Upload folder using huggingface_hub
Browse files- README.md +0 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +587 -0
- training_args.bin +3 -0
README.md
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d527f2f29f56324150701c2b7a31cc58b7e71efd21aa44fd7a37d87604f4642
|
| 3 |
+
size 1093944715
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92d956d5cc93a34ad408d50e3e3f95eae8755e6db418907b501095ac83002518
|
| 3 |
+
size 14391
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bdc8c206d7d4882b3d035e315b6ad0b072e853c267f16b096bb81bdc74fee7c
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,587 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3906,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"embedding_loss": 0.1733,
|
| 14 |
+
"epoch": 0.0002560163850486431,
|
| 15 |
+
"grad_norm": 1.3411569595336914,
|
| 16 |
+
"learning_rate": 0.0,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"embedding_loss": 0.1425,
|
| 21 |
+
"epoch": 0.012800819252432157,
|
| 22 |
+
"grad_norm": 1.5893903970718384,
|
| 23 |
+
"learning_rate": 2.5063938618925833e-06,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"embedding_loss": 0.0954,
|
| 28 |
+
"epoch": 0.025601638504864313,
|
| 29 |
+
"grad_norm": 1.3638615608215332,
|
| 30 |
+
"learning_rate": 5.0639386189258325e-06,
|
| 31 |
+
"step": 100
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"embedding_loss": 0.0483,
|
| 35 |
+
"epoch": 0.03840245775729647,
|
| 36 |
+
"grad_norm": 1.0745915174484253,
|
| 37 |
+
"learning_rate": 7.62148337595908e-06,
|
| 38 |
+
"step": 150
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"embedding_loss": 0.0297,
|
| 42 |
+
"epoch": 0.051203277009728626,
|
| 43 |
+
"grad_norm": 0.8514853715896606,
|
| 44 |
+
"learning_rate": 1.0179028132992328e-05,
|
| 45 |
+
"step": 200
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"embedding_loss": 0.0189,
|
| 49 |
+
"epoch": 0.06400409626216078,
|
| 50 |
+
"grad_norm": 0.7511357069015503,
|
| 51 |
+
"learning_rate": 1.2736572890025576e-05,
|
| 52 |
+
"step": 250
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"embedding_loss": 0.0122,
|
| 56 |
+
"epoch": 0.07680491551459294,
|
| 57 |
+
"grad_norm": 0.525897204875946,
|
| 58 |
+
"learning_rate": 1.5294117647058822e-05,
|
| 59 |
+
"step": 300
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"embedding_loss": 0.0099,
|
| 63 |
+
"epoch": 0.08960573476702509,
|
| 64 |
+
"grad_norm": 0.5095636248588562,
|
| 65 |
+
"learning_rate": 1.7851662404092073e-05,
|
| 66 |
+
"step": 350
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"embedding_loss": 0.0085,
|
| 70 |
+
"epoch": 0.10240655401945725,
|
| 71 |
+
"grad_norm": 1.5715042352676392,
|
| 72 |
+
"learning_rate": 1.9954480796586062e-05,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"embedding_loss": 0.0067,
|
| 77 |
+
"epoch": 0.1152073732718894,
|
| 78 |
+
"grad_norm": 0.27569064497947693,
|
| 79 |
+
"learning_rate": 1.9669985775248933e-05,
|
| 80 |
+
"step": 450
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"embedding_loss": 0.0059,
|
| 84 |
+
"epoch": 0.12800819252432155,
|
| 85 |
+
"grad_norm": 0.09678972512483597,
|
| 86 |
+
"learning_rate": 1.9385490753911807e-05,
|
| 87 |
+
"step": 500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"embedding_loss": 0.0076,
|
| 91 |
+
"epoch": 0.1408090117767537,
|
| 92 |
+
"grad_norm": 0.15479978919029236,
|
| 93 |
+
"learning_rate": 1.910099573257468e-05,
|
| 94 |
