Upload folder using huggingface_hub
Browse files- logs/64_two_latent_masks/two_latent_masks_seed_1340/args.json +1 -0
- logs/64_two_latent_masks/two_latent_masks_seed_1340/dataloader_04374.pt +3 -0
- logs/64_two_latent_masks/two_latent_masks_seed_1340/log2.txt +548 -0
- logs/64_two_latent_masks/two_latent_masks_seed_1340/model_04374.pt +3 -0
- logs/64_two_latent_masks/two_latent_masks_seed_1340/optimizer_04374.pt +3 -0
logs/64_two_latent_masks/two_latent_masks_seed_1340/args.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"hellaswag": true, "attention_kind": "selective", "log_dir": "logs/64_two_latent_masks/two_latent_masks_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": 768, "head_dim": 64, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 4375, "warmup_steps": 250, "group": "64_two_latent_masks", "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, "n_latent_masks": 2, "init_latent_masks_to_identity": true, "init_latent_masks_to_inverse": false, "latent_mask_scale": null, "latent_mask_runtime_multiplier": null, "latent_mask_sigmoid": false, "S_layernorm": false, "one_head_per_latent_mask": 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": 1024, "batch_size": 4, "total_batch_size": 131072, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "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": 0.003, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "two_latent_masks", "latent_mask_precision": "float32"}
|
logs/64_two_latent_masks/two_latent_masks_seed_1340/dataloader_04374.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:994809a5ef065b3b39ff8d94ca69f209f3eedb103954655f09fa348a0918b20e
|
| 3 |
+
size 964
|
logs/64_two_latent_masks/two_latent_masks_seed_1340/log2.txt
ADDED
|
@@ -0,0 +1,548 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
max_steps: 4375
|
| 2 |
+
0 val loss 10.9685
|
| 3 |
+
0 val perplexity 58017.3984
|
| 4 |
+
0 hella 0.2516
|
| 5 |
+
0 train 10.978451 (lr=4.1958e-06) (hash(x)=19978355)
|
| 6 |
+
10 train 9.656172 (lr=4.6154e-05) (hash(x)=24395949)
|
| 7 |
+
20 train 9.163681 (lr=8.8112e-05) (hash(x)=20203571)
|
| 8 |
+
30 train 8.454391 (lr=1.3007e-04) (hash(x)=22227211)
|
| 9 |
+
40 train 7.817407 (lr=1.7203e-04) (hash(x)=16492366)
|
| 10 |
+
50 train 7.571382 (lr=2.1399e-04) (hash(x)=17690059)
|
| 11 |
+
60 train 7.484971 (lr=2.5594e-04) (hash(x)=19389177)
|
| 12 |
+
70 train 7.411749 (lr=2.9790e-04) (hash(x)=16938017)
|
| 13 |
+
80 train 7.265919 (lr=3.3986e-04) (hash(x)=18013370)
|
| 14 |
+
90 train 7.034935 (lr=3.8182e-04) (hash(x)=22273301)
|
| 15 |
+
100 val loss 6.9428
|
| 16 |
+
100 val perplexity 1035.6261
|
| 17 |
+
100 train 6.907804 (lr=4.2378e-04) (hash(x)=21867969)
|
| 18 |
+
110 train 6.813749 (lr=4.6573e-04) (hash(x)=17559385)
|
| 19 |
+
120 train 6.718190 (lr=5.0769e-04) (hash(x)=21415593)
|
| 20 |
+
130 train 6.575587 (lr=5.4965e-04) (hash(x)=17493363)
|
| 21 |
+
140 train 6.441347 (lr=5.9161e-04) (hash(x)=19060392)
|
| 22 |
+
150 train 6.402858 (lr=6.3357e-04) (hash(x)=18669634)
|
| 23 |
+
160 train 6.256875 (lr=6.7552e-04) (hash(x)=19128903)
|
| 24 |
+
170 train 6.311938 (lr=7.1748e-04) (hash(x)=21173132)
|
| 25 |
+
180 train 6.306371 (lr=7.5944e-04) (hash(x)=21482016)
|
| 26 |
+
190 train 6.041762 (lr=8.0140e-04) (hash(x)=19274288)
|
| 27 |
+
200 val loss 6.0194
|
| 28 |
+
200 val perplexity 411.3516
|
| 29 |
+
200 train 5.982449 (lr=8.4336e-04) (hash(x)=17153997)
|
| 30 |
+
210 train 5.894338 (lr=8.8531e-04) (hash(x)=18659881)
|
| 31 |
+
220 train 5.816366 (lr=9.2727e-04) (hash(x)=13041762)
|
| 32 |
+
230 train 5.844281 (lr=9.6923e-04) (hash(x)=19089409)
|
| 33 |
+
240 train 5.906467 (lr=1.0112e-03) (hash(x)=21588907)
|
| 34 |
+
250 hella 0.2374
|
| 35 |
+
250 train 5.859643 (lr=1.0531e-03) (hash(x)=22853384)
|
| 36 |
+
260 train 5.791600 (lr=1.0951e-03) (hash(x)=21258918)
|
| 37 |
+
270 train 5.646438 (lr=1.1371e-03) (hash(x)=18064675)
|
| 38 |
+
280 train 5.632165 (lr=1.1790e-03) (hash(x)=17783654)
|
| 39 |
+
290 train 5.577308 (lr=1.2210e-03) (hash(x)=16206303)
|
| 40 |
+
300 val loss 5.5500
|
| 41 |
+
300 val perplexity 257.2322
|
| 42 |
+
300 train 5.569802 (lr=1.2629e-03) (hash(x)=23548851)
|
| 43 |
+
310 train 5.418448 (lr=1.3049e-03) (hash(x)=17695462)
|
| 44 |
+
320 train 5.482279 (lr=1.3469e-03) (hash(x)=16887671)
|
| 45 |
+
330 train 5.415515 (lr=1.3888e-03) (hash(x)=19956923)
|
| 46 |
+
340 train 5.413233 (lr=1.4308e-03) (hash(x)=16428364)
