E2 run e2_rlce_wrl2_sigma1_fixed_seed42 (epoch_4.pt)
Browse files- e2_rlce_wrl2_sigma1_fixed_seed42/config.json +77 -0
- e2_rlce_wrl2_sigma1_fixed_seed42/grad_diagnostics.json +308 -0
- e2_rlce_wrl2_sigma1_fixed_seed42/model.safetensors +3 -0
- e2_rlce_wrl2_sigma1_fixed_seed42/model_config.json +18 -0
- e2_rlce_wrl2_sigma1_fixed_seed42/test_eval.json +99 -0
- e2_rlce_wrl2_sigma1_fixed_seed42/train_log.json +70 -0
e2_rlce_wrl2_sigma1_fixed_seed42/config.json
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model": {
|
| 3 |
+
"encoder_name": "answerdotai/ModernBERT-large",
|
| 4 |
+
"head_layers": 2,
|
| 5 |
+
"head_dropout": 0.1,
|
| 6 |
+
"num_qtypes": 3,
|
| 7 |
+
"cls_query": false,
|
| 8 |
+
"slot_emb": false,
|
| 9 |
+
"max_len": 512,
|
| 10 |
+
"head_max_len": 192,
|
| 11 |
+
"k_buckets": [
|
| 12 |
+
2,
|
| 13 |
+
5,
|
| 14 |
+
10
|
| 15 |
+
]
|
| 16 |
+
},
|
| 17 |
+
"training": {
|
| 18 |
+
"dataset_name": "LocalLLaMA/typed-decisions",
|
| 19 |
+
"dataset_config": "all",
|
| 20 |
+
"calib_fraction": 0.1,
|
| 21 |
+
"seed": 42,
|
| 22 |
+
"max_examples": null,
|
| 23 |
+
"ckpt_dir": "checkpoints",
|
| 24 |
+
"result_dir": "results",
|
| 25 |
+
"run_name": "e2_rlce_wrl2_sigma1_fixed_seed42",
|
| 26 |
+
"epochs": 4,
|
| 27 |
+
"batch_size": 8,
|
| 28 |
+
"accum_steps": 8,
|
| 29 |
+
"num_workers": 4,
|
| 30 |
+
"pin_memory": true,
|
| 31 |
+
"amp": true,
|
| 32 |
+
"amp_dtype": "bf16",
|
| 33 |
+
"gradient_checkpointing": false,
|
| 34 |
+
"lr": 2.5e-05,
|
| 35 |
+
"lr_head": 0.0001,
|
| 36 |
+
"weight_decay": 0.01,
|
| 37 |
+
"grad_clip": 1.0,
|
| 38 |
+
"log_every": 10,
|
| 39 |
+
"eval_every": 1,
|
| 40 |
+
"ckpt_every": 4,
|
| 41 |
+
"w_rl": 2.0,
|
| 42 |
+
"w_ce": 1.0,
|
| 43 |
+
"num_noise_samples": 4,
|
| 44 |
+
"sigma_start": 1.0,
|
| 45 |
+
"sigma_end": 1.0,
|
| 46 |
+
"anneal_sigma": false,
|
| 47 |
+
"reward_w_spherical": 0.75,
|
| 48 |
+
"reward_w_rps": 1.0,
|
| 49 |
+
"arch": {
|
| 50 |
+
"encoder_name": "answerdotai/ModernBERT-large",
|
| 51 |
+
"head_layers": 2,
|
| 52 |
+
"max_len": 512,
|
| 53 |
+
"head_max_len": 192
|
| 54 |
+
},
|
| 55 |
+
"init_from": "convaiinnovations/laya",
|
| 56 |
+
"log_grad_diagnostics": true,
|
| 57 |
+
"lr_scheduler": {
|
| 58 |
+
"type": "cosine",
|
| 59 |
+
"t_max": "auto",
|
| 60 |
+
"eta_min": 1e-06
|
| 61 |
+
},
|
| 62 |
+
"early_stopping": {
|
| 63 |
+
"enabled": false
|
| 64 |
+
},
|
| 65 |
+
"save_best": false,
|
| 66 |
+
"best_metric": "raw_ece",
|
| 67 |
+
"best_mode": "min",
|
| 68 |
+
"wandb": {
|
| 69 |
+
"enabled": true,
|
| 70 |
