.
|
| 1281 |
-
"learning_rate":
|
| 1282 |
-
"loss": 0.
|
| 1283 |
"step": 182
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.9606299212598425,
|
| 1287 |
-
"grad_norm": 0.
|
| 1288 |
-
"learning_rate":
|
| 1289 |
-
"loss": 0.
|
| 1290 |
"step": 183
|
| 1291 |
},
|
| 1292 |
{
|
| 1293 |
"epoch": 0.9658792650918635,
|
| 1294 |
-
"grad_norm": 0.
|
| 1295 |
-
"learning_rate":
|
| 1296 |
-
"loss": 0.
|
| 1297 |
"step": 184
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.9711286089238845,
|
| 1301 |
-
"grad_norm": 0.
|
| 1302 |
-
"learning_rate":
|
| 1303 |
-
"loss": 0.
|
| 1304 |
"step": 185
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
"epoch": 0.9763779527559056,
|
| 1308 |
-
"grad_norm": 0.
|
| 1309 |
-
"learning_rate":
|
| 1310 |
-
"loss": 0.
|
| 1311 |
"step": 186
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.9816272965879265,
|
| 1315 |
-
"grad_norm": 0.
|
| 1316 |
-
"learning_rate":
|
| 1317 |
-
"loss": 0.
|
| 1318 |
"step": 187
|
| 1319 |
},
|
| 1320 |
{
|
| 1321 |
"epoch": 0.9868766404199475,
|
| 1322 |
-
"grad_norm": 0.
|
| 1323 |
-
"learning_rate":
|
| 1324 |
-
"loss": 0.
|
| 1325 |
"step": 188
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.9921259842519685,
|
| 1329 |
-
"grad_norm": 0.
|
| 1330 |
-
"learning_rate":
|
| 1331 |
-
"loss": 0.
|
| 1332 |
"step": 189
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
"epoch": 0.9973753280839895,
|
| 1336 |
-
"grad_norm": 0.
|
| 1337 |
-
"learning_rate":
|
| 1338 |
-
"loss": 0.
|
| 1339 |
"step": 190
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 1.0,
|
| 1343 |
-
"grad_norm": 0.
|
| 1344 |
-
"learning_rate":
|
| 1345 |
-
"loss": 0.
|
| 1346 |
"step": 191
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
"epoch": 1.005249343832021,
|
| 1350 |
-
"grad_norm": 0.
|
| 1351 |
-
"learning_rate":
|
| 1352 |
-
"loss": 0.
|
| 1353 |
"step": 192
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 1.010498687664042,
|
| 1357 |
-
"grad_norm": 0.
|
| 1358 |
-
"learning_rate":
|
| 1359 |
-
"loss": 0.
|
| 1360 |
"step": 193
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
"epoch": 1.015748031496063,
|
| 1364 |
-
"grad_norm":
|
| 1365 |
-
"learning_rate":
|
| 1366 |
-
"loss": 0.
|
| 1367 |
"step": 194
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
"epoch": 1.020997375328084,
|
| 1371 |
-
"grad_norm": 0.
|
| 1372 |
-
"learning_rate":
|
| 1373 |
-
"loss": 0.
|
| 1374 |
"step": 195
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
"epoch": 1.026246719160105,
|
| 1378 |
-
"grad_norm": 0.
|
| 1379 |
-
"learning_rate":
|
| 1380 |
-
"loss": 0.
|
| 1381 |
"step": 196
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
"epoch": 1.031496062992126,
|
| 1385 |
-
"grad_norm": 0.
|
| 1386 |
-
"learning_rate":
|
| 1387 |
-
"loss": 0.
|
| 1388 |
"step": 197
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
"epoch": 1.036745406824147,
|
| 1392 |
-
"grad_norm": 0.
|
| 1393 |
-
"learning_rate":
|
| 1394 |
-
"loss": 0.
|
| 1395 |
"step": 198
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
"epoch": 1.041994750656168,
|
| 1399 |
-
"grad_norm": 0.
|
| 1400 |
-
"learning_rate":
|
| 1401 |
-
"loss": 0.
|
| 1402 |
"step": 199
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
"epoch": 1.047244094488189,
|
| 1406 |
-
"grad_norm": 0.
|
| 1407 |
-
"learning_rate":
|
| 1408 |
-
"loss": 0.
|
| 1409 |
"step": 200
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
"epoch": 1.05249343832021,
|
| 1413 |
-
"grad_norm": 0.
|
| 1414 |
-
"learning_rate":
|
| 1415 |
-
"loss": 0.
|
| 1416 |
"step": 201
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
"epoch": 1.057742782152231,
|
| 1420 |
-
"grad_norm": 0.
|
| 1421 |
-
"learning_rate":
|
| 1422 |
-
"loss": 0.
|
| 1423 |
"step": 202
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
"epoch": 1.0629921259842519,
|
| 1427 |
-
"grad_norm": 0.
|
| 1428 |
-
"learning_rate":
|
| 1429 |
-
"loss": 0.
|
| 1430 |
"step": 203
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 1.068241469816273,
|
| 1434 |
-
"grad_norm": 0.
|
| 1435 |
-
"learning_rate":
|
| 1436 |
-
"loss": 0.
|
| 1437 |
"step": 204
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
"epoch": 1.073490813648294,
|
| 1441 |
-
"grad_norm": 0.
|
| 1442 |
-
"learning_rate":
|
| 1443 |
-
"loss": 0.
|
| 1444 |
"step": 205
|
| 1445 |
},
|
| 1446 |
{
|
| 1447 |
"epoch": 1.078740157480315,
|
| 1448 |
-
"grad_norm": 0.
|
| 1449 |
-
"learning_rate":
|
| 1450 |
-
"loss": 0.
|
| 1451 |
"step": 206
|
| 1452 |
},
|
| 1453 |
{
|
| 1454 |
"epoch": 1.083989501312336,
|
| 1455 |
-
"grad_norm": 0.
|
| 1456 |
-
"learning_rate":
|
| 1457 |
-
"loss": 0.
|
| 1458 |
"step": 207
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 1.0892388451443569,
|
| 1462 |
-
"grad_norm": 0.
|
| 1463 |
-
"learning_rate":
|
| 1464 |
-
"loss": 0.
|
| 1465 |
"step": 208
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 1.094488188976378,
|
| 1469 |
-
"grad_norm": 0.
|
| 1470 |
-
"learning_rate":
|
| 1471 |
-
"loss": 0.
|
| 1472 |
"step": 209
|
| 1473 |
},
|
| 1474 |
{
|
| 1475 |
"epoch": 1.099737532808399,
|
| 1476 |
-
"grad_norm": 0.
|
| 1477 |
-
"learning_rate":
|
| 1478 |
-
"loss": 0.
|
| 1479 |
"step": 210
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 1.10498687664042,
|
| 1483 |
-
"grad_norm": 0.
|
| 1484 |
-
"learning_rate":
|
| 1485 |
-
"loss": 0.
|
| 1486 |
"step": 211
|
| 1487 |
},
|
| 1488 |
{
|
| 1489 |
"epoch": 1.110236220472441,
|
| 1490 |
-
"grad_norm": 0.
|
| 1491 |
-
"learning_rate":
|
| 1492 |
-
"loss": 0.
|
| 1493 |
"step": 212
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
"epoch": 1.1154855643044619,
|
| 1497 |
-
"grad_norm": 0.
|
| 1498 |
-
"learning_rate":
|
| 1499 |
-
"loss": 0.
|
| 1500 |
"step": 213
|
| 1501 |
},
|
| 1502 |
{
|
| 1503 |
"epoch": 1.120734908136483,
|
| 1504 |
-
"grad_norm": 0.
|
| 1505 |
-
"learning_rate":
|
| 1506 |
-
"loss": 0.
|
| 1507 |
"step": 214
|
| 1508 |
},
|
| 1509 |
{
|
| 1510 |
"epoch": 1.125984251968504,
|
| 1511 |
-
"grad_norm": 0.
|
| 1512 |
-
"learning_rate":
|
| 1513 |
-
"loss": 0.
|
| 1514 |
"step": 215
|
| 1515 |
},
|
| 1516 |
{
|
| 1517 |
"epoch": 1.1312335958005248,
|
| 1518 |
-
"grad_norm": 0.
|
| 1519 |
-
"learning_rate":
|
| 1520 |
-
"loss": 0.
|
| 1521 |
"step": 216
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 1.136482939632546,
|
| 1525 |
-
"grad_norm": 0.
|
| 1526 |
-
"learning_rate":
|
| 1527 |
-
"loss": 0.
|
| 1528 |
"step": 217
|
| 1529 |
},
|
| 1530 |
{
|
| 1531 |
"epoch": 1.141732283464567,
|
| 1532 |
-
"grad_norm": 0.
|
| 1533 |
-
"learning_rate":
|
| 1534 |
-
"loss": 0.
|
| 1535 |
"step": 218
|
| 1536 |
},
|
| 1537 |
{
|
| 1538 |
"epoch": 1.1469816272965878,
|
| 1539 |
-
"grad_norm": 0.
|
| 1540 |
-
"learning_rate":
|
| 1541 |
-
"loss": 0.
|
| 1542 |
"step": 219
|
| 1543 |
},
|
| 1544 |
{
|
| 1545 |
"epoch": 1.152230971128609,
|
| 1546 |
-
"grad_norm": 0.
|
| 1547 |
-
"learning_rate":
|
| 1548 |
-
"loss": 0.
|
| 1549 |
"step": 220
|
| 1550 |
},
|
| 1551 |
{
|
| 1552 |
"epoch": 1.1574803149606299,
|
| 1553 |
-
"grad_norm": 0.
|
| 1554 |
-
"learning_rate":
|
| 1555 |
-
"loss": 0.
|
| 1556 |
"step": 221
|
| 1557 |
},
|
| 1558 |
{
|
| 1559 |
"epoch": 1.162729658792651,
|
| 1560 |
-
"grad_norm": 0.
|
| 1561 |
-
"learning_rate":
|
| 1562 |
-
"loss": 0.
|
| 1563 |
"step": 222
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 1.167979002624672,
|
| 1567 |
-
"grad_norm": 0.
|
| 1568 |
-
"learning_rate":
|
| 1569 |
-
"loss": 0.
|
| 1570 |
"step": 223
|
| 1571 |
},
|
| 1572 |
{
|
| 1573 |
"epoch": 1.1732283464566928,
|
| 1574 |
-
"grad_norm": 0.
|
| 1575 |
-
"learning_rate":
|
| 1576 |
-
"loss": 0.
|
| 1577 |
"step": 224
|
| 1578 |
},
|
| 1579 |
{
|
| 1580 |
"epoch": 1.178477690288714,
|
| 1581 |
-
"grad_norm": 0.
|
| 1582 |
-
"learning_rate":
|
| 1583 |
-
"loss": 0.
|
| 1584 |
"step": 225
|
| 1585 |
},
|
| 1586 |
{
|
| 1587 |
"epoch": 1.1837270341207349,
|
| 1588 |
-
"grad_norm": 0.
|
| 1589 |
-
"learning_rate":
|
| 1590 |
-
"loss": 0.
|
| 1591 |
"step": 226
|
| 1592 |
},
|
| 1593 |
{
|
| 1594 |
"epoch": 1.188976377952756,
|
| 1595 |
-
"grad_norm": 0.
|
| 1596 |
-
"learning_rate":
|
| 1597 |
-
"loss": 0.
|
| 1598 |
"step": 227
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
"epoch": 1.194225721784777,
|
| 1602 |
-
"grad_norm": 0.
|
| 1603 |
-
"learning_rate":
|
| 1604 |
-
"loss": 0.
|
| 1605 |
"step": 228
|
| 1606 |
},
|
| 1607 |
{
|
| 1608 |
"epoch": 1.1994750656167978,
|
| 1609 |
-
"grad_norm": 0.
|
| 1610 |
-
"learning_rate":
|
| 1611 |
-
"loss": 0.
|
| 1612 |
"step": 229
|
| 1613 |
},
|
| 1614 |
{
|
| 1615 |
"epoch": 1.204724409448819,
|
| 1616 |
-
"grad_norm": 0.
|
| 1617 |
-
"learning_rate":
|
| 1618 |
-
"loss": 0.
|
| 1619 |
"step": 230
|
| 1620 |
},
|
| 1621 |
{
|
| 1622 |
"epoch": 1.20997375328084,
|
| 1623 |
-
"grad_norm": 0.
|
| 1624 |
-
"learning_rate":
|
| 1625 |
-
"loss": 0.
|
| 1626 |
"step": 231
|
| 1627 |
},
|
| 1628 |
{
|
| 1629 |
"epoch": 1.2152230971128608,
|
| 1630 |
-
"grad_norm": 0.
|
| 1631 |
-
"learning_rate":
|
| 1632 |
-
"loss": 0.
|
| 1633 |
"step": 232
|
| 1634 |
},
|
| 1635 |
{
|
| 1636 |
"epoch": 1.220472440944882,
|
| 1637 |
-
"grad_norm": 0.
|
| 1638 |
-
"learning_rate":
|
| 1639 |
-
"loss": 0.
|
| 1640 |
"step": 233
|
| 1641 |
},
|
| 1642 |
{
|
| 1643 |
"epoch": 1.2257217847769029,
|
| 1644 |
-
"grad_norm": 0.
