Training in progress, step 7000, checkpoint
Browse files
checkpoint-7000/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"annotation_transformer": false,
|
| 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": "allenai/scibert_scivocab_uncased",
|
| 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-7000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e164680c13f187d4882f0012ff55b2a8f082c18687084501e972efc46ccfee6
|
| 3 |
+
size 1570427506
|
checkpoint-7000/trainer_state.json
ADDED
|
@@ -0,0 +1,600 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.15678896205707119,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 7000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.002239842315101017,
|
| 13 |
+
"grad_norm": 71.10852813720703,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 3.5057,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.004479684630202034,
|
| 20 |
+
"grad_norm": 57.756534576416016,
|
| 21 |
+
"learning_rate": 9.9998756572327e-05,
|
| 22 |
+
"loss": 0.8254,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.006719526945303051,
|
| 27 |
+
"grad_norm": 44.61080551147461,
|
| 28 |
+
"learning_rate": 9.999502635115246e-05,
|
| 29 |
+
"loss": 0.6935,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.008959369260404068,
|
| 34 |
+
"grad_norm": 50.82619094848633,
|
| 35 |
+
"learning_rate": 9.998880952200681e-05,
|
| 36 |
+
"loss": 0.5972,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.011199211575505085,
|
| 41 |
+
"grad_norm": 44.26677703857422,
|
| 42 |
+
"learning_rate": 9.998010639409713e-05,
|
| 43 |
+
"loss": 0.5408,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.013439053890606102,
|
| 48 |
+
"grad_norm": 42.8045654296875,
|
| 49 |
+
"learning_rate": 9.996891740029186e-05,
|
| 50 |
+
"loss": 0.5509,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.01567889620570712,
|
| 55 |
+
"grad_norm": 30.989139556884766,
|
| 56 |
+
"learning_rate": 9.995524309709913e-05,
|
| 57 |
+
"loss": 0.4823,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.017918738520808136,
|
| 62 |
+
"grad_norm": 34.01952362060547,
|
| 63 |
+
"learning_rate": 9.993908416463927e-05,
|
| 64 |
+
"loss": 0.5111,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.020158580835909153,
|
| 69 |
+
"grad_norm": 34.079307556152344,
|
| 70 |
+
"learning_rate": 9.992044140661079e-05,
|
| 71 |
+
"loss": 0.4635,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.02239842315101017,
|
| 76 |
+
"grad_norm": 26.16071128845215,
|
| 77 |
+
"learning_rate": 9.989931575025056e-05,
|
| 78 |
+
"loss": 0.4883,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.02239842315101017,
|
| 83 |
+
"eval_avg_non_pair_similarity": 0.0020343252948339737,
|
| 84 |
+
"eval_avg_pair_similarity": 0.008852629057131708,
|
| 85 |
+
"eval_loss": 0.5431402921676636,
|
| 86 |
+
"eval_runtime": 19.6178,
|
| 87 |
+
"eval_samples_per_second": 25.487,
|
| 88 |
+
"eval_similarity_ratio": 4.35162905343228,
|
| 89 |
+
"eval_steps_per_second": 0.816,
|
| 90 |
+
"step": 1000
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.024638265466111187,
|
| 94 |
+
"grad_norm": 35.38695526123047,
|
| 95 |
+
"learning_rate": 9.987570824628759e-05,
|
| 96 |
+
"loss": 0.4655,
|
| 97 |
+
"step": 1100
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.026878107781212204,
|
| 101 |
+