|
| 47 |
+
350 train 5.316234 (lr=1.4727e-03) (hash(x)=20437634)
|
| 48 |
+
360 train 5.270977 (lr=1.5147e-03) (hash(x)=19622578)
|
| 49 |
+
370 train 5.288632 (lr=1.5566e-03) (hash(x)=21294724)
|
| 50 |
+
380 train 5.289092 (lr=1.5986e-03) (hash(x)=18869667)
|
| 51 |
+
390 train 5.273484 (lr=1.6406e-03) (hash(x)=22256474)
|
| 52 |
+
400 val loss 5.2770
|
| 53 |
+
400 val perplexity 195.7782
|
| 54 |
+
400 train 5.348660 (lr=1.6825e-03) (hash(x)=20048534)
|
| 55 |
+
410 train 5.268780 (lr=1.7245e-03) (hash(x)=23042068)
|
| 56 |
+
420 train 5.229405 (lr=1.7664e-03) (hash(x)=19937603)
|
| 57 |
+
430 train 5.223491 (lr=1.8084e-03) (hash(x)=22150243)
|
| 58 |
+
440 train 5.191227 (lr=1.8503e-03) (hash(x)=23468794)
|
| 59 |
+
450 train 5.135961 (lr=1.8923e-03) (hash(x)=18432856)
|
| 60 |
+
460 train 5.219153 (lr=1.9343e-03) (hash(x)=21494081)
|
| 61 |
+
470 train 5.087043 (lr=1.9762e-03) (hash(x)=18494966)
|
| 62 |
+
480 train 5.003084 (lr=2.0182e-03) (hash(x)=20099891)
|
| 63 |
+
490 train 5.104172 (lr=2.0601e-03) (hash(x)=21778663)
|
| 64 |
+
500 val loss 5.0444
|
| 65 |
+
500 val perplexity 155.1553
|
| 66 |
+
500 hella 0.2438
|
| 67 |
+
500 train 4.963651 (lr=2.1021e-03) (hash(x)=21164498)
|
| 68 |
+
510 train 4.969781 (lr=2.1441e-03) (hash(x)=20617181)
|
| 69 |
+
520 train 5.012531 (lr=2.1860e-03) (hash(x)=24124812)
|
| 70 |
+
530 train 4.955841 (lr=2.2280e-03) (hash(x)=18270550)
|
| 71 |
+
540 train 4.964828 (lr=2.2699e-03) (hash(x)=17643805)
|
| 72 |
+
550 train 4.851285 (lr=2.3119e-03) (hash(x)=19633450)
|
| 73 |
+
560 train 4.850564 (lr=2.3538e-03) (hash(x)=18215455)
|
| 74 |
+
570 train 4.941423 (lr=2.3958e-03) (hash(x)=23240966)
|
| 75 |
+
580 train 4.807738 (lr=2.4378e-03) (hash(x)=20144929)
|
| 76 |
+
590 train 4.858644 (lr=2.4797e-03) (hash(x)=16296841)
|
| 77 |
+
600 val loss 4.7438
|
| 78 |
+
600 val perplexity 114.8657
|
| 79 |
+
600 train 4.777767 (lr=2.5217e-03) (hash(x)=19836338)
|
| 80 |
+
610 train 5.049328 (lr=2.5636e-03) (hash(x)=19928421)
|
| 81 |
+
620 train 4.642990 (lr=2.6056e-03) (hash(x)=17727783)
|
| 82 |
+
630 train 4.732543 (lr=2.6476e-03) (hash(x)=20940923)
|
| 83 |
+
640 train 4.662316 (lr=2.6895e-03) (hash(x)=29374503)
|
| 84 |
+
650 train 4.664038 (lr=2.7315e-03) (hash(x)=18000892)
|
| 85 |
+
660 train 4.661573 (lr=2.7734e-03) (hash(x)=20632048)
|
| 86 |
+
670 train 4.690650 (lr=2.8154e-03) (hash(x)=16874750)
|
| 87 |
+
680 train 4.773664 (lr=2.8573e-03) (hash(x)=16996745)
|
| 88 |
+
690 train 4.576404 (lr=2.8993e-03) (hash(x)=17242267)
|
| 89 |
+
700 val loss 4.5570
|
| 90 |
+
700 val perplexity 95.2968
|
| 91 |
+
700 train 4.468401 (lr=2.9413e-03) (hash(x)=4532237)
|
| 92 |
+
710 train 4.761856 (lr=2.9832e-03) (hash(x)=18455888)
|
| 93 |
+
720 train 4.531286 (lr=3.0000e-03) (hash(x)=22389104)
|
| 94 |
+
730 train 4.522931 (lr=2.9999e-03) (hash(x)=19048232)
|
| 95 |
+
740 train 4.471041 (lr=2.9997e-03) (hash(x)=21823621)
|
| 96 |
+
750 hella 0.2483
|
| 97 |
+
750 train 4.492226 (lr=2.9994e-03) (hash(x)=19834992)
|
| 98 |
+
760 train 4.385906 (lr=2.9990e-03) (hash(x)=20804217)
|
| 99 |
+
770 train 4.286895 (lr=2.9985e-03) (hash(x)=21676803)
|
| 100 |
+
780 train 4.246721 (lr=2.9979e-03) (hash(x)=20436543)
|
| 101 |
+
790 train 4.304165 (lr=2.9972e-03) (hash(x)=24049409)
|
| 102 |
+
800 val loss 4.4360
|
| 103 |
+
800 val perplexity 84.4394
|
| 104 |
+
800 train 4.166337 (lr=2.9964e-03) (hash(x)=21101828)
|
| 105 |
+
810 train 4.382170 (lr=2.9955e-03) (hash(x)=21158051)
|
| 106 |
+
820 train 4.344272 (lr=2.9945e-03) (hash(x)=17068830)
|
| 107 |
+
830 train 4.374517 (lr=2.9934e-03) (hash(x)=20708816)
|
| 108 |
+
840 train 4.442597 (lr=2.9922e-03) (hash(x)=23031486)
|
| 109 |
+
850 train 4.443664 (lr=2.9909e-03) (hash(x)=17808332)
|
| 110 |
+
860 train 4.308631 (lr=2.9896e-03) (hash(x)=19935633)
|
| 111 |
+
870 train 4.360803 (lr=2.9881e-03) (hash(x)=21546448)
|
| 112 |
+
880 train 4.381402 (lr=2.9865e-03) (hash(x)=18928060)
|
| 113 |
+
890 train 4.272405 (lr=2.9848e-03) (hash(x)=19620142)
|
| 114 |
+
900 val loss 4.3013
|
| 115 |
+
900 val perplexity 73.7923
|
| 116 |
+
900 train 4.311539 (lr=2.9830e-03) (hash(x)=15281589)
|
| 117 |
+
910 train 4.234674 (lr=2.9811e-03) (hash(x)=19313761)
|
| 118 |
+
920 train 4.237674 (lr=2.9792e-03) (hash(x)=21534098)
|
| 119 |
+
930 train 4.240538 (lr=2.9771e-03) (hash(x)=22007834)
|
| 120 |
+
940 train 4.344173 (lr=2.9749e-03) (hash(x)=19528278)