|
| 1645 |
-
"learning_rate":
|
| 1646 |
-
"loss": 0.
|
| 1647 |
"step": 234
|
| 1648 |
},
|
| 1649 |
{
|
| 1650 |
"epoch": 1.2309711286089238,
|
| 1651 |
-
"grad_norm": 0.
|
| 1652 |
-
"learning_rate":
|
| 1653 |
-
"loss": 0.
|
| 1654 |
"step": 235
|
| 1655 |
},
|
| 1656 |
{
|
| 1657 |
"epoch": 1.236220472440945,
|
| 1658 |
-
"grad_norm": 0.
|
| 1659 |
-
"learning_rate":
|
| 1660 |
-
"loss": 0.
|
| 1661 |
"step": 236
|
| 1662 |
},
|
| 1663 |
{
|
| 1664 |
"epoch": 1.2414698162729658,
|
| 1665 |
-
"grad_norm": 0.
|
| 1666 |
-
"learning_rate":
|
| 1667 |
-
"loss": 0.
|
| 1668 |
"step": 237
|
| 1669 |
},
|
| 1670 |
{
|
| 1671 |
"epoch": 1.246719160104987,
|
| 1672 |
-
"grad_norm": 0.
|
| 1673 |
-
"learning_rate":
|
| 1674 |
-
"loss": 0.
|
| 1675 |
"step": 238
|
| 1676 |
},
|
| 1677 |
{
|
| 1678 |
"epoch": 1.2519685039370079,
|
| 1679 |
-
"grad_norm": 0.
|
| 1680 |
-
"learning_rate":
|
| 1681 |
-
"loss": 0.
|
| 1682 |
"step": 239
|
| 1683 |
},
|
| 1684 |
{
|
| 1685 |
"epoch": 1.257217847769029,
|
| 1686 |
-
"grad_norm": 0.
|
| 1687 |
-
"learning_rate":
|
| 1688 |
-
"loss": 0.
|
| 1689 |
"step": 240
|
| 1690 |
},
|
| 1691 |
{
|
| 1692 |
"epoch": 1.26246719160105,
|
| 1693 |
-
"grad_norm": 0.
|
| 1694 |
-
"learning_rate":
|
| 1695 |
-
"loss": 0.
|
| 1696 |
"step": 241
|
| 1697 |
},
|
| 1698 |
{
|
| 1699 |
"epoch": 1.2677165354330708,
|
| 1700 |
-
"grad_norm": 0.
|
| 1701 |
-
"learning_rate":
|
| 1702 |
-
"loss": 0.
|
| 1703 |
"step": 242
|
| 1704 |
},
|
| 1705 |
{
|
| 1706 |
"epoch": 1.272965879265092,
|
| 1707 |
-
"grad_norm": 0.
|
| 1708 |
-
"learning_rate":
|
| 1709 |
-
"loss": 0.
|
| 1710 |
"step": 243
|
| 1711 |
},
|
| 1712 |
{
|
| 1713 |
"epoch": 1.2782152230971129,
|
| 1714 |
-
"grad_norm": 0.
|
| 1715 |
-
"learning_rate":
|
| 1716 |
-
"loss": 0.
|
| 1717 |
"step": 244
|
| 1718 |
},
|
| 1719 |
{
|
| 1720 |
"epoch": 1.2834645669291338,
|
| 1721 |
-
"grad_norm": 0.
|
| 1722 |
-
"learning_rate":
|
| 1723 |
-
"loss": 0.
|
| 1724 |
"step": 245
|
| 1725 |
},
|
| 1726 |
{
|
| 1727 |
"epoch": 1.288713910761155,
|
| 1728 |
-
"grad_norm": 0.
|
| 1729 |
-
"learning_rate":
|
| 1730 |
-
"loss": 0.
|
| 1731 |
"step": 246
|
| 1732 |
},
|
| 1733 |
{
|
| 1734 |
"epoch": 1.2939632545931758,
|
| 1735 |
-
"grad_norm": 0.
|
| 1736 |
-
"learning_rate": 1.
|
| 1737 |
-
"loss": 0.
|
| 1738 |
"step": 247
|
| 1739 |
},
|
| 1740 |
{
|
| 1741 |
"epoch": 1.2992125984251968,
|
| 1742 |
-
"grad_norm": 0.
|
| 1743 |
-
"learning_rate": 1.
|
| 1744 |
-
"loss": 0.
|
| 1745 |
"step": 248
|
| 1746 |
},
|
| 1747 |
{
|
| 1748 |
"epoch": 1.304461942257218,
|
| 1749 |
-
"grad_norm": 0.
|
| 1750 |
-
"learning_rate": 1.
|
| 1751 |
-
"loss": 0.
|
| 1752 |
"step": 249
|
| 1753 |
},
|
| 1754 |
{
|
| 1755 |
"epoch": 1.3097112860892388,
|
| 1756 |
-
"grad_norm": 0.
|
| 1757 |
-
"learning_rate": 1.
|
| 1758 |
-
"loss": 0.
|
| 1759 |
"step": 250
|
| 1760 |
},
|
| 1761 |
{
|
| 1762 |
"epoch": 1.3149606299212597,
|
| 1763 |
-
"grad_norm": 0.
|
| 1764 |
-
"learning_rate": 1.
|
| 1765 |
-
"loss": 0.
|
| 1766 |
"step": 251
|
| 1767 |
},
|
| 1768 |
{
|
| 1769 |
"epoch": 1.3202099737532809,
|
| 1770 |
-
"grad_norm": 0.
|
| 1771 |
-
"learning_rate": 1.
|
| 1772 |
-
"loss": 0.
|
| 1773 |
"step": 252
|
| 1774 |
},
|
| 1775 |
{
|
| 1776 |
"epoch": 1.3254593175853018,
|
| 1777 |
-
"grad_norm": 0.
|
| 1778 |
-
"learning_rate": 1.
|
| 1779 |
-
"loss": 0.
|
| 1780 |
"step": 253
|
| 1781 |
},
|
| 1782 |
{
|
| 1783 |
"epoch": 1.330708661417323,
|
| 1784 |
-
"grad_norm": 0.
|
| 1785 |
-
"learning_rate": 1.
|
| 1786 |
-
"loss": 0.
|
| 1787 |
"step": 254
|
| 1788 |
},
|
| 1789 |
{
|
| 1790 |
"epoch": 1.3359580052493438,
|
| 1791 |
-
"grad_norm": 0.
|
| 1792 |
-
"learning_rate": 1.
|
| 1793 |
-
"loss": 0.
|
| 1794 |
"step": 255
|
| 1795 |
},
|
| 1796 |
{
|
| 1797 |
"epoch": 1.341207349081365,
|
| 1798 |
-
"grad_norm": 0.
|
| 1799 |
-
"learning_rate": 1.
|
| 1800 |
-
"loss": 0.
|
| 1801 |
"step": 256
|
| 1802 |
},
|
| 1803 |
{
|
| 1804 |
"epoch": 1.3464566929133859,
|
| 1805 |
-
"grad_norm": 0.
|
| 1806 |
-
"learning_rate": 1.
|
| 1807 |
-
"loss": 0.
|
| 1808 |
"step": 257
|
| 1809 |
},
|
| 1810 |
{
|
| 1811 |
"epoch": 1.3517060367454068,
|
| 1812 |
-
"grad_norm": 0.
|
| 1813 |
-
"learning_rate": 1.
|
| 1814 |
-
"loss": 0.
|
| 1815 |
"step": 258
|
| 1816 |
},
|
| 1817 |
{
|
| 1818 |
"epoch": 1.356955380577428,
|
| 1819 |
-
"grad_norm": 0.
|
| 1820 |
-
"learning_rate": 1.
|
| 1821 |
-
"loss": 0.
|
| 1822 |
"step": 259
|
| 1823 |
},
|
| 1824 |
{
|
| 1825 |
"epoch": 1.3622047244094488,
|
| 1826 |
-
"grad_norm": 0.
|
| 1827 |
-
"learning_rate": 1.
|
| 1828 |
-
"loss": 0.
|
| 1829 |
"step": 260
|
| 1830 |
},
|
| 1831 |
{
|
| 1832 |
"epoch": 1.3674540682414698,
|
| 1833 |
-
"grad_norm": 0.
|
| 1834 |
-
"learning_rate": 1.
|
| 1835 |
-
"loss": 0.
|
| 1836 |
"step": 261
|
| 1837 |
},
|
| 1838 |
{
|
| 1839 |
"epoch": 1.372703412073491,
|
| 1840 |
-
"grad_norm": 0.
|
| 1841 |
-
"learning_rate": 1.
|
| 1842 |
-
"loss": 0.
|
| 1843 |
"step": 262
|
| 1844 |
},
|
| 1845 |
{
|
| 1846 |
"epoch": 1.3779527559055118,
|
| 1847 |
-
"grad_norm": 0.
|
| 1848 |
-
"learning_rate": 1.
|
| 1849 |
-
"loss": 0.
|
| 1850 |
"step": 263
|
| 1851 |
},
|
| 1852 |
{
|
| 1853 |
"epoch": 1.3832020997375327,
|
| 1854 |
-
"grad_norm": 0.
|
| 1855 |
-
"learning_rate": 1.
|
| 1856 |
-
"loss": 0.
|
| 1857 |
"step": 264
|
| 1858 |
},
|
| 1859 |
{
|
| 1860 |
"epoch": 1.3884514435695539,
|
| 1861 |
-
"grad_norm": 0.
|
| 1862 |
-
"learning_rate": 1.
|
| 1863 |
-
"loss": 0.
|
| 1864 |
"step": 265
|
| 1865 |
},
|
| 1866 |
{
|
| 1867 |
"epoch": 1.3937007874015748,
|
| 1868 |
-
"grad_norm": 0.
|
| 1869 |
-
"learning_rate": 1.
|
| 1870 |
-
"loss": 0.
|
| 1871 |
"step": 266
|
| 1872 |
},
|
| 1873 |
{
|
| 1874 |
"epoch": 1.3989501312335957,
|
| 1875 |
-
"grad_norm": 0.
|
| 1876 |
-
"learning_rate":
|
| 1877 |
-
"loss": 0.
|
| 1878 |
"step": 267
|
| 1879 |
},
|
| 1880 |
{
|
| 1881 |
"epoch": 1.4041994750656168,
|
| 1882 |
-
"grad_norm": 0.
|
| 1883 |
-
"learning_rate":
|
| 1884 |
-
"loss": 0.
|
| 1885 |
"step": 268
|
| 1886 |
},
|
| 1887 |
{
|
| 1888 |
"epoch": 1.4094488188976377,
|
| 1889 |
-
"grad_norm": 0.
|
| 1890 |
-
"learning_rate":
|
| 1891 |
-
"loss": 0.
|
| 1892 |
"step": 269
|
| 1893 |
},
|
| 1894 |
{
|
| 1895 |
"epoch": 1.4146981627296589,
|
| 1896 |
-
"grad_norm": 0.
|
| 1897 |
-
"learning_rate":
|
| 1898 |
-
"loss": 0.
|
| 1899 |
"step": 270
|
| 1900 |
},
|
| 1901 |
{
|
| 1902 |
"epoch": 1.4199475065616798,
|
| 1903 |
-
"grad_norm": 0.
|
| 1904 |
-
"learning_rate":
|
| 1905 |
-
"loss": 0.
|
| 1906 |
"step": 271
|
| 1907 |
},
|
| 1908 |
{
|
| 1909 |
"epoch": 1.425196850393701,
|
| 1910 |
-
"grad_norm": 0.
|
| 1911 |
-
"learning_rate":
|
| 1912 |
-
"loss": 0.
|
| 1913 |
"step": 272
|
| 1914 |
},
|
| 1915 |
{
|
| 1916 |
"epoch": 1.4304461942257218,
|
| 1917 |
-
"grad_norm": 0.
|
| 1918 |
-
"learning_rate":
|
| 1919 |
-
"loss": 0.
|
| 1920 |
"step": 273
|
| 1921 |
},
|
| 1922 |
{
|
| 1923 |
"epoch": 1.4356955380577427,
|
| 1924 |
-
"grad_norm": 0.
|
| 1925 |
-
"learning_rate":
|
| 1926 |
-
"loss": 0.
|
| 1927 |
"step": 274
|
| 1928 |
},
|
| 1929 |
{
|
| 1930 |
"epoch": 1.4409448818897639,
|
| 1931 |
-
"grad_norm": 0.
|
| 1932 |
-
"learning_rate":
|
| 1933 |
-
"loss": 0.
|
| 1934 |
"step": 275
|
| 1935 |
},
|
| 1936 |
{
|
| 1937 |
"epoch": 1.4461942257217848,
|
| 1938 |
-
"grad_norm": 0.
|
| 1939 |
-
"learning_rate":
|
| 1940 |
-
"loss": 0.
|
| 1941 |
"step": 276
|
| 1942 |
},
|
| 1943 |
{
|
| 1944 |
"epoch": 1.4514435695538057,
|
| 1945 |
-
"grad_norm": 0.
|
| 1946 |
-
"learning_rate":
|
| 1947 |
-
"loss": 0.
