Training in progress, step 6000, checkpoint
Browse files
checkpoint-6000/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"annotation_transformer": true,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"CAMP"
|
| 5 |
+
],
|
| 6 |
+
"diff": false,
|
| 7 |
+
"dropout": 0.05,
|
| 8 |
+
"hidden_dim": 640,
|
| 9 |
+
"input_dim": 384,
|
| 10 |
+
"intermediate_dim": 2560,
|
| 11 |
+
"kernel_size": 11,
|
| 12 |
+
"latent": false,
|
| 13 |
+
"mlm": false,
|
| 14 |
+
"mnr": false,
|
| 15 |
+
"model_type": "CAMP",
|
| 16 |
+
"nhead": 8,
|
| 17 |
+
"nlp_path": "lhallee/annotation_transformer_uniref90",
|
| 18 |
+
"num_hidden_layers": 1,
|
| 19 |
+
"out_dim": 512,
|
| 20 |
+
"plm_path": "facebook/esm2_t33_650m_UR50D",
|
| 21 |
+
"pooling": "avg",
|
| 22 |
+
"token": null,
|
| 23 |
+
"torch_dtype": "bfloat16",
|
| 24 |
+
"transformers_version": "4.41.2"
|
| 25 |
+
}
|
checkpoint-6000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c2cb8e69bd5cc1eb4c07c41f97ca6e659278830ea847fb3505c2f4049e201301
|
| 3 |
+
size 1384137156
|
checkpoint-6000/trainer_state.json
ADDED
|
@@ -0,0 +1,519 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.37230081906180196,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 6000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0062050136510300325,
|
| 13 |
+
"grad_norm": 101.88961029052734,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 4.5088,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.012410027302060065,
|
| 20 |
+
"grad_norm": 99.7363510131836,
|
| 21 |
+
"learning_rate": 9.999038127056248e-05,
|
| 22 |
+
"loss": 0.8644,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.018615040953090096,
|
| 27 |
+
"grad_norm": 59.41786575317383,
|
| 28 |
+
"learning_rate": 9.996152878304816e-05,
|
| 29 |
+
"loss": 0.7189,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02482005460412013,
|
| 34 |
+
"grad_norm": 61.05416488647461,
|
| 35 |
+
"learning_rate": 9.991345363842789e-05,
|
| 36 |
+
"loss": 0.6253,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03102506825515016,
|
| 41 |
+
"grad_norm": 53.87808609008789,
|
| 42 |
+
"learning_rate": 9.9846174333574e-05,
|
| 43 |
+
"loss": 0.5863,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03723008190618019,
|
| 48 |
+
"grad_norm": 48.58293914794922,
|
| 49 |
+
"learning_rate": 9.975971675414371e-05,
|
| 50 |
+
"loss": 0.5555,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04343509555721023,
|
| 55 |
+
"grad_norm": 42.353153228759766,
|
| 56 |
+
"learning_rate": 9.965411416461959e-05,
|
| 57 |
+
"loss": 0.5566,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04964010920824026,
|
| 62 |
+
"grad_norm": 46.06748580932617,
|
| 63 |
+
"learning_rate": 9.952940719551112e-05,
|
| 64 |
+
"loss": 0.5425,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05584512285927029,
|
| 69 |
+
"grad_norm": 46.486637115478516,
|
| 70 |
+
"learning_rate": 9.938564382772205e-05,
|
| 71 |
+
"loss": 0.5004,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06205013651030032,
|
| 76 |
+
"grad_norm": 43.99421691894531,
|
| 77 |
+
"learning_rate": 9.922287937408994e-05,
|
| 78 |
+
"loss": 0.4722,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.06205013651030032,
|
| 83 |
+
"eval_avg_non_pair_similarity": -0.00011276310920446573,
|
| 84 |
+
"eval_avg_pair_similarity": 0.0020758428336121143,
|
| 85 |
+
"eval_loss": 0.44835320115089417,
|
| 86 |
+
"eval_runtime": 52.6112,
|
| 87 |
+
"eval_samples_per_second": 9.504,
|
| 88 |
+
"eval_similarity_ratio": -18.408882552609725,
|
| 89 |
+
"eval_steps_per_second": 0.304,