"grad_norm": 32.071346282958984,
|
| 102 |
+
"learning_rate": 9.984962006889084e-05,
|
| 103 |
+
"loss": 0.4342,
|
| 104 |
+
"step": 1200
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.029117950096313218,
|
| 108 |
+
"grad_norm": 39.6610221862793,
|
| 109 |
+
"learning_rate": 9.982105251561082e-05,
|
| 110 |
+
"loss": 0.458,
|
| 111 |
+
"step": 1300
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.03135779241141424,
|
| 115 |
+
"grad_norm": 31.493322372436523,
|
| 116 |
+
"learning_rate": 9.979000700731491e-05,
|
| 117 |
+
"loss": 0.4525,
|
| 118 |
+
"step": 1400
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.03359763472651525,
|
| 122 |
+
"grad_norm": 34.453399658203125,
|
| 123 |
+
"learning_rate": 9.975648508811693e-05,
|
| 124 |
+
"loss": 0.41,
|
| 125 |
+
"step": 1500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.03583747704161627,
|
| 129 |
+
"grad_norm": 33.990074157714844,
|
| 130 |
+
"learning_rate": 9.972048842530012e-05,
|
| 131 |
+
"loss": 0.4097,
|
| 132 |
+
"step": 1600
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.038077319356717286,
|
| 136 |
+
"grad_norm": 30.44228172302246,
|
| 137 |
+
"learning_rate": 9.968201880923439e-05,
|
| 138 |
+
"loss": 0.4257,
|
| 139 |
+
"step": 1700
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.040317161671818307,
|
| 143 |
+
"grad_norm": 31.427162170410156,
|
| 144 |
+
"learning_rate": 9.964107815328711e-05,
|
| 145 |
+
"loss": 0.3821,
|
| 146 |
+
"step": 1800
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.04255700398691932,
|
| 150 |
+
"grad_norm": 25.877887725830078,
|
| 151 |
+
"learning_rate": 9.959766849372808e-05,
|
| 152 |
+
"loss": 0.3788,
|
| 153 |
+
"step": 1900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.04479684630202034,
|
| 157 |
+
"grad_norm": 25.36798095703125,
|
| 158 |
+
"learning_rate": 9.955179198962817e-05,
|
| 159 |
+
"loss": 0.3854,
|
| 160 |
+
"step": 2000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.04479684630202034,
|
| 164 |
+
"eval_avg_non_pair_similarity": 0.0016289287904792565,
|
| 165 |
+
"eval_avg_pair_similarity": 0.0032495629731565715,
|
| 166 |
+
"eval_loss": 0.4394480586051941,
|
| 167 |
+
"eval_runtime": 19.5064,
|
| 168 |
+
"eval_samples_per_second": 25.633,
|
| 169 |
+
"eval_similarity_ratio": 1.9949079371360972,
|
| 170 |
+
"eval_steps_per_second": 0.82,
|
| 171 |
+
"step": 2000
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.047036688617121354,
|
| 175 |
+
"grad_norm": 24.537384033203125,
|
| 176 |
+
"learning_rate": 9.950345092275198e-05,
|
| 177 |
+
"loss": 0.3886,
|
| 178 |
+
"step": 2100
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.049276530932222375,
|
| 182 |
+
"grad_norm": 26.4981632232666,
|
| 183 |
+
"learning_rate": 9.945264769744431e-05,
|
| 184 |
+
"loss": 0.3926,
|
| 185 |
+
"step": 2200
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.05151637324732339,
|
| 189 |
+
"grad_norm": 40.585941314697266,
|
| 190 |
+
"learning_rate": 9.939938484051063e-05,
|
| 191 |
+
"loss": 0.3796,
|
| 192 |
+
"step": 2300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.05375621556242441,
|
| 196 |
+
"grad_norm": 26.374149322509766,
|
| 197 |
+
"learning_rate": 9.934366500109132e-05,
|
| 198 |
+
"loss": 0.3723,
|
| 199 |
+
"step": 2400
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.05599605787752542,
