Upload folder using huggingface_hub
Browse files- logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/args.json +1 -0
- logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/dataloader_04374.pt +3 -0
- logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/log2.txt +548 -0
- logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/model_04374.pt +3 -0
- logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/optimizer_04374.pt +3 -0
logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/args.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"hellaswag": true, "attention_kind": "selective", "log_dir": "logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 12, "n_embd": 264, "head_dim": 22, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 4375, "warmup_steps": 250, "group": "fix_1_latent_mask", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1340, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "n_latent_masks", "selection_head_linear_combo_scale": 1.0, "disable_selection_head_linear_combo_bias": false, "assert_latent_matches_no_head": false, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": false, "use_mini_model": false, "upload_to_hf": true, "seq_len": 256, "batch_size": 32, "total_batch_size": 131072, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": 2, "init_latent_masks_to_identity": true, "latent_mask_scale": null, "latent_mask_sigmoid": false, "S_layernorm": false, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": false, "mup_enable_coord_check_logging": false, "max_lr": 0.0025, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile"}
|
logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/dataloader_04374.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6031fd3e2855a036f7a5531cc24555aabd1115f9dd6618b8b2ca6f55279ef0b2
|
| 3 |
+
size 964
|
logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/log2.txt
ADDED
|
@@ -0,0 +1,548 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
max_steps: 4375
|
| 2 |
+
0 val loss 11.2068
|
| 3 |
+
0 val perplexity 73626.5078
|
| 4 |
+
0 hella 0.2453
|
| 5 |
+
0 train 11.205779 (lr=3.4965e-06) (hash(x)=45482580)
|
| 6 |
+
10 train 9.890795 (lr=3.8462e-05) (hash(x)=38414730)
|
| 7 |
+
20 train 9.453151 (lr=7.3427e-05) (hash(x)=39783917)
|
| 8 |
+
30 train 8.800746 (lr=1.0839e-04) (hash(x)=38395733)
|
| 9 |
+
40 train 8.090767 (lr=1.4336e-04) (hash(x)=37486707)
|
| 10 |
+
50 train 7.688941 (lr=1.7832e-04) (hash(x)=39897505)
|
| 11 |
+
60 train 7.564519 (lr=2.1329e-04) (hash(x)=40933473)
|
| 12 |
+
70 train 7.572167 (lr=2.4825e-04) (hash(x)=37690746)
|
| 13 |
+
80 train 7.474277 (lr=2.8322e-04) (hash(x)=38609656)
|
| 14 |
+
90 train 7.251230 (lr=3.1818e-04) (hash(x)=38356571)
|
| 15 |
+
100 val loss 7.1454
|
| 16 |
+
100 val perplexity 1268.2235
|
| 17 |
+
100 train 7.123769 (lr=3.5315e-04) (hash(x)=35980376)
|
| 18 |
+
110 train 7.004111 (lr=3.8811e-04) (hash(x)=42011042)
|
| 19 |
+
120 train 6.897711 (lr=4.2308e-04) (hash(x)=40266823)
|
| 20 |
+
130 train 6.726137 (lr=4.5804e-04) (hash(x)=38645447)
|
| 21 |
+
140 train 6.607510 (lr=4.9301e-04) (hash(x)=36059313)
|
| 22 |
+
150 train 6.547801 (lr=5.2797e-04) (hash(x)=34161947)
|
| 23 |
+
160 train 6.431300 (lr=5.6294e-04) (hash(x)=42263375)
|
| 24 |
+
170 train 6.484595 (lr=5.9790e-04) (hash(x)=42957725)
|
| 25 |
+
180 train 6.499974 (lr=6.3287e-04) (hash(x)=40198018)
|
| 26 |
+
190 train 6.242249 (lr=6.6783e-04) (hash(x)=41666215)
|
| 27 |
+
200 val loss 6.2064
|
| 28 |
+
200 val perplexity 495.9352
|
| 29 |
+
200 train 6.167308 (lr=7.0280e-04) (hash(x)=54060482)
|
| 30 |
+
210 train 6.110983 (lr=7.3776e-04) (hash(x)=42332778)
|
| 31 |
+
220 train 6.029080 (lr=7.7273e-04) (hash(x)=39551486)
|
| 32 |
+
230 train 6.030873 (lr=8.0769e-04) (hash(x)=38462018)
|
| 33 |
+
240 train 6.091299 (lr=8.4266e-04) (hash(x)=36591442)
|
| 34 |
+
250 hella 0.2403
|
| 35 |
+
250 train 6.008242 (lr=8.7762e-04) (hash(x)=40861237)
|
| 36 |
+
260 train 5.954885 (lr=9.1259e-04) (hash(x)=41739369)
|
| 37 |
+
270 train 5.832683 (lr=9.4755e-04) (hash(x)=39415360)
|
| 38 |
+
280 train 5.816064 (lr=9.8252e-04) (hash(x)=44665934)
|
| 39 |
+
290 train 5.746920 (lr=1.0175e-03) (hash(x)=37035579)
|
| 40 |
+
300 val loss 5.7333
|
| 41 |
+
300 val perplexity 308.9760
|
| 42 |
+
300 train 5.758825 (lr=1.0524e-03) (hash(x)=38301011)
|
| 43 |
+
310 train 5.601055 (lr=1.0874e-03) (hash(x)=42850980)
|
| 44 |
+
320 train 5.664670 (lr=1.1224e-03) (hash(x)=37476222)