|
| 90 |
+
"step": 1000
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.06825515016133035,
|
| 94 |
+
"grad_norm": 38.14698791503906,
|
| 95 |
+
"learning_rate": 9.904117645810441e-05,
|
| 96 |
+
"loss": 0.4444,
|
| 97 |
+
"step": 1100
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.07446016381236038,
|
| 101 |
+
"grad_norm": 29.508113861083984,
|
| 102 |
+
"learning_rate": 9.884060498981296e-05,
|
| 103 |
+
"loss": 0.4615,
|
| 104 |
+
"step": 1200
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.08066517746339041,
|
| 108 |
+
"grad_norm": 34.64622497558594,
|
| 109 |
+
"learning_rate": 9.862124213892304e-05,
|
| 110 |
+
"loss": 0.4665,
|
| 111 |
+
"step": 1300
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.08687019111442046,
|
| 115 |
+
"grad_norm": 40.75010299682617,
|
| 116 |
+
"learning_rate": 9.838317230511112e-05,
|
| 117 |
+
"loss": 0.4275,
|
| 118 |
+
"step": 1400
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.09307520476545049,
|
| 122 |
+
"grad_norm": 28.284589767456055,
|
| 123 |
+
"learning_rate": 9.81264870855499e-05,
|
| 124 |
+
"loss": 0.4129,
|
| 125 |
+
"step": 1500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.09928021841648052,
|
| 129 |
+
"grad_norm": 34.2197151184082,
|
| 130 |
+
"learning_rate": 9.785128523966653e-05,
|
| 131 |
+
"loss": 0.4389,
|
| 132 |
+
"step": 1600
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.10548523206751055,
|
| 136 |
+
"grad_norm": 27.808895111083984,
|
| 137 |
+
"learning_rate": 9.755767265114484e-05,
|
| 138 |
+
"loss": 0.4184,
|
| 139 |
+
"step": 1700
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.11169024571854058,
|
| 143 |
+
"grad_norm": 35.00907897949219,
|
| 144 |
+
"learning_rate": 9.724576228718678e-05,
|
| 145 |
+
"loss": 0.3995,
|
| 146 |
+
"step": 1800
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.11789525936957061,
|
| 150 |
+
"grad_norm": 30.57769203186035,
|
| 151 |
+
"learning_rate": 9.691567415504832e-05,
|
| 152 |
+
"loss": 0.415,
|
| 153 |
+
"step": 1900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.12410027302060064,
|
| 157 |
+
"grad_norm": 26.989404678344727,
|
| 158 |
+
"learning_rate": 9.656753525586681e-05,
|
| 159 |
+
"loss": 0.4052,
|
| 160 |
+
"step": 2000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.12410027302060064,
|
| 164 |
+
"eval_avg_non_pair_similarity": 0.0006865022985091086,
|
| 165 |
+
"eval_avg_pair_similarity": 0.007665629971772433,
|
| 166 |
+
"eval_loss": 0.3601702153682709,
|
| 167 |
+
"eval_runtime": 52.565,
|
| 168 |
+
"eval_samples_per_second": 9.512,
|
| 169 |
+
"eval_similarity_ratio": 11.166211662247951,
|
| 170 |
+
"eval_steps_per_second": 0.304,
|
| 171 |
+
"step": 2000
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1303052866716307,
|
| 175 |
+
"grad_norm": 32.37137222290039,
|
| 176 |
+
"learning_rate": 9.620147953579737e-05,
|
| 177 |
+
"loss": 0.3979,
|
| 178 |
+
"step": 2100
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.1365103003226607,
|
| 182 |
+
"grad_norm": 22.6475772857666,
|
| 183 |
+
"learning_rate": 9.581764783447719e-05,
|
| 184 |
+
"loss": 0.3807,
|
| 185 |
+
"step": 2200
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.14271531397369075,
|
| 189 |
+
"grad_norm": 22.612077713012695,
|
| 190 |
+
"learning_rate": 9.54161878308377e-05,
|
| 191 |
+
"loss": 0.3711,
|
| 192 |
+
"step": 2300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.14892032762472077,
|
| 196 |
+
"grad_norm": 30.973310470581055,
|
| 197 |
+
"learning_rate": 9.499725398628507e-05,
|
| 198 |
+
"loss": 0.3658,
|
| 199 |