|
| 1948 |
"step": 277
|
| 1949 |
},
|
| 1950 |
{
|
| 1951 |
"epoch": 1.4566929133858268,
|
| 1952 |
-
"grad_norm": 0.
|
| 1953 |
-
"learning_rate":
|
| 1954 |
-
"loss": 0.
|
| 1955 |
"step": 278
|
| 1956 |
},
|
| 1957 |
{
|
| 1958 |
"epoch": 1.4619422572178478,
|
| 1959 |
-
"grad_norm": 0.
|
| 1960 |
-
"learning_rate":
|
| 1961 |
-
"loss": 0.
|
| 1962 |
"step": 279
|
| 1963 |
},
|
| 1964 |
{
|
| 1965 |
"epoch": 1.4671916010498687,
|
| 1966 |
-
"grad_norm": 0.
|
| 1967 |
-
"learning_rate":
|
| 1968 |
-
"loss": 0.
|
| 1969 |
"step": 280
|
| 1970 |
},
|
| 1971 |
{
|
| 1972 |
"epoch": 1.4724409448818898,
|
| 1973 |
-
"grad_norm": 0.
|
| 1974 |
-
"learning_rate":
|
| 1975 |
-
"loss": 0.
|
| 1976 |
"step": 281
|
| 1977 |
},
|
| 1978 |
{
|
| 1979 |
"epoch": 1.4776902887139107,
|
| 1980 |
-
"grad_norm": 0.
|
| 1981 |
-
"learning_rate":
|
| 1982 |
-
"loss": 0.
|
| 1983 |
"step": 282
|
| 1984 |
},
|
| 1985 |
{
|
| 1986 |
"epoch": 1.4829396325459316,
|
| 1987 |
-
"grad_norm": 0.
|
| 1988 |
-
"learning_rate":
|
| 1989 |
-
"loss": 0.
|
| 1990 |
"step": 283
|
| 1991 |
},
|
| 1992 |
{
|
| 1993 |
"epoch": 1.4881889763779528,
|
| 1994 |
-
"grad_norm": 0.
|
| 1995 |
-
"learning_rate":
|
| 1996 |
-
"loss": 0.
|
| 1997 |
"step": 284
|
| 1998 |
},
|
| 1999 |
{
|
| 2000 |
"epoch": 1.4934383202099737,
|
| 2001 |
-
"grad_norm": 0.
|
| 2002 |
-
"learning_rate":
|
| 2003 |
-
"loss": 0.
|
| 2004 |
"step": 285
|
| 2005 |
}
|
| 2006 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.005249343832020997,
|
| 13 |
+
"grad_norm": 1.1348930782232016,
|
| 14 |
+
"learning_rate": 1.5000000000000002e-07,
|
| 15 |
"loss": 1.1087,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.010498687664041995,
|
| 20 |
+
"grad_norm": 1.123696373079589,
|
| 21 |
+
"learning_rate": 3.0000000000000004e-07,
|
| 22 |
"loss": 1.1356,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.015748031496062992,
|
| 27 |
+
"grad_norm": 1.0989081863562118,
|
| 28 |
+
"learning_rate": 4.5e-07,
|
| 29 |
+
"loss": 1.1158,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.02099737532808399,
|
| 34 |
+
"grad_norm": 1.0628548113414964,
|
| 35 |
+
"learning_rate": 6.000000000000001e-07,
|
| 36 |
+
"loss": 1.0986,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.026246719160104987,
|
| 41 |
+
"grad_norm": 1.0629069543612368,
|
| 42 |
+
"learning_rate": 7.5e-07,
|
| 43 |
+
"loss": 1.0727,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.031496062992125984,
|
| 48 |
+
"grad_norm": 1.1219311917213644,
|
| 49 |
+
"learning_rate": 9e-07,
|
| 50 |
+
"loss": 1.1513,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.03674540682414698,
|
| 55 |
+
"grad_norm": 1.068318638334139,
|
| 56 |
+
"learning_rate": 1.05e-06,
|
| 57 |
+
"loss": 1.0978,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04199475065616798,
|
| 62 |
+
"grad_norm": 1.0335025624008565,
|
| 63 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 64 |
+
"loss": 1.0932,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.047244094488188976,
|
| 69 |
+
"grad_norm": 0.9514112971268772,
|
| 70 |
+
"learning_rate": 1.35e-06,
|
| 71 |
+
"loss": 1.1046,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.05249343832020997,
|
| 76 |
+
"grad_norm": 0.8944230714776324,
|
| 77 |
+
"learning_rate": 1.5e-06,
|
| 78 |
+
"loss": 1.0638,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.05774278215223097,
|
| 83 |
+
"grad_norm": 0.8720343077794245,
|
| 84 |
+
"learning_rate": 1.65e-06,
|
| 85 |
+
"loss": 1.1132,
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.06299212598425197,
|
| 90 |
+
"grad_norm": 0.7519518665820406,
|
| 91 |
+
"learning_rate": 1.8e-06,
|
| 92 |
+
"loss": 1.0788,
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.06824146981627296,
|
| 97 |
+
"grad_norm": 0.7768466543241798,
|
| 98 |
+
"learning_rate": 1.95e-06,
|
| 99 |
+
"loss": 1.0795,
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.07349081364829396,
|
| 104 |
+
"grad_norm": 0.7109922479048013,
|
| 105 |
+
"learning_rate": 2.1e-06,
|
| 106 |
+
"loss": 1.1012,
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.07874015748031496,
|
| 111 |
+
"grad_norm": 0.6312078880187205,
|
| 112 |
+
"learning_rate": 2.25e-06,
|
| 113 |
+
"loss": 1.0851,
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.08398950131233596,
|
| 118 |
+
"grad_norm": 0.5514473048370377,
|
| 119 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 120 |
+
"loss": 1.1041,
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.08923884514435695,
|
| 125 |
+
"grad_norm": 0.6271281070432462,
|
| 126 |
+
"learning_rate": 2.55e-06,
|
| 127 |
+
"loss": 1.0855,
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.09448818897637795,
|
| 132 |
+
"grad_norm": 0.7059888078645049,
|
| 133 |
+
"learning_rate": 2.7e-06,
|
| 134 |
+
"loss": 1.0473,
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.09973753280839895,
|
| 139 |
+
"grad_norm": 0.7226157330393405,
|
| 140 |
+
"learning_rate": 2.85e-06,
|
| 141 |
+
"loss": 1.0665,
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.10498687664041995,
|
| 146 |
+
"grad_norm": 0.7244742832208652,
|
| 147 |
+
"learning_rate": 3e-06,
|
| 148 |
+
"loss": 1.0604,
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.11023622047244094,
|
| 153 |
+
"grad_norm": 0.7088251146482789,
|
| 154 |
+
"learning_rate": 3.1500000000000003e-06,
|
| 155 |
+
"loss": 1.0516,
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.11548556430446194,
|
| 160 |
+
"grad_norm": 0.5987242362229293,
|
| 161 |
+
"learning_rate": 3.3e-06,
|
| 162 |
+
"loss": 1.084,
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.12073490813648294,
|
| 167 |
+
"grad_norm": 0.5730637810768702,
|
| 168 |
+
"learning_rate": 3.45e-06,
|
| 169 |
+
"loss": 1.0621,
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.12598425196850394,
|
| 174 |
+
"grad_norm": 0.5894968443138215,
|
| 175 |
+
"learning_rate": 3.6e-06,
|
| 176 |
+
"loss": 1.0797,
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.13123359580052493,
|
| 181 |
+
"grad_norm": 0.5798124303184627,
|
| 182 |
+
"learning_rate": 3.75e-06,
|
| 183 |
+
"loss": 1.0035,
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.13648293963254593,
|
| 188 |
+
"grad_norm": 0.643205751513686,
|
| 189 |
+
"learning_rate": 3.9e-06,
|
| 190 |
+
"loss": 1.0455,
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.14173228346456693,
|
| 195 |
+
"grad_norm": 0.5621970774702022,
|
| 196 |
+
"learning_rate": 4.05e-06,
|
| 197 |
+
"loss": 1.0576,
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.14698162729658792,
|
| 202 |
+
"grad_norm": 0.5506084571895594,
|
| 203 |
+
"learning_rate": 4.2e-06,
|
| 204 |
+
"loss": 1.0298,
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.15223097112860892,
|
| 209 |
+
"grad_norm": 0.48741149421912777,
|
| 210 |
+
"learning_rate": 4.35e-06,
|
| 211 |
+
"loss": 1.0018,
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.15748031496062992,
|
| 216 |
+
"grad_norm": 0.46403007703544275,
|
| 217 |
+
"learning_rate": 4.5e-06,
|
| 218 |
+
"loss": 0.9872,
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.16272965879265092,
|
| 223 |
+
"grad_norm": 0.4754381818573106,
|
| 224 |
+
"learning_rate": 4.65e-06,
|
| 225 |
+
"loss": 1.0271,
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.1679790026246719,
|
| 230 |
+
"grad_norm": 0.9362850890979981,
|
| 231 |
+
"learning_rate": 4.800000000000001e-06,
|
| 232 |
+
"loss": 1.0437,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.1732283464566929,
|
| 237 |
+
"grad_norm": 0.47391181595772164,
|
| 238 |
+
"learning_rate": 4.95e-06,
|
| 239 |
+
"loss": 1.0437,
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.1784776902887139,
|
| 244 |
+
"grad_norm": 0.5276920454851337,
|
| 245 |
+
"learning_rate": 5.1e-06,
|
| 246 |
+
"loss": 1.0557,
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.1837270341207349,
|
| 251 |
+
"grad_norm": 0.4616075133913133,
|
| 252 |
+
"learning_rate": 5.2500000000000006e-06,
|
| 253 |
+
"loss": 1.0465,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.1889763779527559,
|
| 258 |
+
"grad_norm": 0.4555174555636226,
|
| 259 |
+
"learning_rate": 5.4e-06,
|
| 260 |
+
"loss": 1.0588,
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.1942257217847769,
|
| 265 |
+
"grad_norm": 0.5071864534648831,
|
| 266 |
+
"learning_rate": 5.55e-06,
|
| 267 |
+
"loss": 1.044,
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1994750656167979,
|
| 272 |
+
"grad_norm": 0.4851367263882934,
|
| 273 |
+
"learning_rate": 5.7e-06,
|
| 274 |
+
"loss": 1.0464,
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2047244094488189,
|
| 279 |
+
"grad_norm": 0.44188022228811896,
|
| 280 |
+
"learning_rate": 5.85e-06,
|
| 281 |
+
"loss": 1.0182,
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.2099737532808399,
|
| 286 |
+
"grad_norm": 0.43420740120454643,
|
| 287 |
+
"learning_rate": 6e-06,
|
| 288 |
+
"loss": 1.0188,
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.2152230971128609,
|
| 293 |
+
"grad_norm": 0.4291543441241407,
|
| 294 |
+
"learning_rate": 5.9998719351101036e-06,
|
| 295 |
+
"loss": 1.0245,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.2204724409448819,
|
| 300 |
+
"grad_norm": 0.43326370236005163,
|
| 301 |
+
"learning_rate": 5.999487751374158e-06,
|
| 302 |
+
"loss": 1.0238,
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.22572178477690288,
|
| 307 |
+
"grad_norm": 0.427571644972227,
|
| 308 |
+
"learning_rate": 5.998847481592462e-06,
|
| 309 |
+
"loss": 1.0311,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.23097112860892388,
|
| 314 |
+
"grad_norm": 0.4215063088273006,
|
| 315 |
+
"learning_rate": 5.997951180429069e-06,
|
| 316 |
+
"loss": 0.9925,
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.23622047244094488,
|
| 321 |
+
"grad_norm": 0.4206536914503675,
|
| 322 |
+
"learning_rate": 5.996798924407118e-06,
|
| 323 |
+
"loss": 1.003,
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.24146981627296588,
|
| 328 |
+
"grad_norm": 0.40910969064965136,
|
| 329 |
+
"learning_rate": 5.995390811902302e-06,
|
| 330 |
+
"loss": 0.9949,
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.24671916010498687,
|
| 335 |
+
"grad_norm": 0.4165775049327623,
|
| 336 |
+
"learning_rate": 5.993726963134471e-06,
|
| 337 |
+
"loss": 0.9734,
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.25196850393700787,
|
| 342 |