+
"project": "rlcd-reverse-engineering",
|
| 71 |
+
"entity": null,
|
| 72 |
+
"group": "e2-ratio",
|
| 73 |
+
"log_artifacts": false
|
| 74 |
+
},
|
| 75 |
+
"resume_from": null
|
| 76 |
+
}
|
| 77 |
+
}
|
e2_rlce_wrl2_sigma1_fixed_seed42/grad_diagnostics.json
ADDED
|
@@ -0,0 +1,308 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"step": 10,
|
| 4 |
+
"epoch": 1,
|
| 5 |
+
"sigma": 1.0,
|
| 6 |
+
"grad_cos_rl_ce": 0.6403411626815796,
|
| 7 |
+
"grad_norm_rl": 0.5020345449447632,
|
| 8 |
+
"grad_norm_ce": 0.17158567905426025,
|
| 9 |
+
"grad_norm_ratio_rl_ce": 2.9258534145265447
|
| 10 |
+
},
|
| 11 |
+
{
|
| 12 |
+
"step": 20,
|
| 13 |
+
"epoch": 1,
|
| 14 |
+
"sigma": 1.0,
|
| 15 |
+
"grad_cos_rl_ce": 0.7001619338989258,
|
| 16 |
+
"grad_norm_rl": 0.3677082061767578,
|
| 17 |
+
"grad_norm_ce": 0.14867256581783295,
|
| 18 |
+
"grad_norm_ratio_rl_ce": 2.4732754436168616
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"step": 30,
|
| 22 |
+
"epoch": 1,
|
| 23 |
+
"sigma": 1.0,
|
| 24 |
+
"grad_cos_rl_ce": 0.6850250959396362,
|
| 25 |
+
"grad_norm_rl": 0.33745238184928894,
|
| 26 |
+
"grad_norm_ce": 0.1552022099494934,
|
| 27 |
+
"grad_norm_ratio_rl_ce": 2.1742756237756162
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"step": 40,
|
| 31 |
+
"epoch": 1,
|
| 32 |
+
"sigma": 1.0,
|
| 33 |
+
"grad_cos_rl_ce": 0.6903390884399414,
|
| 34 |
+
"grad_norm_rl": 0.41107022762298584,
|
| 35 |
+
"grad_norm_ce": 0.10673101246356964,
|
| 36 |
+
"grad_norm_ratio_rl_ce": 3.8514600221120916
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"step": 50,
|
| 40 |
+
"epoch": 1,
|
| 41 |
+
"sigma": 1.0,
|
| 42 |
+
"grad_cos_rl_ce": 0.7652473449707031,
|
| 43 |
+
"grad_norm_rl": 0.3909636437892914,
|
| 44 |
+
"grad_norm_ce": 0.09709218889474869,
|
| 45 |
+
"grad_norm_ratio_rl_ce": 4.026726024408715
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"step": 60,
|
| 49 |
+
"epoch": 1,
|
| 50 |
+
"sigma": 1.0,
|
| 51 |
+
"grad_cos_rl_ce": 0.605665922164917,
|
| 52 |
+
"grad_norm_rl": 0.3942667841911316,
|
| 53 |
+
"grad_norm_ce": 0.097779780626297,
|
| 54 |
+
"grad_norm_ratio_rl_ce": 4.032191335118388
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"step": 70,
|
| 58 |
+
"epoch": 1,
|
| 59 |
+
"sigma": 1.0,
|
| 60 |
+
"grad_cos_rl_ce": 0.312794029712677,
|
| 61 |
+
"grad_norm_rl": 0.4949149787425995,
|
| 62 |
+
"grad_norm_ce": 0.10643666237592697,
|
| 63 |
+
"grad_norm_ratio_rl_ce": 4.64985436122183
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"step": 80,
|
| 67 |
+
"epoch": 1,
|
| 68 |
+