|
| 203 |
+
"grad_norm": 31.54728126525879,
|
| 204 |
+
"learning_rate": 9.928549095053001e-05,
|
| 205 |
+
"loss": 0.3639,
|
| 206 |
+
"step": 2500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.058235900192626436,
|
| 210 |
+
"grad_norm": 26.00455093383789,
|
| 211 |
+
"learning_rate": 9.922486558223567e-05,
|
| 212 |
+
"loss": 0.3387,
|
| 213 |
+
"step": 2600
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.060475742507727456,
|
| 217 |
+
"grad_norm": 23.684057235717773,
|
| 218 |
+
"learning_rate": 9.916179191153873e-05,
|
| 219 |
+
"loss": 0.3494,
|
| 220 |
+
"step": 2700
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.06271558482282848,
|
| 224 |
+
"grad_norm": 24.034656524658203,
|
| 225 |
+
"learning_rate": 9.909627307554108e-05,
|
| 226 |
+
"loss": 0.362,
|
| 227 |
+
"step": 2800
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.06495542713792948,
|
| 231 |
+
"grad_norm": 31.07891082763672,
|
| 232 |
+
"learning_rate": 9.902831233296009e-05,
|
| 233 |
+
"loss": 0.3481,
|
| 234 |
+
"step": 2900
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.0671952694530305,
|
| 238 |
+
"grad_norm": 24.183626174926758,
|
| 239 |
+
"learning_rate": 9.895791306396644e-05,
|
| 240 |
+
"loss": 0.3535,
|
| 241 |
+
"step": 3000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.0671952694530305,
|
| 245 |
+
"eval_avg_non_pair_similarity": -0.0007878901720614375,
|
| 246 |
+
"eval_avg_pair_similarity": -0.008789425778668375,
|
| 247 |
+
"eval_loss": 0.4187028110027313,
|
| 248 |
+
"eval_runtime": 19.4839,
|
| 249 |
+
"eval_samples_per_second": 25.662,
|
| 250 |
+
"eval_similarity_ratio": 11.155648452463497,
|
| 251 |
+
"eval_steps_per_second": 0.821,
|
| 252 |
+
"step": 3000
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.06943511176813152,
|
| 256 |
+
"grad_norm": 20.178213119506836,
|
| 257 |
+
"learning_rate": 9.888507877001616e-05,
|
| 258 |
+
"loss": 0.3375,
|
| 259 |
+
"step": 3100
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.07167495408323254,
|
| 263 |
+
"grad_norm": 31.757244110107422,
|
| 264 |
+
"learning_rate": 9.880981307367627e-05,
|
| 265 |
+
"loss": 0.3489,
|
| 266 |
+
"step": 3200
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.07391479639833355,
|
| 270 |
+
"grad_norm": 27.889982223510742,
|
| 271 |
+
"learning_rate": 9.873211971844477e-05,
|
| 272 |
+
"loss": 0.3463,
|
| 273 |
+
"step": 3300
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.07615463871343457,
|
| 277 |
+
"grad_norm": 23.282093048095703,
|
| 278 |
+
"learning_rate": 9.865200256856437e-05,
|
| 279 |
+
"loss": 0.3235,
|
| 280 |
+
"step": 3400
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.07839448102853559,
|
| 284 |
+
"grad_norm": 24.573863983154297,
|
| 285 |
+
"learning_rate": 9.856946560883034e-05,
|
| 286 |
+
"loss": 0.3474,
|
| 287 |
+
"step": 3500
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.08063432334363661,
|
| 291 |
+
"grad_norm": 20.7421932220459,
|
| 292 |
+
"learning_rate": 9.848451294439224e-05,
|
| 293 |
+
"loss": 0.3533,
|
| 294 |
+
"step": 3600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.08287416565873762,
|
| 298 |
+
"grad_norm": 22.051513671875,
|
| 299 |
+
"learning_rate": 9.839714880054987e-05,
|
| 300 |
+
"loss": 0.3277,
|
| 301 |
+
"step": 3700
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 0.08511400797383864,