+
"step": 2400
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1551253412757508,
|
| 203 |
+
"grad_norm": 24.443492889404297,
|
| 204 |
+
"learning_rate": 9.456100748527143e-05,
|
| 205 |
+
"loss": 0.37,
|
| 206 |
+
"step": 2500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.16133035492678083,
|
| 210 |
+
"grad_norm": 33.75666809082031,
|
| 211 |
+
"learning_rate": 9.410761617327921e-05,
|
| 212 |
+
"loss": 0.3512,
|
| 213 |
+
"step": 2600
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.16753536857781087,
|
| 217 |
+
"grad_norm": 23.727365493774414,
|
| 218 |
+
"learning_rate": 9.363725449224282e-05,
|
| 219 |
+
"loss": 0.3318,
|
| 220 |
+
"step": 2700
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.17374038222884092,
|
| 224 |
+
"grad_norm": 28.48720359802246,
|
| 225 |
+
"learning_rate": 9.315010341343213e-05,
|
| 226 |
+
"loss": 0.3729,
|
| 227 |
+
"step": 2800
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.17994539587987093,
|
| 231 |
+
"grad_norm": 28.53714370727539,
|
| 232 |
+
"learning_rate": 9.264635036782405e-05,
|
| 233 |
+
"loss": 0.3369,
|
| 234 |
+
"step": 2900
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.18615040953090098,
|
| 238 |
+
"grad_norm": 21.766740798950195,
|
| 239 |
+
"learning_rate": 9.212618917398855e-05,
|
| 240 |
+
"loss": 0.317,
|
| 241 |
+
"step": 3000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.18615040953090098,
|
| 245 |
+
"eval_avg_non_pair_similarity": 0.00025785160030682527,
|
| 246 |
+
"eval_avg_pair_similarity": 0.0073178326906636355,
|
| 247 |
+
"eval_loss": 0.323689728975296,
|
| 248 |
+
"eval_runtime": 52.8381,
|
| 249 |
+
"eval_samples_per_second": 9.463,
|
| 250 |
+
"eval_similarity_ratio": 28.38001657525464,
|
| 251 |
+
"eval_steps_per_second": 0.303,
|
| 252 |
+
"step": 3000
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.192355423181931,
|
| 256 |
+
"grad_norm": 25.073217391967773,
|
| 257 |
+
"learning_rate": 9.158981996351711e-05,
|
| 258 |
+
"loss": 0.3092,
|
| 259 |
+
"step": 3100
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.19856043683296104,
|
| 263 |
+
"grad_norm": 25.02987289428711,
|
| 264 |
+
"learning_rate": 9.10374491040223e-05,
|
| 265 |
+
"loss": 0.3177,
|
| 266 |
+
"step": 3200
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.20476545048399106,
|
| 270 |
+
"grad_norm": 23.147781372070312,
|
| 271 |
+
"learning_rate": 9.046928911973799e-05,
|
| 272 |
+
"loss": 0.3357,
|
| 273 |
+
"step": 3300
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.2109704641350211,
|
| 277 |
+
"grad_norm": 18.975126266479492,
|
| 278 |
+
"learning_rate": 8.988555860975082e-05,
|
| 279 |
+
"loss": 0.2946,
|
| 280 |
+
"step": 3400
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.21717547778605112,
|
| 284 |
+
"grad_norm": 22.149898529052734,
|
| 285 |
+
"learning_rate": 8.92864821638944e-05,
|
| 286 |
+
"loss": 0.3184,
|
| 287 |
+
"step": 3500
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.22338049143708116,
|
| 291 |
+
"grad_norm": 31.714447021484375,
|
| 292 |
+
"learning_rate": 8.86722902763385e-05,
|
| 293 |
+
"loss": 0.315,
|
| 294 |
+
"step": 3600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.22958550508811118,
|
| 298 |
+
"grad_norm": 22.137062072753906,
|
| 299 |
+
"learning_rate": 8.804321925690672e-05,
|
| 300 |
+
"loss": 0.3007,
|
| 301 |
+
"step": 3700
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 0.23579051873914122,
|
| 305 |
+
"grad_norm": 27.34349250793457,
|
| 306 |
+
"learning_rate": 8.739951114015636e-05,
|
| 307 |
+
"loss": 0.3297,
|
| 308 |