|
| 45 |
+
330 train 5.616964 (lr=1.1573e-03) (hash(x)=53028205)
|
| 46 |
+
340 train 5.606616 (lr=1.1923e-03) (hash(x)=41466008)
|
| 47 |
+
350 train 5.512977 (lr=1.2273e-03) (hash(x)=37802865)
|
| 48 |
+
360 train 5.452415 (lr=1.2622e-03) (hash(x)=42769282)
|
| 49 |
+
370 train 5.482262 (lr=1.2972e-03) (hash(x)=39319256)
|
| 50 |
+
380 train 5.458792 (lr=1.3322e-03) (hash(x)=42637402)
|
| 51 |
+
390 train 5.428664 (lr=1.3671e-03) (hash(x)=31141514)
|
| 52 |
+
400 val loss 5.4183
|
| 53 |
+
400 val perplexity 225.4922
|
| 54 |
+
400 train 5.495156 (lr=1.4021e-03) (hash(x)=38151157)
|
| 55 |
+
410 train 5.438721 (lr=1.4371e-03) (hash(x)=46033439)
|
| 56 |
+
420 train 5.412293 (lr=1.4720e-03) (hash(x)=41365246)
|
| 57 |
+
430 train 5.385355 (lr=1.5070e-03) (hash(x)=42369184)
|
| 58 |
+
440 train 5.376598 (lr=1.5420e-03) (hash(x)=42004840)
|
| 59 |
+
450 train 5.307515 (lr=1.5769e-03) (hash(x)=37181172)
|
| 60 |
+
460 train 5.392766 (lr=1.6119e-03) (hash(x)=31630797)
|
| 61 |
+
470 train 5.266939 (lr=1.6469e-03) (hash(x)=42135747)
|
| 62 |
+
480 train 5.186979 (lr=1.6818e-03) (hash(x)=30023651)
|
| 63 |
+
490 train 5.264831 (lr=1.7168e-03) (hash(x)=40463476)
|
| 64 |
+
500 val loss 5.2205
|
| 65 |
+
500 val perplexity 185.0329
|
| 66 |
+
500 hella 0.2377
|
| 67 |
+
500 train 5.140098 (lr=1.7517e-03) (hash(x)=37089842)
|
| 68 |
+
510 train 5.161531 (lr=1.7867e-03) (hash(x)=41095974)
|
| 69 |
+
520 train 5.207872 (lr=1.8217e-03) (hash(x)=41525258)
|
| 70 |
+
530 train 5.193594 (lr=1.8566e-03) (hash(x)=37697291)
|
| 71 |
+
540 train 5.213845 (lr=1.8916e-03) (hash(x)=43459894)
|
| 72 |
+
550 train 5.077550 (lr=1.9266e-03) (hash(x)=38993956)
|
| 73 |
+
560 train 5.102020 (lr=1.9615e-03) (hash(x)=40677277)
|
| 74 |
+
570 train 5.176210 (lr=1.9965e-03) (hash(x)=43103127)
|
| 75 |
+
580 train 5.096685 (lr=2.0315e-03) (hash(x)=41871776)
|
| 76 |
+
590 train 5.103245 (lr=2.0664e-03) (hash(x)=43230967)
|
| 77 |
+
600 val loss 5.0229
|
| 78 |
+
600 val perplexity 151.8582
|
| 79 |
+
600 train 5.063087 (lr=2.1014e-03) (hash(x)=42226172)
|
| 80 |
+
610 train 5.407247 (lr=2.1364e-03) (hash(x)=43007164)
|
| 81 |
+
620 train 4.914196 (lr=2.1713e-03) (hash(x)=35752341)
|
| 82 |
+
630 train 4.980386 (lr=2.2063e-03) (hash(x)=39453504)
|
| 83 |
+
640 train 4.926097 (lr=2.2413e-03) (hash(x)=43635712)
|
| 84 |
+
650 train 4.933643 (lr=2.2762e-03) (hash(x)=41852456)
|
| 85 |
+
660 train 4.958025 (lr=2.3112e-03) (hash(x)=49125742)
|
| 86 |
+
670 train 4.948305 (lr=2.3462e-03) (hash(x)=34886275)
|
| 87 |
+
680 train 5.026585 (lr=2.3811e-03) (hash(x)=40546482)
|
| 88 |
+
690 train 4.836580 (lr=2.4161e-03) (hash(x)=40850901)
|
| 89 |
+
700 val loss 4.8319
|
| 90 |
+
700 val perplexity 125.4477
|
| 91 |
+
700 train 4.817990 (lr=2.4510e-03) (hash(x)=37426993)
|
| 92 |
+
710 train 4.978594 (lr=2.4860e-03) (hash(x)=42679743)
|
| 93 |
+
720 train 4.784123 (lr=2.5000e-03) (hash(x)=44856043)
|
| 94 |
+
730 train 4.766061 (lr=2.4999e-03) (hash(x)=36457471)
|
| 95 |
+
740 train 4.722989 (lr=2.4997e-03) (hash(x)=47203534)
|
| 96 |
+
750 hella 0.2473
|
| 97 |
+
750 train 4.744538 (lr=2.4995e-03) (hash(x)=35585657)
|
| 98 |
+
760 train 4.654180 (lr=2.4992e-03) (hash(x)=37259264)
|
| 99 |
+
770 train 4.576164 (lr=2.4987e-03) (hash(x)=44161997)
|
| 100 |
+
780 train 4.517097 (lr=2.4982e-03) (hash(x)=41272886)
|
| 101 |
+
790 train 4.518215 (lr=2.4977e-03) (hash(x)=43303662)
|
| 102 |
+
800 val loss 4.7176
|
| 103 |
+
800 val perplexity 111.9005
|
| 104 |
+
800 train 4.423057 (lr=2.4970e-03) (hash(x)=39067231)
|
| 105 |
+
810 train 4.658465 (lr=2.4963e-03) (hash(x)=44046732)
|
| 106 |
+
820 train 4.646113 (lr=2.4954e-03) (hash(x)=39783962)
|
| 107 |
+
830 train 4.705292 (lr=2.4945e-03) (hash(x)=40973974)
|
| 108 |
+
840 train 4.671281 (lr=2.4935e-03) (hash(x)=41619968)
|
| 109 |
+
850 train 4.717260 (lr=2.4925e-03) (hash(x)=36939960)
|
| 110 |
+
860 train 4.605565 (lr=2.4913e-03) (hash(x)=42942377)
|
| 111 |
+
870 train 4.609995 (lr=2.4901e-03) (hash(x)=39659455)
|
| 112 |
+
880 train 4.625117 (lr=2.4887e-03) (hash(x)=40331986)
|
| 113 |
+
890 train 4.534225 (lr=2.4873e-03) (hash(x)=38084814)
|
| 114 |
+
900 val loss 4.5708
|
| 115 |
+
900 val perplexity 96.6242
|
| 116 |
+
900 train 4.561557 (lr=2.4858e-03) (hash(x)=30830367)
|
| 117 |
+
910 train 4.474580 (lr=2.4843e-03) (hash(x)=39664356)
|
| 118 |
+
920 train 4.482002 (lr=2.4826e-03) (hash(x)=39007775)
|
| 119 |