+
"step": 550
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"embedding_loss": 0.0061,
|
| 98 |
+
"epoch": 0.15360983102918588,
|
| 99 |
+
"grad_norm": 0.4873831272125244,
|
| 100 |
+
"learning_rate": 1.8816500711237555e-05,
|
| 101 |
+
"step": 600
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"embedding_loss": 0.0049,
|
| 105 |
+
"epoch": 0.16641065028161803,
|
| 106 |
+
"grad_norm": 0.09139434248209,
|
| 107 |
+
"learning_rate": 1.8532005689900426e-05,
|
| 108 |
+
"step": 650
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"embedding_loss": 0.003,
|
| 112 |
+
"epoch": 0.17921146953405018,
|
| 113 |
+
"grad_norm": 0.05168338865041733,
|
| 114 |
+
"learning_rate": 1.8247510668563304e-05,
|
| 115 |
+
"step": 700
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"embedding_loss": 0.0023,
|
| 119 |
+
"epoch": 0.19201228878648233,
|
| 120 |
+
"grad_norm": 0.05378415808081627,
|
| 121 |
+
"learning_rate": 1.7963015647226174e-05,
|
| 122 |
+
"step": 750
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"embedding_loss": 0.002,
|
| 126 |
+
"epoch": 0.2048131080389145,
|
| 127 |
+
"grad_norm": 0.026590052992105484,
|
| 128 |
+
"learning_rate": 1.7678520625889048e-05,
|
| 129 |
+
"step": 800
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"embedding_loss": 0.0037,
|
| 133 |
+
"epoch": 0.21761392729134665,
|
| 134 |
+
"grad_norm": 0.03566845878958702,
|
| 135 |
+
"learning_rate": 1.739402560455192e-05,
|
| 136 |
+
"step": 850
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"embedding_loss": 0.0007,
|
| 140 |
+
"epoch": 0.2304147465437788,
|
| 141 |
+
"grad_norm": 0.061990801244974136,
|
| 142 |
+
"learning_rate": 1.7109530583214796e-05,
|
| 143 |
+
"step": 900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"embedding_loss": 0.0015,
|
| 147 |
+
"epoch": 0.24321556579621095,
|
| 148 |
+
"grad_norm": 0.027544036507606506,
|
| 149 |
+
"learning_rate": 1.682503556187767e-05,
|
| 150 |
+
"step": 950
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"embedding_loss": 0.0017,
|
| 154 |
+
"epoch": 0.2560163850486431,
|
| 155 |
+
"grad_norm": 0.04399234429001808,
|
| 156 |
+
"learning_rate": 1.654054054054054e-05,
|
| 157 |
+
"step": 1000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"embedding_loss": 0.0014,
|
| 161 |
+
"epoch": 0.26881720430107525,
|
| 162 |
+
"grad_norm": 0.01828560046851635,
|
| 163 |
+
"learning_rate": 1.6256045519203415e-05,
|
| 164 |
+
"step": 1050
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"embedding_loss": 0.0003,
|
| 168 |
+
"epoch": 0.2816180235535074,
|
| 169 |
+
"grad_norm": 0.018775783479213715,
|
| 170 |
+
"learning_rate": 1.597155049786629e-05,
|
| 171 |
+
"step": 1100
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"embedding_loss": 0.0006,
|
| 175 |
+
"epoch": 0.2944188428059396,
|
| 176 |
+
"grad_norm": 0.01576610654592514,
|
| 177 |
+
"learning_rate": 1.5687055476529163e-05,
|
| 178 |
+
"step": 1150
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"embedding_loss": 0.001,
|
| 182 |
+
"epoch": 0.30721966205837176,
|
| 183 |
+
"grad_norm": 0.14026065170764923,
|
| 184 |
+
"learning_rate": 1.5402560455192034e-05,
|
| 185 |
+
"step": 1200
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"embedding_loss": 0.0007,
|
| 189 |
+
"epoch": 0.3200204813108039,
|
| 190 |
+
"grad_norm": 0.01965928263962269,
|
| 191 |
+
"learning_rate": 1.511806543385491e-05,
|
| 192 |
+
"step": 1250
