Upload parameters.json
Browse files- parameters.json +49 -0
parameters.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"force": true,
|
| 3 |
+
"dump_path": "data/distilgpt2-e7",
|
| 4 |
+
"student_type": "gpt2",
|
| 5 |
+
"student_config": "distilgpt2-ja.json",
|
| 6 |
+
"student_pretrained_weights": "data/distilgpt2-e6/checkpoint.pth",
|
| 7 |
+
"teacher_type": "gpt2",
|
| 8 |
+
"teacher_name": "rinna/japanese-gpt2-medium",
|
| 9 |
+
"temperature": 2.0,
|
| 10 |
+
"alpha_ce": 5.0,
|
| 11 |
+
"alpha_mlm": 0.0,
|
| 12 |
+
"alpha_clm": 0.5,
|
| 13 |
+
"alpha_mse": 0.0,
|
| 14 |
+
"alpha_cos": 1.0,
|
| 15 |
+
"mlm": false,
|
| 16 |
+
"mlm_mask_prop": 0.15,
|
| 17 |
+
"word_mask": 0.8,
|
| 18 |
+
"word_keep": 0.1,
|
| 19 |
+
"word_rand": 0.1,
|
| 20 |
+
"mlm_smoothing": 0.7,
|
| 21 |
+
"restrict_ce_to_mask": false,
|
| 22 |
+
"freeze_pos_embs": true,
|
| 23 |
+
"freeze_token_type_embds": false,
|
| 24 |
+
"n_epoch": 2,
|
| 25 |
+
"batch_size": 16,
|
| 26 |
+
"group_by_size": false,
|
| 27 |
+
"gradient_accumulation_steps": 50,
|
| 28 |
+
"warmup_prop": 0.0,
|
| 29 |
+
"weight_decay": 0.0,
|
| 30 |
+
"learning_rate": 4.9e-05,
|
| 31 |
+
"adam_epsilon": 1e-06,
|
| 32 |
+
"max_grad_norm": 5.0,
|
| 33 |
+
"initializer_range": 0.02,
|
| 34 |
+
"fp16": false,
|
| 35 |
+
"fp16_opt_level": "O1",
|
| 36 |
+
"n_gpu": 4,
|
| 37 |
+
"local_rank": 0,
|
| 38 |
+
"seed": 56,
|
| 39 |
+
"log_interval": 500,
|
| 40 |
+
"checkpoint_interval": 4000,
|
| 41 |
+
"world_size": 4,
|
| 42 |
+
"n_gpu_per_node": 4,
|
| 43 |
+
"global_rank": 0,
|
| 44 |
+
"n_nodes": 1,
|
| 45 |
+
"node_id": 0,
|
| 46 |
+
"multi_gpu": true,
|
| 47 |
+
"is_master": true,
|
| 48 |
+
"multi_node": false
|
| 49 |
+
}
|