Commit
•
8385fbd
1
Parent(s):
31287ee
Upload with huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- checkpoint-10000/config.json +28 -0
- checkpoint-10000/optimizer.pt +3 -0
- checkpoint-10000/pytorch_model.bin +3 -0
- checkpoint-10000/rng_state.pth +3 -0
- checkpoint-10000/scaler.pt +3 -0
- checkpoint-10000/scheduler.pt +3 -0
- checkpoint-10000/trainer_state.json +616 -0
- checkpoint-10000/training_args.bin +3 -0
- checkpoint-10700/config.json +28 -0
- checkpoint-10700/optimizer.pt +3 -0
- checkpoint-10700/pytorch_model.bin +3 -0
- checkpoint-10700/rng_state.pth +3 -0
- checkpoint-10700/scaler.pt +3 -0
- checkpoint-10700/scheduler.pt +3 -0
- checkpoint-10700/trainer_state.json +658 -0
- checkpoint-10700/training_args.bin +3 -0
- checkpoint-13200/config.json +28 -0
- checkpoint-13200/optimizer.pt +3 -0
- checkpoint-13200/pytorch_model.bin +3 -0
- checkpoint-13200/rng_state.pth +3 -0
- checkpoint-13200/scaler.pt +3 -0
- checkpoint-13200/scheduler.pt +3 -0
- checkpoint-13200/trainer_state.json +808 -0
- checkpoint-13200/training_args.bin +3 -0
- checkpoint-15600/config.json +28 -0
- checkpoint-15600/optimizer.pt +3 -0
- checkpoint-15600/pytorch_model.bin +3 -0
- checkpoint-15600/rng_state.pth +3 -0
- checkpoint-15600/scaler.pt +3 -0
- checkpoint-15600/scheduler.pt +3 -0
- checkpoint-15600/trainer_state.json +952 -0
- checkpoint-15600/training_args.bin +3 -0
- checkpoint-18000/config.json +28 -0
- checkpoint-18000/optimizer.pt +3 -0
- checkpoint-18000/pytorch_model.bin +3 -0
- checkpoint-18000/rng_state.pth +3 -0
- checkpoint-18000/scaler.pt +3 -0
- checkpoint-18000/scheduler.pt +3 -0
- checkpoint-18000/trainer_state.json +1096 -0
- checkpoint-18000/training_args.bin +3 -0
- checkpoint-20000/config.json +28 -0
- checkpoint-20000/optimizer.pt +3 -0
- checkpoint-20000/pytorch_model.bin +3 -0
- checkpoint-20000/rng_state.pth +3 -0
- checkpoint-20000/scaler.pt +3 -0
- checkpoint-20000/scheduler.pt +3 -0
- checkpoint-20000/trainer_state.json +1216 -0
- checkpoint-20000/training_args.bin +3 -0
- checkpoint-3100/config.json +28 -0
- checkpoint-3100/optimizer.pt +3 -0
checkpoint-10000/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-10000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c5b775aa8a202fd4d36055e1eaf56d5fc6b253a6e317aa79dc35388b41dba4c
|
| 3 |
+
size 883771077
|
checkpoint-10000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ae23bef57bae1b488c85a5990592cf30f4d350cca3e7354e9ff6526ba08e78f
|
| 3 |
+
size 441897977
|
checkpoint-10000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eef4cbb69ab20145d1568f2f44ce9e8cb53c9fd697b6bade8bebff7bfee5082a
|
| 3 |
+
size 14511
|
checkpoint-10000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9be2ad76dcbc923b00e6a142f6db62aad4a46c47bb83864ccb68ddc899d0ce78
|
| 3 |
+
size 557
|
checkpoint-10000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26b421957aeaf8aaf98c5369a533b691198a7ccccd412c8da8f9b0e06d8aaee6
|
| 3 |
+
size 627
|
checkpoint-10000/trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 78.12415130940835,
|
| 5 |
+
"global_step": 10000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.78,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 10.1419,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.56,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 9.2788,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 2.34,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 8.8011,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 3.12,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 8.3846,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.9,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 7.8214,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.68,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 7.3094,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5.47,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 6.7911,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 6.25,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 6.4144,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 7.03,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 6.2168,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 7.81,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 6.0642,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 8.59,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 6.0511,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 9.37,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 5.995,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 10.16,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 5.9539,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 10.93,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 5.8675,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 11.71,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 5.8874,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 12.5,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 5.8578,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 13.28,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 5.8386,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 14.06,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 5.8138,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 14.84,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 5.7399,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 15.62,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 5.7753,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 16.4,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 5.7564,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 17.19,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 5.738,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 17.96,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 5.6753,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 18.74,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 