Upload folder using huggingface_hub
Browse files- attention_kindselective_n_heads4_seed1341/args.json +1 -0
- attention_kindselective_n_heads4_seed1341/dataloader_02500.pt +3 -0
- attention_kindselective_n_heads4_seed1341/dataloader_05000.pt +3 -0
- attention_kindselective_n_heads4_seed1341/dataloader_07500.pt +3 -0
- attention_kindselective_n_heads4_seed1341/dataloader_09999.pt +3 -0
- attention_kindselective_n_heads4_seed1341/log2.txt +303 -0
- attention_kindselective_n_heads4_seed1341/model_02500.pt +3 -0
- attention_kindselective_n_heads4_seed1341/model_05000.pt +3 -0
- attention_kindselective_n_heads4_seed1341/model_07500.pt +3 -0
- attention_kindselective_n_heads4_seed1341/model_09999.pt +3 -0
- attention_kindselective_n_heads4_seed1341/optimizer_02500.pt +3 -0
- attention_kindselective_n_heads4_seed1341/optimizer_05000.pt +3 -0
- attention_kindselective_n_heads4_seed1341/optimizer_07500.pt +3 -0
- attention_kindselective_n_heads4_seed1341/optimizer_09999.pt +3 -0
attention_kindselective_n_heads4_seed1341/args.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"hellaswag": true, "attention_kind": "selective", "log_dir": "wider_is_better_4/attention_kindselective_n_heads4_seed1341", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 4, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 10000, "warmup_steps": 200, "group": "wider_is_better_4", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1341, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 256, "batch_size": 120, "total_batch_size": 30720, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_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": 0.00015, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "1.5e-4_30720_4_1341", "n_embd": 256}
|
attention_kindselective_n_heads4_seed1341/dataloader_02500.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db41c5e5513137877487a93451adf8ec4ed2448ab6e9471ebd5595c8e3293875
|
| 3 |
+
size 964
|
attention_kindselective_n_heads4_seed1341/dataloader_05000.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6df8947c6ab773db1947914387d3db345a84828521d3a64bae9b652e1b0a410
|
| 3 |
+
size 964
|
attention_kindselective_n_heads4_seed1341/dataloader_07500.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:169891a726a7ff746d1a7aa99f459a66d85ceb4e9f2583f790f5b8501f97b6af
|
| 3 |
+
size 964
|
attention_kindselective_n_heads4_seed1341/dataloader_09999.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e685a568a36c792ccbe7b5fcae0b9d630955e589991190bd8902836cea6a91df
|
| 3 |
+
size 964
|
attention_kindselective_n_heads4_seed1341/log2.txt
ADDED
|
@@ -0,0 +1,303 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