+
"step": 3800
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.24199553239017127,
|
| 312 |
+
"grad_norm": 20.715852737426758,
|
| 313 |
+
"learning_rate": 8.674141359225589e-05,
|
| 314 |
+
"loss": 0.2943,
|
| 315 |
+
"step": 3900
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.2482005460412013,
|
| 319 |
+
"grad_norm": 21.26046371459961,
|
| 320 |
+
"learning_rate": 8.60691798156956e-05,
|
| 321 |
+
"loss": 0.284,
|
| 322 |
+
"step": 4000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.2482005460412013,
|
| 326 |
+
"eval_avg_non_pair_similarity": 0.00036512027781798257,
|
| 327 |
+
"eval_avg_pair_similarity": 0.002839699084404856,
|
| 328 |
+
"eval_loss": 0.26940277218818665,
|
| 329 |
+
"eval_runtime": 52.6691,
|
| 330 |
+
"eval_samples_per_second": 9.493,
|
| 331 |
+
"eval_similarity_ratio": 7.777434606961175,
|
| 332 |
+
"eval_steps_per_second": 0.304,
|
| 333 |
+
"step": 4000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.25440555969223133,
|
| 337 |
+
"grad_norm": 22.599706649780273,
|
| 338 |
+
"learning_rate": 8.538306845186808e-05,
|
| 339 |
+
"loss": 0.2942,
|
| 340 |
+
"step": 4100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.2606105733432614,
|
| 344 |
+
"grad_norm": 16.177858352661133,
|
| 345 |
+
"learning_rate": 8.468334348155624e-05,
|
| 346 |
+
"loss": 0.2874,
|
| 347 |
+
"step": 4200
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.26681558699429136,
|
| 351 |
+
"grad_norm": 15.479412078857422,
|
| 352 |
+
"learning_rate": 8.39702741233669e-05,
|
| 353 |
+
"loss": 0.2758,
|
| 354 |
+
"step": 4300
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.2730206006453214,
|
| 358 |
+
"grad_norm": 20.40460777282715,
|
| 359 |
+
"learning_rate": 8.324413473014913e-05,
|
| 360 |
+
"loss": 0.2788,
|
| 361 |
+
"step": 4400
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.27922561429635145,
|
| 365 |
+
"grad_norm": 15.524222373962402,
|
| 366 |
+
"learning_rate": 8.250520468343722e-05,
|
| 367 |
+
"loss": 0.2667,
|
| 368 |
+
"step": 4500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.2854306279473815,
|
| 372 |
+
"grad_norm": 19.124486923217773,
|
| 373 |
+
"learning_rate": 8.175376828595885e-05,
|
| 374 |
+
"loss": 0.2828,
|
| 375 |
+
"step": 4600
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.29163564159841154,
|
| 379 |
+
"grad_norm": 14.729342460632324,
|
| 380 |
+
"learning_rate": 8.09901146522499e-05,
|
| 381 |
+
"loss": 0.2757,
|
| 382 |
+
"step": 4700
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.29784065524944153,
|
| 386 |
+
"grad_norm": 19.547435760498047,
|
| 387 |
+
"learning_rate": 8.021453759741784e-05,
|
| 388 |
+
"loss": 0.3102,
|
| 389 |
+
"step": 4800
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.3040456689004716,
|
| 393 |
+
"grad_norm": 16.095699310302734,
|
| 394 |
+
"learning_rate": 7.942733552409662e-05,
|
| 395 |
+
"loss": 0.2791,
|
| 396 |
+
"step": 4900
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.3102506825515016,
|
| 400 |
+
"grad_norm": 25.538352966308594,
|
| 401 |
+
"learning_rate": 7.862881130763646e-05,
|
| 402 |
+
"loss": 0.2487,
|
| 403 |
+
"step": 5000
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.3102506825515016,
|
| 407 |
+
"eval_avg_non_pair_similarity": -2.6742616108896766e-05,
|
| 408 |
+
"eval_avg_pair_similarity": 0.007982193220406771,
|
| 409 |
+
"eval_loss": 0.3381979465484619,
|
| 410 |
+
"eval_runtime": 52.9305,
|
| 411 |
+
"eval_samples_per_second": 9.446,
|
| 412 |
+
"eval_similarity_ratio": -298.4821375703496,
|
| 413 |
+
"eval_steps_per_second": 0.302,
