Upload config.json with huggingface_hub
Browse files- config.json +46 -0
config.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 503,
|
| 3 |
+
"val_bpb": 0.4306600049435095,
|
| 4 |
+
"model_config": {
|
| 5 |
+
"sequence_len": 2048,
|
| 6 |
+
"vocab_size": 32768,
|
| 7 |
+
"n_layer": 18,
|
| 8 |
+
"n_head": 9,
|
| 9 |
+
"n_kv_head": 9,
|
| 10 |
+
"n_embd": 1152,
|
| 11 |
+
"window_pattern": "L",
|
| 12 |
+
"use_mla": false,
|
| 13 |
+
"kv_lora_rank": 512,
|
| 14 |
+
"qk_lora_rank": 1536,
|
| 15 |
+
"qk_nope_head_dim": 128,
|
| 16 |
+
"qk_rope_head_dim": 64,
|
| 17 |
+
"v_head_dim": 128,
|
| 18 |
+
"use_ssa": true
|
| 19 |
+
},
|
| 20 |
+
"user_config": {
|
| 21 |
+
"run": "nanochat",
|
| 22 |
+
"device_type": "",
|
| 23 |
+
"model_tag": "picochat_long",
|
| 24 |
+
"model_step": null,
|
| 25 |
+
"load_optimizer": 1,
|
| 26 |
+
"num_iterations": -1,
|
| 27 |
+
"max_seq_len": null,
|
| 28 |
+
"device_batch_size": 32,
|
| 29 |
+
"total_batch_size": null,
|
| 30 |
+
"embedding_lr": null,
|
| 31 |
+
"unembedding_lr": null,
|
| 32 |
+
"matrix_lr": null,
|
| 33 |
+
"init_lr_frac": 0.8,
|
| 34 |
+
"warmup_ratio": 0.0,
|
| 35 |
+
"warmdown_ratio": 0.5,
|
| 36 |
+
"final_lr_frac": 0.0,
|
| 37 |
+
"no_muon": false,
|
| 38 |
+
"eval_every": 200,
|
| 39 |
+
"eval_tokens": 20971520,
|
| 40 |
+
"chatcore_every": 200,
|
| 41 |
+
"chatcore_max_cat": -1,
|
| 42 |
+
"chatcore_max_sample": 24,
|
| 43 |
+
"mmlu_epochs": 3,
|
| 44 |
+
"gsm8k_epochs": 4
|
| 45 |
+
}
|
| 46 |
+
}
|