max_steps: 10000
|
| 2 |
+
0 val loss 11.3107
|
| 3 |
+
0 val perplexity 81693.5234
|
| 4 |
+
0 train 11.308768 (lr=7.5000e-07) (hash(x)=145079536)
|
| 5 |
+
100 val loss 8.8391
|
| 6 |
+
100 val perplexity 6898.6484
|
| 7 |
+
100 train 8.772367 (lr=7.5750e-05) (hash(x)=154745873)
|
| 8 |
+
200 val loss 7.6156
|
| 9 |
+
200 val perplexity 2029.7083
|
| 10 |
+
200 train 7.846752 (lr=1.5000e-04) (hash(x)=155800595)
|
| 11 |
+
300 val loss 7.4963
|
| 12 |
+
300 val perplexity 1801.3328
|
| 13 |
+
300 train 7.453661 (lr=1.4997e-04) (hash(x)=148595389)
|
| 14 |
+
400 val loss 7.4522
|
| 15 |
+
400 val perplexity 1723.7068
|
| 16 |
+
400 train 7.349169 (lr=1.4986e-04) (hash(x)=145606733)
|
| 17 |
+
500 val loss 7.4211
|
| 18 |
+
500 val perplexity 1670.7958
|
| 19 |
+
500 train 7.219723 (lr=1.4969e-04) (hash(x)=138221231)
|
| 20 |
+
600 val loss 7.3785
|
| 21 |
+
600 val perplexity 1601.2350
|
| 22 |
+
600 train 7.374033 (lr=1.4945e-04) (hash(x)=150367139)
|
| 23 |
+
700 val loss 7.3640
|
| 24 |
+
700 val perplexity 1578.1882
|
| 25 |
+
700 train 7.532372 (lr=1.4913e-04) (hash(x)=155579314)
|
| 26 |
+
800 val loss 7.3582
|
| 27 |
+
800 val perplexity 1569.0526
|
| 28 |
+
800 train 7.439426 (lr=1.4876e-04) (hash(x)=155747374)
|
| 29 |
+
900 val loss 7.3403
|
| 30 |
+
900 val perplexity 1541.1282
|
| 31 |
+
900 train 7.548749 (lr=1.4831e-04) (hash(x)=159334575)
|
| 32 |
+
1000 val loss 7.2672
|
| 33 |
+
1000 val perplexity 1432.5939
|
| 34 |
+
1000 train 7.151237 (lr=1.4779e-04) (hash(x)=140604760)
|
| 35 |
+
1100 val loss 7.2486
|
| 36 |
+
1100 val perplexity 1406.1088
|
| 37 |
+
1100 train 7.090343 (lr=1.4721e-04) (hash(x)=146208052)
|
| 38 |
+
1200 val loss 7.2222
|
| 39 |
+
1200 val perplexity 1369.4868
|
| 40 |
+
1200 train 7.149972 (lr=1.4656e-04) (hash(x)=148404734)
|
| 41 |
+
1300 val loss 7.1887
|
| 42 |
+
1300 val perplexity 1324.3914
|
| 43 |
+
1300 train 7.119121 (lr=1.4585e-04) (hash(x)=155681970)
|
| 44 |
+
1400 val loss 7.1846
|
| 45 |
+
1400 val perplexity 1318.9463
|
| 46 |
+
1400 train 7.138198 (lr=1.4507e-04) (hash(x)=148115934)
|
| 47 |
+
1500 val loss 7.2087
|
| 48 |
+
1500 val perplexity 1351.1703
|
| 49 |
+
1500 train 7.343921 (lr=1.4422e-04) (hash(x)=157074034)
|
| 50 |
+
1600 val loss 7.1871
|
| 51 |
+
1600 val perplexity 1322.3191
|
| 52 |
+
1600 train 6.987100 (lr=1.4332e-04) (hash(x)=137464699)
|
| 53 |
+
1700 val loss 7.1538
|
| 54 |
+
1700 val perplexity 1278.8964
|
| 55 |
+
1700 train 7.346094 (lr=1.4235e-04) (hash(x)=166955614)
|
| 56 |
+
1800 val loss 7.1954
|
| 57 |
+
1800 val perplexity 1333.2679
|
| 58 |
+
1800 train 7.040450 (lr=1.4131e-04) (hash(x)=143886042)