|
| 121 |
+
950 train 4.255614 (lr=2.9726e-03) (hash(x)=18826968)
|
| 122 |
+
960 train 4.223988 (lr=2.9703e-03) (hash(x)=24147820)
|
| 123 |
+
970 train 4.053637 (lr=2.9678e-03) (hash(x)=22646220)
|
| 124 |
+
980 train 4.075750 (lr=2.9652e-03) (hash(x)=21026041)
|
| 125 |
+
990 train 4.073892 (lr=2.9626e-03) (hash(x)=17067127)
|
| 126 |
+
1000 val loss 4.2388
|
| 127 |
+
1000 val perplexity 69.3262
|
| 128 |
+
1000 hella 0.2507
|
| 129 |
+
1000 train 4.404191 (lr=2.9598e-03) (hash(x)=18557650)
|
| 130 |
+
1010 train 4.279684 (lr=2.9570e-03) (hash(x)=20276417)
|
| 131 |
+
1020 train 4.316249 (lr=2.9540e-03) (hash(x)=25973944)
|
| 132 |
+
1030 train 4.323628 (lr=2.9510e-03) (hash(x)=21253298)
|
| 133 |
+
1040 train 4.170699 (lr=2.9478e-03) (hash(x)=24702365)
|
| 134 |
+
1050 train 4.250101 (lr=2.9446e-03) (hash(x)=19691634)
|
| 135 |
+
1060 train 4.189631 (lr=2.9412e-03) (hash(x)=16659615)
|
| 136 |
+
1070 train 4.156744 (lr=2.9378e-03) (hash(x)=18095563)
|
| 137 |
+
1080 train 4.150320 (lr=2.9343e-03) (hash(x)=18443383)
|
| 138 |
+
1090 train 4.173634 (lr=2.9307e-03) (hash(x)=18330349)
|
| 139 |
+
1100 val loss 4.1539
|
| 140 |
+
1100 val perplexity 63.6789
|
| 141 |
+
1100 train 4.129778 (lr=2.9270e-03) (hash(x)=19031432)
|
| 142 |
+
1110 train 4.281431 (lr=2.9231e-03) (hash(x)=27724663)
|
| 143 |
+
1120 train 4.092184 (lr=2.9192e-03) (hash(x)=20063068)
|
| 144 |
+
1130 train 4.025007 (lr=2.9152e-03) (hash(x)=18456567)
|
| 145 |
+
1140 train 4.045931 (lr=2.9112e-03) (hash(x)=21274378)
|
| 146 |
+
1150 train 4.311396 (lr=2.9070e-03) (hash(x)=20166104)
|
| 147 |
+
1160 train 3.851309 (lr=2.9027e-03) (hash(x)=19016072)
|
| 148 |
+
1170 train 3.907337 (lr=2.8983e-03) (hash(x)=20983879)
|
| 149 |
+
1180 train 3.882984 (lr=2.8939e-03) (hash(x)=22773283)
|
| 150 |
+
1190 train 4.126947 (lr=2.8893e-03) (hash(x)=19469721)
|
| 151 |
+
1200 val loss 4.0984
|
| 152 |
+
1200 val perplexity 60.2423
|
| 153 |
+
1200 train 4.136074 (lr=2.8847e-03) (hash(x)=21239510)
|
| 154 |
+
1210 train 4.058372 (lr=2.8800e-03) (hash(x)=21515094)
|
| 155 |
+
1220 train 4.040838 (lr=2.8751e-03) (hash(x)=16618399)
|
| 156 |
+
1230 train 4.152742 (lr=2.8702e-03) (hash(x)=19402906)
|
| 157 |
+
1240 train 4.222228 (lr=2.8652e-03) (hash(x)=21224237)
|
| 158 |
+
1250 hella 0.2589
|
| 159 |
+
1250 train 4.006453 (lr=2.8601e-03) (hash(x)=16261179)
|
| 160 |
+
1260 train 4.126361 (lr=2.8550e-03) (hash(x)=19045671)
|
| 161 |
+
1270 train 4.058733 (lr=2.8497e-03) (hash(x)=13880880)
|
| 162 |
+
1280 train 4.131390 (lr=2.8443e-03) (hash(x)=22126650)
|
| 163 |
+
1290 train 3.946322 (lr=2.8389e-03) (hash(x)=22005339)
|
| 164 |
+
1300 val loss 4.0249
|
| 165 |
+
1300 val perplexity 55.9738
|
| 166 |
+
1300 train 4.026751 (lr=2.8333e-03) (hash(x)=22293195)
|
| 167 |
+
1310 train 4.056719 (lr=2.8277e-03) (hash(x)=19908750)
|
| 168 |
+
1320 train 3.912369 (lr=2.8220e-03) (hash(x)=24199231)
|
| 169 |
+
1330 train 3.903561 (lr=2.8162e-03) (hash(x)=20450695)
|
| 170 |
+
1340 train 3.832526 (lr=2.8103e-03) (hash(x)=17483865)
|
| 171 |
+
1350 train 3.855026 (lr=2.8044e-03) (hash(x)=20351158)
|
| 172 |
+
1360 train 3.855265 (lr=2.7983e-03) (hash(x)=23798487)
|
| 173 |
+
1370 train 3.805084 (lr=2.7922e-03) (hash(x)=20090485)
|
| 174 |
+
1380 train 3.950567 (lr=2.7860e-03) (hash(x)=24781076)
|
| 175 |
+
1390 train 4.029517 (lr=2.7797e-03) (hash(x)=21234884)
|
| 176 |
+
1400 val loss 3.9745
|
| 177 |
+
1400 val perplexity 53.2218
|
| 178 |
+
1400 train 3.947072 (lr=2.7733e-03) (hash(x)=22226319)
|
| 179 |
+
1410 train 4.195976 (lr=2.7668e-03) (hash(x)=25309506)
|
| 180 |
+
1420 train 4.051446 (lr=2.7603e-03) (hash(x)=22351162)
|
| 181 |
+
1430 train 3.913201 (lr=2.7536e-03) (hash(x)=20164493)
|
| 182 |
+
1440 train 3.974930 (lr=2.7469e-03) (hash(x)=16571343)
|
| 183 |
+
1450 train 4.098995 (lr=2.7401e-03) (hash(x)=17493318)
|
| 184 |
+
1460 train 4.048571 (lr=2.7333e-03) (hash(x)=19324896)
|
| 185 |
+
1470 train 3.966823 (lr=2.7263e-03) (hash(x)=19804956)
|
| 186 |
+
1480 train 3.991376 (lr=2.7193e-03) (hash(x)=19628407)
|
| 187 |
+
1490 train 3.931729 (lr=2.7121e-03) (hash(x)=19770129)
|
| 188 |
+
1500 val loss 3.9456
|
| 189 |
+
1500 val perplexity 51.7050
|
| 190 |
+
1500 hella 0.2542
|
| 191 |
+
1500 train 3.921913 (lr=2.7050e-03) (hash(x)=16945773)
|
| 192 |
+
1510 train 3.943618 (lr=2.6977e-03) (hash(x)=18780367)
|
| 193 |
+
1520 train 3.772817 (lr=2.6903e-03) (hash(x)=19997660)
