Upload folder using huggingface_hub
Browse files
outputs/bb8-char-medium-v006b/results.json
ADDED
|
@@ -0,0 +1,157 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"experiment_name": "bb8-char-medium-v006b",
|
| 3 |
+
"config": {
|
| 4 |
+
"experiment": {
|
| 5 |
+
"objective": "character_language_model_depth_ablation",
|
| 6 |
+
"status": "completed",
|
| 7 |
+
"baseline": "bb8-char-medium-v006",
|
| 8 |
+
"change": "num_layers 4->6, keeping d_model=128"
|
| 9 |
+
},
|
| 10 |
+
"model": {
|
| 11 |
+
"d_model": 128,
|
| 12 |
+
"num_layers": 6,
|
| 13 |
+
"num_heads": 4,
|
| 14 |
+
"d_ff": 512,
|
| 15 |
+
"max_seq_len": 256,
|
| 16 |
+
"dropout": 0.1,
|
| 17 |
+
"activation": "gelu",
|
| 18 |
+
"tie_weights": true
|
| 19 |
+
},
|
| 20 |
+
"tokenizer": {
|
| 21 |
+
"type": "char",
|
| 22 |
+
"vocab_size": null
|
| 23 |
+
},
|
| 24 |
+
"training": {
|
| 25 |
+
"batch_size": 32,
|
| 26 |
+
"num_epochs": 15,
|
| 27 |
+
"lr": 0.0003,
|
| 28 |
+
"weight_decay": 0.1,
|
| 29 |
+
"max_grad_norm": 1.0,
|
| 30 |
+
"warmup_steps": 100,
|
| 31 |
+
"mixed_precision": true,
|
| 32 |
+
"log_interval": 20,
|
| 33 |
+
"eval_interval": 200,
|
| 34 |
+
"save_interval": 1000
|
| 35 |
+
},
|
| 36 |
+
"data": {
|
| 37 |
+
"seq_len": 256,
|
| 38 |
+
"stride": 128,
|
| 39 |
+
"val_fraction": 0.1
|
| 40 |
+
},
|
| 41 |
+
"generation": {
|
| 42 |
+
"max_new_tokens": 200,
|
| 43 |
+
"temperature": 0.8,
|
| 44 |
+
"top_k": 50,
|
| 45 |
+
"top_p": 0.9,
|
| 46 |
+
"repetition_penalty": 1.1
|
| 47 |
+
}
|
| 48 |
+
},
|
| 49 |
+
"tokenizer": {
|
| 50 |
+
"type": "char",
|
| 51 |
+
"vocab_size": 69
|
| 52 |
+
},
|
| 53 |
+
"model_params": 1229184,
|
| 54 |
+
"model_config": {
|
| 55 |
+
"vocab_size": 69,
|
| 56 |
+
"d_model": 128,
|
| 57 |
+
"num_layers": 6,
|
| 58 |
+
"num_heads": 4,
|
| 59 |
+
"d_ff": 512,
|
| 60 |
+
"max_seq_len": 256,
|
| 61 |
+
"dropout": 0.1,
|
| 62 |
+
"activation": "gelu",
|
| 63 |
+
"tie_weights": true
|
| 64 |
+
},
|
| 65 |
+
"training_history": {
|
| 66 |
+
"train_loss": [
|
| 67 |
+
3.0068936980500514,
|
| 68 |
+
2.4787053030364365,
|
| 69 |
+
2.423370155023069,
|
| 70 |
+
2.3880616888708,
|
| 71 |
+
2.3606254665219053,
|
| 72 |
+
2.334074797922251,
|
| 73 |
+
2.308009270259312,
|
| 74 |
+
2.2775011091816184,
|
| 75 |
+
2.245568352329488,
|
| 76 |
+
2.215775710709241,
|
| 77 |
+
2.190500055040632,
|
| 78 |
+
2.1707189939459974,
|
| 79 |
+
2.1555856247337495,
|
| 80 |
+
2.143706644797812,
|
| 81 |
+
2.134737812742895
|
| 82 |
+
],
|
| 83 |
+
"val_loss": [
|
| 84 |
+
2.521340651171548,
|
| 85 |
+
2.440242963177817,
|
| 86 |
+
2.3959703275135587,
|
| 87 |
+
2.3624447073255266,
|
| 88 |
+
2.3333136354173933,
|
| 89 |
+
2.30327901669911,
|
| 90 |
+
2.2726345658302307,
|
| 91 |
+
2.235859819820949,
|
| 92 |
+
2.20341032743454,
|
| 93 |
+
2.1785084434917996,
|
| 94 |
+
2.1539460590907504,
|
| 95 |