|
| 59 |
+
1900 val loss 7.2243
|
| 60 |
+
1900 val perplexity 1372.3500
|
| 61 |
+
1900 train 7.750473 (lr=1.4022e-04) (hash(x)=193610391)
|
| 62 |
+
2000 val loss 7.2533
|
| 63 |
+
2000 val perplexity 1412.8269
|
| 64 |
+
2000 train 7.405033 (lr=1.3907e-04) (hash(x)=163799796)
|
| 65 |
+
2100 val loss 7.2782
|
| 66 |
+
2100 val perplexity 1448.4299
|
| 67 |
+
2100 train 7.240331 (lr=1.3786e-04) (hash(x)=154107339)
|
| 68 |
+
2200 val loss 7.2263
|
| 69 |
+
2200 val perplexity 1375.1130
|
| 70 |
+
2200 train 7.216848 (lr=1.3660e-04) (hash(x)=144592844)
|
| 71 |
+
2300 val loss 7.2623
|
| 72 |
+
2300 val perplexity 1425.5057
|
| 73 |
+
2300 train 7.331630 (lr=1.3527e-04) (hash(x)=154535861)
|
| 74 |
+
2400 val loss 7.2641
|
| 75 |
+
2400 val perplexity 1428.1707
|
| 76 |
+
2400 train 7.981302 (lr=1.3390e-04) (hash(x)=204706354)
|
| 77 |
+
2500 val loss 7.2574
|
| 78 |
+
2500 val perplexity 1418.5338
|
| 79 |
+
2500 train 7.104072 (lr=1.3247e-04) (hash(x)=143783202)
|
| 80 |
+
2600 val loss 7.2129
|
| 81 |
+
2600 val perplexity 1356.7827
|
| 82 |
+
2600 train 7.102783 (lr=1.3099e-04) (hash(x)=150862210)
|
| 83 |
+
2700 val loss 7.1940
|
| 84 |
+
2700 val perplexity 1331.4440
|
| 85 |
+
2700 train 7.248954 (lr=1.2946e-04) (hash(x)=150753426)
|
| 86 |
+
2800 val loss 7.2206
|
| 87 |
+
2800 val perplexity 1367.2833
|
| 88 |
+
2800 train 7.188674 (lr=1.2788e-04) (hash(x)=147766811)
|
| 89 |
+
2900 val loss 7.1182
|
| 90 |
+
2900 val perplexity 1234.2701
|
| 91 |
+
2900 train 7.065569 (lr=1.2626e-04) (hash(x)=151177814)
|
| 92 |
+
3000 val loss 7.0958
|
| 93 |
+
3000 val perplexity 1206.8909
|
| 94 |
+
3000 train 6.964172 (lr=1.2459e-04) (hash(x)=135925327)
|
| 95 |
+
3100 val loss 7.1356
|
| 96 |
+
3100 val perplexity 1255.9456
|
| 97 |
+
3100 train 6.881985 (lr=1.2287e-04) (hash(x)=124842353)
|
| 98 |
+
3200 val loss 7.1582
|
| 99 |
+
3200 val perplexity 1284.6527
|
| 100 |
+
3200 train 7.265501 (lr=1.2112e-04) (hash(x)=160440642)
|
| 101 |
+
3300 val loss 7.1195
|
| 102 |
+
3300 val perplexity 1235.8479
|
| 103 |
+
3300 train 7.041489 (lr=1.1932e-04) (hash(x)=143017131)
|
| 104 |
+
3400 val loss 7.1233
|
| 105 |
+
3400 val perplexity 1240.5908
|
| 106 |
+
3400 train 7.341820 (lr=1.1749e-04) (hash(x)=148143474)
|
| 107 |
+
3500 val loss 7.1400
|
| 108 |
+
3500 val perplexity 1261.4679
|
| 109 |
+
3500 train 6.968966 (lr=1.1562e-04) (hash(x)=148732639)
|
| 110 |
+
3600 val loss 7.1300
|
| 111 |
+
3600 val perplexity 1248.8193
|
| 112 |
+
3600 train 7.117164 (lr=1.1372e-04) (hash(x)=146333803)
|
| 113 |
+
3700 val loss 7.1258
|
| 114 |
+
3700 val perplexity 1243.6559
|
| 115 |
