Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +594 -0
- training_args.bin +3 -0
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3158328
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e92e7c0e65d124154f882db6e6b89ae7dedbcf6ca149012ec2de51653e86644f
|
| 3 |
size 3158328
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0701747dc14bac44bc2768ca59a83df1f871b155a2e2c5492794c93f184fd54
|
| 3 |
+
size 6372346
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae6ca1c4675db6b88190f773cd2468c2e86958ece56a79a0a2ac9f552b528d5d
|
| 3 |
+
size 14244
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59a4acfb0da74c479080613978839dd3cbb4608fd2c07e764b4c844401d8dd5f
|
| 3 |
+
size 988
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71c858e7d80e455b20613a1a6bfebd32fda797d631d427b2c8f97c0128f7aba4
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,594 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05,
|
| 14 |
+
"grad_norm": 0.07341445982456207,
|
| 15 |
+
"learning_rate": 4.93875e-05,
|
| 16 |
+
"loss": 2.2129,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.1,
|
| 21 |
+
"grad_norm": 0.05993746593594551,
|
| 22 |
+
"learning_rate": 4.87625e-05,
|
| 23 |
+
"loss": 1.9761,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.15,
|
| 28 |
+
"grad_norm": 0.09222695976495743,
|
| 29 |
+
"learning_rate": 4.8137500000000005e-05,
|
| 30 |
+
"loss": 2.1171,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.2,
|
| 35 |
+
"grad_norm": 0.10137380659580231,
|
| 36 |
+
"learning_rate": 4.75125e-05,
|
| 37 |
+
"loss": 2.1274,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.25,
|
| 42 |
+
"grad_norm": 0.1008363887667656,
|
| 43 |
+
"learning_rate": 4.68875e-05,
|
| 44 |
+
"loss": 1.9111,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3,
|
| 49 |
+
"grad_norm": 0.09414055943489075,
|
| 50 |
+
"learning_rate": 4.6262500000000006e-05,
|
| 51 |
+
"loss": 1.9712,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.35,
|
| 56 |
+
"grad_norm": 0.08397311717271805,
|
| 57 |
+
"learning_rate": 4.56375e-05,
|
| 58 |
+
"loss": 1.9442,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.4,
|
| 63 |
+
"grad_norm": 0.11413953453302383,
|
| 64 |
+
"learning_rate": 4.50125e-05,
|
| 65 |
+
"loss": 1.8472,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.45,
|
| 70 |
+
"grad_norm": 0.1147843673825264,
|
| 71 |
+
"learning_rate": 4.43875e-05,
|
| 72 |
+
"loss": 1.88,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5,
|
| 77 |
+
"grad_norm": 0.15569846332073212,
|
| 78 |
+
"learning_rate": 4.37625e-05,
|
| 79 |
+
"loss": 1.8664,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.55,
|
| 84 |
+
"grad_norm": 0.1375017762184143,
|
| 85 |
+
"learning_rate": 4.3137500000000005e-05,
|
| 86 |
+
"loss": 1.8417,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6,
|
| 91 |
+
"grad_norm": 0.10806793719530106,
|
| 92 |
+
"learning_rate": 4.2512499999999997e-05,
|
| 93 |
+
"loss": 1.8872,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.65,
|
| 98 |
+
"grad_norm": 0.12453680485486984,
|
| 99 |
+
"learning_rate": 4.18875e-05,
|
| 100 |
+
"loss": 1.8628,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.7,
|
| 105 |
+
"grad_norm": 0.1216558963060379,
|
| 106 |
+
"learning_rate": 4.126250000000001e-05,
|
| 107 |
+