"sigma": 1.0,
|
| 69 |
+
"grad_cos_rl_ce": 0.43239080905914307,
|
| 70 |
+
"grad_norm_rl": 0.4541529715061188,
|
| 71 |
+
"grad_norm_ce": 0.11235574632883072,
|
| 72 |
+
"grad_norm_ratio_rl_ce": 4.042098302448659
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"step": 90,
|
| 76 |
+
"epoch": 2,
|
| 77 |
+
"sigma": 1.0,
|
| 78 |
+
"grad_cos_rl_ce": 0.67087721824646,
|
| 79 |
+
"grad_norm_rl": 0.32067450881004333,
|
| 80 |
+
"grad_norm_ce": 0.08662408590316772,
|
| 81 |
+
"grad_norm_ratio_rl_ce": 3.7019092953951356
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"step": 100,
|
| 85 |
+
"epoch": 2,
|
| 86 |
+
"sigma": 1.0,
|
| 87 |
+
"grad_cos_rl_ce": 0.458447128534317,
|
| 88 |
+
"grad_norm_rl": 0.31714800000190735,
|
| 89 |
+
"grad_norm_ce": 0.07650914043188095,
|
| 90 |
+
"grad_norm_ratio_rl_ce": 4.145230206634938
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"step": 110,
|
| 94 |
+
"epoch": 2,
|
| 95 |
+
"sigma": 1.0,
|
| 96 |
+
"grad_cos_rl_ce": 0.5613347291946411,
|
| 97 |
+
"grad_norm_rl": 0.3095552921295166,
|
| 98 |
+
"grad_norm_ce": 0.0698700025677681,
|
| 99 |
+
"grad_norm_ratio_rl_ce": 4.43044626811447
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"step": 120,
|
| 103 |
+
"epoch": 2,
|
| 104 |
+
"sigma": 1.0,
|
| 105 |
+
"grad_cos_rl_ce": 0.347777783870697,
|
| 106 |
+
"grad_norm_rl": 0.37637409567832947,
|
| 107 |
+
"grad_norm_ce": 0.11371447145938873,
|
| 108 |
+
"grad_norm_ratio_rl_ce": 3.3098170430554683
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"step": 130,
|
| 112 |
+
"epoch": 2,
|
| 113 |
+
"sigma": 1.0,
|
| 114 |
+
"grad_cos_rl_ce": 0.6471930742263794,
|
| 115 |
+
"grad_norm_rl": 0.47030115127563477,
|
| 116 |
+
"grad_norm_ce": 0.07688033580780029,
|
| 117 |
+
"grad_norm_ratio_rl_ce": 6.1173139572566475
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"step": 140,
|
| 121 |
+
"epoch": 2,
|
| 122 |
+
"sigma": 1.0,
|
| 123 |
+
"grad_cos_rl_ce": 0.3619423508644104,
|
| 124 |
+
"grad_norm_rl": 0.3926744759082794,
|
| 125 |
+
"grad_norm_ce": 0.03303954005241394,
|
| 126 |
+
"grad_norm_ratio_rl_ce": 11.88498614948454
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"step": 150,
|
| 130 |
+
"epoch": 2,
|
| 131 |
+
"sigma": 1.0,
|
| 132 |
+
"grad_cos_rl_ce": 0.73228919506073,
|
| 133 |
+
"grad_norm_rl": 0.440285861492157,
|
| 134 |
+
"grad_norm_ce": 0.10116377472877502,
|
| 135 |
+
"grad_norm_ratio_rl_ce": 4.352208709813218
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"step": 160,
|
| 139 |
+
"epoch": 2,
|
| 140 |
+
"sigma": 1.0,
|
| 141 |
+