5.7082,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 19.53,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 5.6991,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 20.31,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 5.6801,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 21.09,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 5.6692,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 21.87,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 5.6063,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 22.65,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 5.6445,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 23.43,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 5.6328,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 24.22,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 5.6217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 24.99,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 5.5601,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 25.78,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 5.5976,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 26.56,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 5.5911,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 27.34,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 5.5738,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 28.12,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 5.566,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 28.9,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 5.5071,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 29.68,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 5.5438,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 30.47,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 5.5366,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 31.25,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 5.5268,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 32.03,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 5.517,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 32.81,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 5.4574,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 33.59,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 5.5002,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 34.37,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 5.4887,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 35.16,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 5.4805,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 35.93,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 5.4265,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 36.71,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 5.4615,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 37.5,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 5.4576,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 38.28,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 5.4421,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 39.06,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 5.4342,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 39.84,
|
| 312 |
+
"learning_rate": 5.1000000000000006e-05,
|
| 313 |
+
"loss": 5.3641,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 40.62,
|
| 318 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 319 |
+
"loss": 5.379,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 41.4,
|
| 324 |
+
"learning_rate": 5.300000000000001e-05,
|
| 325 |
+
"loss": 5.3638,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 42.19,
|
| 330 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 331 |
+
"loss": 5.3441,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 42.96,
|
| 336 |
+
"learning_rate": 5.500000000000001e-05,
|
| 337 |
+
"loss": 5.2759,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 43.74,
|
| 342 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 343 |
+
"loss": 5.3011,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 44.53,
|
| 348 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 349 |
+
"loss": 5.2758,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 45.31,
|
| 354 |
+
"learning_rate": 5.8e-05,
|
| 355 |
+
"loss": 5.2559,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 46.09,
|
| 360 |
+
"learning_rate": 5.9e-05,
|
| 361 |
+
"loss": 5.2326,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 46.87,
|
| 366 |
+
"learning_rate": 6e-05,
|
| 367 |
+
"loss": 5.1616,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 47.65,
|
| 372 |
+
"learning_rate": 6.1e-05,
|
| 373 |
+
"loss": 5.1753,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 48.43,
|
| 378 |
+
"learning_rate": 6.2e-05,
|
| 379 |
+
"loss": 5.1378,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 49.22,
|
| 384 |
+
"learning_rate": 6.3e-05,
|
| 385 |
+
"loss": 5.1,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 49.99,
|
| 390 |
+
"learning_rate": 6.400000000000001e-05,
|
| 391 |
+
"loss": 5.015,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 50.78,
|
| 396 |
+
"learning_rate": 6.500000000000001e-05,
|
| 397 |
+
"loss": 4.9758,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 51.56,
|
| 402 |
+
"learning_rate": 6.6e-05,
|
| 403 |
+
"loss": 4.8417,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 52.34,
|
| 408 |
+
"learning_rate": 6.7e-05,
|
| 409 |
+
"loss": 4.7116,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 53.12,
|
| 414 |
+
"learning_rate": 6.800000000000001e-05,
|
| 415 |
+
"loss": 4.5582,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 53.9,
|
| 420 |
+
"learning_rate": 6.9e-05,
|
| 421 |
+
"loss": 4.3437,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 54.68,
|
| 426 |
+
"learning_rate": 7e-05,
|
| 427 |
+
"loss": 4.2114,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 55.47,
|
| 432 |
+
"learning_rate": 7.1e-05,
|
| 433 |
+
"loss": 4.1021,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 56.25,