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"embedding_loss": 0.0002,
|
| 196 |
+
"epoch": 0.33282130056323606,
|
| 197 |
+
"grad_norm": 0.014394218102097511,
|
| 198 |
+
"learning_rate": 1.4833570412517782e-05,
|
| 199 |
+
"step": 1300
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"embedding_loss": 0.0012,
|
| 203 |
+
"epoch": 0.3456221198156682,
|
| 204 |
+
"grad_norm": 0.018288280814886093,
|
| 205 |
+
"learning_rate": 1.4549075391180656e-05,
|
| 206 |
+
"step": 1350
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"embedding_loss": 0.0006,
|
| 210 |
+
"epoch": 0.35842293906810035,
|
| 211 |
+
"grad_norm": 0.017536135390400887,
|
| 212 |
+
"learning_rate": 1.4264580369843529e-05,
|
| 213 |
+
"step": 1400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"embedding_loss": 0.0003,
|
| 217 |
+
"epoch": 0.3712237583205325,
|
| 218 |
+
"grad_norm": 0.012779198586940765,
|
| 219 |
+
"learning_rate": 1.3980085348506403e-05,
|
| 220 |
+
"step": 1450
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"embedding_loss": 0.0005,
|
| 224 |
+
"epoch": 0.38402457757296465,
|
| 225 |
+
"grad_norm": 0.19666947424411774,
|
| 226 |
+
"learning_rate": 1.3695590327169275e-05,
|
| 227 |
+
"step": 1500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"embedding_loss": 0.0002,
|
| 231 |
+
"epoch": 0.3968253968253968,
|
| 232 |
+
"grad_norm": 0.008438820950686932,
|
| 233 |
+
"learning_rate": 1.341109530583215e-05,
|
| 234 |
+
"step": 1550
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"embedding_loss": 0.0004,
|
| 238 |
+
"epoch": 0.409626216077829,
|
| 239 |
+
"grad_norm": 0.009661088697612286,
|
| 240 |
+
"learning_rate": 1.3126600284495022e-05,
|
| 241 |
+
"step": 1600
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"embedding_loss": 0.0009,
|
| 245 |
+
"epoch": 0.42242703533026116,
|
| 246 |
+
"grad_norm": 0.02346787601709366,
|
| 247 |
+
"learning_rate": 1.2842105263157896e-05,
|
| 248 |
+
"step": 1650
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"embedding_loss": 0.0007,
|
| 252 |
+
"epoch": 0.4352278545826933,
|
| 253 |
+
"grad_norm": 0.0062308646738529205,
|
| 254 |
+
"learning_rate": 1.255761024182077e-05,
|
| 255 |
+
"step": 1700
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"embedding_loss": 0.0003,
|
| 259 |
+
"epoch": 0.44802867383512546,
|
| 260 |
+
"grad_norm": 0.013097619637846947,
|
| 261 |
+
"learning_rate": 1.2273115220483642e-05,
|
| 262 |
+
"step": 1750
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"embedding_loss": 0.0001,
|
| 266 |
+
"epoch": 0.4608294930875576,
|
| 267 |
+
"grad_norm": 0.008484157733619213,
|
| 268 |
+
"learning_rate": 1.1988620199146516e-05,
|
| 269 |
+
"step": 1800
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"embedding_loss": 0.0006,
|
| 273 |
+
"epoch": 0.47363031233998976,
|
| 274 |
+
"grad_norm": 0.010278033092617989,
|
| 275 |
+
"learning_rate": 1.1704125177809389e-05,
|
| 276 |
+
"step": 1850
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"embedding_loss": 0.0004,
|
| 280 |
+
"epoch": 0.4864311315924219,
|
| 281 |
+
"grad_norm": 0.012740347534418106,
|
| 282 |
+
"learning_rate": 1.1419630156472263e-05,
|
| 283 |
+
"step": 1900
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"embedding_loss": 0.0004,
|
| 287 |
+
"epoch": 0.49923195084485406,
|
| 288 |
+
"grad_norm": 0.022875774651765823,
|
| 289 |
+
"learning_rate": 1.1135135135135135e-05,
|
| 290 |
+
"step": 1950