|
| 414 |
+
"step": 5000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.31645569620253167,
|
| 418 |
+
"grad_norm": 19.506053924560547,
|
| 419 |
+
"learning_rate": 7.781927217957286e-05,
|
| 420 |
+
"loss": 0.2534,
|
| 421 |
+
"step": 5100
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 0.32266070985356166,
|
| 425 |
+
"grad_norm": 17.35991668701172,
|
| 426 |
+
"learning_rate": 7.699902960941952e-05,
|
| 427 |
+
"loss": 0.2591,
|
| 428 |
+
"step": 5200
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.3288657235045917,
|
| 432 |
+
"grad_norm": 18.530628204345703,
|
| 433 |
+
"learning_rate": 7.616839918483061e-05,
|
| 434 |
+
"loss": 0.2504,
|
| 435 |
+
"step": 5300
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.33507073715562175,
|
| 439 |
+
"grad_norm": 16.918973922729492,
|
| 440 |
+
"learning_rate": 7.532770049017883e-05,
|
| 441 |
+
"loss": 0.2697,
|
| 442 |
+
"step": 5400
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.3412757508066518,
|
| 446 |
+
"grad_norm": 15.549245834350586,
|
| 447 |
+
"learning_rate": 7.447725698359547e-05,
|
| 448 |
+
"loss": 0.269,
|
| 449 |
+
"step": 5500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.34748076445768183,
|
| 453 |
+
"grad_norm": 14.331618309020996,
|
| 454 |
+
"learning_rate": 7.361739587252019e-05,
|
| 455 |
+
"loss": 0.2466,
|
| 456 |
+
"step": 5600
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.3536857781087118,
|
| 460 |
+
"grad_norm": 13.95034122467041,
|
| 461 |
+
"learning_rate": 7.274844798780826e-05,
|
| 462 |
+
"loss": 0.2491,
|
| 463 |
+
"step": 5700
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.35989079175974187,
|
| 467 |
+
"grad_norm": 13.919034004211426,
|
| 468 |
+
"learning_rate": 7.18707476564436e-05,
|
| 469 |
+
"loss": 0.2189,
|
| 470 |
+
"step": 5800
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.3660958054107719,
|
| 474 |
+
"grad_norm": 15.497775077819824,
|
| 475 |
+
"learning_rate": 7.098463257290679e-05,
|
| 476 |
+
"loss": 0.1965,
|
| 477 |
+
"step": 5900
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.37230081906180196,
|
| 481 |
+
"grad_norm": 17.710037231445312,
|
| 482 |
+
"learning_rate": 7.009044366924739e-05,
|
| 483 |
+
"loss": 0.2509,
|
| 484 |
+
"step": 6000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.37230081906180196,
|
| 488 |
+
"eval_avg_non_pair_similarity": 2.055158477392279e-05,
|
| 489 |
+
"eval_avg_pair_similarity": 0.006798229112289846,
|
| 490 |
+
"eval_loss": 0.24542327225208282,
|
| 491 |
+
"eval_runtime": 52.6713,
|
| 492 |
+
"eval_samples_per_second": 9.493,
|
| 493 |
+
"eval_similarity_ratio": 330.78855898820456,
|
| 494 |
+
"eval_steps_per_second": 0.304,
|
| 495 |
+
"step": 6000
|
| 496 |
+
}
|
| 497 |
+
],
|
| 498 |
+
"logging_steps": 100,
|
| 499 |
+
"max_steps": 16116,
|
| 500 |
+
"num_input_tokens_seen": 0,
|
| 501 |
+
"num_train_epochs": 1,
|
| 502 |
+
"save_steps": 1000,
|
| 503 |
+
"stateful_callbacks": {
|
| 504 |
+
"TrainerControl": {
|
| 505 |
+
"args": {
|
| 506 |
+
"should_epoch_stop": false,
|
| 507 |
+
"should_evaluate": false,
|
| 508 |
+
"should_log": false,
|
| 509 |
+
"should_save": true,
|
| 510 |
+
"should_training_stop": false
|
| 511 |
+
},
|
| 512 |
+
"attributes": {}
|
| 513 |
+
}
|
| 514 |
+
},
|
| 515 |
+
"total_flos": 0.0,
|
| 516 |
+
"train_batch_size": 32,
|
| 517 |
+
"trial_name": null,
|
| 518 |
+
"trial_params": null
|
| 519 |
+
}
|
checkpoint-6000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45f6f301aa93fd04cb88245f1be6b5d49fae9be8dd0abe1d8a346a6aadd21663
|
| 3 |
+
size 5112
|