+
"grad_norm": 0.3832235501001726,
|
| 343 |
+
"learning_rate": 5.9918075201573645e-06,
|
| 344 |
+
"loss": 0.9485,
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.2572178477690289,
|
| 349 |
+
"grad_norm": 0.37002495168808525,
|
| 350 |
+
"learning_rate": 5.9896326468464835e-06,
|
| 351 |
+
"loss": 0.9358,
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.26246719160104987,
|
| 356 |
+
"grad_norm": 0.44836853406053057,
|
| 357 |
+
"learning_rate": 5.987202528885104e-06,
|
| 358 |
+
"loss": 0.9982,
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.2677165354330709,
|
| 363 |
+
"grad_norm": 0.4080608606117312,
|
| 364 |
+
"learning_rate": 5.984517373748417e-06,
|
| 365 |
+
"loss": 1.0129,
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.27296587926509186,
|
| 370 |
+
"grad_norm": 0.4001550595702573,
|
| 371 |
+
"learning_rate": 5.981577410685822e-06,
|
| 372 |
+
"loss": 0.9788,
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.2782152230971129,
|
| 377 |
+
"grad_norm": 0.41021488877460305,
|
| 378 |
+
"learning_rate": 5.978382890701347e-06,
|
| 379 |
+
"loss": 1.0262,
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.28346456692913385,
|
| 384 |
+
"grad_norm": 0.39997016380492506,
|
| 385 |
+
"learning_rate": 5.9749340865322284e-06,
|
| 386 |
+
"loss": 1.0275,
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.2887139107611549,
|
| 391 |
+
"grad_norm": 0.3839823787027912,
|
| 392 |
+
"learning_rate": 5.971231292625615e-06,
|
| 393 |
+
"loss": 0.9374,
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.29396325459317585,
|
| 398 |
+
"grad_norm": 0.4125068495663659,
|
| 399 |
+
"learning_rate": 5.967274825113438e-06,
|
| 400 |
+
"loss": 0.9954,
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2992125984251969,
|
| 405 |
+
"grad_norm": 0.3908377197765856,
|
| 406 |
+
"learning_rate": 5.963065021785414e-06,
|
| 407 |
+
"loss": 0.9671,
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.30446194225721784,
|
| 412 |
+
"grad_norm": 0.3850488592862481,
|
| 413 |
+
"learning_rate": 5.958602242060207e-06,
|
| 414 |
+
"loss": 0.9657,
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.30971128608923887,
|
| 419 |
+
"grad_norm": 0.3877990366088493,
|
| 420 |
+
"learning_rate": 5.95388686695475e-06,
|
| 421 |
+
"loss": 0.9678,
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.31496062992125984,
|
| 426 |
+
"grad_norm": 0.40470471194287355,
|
| 427 |
+
"learning_rate": 5.948919299051706e-06,
|
| 428 |
+
"loss": 1.0149,
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.32020997375328086,
|
| 433 |
+
"grad_norm": 0.42889495063392963,
|
| 434 |
+
"learning_rate": 5.943699962465096e-06,
|
| 435 |
+
"loss": 1.033,
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.32545931758530183,
|
| 440 |
+
"grad_norm": 0.39164358737100274,
|
| 441 |
+
"learning_rate": 5.9382293028040985e-06,
|
| 442 |
+
"loss": 0.9761,
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.33070866141732286,
|
| 447 |
+
"grad_norm": 0.3869342590567232,
|
| 448 |
+
"learning_rate": 5.9325077871349975e-06,
|
| 449 |
+
"loss": 0.9982,
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.3359580052493438,
|
| 454 |
+
"grad_norm": 0.39264627926569035,
|
| 455 |
+
"learning_rate": 5.9265359039413105e-06,
|
| 456 |
+
"loss": 0.9667,
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.34120734908136485,
|
| 461 |
+
"grad_norm": 0.3887717698297268,
|
| 462 |
+
"learning_rate": 5.920314163082079e-06,
|
| 463 |
+
"loss": 0.9806,
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.3464566929133858,
|
| 468 |
+
"grad_norm": 0.40896336915084297,
|
| 469 |
+
"learning_rate": 5.913843095748342e-06,
|
| 470 |
+
"loss": 1.0135,
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.35170603674540685,
|
| 475 |
+
"grad_norm": 0.3610209560875707,
|
| 476 |
+
"learning_rate": 5.907123254417783e-06,
|
| 477 |
+
"loss": 0.956,
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.3569553805774278,
|
| 482 |
+
"grad_norm": 0.38154744815823505,
|
| 483 |
+
"learning_rate": 5.9001552128075625e-06,
|
| 484 |
+
"loss": 1.0045,
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.36220472440944884,
|
| 489 |
+
"grad_norm": 0.4094826396119445,
|
| 490 |
+
"learning_rate": 5.892939565825335e-06,
|
| 491 |
+
"loss": 1.0069,
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.3674540682414698,
|
| 496 |
+
"grad_norm": 0.39129138622932325,
|
| 497 |
+
"learning_rate": 5.885476929518457e-06,
|
| 498 |
+
"loss": 0.9525,
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.37270341207349084,
|
| 503 |
+
"grad_norm": 0.3712890701175899,
|
| 504 |
+
"learning_rate": 5.8777679410213956e-06,
|
| 505 |
+
"loss": 0.9792,
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.3779527559055118,
|
| 510 |
+
"grad_norm": 0.4086264062600148,
|
| 511 |
+
"learning_rate": 5.869813258501323e-06,
|
| 512 |
+
"loss": 0.9926,
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.38320209973753283,
|
| 517 |
+
"grad_norm": 0.368975878599487,
|
| 518 |
+
"learning_rate": 5.861613561101934e-06,
|
| 519 |
+
"loss": 0.9643,
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.3884514435695538,
|
| 524 |
+
"grad_norm": 0.36792811629461203,
|
| 525 |
+
"learning_rate": 5.853169548885461e-06,
|
| 526 |
+
"loss": 0.9867,
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.3937007874015748,
|
| 531 |
+
"grad_norm": 0.3566251893981936,
|
| 532 |
+
"learning_rate": 5.844481942772898e-06,
|
| 533 |
+
"loss": 1.0069,
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3989501312335958,
|
| 538 |
+
"grad_norm": 0.4578529359685586,
|
| 539 |
+
"learning_rate": 5.835551484482459e-06,
|
| 540 |
+
"loss": 1.0173,
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.4041994750656168,
|
| 545 |
+
"grad_norm": 0.3935925285922137,
|
| 546 |
+
"learning_rate": 5.826378936466249e-06,
|
| 547 |
+
"loss": 0.9743,
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.4094488188976378,
|
| 552 |
+
"grad_norm": 0.4109939217838428,
|
| 553 |
+
"learning_rate": 5.81696508184517e-06,
|
| 554 |
+
"loss": 0.9866,
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.4146981627296588,
|
| 559 |
+
"grad_norm": 0.3839870332489822,
|
| 560 |
+
"learning_rate": 5.807310724342058e-06,
|
| 561 |
+
"loss": 0.9516,
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.4199475065616798,
|
| 566 |
+
"grad_norm": 0.3774576797883406,
|
| 567 |
+
"learning_rate": 5.797416688213067e-06,
|
| 568 |
+
"loss": 0.9895,
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.4251968503937008,
|
| 573 |
+
"grad_norm": 0.3817468964498129,
|
| 574 |
+
"learning_rate": 5.787283818177297e-06,
|
| 575 |
+
"loss": 0.9632,
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.4304461942257218,
|
| 580 |
+
"grad_norm": 0.60843002346461,
|
| 581 |
+
"learning_rate": 5.776912979344669e-06,
|
| 582 |
+
"loss": 1.0166,
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.4356955380577428,
|
| 587 |
+
"grad_norm": 0.3858713700245362,
|
| 588 |
+
"learning_rate": 5.766305057142073e-06,
|
| 589 |
+
"loss": 0.9976,
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.4409448818897638,
|
| 594 |
+
"grad_norm": 0.3724153436541016,
|
| 595 |
+
"learning_rate": 5.755460957237769e-06,
|
| 596 |
+
"loss": 0.9645,
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.4461942257217848,
|
| 601 |
+
"grad_norm": 0.38201105695018567,
|
| 602 |
+
"learning_rate": 5.744381605464064e-06,
|
| 603 |
+
"loss": 0.9899,
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.45144356955380577,
|
| 608 |
+
"grad_norm": 0.38383930861007165,
|
| 609 |
+
"learning_rate": 5.7330679477382655e-06,
|
| 610 |
+
"loss": 0.9919,
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.4566929133858268,
|
| 615 |
+
"grad_norm": 0.4078870418259581,
|
| 616 |
+
"learning_rate": 5.7215209499819296e-06,
|
| 617 |
+
"loss": 0.9797,
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.46194225721784776,
|
| 622 |
+
"grad_norm": 0.38463767466523974,
|
| 623 |
+
"learning_rate": 5.709741598038387e-06,
|
| 624 |
+
"loss": 0.9597,
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.4671916010498688,
|
| 629 |
+
"grad_norm": 0.36309855116472584,
|
| 630 |
+
"learning_rate": 5.697730897588577e-06,
|
| 631 |
+
"loss": 0.9737,
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.47244094488188976,
|
| 636 |
+
"grad_norm": 0.4106701446638758,
|
| 637 |
+
"learning_rate": 5.685489874065187e-06,
|
| 638 |
+
"loss": 0.9683,
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.4776902887139108,
|
| 643 |
+
"grad_norm": 0.37110409255145443,
|
| 644 |
+
"learning_rate": 5.673019572565103e-06,
|
| 645 |
+
"loss": 1.0418,
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.48293963254593175,
|
| 650 |
+
"grad_norm": 0.3558357783330656,
|
| 651 |
+
"learning_rate": 5.660321057760186e-06,
|
| 652 |
+
"loss": 1.0055,
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.4881889763779528,
|
| 657 |
+
"grad_norm": 0.40499489938404787,
|
| 658 |
+
"learning_rate": 5.6473954138063674e-06,
|
| 659 |
+
"loss": 1.0113,
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.49343832020997375,
|
| 664 |
+
"grad_norm": 0.39428526462199764,
|
| 665 |
+
"learning_rate": 5.634243744251094e-06,
|
| 666 |
+
"loss": 0.9875,
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.49868766404199477,
|
| 671 |
+
"grad_norm": 0.3711741011240413,
|
| 672 |
+
"learning_rate": 5.620867171939109e-06,
|
| 673 |
+
"loss": 0.9749,
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.5039370078740157,
|
| 678 |
+
"grad_norm": 0.3961340085644134,
|
| 679 |
+
"learning_rate": 5.607266838916585e-06,
|
| 680 |
+
"loss": 0.982,
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.5091863517060368,
|
| 685 |
+
"grad_norm": 0.3784646685814138,
|
| 686 |
+
"learning_rate": 5.593443906333624e-06,
|
| 687 |
+
"loss": 0.9957,
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.5144356955380578,
|
| 692 |
+
"grad_norm": 0.3750460397069026,
|
| 693 |
+
"learning_rate": 5.579399554345118e-06,
|
| 694 |
+
"loss": 0.9755,
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.5196850393700787,
|
| 699 |
+
"grad_norm": 0.3746718538274792,
|
| 700 |
+
"learning_rate": 5.565134982009994e-06,
|
| 701 |
+
"loss": 0.9736,
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.5249343832020997,
|
| 706 |
+
"grad_norm": 0.38418890409196027,
|
| 707 |