|
| 438 |
+
"learning_rate": 7.2e-05,
|
| 439 |
+
"loss": 4.0074,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 57.03,
|
| 444 |
+
"learning_rate": 7.3e-05,
|
| 445 |
+
"loss": 3.9346,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 57.81,
|
| 450 |
+
"learning_rate": 7.4e-05,
|
| 451 |
+
"loss": 3.8289,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 58.59,
|
| 456 |
+
"learning_rate": 7.500000000000001e-05,
|
| 457 |
+
"loss": 3.8105,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 59.37,
|
| 462 |
+
"learning_rate": 7.6e-05,
|
| 463 |
+
"loss": 3.755,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 60.16,
|
| 468 |
+
"learning_rate": 7.7e-05,
|
| 469 |
+
"loss": 3.7105,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 60.93,
|
| 474 |
+
"learning_rate": 7.800000000000001e-05,
|
| 475 |
+
"loss": 3.6394,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 61.71,
|
| 480 |
+
"learning_rate": 7.900000000000001e-05,
|
| 481 |
+
"loss": 3.6262,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 62.5,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 3.5924,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 63.28,
|
| 492 |
+
"learning_rate": 8.1e-05,
|
| 493 |
+
"loss": 3.558,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 64.06,
|
| 498 |
+
"learning_rate": 8.2e-05,
|
| 499 |
+
"loss": 3.5255,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 64.84,
|
| 504 |
+
"learning_rate": 8.3e-05,
|
| 505 |
+
"loss": 3.4602,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 65.62,
|
| 510 |
+
"learning_rate": 8.4e-05,
|
| 511 |
+
"loss": 3.4641,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 66.4,
|
| 516 |
+
"learning_rate": 8.5e-05,
|
| 517 |
+
"loss": 3.435,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 67.19,
|
| 522 |
+
"learning_rate": 8.6e-05,
|
| 523 |
+
"loss": 3.408,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 67.96,
|
| 528 |
+
"learning_rate": 8.7e-05,
|
| 529 |
+
"loss": 3.3594,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 68.74,
|
| 534 |
+
"learning_rate": 8.800000000000001e-05,
|
| 535 |
+
"loss": 3.3593,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 69.53,
|
| 540 |
+
"learning_rate": 8.900000000000001e-05,
|
| 541 |
+
"loss": 3.3372,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 70.31,
|
| 546 |
+
"learning_rate": 9e-05,
|
| 547 |
+
"loss": 3.3217,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 71.09,
|
| 552 |
+
"learning_rate": 9.1e-05,
|
| 553 |
+
"loss": 3.2985,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 71.87,
|
| 558 |
+
"learning_rate": 9.200000000000001e-05,
|
| 559 |
+
"loss": 3.2509,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 72.65,
|
| 564 |
+
"learning_rate": 9.300000000000001e-05,
|
| 565 |
+
"loss": 3.2584,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 73.43,
|
| 570 |
+
"learning_rate": 9.4e-05,
|
| 571 |
+
"loss": 3.2386,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 74.22,
|
| 576 |
+
"learning_rate": 9.5e-05,
|
| 577 |
+
"loss": 3.2232,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 74.99,
|
| 582 |
+
"learning_rate": 9.6e-05,
|
| 583 |
+
"loss": 3.1786,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 75.78,
|
| 588 |
+
"learning_rate": 9.7e-05,
|
| 589 |
+
"loss": 3.1855,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 76.56,
|
| 594 |
+
"learning_rate": 9.8e-05,
|
| 595 |
+
"loss": 3.1737,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 77.34,
|
| 600 |
+
"learning_rate": 9.900000000000001e-05,
|
| 601 |
+
"loss": 3.1565,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 78.12,
|
| 606 |
+
"learning_rate": 0.0001,
|
| 607 |
+
"loss": 3.1442,
|
| 608 |
+
"step": 10000
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 20480,
|
| 612 |
+
"num_train_epochs": 160,
|
| 613 |
+
"total_flos": 6.779220101431296e+17,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
checkpoint-10000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6de0a5888f9534c475c0b5c0d6fa6d4920e1a50a89517b67872f8b6ca51ef166
|
| 3 |
+
size 3579
|
checkpoint-10700/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-10700/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a62fa10a633b9cc8c455879067e98c4a60dee9731426311513762c8fcc5cc383
|
| 3 |
+
size 883771077
|
checkpoint-10700/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4a60cffa7310860b0eb84d368758a8f86c4d12451fc09d6cc90a802ff691a75
|
| 3 |
+
size 441897977
|
checkpoint-10700/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7da8e8fc3d9bda7e0d5df98b0773b928133c5ee23feebab467973b5533189833
|
| 3 |
+
size 14511
|
checkpoint-10700/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:680d3cfc1ccd4ccf0f6548860547f0c185be680cb7d873b614410fff2d70ee99
|
| 3 |
+
size 557
|
checkpoint-10700/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11681a532209a6e5b24701b99eb2918864d8c52d36263c42e8a41b1b76442615
|
| 3 |
+
size 627
|
checkpoint-10700/trainer_state.json
ADDED
|
@@ -0,0 +1,658 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 83.58971871968963,
|
| 5 |
+
