Upload meta_000765.json with huggingface_hub
Browse files- meta_000765.json +37 -0
meta_000765.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 765,
|
| 3 |
+
"val_bpb": 0.45069642589205333,
|
| 4 |
+
"model_config": {
|
| 5 |
+
"sequence_len": 2048,
|
| 6 |
+
"vocab_size": 65536,
|
| 7 |
+
"n_layer": 20,
|
| 8 |
+
"n_head": 10,
|
| 9 |
+
"n_kv_head": 10,
|
| 10 |
+
"n_embd": 1280
|
| 11 |
+
},
|
| 12 |
+
"user_config": {
|
| 13 |
+
"run": "nanochat1diloco-mid-normal",
|
| 14 |
+
"max_iterations": -1,
|
| 15 |
+
"dtype": "bfloat16",
|
| 16 |
+
"max_seq_len": 2048,
|
| 17 |
+
"device_batch_size": 32,
|
| 18 |
+
"unembedding_lr": 0.004,
|
| 19 |
+
"embedding_lr": 0.2,
|
| 20 |
+
"matrix_lr": 0.02,
|
| 21 |
+
"init_lr_frac": 1.0,
|
| 22 |
+
"weight_decay": 0.0,
|
| 23 |
+
"eval_every": 150,
|
| 24 |
+
"eval_tokens": 10485760,
|
| 25 |
+
"total_batch_size": 524288,
|
| 26 |
+
"use_diloco": 0,
|
| 27 |
+
"base_diloco": 1,
|
| 28 |
+
"diloco_H": 50,
|
| 29 |
+
"diloco_outer_lr": 0.8,
|
| 30 |
+
"diloco_outer_momentum": 0.9,
|
| 31 |
+
"dry_run": 0
|
| 32 |
+
},
|
| 33 |
+
"device_batch_size": 32,
|
| 34 |
+
"max_seq_len": 2048,
|
| 35 |
+
"use_diloco": 0,
|
| 36 |
+
"base_diloco": 1
|
| 37 |
+
}
|