+
930 train 4.504342 (lr=2.4809e-03) (hash(x)=39319254)
|
| 120 |
+
940 train 4.548746 (lr=2.4791e-03) (hash(x)=36347051)
|
| 121 |
+
950 train 4.504550 (lr=2.4772e-03) (hash(x)=39243577)
|
| 122 |
+
960 train 4.527729 (lr=2.4752e-03) (hash(x)=45603839)
|
| 123 |
+
970 train 4.339453 (lr=2.4732e-03) (hash(x)=40036075)
|
| 124 |
+
980 train 4.389059 (lr=2.4710e-03) (hash(x)=43068524)
|
| 125 |
+
990 train 4.377832 (lr=2.4688e-03) (hash(x)=33798472)
|
| 126 |
+
1000 val loss 4.5367
|
| 127 |
+
1000 val perplexity 93.3818
|
| 128 |
+
1000 hella 0.2507
|
| 129 |
+
1000 train 4.741880 (lr=2.4665e-03) (hash(x)=37728665)
|
| 130 |
+
1010 train 4.606746 (lr=2.4641e-03) (hash(x)=43678636)
|
| 131 |
+
1020 train 4.632055 (lr=2.4617e-03) (hash(x)=48464376)
|
| 132 |
+
1030 train 4.589381 (lr=2.4591e-03) (hash(x)=43771461)
|
| 133 |
+
1040 train 4.449974 (lr=2.4565e-03) (hash(x)=44643085)
|
| 134 |
+
1050 train 4.531778 (lr=2.4538e-03) (hash(x)=39201716)
|
| 135 |
+
1060 train 4.438889 (lr=2.4510e-03) (hash(x)=36396343)
|
| 136 |
+
1070 train 4.435174 (lr=2.4482e-03) (hash(x)=37504111)
|
| 137 |
+
1080 train 4.473915 (lr=2.4452e-03) (hash(x)=37454537)
|
| 138 |
+
1090 train 4.456357 (lr=2.4422e-03) (hash(x)=36888183)
|
| 139 |
+
1100 val loss 4.4493
|
| 140 |
+
1100 val perplexity 85.5665
|
| 141 |
+
1100 train 4.399782 (lr=2.4391e-03) (hash(x)=39549045)
|
| 142 |
+
1110 train 4.590610 (lr=2.4360e-03) (hash(x)=56818420)
|
| 143 |
+
1120 train 4.400492 (lr=2.4327e-03) (hash(x)=36573357)
|
| 144 |
+
1130 train 4.340428 (lr=2.4294e-03) (hash(x)=38540878)
|
| 145 |
+
1140 train 4.328143 (lr=2.4260e-03) (hash(x)=38645859)
|
| 146 |
+
1150 train 4.538775 (lr=2.4225e-03) (hash(x)=41972503)
|
| 147 |
+
1160 train 4.157328 (lr=2.4189e-03) (hash(x)=38151868)
|
| 148 |
+
1170 train 4.188759 (lr=2.4153e-03) (hash(x)=42375436)
|
| 149 |
+
1180 train 4.169129 (lr=2.4116e-03) (hash(x)=42868028)
|
| 150 |
+
1190 train 4.423557 (lr=2.4078e-03) (hash(x)=41060868)
|
| 151 |
+
1200 val loss 4.4131
|
| 152 |
+
1200 val perplexity 82.5226
|
| 153 |
+
1200 train 4.431404 (lr=2.4039e-03) (hash(x)=40852814)
|
| 154 |
+
1210 train 4.406247 (lr=2.4000e-03) (hash(x)=43365554)
|
| 155 |
+
1220 train 4.346675 (lr=2.3960e-03) (hash(x)=39121134)
|
| 156 |
+
1230 train 4.438187 (lr=2.3919e-03) (hash(x)=40565426)
|
| 157 |
+
1240 train 4.501116 (lr=2.3877e-03) (hash(x)=43378926)
|
| 158 |
+
1250 hella 0.2508
|
| 159 |
+
1250 train 4.356571 (lr=2.3834e-03) (hash(x)=35227381)
|
| 160 |
+
1260 train 4.456383 (lr=2.3791e-03) (hash(x)=41267226)
|
| 161 |
+
1270 train 4.355784 (lr=2.3747e-03) (hash(x)=31395565)
|
| 162 |
+
1280 train 4.467368 (lr=2.3703e-03) (hash(x)=41470963)
|
| 163 |
+
1290 train 4.288065 (lr=2.3657e-03) (hash(x)=40729366)
|
| 164 |
+
1300 val loss 4.3481
|
| 165 |
+
1300 val perplexity 77.3305
|
| 166 |
+
1300 train 4.320520 (lr=2.3611e-03) (hash(x)=39916663)
|
| 167 |
+
1310 train 4.377025 (lr=2.3564e-03) (hash(x)=42179962)
|
| 168 |
+
1320 train 4.233400 (lr=2.3517e-03) (hash(x)=44991932)
|
| 169 |
+
1330 train 4.209435 (lr=2.3469e-03) (hash(x)=38925911)
|
| 170 |
+
1340 train 4.159973 (lr=2.3420e-03) (hash(x)=38565560)
|
| 171 |
+
1350 train 4.202387 (lr=2.3370e-03) (hash(x)=43384902)
|
| 172 |
+
1360 train 4.152455 (lr=2.3319e-03) (hash(x)=45422261)
|
| 173 |
+
1370 train 4.133152 (lr=2.3268e-03) (hash(x)=40167325)
|
| 174 |
+
1380 train 4.288229 (lr=2.3216e-03) (hash(x)=51258162)
|
| 175 |
+
1390 train 4.352070 (lr=2.3164e-03) (hash(x)=43229106)
|
| 176 |
+
1400 val loss 4.3079
|
| 177 |
+
1400 val perplexity 74.2812
|
| 178 |
+
1400 train 4.257721 (lr=2.3111e-03) (hash(x)=42537494)
|
| 179 |
+
1410 train 4.557415 (lr=2.3057e-03) (hash(x)=51236490)
|
| 180 |
+
1420 train 4.405765 (lr=2.3002e-03) (hash(x)=45243347)
|
| 181 |
+
1430 train 4.240701 (lr=2.2947e-03) (hash(x)=37658566)
|
| 182 |
+
1440 train 4.305216 (lr=2.2891e-03) (hash(x)=38856106)
|
| 183 |
+
1450 train 4.370714 (lr=2.2834e-03) (hash(x)=36649488)
|
| 184 |
+
1460 train 4.367593 (lr=2.2777e-03) (hash(x)=40445407)
|
| 185 |
+
1470 train 4.303241 (lr=2.2719e-03) (hash(x)=39916387)
|
| 186 |
+
1480 train 4.272441 (lr=2.2661e-03) (hash(x)=40934555)
|
| 187 |
+
1490 train 4.238917 (lr=2.2601e-03) (hash(x)=42276527)
|
| 188 |
+
1500 val loss 4.2897
|
| 189 |
+
1500 val perplexity 72.9480
|
| 190 |
+
1500 hella 0.2489
|
| 191 |
+
1500 train 4.232102 (lr=2.2541e-03) (hash(x)=37973883)
|
| 192 |
+
