Upload folder using huggingface_hub
Browse files- config.json +57 -0
- state_step1808.pt +3 -0
config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"data": {
|
| 3 |
+
"train_files": "data/tinystories_gpt2/roneneldan_TinyStories_train_*.bin",
|
| 4 |
+
"val_files": "data/tinystories_gpt2/roneneldan_TinyStories_validation_*.bin",
|
| 5 |
+
"batch_size": 512,
|
| 6 |
+
"device_batch_size": 16
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"dim": 768,
|
| 10 |
+
"n_layers": 2,
|
| 11 |
+
"n_heads": 8,
|
| 12 |
+
"n_kv_heads": 8,
|
| 13 |
+
"vocab_size": 50257,
|
| 14 |
+
"multiple_of": 256,
|
| 15 |
+
"ffn_dim_multiplier": 4,
|
| 16 |
+
"norm_eps": 1e-5,
|
| 17 |
+
"rope_theta": 10000,
|
| 18 |
+
"use_scaled_rope": true,
|
| 19 |
+
"max_seq_len": 512,
|
| 20 |
+
"zero_init_masks": true
|
| 21 |
+
},
|
| 22 |
+
"optimizer": {
|
| 23 |
+
"default": {
|
| 24 |
+
"lr": 0.005,
|
| 25 |
+
"beta1": 0.9,
|
| 26 |
+
"beta2": 0.999,
|
| 27 |
+
"eps": 1e-8,
|
| 28 |
+
"weight_decay": 0.01
|
| 29 |
+
},
|
| 30 |
+
"masks": {
|
| 31 |
+
"lr": 0.005,
|
| 32 |
+
"beta1": 0.9,
|
| 33 |
+
"beta2": 0.999,
|
| 34 |
+
"eps": 1e-8,
|
| 35 |
+
"weight_decay": 0.01
|
| 36 |
+
},
|
| 37 |
+
"norms": {
|
| 38 |
+
"lr": 0.05,
|
| 39 |
+
"beta1": 0.9,
|
| 40 |
+
"beta2": 0.999,
|
| 41 |
+
"eps": 1e-8,
|
| 42 |
+
"weight_decay": 0.01
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"scheduler": { "warmup_steps": 0.01, "start_factor": 0.1 },
|
| 46 |
+
"gates": {},
|
| 47 |
+
"gates_zero_eps": 1e-8,
|
| 48 |
+
"seed": 0,
|
| 49 |
+
"project": "tinystories_baselines",
|
| 50 |
+
"run_id": null,
|
| 51 |
+
"logdir": "logs/tinystories",
|
| 52 |
+
"log_gradients": false,
|
| 53 |
+
"log_params": false,
|
| 54 |
+
"log_every_steps": 1,
|
| 55 |
+
"val_every_steps": 100,
|
| 56 |
+
"save_every_steps": -1
|
| 57 |
+
}
|
state_step1808.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8da6bd5f34a0a0acd0804ff200fea33e1ddd99b98b4695d60b203cf73dcbcbb3
|
| 3 |
+
size 921383641
|