Stojancho Tudjarski
commited on
Commit
·
5c6c765
1
Parent(s):
e4edad2
commit from macedonizer
Browse files- config.json +25 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- scheduler.pt +3 -0
- trainer_state.json +856 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 514,
|
| 16 |
+
"model_type": "roberta",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 6,
|
| 19 |
+
"pad_token_id": 1,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"transformers_version": "4.4.2",
|
| 22 |
+
"type_vocab_size": 1,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 52000
|
| 25 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee73c73a013e5106ea75ab7b12b5bc5a769b46e652fcff7d028464dfc3b159ef
|
| 3 |
+
size 668102115
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7084e2d3d3c1ddc9ca78d02f8d1aa6212d4413882334aedf22d29a957926f67
|
| 3 |
+
size 334067506
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3260ac8e6fd861ac94b726d086012dae1d320db3cf849044d311ba5da2008b7
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,856 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.840146837977641,
|
| 5 |
+
"global_step": 70000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 4.9791423327215084e-05,
|
| 13 |
+
"loss": 8.0732,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.08,
|
| 18 |
+
"learning_rate": 4.958284665443017e-05,
|
| 19 |
+
"loss": 7.3801,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.13,
|
| 24 |
+
"learning_rate": 4.9374269981645254e-05,
|
| 25 |
+
"loss": 7.0667,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.17,
|
| 30 |
+
"learning_rate": 4.9165693308860336e-05,
|
| 31 |
+
"loss": 6.906,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.21,
|
| 36 |
+
"learning_rate": 4.8957116636075424e-05,
|
| 37 |
+
"loss": 6.7114,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.25,
|
| 42 |
+
"learning_rate": 4.8748539963290505e-05,
|
| 43 |
+
"loss": 6.6333,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.29,
|
| 48 |
+
"learning_rate": 4.8539963290505594e-05,
|
| 49 |
+
"loss": 6.4948,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.33,
|
| 54 |
+
"learning_rate": 4.8331386617720675e-05,
|
| 55 |
+
"loss": 6.3157,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.38,
|
| 60 |
+
"learning_rate": 4.812280994493576e-05,
|
| 61 |
+
"loss": 6.1185,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.42,
|
| 66 |
+
"learning_rate": 4.7914233272150845e-05,
|
| 67 |
+
"loss": 5.9335,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.46,
|
| 72 |
+
"learning_rate": 4.770565659936593e-05,
|
| 73 |
+
"loss": 5.7535,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.5,
|
| 78 |
+
"learning_rate": 4.749707992658101e-05,
|
| 79 |
+
"loss": 5.5631,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.54,
|
| 84 |
+
"learning_rate": 4.72885032537961e-05,
|
| 85 |
+
"loss": 5.3707,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.58,
|
| 90 |
+
"learning_rate": 4.707992658101118e-05,
|
| 91 |
+
"loss": 5.2234,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.63,
|
| 96 |
+
"learning_rate": 4.687134990822627e-05,
|
| 97 |
+
"loss": 5.1324,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.67,
|
| 102 |
+
"learning_rate": 4.666277323544135e-05,
|
| 103 |
+
"loss": 5.023,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.71,
|
| 108 |
+
"learning_rate": 4.645419656265643e-05,
|
| 109 |
+
"loss": 4.9164,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.75,
|
| 114 |
+
"learning_rate": 4.624561988987152e-05,
|
| 115 |
+
"loss": 4.8293,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.79,
|
| 120 |
+
"learning_rate": 4.60370432170866e-05,
|
| 121 |
+
"loss": 4.7737,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.83,
|
| 126 |
+
"learning_rate": 4.582846654430168e-05,
|
| 127 |
+
"loss": 4.6915,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.88,
|
| 132 |
+
"learning_rate": 4.561988987151678e-05,
|
| 133 |
+
"loss": 4.6345,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.92,
|
| 138 |
+
"learning_rate": 4.541131319873186e-05,
|
| 139 |
+
"loss": 4.6052,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.96,
|