+
3700 train 6.980933 (lr=1.1179e-04) (hash(x)=143611445)
|
| 116 |
+
3800 val loss 7.1099
|
| 117 |
+
3800 val perplexity 1224.0386
|
| 118 |
+
3800 train 6.920319 (lr=1.0982e-04) (hash(x)=141272115)
|
| 119 |
+
3900 val loss 7.0775
|
| 120 |
+
3900 val perplexity 1185.0203
|
| 121 |
+
3900 train 7.024056 (lr=1.0783e-04) (hash(x)=149161848)
|
| 122 |
+
4000 val loss 7.0654
|
| 123 |
+
4000 val perplexity 1170.7074
|
| 124 |
+
4000 train 7.113854 (lr=1.0581e-04) (hash(x)=152036955)
|
| 125 |
+
4100 val loss 7.0592
|
| 126 |
+
4100 val perplexity 1163.5604
|
| 127 |
+
4100 train 7.113382 (lr=1.0377e-04) (hash(x)=153260968)
|
| 128 |
+
4200 val loss 7.0436
|
| 129 |
+
4200 val perplexity 1145.5483
|
| 130 |
+
4200 train 6.939513 (lr=1.0171e-04) (hash(x)=147629478)
|
| 131 |
+
4300 val loss 7.0205
|
| 132 |
+
4300 val perplexity 1119.3015
|
| 133 |
+
4300 train 7.055896 (lr=9.9622e-05) (hash(x)=161701080)
|
| 134 |
+
4400 val loss 7.0205
|
| 135 |
+
4400 val perplexity 1119.4014
|
| 136 |
+
4400 train 7.079413 (lr=9.7520e-05) (hash(x)=155290524)
|
| 137 |
+
4500 val loss 7.0365
|
| 138 |
+
4500 val perplexity 1137.4241
|
| 139 |
+
4500 train 6.863458 (lr=9.5403e-05) (hash(x)=144726593)
|
| 140 |
+
4600 val loss 7.0281
|
| 141 |
+
4600 val perplexity 1127.8313
|
| 142 |
+
4600 train 6.985920 (lr=9.3273e-05) (hash(x)=149517202)
|
| 143 |
+
4700 val loss 6.9983
|
| 144 |
+
4700 val perplexity 1094.7998
|
| 145 |
+
4700 train 7.000892 (lr=9.1132e-05) (hash(x)=139403008)
|
| 146 |
+
4800 val loss 6.9746
|
| 147 |
+
4800 val perplexity 1069.1084
|
| 148 |
+
4800 train 7.027068 (lr=8.8982e-05) (hash(x)=153387238)
|
| 149 |
+
4900 val loss 6.9510
|
| 150 |
+
4900 val perplexity 1044.2091
|
| 151 |
+
4900 train 6.737017 (lr=8.6825e-05) (hash(x)=139881569)
|
| 152 |
+
5000 val loss 6.9486
|
| 153 |
+
5000 val perplexity 1041.6897
|
| 154 |
+
5000 train 6.945867 (lr=8.4663e-05) (hash(x)=152602580)
|
| 155 |
+
5100 val loss 6.9585
|
| 156 |
+
5100 val perplexity 1052.0623
|
| 157 |
+
5100 train 7.015290 (lr=8.2500e-05) (hash(x)=150598460)
|
| 158 |
+
5200 val loss 6.9349
|
| 159 |
+
5200 val perplexity 1027.5084
|
| 160 |
+
5200 train 6.861823 (lr=8.0337e-05) (hash(x)=147660496)
|
| 161 |
+
5300 val loss 6.9434
|
| 162 |
+
5300 val perplexity 1036.3123
|
| 163 |
+
5300 train 6.717848 (lr=7.8175e-05) (hash(x)=147693848)
|
| 164 |
+
5400 val loss 6.9382
|
| 165 |
+
5400 val perplexity 1030.9615
|
| 166 |
+
5400 train 6.709671 (lr=7.6018e-05) (hash(x)=136048145)
|
| 167 |
+
5500 val loss 6.9367
|
| 168 |
+
5500 val perplexity 1029.3636
|
| 169 |
+
5500 train 6.882251 (lr=7.3868e-05) (hash(x)=140313123)
|
| 170 |