+
"learning_rate": 5.550651407188843e-06,
|
| 708 |
+
"loss": 0.9506,
|
| 709 |
"step": 100
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.5301837270341208,
|
| 713 |
+
"grad_norm": 0.422976375435725,
|
| 714 |
+
"learning_rate": 5.535950066439941e-06,
|
| 715 |
+
"loss": 1.0141,
|
| 716 |
"step": 101
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.5354330708661418,
|
| 720 |
+
"grad_norm": 0.38354451243133536,
|
| 721 |
+
"learning_rate": 5.521032214913679e-06,
|
| 722 |
+
"loss": 0.9618,
|
| 723 |
"step": 102
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.5406824146981627,
|
| 727 |
+
"grad_norm": 0.38257660011773076,
|
| 728 |
+
"learning_rate": 5.505899126245397e-06,
|
| 729 |
+
"loss": 0.939,
|
| 730 |
"step": 103
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.5459317585301837,
|
| 734 |
+
"grad_norm": 0.3768438915225408,
|
| 735 |
+
"learning_rate": 5.490552092446652e-06,
|
| 736 |
+
"loss": 0.9675,
|
| 737 |
"step": 104
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.5511811023622047,
|
| 741 |
+
"grad_norm": 0.3749655286727107,
|
| 742 |
+
"learning_rate": 5.474992423794907e-06,
|
| 743 |
+
"loss": 0.9592,
|
| 744 |
"step": 105
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.5564304461942258,
|
| 748 |
+
"grad_norm": 0.38461916993489687,
|
| 749 |
+
"learning_rate": 5.459221448721664e-06,
|
| 750 |
+
"loss": 0.9623,
|
| 751 |
"step": 106
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.5616797900262467,
|
| 755 |
+
"grad_norm": 0.35648642966931204,
|
| 756 |
+
"learning_rate": 5.443240513699045e-06,
|
| 757 |
+
"loss": 0.985,
|
| 758 |
"step": 107
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.5669291338582677,
|
| 762 |
+
"grad_norm": 0.4051560712719681,
|
| 763 |
+
"learning_rate": 5.427050983124842e-06,
|
| 764 |
+
"loss": 0.9407,
|
| 765 |
"step": 108
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.5721784776902887,
|
| 769 |
+
"grad_norm": 0.3769879713701903,
|
| 770 |
+
"learning_rate": 5.410654239206021e-06,
|
| 771 |
+
"loss": 0.968,
|
| 772 |
"step": 109
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.5774278215223098,
|
| 776 |
+
"grad_norm": 0.3746822083724367,
|
| 777 |
+
"learning_rate": 5.394051681840719e-06,
|
| 778 |
+
"loss": 0.9497,
|
| 779 |
"step": 110
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.5826771653543307,
|
| 783 |
+
"grad_norm": 0.3987231911136733,
|
| 784 |
+
"learning_rate": 5.3772447284987216e-06,
|
| 785 |
+
"loss": 0.961,
|
| 786 |
"step": 111
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.5879265091863517,
|
| 790 |
+
"grad_norm": 0.37848222525971176,
|
| 791 |
+
"learning_rate": 5.36023481410045e-06,
|
| 792 |
+
"loss": 0.9707,
|
| 793 |
"step": 112
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.5931758530183727,
|
| 797 |
+
"grad_norm": 0.3794904855253974,
|
| 798 |
+
"learning_rate": 5.343023390894446e-06,
|
| 799 |
+
"loss": 0.9714,
|
| 800 |
"step": 113
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.5984251968503937,
|
| 804 |
+
"grad_norm": 0.37452267525256994,
|
| 805 |
+
"learning_rate": 5.325611928333389e-06,
|
| 806 |
+
"loss": 0.9406,
|
| 807 |
"step": 114
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.6036745406824147,
|
| 811 |
+
"grad_norm": 0.39474437059829304,
|
| 812 |
+
"learning_rate": 5.308001912948637e-06,
|
| 813 |
+
"loss": 0.9626,
|
| 814 |
"step": 115
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.6089238845144357,
|
| 818 |
+
"grad_norm": 0.4023921986663554,
|
| 819 |
+
"learning_rate": 5.290194848223309e-06,
|
| 820 |
+
"loss": 0.9889,
|
| 821 |
"step": 116
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.6141732283464567,
|
| 825 |
+
"grad_norm": 0.39963771712171875,
|
| 826 |
+
"learning_rate": 5.272192254463929e-06,
|
| 827 |
+
"loss": 0.9639,
|
| 828 |
"step": 117
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.6194225721784777,
|
| 832 |
+
"grad_norm": 0.3893586064595733,
|
| 833 |
+
"learning_rate": 5.2539956686706205e-06,
|
| 834 |
+
"loss": 0.9469,
|
| 835 |
"step": 118
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.6246719160104987,
|
| 839 |
+
"grad_norm": 0.4651495625439333,
|
| 840 |
+
"learning_rate": 5.2356066444058875e-06,
|
| 841 |
+
"loss": 0.9658,
|
| 842 |
"step": 119
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.6299212598425197,
|
| 846 |
+
"grad_norm": 0.39599728107932586,
|
| 847 |
+
"learning_rate": 5.217026751661978e-06,
|
| 848 |
+
"loss": 1.0137,
|
| 849 |
"step": 120
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.6351706036745407,
|
| 853 |
+
"grad_norm": 0.406988761369817,
|
| 854 |
+
"learning_rate": 5.198257576726835e-06,
|
| 855 |
+
"loss": 0.9306,
|
| 856 |
"step": 121
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.6404199475065617,
|
| 860 |
+
"grad_norm": 0.3611939094322339,
|
| 861 |
+
"learning_rate": 5.179300722048673e-06,
|
| 862 |
+
"loss": 0.9462,
|
| 863 |
"step": 122
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.6456692913385826,
|
| 867 |
+
"grad_norm": 0.3809841775392484,
|
| 868 |
+
"learning_rate": 5.1601578060991645e-06,
|
| 869 |
+
"loss": 0.953,
|
| 870 |
"step": 123
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.6509186351706037,
|
| 874 |
+
"grad_norm": 0.46022843064705843,
|
| 875 |
+
"learning_rate": 5.1408304632352575e-06,
|
| 876 |
+
"loss": 0.9422,
|
| 877 |
"step": 124
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.6561679790026247,
|
| 881 |
+
"grad_norm": 0.3979704646560941,
|
| 882 |
+
"learning_rate": 5.1213203435596425e-06,
|
| 883 |
+
"loss": 0.9751,
|
| 884 |
"step": 125
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.6614173228346457,
|
| 888 |
+
"grad_norm": 0.39388496260457084,
|
| 889 |
+
"learning_rate": 5.101629112779873e-06,
|
| 890 |
+
"loss": 0.9722,
|
| 891 |
"step": 126
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.6666666666666666,
|
| 895 |
+
"grad_norm": 0.3899148438115094,
|
| 896 |
+
"learning_rate": 5.08175845206615e-06,
|
| 897 |
+
"loss": 0.9652,
|
| 898 |
"step": 127
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.6719160104986877,
|
| 902 |
+
"grad_norm": 0.37391882787694275,
|
| 903 |
+
"learning_rate": 5.061710057907788e-06,
|
| 904 |
+
"loss": 0.9621,
|
| 905 |
"step": 128
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.6771653543307087,
|
| 909 |
+
"grad_norm": 0.39500875865406576,
|
| 910 |
+
"learning_rate": 5.041485641968385e-06,
|
| 911 |
+
"loss": 0.9899,
|
| 912 |
"step": 129
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.6824146981627297,
|
| 916 |
+
"grad_norm": 0.37540362490802714,
|
| 917 |
+
"learning_rate": 5.021086930939672e-06,
|
| 918 |
+
"loss": 0.9472,
|
| 919 |
"step": 130
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.6876640419947506,
|
| 923 |
+
"grad_norm": 0.3940788728379769,
|
| 924 |
+
"learning_rate": 5.000515666394105e-06,
|
| 925 |
+
"loss": 0.9479,
|
| 926 |
"step": 131
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.6929133858267716,
|
| 930 |
+
"grad_norm": 0.3919125365655477,
|
| 931 |
+
"learning_rate": 4.979773604636169e-06,
|
| 932 |
+
"loss": 0.9624,
|
| 933 |
"step": 132
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.6981627296587927,
|
| 937 |
+
"grad_norm": 0.3804552314744538,
|
| 938 |
+
"learning_rate": 4.958862516552433e-06,
|
| 939 |
+
"loss": 0.9806,
|
| 940 |
"step": 133
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.7034120734908137,
|
| 944 |
+
"grad_norm": 0.3674434286105591,
|
| 945 |
+
"learning_rate": 4.937784187460362e-06,
|
| 946 |
+
"loss": 0.9511,
|
| 947 |
"step": 134
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.7086614173228346,
|
| 951 |
+
"grad_norm": 0.4109777494732396,
|
| 952 |
+
"learning_rate": 4.916540416955884e-06,
|
| 953 |
+
"loss": 0.9943,
|
| 954 |
"step": 135
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.7139107611548556,
|
| 958 |
+
"grad_norm": 0.40231567788837497,
|
| 959 |
+
"learning_rate": 4.895133018759753e-06,
|
| 960 |
+
"loss": 0.9798,
|
| 961 |
"step": 136
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.7191601049868767,
|
| 965 |
+
"grad_norm": 0.3721834479908975,
|
| 966 |
+
"learning_rate": 4.873563820562698e-06,
|
| 967 |
+
"loss": 0.9504,
|
| 968 |
"step": 137
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.7244094488188977,
|
| 972 |
+
"grad_norm": 0.36127526200518306,
|
| 973 |
+
"learning_rate": 4.851834663869379e-06,
|
| 974 |
+
"loss": 0.9517,
|
| 975 |
"step": 138
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.7296587926509186,
|
| 979 |
+
"grad_norm": 0.3513827139135777,
|
| 980 |
+
"learning_rate": 4.82994740384117e-06,
|
| 981 |
+
"loss": 0.9835,
|
| 982 |
"step": 139
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.7349081364829396,
|
| 986 |
+
"grad_norm": 0.36760728272750326,
|
| 987 |
+
"learning_rate": 4.80790390913777e-06,
|
| 988 |
+
"loss": 0.9503,
|
| 989 |
"step": 140
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.7401574803149606,
|
| 993 |
+
"grad_norm": 0.36275280721999276,
|
| 994 |
+
"learning_rate": 4.785706061757656e-06,
|
| 995 |
+
"loss": 0.9743,
|
| 996 |
"step": 141
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.7454068241469817,
|
| 1000 |
+
"grad_norm": 0.3733380512329921,
|
| 1001 |
+
"learning_rate": 4.763355756877419e-06,
|
| 1002 |
+
"loss": 0.9384,
|
| 1003 |
"step": 142
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.7506561679790026,
|
| 1007 |
+
"grad_norm": 0.3801691027568987,
|
| 1008 |
+
"learning_rate": 4.740854902689947e-06,
|
| 1009 |
+
"loss": 0.9296,
|
| 1010 |
"step": 143
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.7559055118110236,
|
| 1014 |
+
"grad_norm": 0.39053906811778566,
|
| 1015 |
+
"learning_rate": 4.718205420241516e-06,
|
| 1016 |
+
"loss": 0.9488,
|
| 1017 |
"step": 144
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.7611548556430446,
|
| 1021 |
+
"grad_norm": 0.3923993707534958,
|
| 1022 |
+
"learning_rate": 4.695409243267776e-06,
|
| 1023 |
+
"loss": 0.9383,
|
| 1024 |
"step": 145
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.7664041994750657,
|
| 1028 |
+
"grad_norm": 0.364792552828712,
|
| 1029 |
+
"learning_rate": 4.672468318028657e-06,
|
| 1030 |
+
"loss": 0.9193,
|
| 1031 |
"step": 146
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.7716535433070866,
|
| 1035 |
+
"grad_norm": 0.35070825551906964,
|
| 1036 |
+