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"embedding_loss": 0.0001,
|
| 294 |
+
"epoch": 0.5120327700972862,
|
| 295 |
+
"grad_norm": 0.007295957300812006,
|
| 296 |
+
"learning_rate": 1.085064011379801e-05,
|
| 297 |
+
"step": 2000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"embedding_loss": 0.0004,
|
| 301 |
+
"epoch": 0.5248335893497184,
|
| 302 |
+
"grad_norm": 0.008994905278086662,
|
| 303 |
+
"learning_rate": 1.0566145092460882e-05,
|
| 304 |
+
"step": 2050
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"embedding_loss": 0.0003,
|
| 308 |
+
"epoch": 0.5376344086021505,
|
| 309 |
+
"grad_norm": 0.01373240165412426,
|
| 310 |
+
"learning_rate": 1.0281650071123756e-05,
|
| 311 |
+
"step": 2100
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"embedding_loss": 0.0012,
|
| 315 |
+
"epoch": 0.5504352278545827,
|
| 316 |
+
"grad_norm": 0.017664149403572083,
|
| 317 |
+
"learning_rate": 9.99715504978663e-06,
|
| 318 |
+
"step": 2150
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"embedding_loss": 0.0006,
|
| 322 |
+
"epoch": 0.5632360471070148,
|
| 323 |
+
"grad_norm": 0.026641126722097397,
|
| 324 |
+
"learning_rate": 9.712660028449504e-06,
|
| 325 |
+
"step": 2200
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"embedding_loss": 0.0001,
|
| 329 |
+
"epoch": 0.576036866359447,
|
| 330 |
+
"grad_norm": 0.015163728035986423,
|
| 331 |
+
"learning_rate": 9.428165007112376e-06,
|
| 332 |
+
"step": 2250
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"embedding_loss": 0.0003,
|
| 336 |
+
"epoch": 0.5888376856118792,
|
| 337 |
+
"grad_norm": 0.014809815213084221,
|
| 338 |
+
"learning_rate": 9.14366998577525e-06,
|
| 339 |
+
"step": 2300
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"embedding_loss": 0.0007,
|
| 343 |
+
"epoch": 0.6016385048643114,
|
| 344 |
+
"grad_norm": 0.007227804511785507,
|
| 345 |
+
"learning_rate": 8.859174964438123e-06,
|
| 346 |
+
"step": 2350
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"embedding_loss": 0.0003,
|
| 350 |
+
"epoch": 0.6144393241167435,
|
| 351 |
+
"grad_norm": 0.010935621336102486,
|
| 352 |
+
"learning_rate": 8.574679943100997e-06,
|
| 353 |
+
"step": 2400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"embedding_loss": 0.0004,
|
| 357 |
+
"epoch": 0.6272401433691757,
|
| 358 |
+
"grad_norm": 0.0073186722584068775,
|
| 359 |
+
"learning_rate": 8.29018492176387e-06,
|
| 360 |
+
"step": 2450
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"embedding_loss": 0.0006,
|
| 364 |
+
"epoch": 0.6400409626216078,
|
| 365 |
+
"grad_norm": 0.006618503015488386,
|
| 366 |
+
"learning_rate": 8.005689900426743e-06,
|
| 367 |
+
"step": 2500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"embedding_loss": 0.0001,
|
| 371 |
+
"epoch": 0.65284178187404,
|
| 372 |
+
"grad_norm": 0.008351747877895832,
|
| 373 |
+
"learning_rate": 7.721194879089616e-06,
|
| 374 |
+
"step": 2550
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"embedding_loss": 0.0003,
|
| 378 |
+
"epoch": 0.6656426011264721,
|
| 379 |
+
"grad_norm": 0.008504342287778854,
|
| 380 |
+
"learning_rate": 7.43669985775249e-06,
|
| 381 |
+
"step": 2600
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"embedding_loss": 0.0006,
|
| 385 |
+
"epoch": 0.6784434203789043,
|
| 386 |
+
"grad_norm": 0.015397731214761734,
|
| 387 |
+
"learning_rate": 7.152204836415363e-06,
|
| 388 |
+
"step": 2650
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"embedding_loss": 0.001,