| 144 |
+
"learning_rate": 4.520273652594694e-05,
|
| 145 |
+
"loss": 4.5157,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.0,
|
| 150 |
+
"learning_rate": 4.499415985316203e-05,
|
| 151 |
+
"loss": 4.4535,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.04,
|
| 156 |
+
"learning_rate": 4.478558318037711e-05,
|
| 157 |
+
"loss": 4.4238,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.08,
|
| 162 |
+
"learning_rate": 4.45770065075922e-05,
|
| 163 |
+
"loss": 4.3574,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.13,
|
| 168 |
+
"learning_rate": 4.436842983480728e-05,
|
| 169 |
+
"loss": 4.3355,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.17,
|
| 174 |
+
"learning_rate": 4.415985316202236e-05,
|
| 175 |
+
"loss": 4.3007,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.21,
|
| 180 |
+
"learning_rate": 4.395127648923745e-05,
|
| 181 |
+
"loss": 4.2482,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.25,
|
| 186 |
+
"learning_rate": 4.374269981645253e-05,
|
| 187 |
+
"loss": 4.1904,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.29,
|
| 192 |
+
"learning_rate": 4.353412314366761e-05,
|
| 193 |
+
"loss": 4.1877,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.33,
|
| 198 |
+
"learning_rate": 4.33255464708827e-05,
|
| 199 |
+
"loss": 4.1422,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.38,
|
| 204 |
+
"learning_rate": 4.311696979809778e-05,
|
| 205 |
+
"loss": 4.0952,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.42,
|
| 210 |
+
"learning_rate": 4.2908393125312865e-05,
|
| 211 |
+
"loss": 4.0837,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.46,
|
| 216 |
+
"learning_rate": 4.269981645252795e-05,
|
| 217 |
+
"loss": 4.0552,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.5,
|
| 222 |
+
"learning_rate": 4.2491239779743035e-05,
|
| 223 |
+
"loss": 4.0141,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.54,
|
| 228 |
+
"learning_rate": 4.228266310695812e-05,
|
| 229 |
+
"loss": 3.9979,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.59,
|
| 234 |
+
"learning_rate": 4.2074086434173204e-05,
|
| 235 |
+
"loss": 3.9504,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.63,
|
| 240 |
+
"learning_rate": 4.1865509761388286e-05,
|
| 241 |
+
"loss": 3.9522,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.67,
|
| 246 |
+
"learning_rate": 4.1656933088603374e-05,
|
| 247 |
+
"loss": 3.9057,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.71,
|
| 252 |
+
"learning_rate": 4.1448356415818456e-05,
|
| 253 |
+
"loss": 3.8997,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.75,
|
| 258 |
+
"learning_rate": 4.123977974303354e-05,
|
| 259 |
+
"loss": 3.8594,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.79,
|
| 264 |
+
"learning_rate": 4.1031203070248626e-05,
|
| 265 |
+
"loss": 3.823,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.84,
|
| 270 |
+
"learning_rate": 4.082262639746371e-05,
|
| 271 |
+
"loss": 3.8168,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.88,
|
| 276 |
+
"learning_rate": 4.0614049724678796e-05,
|
| 277 |
+
"loss": 3.7902,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.92,
|
| 282 |
+
"learning_rate": 4.040547305189388e-05,
|
| 283 |
+
"loss": 3.7825,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.96,
|
| 288 |
+
"learning_rate": 4.019689637910896e-05,
|
| 289 |
+
"loss": 3.7529,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.0,
|
| 294 |
+
"learning_rate": 3.998831970632405e-05,
|
| 295 |
+
"loss": 3.7309,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.04,
|
| 300 |
+
"learning_rate": 3.977974303353913e-05,
|
| 301 |
+
"loss": 3.6979,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.09,
|
| 306 |
+
"learning_rate": 3.957116636075421e-05,
|
| 307 |
+
"loss": 3.6644,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 2.13,
|
| 312 |
+
"learning_rate": 3.93625896879693e-05,
|
| 313 |
+
"loss": 3.6569,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.17,
|
| 318 |
+
"learning_rate": 3.915401301518438e-05,
|
| 319 |
+
"loss": 3.6378,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.21,
|
| 324 |
+
"learning_rate": 3.894543634239947e-05,
|