1510 train 4.294239 (lr=2.2481e-03) (hash(x)=38929286)
|
| 193 |
+
1520 train 4.105873 (lr=2.2420e-03) (hash(x)=36855837)
|
| 194 |
+
1530 train 4.353456 (lr=2.2358e-03) (hash(x)=42146431)
|
| 195 |
+
1540 train 4.294855 (lr=2.2295e-03) (hash(x)=41898411)
|
| 196 |
+
1550 train 4.203965 (lr=2.2232e-03) (hash(x)=44906272)
|
| 197 |
+
1560 train 4.233556 (lr=2.2168e-03) (hash(x)=40414353)
|
| 198 |
+
1570 train 4.262726 (lr=2.2104e-03) (hash(x)=38723190)
|
| 199 |
+
1580 train 4.182793 (lr=2.2039e-03) (hash(x)=39065271)
|
| 200 |
+
1590 train 4.179083 (lr=2.1973e-03) (hash(x)=40000886)
|
| 201 |
+
1600 val loss 4.2609
|
| 202 |
+
1600 val perplexity 70.8719
|
| 203 |
+
1600 train 4.098089 (lr=2.1907e-03) (hash(x)=37898571)
|
| 204 |
+
1610 train 4.200827 (lr=2.1840e-03) (hash(x)=40447863)
|
| 205 |
+
1620 train 4.172432 (lr=2.1773e-03) (hash(x)=40119318)
|
| 206 |
+
1630 train 4.102351 (lr=2.1705e-03) (hash(x)=40664074)
|
| 207 |
+
1640 train 4.072355 (lr=2.1636e-03) (hash(x)=37404741)
|
| 208 |
+
1650 train 4.230998 (lr=2.1567e-03) (hash(x)=39214928)
|
| 209 |
+
1660 train 4.258530 (lr=2.1497e-03) (hash(x)=35005019)
|
| 210 |
+
1670 train 4.301500 (lr=2.1427e-03) (hash(x)=37675832)
|
| 211 |
+
1680 train 4.245173 (lr=2.1356e-03) (hash(x)=40489680)
|
| 212 |
+
1690 train 4.163901 (lr=2.1285e-03) (hash(x)=45537879)
|
| 213 |
+
1700 val loss 4.2264
|
| 214 |
+
1700 val perplexity 68.4712
|
| 215 |
+
1700 train 4.143909 (lr=2.1213e-03) (hash(x)=38866100)
|
| 216 |
+
1710 train 4.206710 (lr=2.1140e-03) (hash(x)=44726254)
|
| 217 |
+
1720 train 4.177144 (lr=2.1067e-03) (hash(x)=36917792)
|
| 218 |
+
1730 train 4.168089 (lr=2.0994e-03) (hash(x)=42814805)
|
| 219 |
+
1740 train 4.240849 (lr=2.0920e-03) (hash(x)=51812216)
|
| 220 |
+
1750 hella 0.2516
|
| 221 |
+
1750 train 4.115794 (lr=2.0845e-03) (hash(x)=33549014)
|
| 222 |
+
1760 train 4.128290 (lr=2.0770e-03) (hash(x)=40354215)
|
| 223 |
+
1770 train 4.143920 (lr=2.0694e-03) (hash(x)=40718606)
|
| 224 |
+
1780 train 4.080493 (lr=2.0618e-03) (hash(x)=38274164)
|
| 225 |
+
1790 train 4.251867 (lr=2.0541e-03) (hash(x)=42531471)
|
| 226 |
+
1800 val loss 4.2075
|
| 227 |
+
1800 val perplexity 67.1900
|
| 228 |
+
1800 train 4.185470 (lr=2.0464e-03) (hash(x)=35616519)
|
| 229 |
+
1810 train 4.240880 (lr=2.0386e-03) (hash(x)=33803118)
|
| 230 |
+
1820 train 4.242918 (lr=2.0308e-03) (hash(x)=36973525)
|
| 231 |
+
1830 train 4.183969 (lr=2.0229e-03) (hash(x)=38162549)
|
| 232 |
+
1840 train 4.148647 (lr=2.0150e-03) (hash(x)=33937159)
|
| 233 |
+
1850 train 4.255347 (lr=2.0070e-03) (hash(x)=39887546)
|
| 234 |
+
1860 train 4.110179 (lr=1.9990e-03) (hash(x)=37818525)
|
| 235 |
+
1870 train 4.141788 (lr=1.9910e-03) (hash(x)=37250478)
|
| 236 |
+
1880 train 4.033772 (lr=1.9829e-03) (hash(x)=36412167)
|
| 237 |
+
1890 train 4.145341 (lr=1.9747e-03) (hash(x)=40261189)
|
| 238 |
+
1900 val loss 4.2019
|
| 239 |
+
1900 val perplexity 66.8117
|
| 240 |
+
1900 train 4.101543 (lr=1.9665e-03) (hash(x)=38654303)
|
| 241 |
+
1910 train 4.075214 (lr=1.9583e-03) (hash(x)=37929515)
|
| 242 |
+
1920 train 4.249940 (lr=1.9500e-03) (hash(x)=45580146)
|
| 243 |
+
1930 train 4.194916 (lr=1.9417e-03) (hash(x)=32386330)
|
| 244 |
+
1940 train 4.227789 (lr=1.9333e-03) (hash(x)=36331864)
|
| 245 |
+
1950 train 4.210799 (lr=1.9249e-03) (hash(x)=41181727)
|
| 246 |
+
1960 train 4.040340 (lr=1.9165e-03) (hash(x)=42705152)
|
| 247 |
+
1970 train 4.149493 (lr=1.9080e-03) (hash(x)=40251511)
|
| 248 |
+
1980 train 4.067660 (lr=1.8995e-03) (hash(x)=37525551)
|
| 249 |
+
1990 train 4.142854 (lr=1.8909e-03) (hash(x)=41849618)
|
| 250 |
+
2000 val loss 4.1674
|
| 251 |
+
2000 val perplexity 64.5449
|
| 252 |
+
2000 hella 0.2532
|
| 253 |
+
2000 train 4.143337 (lr=1.8823e-03) (hash(x)=37642582)
|
| 254 |
+
2010 train 4.143160 (lr=1.8737e-03) (hash(x)=42131121)
|
| 255 |
+
2020 train 4.026949 (lr=1.8650e-03) (hash(x)=39000209)
|
| 256 |
+
2030 train 4.118140 (lr=1.8563e-03) (hash(x)=43641355)
|
| 257 |
+
2040 train 4.010603 (lr=1.8475e-03) (hash(x)=47910507)
|
| 258 |
+
2050 train 4.011080 (lr=1.8387e-03) (hash(x)=36670359)
|
| 259 |
+
2060 train 4.207928 (lr=1.8299e-03) (hash(x)=36477755)
|
| 260 |
+
2070 train 4.299851 (lr=1.8211e-03) (hash(x)=43262487)
|
| 261 |
+
2080 train 4.111676 (lr=1.8122e-03) (hash(x)=49546029)
|
| 262 |
+
2090 train 4.241959 (lr=1.8033e-03) (hash(x)=39924731)
|
| 263 |
+
2100 val loss 4.1547
|
| 264 |
+
2100 val perplexity 63.7336
|
| 265 |