+
5600 val loss 6.9212
|
| 171 |
+
5600 val perplexity 1013.5551
|
| 172 |
+
5600 train 6.946658 (lr=7.1727e-05) (hash(x)=156997993)
|
| 173 |
+
5700 val loss 6.9164
|
| 174 |
+
5700 val perplexity 1008.6826
|
| 175 |
+
5700 train 6.920675 (lr=6.9597e-05) (hash(x)=143513880)
|
| 176 |
+
5800 val loss 6.9113
|
| 177 |
+
5800 val perplexity 1003.5991
|
| 178 |
+
5800 train 7.050395 (lr=6.7480e-05) (hash(x)=153736628)
|
| 179 |
+
5900 val loss 6.9003
|
| 180 |
+
5900 val perplexity 992.5403
|
| 181 |
+
5900 train 7.066628 (lr=6.5378e-05) (hash(x)=166385417)
|
| 182 |
+
6000 val loss 6.9046
|
| 183 |
+
6000 val perplexity 996.8475
|
| 184 |
+
6000 train 7.093999 (lr=6.3294e-05) (hash(x)=153216517)
|
| 185 |
+
6100 val loss 6.9056
|
| 186 |
+
6100 val perplexity 997.8129
|
| 187 |
+
6100 train 6.820673 (lr=6.1230e-05) (hash(x)=145387545)
|
| 188 |
+
6200 val loss 6.8980
|
| 189 |
+
6200 val perplexity 990.2825
|
| 190 |
+
6200 train 7.000422 (lr=5.9188e-05) (hash(x)=156306460)
|
| 191 |
+
6300 val loss 6.8904
|
| 192 |
+
6300 val perplexity 982.8094
|
| 193 |
+
6300 train 6.779611 (lr=5.7169e-05) (hash(x)=148610447)
|
| 194 |
+
6400 val loss 6.8853
|
| 195 |
+
6400 val perplexity 977.7564
|
| 196 |
+
6400 train 7.063712 (lr=5.5177e-05) (hash(x)=158343535)
|
| 197 |
+
6500 val loss 6.8739
|
| 198 |
+
6500 val perplexity 966.6749
|
| 199 |
+
6500 train 7.008237 (lr=5.3213e-05) (hash(x)=157742557)
|
| 200 |
+
6600 val loss 6.8612
|
| 201 |
+
6600 val perplexity 954.4707
|
| 202 |
+
6600 train 6.838261 (lr=5.1279e-05) (hash(x)=150457308)
|
| 203 |
+
6700 val loss 6.8735
|
| 204 |
+
6700 val perplexity 966.3348
|
| 205 |
+
6700 train 6.790349 (lr=4.9377e-05) (hash(x)=142426205)
|
| 206 |
+
6800 val loss 6.8622
|
| 207 |
+
6800 val perplexity 955.5117
|
| 208 |
+
6800 train 6.693820 (lr=4.7509e-05) (hash(x)=151522525)
|
| 209 |
+
6900 val loss 6.8624
|
| 210 |
+
6900 val perplexity 955.6771
|
| 211 |
+
6900 train 6.791017 (lr=4.5676e-05) (hash(x)=152029630)
|
| 212 |
+
7000 val loss 6.8435
|
| 213 |
+
7000 val perplexity 937.7366
|
| 214 |
+
7000 train 6.719049 (lr=4.3882e-05) (hash(x)=142978954)
|
| 215 |
+
7100 val loss 6.8394
|
| 216 |
+
7100 val perplexity 933.9645
|
| 217 |
+
7100 train 6.734494 (lr=4.2128e-05) (hash(x)=148064359)
|
| 218 |
+
7200 val loss 6.8358
|
| 219 |
+
7200 val perplexity 930.6148
|
| 220 |
+
7200 train 6.782277 (lr=4.0414e-05) (hash(x)=155771845)
|
| 221 |
+
7300 val loss 6.8361
|
| 222 |
+
7300 val perplexity 930.8500
|
| 223 |
+
7300 train 6.662339 (lr=3.8745e-05) (hash(x)=141639844)
|
| 224 |
+
7400 val loss 6.8391
|
| 225 |
+
7400 val perplexity 933.6702
|
| 226 |
+