"grad_cos_rl_ce": 0.6286622285842896,
|
| 142 |
+
"grad_norm_rl": 0.35417866706848145,
|
| 143 |
+
"grad_norm_ce": 0.10011424124240875,
|
| 144 |
+
"grad_norm_ratio_rl_ce": 3.537745106721641
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"step": 170,
|
| 148 |
+
"epoch": 2,
|
| 149 |
+
"sigma": 1.0,
|
| 150 |
+
"grad_cos_rl_ce": 0.745894193649292,
|
| 151 |
+
"grad_norm_rl": 0.30267196893692017,
|
| 152 |
+
"grad_norm_ce": 0.1110180914402008,
|
| 153 |
+
"grad_norm_ratio_rl_ce": 2.7263301414252155
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"step": 180,
|
| 157 |
+
"epoch": 3,
|
| 158 |
+
"sigma": 1.0,
|
| 159 |
+
"grad_cos_rl_ce": 0.5900934934616089,
|
| 160 |
+
"grad_norm_rl": 0.49082449078559875,
|
| 161 |
+
"grad_norm_ce": 0.0876513347029686,
|
| 162 |
+
"grad_norm_ratio_rl_ce": 5.599737784357725
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"step": 190,
|
| 166 |
+
"epoch": 3,
|
| 167 |
+
"sigma": 1.0,
|
| 168 |
+
"grad_cos_rl_ce": 0.5787338614463806,
|
| 169 |
+
"grad_norm_rl": 0.4556364119052887,
|
| 170 |
+
"grad_norm_ce": 0.07928700000047684,
|
| 171 |
+
"grad_norm_ratio_rl_ce": 5.746672366246024
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"step": 200,
|
| 175 |
+
"epoch": 3,
|
| 176 |
+
"sigma": 1.0,
|
| 177 |
+
"grad_cos_rl_ce": 0.7399789094924927,
|
| 178 |
+
"grad_norm_rl": 0.44587916135787964,
|
| 179 |
+
"grad_norm_ce": 0.09872928261756897,
|
| 180 |
+
"grad_norm_ratio_rl_ce": 4.516179491397773
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"step": 210,
|
| 184 |
+
"epoch": 3,
|
| 185 |
+
"sigma": 1.0,
|
| 186 |
+
"grad_cos_rl_ce": 0.5722417831420898,
|
| 187 |
+
"grad_norm_rl": 0.39616018533706665,
|
| 188 |
+
"grad_norm_ce": 0.1082942932844162,
|
| 189 |
+
"grad_norm_ratio_rl_ce": 3.658181546987158
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"step": 220,
|
| 193 |
+
"epoch": 3,
|
| 194 |
+
"sigma": 1.0,
|
| 195 |
+
"grad_cos_rl_ce": 0.4820224940776825,
|
| 196 |
+
"grad_norm_rl": 0.38037729263305664,
|
| 197 |
+
"grad_norm_ce": 0.051550909876823425,
|
| 198 |
+
"grad_norm_ratio_rl_ce": 7.378672724534567
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"step": 230,
|
| 202 |
+
"epoch": 3,
|
| 203 |
+
"sigma": 1.0,
|
| 204 |
+
"grad_cos_rl_ce": 0.26657313108444214,
|
| 205 |
+
"grad_norm_rl": 0.35336142778396606,
|
| 206 |
+
"grad_norm_ce": 0.04918649420142174,
|
| 207 |
+
"grad_norm_ratio_rl_ce": 7.184114938888085
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"step": 240,
|
| 211 |
+
"epoch": 3,
|
| 212 |
+
"sigma": 1.0,
|
| 213 |