+
2.1395692995616367,
|
| 96 |
+
2.1260189584323337,
|
| 97 |
+
2.1169184276035855,
|
| 98 |
+
2.11043712922505
|
| 99 |
+
],
|
| 100 |
+
"perplexity": [
|
| 101 |
+
12.445270250489832,
|
| 102 |
+
11.475828607893495,
|
| 103 |
+
10.978845968575628,
|
| 104 |
+
10.616874903748347,
|
| 105 |
+
10.31205537332526,
|
| 106 |
+
10.00694164525819,
|
| 107 |
+
9.704935454290482,
|
| 108 |
+
9.354521598420563,
|
| 109 |
+
9.055844292236568,
|
| 110 |
+
8.833121328525795,
|
| 111 |
+
8.618801682758889,
|
| 112 |
+
8.49577770559538,
|
| 113 |
+
8.381433470793056,
|
| 114 |
+
8.305504000383168,
|
| 115 |
+
8.251847620107325
|
| 116 |
+
],
|
| 117 |
+
"lr": [
|
| 118 |
+
0.0002989055403767918,
|
| 119 |
+
0.00029214898909133585,
|
| 120 |
+
0.00027952140501033925,
|
| 121 |
+
0.0002616058597004563,
|
| 122 |
+
0.00023922959335302468,
|
| 123 |
+
0.00021342581743520874,
|
| 124 |
+
0.00018538600673375218,
|
| 125 |
+
0.00015640488367928905,
|
| 126 |
+
0.00012782063526898158,
|
| 127 |
+
0.00010095312303733438,
|
| 128 |
+
7.704293919490188e-05,
|
| 129 |
+
5.719412298328191e-05,
|
| 130 |
+
4.232318228636624e-05,
|
| 131 |
+
3.3116774396171685e-05,
|
| 132 |
+
2.9999999999999997e-05
|
| 133 |
+
]
|
| 134 |
+
},
|
| 135 |
+
"evaluation": {
|
| 136 |
+
"train": {
|
| 137 |
+
"loss": 2.0324661191628905,
|
| 138 |
+
"perplexity": 7.63288677556304,
|
| 139 |
+
"accuracy": 0.39493582589285714,
|
| 140 |
+
"bits_per_token": 2.932228790891139,
|
| 141 |
+
"bits_per_char": 2.932228790891139
|
| 142 |
+
},
|
| 143 |
+
"val": {
|
| 144 |
+
"loss": 2.109019522283269,
|
| 145 |
+
"perplexity": 8.240158031217348,
|
| 146 |
+
"accuracy": 0.37666576867816093,
|
| 147 |
+
"bits_per_token": 3.0426720059360832,
|
| 148 |
+
"bits_per_char": 3.0426720059360832
|
| 149 |
+
}
|
| 150 |
+
},
|
| 151 |
+
"generated_samples": {
|
| 152 |
+
"greedy": "ROMEO:\nThe that the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the\nthat the the the the the the the the the the the the the the the\nt",
|
| 153 |
+
"temperature": "ROMEO:\nThis sen raucks\nWit morhe he goos sond cand, be nowse that the us light badauns,\nThim: seelesh hear the heifer.\n\nPENIUS:\nTheace.\n\nCAULLIO:\nLe you thou notw not I your, ase bandy fort brarie\nWhis spri",
|
| 154 |
+
"top_k": "ROMEO:\nNor to thou is rube,\nThe to hand is my gone prie I hean the comel bat tearlave\nBas gasof the knound, wom ay at live thee pa scaclince\nAll thou wor of mirt beste\nAnd not to sur the is to sear fry aur ",
|
| 155 |
+
"top_p": "ROMEO:\nBy you thall this be that the alle deake allind fraly\nHerevel me thingh to no sis thatht athould of to iste,\nI is in whis ourd destring send.\n\nPOLAR:\nWhat o by cond the on was apour alt thaver'd.\nMEN"
|
| 156 |
+
}
|
| 157 |
+
}
|
outputs/bb8-char-medium-v006b/run.json
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_id": "bb8-char-medium-v006b",