7400 train 6.885146 (lr=3.7120e-05) (hash(x)=153466021)
|
| 227 |
+
7500 val loss 6.8325
|
| 228 |
+
7500 val perplexity 927.5389
|
| 229 |
+
7500 train 6.823289 (lr=3.5541e-05) (hash(x)=148843587)
|
| 230 |
+
7600 val loss 6.8233
|
| 231 |
+
7600 val perplexity 919.0372
|
| 232 |
+
7600 train 6.953442 (lr=3.4011e-05) (hash(x)=152053933)
|
| 233 |
+
7700 val loss 6.8180
|
| 234 |
+
7700 val perplexity 914.1565
|
| 235 |
+
7700 train 6.620625 (lr=3.2531e-05) (hash(x)=141564003)
|
| 236 |
+
7800 val loss 6.8160
|
| 237 |
+
7800 val perplexity 912.3558
|
| 238 |
+
7800 train 6.709721 (lr=3.1102e-05) (hash(x)=148617843)
|
| 239 |
+
7900 val loss 6.8102
|
| 240 |
+
7900 val perplexity 907.0121
|
| 241 |
+
7900 train 6.794557 (lr=2.9726e-05) (hash(x)=146417668)
|
| 242 |
+
8000 val loss 6.8158
|
| 243 |
+
8000 val perplexity 912.1326
|
| 244 |
+
8000 train 6.651595 (lr=2.8405e-05) (hash(x)=150165488)
|
| 245 |
+
8100 val loss 6.8004
|
| 246 |
+
8100 val perplexity 898.2380
|
| 247 |
+
8100 train 6.779853 (lr=2.7138e-05) (hash(x)=144818855)
|
| 248 |
+
8200 val loss 6.7972
|
| 249 |
+
8200 val perplexity 895.3443
|
| 250 |
+
8200 train 6.843230 (lr=2.5929e-05) (hash(x)=143074954)
|
| 251 |
+
8300 val loss 6.7966
|
| 252 |
+
8300 val perplexity 894.8095
|
| 253 |
+
8300 train 6.712090 (lr=2.4778e-05) (hash(x)=145172494)
|
| 254 |
+
8400 val loss 6.7914
|
| 255 |
+
8400 val perplexity 890.1903
|
| 256 |
+
8400 train 6.796282 (lr=2.3686e-05) (hash(x)=152284393)
|
| 257 |
+
8500 val loss 6.7894
|
| 258 |
+
8500 val perplexity 888.4059
|
| 259 |
+
8500 train 6.656650 (lr=2.2655e-05) (hash(x)=147044705)
|
| 260 |
+
8600 val loss 6.7893
|
| 261 |
+
8600 val perplexity 888.3284
|
| 262 |
+
8600 train 6.691913 (lr=2.1685e-05) (hash(x)=147762608)
|
| 263 |
+
8700 val loss 6.7876
|
| 264 |
+
8700 val perplexity 886.8052
|
| 265 |
+
8700 train 6.776985 (lr=2.0777e-05) (hash(x)=141994567)
|
| 266 |
+
8800 val loss 6.7824
|
| 267 |
+
8800 val perplexity 882.1815
|
| 268 |
+
8800 train 6.821270 (lr=1.9933e-05) (hash(x)=152552896)
|
| 269 |
+
8900 val loss 6.7805
|
| 270 |
+
8900 val perplexity 880.5118
|
| 271 |
+
8900 train 6.645514 (lr=1.9153e-05) (hash(x)=139714454)
|
| 272 |
+
9000 val loss 6.7779
|
| 273 |
+
9000 val perplexity 878.2654
|
| 274 |
+
9000 train 6.816576 (lr=1.8439e-05) (hash(x)=152274419)
|
| 275 |
+
9100 val loss 6.7756
|
| 276 |
+
9100 val perplexity 876.1732
|
| 277 |
+
9100 train 6.920788 (lr=1.7790e-05) (hash(x)=160536885)
|
| 278 |
+
9200 val loss 6.7732
|
| 279 |
+
9200 val perplexity 874.1129
|
| 280 |
+
9200 train 6.611001 (lr=1.7208e-05) (hash(x)=138386899)
|
| 281 |
+
9300 val loss 6.7720