|
| 392 |
+
"epoch": 0.6912442396313364,
|
| 393 |
+
"grad_norm": 0.004343625158071518,
|
| 394 |
+
"learning_rate": 6.867709815078236e-06,
|
| 395 |
+
"step": 2700
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"embedding_loss": 0.0004,
|
| 399 |
+
"epoch": 0.7040450588837686,
|
| 400 |
+
"grad_norm": 0.010210598818957806,
|
| 401 |
+
"learning_rate": 6.58321479374111e-06,
|
| 402 |
+
"step": 2750
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"embedding_loss": 0.0008,
|
| 406 |
+
"epoch": 0.7168458781362007,
|
| 407 |
+
"grad_norm": 0.021265419200062752,
|
| 408 |
+
"learning_rate": 6.2987197724039836e-06,
|
| 409 |
+
"step": 2800
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"embedding_loss": 0.0003,
|
| 413 |
+
"epoch": 0.7296466973886329,
|
| 414 |
+
"grad_norm": 0.006350652314722538,
|
| 415 |
+
"learning_rate": 6.014224751066858e-06,
|
| 416 |
+
"step": 2850
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"embedding_loss": 0.0007,
|
| 420 |
+
"epoch": 0.742447516641065,
|
| 421 |
+
"grad_norm": 0.042495131492614746,
|
| 422 |
+
"learning_rate": 5.729729729729731e-06,
|
| 423 |
+
"step": 2900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"embedding_loss": 0.0007,
|
| 427 |
+
"epoch": 0.7552483358934972,
|
| 428 |
+
"grad_norm": 0.003976090345531702,
|
| 429 |
+
"learning_rate": 5.445234708392604e-06,
|
| 430 |
+
"step": 2950
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"embedding_loss": 0.0007,
|
| 434 |
+
"epoch": 0.7680491551459293,
|
| 435 |
+
"grad_norm": 0.008590229786932468,
|
| 436 |
+
"learning_rate": 5.160739687055477e-06,
|
| 437 |
+
"step": 3000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"embedding_loss": 0.0003,
|
| 441 |
+
"epoch": 0.7808499743983615,
|
| 442 |
+
"grad_norm": 0.006639127153903246,
|
| 443 |
+
"learning_rate": 4.8762446657183506e-06,
|
| 444 |
+
"step": 3050
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"embedding_loss": 0.0003,
|
| 448 |
+
"epoch": 0.7936507936507936,
|
| 449 |
+
"grad_norm": 0.006947138346731663,
|
| 450 |
+
"learning_rate": 4.591749644381224e-06,
|
| 451 |
+
"step": 3100
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"embedding_loss": 0.0003,
|
| 455 |
+
"epoch": 0.8064516129032258,
|
| 456 |
+
"grad_norm": 0.00519227422773838,
|
| 457 |
+
"learning_rate": 4.307254623044097e-06,
|
| 458 |
+
"step": 3150
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"embedding_loss": 0.0007,
|
| 462 |
+
"epoch": 0.819252432155658,
|
| 463 |
+
"grad_norm": 0.013285805471241474,
|
| 464 |
+
"learning_rate": 4.02275960170697e-06,
|
| 465 |
+
"step": 3200
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"embedding_loss": 0.0001,
|
| 469 |
+
"epoch": 0.8320532514080902,
|
| 470 |
+
"grad_norm": 0.010509872809052467,
|
| 471 |
+
"learning_rate": 3.7382645803698435e-06,
|
| 472 |
+
"step": 3250
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"embedding_loss": 0.0006,
|
| 476 |
+
"epoch": 0.8448540706605223,
|
| 477 |
+
"grad_norm": 0.0032197178807109594,
|
| 478 |
+
"learning_rate": 3.4537695590327167e-06,
|
| 479 |
+
"step": 3300
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"embedding_loss": 0.0007,
|
| 483 |
+
"epoch": 0.8576548899129545,
|
| 484 |
+
"grad_norm": 0.007785377558320761,
|
| 485 |
+
"learning_rate": 3.169274537695591e-06,
|
| 486 |
+
"step": 3350
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"embedding_loss": 0.0003,