|
| 194 |
+
1530 train 4.048918 (lr=2.6829e-03) (hash(x)=17063727)
|
| 195 |
+
1540 train 3.955954 (lr=2.6754e-03) (hash(x)=20300139)
|
| 196 |
+
1550 train 3.847311 (lr=2.6678e-03) (hash(x)=21541648)
|
| 197 |
+
1560 train 3.920349 (lr=2.6602e-03) (hash(x)=25169762)
|
| 198 |
+
1570 train 3.907814 (lr=2.6525e-03) (hash(x)=18020155)
|
| 199 |
+
1580 train 3.842175 (lr=2.6447e-03) (hash(x)=21439415)
|
| 200 |
+
1590 train 3.863313 (lr=2.6368e-03) (hash(x)=20345952)
|
| 201 |
+
1600 val loss 3.9078
|
| 202 |
+
1600 val perplexity 49.7881
|
| 203 |
+
1600 train 3.770043 (lr=2.6289e-03) (hash(x)=19677331)
|
| 204 |
+
1610 train 3.891068 (lr=2.6208e-03) (hash(x)=23142852)
|
| 205 |
+
1620 train 3.876302 (lr=2.6128e-03) (hash(x)=18712875)
|
| 206 |
+
1630 train 3.756347 (lr=2.6046e-03) (hash(x)=17940151)
|
| 207 |
+
1640 train 3.728640 (lr=2.5964e-03) (hash(x)=18928868)
|
| 208 |
+
1650 train 3.887410 (lr=2.5881e-03) (hash(x)=21562811)
|
| 209 |
+
1660 train 3.982477 (lr=2.5797e-03) (hash(x)=19917546)
|
| 210 |
+
1670 train 3.929292 (lr=2.5713e-03) (hash(x)=19862190)
|
| 211 |
+
1680 train 3.952430 (lr=2.5628e-03) (hash(x)=19510122)
|
| 212 |
+
1690 train 3.837917 (lr=2.5542e-03) (hash(x)=18819854)
|
| 213 |
+
1700 val loss 3.8720
|
| 214 |
+
1700 val perplexity 48.0364
|
| 215 |
+
1700 train 3.821144 (lr=2.5455e-03) (hash(x)=20098069)
|
| 216 |
+
1710 train 3.829283 (lr=2.5368e-03) (hash(x)=19180845)
|
| 217 |
+
1720 train 3.883614 (lr=2.5281e-03) (hash(x)=22610840)
|
| 218 |
+
1730 train 3.880691 (lr=2.5192e-03) (hash(x)=17804171)
|
| 219 |
+
1740 train 3.842907 (lr=2.5103e-03) (hash(x)=18275747)
|
| 220 |
+
1750 hella 0.2553
|
| 221 |
+
1750 train 3.776542 (lr=2.5014e-03) (hash(x)=20884937)
|
| 222 |
+
1760 train 3.776291 (lr=2.4924e-03) (hash(x)=18783179)
|
| 223 |
+
1770 train 3.826714 (lr=2.4833e-03) (hash(x)=20225710)
|
| 224 |
+
1780 train 3.761395 (lr=2.4741e-03) (hash(x)=19910233)
|
| 225 |
+
1790 train 3.868055 (lr=2.4649e-03) (hash(x)=19722753)
|
| 226 |
+
1800 val loss 3.8462
|
| 227 |
+
1800 val perplexity 46.8146
|
| 228 |
+
1800 train 3.855605 (lr=2.4556e-03) (hash(x)=21871844)
|
| 229 |
+
1810 train 3.859301 (lr=2.4463e-03) (hash(x)=19497156)
|
| 230 |
+
1820 train 3.889256 (lr=2.4369e-03) (hash(x)=18788042)
|
| 231 |
+
1830 train 3.846570 (lr=2.4275e-03) (hash(x)=19961057)
|
| 232 |
+
1840 train 3.807300 (lr=2.4180e-03) (hash(x)=17863548)
|
| 233 |
+
1850 train 3.899056 (lr=2.4084e-03) (hash(x)=17751881)
|
| 234 |
+
1860 train 3.775853 (lr=2.3988e-03) (hash(x)=19219734)
|
| 235 |
+
1870 train 3.769555 (lr=2.3891e-03) (hash(x)=19455814)
|
| 236 |
+
1880 train 3.814949 (lr=2.3794e-03) (hash(x)=20580620)
|
| 237 |
+
1890 train 3.788155 (lr=2.3696e-03) (hash(x)=17077823)
|
| 238 |
+
1900 val loss 3.8331
|
| 239 |
+
1900 val perplexity 46.2054
|
| 240 |
+
1900 train 3.797397 (lr=2.3598e-03) (hash(x)=17322645)
|
| 241 |
+
1910 train 3.710891 (lr=2.3499e-03) (hash(x)=20740743)
|
| 242 |
+
1920 train 3.898655 (lr=2.3400e-03) (hash(x)=20599034)
|
| 243 |
+
1930 train 3.836424 (lr=2.3300e-03) (hash(x)=21358394)
|
| 244 |
+
1940 train 3.908106 (lr=2.3200e-03) (hash(x)=19899662)
|
| 245 |
+
1950 train 3.839334 (lr=2.3099e-03) (hash(x)=15623618)
|
| 246 |
+
1960 train 3.705758 (lr=2.2998e-03) (hash(x)=19094692)
|
| 247 |
+
1970 train 3.797168 (lr=2.2896e-03) (hash(x)=21521915)
|
| 248 |
+
1980 train 3.752630 (lr=2.2793e-03) (hash(x)=18396717)
|
| 249 |
+
1990 train 3.806487 (lr=2.2691e-03) (hash(x)=20176898)
|
| 250 |
+
2000 val loss 3.7904
|
| 251 |
+
2000 val perplexity 44.2749
|
| 252 |
+
2000 hella 0.2633
|
| 253 |
+
2000 train 3.754010 (lr=2.2588e-03) (hash(x)=17939830)
|
| 254 |
+
2010 train 3.803909 (lr=2.2484e-03) (hash(x)=18313312)
|
| 255 |
+
2020 train 3.707282 (lr=2.2380e-03) (hash(x)=22132416)
|
| 256 |
+
2030 train 3.743278 (lr=2.2275e-03) (hash(x)=18403620)
|
| 257 |
+
2040 train 3.642592 (lr=2.2170e-03) (hash(x)=19852232)
|
| 258 |
+
2050 train 3.648277 (lr=2.2065e-03) (hash(x)=21010646)
|
| 259 |
+
2060 train 3.834798 (lr=2.1959e-03) (hash(x)=21119229)
|
| 260 |
+
2070 train 3.902164 (lr=2.1853e-03) (hash(x)=22447456)
|
| 261 |
+
2080 train 3.778033 (lr=2.1746e-03) (hash(x)=19656298)
|
| 262 |
+
2090 train 3.838660 (lr=2.1639e-03) (hash(x)=17300349)
|
| 263 |
+
2100 val loss 3.7734
|
| 264 |
+
2100 val perplexity 43.5262
|
| 265 |
+
2100 train 3.731896 (lr=2.1532e-03) (hash(x)=19489545)
|