|
| 305 |
+
"grad_norm": 22.470027923583984,
|
| 306 |
+
"learning_rate": 9.830737752254294e-05,
|
| 307 |
+
"loss": 0.3261,
|
| 308 |
+
"step": 3800
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.08735385028893966,
|
| 312 |
+
"grad_norm": 20.80890464782715,
|
| 313 |
+
"learning_rate": 9.821520357533513e-05,
|
| 314 |
+
"loss": 0.3474,
|
| 315 |
+
"step": 3900
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.08959369260404068,
|
| 319 |
+
"grad_norm": 20.187280654907227,
|
| 320 |
+
"learning_rate": 9.812063154339191e-05,
|
| 321 |
+
"loss": 0.2984,
|
| 322 |
+
"step": 4000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.08959369260404068,
|
| 326 |
+
"eval_avg_non_pair_similarity": -0.0008396646829899625,
|
| 327 |
+
"eval_avg_pair_similarity": -0.011315496074035763,
|
| 328 |
+
"eval_loss": 0.32538020610809326,
|
| 329 |
+
"eval_runtime": 19.4916,
|
| 330 |
+
"eval_samples_per_second": 25.652,
|
| 331 |
+
"eval_similarity_ratio": 13.476208185560939,
|
| 332 |
+
"eval_steps_per_second": 0.821,
|
| 333 |
+
"step": 4000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.09183353491914169,
|
| 337 |
+
"grad_norm": 26.496036529541016,
|
| 338 |
+
"learning_rate": 9.802366613045254e-05,
|
| 339 |
+
"loss": 0.3326,
|
| 340 |
+
"step": 4100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.09407337723424271,
|
| 344 |
+
"grad_norm": 23.102359771728516,
|
| 345 |
+
"learning_rate": 9.792431215929613e-05,
|
| 346 |
+
"loss": 0.3341,
|
| 347 |
+
"step": 4200
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.09631321954934373,
|
| 351 |
+
"grad_norm": 21.27369499206543,
|
| 352 |
+
"learning_rate": 9.782257457150177e-05,
|
| 353 |
+
"loss": 0.3254,
|
| 354 |
+
"step": 4300
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.09855306186444475,
|
| 358 |
+
"grad_norm": 19.25406265258789,
|
| 359 |
+
"learning_rate": 9.771845842720274e-05,
|
| 360 |
+
"loss": 0.3334,
|
| 361 |
+
"step": 4400
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.10079290417954576,
|
| 365 |
+
"grad_norm": 15.666335105895996,
|
| 366 |
+
"learning_rate": 9.761196890483482e-05,
|
| 367 |
+
"loss": 0.3064,
|
| 368 |
+
"step": 4500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.10303274649464678,
|
| 372 |
+
"grad_norm": 24.592592239379883,
|
| 373 |
+
"learning_rate": 9.75031113008788e-05,
|
| 374 |
+
"loss": 0.2902,
|
| 375 |
+
"step": 4600
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.1052725888097478,
|
| 379 |
+
"grad_norm": 20.0572566986084,
|
| 380 |
+
"learning_rate": 9.739189102959696e-05,
|
| 381 |
+
"loss": 0.3121,
|
| 382 |
+
"step": 4700
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.10751243112484882,
|
| 386 |
+
"grad_norm": 19.887725830078125,
|
| 387 |
+
"learning_rate": 9.727831362276381e-05,
|
| 388 |
+
"loss": 0.3014,
|
| 389 |
+
"step": 4800
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.10975227343994982,
|
| 393 |
+
"grad_norm": 15.561097145080566,
|
| 394 |
+
"learning_rate": 9.716238472939101e-05,
|
| 395 |
+
"loss": 0.316,
|
| 396 |
+
"step": 4900
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.11199211575505084,
|
| 400 |
+
"grad_norm": 17.96786880493164,
|
| 401 |
+
"learning_rate": 9.704411011544629e-05,
|
| 402 |
+
"loss": 0.3215,
|
| 403 |
+
"step": 5000