"loss": 1.9252,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.75,
|
| 112 |
+
"grad_norm": 0.10894916951656342,
|
| 113 |
+
"learning_rate": 4.06375e-05,
|
| 114 |
+
"loss": 1.8206,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.8,
|
| 119 |
+
"grad_norm": 0.11225900053977966,
|
| 120 |
+
"learning_rate": 4.0012500000000004e-05,
|
| 121 |
+
"loss": 1.8024,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.85,
|
| 126 |
+
"grad_norm": 0.0950043648481369,
|
| 127 |
+
"learning_rate": 3.93875e-05,
|
| 128 |
+
"loss": 1.8965,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.9,
|
| 133 |
+
"grad_norm": 0.15110863745212555,
|
| 134 |
+
"learning_rate": 3.87625e-05,
|
| 135 |
+
"loss": 1.892,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.95,
|
| 140 |
+
"grad_norm": 0.11794973164796829,
|
| 141 |
+
"learning_rate": 3.8137500000000005e-05,
|
| 142 |
+
"loss": 1.846,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.0,
|
| 147 |
+
"grad_norm": 0.09780783951282501,
|
| 148 |
+
"learning_rate": 3.7512500000000004e-05,
|
| 149 |
+
"loss": 1.8793,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.05,
|
| 154 |
+
"grad_norm": 0.10895536839962006,
|
| 155 |
+
"learning_rate": 3.68875e-05,
|
| 156 |
+
"loss": 1.8671,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.1,
|
| 161 |
+
"grad_norm": 0.11591296643018723,
|
| 162 |
+
"learning_rate": 3.62625e-05,
|
| 163 |
+
"loss": 1.8809,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.15,
|
| 168 |
+
"grad_norm": 0.12027107179164886,
|
| 169 |
+
"learning_rate": 3.56375e-05,
|
| 170 |
+
"loss": 1.8601,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.2,
|
| 175 |
+
"grad_norm": 0.17965424060821533,
|
| 176 |
+
"learning_rate": 3.5012500000000004e-05,
|
| 177 |
+
"loss": 1.79,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.25,
|
| 182 |
+
"grad_norm": 0.10349484533071518,
|
| 183 |
+
"learning_rate": 3.43875e-05,
|
| 184 |
+
"loss": 1.911,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.3,
|
| 189 |
+
"grad_norm": 0.11761331558227539,
|
| 190 |
+
"learning_rate": 3.37625e-05,
|
| 191 |
+
"loss": 1.8324,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.35,
|
| 196 |
+
"grad_norm": 0.10469675809144974,
|
| 197 |
+
"learning_rate": 3.31375e-05,
|
| 198 |
+
"loss": 1.8358,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.4,
|
| 203 |
+
"grad_norm": 0.1126856803894043,
|
| 204 |
+
"learning_rate": 3.2512500000000004e-05,
|
| 205 |
+
"loss": 1.7683,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.45,
|
| 210 |
+
"grad_norm": 0.13452214002609253,
|
| 211 |
+
"learning_rate": 3.18875e-05,
|
| 212 |
+
"loss": 1.8371,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.5,
|
| 217 |
+
"grad_norm": 0.13590937852859497,
|
| 218 |
+
"learning_rate": 3.12625e-05,
|
| 219 |
+
"loss": 1.8016,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.55,
|
| 224 |
+
"grad_norm": 0.16644452512264252,
|
| 225 |
+
"learning_rate": 3.06375e-05,
|
| 226 |
+
"loss": 1.828,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.6,
|
| 231 |
+
"grad_norm": 0.1445678323507309,
|
| 232 |
+
"learning_rate": 3.0012499999999998e-05,
|
| 233 |
+
"loss": 1.8417,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.65,
|
| 238 |
+
"grad_norm": 0.10797803103923798,
|
| 239 |
+
"learning_rate": 2.9387500000000003e-05,
|
| 240 |
+