| 266 |
+
2110 train 3.666215 (lr=2.1424e-03) (hash(x)=18592164)
|
| 267 |
+
2120 train 3.752768 (lr=2.1316e-03) (hash(x)=18607056)
|
| 268 |
+
2130 train 3.770085 (lr=2.1208e-03) (hash(x)=18652918)
|
| 269 |
+
2140 train 3.789872 (lr=2.1099e-03) (hash(x)=21040103)
|
| 270 |
+
2150 train 3.647597 (lr=2.0990e-03) (hash(x)=16399764)
|
| 271 |
+
2160 train 3.707310 (lr=2.0881e-03) (hash(x)=17535477)
|
| 272 |
+
2170 train 3.627285 (lr=2.0771e-03) (hash(x)=19890415)
|
| 273 |
+
2180 train 3.745899 (lr=2.0661e-03) (hash(x)=16652534)
|
| 274 |
+
2190 train 3.696486 (lr=2.0550e-03) (hash(x)=22131443)
|
| 275 |
+
2200 val loss 3.7537
|
| 276 |
+
2200 val perplexity 42.6775
|
| 277 |
+
2200 train 3.801488 (lr=2.0440e-03) (hash(x)=19148364)
|
| 278 |
+
2210 train 3.879743 (lr=2.0329e-03) (hash(x)=15715527)
|
| 279 |
+
2220 train 3.723675 (lr=2.0217e-03) (hash(x)=23414699)
|
| 280 |
+
2230 train 3.796094 (lr=2.0106e-03) (hash(x)=19689870)
|
| 281 |
+
2240 train 3.805126 (lr=1.9994e-03) (hash(x)=22678814)
|
| 282 |
+
2250 hella 0.2595
|
| 283 |
+
2250 train 3.671422 (lr=1.9882e-03) (hash(x)=18976933)
|
| 284 |
+
2260 train 3.762718 (lr=1.9770e-03) (hash(x)=20397863)
|
| 285 |
+
2270 train 3.688645 (lr=1.9657e-03) (hash(x)=19308365)
|
| 286 |
+
2280 train 3.694260 (lr=1.9544e-03) (hash(x)=18078666)
|
| 287 |
+
2290 train 3.718165 (lr=1.9431e-03) (hash(x)=20419667)
|
| 288 |
+
2300 val loss 3.7214
|
| 289 |
+
2300 val perplexity 41.3235
|
| 290 |
+
2300 train 3.694875 (lr=1.9318e-03) (hash(x)=16788415)
|
| 291 |
+
2310 train 3.650484 (lr=1.9205e-03) (hash(x)=18566522)
|
| 292 |
+
2320 train 3.715640 (lr=1.9091e-03) (hash(x)=21387107)
|
| 293 |
+
2330 train 3.847509 (lr=1.8977e-03) (hash(x)=16815344)
|
| 294 |
+
2340 train 3.812802 (lr=1.8863e-03) (hash(x)=19301309)
|
| 295 |
+
2350 train 3.764314 (lr=1.8749e-03) (hash(x)=18411578)
|
| 296 |
+
2360 train 3.937851 (lr=1.8635e-03) (hash(x)=17212565)
|
| 297 |
+
2370 train 3.742077 (lr=1.8520e-03) (hash(x)=19083591)
|
| 298 |
+
2380 train 3.661600 (lr=1.8406e-03) (hash(x)=19095330)
|
| 299 |
+
2390 train 3.688589 (lr=1.8291e-03) (hash(x)=22137679)
|
| 300 |
+
2400 val loss 3.7018
|
| 301 |
+
2400 val perplexity 40.5184
|
| 302 |
+
2400 train 3.677315 (lr=1.8176e-03) (hash(x)=19988438)
|
| 303 |
+
2410 train 3.771107 (lr=1.8061e-03) (hash(x)=18845683)
|
| 304 |
+
2420 train 3.703688 (lr=1.7946e-03) (hash(x)=18942135)
|
| 305 |
+
2430 train 3.618211 (lr=1.7830e-03) (hash(x)=19844600)
|
| 306 |
+
2440 train 3.596223 (lr=1.7715e-03) (hash(x)=20037517)
|
| 307 |
+
2450 train 3.578623 (lr=1.7600e-03) (hash(x)=18071985)
|
| 308 |
+
2460 train 3.605704 (lr=1.7484e-03) (hash(x)=19498889)
|
| 309 |
+
2470 train 3.823439 (lr=1.7368e-03) (hash(x)=21020015)
|
| 310 |
+
2480 train 3.817960 (lr=1.7253e-03) (hash(x)=24541400)
|
| 311 |
+
2490 train 3.770563 (lr=1.7137e-03) (hash(x)=17117268)
|
| 312 |
+
2500 val loss 3.6837
|
| 313 |
+
2500 val perplexity 39.7928
|
| 314 |
+
2500 hella 0.2641
|
| 315 |
+
2500 train 3.708674 (lr=1.7021e-03) (hash(x)=20019265)
|
| 316 |
+
2510 train 3.794827 (lr=1.6906e-03) (hash(x)=15424606)
|
| 317 |
+
2520 train 3.762325 (lr=1.6790e-03) (hash(x)=16589718)
|
| 318 |
+
2530 train 3.667368 (lr=1.6674e-03) (hash(x)=21793520)
|
| 319 |
+
2540 train 3.634705 (lr=1.6558e-03) (hash(x)=26991001)
|
| 320 |
+
2550 train 3.580152 (lr=1.6442e-03) (hash(x)=19542562)
|
| 321 |
+
2560 train 3.690336 (lr=1.6326e-03) (hash(x)=16342890)
|
| 322 |
+
2570 train 3.541937 (lr=1.6210e-03) (hash(x)=22032122)
|
| 323 |
+
2580 train 3.536467 (lr=1.6094e-03) (hash(x)=20442648)
|
| 324 |
+
2590 train 3.500366 (lr=1.5979e-03) (hash(x)=20229881)
|
| 325 |
+
2600 val loss 3.6618
|
| 326 |
+
2600 val perplexity 38.9320
|
| 327 |
+
2600 train 3.596477 (lr=1.5863e-03) (hash(x)=20759842)
|
| 328 |
+
2610 train 3.539990 (lr=1.5747e-03) (hash(x)=18590738)
|
| 329 |
+
2620 train 3.722031 (lr=1.5632e-03) (hash(x)=20016008)
|
| 330 |
+
2630 train 3.675484 (lr=1.5516e-03) (hash(x)=20966006)
|
| 331 |
+
2640 train 3.749726 (lr=1.5400e-03) (hash(x)=20480389)
|
| 332 |
+
2650 train 3.624381 (lr=1.5285e-03) (hash(x)=17625327)
|
| 333 |
+
2660 train 3.584452 (lr=1.5170e-03) (hash(x)=17462746)
|
| 334 |
+
2670 train 3.666282 (lr=1.5054e-03) (hash(x)=24540619)
|
| 335 |
+
2680 train 3.595082 (lr=1.4939e-03) (hash(x)=18300466)
|
| 336 |
+
2690 train 3.562386 (lr=1.4824e-03) (hash(x)=20785354)
|
| 337 |
+
2700 val loss 3.6396