+
"grad_cos_rl_ce": 0.42845308780670166,
|
| 214 |
+
"grad_norm_rl": 0.2999342083930969,
|
| 215 |
+
"grad_norm_ce": 0.07464830577373505,
|
| 216 |
+
"grad_norm_ratio_rl_ce": 4.017964041973322
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"step": 250,
|
| 220 |
+
"epoch": 3,
|
| 221 |
+
"sigma": 1.0,
|
| 222 |
+
"grad_cos_rl_ce": 0.4402986764907837,
|
| 223 |
+
"grad_norm_rl": 0.5924450755119324,
|
| 224 |
+
"grad_norm_ce": 0.11506222933530807,
|
| 225 |
+
"grad_norm_ratio_rl_ce": 5.148910106595112
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"step": 260,
|
| 229 |
+
"epoch": 4,
|
| 230 |
+
"sigma": 1.0,
|
| 231 |
+
"grad_cos_rl_ce": 0.6791796684265137,
|
| 232 |
+
"grad_norm_rl": 0.4480166435241699,
|
| 233 |
+
"grad_norm_ce": 0.08907032012939453,
|
| 234 |
+
"grad_norm_ratio_rl_ce": 5.02992066126321
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"step": 270,
|
| 238 |
+
"epoch": 4,
|
| 239 |
+
"sigma": 1.0,
|
| 240 |
+
"grad_cos_rl_ce": 0.03763109818100929,
|
| 241 |
+
"grad_norm_rl": 0.5576416254043579,
|
| 242 |
+
"grad_norm_ce": 0.05908827483654022,
|
| 243 |
+
"grad_norm_ratio_rl_ce": 9.437432840051576
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"step": 280,
|
| 247 |
+
"epoch": 4,
|
| 248 |
+
"sigma": 1.0,
|
| 249 |
+
"grad_cos_rl_ce": -0.08370169252157211,
|
| 250 |
+
"grad_norm_rl": 0.4843883514404297,
|
| 251 |
+
"grad_norm_ce": 0.06895625591278076,
|
| 252 |
+
"grad_norm_ratio_rl_ce": 7.024574420819885
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"step": 290,
|
| 256 |
+
"epoch": 4,
|
| 257 |
+
"sigma": 1.0,
|
| 258 |
+
"grad_cos_rl_ce": 0.7298516631126404,
|
| 259 |
+
"grad_norm_rl": 0.34276849031448364,
|
| 260 |
+
"grad_norm_ce": 0.0673132836818695,
|
| 261 |
+
"grad_norm_ratio_rl_ce": 5.092137414279889
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"step": 300,
|
| 265 |
+
"epoch": 4,
|
| 266 |
+
"sigma": 1.0,
|
| 267 |
+
"grad_cos_rl_ce": 0.7453404664993286,
|
| 268 |
+
"grad_norm_rl": 0.439138263463974,
|
| 269 |
+
"grad_norm_ce": 0.06507617235183716,
|
| 270 |
+
"grad_norm_ratio_rl_ce": 6.7480653454808905
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"step": 310,
|
| 274 |
+
"epoch": 4,
|
| 275 |
+
"sigma": 1.0,
|
| 276 |
+
"grad_cos_rl_ce": 0.16622745990753174,
|
| 277 |
+
"grad_norm_rl": 0.4058675169944763,
|
| 278 |
+
"grad_norm_ce": 0.07220175117254257,
|
| 279 |
+
"grad_norm_ratio_rl_ce": 5.621297411811566
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"step": 320,
|
| 283 |
+
"epoch": 4,
|
| 284 |
+
"sigma": 1.0,
|
| 285 |