|
| 282 |
+
9300 val perplexity 873.0236
|
| 283 |
+
9300 train 6.784839 (lr=1.6692e-05) (hash(x)=149957457)
|
| 284 |
+
9400 val loss 6.7659
|
| 285 |
+
9400 val perplexity 867.7105
|
| 286 |
+
9400 train 6.830824 (lr=1.6245e-05) (hash(x)=161671548)
|
| 287 |
+
9500 val loss 6.7652
|
| 288 |
+
9500 val perplexity 867.1352
|
| 289 |
+
9500 train 6.889501 (lr=1.5865e-05) (hash(x)=171857400)
|
| 290 |
+
9600 val loss 6.7605
|
| 291 |
+
9600 val perplexity 863.0999
|
| 292 |
+
9600 train 6.499973 (lr=1.5554e-05) (hash(x)=135378471)
|
| 293 |
+
9700 val loss 6.7598
|
| 294 |
+
9700 val perplexity 862.4960
|
| 295 |
+
9700 train 6.683197 (lr=1.5312e-05) (hash(x)=149514591)
|
| 296 |
+
9800 val loss 6.7579
|
| 297 |
+
9800 val perplexity 860.8607
|
| 298 |
+
9800 train 6.735991 (lr=1.5139e-05) (hash(x)=152679072)
|
| 299 |
+
9900 val loss 6.7552
|
| 300 |
+
9900 val perplexity 858.5027
|
| 301 |
+
9900 train 6.687136 (lr=1.5035e-05) (hash(x)=150415193)
|
| 302 |
+
9999 val loss 6.7556
|
| 303 |
+
9999 val perplexity 858.8172
|
attention_kindselective_n_heads4_seed1341/model_02500.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1a36134feee8c838ca59f4f4cf424e2fca8a150b9359d0b271c8fcf70e2accd
|
| 3 |
+
size 92843394
|
attention_kindselective_n_heads4_seed1341/model_05000.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e6706f1ac208ee2aa2c18be2455c671f7c4625e8295b4a3ad2f06915da9e3a6
|
| 3 |
+
size 92843394
|
attention_kindselective_n_heads4_seed1341/model_07500.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aeea8582c4be0c746ae7bfb25d547cfa9115d740fb997f5b01c137d6c55be2c4
|
| 3 |
+
size 92843394
|
attention_kindselective_n_heads4_seed1341/model_09999.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24392d1319c6233ffb98a24dbd2887950aad5ebd6d687d9b5a14b369e8b2ee0c
|
| 3 |
+
size 92843394
|
attention_kindselective_n_heads4_seed1341/optimizer_02500.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b3ee680be0c5dba7e1c48deb68b6a24e64a79ea7a8d3a868ed2524130474d23
|
| 3 |
+
size 179406214
|
attention_kindselective_n_heads4_seed1341/optimizer_05000.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b4942203e4b5d098edd126d941bf815c57f9948a4c52e565b12e9801fdf55ed
|
| 3 |
+
size 179406214
|
attention_kindselective_n_heads4_seed1341/optimizer_07500.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:961e8ef3395d436796afd9a5a48e3c262a1face942572c935f59e061c05f4621
|
| 3 |
+
size 179406214
|
attention_kindselective_n_heads4_seed1341/optimizer_09999.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3632ba34851aa4a52a98c6a9cd20d04bbf16ea432d3fccfc412c40d99e2526ef
|
| 3 |
+
size 179406214
|