"learning_rate": 4.649384603142202e-06,
|
| 1037 |
+
"loss": 0.9164,
|
| 1038 |
"step": 147
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.7769028871391076,
|
| 1042 |
+
"grad_norm": 0.37099778180795795,
|
| 1043 |
+
"learning_rate": 4.626160069417348e-06,
|
| 1044 |
+
"loss": 0.9425,
|
| 1045 |
"step": 148
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.7821522309711286,
|
| 1049 |
+
"grad_norm": 0.36954118968922517,
|
| 1050 |
+
"learning_rate": 4.602796699685665e-06,
|
| 1051 |
+
"loss": 0.9265,
|
| 1052 |
"step": 149
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.7874015748031497,
|
| 1056 |
+
"grad_norm": 0.4076466706382121,
|
| 1057 |
+
"learning_rate": 4.579296488632067e-06,
|
| 1058 |
+
"loss": 1.0133,
|
| 1059 |
"step": 150
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.7926509186351706,
|
| 1063 |
+
"grad_norm": 0.4015334925568992,
|
| 1064 |
+
"learning_rate": 4.5556614426245165e-06,
|
| 1065 |
+
"loss": 0.9486,
|
| 1066 |
"step": 151
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.7979002624671916,
|
| 1070 |
+
"grad_norm": 0.39628644809730684,
|
| 1071 |
+
"learning_rate": 4.5318935795427206e-06,
|
| 1072 |
+
"loss": 0.9605,
|
| 1073 |
"step": 152
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.8031496062992126,
|
| 1077 |
+
"grad_norm": 0.36792154742540445,
|
| 1078 |
+
"learning_rate": 4.507994928605862e-06,
|
| 1079 |
+
"loss": 0.9287,
|
| 1080 |
"step": 153
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.8083989501312336,
|
| 1084 |
+
"grad_norm": 0.3887839296706913,
|
| 1085 |
+
"learning_rate": 4.483967530199337e-06,
|
| 1086 |
+
"loss": 0.951,
|
| 1087 |
"step": 154
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.8136482939632546,
|
| 1091 |
+
"grad_norm": 0.36716852968968616,
|
| 1092 |
+
"learning_rate": 4.459813435700569e-06,
|
| 1093 |
+
"loss": 0.9702,
|
| 1094 |
"step": 155
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.8188976377952756,
|
| 1098 |
+
"grad_norm": 0.3533521076976156,
|
| 1099 |
+
"learning_rate": 4.4355347073038595e-06,
|
| 1100 |
+
"loss": 0.9612,
|
| 1101 |
"step": 156
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.8241469816272966,
|
| 1105 |
+
"grad_norm": 0.3499649930079787,
|
| 1106 |
+
"learning_rate": 4.411133417844328e-06,
|
| 1107 |
+
"loss": 0.9599,
|
| 1108 |
"step": 157
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.8293963254593176,
|
| 1112 |
+
"grad_norm": 0.38582146832565867,
|
| 1113 |
+
"learning_rate": 4.38661165062094e-06,
|
| 1114 |
+
"loss": 0.9894,
|
| 1115 |
"step": 158
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.8346456692913385,
|
| 1119 |
+
"grad_norm": 0.39040836855795735,
|
| 1120 |
+
"learning_rate": 4.36197149921864e-06,
|
| 1121 |
+
"loss": 0.9747,
|
| 1122 |
"step": 159
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.8398950131233596,
|
| 1126 |
+
"grad_norm": 0.3798580758700489,
|
| 1127 |
+
"learning_rate": 4.3372150673296155e-06,
|
| 1128 |
+
"loss": 0.9654,
|
| 1129 |
"step": 160
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.8451443569553806,
|
| 1133 |
+
"grad_norm": 0.3764456540061034,
|
| 1134 |
+
"learning_rate": 4.3123444685736795e-06,
|
| 1135 |
+
"loss": 0.9823,
|
| 1136 |
"step": 161
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.8503937007874016,
|
| 1140 |
+
"grad_norm": 0.3771195417830333,
|
| 1141 |
+
"learning_rate": 4.287361826317827e-06,
|
| 1142 |
+
"loss": 0.9456,
|
| 1143 |
"step": 162
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.8556430446194225,
|
| 1147 |
+
"grad_norm": 0.37650137746409273,
|
| 1148 |
+
"learning_rate": 4.262269273494946e-06,
|
| 1149 |
+
"loss": 1.0022,
|
| 1150 |
"step": 163
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.8608923884514436,
|
| 1154 |
+
"grad_norm": 0.38148353077474145,
|
| 1155 |
+
"learning_rate": 4.237068952421711e-06,
|
| 1156 |
+
"loss": 0.964,
|
| 1157 |
"step": 164
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.8661417322834646,
|
| 1161 |
+
"grad_norm": 0.3982519128695332,
|
| 1162 |
+
"learning_rate": 4.2117630146156845e-06,
|
| 1163 |
+
"loss": 0.9673,
|
| 1164 |
"step": 165
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.8713910761154856,
|
| 1168 |
+
"grad_norm": 0.36000775624632003,
|
| 1169 |
+
"learning_rate": 4.186353620611627e-06,
|
| 1170 |
+
"loss": 0.9359,
|
| 1171 |
"step": 166
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.8766404199475065,
|
| 1175 |
+
"grad_norm": 0.36850454735662447,
|
| 1176 |
+
"learning_rate": 4.160842939777036e-06,
|
| 1177 |
+
"loss": 0.9422,
|
| 1178 |
"step": 167
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.8818897637795275,
|
| 1182 |
+
"grad_norm": 0.37804115639757085,
|
| 1183 |
+
"learning_rate": 4.135233150126931e-06,
|
| 1184 |
+
"loss": 0.9454,
|
| 1185 |
"step": 168
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.8871391076115486,
|
| 1189 |
+
"grad_norm": 0.3689383402086321,
|
| 1190 |
+
"learning_rate": 4.109526438137908e-06,
|
| 1191 |
+
"loss": 0.9455,
|
| 1192 |
"step": 169
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.8923884514435696,
|
| 1196 |
+
"grad_norm": 0.46527154775209717,
|
| 1197 |
+
"learning_rate": 4.08372499856146e-06,
|
| 1198 |
+
"loss": 0.9386,
|
| 1199 |
"step": 170
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.8976377952755905,
|
| 1203 |
+
"grad_norm": 0.45653306710128705,
|
| 1204 |
+
"learning_rate": 4.0578310342365975e-06,
|
| 1205 |
+
"loss": 0.9616,
|
| 1206 |
"step": 171
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.9028871391076115,
|
| 1210 |
+
"grad_norm": 0.3773630567359451,
|
| 1211 |
+
"learning_rate": 4.031846755901785e-06,
|
| 1212 |
+
"loss": 0.9285,
|
| 1213 |
"step": 172
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.9081364829396326,
|
| 1217 |
+
"grad_norm": 0.3644595191521506,
|
| 1218 |
+
"learning_rate": 4.005774382006182e-06,
|
| 1219 |
+
"loss": 0.9663,
|
| 1220 |
"step": 173
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
"epoch": 0.9133858267716536,
|
| 1224 |
+
"grad_norm": 0.3539767481135477,
|
| 1225 |
+
"learning_rate": 3.97961613852025e-06,
|
| 1226 |
+
"loss": 0.9564,
|
| 1227 |
"step": 174
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.9186351706036745,
|
| 1231 |
+
"grad_norm": 0.3819676152776953,
|
| 1232 |
+
"learning_rate": 3.953374258745705e-06,
|
| 1233 |
+
"loss": 0.9607,
|
| 1234 |
"step": 175
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
"epoch": 0.9238845144356955,
|
| 1238 |
+
"grad_norm": 0.38397675786726637,
|
| 1239 |
+
"learning_rate": 3.927050983124842e-06,
|
| 1240 |
+
"loss": 0.9539,
|
| 1241 |
"step": 176
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.9291338582677166,
|
| 1245 |
+
"grad_norm": 0.3979084367711538,
|
| 1246 |
+
"learning_rate": 3.900648559049258e-06,
|
| 1247 |
+
"loss": 0.9505,
|
| 1248 |
"step": 177
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
"epoch": 0.9343832020997376,
|
| 1252 |
+
"grad_norm": 0.3756154385935223,
|
| 1253 |
+
"learning_rate": 3.874169240667974e-06,
|
| 1254 |
+
"loss": 0.9519,
|
| 1255 |
"step": 178
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.9396325459317585,
|
| 1259 |
+
"grad_norm": 0.40551973597201274,
|
| 1260 |
+
"learning_rate": 3.847615288694985e-06,
|
| 1261 |
+
"loss": 0.9727,
|
| 1262 |
"step": 179
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
"epoch": 0.9448818897637795,
|
| 1266 |
+
"grad_norm": 0.4149625851710124,
|
| 1267 |
+
"learning_rate": 3.820988970216249e-06,
|
| 1268 |
+
"loss": 0.9464,
|
| 1269 |
"step": 180
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.9501312335958005,
|
| 1273 |
+
"grad_norm": 0.35739115830542967,
|
| 1274 |
+
"learning_rate": 3.7942925584961272e-06,
|
| 1275 |
+
"loss": 0.9427,
|
| 1276 |
"step": 181
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
"epoch": 0.9553805774278216,
|
| 1280 |
+
"grad_norm": 0.3759540038847051,
|
| 1281 |
+
"learning_rate": 3.767528332783307e-06,
|
| 1282 |
+
"loss": 0.9679,
|
| 1283 |
"step": 182
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.9606299212598425,
|
| 1287 |
+
"grad_norm": 0.3525867658299593,
|
| 1288 |
+
"learning_rate": 3.740698578116199e-06,
|
| 1289 |
+
"loss": 0.9183,
|
| 1290 |
"step": 183
|
| 1291 |
},
|
| 1292 |
{
|
| 1293 |
"epoch": 0.9658792650918635,
|
| 1294 |
+
"grad_norm": 0.3557123352774738,
|
| 1295 |
+
"learning_rate": 3.7138055851278564e-06,
|
| 1296 |
+
"loss": 0.9383,
|
| 1297 |
"step": 184
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.9711286089238845,
|
| 1301 |
+
"grad_norm": 0.3623514252763418,
|
| 1302 |
+
"learning_rate": 3.6868516498504025e-06,
|
| 1303 |
+
"loss": 0.9246,
|
| 1304 |
"step": 185
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
"epoch": 0.9763779527559056,
|
| 1308 |
+
"grad_norm": 0.38495496418054853,
|
| 1309 |
+
"learning_rate": 3.6598390735190066e-06,
|
| 1310 |
+
"loss": 0.9612,
|
| 1311 |
"step": 186
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.9816272965879265,
|
| 1315 |
+
"grad_norm": 0.3648599004428126,
|
| 1316 |
+
"learning_rate": 3.63277016237541e-06,
|
| 1317 |
+
"loss": 0.9293,
|
| 1318 |
"step": 187
|
| 1319 |
},
|
| 1320 |
{
|
| 1321 |
"epoch": 0.9868766404199475,
|
| 1322 |
+
"grad_norm": 0.38871547084803876,
|
| 1323 |
+
"learning_rate": 3.6056472274710305e-06,
|
| 1324 |
+
"loss": 0.9973,
|
| 1325 |
"step": 188
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.9921259842519685,
|
| 1329 |
+
"grad_norm": 0.38590844403642666,
|
| 1330 |
+
"learning_rate": 3.578472584469651e-06,
|
| 1331 |
+
"loss": 0.9457,
|
| 1332 |
"step": 189
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
"epoch": 0.9973753280839895,
|
| 1336 |
+
"grad_norm": 0.3872507088649178,
|
| 1337 |
+
"learning_rate": 3.5512485534497116e-06,
|
| 1338 |
+
"loss": 0.9462,
|
| 1339 |
"step": 190
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 1.0,
|
| 1343 |
+
"grad_norm": 0.3872507088649178,
|
| 1344 |
+
"learning_rate": 3.523977458706237e-06,
|
| 1345 |
+
"loss": 0.9693,
|
| 1346 |
"step": 191
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
"epoch": 1.005249343832021,
|
| 1350 |
+
"grad_norm": 0.6232728744646114,
|
| 1351 |
+
"learning_rate": 3.49666162855239e-06,
|
| 1352 |
+
"loss": 0.887,
|
| 1353 |
"step": 192
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 1.010498687664042,
|
| 1357 |
+
"grad_norm": 0.4149641950734625,