+
2100 train 4.112255 (lr=1.7943e-03) (hash(x)=39921304)
|
| 266 |
+
2110 train 4.053962 (lr=1.7854e-03) (hash(x)=38152788)
|
| 267 |
+
2120 train 4.108117 (lr=1.7764e-03) (hash(x)=37977911)
|
| 268 |
+
2130 train 4.070705 (lr=1.7673e-03) (hash(x)=34748760)
|
| 269 |
+
2140 train 4.125025 (lr=1.7583e-03) (hash(x)=40509369)
|
| 270 |
+
2150 train 4.088189 (lr=1.7492e-03) (hash(x)=37654262)
|
| 271 |
+
2160 train 4.048430 (lr=1.7400e-03) (hash(x)=38139543)
|
| 272 |
+
2170 train 3.955610 (lr=1.7309e-03) (hash(x)=42501806)
|
| 273 |
+
2180 train 4.108547 (lr=1.7217e-03) (hash(x)=40085092)
|
| 274 |
+
2190 train 4.045713 (lr=1.7125e-03) (hash(x)=50480193)
|
| 275 |
+
2200 val loss 4.1363
|
| 276 |
+
2200 val perplexity 62.5703
|
| 277 |
+
2200 train 4.136502 (lr=1.7033e-03) (hash(x)=40604084)
|
| 278 |
+
2210 train 4.244066 (lr=1.6941e-03) (hash(x)=41555823)
|
| 279 |
+
2220 train 4.068172 (lr=1.6848e-03) (hash(x)=50441765)
|
| 280 |
+
2230 train 4.160952 (lr=1.6755e-03) (hash(x)=39796580)
|
| 281 |
+
2240 train 4.187014 (lr=1.6662e-03) (hash(x)=44127022)
|
| 282 |
+
2250 hella 0.2537
|
| 283 |
+
2250 train 4.090375 (lr=1.6568e-03) (hash(x)=37026826)
|
| 284 |
+
2260 train 4.127663 (lr=1.6475e-03) (hash(x)=42133839)
|
| 285 |
+
2270 train 4.090043 (lr=1.6381e-03) (hash(x)=38500664)
|
| 286 |
+
2280 train 4.076827 (lr=1.6287e-03) (hash(x)=40538661)
|
| 287 |
+
2290 train 4.079439 (lr=1.6193e-03) (hash(x)=51509210)
|
| 288 |
+
2300 val loss 4.1132
|
| 289 |
+
2300 val perplexity 61.1432
|
| 290 |
+
2300 train 4.084836 (lr=1.6098e-03) (hash(x)=41952328)
|
| 291 |
+
2310 train 4.009544 (lr=1.6004e-03) (hash(x)=39758123)
|
| 292 |
+
2320 train 4.128554 (lr=1.5909e-03) (hash(x)=51089268)
|
| 293 |
+
2330 train 4.182822 (lr=1.5814e-03) (hash(x)=39767618)
|
| 294 |
+
2340 train 4.225821 (lr=1.5719e-03) (hash(x)=40409617)
|
| 295 |
+
2350 train 4.063956 (lr=1.5624e-03) (hash(x)=40349634)
|
| 296 |
+
2360 train 4.335771 (lr=1.5529e-03) (hash(x)=31841172)
|
| 297 |
+
2370 train 4.128212 (lr=1.5434e-03) (hash(x)=42720539)
|
| 298 |
+
2380 train 3.994942 (lr=1.5338e-03) (hash(x)=40998632)
|
| 299 |
+
2390 train 4.098209 (lr=1.5242e-03) (hash(x)=40615413)
|
| 300 |
+
2400 val loss 4.0955
|
| 301 |
+
2400 val perplexity 60.0697
|
| 302 |
+
2400 train 4.014865 (lr=1.5147e-03) (hash(x)=39373658)
|
| 303 |
+
2410 train 4.061779 (lr=1.5051e-03) (hash(x)=35480858)
|
| 304 |
+
2420 train 4.051028 (lr=1.4955e-03) (hash(x)=32241095)
|
| 305 |
+
2430 train 4.036877 (lr=1.4859e-03) (hash(x)=36669715)
|
| 306 |
+
2440 train 4.036469 (lr=1.4763e-03) (hash(x)=45768335)
|
| 307 |
+
2450 train 3.945927 (lr=1.4666e-03) (hash(x)=39395055)
|
| 308 |
+
2460 train 3.972100 (lr=1.4570e-03) (hash(x)=34899269)
|
| 309 |
+
2470 train 4.165618 (lr=1.4474e-03) (hash(x)=43642420)
|
| 310 |
+
2480 train 4.262723 (lr=1.4377e-03) (hash(x)=47099507)
|
| 311 |
+
2490 train 4.137328 (lr=1.4281e-03) (hash(x)=38825558)
|
| 312 |
+
2500 val loss 4.0814
|
| 313 |
+
2500 val perplexity 59.2311
|
| 314 |
+
2500 hella 0.2522
|
| 315 |
+
2500 train 4.115538 (lr=1.4184e-03) (hash(x)=39833804)
|
| 316 |
+
2510 train 4.207315 (lr=1.4088e-03) (hash(x)=35051654)
|
| 317 |
+
2520 train 4.106496 (lr=1.3991e-03) (hash(x)=40291109)
|
| 318 |
+
2530 train 4.017591 (lr=1.3895e-03) (hash(x)=36915768)
|
| 319 |
+
2540 train 3.995523 (lr=1.3798e-03) (hash(x)=37016308)
|
| 320 |
+
2550 train 3.960024 (lr=1.3702e-03) (hash(x)=43221777)
|
| 321 |
+
2560 train 4.010394 (lr=1.3605e-03) (hash(x)=37233207)
|
| 322 |
+
2570 train 3.910670 (lr=1.3509e-03) (hash(x)=37510626)
|
| 323 |
+
2580 train 3.962257 (lr=1.3412e-03) (hash(x)=39624656)
|
| 324 |
+
2590 train 3.816870 (lr=1.3316e-03) (hash(x)=37647501)
|
| 325 |
+
2600 val loss 4.0713
|
| 326 |
+
2600 val perplexity 58.6352
|
| 327 |
+
2600 train 3.950845 (lr=1.3219e-03) (hash(x)=37200138)
|
| 328 |
+
2610 train 3.960705 (lr=1.3123e-03) (hash(x)=43066970)
|
| 329 |
+
2620 train 4.103157 (lr=1.3026e-03) (hash(x)=40503799)
|
| 330 |
+
2630 train 4.034739 (lr=1.2930e-03) (hash(x)=45534660)
|
| 331 |
+
2640 train 4.252683 (lr=1.2834e-03) (hash(x)=52358353)
|
| 332 |
+
2650 train 4.029259 (lr=1.2737e-03) (hash(x)=41843599)
|
| 333 |
+
2660 train 4.018857 (lr=1.2641e-03) (hash(x)=40519770)
|
| 334 |
+
2670 train 4.092648 (lr=1.2545e-03) (hash(x)=37185602)
|
| 335 |
+
2680 train 3.957591 (lr=1.2449e-03) (hash(x)=43805719)
|
| 336 |
+
2690 train 3.969815 (lr=1.2353e-03) (hash(x)=39458250)