+
"grad_cos_rl_ce": 0.07581481337547302,
|
| 286 |
+
"grad_norm_rl": 0.3595651090145111,
|
| 287 |
+
"grad_norm_ce": 0.05238677188754082,
|
| 288 |
+
"grad_norm_ratio_rl_ce": 6.863662257838543
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"step": 330,
|
| 292 |
+
"epoch": 4,
|
| 293 |
+
"sigma": 1.0,
|
| 294 |
+
"grad_cos_rl_ce": 0.5164712071418762,
|
| 295 |
+
"grad_norm_rl": 0.4224754869937897,
|
| 296 |
+
"grad_norm_ce": 0.06374914199113846,
|
| 297 |
+
"grad_norm_ratio_rl_ce": 6.627155657287379
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"step": 340,
|
| 301 |
+
"epoch": 4,
|
| 302 |
+
"sigma": 1.0,
|
| 303 |
+
"grad_cos_rl_ce": 0.02586999535560608,
|
| 304 |
+
"grad_norm_rl": 0.44144001603126526,
|
| 305 |
+
"grad_norm_ce": 0.05171021819114685,
|
| 306 |
+
"grad_norm_ratio_rl_ce": 8.536804358463195
|
| 307 |
+
}
|
| 308 |
+
]
|
e2_rlce_wrl2_sigma1_fixed_seed42/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:437dbc8bf474666f00cc393d300f0a76468b8e372f0ff4d652521f8163ee1bb1
|
| 3 |
+
size 1684141044
|
e2_rlce_wrl2_sigma1_fixed_seed42/model_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"encoder_name": "answerdotai/ModernBERT-large",
|
| 3 |
+
"head_layers": 2,
|
| 4 |
+
"head_dropout": 0.1,
|
| 5 |
+
"num_qtypes": 3,
|
| 6 |
+
"cls_query": false,
|
| 7 |
+
"slot_emb": false,
|
| 8 |
+
"max_len": 512,
|
| 9 |
+
"head_max_len": 192,
|
| 10 |
+
"k_buckets": [
|
| 11 |
+
2,
|
| 12 |
+
5,
|
| 13 |
+
10
|
| 14 |
+
],
|
| 15 |
+
"source_checkpoint": "epoch_4.pt",
|
| 16 |
+
"epoch": 4,
|
| 17 |
+
"step": 340
|
| 18 |
+
}
|
e2_rlce_wrl2_sigma1_fixed_seed42/test_eval.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"raw": {
|
| 3 |
+
"ece": 0.07199750246107578,
|
| 4 |
+
"brier": 0.06645381570230598,
|
| 5 |
+
"nll": 0.8978294851486571,
|
| 6 |
+
"accuracy": 0.7695,
|
| 7 |
+
"soft_accuracy": 0.5429161656722427,
|
| 8 |
+
"mean_confidence": 0.6985145596116781,
|
| 9 |
+
"mean_target_max": 0.6589462240487337,
|
| 10 |
+
"score_mae": 0.22932341744424775,
|
| 11 |
+
"n": 2000,
|
| 12 |
+
"by_type": {
|
| 13 |
+
"choice": {
|
| 14 |
+
"ece": 0.07335309053460756,
|
| 15 |
+
"brier": 0.08596291952401468,
|
| 16 |
+
"nll": 1.0393390773640325,
|
| 17 |
+
"accuracy": 0.7483333333333333,
|
| 18 |
+
"soft_accuracy": 0.49986916556954386,
|
| 19 |
+
"mean_confidence": 0.677424898892641,
|
| 20 |
+
"mean_target_max": 0.6345381752649943,
|
| 21 |
+
"score_mae": null,
|
| 22 |
+
"n": 600
|
| 23 |
+
},
|
| 24 |
+
"score": {
|
| 25 |
+
"ece": 0.08369359292089938,
|
| 26 |
+
"brier": 0.06953511679139637,
