Upload config.json with huggingface_hub
Browse files- config.json +48 -0
config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "bigru_t_version",
|
| 3 |
+
"arch": "BiGRU_T_version (4 lemas)",
|
| 4 |
+
"vocab_size": 16384,
|
| 5 |
+
"d_model": 64,
|
| 6 |
+
"max_seq_len": 32,
|
| 7 |
+
"pad_token_id": 1,
|
| 8 |
+
"input_dim": 64,
|
| 9 |
+
"bigru_hidden": 16,
|
| 10 |
+
"d_transformer": 32,
|
| 11 |
+
"nhead_tu": 2,
|
| 12 |
+
"d_ff_tu": 64,
|
| 13 |
+
"output_dim_u8cell": 32,
|
| 14 |
+
"max_modules": 4,
|
| 15 |
+
"lambda_ent": 0.01,
|
| 16 |
+
"cache_len": 8,
|
| 17 |
+
"d_cache": 64,
|
| 18 |
+
"nhead_orq": 2,
|
| 19 |
+
"d_ff_orq": 128,
|
| 20 |
+
"trainT_dim": 64,
|
| 21 |
+
"nhead_train": 2,
|
| 22 |
+
"d_ff_train": 128,
|
| 23 |
+
"num_layers_train": 1,
|
| 24 |
+
"hypT_dim": 64,
|
| 25 |
+
"nhead_hyp": 2,
|
| 26 |
+
"d_ff_hyp": 128,
|
| 27 |
+
"num_layers_hyp": 1,
|
| 28 |
+
"num_bits": 8,
|
| 29 |
+
"dropout": 0.1,
|
| 30 |
+
"total_params": 4045188,
|
| 31 |
+
"total_params_M": 4.045188,
|
| 32 |
+
"training": {
|
| 33 |
+
"version": "BiGRU_T_version",
|
| 34 |
+
"epochs": 2,
|
| 35 |
+
"global_step": 7,
|
| 36 |
+
"best_loss": 7.62284517288208,
|
| 37 |
+
"final_loss": 7.702183723449707,
|
| 38 |
+
"final_avg_loss": 7.6625144481658936,
|
| 39 |
+
"final_perplexity": 2127.09919063071,
|
| 40 |
+
"killed": false,
|
| 41 |
+
"kill_reason": null,
|
| 42 |
+
"elapsed_s": 45.323012351989746,
|
| 43 |
+
"datasets": [
|
| 44 |
+
"CEIA-POSITIVO/ultrachat_br_clustred_balanced_v1"
|
| 45 |
+
],
|
| 46 |
+
"max_samples_per_dataset": 8
|
| 47 |
+
}
|
| 48 |
+
}
|