Delete experiments/ib80climb20-d12-1ep-26sh-r12
#1
by osvoorhe - opened
experiments/ib80climb20-d12-1ep-26sh-r12/config.json
DELETED
|
@@ -1,99 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"schema_version": 1,
|
| 3 |
-
"stage": "base",
|
| 4 |
-
"experiment_id": "ib80climb20-d12-1ep-26sh-r12",
|
| 5 |
-
"datasets": [
|
| 6 |
-
{
|
| 7 |
-
"name": "institutional-books",
|
| 8 |
-
"repo": "jbduran/think-dataset",
|
| 9 |
-
"revision": "main",
|
| 10 |
-
"ratio": 0.8,
|
| 11 |
-
"train_shards": [
|
| 12 |
-
0,
|
| 13 |
-
1,
|
| 14 |
-
2,
|
| 15 |
-
3,
|
| 16 |
-
4,
|
| 17 |
-
5,
|
| 18 |
-
6,
|
| 19 |
-
7,
|
| 20 |
-
8,
|
| 21 |
-
9,
|
| 22 |
-
10,
|
| 23 |
-
11,
|
| 24 |
-
12,
|
| 25 |
-
13,
|
| 26 |
-
14,
|
| 27 |
-
15,
|
| 28 |
-
16,
|
| 29 |
-
17,
|
| 30 |
-
18,
|
| 31 |
-
19,
|
| 32 |
-
20
|
| 33 |
-
],
|
| 34 |
-
"validation_shard": 472,
|
| 35 |
-
"download_workers": 4
|
| 36 |
-
},
|
| 37 |
-
{
|
| 38 |
-
"name": "climbmix",
|
| 39 |
-
"repo": "karpathy/climbmix-400b-shuffle",
|
| 40 |
-
"revision": "main",
|
| 41 |
-
"ratio": 0.2,
|
| 42 |
-
"train_shards": [
|
| 43 |
-
0,
|
| 44 |
-
1,
|
| 45 |
-
2,
|
| 46 |
-
3,
|
| 47 |
-
4
|
| 48 |
-
],
|
| 49 |
-
"validation_shard": 99,
|
| 50 |
-
"download_workers": 4
|
| 51 |
-
}
|
| 52 |
-
],
|
| 53 |
-
"tokenizer": {
|
| 54 |
-
"mode": "train",
|
| 55 |
-
"max_chars": 2000000000,
|
| 56 |
-
"doc_cap": 10000,
|
| 57 |
-
"vocab_size": 32768
|
| 58 |
-
},
|
| 59 |
-
"pretokenize": {
|
| 60 |
-
"enabled": true,
|
| 61 |
-
"slack": 1.03,
|
| 62 |
-
"val_tokens": 20971520,
|
| 63 |
-
"shard_tokens": 100000000,
|
| 64 |
-
"tokenizer_threads": 8
|
| 65 |
-
},
|
| 66 |
-
"training": {
|
| 67 |
-
"depth": 12,
|
| 68 |
-
"scaling_params": 110100912,
|
| 69 |
-
"target_param_data_ratio": 12.0,
|
| 70 |
-
"window_pattern": "L",
|
| 71 |
-
"device_batch_size": 16,
|
| 72 |
-
"total_batch_size": 524288,
|
| 73 |
-
"save_every": 500,
|
| 74 |
-
"eval_every": 250,
|
| 75 |
-
"eval_tokens": 2097152,
|
| 76 |
-
"core_metric_every": -1,
|
| 77 |
-
"sample_every": -1
|
| 78 |
-
},
|
| 79 |
-
"artifacts": {
|
| 80 |
-
"repo": "jbduran/think.nano"
|
| 81 |
-
},
|
| 82 |
-
"wandb": {
|
| 83 |
-
"entity": "jbduran-thinkingmachinesncsu",
|
| 84 |
-
"project": "think.nano",
|
| 85 |
-
"name": "ib80climb20-d12-1ep-26sh-r12",
|
| 86 |
-
"group": "think-d12",
|
| 87 |
-
"tags": [
|
| 88 |
-
"think-dataset",
|
| 89 |
-
"climbmix",
|
| 90 |
-
"d12",
|
| 91 |
-
"1ep",
|
| 92 |
-
"26sh",
|
| 93 |
-
"r12",
|
| 94 |
-
"mix-80-20"
|
| 95 |
-
]
|
| 96 |
-
},
|
| 97 |
-
"config_fingerprint": "b7fbe90857fedc42",
|
| 98 |
-
"artifact_path": "experiments/ib80climb20-d12-1ep-26sh-r12"
|
| 99 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
experiments/ib80climb20-d12-1ep-26sh-r12/run.json
DELETED
|
@@ -1,10 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"experiment_id": "ib80climb20-d12-1ep-26sh-r12",
|
| 3 |
-
"stage": "base",
|
| 4 |
-
"base_experiment_id": "ib80climb20-d12-1ep-26sh-r12",
|
| 5 |
-
"parent_experiment_id": null,
|
| 6 |
-
"parent_checkpoint_step": null,
|
| 7 |
-
"config_fingerprint": "b7fbe90857fedc42",
|
| 8 |
-
"wandb_run_id": "ad5162da",
|
| 9 |
-
"created_at": 1781540222
|
| 10 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|