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.11199211575505084,
|
| 407 |
+
"eval_avg_non_pair_similarity": 0.00024146916974524413,
|
| 408 |
+
"eval_avg_pair_similarity": -0.009633154251612723,
|
| 409 |
+
"eval_loss": 0.2951599061489105,
|
| 410 |
+
"eval_runtime": 19.5025,
|
| 411 |
+
"eval_samples_per_second": 25.638,
|
| 412 |
+
"eval_similarity_ratio": -39.893930400207765,
|
| 413 |
+
"eval_steps_per_second": 0.82,
|
| 414 |
+
"step": 5000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.11423195807015186,
|
| 418 |
+
"grad_norm": 14.672723770141602,
|
| 419 |
+
"learning_rate": 9.692349566356677e-05,
|
| 420 |
+
"loss": 0.2954,
|
| 421 |
+
"step": 5100
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 0.11647180038525287,
|
| 425 |
+
"grad_norm": 17.881025314331055,
|
| 426 |
+
"learning_rate": 9.680054737276638e-05,
|
| 427 |
+
"loss": 0.2968,
|
| 428 |
+
"step": 5200
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.11871164270035389,
|
| 432 |
+
"grad_norm": 16.55130958557129,
|
| 433 |
+
"learning_rate": 9.667527135813737e-05,
|
| 434 |
+
"loss": 0.2707,
|
| 435 |
+
"step": 5300
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.12095148501545491,
|
| 439 |
+
"grad_norm": 21.650575637817383,
|
| 440 |
+
"learning_rate": 9.654767385054627e-05,
|
| 441 |
+
"loss": 0.3068,
|
| 442 |
+
"step": 5400
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.12319132733055593,
|
| 446 |
+
"grad_norm": 20.990921020507812,
|
| 447 |
+
"learning_rate": 9.641776119632397e-05,
|
| 448 |
+
"loss": 0.3037,
|
| 449 |
+
"step": 5500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.12543116964565695,
|
| 453 |
+
"grad_norm": 21.97833824157715,
|
| 454 |
+
"learning_rate": 9.628553985695005e-05,
|
| 455 |
+
"loss": 0.3307,
|
| 456 |
+
"step": 5600
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.12767101196075797,
|
| 460 |
+
"grad_norm": 20.353246688842773,
|
| 461 |
+
"learning_rate": 9.61510164087314e-05,
|
| 462 |
+
"loss": 0.2907,
|
| 463 |
+
"step": 5700
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.12991085427585897,
|
| 467 |
+
"grad_norm": 16.254976272583008,
|
| 468 |
+
"learning_rate": 9.601419754247514e-05,
|
| 469 |
+
"loss": 0.3025,
|
| 470 |
+
"step": 5800
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.13215069659096,
|
| 474 |
+
"grad_norm": 24.14662742614746,
|
| 475 |
+
"learning_rate": 9.587509006315585e-05,
|
| 476 |
+
"loss": 0.2676,
|
| 477 |
+
"step": 5900
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.134390538906061,
|
| 481 |
+
"grad_norm": 15.861074447631836,
|
| 482 |
+
"learning_rate": 9.573370088957712e-05,
|
| 483 |
+
"loss": 0.2953,
|
| 484 |
+
"step": 6000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.134390538906061,
|
| 488 |
+
"eval_avg_non_pair_similarity": -0.00021786548842126337,
|
| 489 |
+
"eval_avg_pair_similarity": -0.001965825233142823,
|
| 490 |
+
"eval_loss": 0.31484127044677734,
|
| 491 |
+
"eval_runtime": 19.5134,
|
| 492 |
+
"eval_samples_per_second": 25.623,
|
| 493 |
+
"eval_similarity_ratio": 9.023114433533939,
|
| 494 |
+
"eval_steps_per_second": 0.82,
|
| 495 |
+
"step": 6000
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.13663038122116203,
|
| 499 |
+
"grad_norm": 14.910375595092773,
|
| 500 |
+
"learning_rate": 9.559003705402737e-05,
|
| 501 |
+
"loss": 0.2846,
|
| 502 |