|
| 338 |
+
2700 val perplexity 38.0780
|
| 339 |
+
2700 train 3.664153 (lr=1.4709e-03) (hash(x)=19689301)
|
| 340 |
+
2710 train 3.563334 (lr=1.4594e-03) (hash(x)=18041572)
|
| 341 |
+
2720 train 3.585707 (lr=1.4480e-03) (hash(x)=23202500)
|
| 342 |
+
2730 train 3.555113 (lr=1.4365e-03) (hash(x)=25706624)
|
| 343 |
+
2740 train 3.567460 (lr=1.4251e-03) (hash(x)=24172973)
|
| 344 |
+
2750 hella 0.2641
|
| 345 |
+
2750 train 3.596277 (lr=1.4137e-03) (hash(x)=22379491)
|
| 346 |
+
2760 train 3.627256 (lr=1.4023e-03) (hash(x)=21701022)
|
| 347 |
+
2770 train 3.677678 (lr=1.3909e-03) (hash(x)=21951124)
|
| 348 |
+
2780 train 3.625000 (lr=1.3795e-03) (hash(x)=19525099)
|
| 349 |
+
2790 train 3.642115 (lr=1.3682e-03) (hash(x)=18035060)
|
| 350 |
+
2800 val loss 3.6178
|
| 351 |
+
2800 val perplexity 37.2555
|
| 352 |
+
2800 train 3.646967 (lr=1.3569e-03) (hash(x)=19475507)
|
| 353 |
+
2810 train 3.615546 (lr=1.3456e-03) (hash(x)=20631788)
|
| 354 |
+
2820 train 3.626665 (lr=1.3343e-03) (hash(x)=21229388)
|
| 355 |
+
2830 train 3.611261 (lr=1.3230e-03) (hash(x)=17707440)
|
| 356 |
+
2840 train 3.640990 (lr=1.3118e-03) (hash(x)=17822842)
|
| 357 |
+
2850 train 3.585284 (lr=1.3006e-03) (hash(x)=20668384)
|
| 358 |
+
2860 train 3.543253 (lr=1.2894e-03) (hash(x)=21880436)
|
| 359 |
+
2870 train 3.562316 (lr=1.2783e-03) (hash(x)=22236229)
|
| 360 |
+
2880 train 3.508169 (lr=1.2671e-03) (hash(x)=22480364)
|
| 361 |
+
2890 train 3.450295 (lr=1.2560e-03) (hash(x)=20782341)
|
| 362 |
+
2900 val loss 3.6097
|
| 363 |
+
2900 val perplexity 36.9537
|
| 364 |
+
2900 train 3.798346 (lr=1.2450e-03) (hash(x)=20786150)
|
| 365 |
+
2910 train 3.583338 (lr=1.2339e-03) (hash(x)=20842137)
|
| 366 |
+
2920 train 3.621460 (lr=1.2229e-03) (hash(x)=23939744)
|
| 367 |
+
2930 train 3.701345 (lr=1.2119e-03) (hash(x)=20246627)
|
| 368 |
+
2940 train 3.663435 (lr=1.2010e-03) (hash(x)=20386931)
|
| 369 |
+
2950 train 3.599035 (lr=1.1901e-03) (hash(x)=19517612)
|
| 370 |
+
2960 train 3.574937 (lr=1.1792e-03) (hash(x)=19643612)
|
| 371 |
+
2970 train 3.504542 (lr=1.1684e-03) (hash(x)=17735841)
|
| 372 |
+
2980 train 3.605876 (lr=1.1576e-03) (hash(x)=18179947)
|
| 373 |
+
2990 train 3.510193 (lr=1.1468e-03) (hash(x)=19998132)
|
| 374 |
+
3000 val loss 3.5815
|
| 375 |
+
3000 val perplexity 35.9277
|
| 376 |
+
3000 hella 0.2606
|
| 377 |
+
3000 train 3.471521 (lr=1.1361e-03) (hash(x)=19400108)
|
| 378 |
+
3010 train 3.454834 (lr=1.1254e-03) (hash(x)=19713177)
|
| 379 |
+
3020 train 3.468912 (lr=1.1147e-03) (hash(x)=19871710)
|
| 380 |
+
3030 train 3.431635 (lr=1.1041e-03) (hash(x)=16799044)
|
| 381 |
+
3040 train 3.935755 (lr=1.0935e-03) (hash(x)=19431657)
|
| 382 |
+
3050 train 3.703310 (lr=1.0830e-03) (hash(x)=22982891)
|
| 383 |
+
3060 train 3.728739 (lr=1.0725e-03) (hash(x)=21562494)
|
| 384 |
+
3070 train 3.626504 (lr=1.0620e-03) (hash(x)=33329781)
|
| 385 |
+
3080 train 3.637386 (lr=1.0516e-03) (hash(x)=17296692)
|
| 386 |
+
3090 train 3.650139 (lr=1.0412e-03) (hash(x)=19390072)
|
| 387 |
+
3100 val loss 3.5636
|
| 388 |
+
3100 val perplexity 35.2911
|
| 389 |
+
3100 train 3.774023 (lr=1.0309e-03) (hash(x)=15455793)
|
| 390 |
+
3110 train 3.636648 (lr=1.0207e-03) (hash(x)=20382524)
|
| 391 |
+
3120 train 3.496571 (lr=1.0104e-03) (hash(x)=21248576)
|
| 392 |
+
3130 train 3.531970 (lr=1.0002e-03) (hash(x)=20367412)
|
| 393 |
+
3140 train 3.479142 (lr=9.9011e-04) (hash(x)=17826969)
|
| 394 |
+
3150 train 3.564704 (lr=9.8002e-04) (hash(x)=21631872)
|
| 395 |
+
3160 train 3.650562 (lr=9.6999e-04) (hash(x)=19506122)
|
| 396 |
+
3170 train 3.692784 (lr=9.6000e-04) (hash(x)=21748750)
|
| 397 |
+
3180 train 3.622458 (lr=9.5007e-04) (hash(x)=20805736)
|
| 398 |
+
3190 train 3.668063 (lr=9.4019e-04) (hash(x)=19850067)
|
| 399 |
+
3200 val loss 3.5446
|
| 400 |
+
3200 val perplexity 34.6260
|
| 401 |
+
3200 train 3.589293 (lr=9.3036e-04) (hash(x)=21173706)
|
| 402 |
+
3210 train 3.593690 (lr=9.2058e-04) (hash(x)=21256910)
|
| 403 |
+
3220 train 3.645723 (lr=9.1085e-04) (hash(x)=20765073)
|
| 404 |
+
3230 train 3.608025 (lr=9.0118e-04) (hash(x)=17737503)
|
| 405 |
+
3240 train 3.598701 (lr=8.9157e-04) (hash(x)=19598782)
|
| 406 |
+
3250 hella 0.2651
|
| 407 |
+
3250 train 3.566664 (lr=8.8201e-04) (hash(x)=24566640)
|
| 408 |
+
3260 train 3.581837 (lr=8.7251e-04) (hash(x)=22002791)
|
| 409 |
+
3270 train 3.476205 (lr=8.6307e-04) (hash(x)=20295322)
|
| 410 |
+