"loss": 1.8364,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.7,
|
| 245 |
+
"grad_norm": 0.14495541155338287,
|
| 246 |
+
"learning_rate": 2.8762500000000005e-05,
|
| 247 |
+
"loss": 1.7685,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.75,
|
| 252 |
+
"grad_norm": 0.19478319585323334,
|
| 253 |
+
"learning_rate": 2.81375e-05,
|
| 254 |
+
"loss": 1.8443,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.8,
|
| 259 |
+
"grad_norm": 0.1155581995844841,
|
| 260 |
+
"learning_rate": 2.75125e-05,
|
| 261 |
+
"loss": 1.8308,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.85,
|
| 266 |
+
"grad_norm": 0.16222324967384338,
|
| 267 |
+
"learning_rate": 2.68875e-05,
|
| 268 |
+
"loss": 1.8612,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.9,
|
| 273 |
+
"grad_norm": 0.12850694358348846,
|
| 274 |
+
"learning_rate": 2.62625e-05,
|
| 275 |
+
"loss": 1.7941,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.95,
|
| 280 |
+
"grad_norm": 0.15454839169979095,
|
| 281 |
+
"learning_rate": 2.5637500000000003e-05,
|
| 282 |
+
"loss": 1.8409,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.0,
|
| 287 |
+
"grad_norm": 0.13590778410434723,
|
| 288 |
+
"learning_rate": 2.50125e-05,
|
| 289 |
+
"loss": 1.8786,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.05,
|
| 294 |
+
"grad_norm": 0.16667228937149048,
|
| 295 |
+
"learning_rate": 2.4387500000000003e-05,
|
| 296 |
+
"loss": 1.8638,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.1,
|
| 301 |
+
"grad_norm": 0.1315215826034546,
|
| 302 |
+
"learning_rate": 2.37625e-05,
|
| 303 |
+
"loss": 1.8444,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.15,
|
| 308 |
+
"grad_norm": 0.12784986197948456,
|
| 309 |
+
"learning_rate": 2.31375e-05,
|
| 310 |
+
"loss": 1.8937,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.2,
|
| 315 |
+
"grad_norm": 0.1757715791463852,
|
| 316 |
+
"learning_rate": 2.2512500000000002e-05,
|
| 317 |
+
"loss": 1.8579,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.25,
|
| 322 |
+
"grad_norm": 0.13813862204551697,
|
| 323 |
+
"learning_rate": 2.18875e-05,
|
| 324 |
+
"loss": 1.8044,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.3,
|
| 329 |
+
"grad_norm": 0.1028163880109787,
|
| 330 |
+
"learning_rate": 2.1262500000000002e-05,
|
| 331 |
+
"loss": 1.8371,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.35,
|
| 336 |
+
"grad_norm": 0.1259194314479828,
|
| 337 |
+
"learning_rate": 2.06375e-05,
|
| 338 |
+
"loss": 1.7808,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.4,
|
| 343 |
+
"grad_norm": 0.14981134235858917,
|
| 344 |
+
"learning_rate": 2.0012500000000002e-05,
|
| 345 |
+
"loss": 1.791,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.45,
|
| 350 |
+
"grad_norm": 0.17923866212368011,
|
| 351 |
+
"learning_rate": 1.93875e-05,
|
| 352 |
+
"loss": 1.7812,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.5,
|
| 357 |
+
"grad_norm": 0.1661010980606079,
|
| 358 |
+
"learning_rate": 1.87625e-05,
|
| 359 |
+
"loss": 1.8722,
|
| 360 |
+
"step": 2500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.55,
|
| 364 |
+
"grad_norm": 0.1746157556772232,
|
| 365 |
+
"learning_rate": 1.81375e-05,
|
| 366 |
+
"loss": 1.7818,
|
| 367 |
+
"step": 2550
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.6,
|
| 371 |
+
"grad_norm": 0.13667204976081848,