| 325 |
+
"loss": 3.6143,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.25,
|
| 330 |
+
"learning_rate": 3.873685966961455e-05,
|
| 331 |
+
"loss": 3.5909,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.29,
|
| 336 |
+
"learning_rate": 3.852828299682963e-05,
|
| 337 |
+
"loss": 3.6046,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.34,
|
| 342 |
+
"learning_rate": 3.831970632404472e-05,
|
| 343 |
+
"loss": 3.577,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.38,
|
| 348 |
+
"learning_rate": 3.81111296512598e-05,
|
| 349 |
+
"loss": 3.5381,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.42,
|
| 354 |
+
"learning_rate": 3.7902552978474884e-05,
|
| 355 |
+
"loss": 3.5436,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.46,
|
| 360 |
+
"learning_rate": 3.769397630568997e-05,
|
| 361 |
+
"loss": 3.5292,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.5,
|
| 366 |
+
"learning_rate": 3.7485399632905054e-05,
|
| 367 |
+
"loss": 3.5224,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.54,
|
| 372 |
+
"learning_rate": 3.727682296012014e-05,
|
| 373 |
+
"loss": 3.5099,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.59,
|
| 378 |
+
"learning_rate": 3.706824628733523e-05,
|
| 379 |
+
"loss": 3.4796,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.63,
|
| 384 |
+
"learning_rate": 3.685966961455031e-05,
|
| 385 |
+
"loss": 3.475,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.67,
|
| 390 |
+
"learning_rate": 3.66510929417654e-05,
|
| 391 |
+
"loss": 3.4557,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.71,
|
| 396 |
+
"learning_rate": 3.644251626898048e-05,
|
| 397 |
+
"loss": 3.4315,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.75,
|
| 402 |
+
"learning_rate": 3.6233939596195564e-05,
|
| 403 |
+
"loss": 3.4356,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.79,
|
| 408 |
+
"learning_rate": 3.602536292341065e-05,
|
| 409 |
+
"loss": 3.4244,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.84,
|
| 414 |
+
"learning_rate": 3.5816786250625733e-05,
|
| 415 |
+
"loss": 3.4026,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.88,
|
| 420 |
+
"learning_rate": 3.5608209577840815e-05,
|
| 421 |
+
"loss": 3.42,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.92,
|
| 426 |
+
"learning_rate": 3.5399632905055903e-05,
|
| 427 |
+
"loss": 3.3832,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.96,
|
| 432 |
+
"learning_rate": 3.5191056232270985e-05,
|
| 433 |
+
"loss": 3.3545,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.0,
|
| 438 |
+
"learning_rate": 3.4982479559486067e-05,
|
| 439 |
+
"loss": 3.3327,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 3.05,
|
| 444 |
+
"learning_rate": 3.4773902886701155e-05,
|
| 445 |
+
"loss": 3.3129,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 3.09,
|
| 450 |
+
"learning_rate": 3.4565326213916237e-05,
|
| 451 |
+
"loss": 3.3115,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 3.13,
|
| 456 |
+
"learning_rate": 3.4356749541131325e-05,
|
| 457 |
+
"loss": 3.2831,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.17,
|
| 462 |
+
"learning_rate": 3.4148172868346407e-05,
|
| 463 |
+
"loss": 3.2743,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.21,
|
| 468 |
+
"learning_rate": 3.393959619556149e-05,
|
| 469 |
+
"loss": 3.2856,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.25,
|
| 474 |
+
"learning_rate": 3.3731019522776576e-05,
|
| 475 |
+
"loss": 3.267,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.3,
|
| 480 |
+
"learning_rate": 3.352244284999166e-05,
|
| 481 |
+
"loss": 3.264,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 3.34,
|
| 486 |
+
"learning_rate": 3.331386617720674e-05,
|
| 487 |
+
"loss": 3.2614,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.38,
|
| 492 |
+
"learning_rate": 3.310528950442183e-05,
|
| 493 |
+
"loss": 3.2554,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 3.42,
|
| 498 |
+
"learning_rate": 3.289671283163691e-05,
|
| 499 |
+
"loss": 3.2308,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.46,
|
| 504 |
+
"learning_rate": 3.2688136158852e-05,
|
| 505 |
+