3280 train 3.445398 (lr=8.5368e-04) (hash(x)=26247277)
|
| 411 |
+
3290 train 3.474231 (lr=8.4435e-04) (hash(x)=20140665)
|
| 412 |
+
3300 val loss 3.5332
|
| 413 |
+
3300 val perplexity 34.2324
|
| 414 |
+
3300 train 3.530115 (lr=8.3508e-04) (hash(x)=23553774)
|
| 415 |
+
3310 train 3.643489 (lr=8.2588e-04) (hash(x)=17529719)
|
| 416 |
+
3320 train 3.516503 (lr=8.1673e-04) (hash(x)=18795190)
|
| 417 |
+
3330 train 3.615374 (lr=8.0764e-04) (hash(x)=20419570)
|
| 418 |
+
3340 train 3.547208 (lr=7.9862e-04) (hash(x)=19596575)
|
| 419 |
+
3350 train 3.557554 (lr=7.8966e-04) (hash(x)=17475331)
|
| 420 |
+
3360 train 3.439014 (lr=7.8076e-04) (hash(x)=18856371)
|
| 421 |
+
3370 train 3.606588 (lr=7.7192e-04) (hash(x)=19732920)
|
| 422 |
+
3380 train 3.545567 (lr=7.6315e-04) (hash(x)=19151956)
|
| 423 |
+
3390 train 3.517526 (lr=7.5445e-04) (hash(x)=22601715)
|
| 424 |
+
3400 val loss 3.5163
|
| 425 |
+
3400 val perplexity 33.6609
|
| 426 |
+
3400 train 3.440074 (lr=7.4581e-04) (hash(x)=20446586)
|
| 427 |
+
3410 train 3.597298 (lr=7.3724e-04) (hash(x)=27373387)
|
| 428 |
+
3420 train 3.450697 (lr=7.2874e-04) (hash(x)=16894087)
|
| 429 |
+
3430 train 3.472631 (lr=7.2030e-04) (hash(x)=19147767)
|
| 430 |
+
3440 train 3.700382 (lr=7.1193e-04) (hash(x)=19394634)
|
| 431 |
+
3450 train 3.557866 (lr=7.0363e-04) (hash(x)=20804137)
|
| 432 |
+
3460 train 3.657521 (lr=6.9541e-04) (hash(x)=25730788)
|
| 433 |
+
3470 train 3.521112 (lr=6.8725e-04) (hash(x)=18471384)
|
| 434 |
+
3480 train 3.475666 (lr=6.7916e-04) (hash(x)=19884759)
|
| 435 |
+
3490 train 3.518395 (lr=6.7114e-04) (hash(x)=21527738)
|
| 436 |
+
3500 val loss 3.4991
|
| 437 |
+
3500 val perplexity 33.0859
|
| 438 |
+
3500 hella 0.2647
|
| 439 |
+
3500 train 3.606184 (lr=6.6320e-04) (hash(x)=21108301)
|
| 440 |
+
3510 train 3.511814 (lr=6.5533e-04) (hash(x)=19438613)
|
| 441 |
+
3520 train 3.512776 (lr=6.4753e-04) (hash(x)=20391609)
|
| 442 |
+
3530 train 3.409636 (lr=6.3981e-04) (hash(x)=21474987)
|
| 443 |
+
3540 train 3.458170 (lr=6.3216e-04) (hash(x)=20544629)
|
| 444 |
+
3550 train 3.520260 (lr=6.2458e-04) (hash(x)=18365326)
|
| 445 |
+
3560 train 3.474289 (lr=6.1708e-04) (hash(x)=18443516)
|
| 446 |
+
3570 train 3.516049 (lr=6.0966e-04) (hash(x)=18235257)
|
| 447 |
+
3580 train 3.538687 (lr=6.0231e-04) (hash(x)=20958034)
|
| 448 |
+
3590 train 3.519769 (lr=5.9504e-04) (hash(x)=20598224)
|
| 449 |
+
3600 val loss 3.4884
|
| 450 |
+
3600 val perplexity 32.7323
|
| 451 |
+
3600 train 3.580956 (lr=5.8785e-04) (hash(x)=18613771)
|
| 452 |
+
3610 train 3.674108 (lr=5.8074e-04) (hash(x)=19248168)
|
| 453 |
+
3620 train 3.569702 (lr=5.7370e-04) (hash(x)=20935084)
|
| 454 |
+
3630 train 3.670843 (lr=5.6675e-04) (hash(x)=20280969)
|
| 455 |
+
3640 train 3.619889 (lr=5.5987e-04) (hash(x)=20735321)
|
| 456 |
+
3650 train 3.508115 (lr=5.5308e-04) (hash(x)=20760704)
|
| 457 |
+
3660 train 3.479910 (lr=5.4636e-04) (hash(x)=19656305)
|
| 458 |
+
3670 train 3.478753 (lr=5.3973e-04) (hash(x)=19035247)
|
| 459 |
+
3680 train 3.537905 (lr=5.3318e-04) (hash(x)=17935451)
|
| 460 |
+
3690 train 3.491722 (lr=5.2671e-04) (hash(x)=22685971)
|
| 461 |
+
3700 val loss 3.4792
|
| 462 |
+
3700 val perplexity 32.4323
|
| 463 |
+
3700 train 3.492342 (lr=5.2033e-04) (hash(x)=18466220)
|
| 464 |
+
3710 train 3.350476 (lr=5.1402e-04) (hash(x)=19986945)
|
| 465 |
+
3720 train 3.424528 (lr=5.0780e-04) (hash(x)=18017736)
|
| 466 |
+
3730 train 3.472725 (lr=5.0167e-04) (hash(x)=19998802)
|
| 467 |
+
3740 train 3.564755 (lr=4.9562e-04) (hash(x)=20145468)
|
| 468 |
+
3750 hella 0.2683
|
| 469 |
+
3750 train 3.475979 (lr=4.8965e-04) (hash(x)=16764686)
|
| 470 |
+
3760 train 3.498098 (lr=4.8377e-04) (hash(x)=18833432)
|
| 471 |
+
3770 train 3.424459 (lr=4.7798e-04) (hash(x)=16701754)
|
| 472 |
+
3780 train 3.534266 (lr=4.7227e-04) (hash(x)=19608729)
|
| 473 |
+
3790 train 3.573368 (lr=4.6665e-04) (hash(x)=19966203)
|
| 474 |
+
3800 val loss 3.4601
|
| 475 |
+
3800 val perplexity 31.8196
|
| 476 |
+
3800 train 3.475248 (lr=4.6112e-04) (hash(x)=21276630)
|
| 477 |
+
3810 train 3.583542 (lr=4.5567e-04) (hash(x)=19385436)
|
| 478 |
+
3820 train 3.457036 (lr=4.5031e-04) (hash(x)=20894001)
|
| 479 |
+
3830 train 3.417060 (lr=4.4504e-04) (hash(x)=22827106)
|
| 480 |
+
3840 train 3.480549 (lr=4.3986e-04) (hash(x)=19263023)
|
| 481 |
+
3850 train 3.464409 (lr=4.3477e-04) (hash(x)=19916874)
|
| 482 |
+