+
"step": 6100
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 0.13887022353626305,
|
| 506 |
+
"grad_norm": 17.768138885498047,
|
| 507 |
+
"learning_rate": 9.544410570193014e-05,
|
| 508 |
+
"loss": 0.3031,
|
| 509 |
+
"step": 6200
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.14111006585136407,
|
| 513 |
+
"grad_norm": 13.620220184326172,
|
| 514 |
+
"learning_rate": 9.529591409148874e-05,
|
| 515 |
+
"loss": 0.2716,
|
| 516 |
+
"step": 6300
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.1433499081664651,
|
| 520 |
+
"grad_norm": 14.697577476501465,
|
| 521 |
+
"learning_rate": 9.514546959332509e-05,
|
| 522 |
+
"loss": 0.2759,
|
| 523 |
+
"step": 6400
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.1455897504815661,
|
| 527 |
+
"grad_norm": 21.234310150146484,
|
| 528 |
+
"learning_rate": 9.499277969011334e-05,
|
| 529 |
+
"loss": 0.2845,
|
| 530 |
+
"step": 6500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.1478295927966671,
|
| 534 |
+
"grad_norm": 16.50038719177246,
|
| 535 |
+
"learning_rate": 9.483785197620747e-05,
|
| 536 |
+
"loss": 0.2471,
|
| 537 |
+
"step": 6600
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.15006943511176812,
|
| 541 |
+
"grad_norm": 16.261537551879883,
|
| 542 |
+
"learning_rate": 9.468069415726377e-05,
|
| 543 |
+
"loss": 0.2978,
|
| 544 |
+
"step": 6700
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.15230927742686914,
|
| 548 |
+
"grad_norm": 16.553571701049805,
|
| 549 |
+
"learning_rate": 9.452131404985752e-05,
|
| 550 |
+
"loss": 0.311,
|
| 551 |
+
"step": 6800
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.15454911974197016,
|
| 555 |
+
"grad_norm": 18.445993423461914,
|
| 556 |
+
"learning_rate": 9.43597195810941e-05,
|
| 557 |
+
"loss": 0.2777,
|
| 558 |
+
"step": 6900
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.15678896205707119,
|
| 562 |
+
"grad_norm": 13.675585746765137,
|
| 563 |
+
"learning_rate": 9.419591878821496e-05,
|
| 564 |
+
"loss": 0.2417,
|
| 565 |
+
"step": 7000
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.15678896205707119,
|
| 569 |
+
"eval_avg_non_pair_similarity": 0.0003613333898595999,
|
| 570 |
+
"eval_avg_pair_similarity": 0.005676110625499859,
|
| 571 |
+
"eval_loss": 0.3019685447216034,
|
| 572 |
+
"eval_runtime": 19.5219,
|
| 573 |
+
"eval_samples_per_second": 25.612,
|
| 574 |
+
"eval_similarity_ratio": 15.708790786551377,
|
| 575 |
+
"eval_steps_per_second": 0.82,
|
| 576 |
+
"step": 7000
|
| 577 |
+
}
|
| 578 |
+
],
|
| 579 |
+
"logging_steps": 100,
|
| 580 |
+
"max_steps": 44646,
|
| 581 |
+
"num_input_tokens_seen": 0,
|
| 582 |
+
"num_train_epochs": 1,
|
| 583 |
+
"save_steps": 1000,
|
| 584 |
+
"stateful_callbacks": {
|
| 585 |
+
"TrainerControl": {
|
| 586 |
+
"args": {
|
| 587 |
+
"should_epoch_stop": false,
|
| 588 |
+
"should_evaluate": false,
|
| 589 |
+
"should_log": false,
|
| 590 |
+
"should_save": true,
|
| 591 |
+
"should_training_stop": false
|
| 592 |
+
},
|
| 593 |
+
"attributes": {}
|
| 594 |
+
}
|
| 595 |
+
},
|
| 596 |
+
"total_flos": 0.0,
|
| 597 |
+
"train_batch_size": 32,
|
| 598 |
+
"trial_name": null,
|
| 599 |
+
"trial_params": null
|
| 600 |
+
}
|
checkpoint-7000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c6bbe0930b5977d5ef83c9c6a2872aea787e9f64c3d133a00319e43c833ad34
|
| 3 |
+
size 5112
|