|
| 1358 |
+
"learning_rate": 3.469303395120693e-06,
|
| 1359 |
+
"loss": 0.8826,
|
| 1360 |
"step": 193
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
"epoch": 1.015748031496063,
|
| 1364 |
+
"grad_norm": 0.37273340109017755,
|
| 1365 |
+
"learning_rate": 3.441905094163913e-06,
|
| 1366 |
+
"loss": 0.8893,
|
| 1367 |
"step": 194
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
"epoch": 1.020997375328084,
|
| 1371 |
+
"grad_norm": 0.4113832689982837,
|
| 1372 |
+
"learning_rate": 3.414469064855647e-06,
|
| 1373 |
+
"loss": 0.9205,
|
| 1374 |
"step": 195
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
"epoch": 1.026246719160105,
|
| 1378 |
+
"grad_norm": 0.49485155842511663,
|
| 1379 |
+
"learning_rate": 3.3869976495906104e-06,
|
| 1380 |
+
"loss": 0.9074,
|
| 1381 |
"step": 196
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
"epoch": 1.031496062992126,
|
| 1385 |
+
"grad_norm": 0.3736781934252868,
|
| 1386 |
+
"learning_rate": 3.3594931937846498e-06,
|
| 1387 |
+
"loss": 0.8966,
|
| 1388 |
"step": 197
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
"epoch": 1.036745406824147,
|
| 1392 |
+
"grad_norm": 0.3758650059773124,
|
| 1393 |
+
"learning_rate": 3.3319580456745023e-06,
|
| 1394 |
+
"loss": 0.8759,
|
| 1395 |
"step": 198
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
"epoch": 1.041994750656168,
|
| 1399 |
+
"grad_norm": 0.4056031624712629,
|
| 1400 |
+
"learning_rate": 3.3043945561173092e-06,
|
| 1401 |
+
"loss": 0.8788,
|
| 1402 |
"step": 199
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
"epoch": 1.047244094488189,
|
| 1406 |
+
"grad_norm": 0.36344982085137467,
|
| 1407 |
+
"learning_rate": 3.2768050783899063e-06,
|
| 1408 |
+
"loss": 0.873,
|
| 1409 |
"step": 200
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
"epoch": 1.05249343832021,
|
| 1413 |
+
"grad_norm": 0.3760103676246,
|
| 1414 |
+
"learning_rate": 3.249191967987912e-06,
|
| 1415 |
+
"loss": 0.899,
|
| 1416 |
"step": 201
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
"epoch": 1.057742782152231,
|
| 1420 |
+
"grad_norm": 0.39433477834527153,
|
| 1421 |
+
"learning_rate": 3.221557582424622e-06,
|
| 1422 |
+
"loss": 0.9019,
|
| 1423 |
"step": 202
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
"epoch": 1.0629921259842519,
|
| 1427 |
+
"grad_norm": 0.3595753440791428,
|
| 1428 |
+
"learning_rate": 3.1939042810297328e-06,
|
| 1429 |
+
"loss": 0.8781,
|
| 1430 |
"step": 203
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 1.068241469816273,
|
| 1434 |
+
"grad_norm": 0.3743448170598354,
|
| 1435 |
+
"learning_rate": 3.16623442474791e-06,
|
| 1436 |
+
"loss": 0.8689,
|
| 1437 |
"step": 204
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
"epoch": 1.073490813648294,
|
| 1441 |
+
"grad_norm": 0.3618551186966609,
|
| 1442 |
+
"learning_rate": 3.138550375937219e-06,
|
| 1443 |
+
"loss": 0.9094,
|
| 1444 |
"step": 205
|
| 1445 |
},
|
| 1446 |
{
|
| 1447 |
"epoch": 1.078740157480315,
|
| 1448 |
+
"grad_norm": 0.36577516842050983,
|
| 1449 |
+
"learning_rate": 3.1108544981674356e-06,
|
| 1450 |
+
"loss": 0.8668,
|
| 1451 |
"step": 206
|
| 1452 |
},
|
| 1453 |
{
|
| 1454 |
"epoch": 1.083989501312336,
|
| 1455 |
+
"grad_norm": 0.3985134455319658,
|
| 1456 |
+
"learning_rate": 3.0831491560182495e-06,
|
| 1457 |
+
"loss": 0.9016,
|
| 1458 |
"step": 207
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 1.0892388451443569,
|
| 1462 |
+
"grad_norm": 0.37808489525197075,
|
| 1463 |
+
"learning_rate": 3.0554367148773897e-06,
|
| 1464 |
+
"loss": 0.895,
|
| 1465 |
"step": 208
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 1.094488188976378,
|
| 1469 |
+
"grad_norm": 0.4112784941005797,
|
| 1470 |
+
"learning_rate": 3.027719540738673e-06,
|
| 1471 |
+
"loss": 0.859,
|
| 1472 |
"step": 209
|
| 1473 |
},
|
| 1474 |
{
|
| 1475 |
"epoch": 1.099737532808399,
|
| 1476 |
+
"grad_norm": 0.3830296759827936,
|
| 1477 |
+
"learning_rate": 3e-06,
|
| 1478 |
+
"loss": 0.8569,
|
| 1479 |
"step": 210
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 1.10498687664042,
|
| 1483 |
+
"grad_norm": 0.3930755503999148,
|
| 1484 |
+
"learning_rate": 2.972280459261328e-06,
|
| 1485 |
+
"loss": 0.8774,
|
| 1486 |
"step": 211
|
| 1487 |
},
|
| 1488 |
{
|
| 1489 |
"epoch": 1.110236220472441,
|
| 1490 |
+
"grad_norm": 0.36738851637178116,
|
| 1491 |
+
"learning_rate": 2.944563285122611e-06,
|
| 1492 |
+
"loss": 0.9086,
|
| 1493 |
"step": 212
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
"epoch": 1.1154855643044619,
|
| 1497 |
+
"grad_norm": 0.3897160841039193,
|
| 1498 |
+
"learning_rate": 2.9168508439817515e-06,
|
| 1499 |
+
"loss": 0.889,
|
| 1500 |
"step": 213
|
| 1501 |
},
|
| 1502 |
{
|
| 1503 |
"epoch": 1.120734908136483,
|
| 1504 |
+
"grad_norm": 0.39858146379374537,
|
| 1505 |
+
"learning_rate": 2.889145501832566e-06,
|
| 1506 |
+
"loss": 0.8964,
|
| 1507 |
"step": 214
|
| 1508 |
},
|
| 1509 |
{
|
| 1510 |
"epoch": 1.125984251968504,
|
| 1511 |
+
"grad_norm": 0.3739395525411432,
|
| 1512 |
+
"learning_rate": 2.861449624062782e-06,
|
| 1513 |
+
"loss": 0.8884,
|
| 1514 |
"step": 215
|
| 1515 |
},
|
| 1516 |
{
|
| 1517 |
"epoch": 1.1312335958005248,
|
| 1518 |
+
"grad_norm": 0.3755768464864809,
|
| 1519 |
+
"learning_rate": 2.83376557525209e-06,
|
| 1520 |
+
"loss": 0.851,
|
| 1521 |
"step": 216
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 1.136482939632546,
|
| 1525 |
+
"grad_norm": 0.38260315757882735,
|
| 1526 |
+
"learning_rate": 2.8060957189702674e-06,
|
| 1527 |
+
"loss": 0.9152,
|
| 1528 |
"step": 217
|
| 1529 |
},
|
| 1530 |
{
|
| 1531 |
"epoch": 1.141732283464567,
|
| 1532 |
+
"grad_norm": 0.4205379839527009,
|
| 1533 |
+
"learning_rate": 2.7784424175753784e-06,
|
| 1534 |
+
"loss": 0.8683,
|
| 1535 |
"step": 218
|
| 1536 |
},
|
| 1537 |
{
|
| 1538 |
"epoch": 1.1469816272965878,
|
| 1539 |
+
"grad_norm": 0.38325260941818995,
|
| 1540 |
+
"learning_rate": 2.7508080320120888e-06,
|
| 1541 |
+
"loss": 0.8943,
|
| 1542 |
"step": 219
|
| 1543 |
},
|
| 1544 |
{
|
| 1545 |
"epoch": 1.152230971128609,
|
| 1546 |
+
"grad_norm": 0.3763198826603672,
|
| 1547 |
+
"learning_rate": 2.7231949216100943e-06,
|
| 1548 |
+
"loss": 0.8676,
|
| 1549 |
"step": 220
|
| 1550 |
},
|
| 1551 |
{
|
| 1552 |
"epoch": 1.1574803149606299,
|
| 1553 |
+
"grad_norm": 0.3767162287387105,
|
| 1554 |
+
"learning_rate": 2.6956054438826918e-06,
|
| 1555 |
+
"loss": 0.8482,
|
| 1556 |
"step": 221
|
| 1557 |
},
|
| 1558 |
{
|
| 1559 |
"epoch": 1.162729658792651,
|
| 1560 |
+
"grad_norm": 0.3486273740901837,
|
| 1561 |
+
"learning_rate": 2.668041954325498e-06,
|
| 1562 |
+
"loss": 0.8879,
|
| 1563 |
"step": 222
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 1.167979002624672,
|
| 1567 |
+
"grad_norm": 0.39084218665366566,
|
| 1568 |
+
"learning_rate": 2.640506806215351e-06,
|
| 1569 |
+
"loss": 0.8679,
|
| 1570 |
"step": 223
|
| 1571 |
},
|
| 1572 |
{
|
| 1573 |
"epoch": 1.1732283464566928,
|
| 1574 |
+
"grad_norm": 0.3538552501730603,
|
| 1575 |
+
"learning_rate": 2.613002350409391e-06,
|
| 1576 |
+
"loss": 0.8871,
|
| 1577 |
"step": 224
|
| 1578 |
},
|
| 1579 |
{
|
| 1580 |
"epoch": 1.178477690288714,
|
| 1581 |
+
"grad_norm": 0.36544200913577,
|
| 1582 |
+
"learning_rate": 2.585530935144354e-06,
|
| 1583 |
+
"loss": 0.8616,
|
| 1584 |
"step": 225
|
| 1585 |
},
|
| 1586 |
{
|
| 1587 |
"epoch": 1.1837270341207349,
|
| 1588 |
+
"grad_norm": 0.3985990462573467,
|
| 1589 |
+
"learning_rate": 2.558094905836087e-06,
|
| 1590 |
+
"loss": 0.8917,
|
| 1591 |
"step": 226
|
| 1592 |
},
|
| 1593 |
{
|
| 1594 |
"epoch": 1.188976377952756,
|
| 1595 |
+
"grad_norm": 0.42608518999556655,
|
| 1596 |
+
"learning_rate": 2.5306966048793067e-06,
|
| 1597 |
+
"loss": 0.8817,
|
| 1598 |
"step": 227
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
"epoch": 1.194225721784777,
|
| 1602 |
+
"grad_norm": 0.37952769789031354,
|
| 1603 |
+
"learning_rate": 2.5033383714476097e-06,
|
| 1604 |
+
"loss": 0.8985,
|
| 1605 |
"step": 228
|
| 1606 |
},
|
| 1607 |
{
|
| 1608 |
"epoch": 1.1994750656167978,
|
| 1609 |
+
"grad_norm": 0.40804864076806885,
|
| 1610 |
+
"learning_rate": 2.4760225412937633e-06,
|
| 1611 |
+
"loss": 0.9073,
|
| 1612 |
"step": 229
|
| 1613 |
},
|
| 1614 |
{
|
| 1615 |
"epoch": 1.204724409448819,
|
| 1616 |
+
"grad_norm": 0.4167713152946991,
|
| 1617 |
+
"learning_rate": 2.4487514465502885e-06,
|
| 1618 |
+
"loss": 0.8566,
|
| 1619 |
"step": 230
|
| 1620 |
},
|
| 1621 |
{
|
| 1622 |
"epoch": 1.20997375328084,
|
| 1623 |
+
"grad_norm": 0.4022153540631621,
|
| 1624 |
+
"learning_rate": 2.42152741553035e-06,
|
| 1625 |
+
"loss": 0.8713,
|
| 1626 |
"step": 231
|
| 1627 |
},
|
| 1628 |
{
|
| 1629 |
"epoch": 1.2152230971128608,
|
| 1630 |
+
"grad_norm": 0.4222065137992956,
|
| 1631 |
+
"learning_rate": 2.39435277252897e-06,
|
| 1632 |
+
"loss": 0.9035,
|
| 1633 |
"step": 232
|
| 1634 |
},
|
| 1635 |
{
|
| 1636 |
"epoch": 1.220472440944882,
|
| 1637 |
+
"grad_norm": 0.3666365807384159,
|
| 1638 |
+
"learning_rate": 2.3672298376245908e-06,
|
| 1639 |
+
"loss": 0.8637,
|
| 1640 |
"step": 233
|
| 1641 |
},
|
| 1642 |
{
|
| 1643 |
"epoch": 1.2257217847769029,
|
| 1644 |
+
"grad_norm": 0.3976853335036615,
|
| 1645 |
+
"learning_rate": 2.3401609264809953e-06,
|
| 1646 |
+
"loss": 0.9398,
|
| 1647 |
"step": 234
|
| 1648 |
},
|
| 1649 |
{
|
| 1650 |
"epoch": 1.2309711286089238,
|
| 1651 |
+
"grad_norm": 0.37956934109451046,
|
| 1652 |
+
"learning_rate": 2.3131483501495985e-06,
|
| 1653 |
+
"loss": 0.8353,
|
| 1654 |
"step": 235
|
| 1655 |
},
|
| 1656 |
{
|
| 1657 |
"epoch": 1.236220472440945,
|
| 1658 |
+
"grad_norm": 0.33722056538083744,
|
| 1659 |
+
"learning_rate": 2.2861944148721446e-06,
|
| 1660 |
+
"loss": 0.8786,
|
| 1661 |
"step": 236
|
| 1662 |
},
|
| 1663 |
{
|
| 1664 |
"epoch": 1.2414698162729658,
|
| 1665 |
+
"grad_norm": 0.49777382093647954,
|
| 1666 |
+
"learning_rate": 2.2593014218838e-06,
|
| 1667 |
+
"loss": 0.8834,
|
| 1668 |
"step": 237
|
| 1669 |
},
|
| 1670 |
{
|
| 1671 |
"epoch": 1.246719160104987,
|
| 1672 |
+
"grad_norm": 0.35315516410389436,
|
| 1673 |
+
"learning_rate": 2.232471667216693e-06,
|
| 1674 |
+
"loss": 0.8442,
|
| 1675 |
"step": 238
|
| 1676 |
},
|
| 1677 |
{
|
| 1678 |
"epoch": 1.2519685039370079,
|
| 1679 |
+
"grad_norm": 0.3816124424363711,
|
| 1680 |
+
"learning_rate": 2.2057074415038725e-06,
|
| 1681 |
+
"loss": 0.8573,
