Upload dense/config.json with huggingface_hub
Browse files- dense/config.json +49 -0
dense/config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_name": "AstroPT-0089.1M",
|
| 3 |
+
"time": 1740574257,
|
| 4 |
+
"log_via_wandb": true,
|
| 5 |
+
"log_emissions": false,
|
| 6 |
+
"out_dir": "logs/astropt070M_dense",
|
| 7 |
+
"eval_interval": 100,
|
| 8 |
+
"log_interval": 10,
|
| 9 |
+
"checkpoint_interval": 500,
|
| 10 |
+
"eval_iters": 100,
|
| 11 |
+
"eval_only": false,
|
| 12 |
+
"always_save_checkpoint": false,
|
| 13 |
+
"init_from": "scratch",
|
| 14 |
+
"hf_url": "smith42/galaxies",
|
| 15 |
+
"stream_hf_dataset": false,
|
| 16 |
+
"gradient_accumulation_steps": 40,
|
| 17 |
+
"batch_size": 64,
|
| 18 |
+
"spiral": true,
|
| 19 |
+
"block_size": 256,
|
| 20 |
+
"image_size": 256,
|
| 21 |
+
"num_workers": 64,
|
| 22 |
+
"n_layer": 12,
|
| 23 |
+
"n_head": 12,
|
| 24 |
+
"n_embd": 768,
|
| 25 |
+
"n_chan": 3,
|
| 26 |
+
"dropout": 0.0,
|
| 27 |
+
"patch_size": 16,
|
| 28 |
+
"bias": false,
|
| 29 |
+
"attn_type": "causal",
|
| 30 |
+
"k_ratio": 0.0,
|
| 31 |
+
"learning_rate": 0.0006,
|
| 32 |
+
"max_iters": 3500,
|
| 33 |
+
"weight_decay": 0.1,
|
| 34 |
+
"beta1": 0.9,
|
| 35 |
+
"beta2": 0.95,
|
| 36 |
+
"grad_clip": 1.0,
|
| 37 |
+
"decay_lr": true,
|
| 38 |
+
"warmup_iters": 2000,
|
| 39 |
+
"lr_decay_iters": 3300.0000000000005,
|
| 40 |
+
"min_lr": "5.9999999999999995e-05",
|
| 41 |
+
"backend": "nccl",
|
| 42 |
+
"device": "cuda",
|
| 43 |
+
"dtype": "bfloat16",
|
| 44 |
+
"compile": true,
|
| 45 |
+
"model_type": "astroPT",
|
| 46 |
+
"architectures": [
|
| 47 |
+
"astroPT"
|
| 48 |
+
]
|
| 49 |
+
}
|