"global_step": 10700,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.78,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 10.1419,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.56,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 9.2788,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 2.34,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 8.8011,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 3.12,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 8.3846,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.9,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 7.8214,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.68,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 7.3094,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5.47,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 6.7911,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 6.25,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 6.4144,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 7.03,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 6.2168,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 7.81,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 6.0642,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 8.59,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 6.0511,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 9.37,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 5.995,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 10.16,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 5.9539,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 10.93,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 5.8675,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 11.71,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 5.8874,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 12.5,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 5.8578,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 13.28,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 5.8386,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 14.06,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 5.8138,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 14.84,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 5.7399,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 15.62,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 5.7753,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 16.4,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 5.7564,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 17.19,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 5.738,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 17.96,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 5.6753,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 18.74,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 5.7082,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 19.53,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 5.6991,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 20.31,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 5.6801,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 21.09,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 5.6692,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 21.87,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 5.6063,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 22.65,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 5.6445,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 23.43,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 5.6328,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 24.22,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 5.6217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 24.99,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 5.5601,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 25.78,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 5.5976,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 26.56,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 5.5911,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 27.34,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 5.5738,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 28.12,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 5.566,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 28.9,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 5.5071,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 29.68,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 5.5438,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 30.47,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 5.5366,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 31.25,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 5.5268,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 32.03,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 5.517,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 32.81,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 5.4574,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 33.59,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 5.5002,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 34.37,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 5.4887,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 35.16,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 5.4805,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 35.93,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 5.4265,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 36.71,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 5.4615,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 37.5,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 5.4576,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 38.28,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 5.4421,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 39.06,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 5.4342,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 39.84,
|
| 312 |
+
"learning_rate": 5.1000000000000006e-05,
|
| 313 |
+
"loss": 5.3641,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 40.62,
|
| 318 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 319 |
+
"loss": 5.379,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 41.4,
|
| 324 |
+
"learning_rate": 5.300000000000001e-05,
|
| 325 |
+
"loss": 5.3638,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 42.19,
|
| 330 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 331 |
+
"loss": 5.3441,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 42.96,
|
| 336 |
+
"learning_rate": 5.500000000000001e-05,
|
| 337 |
+
"loss": 5.2759,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 43.74,
|