|
| 490 |
+
"epoch": 0.8704557091653866,
|
| 491 |
+
"grad_norm": 0.0052650910802185535,
|
| 492 |
+
"learning_rate": 2.884779516358464e-06,
|
| 493 |
+
"step": 3400
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"embedding_loss": 0.0009,
|
| 497 |
+
"epoch": 0.8832565284178188,
|
| 498 |
+
"grad_norm": 0.020432407036423683,
|
| 499 |
+
"learning_rate": 2.6002844950213373e-06,
|
| 500 |
+
"step": 3450
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"embedding_loss": 0.0003,
|
| 504 |
+
"epoch": 0.8960573476702509,
|
| 505 |
+
"grad_norm": 0.009411387145519257,
|
| 506 |
+
"learning_rate": 2.3157894736842105e-06,
|
| 507 |
+
"step": 3500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"embedding_loss": 0.0004,
|
| 511 |
+
"epoch": 0.9088581669226831,
|
| 512 |
+
"grad_norm": 0.007589740678668022,
|
| 513 |
+
"learning_rate": 2.031294452347084e-06,
|
| 514 |
+
"step": 3550
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"embedding_loss": 0.0004,
|
| 518 |
+
"epoch": 0.9216589861751152,
|
| 519 |
+
"grad_norm": 0.12826202809810638,
|
| 520 |
+
"learning_rate": 1.7467994310099576e-06,
|
| 521 |
+
"step": 3600
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"embedding_loss": 0.0001,
|
| 525 |
+
"epoch": 0.9344598054275474,
|
| 526 |
+
"grad_norm": 0.012935018166899681,
|
| 527 |
+
"learning_rate": 1.4623044096728308e-06,
|
| 528 |
+
"step": 3650
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"embedding_loss": 0.0001,
|
| 532 |
+
"epoch": 0.9472606246799795,
|
| 533 |
+
"grad_norm": 0.008784984238445759,
|
| 534 |
+
"learning_rate": 1.1778093883357043e-06,
|
| 535 |
+
"step": 3700
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"embedding_loss": 0.0007,
|
| 539 |
+
"epoch": 0.9600614439324117,
|
| 540 |
+
"grad_norm": 0.004482260439544916,
|
| 541 |
+
"learning_rate": 8.933143669985776e-07,
|
| 542 |
+
"step": 3750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"embedding_loss": 0.0006,
|
| 546 |
+
"epoch": 0.9728622631848438,
|
| 547 |
+
"grad_norm": 0.004357804544270039,
|
| 548 |
+
"learning_rate": 6.08819345661451e-07,
|
| 549 |
+
"step": 3800
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"embedding_loss": 0.0005,
|
| 553 |
+
"epoch": 0.985663082437276,
|
| 554 |
+
"grad_norm": 0.014645060524344444,
|
| 555 |
+
"learning_rate": 3.2432432432432436e-07,
|
| 556 |
+
"step": 3850
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"embedding_loss": 0.0003,
|
| 560 |
+
"epoch": 0.9984639016897081,
|
| 561 |
+
"grad_norm": 0.005299082491546869,
|
| 562 |
+
"learning_rate": 3.9829302987197725e-08,
|
| 563 |
+
"step": 3900
|
| 564 |
+
}
|
| 565 |
+
],
|
| 566 |
+
"logging_steps": 50,
|
| 567 |
+
"max_steps": 3906,
|
| 568 |
+
"num_input_tokens_seen": 0,
|
| 569 |
+
"num_train_epochs": 1,
|
| 570 |
+
"save_steps": 500,
|
| 571 |
+
"stateful_callbacks": {
|
| 572 |
+
"TrainerControl": {
|
| 573 |
+
"args": {
|
| 574 |
+
"should_epoch_stop": false,
|
| 575 |
+
"should_evaluate": false,
|
| 576 |
+
"should_log": false,
|
| 577 |
+
"should_save": true,
|
| 578 |
+
"should_training_stop": true
|
| 579 |
+
},
|
| 580 |
+
"attributes": {}
|
| 581 |
+
}
|
| 582 |
+
},
|
| 583 |
+
"total_flos": 0.0,
|
| 584 |
+
"train_batch_size": 64,
|
| 585 |
+
"trial_name": null,
|
| 586 |
+
"trial_params": null
|
| 587 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a05cf3f11eec5c5eed6224b136087f0c4182dc0d1a243494d5822b939bd049aa
|
| 3 |
+
size 5969
|