|
| 3 |
+
"started_at": "2026-09-27T20:51:16.241359+00:00",
|
| 4 |
+
"completed_at": "2026-09-27T20:54:35.125897+00:00",
|
| 5 |
+
"duration_seconds": 198.88,
|
| 6 |
+
"git_commit": "5e22799e0b14096994bfe5b2f4432344ab31911b",
|
| 7 |
+
"git_dirty": true,
|
| 8 |
+
"seed": 42,
|
| 9 |
+
"config_sha256": "a1cc2d98fab2db2606502f418be3a9a07f302737be8e20c361be972b3ee90f42",
|
| 10 |
+
"experiment": {
|
| 11 |
+
"objective": "character_language_model_depth_ablation",
|
| 12 |
+
"status": "completed",
|
| 13 |
+
"baseline": "bb8-char-medium-v006",
|
| 14 |
+
"change": "num_layers 4->6, keeping d_model=128"
|
| 15 |
+
},
|
| 16 |
+
"dataset": {
|
| 17 |
+
"path": "data/tiny_shakespeare.txt",
|
| 18 |
+
"sha256": "86c4e6aa9db7c042ec79f339dcb96d42b0075e16b8fc2e86bf0ca57e2dc565ed",
|
| 19 |
+
"characters": 1115394,
|
| 20 |
+
"manifest": null
|
| 21 |
+
},
|
| 22 |
+
"tokenizer": {
|
| 23 |
+
"type": "char",
|
| 24 |
+
"vocab_size": 69
|
| 25 |
+
},
|
| 26 |
+
"model": {
|
| 27 |
+
"parameters": 1229184,
|
| 28 |
+
"vocab_size": 69,
|
| 29 |
+
"d_model": 128,
|
| 30 |
+
"num_layers": 6,
|
| 31 |
+
"num_heads": 4,
|
| 32 |
+
"d_ff": 512,
|
| 33 |
+
"max_seq_len": 256,
|
| 34 |
+
"dropout": 0.1,
|
| 35 |
+
"activation": "gelu",
|
| 36 |
+
"tie_weights": true
|
| 37 |
+
},
|
| 38 |
+
"training": {
|
| 39 |
+
"batch_size": 32,
|
| 40 |
+
"num_epochs": 15,
|
| 41 |
+
"lr": 0.0003,
|
| 42 |
+
"weight_decay": 0.1,
|
| 43 |
+
"max_grad_norm": 1.0,
|
| 44 |
+
"warmup_steps": 100,
|
| 45 |
+
"mixed_precision": true,
|
| 46 |
+
"log_interval": 20,
|
| 47 |
+
"eval_interval": 200,
|
| 48 |
+
"save_interval": 1000
|
| 49 |
+
},
|
| 50 |
+
"runtime": {
|
| 51 |
+
"device": "cuda",
|
| 52 |
+
"pytorch": "2.13.0+cu130",
|
| 53 |
+
"cuda": "13.0",
|
| 54 |
+
"gpu": "NVIDIA GeForce RTX 4070 Laptop GPU",
|
| 55 |
+
"gpu_memory_gb": 8.0
|
| 56 |
+
},
|
| 57 |
+
"evaluation": {
|
| 58 |
+
"train": {
|
| 59 |
+
"loss": 2.0324661191628905,
|
| 60 |
+
"perplexity": 7.63288677556304,
|
| 61 |
+
"accuracy": 0.39493582589285714,
|
| 62 |
+
"bits_per_token": 2.932228790891139,
|
| 63 |
+
"bits_per_char": 2.932228790891139
|
| 64 |
+
},
|
| 65 |
+
"val": {
|
| 66 |
+
"loss": 2.109019522283269,
|
| 67 |
+
"perplexity": 8.240158031217348,
|
| 68 |
+
"accuracy": 0.37666576867816093,
|
| 69 |
+
"bits_per_token": 3.0426720059360832,
|
| 70 |
+
"bits_per_char": 3.0426720059360832
|
| 71 |
+
}
|
| 72 |
+
},
|
| 73 |
+
"artifacts": {
|
| 74 |
+
"checkpoint": "checkpoints/bb8-char-medium-v006b/final_model.pt",
|
| 75 |
+
"tokenizer": "checkpoints/bb8-char-medium-v006b/tokenizer.json",
|
| 76 |
+
"config": "checkpoints/bb8-char-medium-v006b/config.yaml",
|
| 77 |
+
"results": "outputs/bb8-char-medium-v006b/results.json",
|
| 78 |
+
"output_dir": "outputs\\bb8-char-medium-v006b"
|
| 79 |
+
}
|
| 80 |
+
}
|