|
| 372 |
+
"learning_rate": 1.7512500000000002e-05,
|
| 373 |
+
"loss": 1.8312,
|
| 374 |
+
"step": 2600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.65,
|
| 378 |
+
"grad_norm": 0.149635910987854,
|
| 379 |
+
"learning_rate": 1.68875e-05,
|
| 380 |
+
"loss": 1.813,
|
| 381 |
+
"step": 2650
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.7,
|
| 385 |
+
"grad_norm": 0.1654106080532074,
|
| 386 |
+
"learning_rate": 1.62625e-05,
|
| 387 |
+
"loss": 1.8739,
|
| 388 |
+
"step": 2700
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 2.75,
|
| 392 |
+
"grad_norm": 0.14472809433937073,
|
| 393 |
+
"learning_rate": 1.56375e-05,
|
| 394 |
+
"loss": 1.7726,
|
| 395 |
+
"step": 2750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 2.8,
|
| 399 |
+
"grad_norm": 0.14457519352436066,
|
| 400 |
+
"learning_rate": 1.5012500000000002e-05,
|
| 401 |
+
"loss": 1.7593,
|
| 402 |
+
"step": 2800
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 2.85,
|
| 406 |
+
"grad_norm": 0.12135745584964752,
|
| 407 |
+
"learning_rate": 1.43875e-05,
|
| 408 |
+
"loss": 1.8547,
|
| 409 |
+
"step": 2850
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 2.9,
|
| 413 |
+
"grad_norm": 0.18297483026981354,
|
| 414 |
+
"learning_rate": 1.37625e-05,
|
| 415 |
+
"loss": 1.852,
|
| 416 |
+
"step": 2900
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.95,
|
| 420 |
+
"grad_norm": 0.13699814677238464,
|
| 421 |
+
"learning_rate": 1.31375e-05,
|
| 422 |
+
"loss": 1.8064,
|
| 423 |
+
"step": 2950
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.0,
|
| 427 |
+
"grad_norm": 0.1180616170167923,
|
| 428 |
+
"learning_rate": 1.25125e-05,
|
| 429 |
+
"loss": 1.8457,
|
| 430 |
+
"step": 3000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 3.05,
|
| 434 |
+
"grad_norm": 0.13587267696857452,
|
| 435 |
+
"learning_rate": 1.18875e-05,
|
| 436 |
+
"loss": 1.8615,
|
| 437 |
+
"step": 3050
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.1,
|
| 441 |
+
"grad_norm": 0.13397973775863647,
|
| 442 |
+
"learning_rate": 1.1262500000000001e-05,
|
| 443 |
+
"loss": 1.8077,
|
| 444 |
+
"step": 3100
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.15,
|
| 448 |
+
"grad_norm": 0.13365299999713898,
|
| 449 |
+
"learning_rate": 1.0637500000000001e-05,
|
| 450 |
+
"loss": 1.9439,
|
| 451 |
+
"step": 3150
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 3.2,
|
| 455 |
+
"grad_norm": 0.1443161964416504,
|
| 456 |
+
"learning_rate": 1.0012500000000001e-05,
|
| 457 |
+
"loss": 1.8582,
|
| 458 |
+
"step": 3200
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.25,
|
| 462 |
+
"grad_norm": 0.14696797728538513,
|
| 463 |
+
"learning_rate": 9.387500000000001e-06,
|
| 464 |
+
"loss": 1.781,
|
| 465 |
+
"step": 3250
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.3,
|
| 469 |
+
"grad_norm": 0.15039022266864777,
|
| 470 |
+
"learning_rate": 8.7625e-06,
|
| 471 |
+
"loss": 1.8631,
|
| 472 |
+
"step": 3300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 3.35,
|
| 476 |
+
"grad_norm": 0.1311933696269989,
|
| 477 |
+
"learning_rate": 8.137500000000001e-06,
|
| 478 |
+
"loss": 1.8544,
|
| 479 |
+
"step": 3350
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 3.4,
|
| 483 |
+
"grad_norm": 0.15743456780910492,
|
| 484 |
+
"learning_rate": 7.5125000000000005e-06,
|
| 485 |
+
