Upload nova1_p2_final_s0115357_config.json with huggingface_hub
Browse files
nova1_p2_final_s0115357_config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"vocab_size": 50304,
|
| 3 |
+
"d_model": 2048,
|
| 4 |
+
"n_heads": 16,
|
| 5 |
+
"n_kv_heads": 8,
|
| 6 |
+
"n_layers": 24,
|
| 7 |
+
"ffn_mult": 2.6666666666666665,
|
| 8 |
+
"max_len": 2048,
|
| 9 |
+
"mod_every_n": 2,
|
| 10 |
+
"mod_capacity": 0.5,
|
| 11 |
+
"lr": 0.0001,
|
| 12 |
+
"min_lr": 1e-05,
|
| 13 |
+
"weight_decay": 0.1,
|
| 14 |
+
"beta1": 0.9,
|
| 15 |
+
"beta2": 0.95,
|
| 16 |
+
"grad_clip": 1.0,
|
| 17 |
+
"warmup_steps": 500,
|
| 18 |
+
"micro_batch": 4,
|
| 19 |
+
"grad_accum": 8,
|
| 20 |
+
"log_every": 10,
|
| 21 |
+
"ckpt_every": 1000,
|
| 22 |
+
"eval_every": 500,
|
| 23 |
+
"ckpt_keep": 1,
|
| 24 |
+
"seed": 42,
|
| 25 |
+
"training_phase": 2,
|
| 26 |
+
"final_step": 115357,
|
| 27 |
+
"final_epoch": 0,
|
| 28 |
+
"final_val_loss": 1.8969886243343352,
|
| 29 |
+
"phase": 2
|
| 30 |
+
}
|