"loss": 3.2329,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.5,
|
| 510 |
+
"learning_rate": 3.247955948606708e-05,
|
| 511 |
+
"loss": 3.2174,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.55,
|
| 516 |
+
"learning_rate": 3.227098281328216e-05,
|
| 517 |
+
"loss": 3.1932,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.59,
|
| 522 |
+
"learning_rate": 3.206240614049725e-05,
|
| 523 |
+
"loss": 3.1748,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 3.63,
|
| 528 |
+
"learning_rate": 3.185382946771233e-05,
|
| 529 |
+
"loss": 3.1664,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.67,
|
| 534 |
+
"learning_rate": 3.164525279492741e-05,
|
| 535 |
+
"loss": 3.1862,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.71,
|
| 540 |
+
"learning_rate": 3.14366761221425e-05,
|
| 541 |
+
"loss": 3.1649,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.75,
|
| 546 |
+
"learning_rate": 3.122809944935758e-05,
|
| 547 |
+
"loss": 3.1528,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 3.8,
|
| 552 |
+
"learning_rate": 3.101952277657267e-05,
|
| 553 |
+
"loss": 3.1512,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.84,
|
| 558 |
+
"learning_rate": 3.081094610378775e-05,
|
| 559 |
+
"loss": 3.1362,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 3.88,
|
| 564 |
+
"learning_rate": 3.0602369431002834e-05,
|
| 565 |
+
"loss": 3.1517,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 3.92,
|
| 570 |
+
"learning_rate": 3.039379275821792e-05,
|
| 571 |
+
"loss": 3.1104,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.96,
|
| 576 |
+
"learning_rate": 3.0185216085433004e-05,
|
| 577 |
+
"loss": 3.1216,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 4.0,
|
| 582 |
+
"learning_rate": 2.997663941264809e-05,
|
| 583 |
+
"loss": 3.1177,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.05,
|
| 588 |
+
"learning_rate": 2.9768062739863174e-05,
|
| 589 |
+
"loss": 3.0815,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 4.09,
|
| 594 |
+
"learning_rate": 2.9559486067078256e-05,
|
| 595 |
+
"loss": 3.0839,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 4.13,
|
| 600 |
+
"learning_rate": 2.935090939429334e-05,
|
| 601 |
+
"loss": 3.0499,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 4.17,
|
| 606 |
+
"learning_rate": 2.9142332721508426e-05,
|
| 607 |
+
"loss": 3.0597,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 4.21,
|
| 612 |
+
"learning_rate": 2.8933756048723514e-05,
|
| 613 |
+
"loss": 3.0588,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.25,
|
| 618 |
+
"learning_rate": 2.87251793759386e-05,
|
| 619 |
+
"loss": 3.0393,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 4.3,
|
| 624 |
+
"learning_rate": 2.8516602703153684e-05,
|
| 625 |
+
"loss": 3.041,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.34,
|
| 630 |
+
"learning_rate": 2.8308026030368766e-05,
|
| 631 |
+
"loss": 3.0235,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 4.38,
|
| 636 |
+
"learning_rate": 2.809944935758385e-05,
|
| 637 |
+
"loss": 3.0257,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 4.42,
|
| 642 |
+
"learning_rate": 2.7890872684798936e-05,
|
| 643 |
+
"loss": 3.0206,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 4.46,
|
| 648 |
+
"learning_rate": 2.768229601201402e-05,
|
| 649 |
+
"loss": 3.0039,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 4.51,
|
| 654 |
+
"learning_rate": 2.7473719339229102e-05,
|
| 655 |
+
"loss": 2.9904,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 4.55,
|
| 660 |
+
"learning_rate": 2.7265142666444187e-05,
|
| 661 |
+
"loss": 2.9984,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 4.59,
|
| 666 |
+
"learning_rate": 2.7056565993659272e-05,
|
| 667 |
+
"loss": 2.9806,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.63,
|
| 672 |
+
"learning_rate": 2.6847989320874357e-05,
|
| 673 |
+
"loss": 2.9864,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 4.67,
|
| 678 |
+
"learning_rate": 2.663941264808944e-05,
|
| 679 |
+
"loss": 2.9916,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 4.71,
|
| 684 |
+
"learning_rate": 2.6430835975304524e-05,
|
| 685 |
+
"loss": 2.9799,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 4.76,