|
| 1682 |
"step": 239
|
| 1683 |
},
|
| 1684 |
{
|
| 1685 |
"epoch": 1.257217847769029,
|
| 1686 |
+
"grad_norm": 0.36319142999803095,
|
| 1687 |
+
"learning_rate": 2.1790110297837514e-06,
|
| 1688 |
+
"loss": 0.8481,
|
| 1689 |
"step": 240
|
| 1690 |
},
|
| 1691 |
{
|
| 1692 |
"epoch": 1.26246719160105,
|
| 1693 |
+
"grad_norm": 0.34672889281207053,
|
| 1694 |
+
"learning_rate": 2.152384711305015e-06,
|
| 1695 |
+
"loss": 0.8623,
|
| 1696 |
"step": 241
|
| 1697 |
},
|
| 1698 |
{
|
| 1699 |
"epoch": 1.2677165354330708,
|
| 1700 |
+
"grad_norm": 0.37448151544392105,
|
| 1701 |
+
"learning_rate": 2.1258307593320262e-06,
|
| 1702 |
+
"loss": 0.8751,
|
| 1703 |
"step": 242
|
| 1704 |
},
|
| 1705 |
{
|
| 1706 |
"epoch": 1.272965879265092,
|
| 1707 |
+
"grad_norm": 0.37082567424502005,
|
| 1708 |
+
"learning_rate": 2.099351440950742e-06,
|
| 1709 |
+
"loss": 0.8914,
|
| 1710 |
"step": 243
|
| 1711 |
},
|
| 1712 |
{
|
| 1713 |
"epoch": 1.2782152230971129,
|
| 1714 |
+
"grad_norm": 0.39074992783073415,
|
| 1715 |
+
"learning_rate": 2.072949016875158e-06,
|
| 1716 |
+
"loss": 0.9222,
|
| 1717 |
"step": 244
|
| 1718 |
},
|
| 1719 |
{
|
| 1720 |
"epoch": 1.2834645669291338,
|
| 1721 |
+
"grad_norm": 0.4150437401629804,
|
| 1722 |
+
"learning_rate": 2.046625741254295e-06,
|
| 1723 |
+
"loss": 0.9475,
|
| 1724 |
"step": 245
|
| 1725 |
},
|
| 1726 |
{
|
| 1727 |
"epoch": 1.288713910761155,
|
| 1728 |
+
"grad_norm": 0.4504166670407193,
|
| 1729 |
+
"learning_rate": 2.0203838614797505e-06,
|
| 1730 |
+
"loss": 0.9026,
|
| 1731 |
"step": 246
|
| 1732 |
},
|
| 1733 |
{
|
| 1734 |
"epoch": 1.2939632545931758,
|
| 1735 |
+
"grad_norm": 0.38345958484903814,
|
| 1736 |
+
"learning_rate": 1.994225617993819e-06,
|
| 1737 |
+
"loss": 0.9074,
|
| 1738 |
"step": 247
|
| 1739 |
},
|
| 1740 |
{
|
| 1741 |
"epoch": 1.2992125984251968,
|
| 1742 |
+
"grad_norm": 0.37086048031752866,
|
| 1743 |
+
"learning_rate": 1.9681532440982154e-06,
|
| 1744 |
+
"loss": 0.8755,
|
| 1745 |
"step": 248
|
| 1746 |
},
|
| 1747 |
{
|
| 1748 |
"epoch": 1.304461942257218,
|
| 1749 |
+
"grad_norm": 0.3775524407980251,
|
| 1750 |
+
"learning_rate": 1.942168965763402e-06,
|
| 1751 |
+
"loss": 0.8986,
|
| 1752 |
"step": 249
|
| 1753 |
},
|
| 1754 |
{
|
| 1755 |
"epoch": 1.3097112860892388,
|
| 1756 |
+
"grad_norm": 0.364796377340789,
|
| 1757 |
+
"learning_rate": 1.916275001438541e-06,
|
| 1758 |
+
"loss": 0.867,
|
| 1759 |
"step": 250
|
| 1760 |
},
|
| 1761 |
{
|
| 1762 |
"epoch": 1.3149606299212597,
|
| 1763 |
+
"grad_norm": 0.3705604843330414,
|
| 1764 |
+
"learning_rate": 1.8904735618620928e-06,
|
| 1765 |
+
"loss": 0.8875,
|
| 1766 |
"step": 251
|
| 1767 |
},
|
| 1768 |
{
|
| 1769 |
"epoch": 1.3202099737532809,
|
| 1770 |
+
"grad_norm": 0.3847344001283667,
|
| 1771 |
+
"learning_rate": 1.8647668498730693e-06,
|
| 1772 |
+
"loss": 0.8678,
|
| 1773 |
"step": 252
|
| 1774 |
},
|
| 1775 |
{
|
| 1776 |
"epoch": 1.3254593175853018,
|
| 1777 |
+
"grad_norm": 0.3507183610862785,
|
| 1778 |
+
"learning_rate": 1.8391570602229647e-06,
|
| 1779 |
+
"loss": 0.8895,
|
| 1780 |
"step": 253
|
| 1781 |
},
|
| 1782 |
{
|
| 1783 |
"epoch": 1.330708661417323,
|
| 1784 |
+
"grad_norm": 0.34464955572346173,
|
| 1785 |
+
"learning_rate": 1.8136463793883725e-06,
|
| 1786 |
+
"loss": 0.9112,
|
| 1787 |
"step": 254
|
| 1788 |
},
|
| 1789 |
{
|
| 1790 |
"epoch": 1.3359580052493438,
|
| 1791 |
+
"grad_norm": 0.3804540728076062,
|
| 1792 |
+
"learning_rate": 1.7882369853843155e-06,
|
| 1793 |
+
"loss": 0.8818,
|
| 1794 |
"step": 255
|
| 1795 |
},
|
| 1796 |
{
|
| 1797 |
"epoch": 1.341207349081365,
|
| 1798 |
+
"grad_norm": 0.38671544491057547,
|
| 1799 |
+
"learning_rate": 1.76293104757829e-06,
|
| 1800 |
+
"loss": 0.8712,
|
| 1801 |
"step": 256
|
| 1802 |
},
|
| 1803 |
{
|
| 1804 |
"epoch": 1.3464566929133859,
|
| 1805 |
+
"grad_norm": 0.35028636565033566,
|
| 1806 |
+
"learning_rate": 1.7377307265050559e-06,
|
| 1807 |
+
"loss": 0.8795,
|
| 1808 |
"step": 257
|
| 1809 |
},
|
| 1810 |
{
|
| 1811 |
"epoch": 1.3517060367454068,
|
| 1812 |
+
"grad_norm": 0.3596694021401425,
|
| 1813 |
+
"learning_rate": 1.7126381736821732e-06,
|
| 1814 |
+
"loss": 0.8791,
|
| 1815 |
"step": 258
|
| 1816 |
},
|
| 1817 |
{
|
| 1818 |
"epoch": 1.356955380577428,
|
| 1819 |
+
"grad_norm": 0.3833574983214166,
|
| 1820 |
+
"learning_rate": 1.6876555314263213e-06,
|
| 1821 |
+
"loss": 0.9108,
|
| 1822 |
"step": 259
|
| 1823 |
},
|
| 1824 |
{
|
| 1825 |
"epoch": 1.3622047244094488,
|
| 1826 |
+
"grad_norm": 0.3701840047085969,
|
| 1827 |
+
"learning_rate": 1.6627849326703855e-06,
|
| 1828 |
+
"loss": 0.8695,
|
| 1829 |
"step": 260
|
| 1830 |
},
|
| 1831 |
{
|
| 1832 |
"epoch": 1.3674540682414698,
|
| 1833 |
+
"grad_norm": 0.36098816535443995,
|
| 1834 |
+
"learning_rate": 1.6380285007813598e-06,
|
| 1835 |
+
"loss": 0.876,
|
| 1836 |
"step": 261
|
| 1837 |
},
|
| 1838 |
{
|
| 1839 |
"epoch": 1.372703412073491,
|
| 1840 |
+
"grad_norm": 0.3900890284585014,
|
| 1841 |
+
"learning_rate": 1.6133883493790609e-06,
|
| 1842 |
+
"loss": 0.8498,
|
| 1843 |
"step": 262
|
| 1844 |
},
|
| 1845 |
{
|
| 1846 |
"epoch": 1.3779527559055118,
|
| 1847 |
+
"grad_norm": 0.34906551126755136,
|
| 1848 |
+
"learning_rate": 1.5888665821556724e-06,
|
| 1849 |
+
"loss": 0.8513,
|
| 1850 |
"step": 263
|
| 1851 |
},
|
| 1852 |
{
|
| 1853 |
"epoch": 1.3832020997375327,
|
| 1854 |
+
"grad_norm": 0.3753732283477496,
|
| 1855 |
+
"learning_rate": 1.5644652926961407e-06,
|
| 1856 |
+
"loss": 0.8714,
|
| 1857 |
"step": 264
|
| 1858 |
},
|
| 1859 |
{
|
| 1860 |
"epoch": 1.3884514435695539,
|
| 1861 |
+
"grad_norm": 0.34748864593560347,
|
| 1862 |
+
"learning_rate": 1.5401865642994315e-06,
|
| 1863 |
+
"loss": 0.9124,
|
| 1864 |
"step": 265
|
| 1865 |
},
|
| 1866 |
{
|
| 1867 |
"epoch": 1.3937007874015748,
|
| 1868 |
+
"grad_norm": 0.36698053817770165,
|
| 1869 |
+
"learning_rate": 1.5160324698006642e-06,
|
| 1870 |
+
"loss": 0.8814,
|
| 1871 |
"step": 266
|
| 1872 |
},
|
| 1873 |
{
|
| 1874 |
"epoch": 1.3989501312335957,
|
| 1875 |
+
"grad_norm": 0.4000964153653425,
|
| 1876 |
+
"learning_rate": 1.4920050713941398e-06,
|
| 1877 |
+
"loss": 0.9082,
|
| 1878 |
"step": 267
|
| 1879 |
},
|
| 1880 |
{
|
| 1881 |
"epoch": 1.4041994750656168,
|
| 1882 |
+
"grad_norm": 0.3985391177875817,
|
| 1883 |
+
"learning_rate": 1.4681064204572798e-06,
|
| 1884 |
+
"loss": 0.8749,
|
| 1885 |
"step": 268
|
| 1886 |
},
|
| 1887 |
{
|
| 1888 |
"epoch": 1.4094488188976377,
|
| 1889 |
+
"grad_norm": 0.3578122677174226,
|
| 1890 |
+
"learning_rate": 1.4443385573754837e-06,
|
| 1891 |
+
"loss": 0.8608,
|
| 1892 |
"step": 269
|
| 1893 |
},
|
| 1894 |
{
|
| 1895 |
"epoch": 1.4146981627296589,
|
| 1896 |
+
"grad_norm": 0.3576093239254431,
|
| 1897 |
+
"learning_rate": 1.4207035113679322e-06,
|
| 1898 |
+
"loss": 0.8798,
|
| 1899 |
"step": 270
|
| 1900 |
},
|
| 1901 |
{
|
| 1902 |
"epoch": 1.4199475065616798,
|
| 1903 |
+
"grad_norm": 0.35299639204379674,
|
| 1904 |
+
"learning_rate": 1.3972033003143348e-06,
|
| 1905 |
+
"loss": 0.8972,
|
| 1906 |
"step": 271
|
| 1907 |
},
|
| 1908 |
{
|
| 1909 |
"epoch": 1.425196850393701,
|
| 1910 |
+
"grad_norm": 0.3937775289907907,
|
| 1911 |
+
"learning_rate": 1.3738399305826516e-06,
|
| 1912 |
+
"loss": 0.8736,
|
| 1913 |
"step": 272
|
| 1914 |
},
|
| 1915 |
{
|
| 1916 |
"epoch": 1.4304461942257218,
|
| 1917 |
+
"grad_norm": 0.3691998032129419,
|
| 1918 |
+
"learning_rate": 1.3506153968577983e-06,
|
| 1919 |
+
"loss": 0.8667,
|
| 1920 |
"step": 273
|
| 1921 |
},
|
| 1922 |
{
|
| 1923 |
"epoch": 1.4356955380577427,
|
| 1924 |
+
"grad_norm": 0.35764876894907843,
|
| 1925 |
+
"learning_rate": 1.3275316819713435e-06,
|
| 1926 |
+
"loss": 0.882,
|
| 1927 |
"step": 274
|
| 1928 |
},
|
| 1929 |
{
|
| 1930 |
"epoch": 1.4409448818897639,
|
| 1931 |
+
"grad_norm": 0.3859579688778526,
|
| 1932 |
+
"learning_rate": 1.3045907567322243e-06,
|
| 1933 |
+
"loss": 0.844,
|
| 1934 |
"step": 275
|
| 1935 |
},
|
| 1936 |
{
|
| 1937 |
"epoch": 1.4461942257217848,
|
| 1938 |
+
"grad_norm": 0.3736621084680505,
|
| 1939 |
+
"learning_rate": 1.2817945797584844e-06,
|
| 1940 |
+
"loss": 0.8525,
|
| 1941 |
"step": 276
|
| 1942 |
},
|
| 1943 |
{
|
| 1944 |
"epoch": 1.4514435695538057,
|
| 1945 |
+
"grad_norm": 0.36602372507940695,
|
| 1946 |
+
"learning_rate": 1.2591450973100532e-06,
|
| 1947 |
+
"loss": 0.8577,
|
| 1948 |
"step": 277
|
| 1949 |
},
|
| 1950 |
{
|
| 1951 |
"epoch": 1.4566929133858268,
|
| 1952 |
+
"grad_norm": 0.37926054124030645,
|
| 1953 |
+
"learning_rate": 1.236644243122581e-06,
|
| 1954 |
+
"loss": 0.8837,
|
| 1955 |
"step": 278
|
| 1956 |
},
|
| 1957 |
{
|
| 1958 |
"epoch": 1.4619422572178478,
|
| 1959 |
+
"grad_norm": 0.3680022216795608,
|
| 1960 |
+
"learning_rate": 1.214293938242344e-06,
|
| 1961 |
+
"loss": 0.8984,
|
| 1962 |
"step": 279
|
| 1963 |
},
|
| 1964 |
{
|
| 1965 |
"epoch": 1.4671916010498687,
|
| 1966 |
+
"grad_norm": 0.37824901927870175,
|
| 1967 |
+
"learning_rate": 1.1920960908622313e-06,
|
| 1968 |
+
"loss": 0.8745,
|
| 1969 |
"step": 280
|
| 1970 |
},
|
| 1971 |
{
|
| 1972 |
"epoch": 1.4724409448818898,
|
| 1973 |
+
"grad_norm": 0.3489273490529577,
|
| 1974 |
+
"learning_rate": 1.17005259615883e-06,
|
| 1975 |
+
"loss": 0.8628,
|
| 1976 |
"step": 281
|
| 1977 |
},
|
| 1978 |
{
|
| 1979 |
"epoch": 1.4776902887139107,
|
| 1980 |
+
"grad_norm": 0.3735770062938505,
|
| 1981 |
+
"learning_rate": 1.1481653361306215e-06,
|
| 1982 |
+
"loss": 0.8619,
|
| 1983 |
"step": 282
|
| 1984 |
},
|
| 1985 |
{
|
| 1986 |
"epoch": 1.4829396325459316,
|
| 1987 |
+
"grad_norm": 0.3458041443504503,
|
| 1988 |
+
"learning_rate": 1.1264361794373032e-06,
|
| 1989 |
+
"loss": 0.8761,
|
| 1990 |
"step": 283
|
| 1991 |
},
|
| 1992 |
{
|
| 1993 |
"epoch": 1.4881889763779528,
|
| 1994 |
+
"grad_norm": 0.35998420937846626,
|
| 1995 |
+
"learning_rate": 1.104866981240248e-06,
|
| 1996 |
+
"loss": 0.8844,
|
| 1997 |
"step": 284
|
| 1998 |
},
|
| 1999 |
{
|
| 2000 |
"epoch": 1.4934383202099737,
|
| 2001 |
+
"grad_norm": 0.4029178073367971,
|
| 2002 |
+
"learning_rate": 1.0834595830441168e-06,
|
| 2003 |
+
"loss": 0.8511,
|
| 2004 |
"step": 285
|
| 2005 |
}
|
| 2006 |
],
|
checkpoint-285/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8376
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b860c7e838727b1a9d8001f5c7a769bd0d63566ea45620719245b6beb59f1cd9
|
| 3 |
size 8376
|