|
| 27 |
+
"nll": 1.0766921566054224,
|
| 28 |
+
"accuracy": 0.725,
|
| 29 |
+
"soft_accuracy": 0.46508267121389507,
|
| 30 |
+
"mean_confidence": 0.6444813898205757,
|
| 31 |
+
"mean_target_max": 0.5928848401084542,
|
| 32 |
+
"score_mae": 0.22932341744424775,
|
| 33 |
+
"n": 800
|
| 34 |
+
},
|
| 35 |
+
"noul": {
|
| 36 |
+
"ece": 0.06794380098581314,
|
| 37 |
+
"brier": 0.042836310428476757,
|
| 38 |
+
"nll": 0.517836330990928,
|
| 39 |
+
"accuracy": 0.85,
|
| 40 |
+
"soft_accuracy": 0.6897411583860715,
|
| 41 |
+
"mean_confidence": 0.7916484467188517,
|
| 42 |
+
"mean_target_max": 0.7714361180861791,
|
| 43 |
+
"score_mae": null,
|
| 44 |
+
"n": 600
|
| 45 |
+
}
|
| 46 |
+
}
|
| 47 |
+
},
|
| 48 |
+
"post_temperature": {
|
| 49 |
+
"ece": 0.13756407079100608,
|
| 50 |
+
"brier": 0.05274168649779342,
|
| 51 |
+
"nll": 0.8639668343029916,
|
| 52 |
+
"accuracy": 0.7695,
|
| 53 |
+
"soft_accuracy": 0.5127460444420576,
|
| 54 |
+
"mean_confidence": 0.6319359563440085,
|
| 55 |
+
"mean_target_max": 0.6589462240487337,
|
| 56 |
+
"score_mae": 0.21926877548685297,
|
| 57 |
+
"n": 2000,
|
| 58 |
+
"by_type": {
|
| 59 |
+
"choice": {
|
| 60 |
+
"ece": 0.15261346027255057,
|
| 61 |
+
"brier": 0.06783732821277833,
|
| 62 |
+
"nll": 0.9961423465857903,
|
| 63 |
+
"accuracy": 0.7483333333333333,
|
| 64 |
+
"soft_accuracy": 0.46264451851447425,
|
| 65 |
+
"mean_confidence": 0.5957198548813661,
|
| 66 |
+
"mean_target_max": 0.6345381752649943,
|
| 67 |
+
"score_mae": null,
|
| 68 |
+
"n": 600
|
| 69 |
+
},
|
| 70 |
+
"score": {
|
| 71 |
+
"ece": 0.14633062213659284,
|
| 72 |
+
"brier": 0.05423825833664523,
|
| 73 |
+
"nll": 1.0376627109292895,
|
| 74 |
+
"accuracy": 0.725,
|
| 75 |
+
"soft_accuracy": 0.43747206911444664,
|
| 76 |
+
"mean_confidence": 0.5786693847179413,
|
| 77 |
+
"mean_target_max": 0.5928848401084542,
|
| 78 |
+
"score_mae": 0.21926877548685297,
|
| 79 |
+
"n": 800
|
| 80 |
+
},
|
| 81 |
+
"noul": {
|
| 82 |
+
"ece": 0.11693950355052948,
|
| 83 |
+
"brier": 0.03565061566433945,
|
| 84 |
+
"nll": 0.5001968198517959,
|
| 85 |
+
"accuracy": 0.85,
|
| 86 |
+
"soft_accuracy": 0.6632128708064556,
|
| 87 |
+
"mean_confidence": 0.7391741533080737,
|
| 88 |
+
"mean_target_max": 0.7714361180861791,
|
| 89 |
+
"score_mae": null,
|
| 90 |
+
"n": 600
|
| 91 |
+
}
|
| 92 |
+
}
|
| 93 |
+
},
|
| 94 |
+
"fitted_temperature": {
|
| 95 |
+
"type0_bucket1": 1.3946102857589722,
|
| 96 |
+
"type2_bucket0": 1.4911189079284668,
|
| 97 |
+
"type1_bucket1": 1.3153350353240967
|
| 98 |
+
}
|
| 99 |
+
}
|