|
| 690 |
+
"learning_rate": 2.622225930251961e-05,
|
| 691 |
+
"loss": 2.9744,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 4.8,
|
| 696 |
+
"learning_rate": 2.6013682629734694e-05,
|
| 697 |
+
"loss": 2.9689,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.84,
|
| 702 |
+
"learning_rate": 2.5805105956949775e-05,
|
| 703 |
+
"loss": 2.9755,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 4.88,
|
| 708 |
+
"learning_rate": 2.559652928416486e-05,
|
| 709 |
+
"loss": 2.9355,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.92,
|
| 714 |
+
"learning_rate": 2.5387952611379945e-05,
|
| 715 |
+
"loss": 2.9549,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.96,
|
| 720 |
+
"learning_rate": 2.517937593859503e-05,
|
| 721 |
+
"loss": 2.9552,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 5.01,
|
| 726 |
+
"learning_rate": 2.497079926581011e-05,
|
| 727 |
+
"loss": 2.9608,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 5.05,
|
| 732 |
+
"learning_rate": 2.4762222593025197e-05,
|
| 733 |
+
"loss": 2.9073,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 5.09,
|
| 738 |
+
"learning_rate": 2.455364592024028e-05,
|
| 739 |
+
"loss": 2.9192,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 5.13,
|
| 744 |
+
"learning_rate": 2.4345069247455367e-05,
|
| 745 |
+
"loss": 2.922,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 5.17,
|
| 750 |
+
"learning_rate": 2.4136492574670448e-05,
|
| 751 |
+
"loss": 2.8877,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 5.21,
|
| 756 |
+
"learning_rate": 2.3927915901885533e-05,
|
| 757 |
+
"loss": 2.9045,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.26,
|
| 762 |
+
"learning_rate": 2.3719339229100618e-05,
|
| 763 |
+
"loss": 2.8876,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 5.3,
|
| 768 |
+
"learning_rate": 2.3510762556315703e-05,
|
| 769 |
+
"loss": 2.8986,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 5.34,
|
| 774 |
+
"learning_rate": 2.3302185883530785e-05,
|
| 775 |
+
"loss": 2.8924,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 5.38,
|
| 780 |
+
"learning_rate": 2.309360921074587e-05,
|
| 781 |
+
"loss": 2.8761,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 5.42,
|
| 786 |
+
"learning_rate": 2.2885032537960955e-05,
|
| 787 |
+
"loss": 2.8782,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 5.46,
|
| 792 |
+
"learning_rate": 2.267645586517604e-05,
|
| 793 |
+
"loss": 2.878,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 5.51,
|
| 798 |
+
"learning_rate": 2.2467879192391125e-05,
|
| 799 |
+
"loss": 2.8706,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 5.55,
|
| 804 |
+
"learning_rate": 2.225930251960621e-05,
|
| 805 |
+
"loss": 2.8647,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 5.59,
|
| 810 |
+
"learning_rate": 2.2050725846821295e-05,
|
| 811 |
+
"loss": 2.8407,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 5.63,
|
| 816 |
+
"learning_rate": 2.1842149174036376e-05,
|
| 817 |
+
"loss": 2.86,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 5.67,
|
| 822 |
+
"learning_rate": 2.163357250125146e-05,
|
| 823 |
+
"loss": 2.8585,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 5.72,
|
| 828 |
+
"learning_rate": 2.1424995828466546e-05,
|
| 829 |
+
"loss": 2.8472,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 5.76,
|
| 834 |
+
"learning_rate": 2.121641915568163e-05,
|
| 835 |
+
"loss": 2.8495,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 5.8,
|
| 840 |
+
"learning_rate": 2.1007842482896713e-05,
|
| 841 |
+
"loss": 2.8375,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 5.84,
|
| 846 |
+
"learning_rate": 2.0799265810111798e-05,
|
| 847 |
+
"loss": 2.8456,
|
| 848 |
+
"step": 70000
|
| 849 |
+
}
|
| 850 |
+
],
|
| 851 |
+
"max_steps": 119860,
|
| 852 |
+
"num_train_epochs": 10,
|
| 853 |
+
"total_flos": 2.8729170317888717e+17,
|
| 854 |
+
"trial_name": null,
|
| 855 |
+
"trial_params": null
|
| 856 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef99a94cdac6dbe7a598d14bbe06802a17e3accbb209ee42d18bc203ee14ed55
|
| 3 |
+
size 2287
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|