3860 train 3.294808 (lr=4.2977e-04) (hash(x)=20568352)
|
| 483 |
+
3870 train 3.259605 (lr=4.2486e-04) (hash(x)=24120742)
|
| 484 |
+
3880 train 3.304085 (lr=4.2004e-04) (hash(x)=16266481)
|
| 485 |
+
3890 train 3.379620 (lr=4.1530e-04) (hash(x)=19310691)
|
| 486 |
+
3900 val loss 3.4636
|
| 487 |
+
3900 val perplexity 31.9307
|
| 488 |
+
3900 train 3.237922 (lr=4.1066e-04) (hash(x)=17735115)
|
| 489 |
+
3910 train 3.448336 (lr=4.0611e-04) (hash(x)=18175918)
|
| 490 |
+
3920 train 3.441614 (lr=4.0166e-04) (hash(x)=21662789)
|
| 491 |
+
3930 train 3.475464 (lr=3.9729e-04) (hash(x)=21034272)
|
| 492 |
+
3940 train 3.476670 (lr=3.9302e-04) (hash(x)=19724980)
|
| 493 |
+
3950 train 3.457575 (lr=3.8884e-04) (hash(x)=20066944)
|
| 494 |
+
3960 train 3.560726 (lr=3.8475e-04) (hash(x)=22065490)
|
| 495 |
+
3970 train 3.483026 (lr=3.8076e-04) (hash(x)=20065365)
|
| 496 |
+
3980 train 3.551414 (lr=3.7685e-04) (hash(x)=21419563)
|
| 497 |
+
3990 train 3.535141 (lr=3.7305e-04) (hash(x)=21180852)
|
| 498 |
+
4000 val loss 3.4428
|
| 499 |
+
4000 val perplexity 31.2731
|
| 500 |
+
4000 hella 0.2675
|
| 501 |
+
4000 train 3.519217 (lr=3.6933e-04) (hash(x)=18548544)
|
| 502 |
+
4010 train 3.510277 (lr=3.6572e-04) (hash(x)=20609113)
|
| 503 |
+
4020 train 3.398473 (lr=3.6219e-04) (hash(x)=16977034)
|
| 504 |
+
4030 train 3.426112 (lr=3.5876e-04) (hash(x)=19248645)
|
| 505 |
+
4040 train 3.130782 (lr=3.5543e-04) (hash(x)=16881719)
|
| 506 |
+
4050 train 3.204061 (lr=3.5219e-04) (hash(x)=18497713)
|
| 507 |
+
4060 train 3.148049 (lr=3.4905e-04) (hash(x)=20036068)
|
| 508 |
+
4070 train 3.120335 (lr=3.4600e-04) (hash(x)=21636163)
|
| 509 |
+
4080 train 3.326336 (lr=3.4305e-04) (hash(x)=23072887)
|
| 510 |
+
4090 train 3.436313 (lr=3.4019e-04) (hash(x)=27646813)
|
| 511 |
+
4100 val loss 3.4398
|
| 512 |
+
4100 val perplexity 31.1809
|
| 513 |
+
4100 train 3.583627 (lr=3.3744e-04) (hash(x)=20052100)
|
| 514 |
+
4110 train 3.485710 (lr=3.3477e-04) (hash(x)=21623032)
|
| 515 |
+
4120 train 3.508895 (lr=3.3221e-04) (hash(x)=17755383)
|
| 516 |
+
4130 train 3.506216 (lr=3.2974e-04) (hash(x)=20577475)
|
| 517 |
+
4140 train 3.508447 (lr=3.2737e-04) (hash(x)=17308418)
|
| 518 |
+
4150 train 3.554176 (lr=3.2510e-04) (hash(x)=20786408)
|
| 519 |
+
4160 train 3.490669 (lr=3.2292e-04) (hash(x)=20006893)
|
| 520 |
+
4170 train 3.448926 (lr=3.2085e-04) (hash(x)=20283286)
|
| 521 |
+
4180 train 3.444695 (lr=3.1887e-04) (hash(x)=17799852)
|
| 522 |
+
4190 train 3.376983 (lr=3.1699e-04) (hash(x)=20291468)
|
| 523 |
+
4200 val loss 3.4307
|
| 524 |
+
4200 val perplexity 30.8984
|
| 525 |
+
4200 train 3.427690 (lr=3.1520e-04) (hash(x)=16871586)
|
| 526 |
+
4210 train 3.497310 (lr=3.1352e-04) (hash(x)=21608030)
|
| 527 |
+
4220 train 3.219290 (lr=3.1193e-04) (hash(x)=22256819)
|
| 528 |
+
4230 train 3.291659 (lr=3.1044e-04) (hash(x)=20929071)
|
| 529 |
+
4240 train 3.307050 (lr=3.0905e-04) (hash(x)=22408188)
|
| 530 |
+
4250 hella 0.2690
|
| 531 |
+
4250 train 3.215563 (lr=3.0776e-04) (hash(x)=19862925)
|
| 532 |
+
4260 train 3.154764 (lr=3.0657e-04) (hash(x)=17765583)
|
| 533 |
+
4270 train 3.427227 (lr=3.0548e-04) (hash(x)=21260687)
|
| 534 |
+
4280 train 3.378080 (lr=3.0449e-04) (hash(x)=20333361)
|
| 535 |
+
4290 train 3.431731 (lr=3.0359e-04) (hash(x)=21242172)
|
| 536 |
+
4300 val loss 3.4236
|
| 537 |
+
4300 val perplexity 30.6789
|
| 538 |
+
4300 train 3.388053 (lr=3.0280e-04) (hash(x)=19835810)
|
| 539 |
+
4310 train 3.419990 (lr=3.0210e-04) (hash(x)=32580183)
|
| 540 |
+
4320 train 3.445674 (lr=3.0150e-04) (hash(x)=17944217)
|
| 541 |
+
4330 train 3.441934 (lr=3.0101e-04) (hash(x)=20116568)
|
| 542 |
+
4340 train 3.436782 (lr=3.0061e-04) (hash(x)=22423070)
|
| 543 |
+
4350 train 3.497629 (lr=3.0031e-04) (hash(x)=15493166)
|
| 544 |
+
4360 train 3.405102 (lr=3.0011e-04) (hash(x)=20130986)
|
| 545 |
+
4370 train 3.491760 (lr=3.0001e-04) (hash(x)=20037228)
|
| 546 |
+
4374 val loss 3.4226
|
| 547 |
+
4374 val perplexity 30.6489
|
| 548 |
+
4374 hella 0.2700
|
logs/64_two_latent_masks/two_latent_masks_seed_1340/model_04374.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2456a736f60bc6a66a4e75ee44c6836ce0f9b2300fbda6d7d8383dab8f2023cd
|
| 3 |
+
size 555249650
|
logs/64_two_latent_masks/two_latent_masks_seed_1340/optimizer_04374.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3a25d7092d56a3a30b1a603e0e6d3b53c1f08fb237f48644411c2807131c979
|
| 3 |
+
size 1009849398
|