|
| 337 |
+
2700 val loss 4.0476
|
| 338 |
+
2700 val perplexity 57.2623
|
| 339 |
+
2700 train 4.071019 (lr=1.2258e-03) (hash(x)=39149255)
|
| 340 |
+
2710 train 3.912796 (lr=1.2162e-03) (hash(x)=43544861)
|
| 341 |
+
2720 train 4.017031 (lr=1.2066e-03) (hash(x)=36858197)
|
| 342 |
+
2730 train 3.935599 (lr=1.1971e-03) (hash(x)=37971712)
|
| 343 |
+
2740 train 3.962466 (lr=1.1876e-03) (hash(x)=42416578)
|
| 344 |
+
2750 hella 0.2561
|
| 345 |
+
2750 train 3.910066 (lr=1.1781e-03) (hash(x)=53554622)
|
| 346 |
+
2760 train 4.049000 (lr=1.1686e-03) (hash(x)=40502545)
|
| 347 |
+
2770 train 4.015411 (lr=1.1591e-03) (hash(x)=36086417)
|
| 348 |
+
2780 train 4.037203 (lr=1.1496e-03) (hash(x)=36242982)
|
| 349 |
+
2790 train 4.114482 (lr=1.1402e-03) (hash(x)=41433780)
|
| 350 |
+
2800 val loss 4.0271
|
| 351 |
+
2800 val perplexity 56.0956
|
| 352 |
+
2800 train 3.984597 (lr=1.1307e-03) (hash(x)=48553484)
|
| 353 |
+
2810 train 3.922276 (lr=1.1213e-03) (hash(x)=39928650)
|
| 354 |
+
2820 train 4.030788 (lr=1.1119e-03) (hash(x)=37891724)
|
| 355 |
+
2830 train 4.008041 (lr=1.1025e-03) (hash(x)=34353412)
|
| 356 |
+
2840 train 4.024433 (lr=1.0932e-03) (hash(x)=38486611)
|
| 357 |
+
2850 train 3.947093 (lr=1.0838e-03) (hash(x)=32706934)
|
| 358 |
+
2860 train 3.895869 (lr=1.0745e-03) (hash(x)=38693681)
|
| 359 |
+
2870 train 3.962201 (lr=1.0652e-03) (hash(x)=38430800)
|
| 360 |
+
2880 train 3.936229 (lr=1.0559e-03) (hash(x)=40619559)
|
| 361 |
+
2890 train 3.911396 (lr=1.0467e-03) (hash(x)=42741066)
|
| 362 |
+
2900 val loss 4.0297
|
| 363 |
+
2900 val perplexity 56.2419
|
| 364 |
+
2900 train 4.159233 (lr=1.0375e-03) (hash(x)=38216091)
|
| 365 |
+
2910 train 3.969090 (lr=1.0283e-03) (hash(x)=35447832)
|
| 366 |
+
2920 train 4.075551 (lr=1.0191e-03) (hash(x)=39825190)
|
| 367 |
+
2930 train 4.091529 (lr=1.0100e-03) (hash(x)=37789121)
|
| 368 |
+
2940 train 3.966560 (lr=1.0008e-03) (hash(x)=44499116)
|
| 369 |
+
2950 train 4.050604 (lr=9.9174e-04) (hash(x)=39669860)
|
| 370 |
+
2960 train 3.993392 (lr=9.8268e-04) (hash(x)=51683741)
|
| 371 |
+
2970 train 3.863164 (lr=9.7364e-04) (hash(x)=36399721)
|
| 372 |
+
2980 train 3.956916 (lr=9.6463e-04) (hash(x)=42629700)
|
| 373 |
+
2990 train 3.842618 (lr=9.5566e-04) (hash(x)=39263773)
|
| 374 |
+
3000 val loss 4.0052
|
| 375 |
+
3000 val perplexity 54.8827
|
| 376 |
+
3000 hella 0.2525
|
| 377 |
+
3000 train 3.882972 (lr=9.4671e-04) (hash(x)=37920485)
|
| 378 |
+
3010 train 3.928420 (lr=9.3780e-04) (hash(x)=40835161)
|
| 379 |
+
3020 train 3.892507 (lr=9.2892e-04) (hash(x)=38381159)
|
| 380 |
+
3030 train 3.886180 (lr=9.2007e-04) (hash(x)=35846270)
|
| 381 |
+
3040 train 4.276841 (lr=9.1125e-04) (hash(x)=47516567)
|
| 382 |
+
3050 train 4.114841 (lr=9.0247e-04) (hash(x)=35600311)
|
| 383 |
+
3060 train 4.147911 (lr=8.9373e-04) (hash(x)=37775318)
|
| 384 |
+
3070 train 3.997110 (lr=8.8502e-04) (hash(x)=39881333)
|
| 385 |
+
3080 train 4.010068 (lr=8.7634e-04) (hash(x)=37310168)
|
| 386 |
+
3090 train 3.931794 (lr=8.6771e-04) (hash(x)=43669978)
|
| 387 |
+
3100 val loss 3.9950
|
| 388 |
+
3100 val perplexity 54.3274
|
| 389 |
+
3100 train 4.399698 (lr=8.5911e-04) (hash(x)=46356797)
|
| 390 |
+
3110 train 4.036257 (lr=8.5054e-04) (hash(x)=40814232)
|
| 391 |
+
3120 train 3.877808 (lr=8.4202e-04) (hash(x)=42504837)
|
| 392 |
+
3130 train 4.014688 (lr=8.3353e-04) (hash(x)=39693594)
|
| 393 |
+
3140 train 3.915281 (lr=8.2509e-04) (hash(x)=40277645)
|
| 394 |
+
3150 train 3.925253 (lr=8.1669e-04) (hash(x)=40988003)
|
| 395 |
+
3160 train 4.043474 (lr=8.0832e-04) (hash(x)=38875266)
|
| 396 |
+
3170 train 4.000197 (lr=8.0000e-04) (hash(x)=45232173)
|
| 397 |
+
3180 train 4.052258 (lr=7.9172e-04) (hash(x)=39213336)
|
| 398 |
+
3190 train 4.111355 (lr=7.8349e-04) (hash(x)=42118576)
|
| 399 |
+
3200 val loss 3.9734
|
| 400 |
+
3200 val perplexity 53.1662
|
| 401 |
+
3200 train 3.970403 (lr=7.7530e-04) (hash(x)=32884223)
|
| 402 |
+
3210 train 4.015232 (lr=7.6715e-04) (hash(x)=41276800)
|
| 403 |
+
3220 train 4.064292 (lr=7.5905e-04) (hash(x)=40284461)
|
| 404 |
+
3230 train 4.040367 (lr=7.5099e-04) (hash(x)=40566734)
|
| 405 |
+
3240 train 4.022342 (lr=7.4298e-04) (hash(x)=36484570)
|
| 406 |
+
3250 hella 0.2514
|
| 407 |
+
3250 train 3.995509 (lr=7.3501e-04) (hash(x)=41642338)
|
| 408 |
+
3260 train 3.957680 (lr=7.2709e-04) (hash(x)=43883570)
|
| 409 |
+
3270 train 3.902362 (lr=7.1922e-04) (hash(x)=40432560)
|