| 342 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 343 |
+
"loss": 5.3011,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 44.53,
|
| 348 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 349 |
+
"loss": 5.2758,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 45.31,
|
| 354 |
+
"learning_rate": 5.8e-05,
|
| 355 |
+
"loss": 5.2559,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 46.09,
|
| 360 |
+
"learning_rate": 5.9e-05,
|
| 361 |
+
"loss": 5.2326,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 46.87,
|
| 366 |
+
"learning_rate": 6e-05,
|
| 367 |
+
"loss": 5.1616,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 47.65,
|
| 372 |
+
"learning_rate": 6.1e-05,
|
| 373 |
+
"loss": 5.1753,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 48.43,
|
| 378 |
+
"learning_rate": 6.2e-05,
|
| 379 |
+
"loss": 5.1378,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 49.22,
|
| 384 |
+
"learning_rate": 6.3e-05,
|
| 385 |
+
"loss": 5.1,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 49.99,
|
| 390 |
+
"learning_rate": 6.400000000000001e-05,
|
| 391 |
+
"loss": 5.015,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 50.78,
|
| 396 |
+
"learning_rate": 6.500000000000001e-05,
|
| 397 |
+
"loss": 4.9758,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 51.56,
|
| 402 |
+
"learning_rate": 6.6e-05,
|
| 403 |
+
"loss": 4.8417,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 52.34,
|
| 408 |
+
"learning_rate": 6.7e-05,
|
| 409 |
+
"loss": 4.7116,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 53.12,
|
| 414 |
+
"learning_rate": 6.800000000000001e-05,
|
| 415 |
+
"loss": 4.5582,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 53.9,
|
| 420 |
+
"learning_rate": 6.9e-05,
|
| 421 |
+
"loss": 4.3437,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 54.68,
|
| 426 |
+
"learning_rate": 7e-05,
|
| 427 |
+
"loss": 4.2114,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 55.47,
|
| 432 |
+
"learning_rate": 7.1e-05,
|
| 433 |
+
"loss": 4.1021,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 56.25,
|
| 438 |
+
"learning_rate": 7.2e-05,
|
| 439 |
+
"loss": 4.0074,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 57.03,
|
| 444 |
+
"learning_rate": 7.3e-05,
|
| 445 |
+
"loss": 3.9346,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 57.81,
|
| 450 |
+
"learning_rate": 7.4e-05,
|
| 451 |
+
"loss": 3.8289,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 58.59,
|
| 456 |
+
"learning_rate": 7.500000000000001e-05,
|
| 457 |
+
"loss": 3.8105,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 59.37,
|
| 462 |
+
"learning_rate": 7.6e-05,
|
| 463 |
+
"loss": 3.755,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 60.16,
|
| 468 |
+
"learning_rate": 7.7e-05,
|
| 469 |
+
"loss": 3.7105,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 60.93,
|
| 474 |
+
"learning_rate": 7.800000000000001e-05,
|
| 475 |
+
"loss": 3.6394,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 61.71,
|
| 480 |
+
"learning_rate": 7.900000000000001e-05,
|
| 481 |
+
"loss": 3.6262,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 62.5,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 3.5924,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 63.28,
|
| 492 |
+
"learning_rate": 8.1e-05,
|
| 493 |
+
"loss": 3.558,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 64.06,
|
| 498 |
+
"learning_rate": 8.2e-05,
|
| 499 |
+
"loss": 3.5255,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 64.84,
|
| 504 |
+
"learning_rate": 8.3e-05,
|
| 505 |
+
"loss": 3.4602,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 65.62,
|
| 510 |
+
"learning_rate": 8.4e-05,
|
| 511 |
+
"loss": 3.4641,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 66.4,
|
| 516 |
+
"learning_rate": 8.5e-05,
|
| 517 |
+
"loss": 3.435,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 67.19,
|
| 522 |
+
"learning_rate": 8.6e-05,
|
| 523 |
+
"loss": 3.408,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 67.96,
|
| 528 |
+
"learning_rate": 8.7e-05,
|
| 529 |
+
"loss": 3.3594,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 68.74,
|
| 534 |
+
"learning_rate": 8.800000000000001e-05,
|
| 535 |
+
"loss": 3.3593,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 69.53,
|
| 540 |
+
"learning_rate": 8.900000000000001e-05,
|
| 541 |
+
"loss": 3.3372,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 70.31,
|
| 546 |
+
"learning_rate": 9e-05,
|
| 547 |
+
"loss": 3.3217,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 71.09,
|
| 552 |
+
"learning_rate": 9.1e-05,
|
| 553 |
+
"loss": 3.2985,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 71.87,
|
| 558 |
+
"learning_rate": 9.200000000000001e-05,
|
| 559 |
+
"loss": 3.2509,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 72.65,
|
| 564 |
+
"learning_rate": 9.300000000000001e-05,
|
| 565 |
+
"loss": 3.2584,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 73.43,
|
| 570 |
+
"learning_rate": 9.4e-05,
|
| 571 |
+
"loss": 3.2386,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 74.22,
|
| 576 |
+
"learning_rate": 9.5e-05,
|
| 577 |
+
"loss": 3.2232,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 74.99,
|
| 582 |
+
"learning_rate": 9.6e-05,
|
| 583 |
+
"loss": 3.1786,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 75.78,
|
| 588 |
+
"learning_rate": 9.7e-05,
|
| 589 |
+
"loss": 3.1855,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 76.56,
|
| 594 |
+
"learning_rate": 9.8e-05,
|
| 595 |
+
"loss": 3.1737,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 77.34,
|
| 600 |
+
"learning_rate": 9.900000000000001e-05,
|
| 601 |
+
"loss": 3.1565,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 78.12,
|
| 606 |
+
"learning_rate": 0.0001,
|
| 607 |
+
"loss": 3.1442,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 78.9,
|
| 612 |
+
"learning_rate": 9.904580152671757e-05,
|
| 613 |
+
"loss": 3.1003,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 79.68,
|
| 618 |
+