e2_rlce_wrl2_sigma1_fixed_seed42/train_log.json
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"step": 85,
|
| 5 |
+
"loss": 0.15223090385689456,
|
| 6 |
+
"sigma": 1.0,
|
| 7 |
+
"epoch_seconds": 157.04513931274414,
|
| 8 |
+
"train_rows": 5400,
|
| 9 |
+
"raw_ece": 0.03326516054570675,
|
| 10 |
+
"raw_brier": 0.10215330210793581,
|
| 11 |
+
"raw_nll": 0.9236466180106314,
|
| 12 |
+
"raw_accuracy": 0.6766666666666666,
|
| 13 |
+
"raw_soft_accuracy": 0.5181662528589368,
|
| 14 |
+
"raw_mean_confidence": 0.6689256232976913,
|
| 15 |
+
"raw_mean_target_max": 0.6725333379705747,
|
| 16 |
+
"raw_score_mae": 0.29017526041328284,
|
| 17 |
+
"raw_n": 600
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 2,
|
| 21 |
+
"step": 170,
|
| 22 |
+
"loss": -0.060645011870884426,
|
| 23 |
+
"sigma": 1.0,
|
| 24 |
+
"epoch_seconds": 156.7033338546753,
|
| 25 |
+
"train_rows": 5400,
|
| 26 |
+
"raw_ece": 0.10309493213891985,
|
| 27 |
+
"raw_brier": 0.08618591079543382,
|
| 28 |
+
"raw_nll": 0.89167972818017,
|
| 29 |
+
"raw_accuracy": 0.74,
|
| 30 |
+
"raw_soft_accuracy": 0.5099005859345198,
|
| 31 |
+
"raw_mean_confidence": 0.6386463411152363,
|
| 32 |
+
"raw_mean_target_max": 0.6725333379705747,
|
| 33 |
+
"raw_score_mae": 0.28722960625697547,
|
| 34 |
+
"raw_n": 600
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 3,
|
| 38 |
+
"step": 255,
|
| 39 |
+
"loss": -0.16430875096543163,
|
| 40 |
+
"sigma": 1.0,
|
| 41 |
+
"epoch_seconds": 156.8451325893402,
|
| 42 |
+
"train_rows": 5400,
|
| 43 |
+
"raw_ece": 0.07621498256921769,
|
| 44 |
+
"raw_brier": 0.07245252084390698,
|
| 45 |
+
"raw_nll": 0.8659721593504461,
|
| 46 |
+
"raw_accuracy": 0.765,
|
| 47 |
+
"raw_soft_accuracy": 0.5422361444930236,
|
| 48 |
+
"raw_mean_confidence": 0.6917001515626907,
|
| 49 |
+
"raw_mean_target_max": 0.6725333379705747,
|
| 50 |
+
"raw_score_mae": 0.2330165967248225,
|
| 51 |
+
"raw_n": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4,
|
| 55 |
+
"step": 340,
|
| 56 |
+
"loss": -0.22030044931407067,
|
| 57 |
+
"sigma": 1.0,
|
| 58 |
+
"epoch_seconds": 156.95254492759705,
|
| 59 |
+
"train_rows": 5400,
|
| 60 |
+
"raw_ece": 0.07950104941924414,
|
| 61 |
+
"raw_brier": 0.06905624482932145,
|
| 62 |
+
"raw_nll": 0.8632122053353426,
|
| 63 |
+
"raw_accuracy": 0.77,
|
| 64 |
+
"raw_soft_accuracy": 0.5465008930241069,
|
| 65 |
+
"raw_mean_confidence": 0.6925134967764218,
|
| 66 |
+
"raw_mean_target_max": 0.6725333379705747,
|
| 67 |
+
"raw_score_mae": 0.2308807331379891,
|
| 68 |
+
"raw_n": 600
|
| 69 |
+
}
|
| 70 |
+
]
|