| 410 |
+
3280 train 3.866064 (lr=7.1140e-04) (hash(x)=38002717)
|
| 411 |
+
3290 train 3.816070 (lr=7.0363e-04) (hash(x)=41926004)
|
| 412 |
+
3300 val loss 3.9663
|
| 413 |
+
3300 val perplexity 52.7862
|
| 414 |
+
3300 train 3.969190 (lr=6.9590e-04) (hash(x)=46015509)
|
| 415 |
+
3310 train 4.003609 (lr=6.8823e-04) (hash(x)=40112249)
|
| 416 |
+
3320 train 3.982442 (lr=6.8061e-04) (hash(x)=49162296)
|
| 417 |
+
3330 train 4.038400 (lr=6.7303e-04) (hash(x)=46606969)
|
| 418 |
+
3340 train 4.053986 (lr=6.6551e-04) (hash(x)=31128992)
|
| 419 |
+
3350 train 4.030886 (lr=6.5805e-04) (hash(x)=41232534)
|
| 420 |
+
3360 train 3.874255 (lr=6.5063e-04) (hash(x)=41599699)
|
| 421 |
+
3370 train 4.047343 (lr=6.4327e-04) (hash(x)=40885280)
|
| 422 |
+
3380 train 3.961383 (lr=6.3596e-04) (hash(x)=37169148)
|
| 423 |
+
3390 train 3.867796 (lr=6.2871e-04) (hash(x)=46581889)
|
| 424 |
+
3400 val loss 3.9530
|
| 425 |
+
3400 val perplexity 52.0920
|
| 426 |
+
3400 train 3.920844 (lr=6.2151e-04) (hash(x)=37612074)
|
| 427 |
+
3410 train 3.991148 (lr=6.1437e-04) (hash(x)=38763316)
|
| 428 |
+
3420 train 3.808919 (lr=6.0728e-04) (hash(x)=40562379)
|
| 429 |
+
3430 train 3.922320 (lr=6.0025e-04) (hash(x)=44724867)
|
| 430 |
+
3440 train 4.121221 (lr=5.9328e-04) (hash(x)=39386624)
|
| 431 |
+
3450 train 4.070877 (lr=5.8636e-04) (hash(x)=38652923)
|
| 432 |
+
3460 train 4.051717 (lr=5.7950e-04) (hash(x)=29587379)
|
| 433 |
+
3470 train 3.955649 (lr=5.7271e-04) (hash(x)=38228776)
|
| 434 |
+
3480 train 3.934286 (lr=5.6597e-04) (hash(x)=30076039)
|
| 435 |
+
3490 train 3.995536 (lr=5.5929e-04) (hash(x)=38398908)
|
| 436 |
+
3500 val loss 3.9405
|
| 437 |
+
3500 val perplexity 51.4445
|
| 438 |
+
3500 hella 0.2535
|
| 439 |
+
3500 train 4.004880 (lr=5.5267e-04) (hash(x)=39259918)
|
| 440 |
+
3510 train 3.904298 (lr=5.4611e-04) (hash(x)=40664091)
|
| 441 |
+
3520 train 3.938787 (lr=5.3961e-04) (hash(x)=40309647)
|
| 442 |
+
3530 train 3.909218 (lr=5.3317e-04) (hash(x)=38424801)
|
| 443 |
+
3540 train 3.829846 (lr=5.2680e-04) (hash(x)=51322307)
|
| 444 |
+
3550 train 3.893046 (lr=5.2048e-04) (hash(x)=38192628)
|
| 445 |
+
3560 train 3.944994 (lr=5.1424e-04) (hash(x)=41563952)
|
| 446 |
+
3570 train 3.894393 (lr=5.0805e-04) (hash(x)=39508843)
|
| 447 |
+
3580 train 3.964689 (lr=5.0193e-04) (hash(x)=41260225)
|
| 448 |
+
3590 train 3.950673 (lr=4.9587e-04) (hash(x)=49098107)
|
| 449 |
+
3600 val loss 3.9305
|
| 450 |
+
3600 val perplexity 50.9329
|
| 451 |
+
3600 train 3.930841 (lr=4.8988e-04) (hash(x)=41194370)
|
| 452 |
+
3610 train 4.049385 (lr=4.8395e-04) (hash(x)=40399152)
|
| 453 |
+
3620 train 4.012703 (lr=4.7809e-04) (hash(x)=41625018)
|
| 454 |
+
3630 train 4.036594 (lr=4.7229e-04) (hash(x)=44077942)
|
| 455 |
+
3640 train 4.037304 (lr=4.6656e-04) (hash(x)=42427512)
|
| 456 |
+
3650 train 4.014557 (lr=4.6090e-04) (hash(x)=40067454)
|
| 457 |
+
3660 train 3.938805 (lr=4.5530e-04) (hash(x)=37969892)
|
| 458 |
+
3670 train 3.910912 (lr=4.4978e-04) (hash(x)=43670246)
|
| 459 |
+
3680 train 3.990559 (lr=4.4432e-04) (hash(x)=37031732)
|
| 460 |
+
3690 train 3.865796 (lr=4.3893e-04) (hash(x)=39237187)
|
| 461 |
+
3700 val loss 3.9248
|
| 462 |
+
3700 val perplexity 50.6411
|
| 463 |
+
3700 train 3.908735 (lr=4.3360e-04) (hash(x)=39521416)
|
| 464 |
+
3710 train 3.807847 (lr=4.2835e-04) (hash(x)=41147425)
|
| 465 |
+
3720 train 3.799598 (lr=4.2317e-04) (hash(x)=40204658)
|
| 466 |
+
3730 train 3.892113 (lr=4.1806e-04) (hash(x)=39420983)
|
| 467 |
+
3740 train 3.938168 (lr=4.1302e-04) (hash(x)=51155740)
|
| 468 |
+
3750 hella 0.2523
|
| 469 |
+
3750 train 3.928628 (lr=4.0804e-04) (hash(x)=41654187)
|
| 470 |
+
3760 train 3.885915 (lr=4.0314e-04) (hash(x)=40598339)
|
| 471 |
+
3770 train 3.845315 (lr=3.9832e-04) (hash(x)=44950116)
|
| 472 |
+
3780 train 3.907977 (lr=3.9356e-04) (hash(x)=40660326)
|
| 473 |
+
3790 train 3.950784 (lr=3.8888e-04) (hash(x)=37805851)
|
| 474 |
+
3800 val loss 3.9071
|
| 475 |
+
3800 val perplexity 49.7562
|
| 476 |
+
3800 train 3.924257 (lr=3.8426e-04) (hash(x)=38064443)
|
| 477 |
+
3810 train 3.995887 (lr=3.7973e-04) (hash(x)=39161498)
|
| 478 |
+
3820 train 3.842611 (lr=3.7526e-04) (hash(x)=40626649)
|
| 479 |
+
3830 train 3.915209 (lr=3.7087e-04) (hash(x)=36894771)
|
| 480 |
+
3840 train 3.858629 (lr=3.6655e-04) (hash(x)=33672141)
|
| 481 |
+
3850 train 3.817992 (lr=3.6231e-04) (hash(x)=41751813)
|
| 482 |
+
3860 train 3.688847 (lr=3.5814e-04) (hash(x)=33784172)