"loss": 1.7619,
|
| 486 |
+
"step": 3400
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 3.45,
|
| 490 |
+
"grad_norm": 0.1521286964416504,
|
| 491 |
+
"learning_rate": 6.8875000000000005e-06,
|
| 492 |
+
"loss": 1.8013,
|
| 493 |
+
"step": 3450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 3.5,
|
| 497 |
+
"grad_norm": 0.22098106145858765,
|
| 498 |
+
"learning_rate": 6.262500000000001e-06,
|
| 499 |
+
"loss": 1.799,
|
| 500 |
+
"step": 3500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.55,
|
| 504 |
+
"grad_norm": 0.17960400879383087,
|
| 505 |
+
"learning_rate": 5.637500000000001e-06,
|
| 506 |
+
"loss": 1.7757,
|
| 507 |
+
"step": 3550
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.6,
|
| 511 |
+
"grad_norm": 0.13921096920967102,
|
| 512 |
+
"learning_rate": 5.012500000000001e-06,
|
| 513 |
+
"loss": 1.8261,
|
| 514 |
+
"step": 3600
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 3.65,
|
| 518 |
+
"grad_norm": 0.1500893533229828,
|
| 519 |
+
"learning_rate": 4.3875e-06,
|
| 520 |
+
"loss": 1.8084,
|
| 521 |
+
"step": 3650
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 3.7,
|
| 525 |
+
"grad_norm": 0.16856829822063446,
|
| 526 |
+
"learning_rate": 3.7625e-06,
|
| 527 |
+
"loss": 1.8691,
|
| 528 |
+
"step": 3700
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 3.75,
|
| 532 |
+
"grad_norm": 0.14953631162643433,
|
| 533 |
+
"learning_rate": 3.1375e-06,
|
| 534 |
+
"loss": 1.7682,
|
| 535 |
+
"step": 3750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 3.8,
|
| 539 |
+
"grad_norm": 0.14736834168434143,
|
| 540 |
+
"learning_rate": 2.5125e-06,
|
| 541 |
+
"loss": 1.7549,
|
| 542 |
+
"step": 3800
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.85,
|
| 546 |
+
"grad_norm": 0.1247793510556221,
|
| 547 |
+
"learning_rate": 1.8875e-06,
|
| 548 |
+
"loss": 1.8505,
|
| 549 |
+
"step": 3850
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 3.9,
|
| 553 |
+
"grad_norm": 0.1904279887676239,
|
| 554 |
+
"learning_rate": 1.2625000000000002e-06,
|
| 555 |
+
"loss": 1.8485,
|
| 556 |
+
"step": 3900
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.95,
|
| 560 |
+
"grad_norm": 0.13930699229240417,
|
| 561 |
+
"learning_rate": 6.375e-07,
|
| 562 |
+
"loss": 1.8034,
|
| 563 |
+
"step": 3950
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.0,
|
| 567 |
+
"grad_norm": 0.1202196553349495,
|
| 568 |
+
"learning_rate": 1.2500000000000001e-08,
|
| 569 |
+
"loss": 1.8422,
|
| 570 |
+
"step": 4000
|
| 571 |
+
}
|
| 572 |
+
],
|
| 573 |
+
"logging_steps": 50,
|
| 574 |
+
"max_steps": 4000,
|
| 575 |
+
"num_input_tokens_seen": 0,
|
| 576 |
+
"num_train_epochs": 4,
|
| 577 |
+
"save_steps": 10,
|
| 578 |
+
"stateful_callbacks": {
|
| 579 |
+
"TrainerControl": {
|
| 580 |
+
"args": {
|
| 581 |
+
"should_epoch_stop": false,
|
| 582 |
+
"should_evaluate": false,
|
| 583 |
+
"should_log": false,
|
| 584 |
+
"should_save": true,
|
| 585 |
+
"should_training_stop": true
|
| 586 |
+
},
|
| 587 |
+
"attributes": {}
|
| 588 |
+
}
|
| 589 |
+
},
|
| 590 |
+
"total_flos": 5.360735748096e+16,
|
| 591 |
+
"train_batch_size": 10,
|
| 592 |
+
"trial_name": null,
|
| 593 |
+
"trial_params": null
|
| 594 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76d1d18100c90f968b787f300279b75f704d674df1dc4e0364972389fca68a1f
|
| 3 |
+
size 5304
|