"learning_rate": 9.809160305343512e-05,
|
| 619 |
+
"loss": 3.1137,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 80.47,
|
| 624 |
+
"learning_rate": 9.713740458015268e-05,
|
| 625 |
+
"loss": 3.0958,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 81.25,
|
| 630 |
+
"learning_rate": 9.618320610687024e-05,
|
| 631 |
+
"loss": 3.0853,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 82.03,
|
| 636 |
+
"learning_rate": 9.522900763358779e-05,
|
| 637 |
+
"loss": 3.0704,
|
| 638 |
+
"step": 10500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 82.81,
|
| 642 |
+
"learning_rate": 9.427480916030534e-05,
|
| 643 |
+
"loss": 3.03,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 83.59,
|
| 648 |
+
"learning_rate": 9.33206106870229e-05,
|
| 649 |
+
"loss": 3.0428,
|
| 650 |
+
"step": 10700
|
| 651 |
+
}
|
| 652 |
+
],
|
| 653 |
+
"max_steps": 20480,
|
| 654 |
+
"num_train_epochs": 160,
|
| 655 |
+
"total_flos": 7.253523355926528e+17,
|
| 656 |
+
"trial_name": null,
|
| 657 |
+
"trial_params": null
|
| 658 |
+
}
|
checkpoint-10700/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6de0a5888f9534c475c0b5c0d6fa6d4920e1a50a89517b67872f8b6ca51ef166
|
| 3 |
+
size 3579
|
checkpoint-13200/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-13200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e38f85e320af3c483e4c5629487359bc2beb6cc4a5fb1be4e62a40daf3d86ea6
|
| 3 |
+
size 883771077
|
checkpoint-13200/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:232ff547d71f660c54653970cfb25dde6796c4b763e25513c7477f785bb05974
|
| 3 |
+
size 441897977
|
checkpoint-13200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:65853d8eba688cfb8801838aa8cddba4a37b866067cfbe2eb1335e92205cf0d6
|
| 3 |
+
size 14511
|
checkpoint-13200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0961d88247b5b0b3e4d659805dd764b2ea62614b66a9b6bd857b1aeb7de0599a
|
| 3 |
+
size 557
|
checkpoint-13200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eea0a608bab19c5f6d0c9e18d30677a93977ade21a6be290b4d4bd8aa2d347ca
|
| 3 |
+
size 627
|
checkpoint-13200/trainer_state.json
ADDED
|
@@ -0,0 +1,808 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 103.12415130940835,
|
| 5 |
+
"global_step": 13200,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.78,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 10.1419,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.56,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 9.2788,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 2.34,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 8.8011,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 3.12,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 8.3846,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.9,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 7.8214,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.68,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 7.3094,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5.47,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 6.7911,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 6.25,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 6.4144,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 7.03,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 6.2168,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 7.81,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 6.0642,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 8.59,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 6.0511,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 9.37,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 5.995,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 10.16,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 5.9539,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 10.93,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 5.8675,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 11.71,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 5.8874,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 12.5,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 5.8578,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 13.28,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 5.8386,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 14.06,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 5.8138,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 14.84,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 5.7399,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 15.62,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 5.7753,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 16.4,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 5.7564,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 17.19,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 5.738,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 17.96,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 5.6753,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 18.74,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 5.7082,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 19.53,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 5.6991,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 20.31,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 5.6801,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 21.09,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 5.6692,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 21.87,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 5.6063,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 22.65,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 5.6445,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 23.43,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 5.6328,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 24.22,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 5.6217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 