|
| 483 |
+
3870 train 3.617184 (lr=3.5405e-04) (hash(x)=43110493)
|
| 484 |
+
3880 train 3.719736 (lr=3.5003e-04) (hash(x)=36585111)
|
| 485 |
+
3890 train 3.725897 (lr=3.4609e-04) (hash(x)=40134264)
|
| 486 |
+
3900 val loss 3.9159
|
| 487 |
+
3900 val perplexity 50.1963
|
| 488 |
+
3900 train 3.718673 (lr=3.4222e-04) (hash(x)=39815215)
|
| 489 |
+
3910 train 3.857253 (lr=3.3843e-04) (hash(x)=42081557)
|
| 490 |
+
3920 train 3.793778 (lr=3.3471e-04) (hash(x)=48184973)
|
| 491 |
+
3930 train 3.991498 (lr=3.3108e-04) (hash(x)=42416681)
|
| 492 |
+
3940 train 3.911790 (lr=3.2752e-04) (hash(x)=38587062)
|
| 493 |
+
3950 train 3.913422 (lr=3.2403e-04) (hash(x)=37296617)
|
| 494 |
+
3960 train 3.887066 (lr=3.2063e-04) (hash(x)=37959330)
|
| 495 |
+
3970 train 3.914000 (lr=3.1730e-04) (hash(x)=35175634)
|
| 496 |
+
3980 train 3.963420 (lr=3.1405e-04) (hash(x)=37219362)
|
| 497 |
+
3990 train 3.981884 (lr=3.1087e-04) (hash(x)=38906132)
|
| 498 |
+
4000 val loss 3.8949
|
| 499 |
+
4000 val perplexity 49.1488
|
| 500 |
+
4000 hella 0.2559
|
| 501 |
+
4000 train 3.904455 (lr=3.0778e-04) (hash(x)=39940517)
|
| 502 |
+
4010 train 3.916798 (lr=3.0476e-04) (hash(x)=39952437)
|
| 503 |
+
4020 train 3.836887 (lr=3.0183e-04) (hash(x)=42000930)
|
| 504 |
+
4030 train 3.845025 (lr=2.9897e-04) (hash(x)=41460672)
|
| 505 |
+
4040 train 3.591413 (lr=2.9619e-04) (hash(x)=43702329)
|
| 506 |
+
4050 train 3.684855 (lr=2.9349e-04) (hash(x)=41740121)
|
| 507 |
+
4060 train 3.686849 (lr=2.9087e-04) (hash(x)=37066349)
|
| 508 |
+
4070 train 3.526846 (lr=2.8833e-04) (hash(x)=38359869)
|
| 509 |
+
4080 train 3.771048 (lr=2.8587e-04) (hash(x)=35131370)
|
| 510 |
+
4090 train 3.829440 (lr=2.8350e-04) (hash(x)=39918811)
|
| 511 |
+
4100 val loss 3.8934
|
| 512 |
+
4100 val perplexity 49.0797
|
| 513 |
+
4100 train 3.991030 (lr=2.8120e-04) (hash(x)=47036374)
|
| 514 |
+
4110 train 3.915308 (lr=2.7898e-04) (hash(x)=37657523)
|
| 515 |
+
4120 train 3.958159 (lr=2.7684e-04) (hash(x)=37218703)
|
| 516 |
+
4130 train 3.975321 (lr=2.7479e-04) (hash(x)=41350513)
|
| 517 |
+
4140 train 3.998177 (lr=2.7281e-04) (hash(x)=38283785)
|
| 518 |
+
4150 train 3.977302 (lr=2.7092e-04) (hash(x)=32927892)
|
| 519 |
+
4160 train 3.858976 (lr=2.6910e-04) (hash(x)=42011933)
|
| 520 |
+
4170 train 3.876726 (lr=2.6737e-04) (hash(x)=41854594)
|
| 521 |
+
4180 train 3.885351 (lr=2.6572e-04) (hash(x)=36737064)
|
| 522 |
+
4190 train 3.798674 (lr=2.6415e-04) (hash(x)=36326176)
|
| 523 |
+
4200 val loss 3.8857
|
| 524 |
+
4200 val perplexity 48.7007
|
| 525 |
+
4200 train 3.871173 (lr=2.6267e-04) (hash(x)=39106683)
|
| 526 |
+
4210 train 3.921570 (lr=2.6126e-04) (hash(x)=39759977)
|
| 527 |
+
4220 train 3.662777 (lr=2.5994e-04) (hash(x)=43666151)
|
| 528 |
+
4230 train 3.609470 (lr=2.5870e-04) (hash(x)=38154367)
|
| 529 |
+
4240 train 3.722245 (lr=2.5754e-04) (hash(x)=38284168)
|
| 530 |
+
4250 hella 0.2553
|
| 531 |
+
4250 train 3.655302 (lr=2.5647e-04) (hash(x)=38888879)
|
| 532 |
+
4260 train 3.573826 (lr=2.5548e-04) (hash(x)=36846417)
|
| 533 |
+
4270 train 3.977118 (lr=2.5457e-04) (hash(x)=43125375)
|
| 534 |
+
4280 train 3.931916 (lr=2.5374e-04) (hash(x)=43556412)
|
| 535 |
+
4290 train 3.862770 (lr=2.5299e-04) (hash(x)=48598302)
|
| 536 |
+
4300 val loss 3.8808
|
| 537 |
+
4300 val perplexity 48.4646
|
| 538 |
+
4300 train 3.847933 (lr=2.5233e-04) (hash(x)=42764385)
|
| 539 |
+
4310 train 3.902875 (lr=2.5175e-04) (hash(x)=41190028)
|
| 540 |
+
4320 train 3.869807 (lr=2.5125e-04) (hash(x)=35767770)
|
| 541 |
+
4330 train 3.944903 (lr=2.5084e-04) (hash(x)=36760476)
|
| 542 |
+
4340 train 3.886782 (lr=2.5051e-04) (hash(x)=40618921)
|
| 543 |
+
4350 train 3.987609 (lr=2.5026e-04) (hash(x)=32300164)
|
| 544 |
+
4360 train 3.842880 (lr=2.5009e-04) (hash(x)=41400085)
|
| 545 |
+
4370 train 3.883012 (lr=2.5001e-04) (hash(x)=40251828)
|
| 546 |
+
4374 val loss 3.8809
|
| 547 |
+
4374 val perplexity 48.4666
|
| 548 |
+
4374 hella 0.2523
|
logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/model_04374.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:377ceefc8342ebafee0454bd034e1b4f9b6ddba2975918dc37c67165b84a0b43
|
| 3 |
+
size 97707314
|
logs/fix_1_latent_mask/1_latent_mask_lr_25e-4_n_latent_masks_2_no_compile_seed_1340/optimizer_04374.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6226fd2a11fa5338dd33c7b1a59f9a191070413b84ecd3db47cf2ce0e6671941
|
| 3 |
+
size 189136950
|