24.99,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 5.5601,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 25.78,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 5.5976,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 26.56,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 5.5911,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 27.34,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 5.5738,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 28.12,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 5.566,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 28.9,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 5.5071,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 29.68,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 5.5438,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 30.47,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 5.5366,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 31.25,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 5.5268,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 32.03,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 5.517,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 32.81,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 5.4574,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 33.59,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 5.5002,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 34.37,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 5.4887,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 35.16,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 5.4805,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 35.93,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 5.4265,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 36.71,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 5.4615,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 37.5,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 5.4576,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 38.28,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 5.4421,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 39.06,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 5.4342,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 39.84,
|
| 312 |
+
"learning_rate": 5.1000000000000006e-05,
|
| 313 |
+
"loss": 5.3641,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 40.62,
|
| 318 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 319 |
+
"loss": 5.379,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 41.4,
|
| 324 |
+
"learning_rate": 5.300000000000001e-05,
|
| 325 |
+
"loss": 5.3638,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 42.19,
|
| 330 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 331 |
+
"loss": 5.3441,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 42.96,
|
| 336 |
+
"learning_rate": 5.500000000000001e-05,
|
| 337 |
+
"loss": 5.2759,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 43.74,
|
| 342 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 343 |
+
"loss": 5.3011,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 44.53,
|
| 348 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 349 |
+
"loss": 5.2758,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 45.31,
|
| 354 |
+
"learning_rate": 5.8e-05,
|
| 355 |
+
"loss": 5.2559,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 46.09,
|
| 360 |
+
"learning_rate": 5.9e-05,
|
| 361 |
+
"loss": 5.2326,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 46.87,
|
| 366 |
+
"learning_rate": 6e-05,
|
| 367 |
+
"loss": 5.1616,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 47.65,
|
| 372 |
+
"learning_rate": 6.1e-05,
|
| 373 |
+
"loss": 5.1753,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 48.43,
|
| 378 |
+
"learning_rate": 6.2e-05,
|
| 379 |
+
"loss": 5.1378,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 49.22,
|
| 384 |
+
"learning_rate": 6.3e-05,
|
| 385 |
+
"loss": 5.1,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 49.99,
|
| 390 |
+
"learning_rate": 6.400000000000001e-05,
|
| 391 |
+
"loss": 5.015,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 50.78,
|
| 396 |
+
"learning_rate": 6.500000000000001e-05,
|
| 397 |
+
"loss": 4.9758,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 51.56,
|
| 402 |
+
"learning_rate": 6.6e-05,
|
| 403 |
+
"loss": 4.8417,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 52.34,
|
| 408 |
+
"learning_rate": 6.7e-05,
|
| 409 |
+
"loss": 4.7116,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 53.12,
|
| 414 |
+
"learning_rate": 6.800000000000001e-05,
|
| 415 |
+
"loss": 4.5582,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 53.9,
|
| 420 |
+
"learning_rate": 6.9e-05,
|
| 421 |
+
"loss": 4.3437,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 54.68,
|
| 426 |
+
"learning_rate": 7e-05,
|
| 427 |
+
"loss": 4.2114,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 55.47,
|
| 432 |
+
"learning_rate": 7.1e-05,
|
| 433 |
+
"loss": 4.1021,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 56.25,
|
| 438 |
+
"learning_rate": 7.2e-05,
|
| 439 |
+
"loss": 4.0074,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 57.03,
|
| 444 |
+
"learning_rate": 7.3e-05,
|
| 445 |
+
"loss": 3.9346,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 57.81,
|
| 450 |
+
"learning_rate": 7.4e-05,
|
| 451 |
+
"loss": 3.8289,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 58.59,
|
| 456 |
+
"learning_rate": 7.500000000000001e-05,
|
| 457 |
+
"loss": 3.8105,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 59.37,
|
| 462 |
+
"learning_rate": 7.6e-05,
|
| 463 |
+
"loss": 3.755,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 60.16,
|
| 468 |
+
"learning_rate": 7.7e-05,
|
| 469 |
+
"loss": 3.7105,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 60.93,
|
| 474 |
+
"learning_rate": 7.800000000000001e-05,
|
| 475 |
+
"loss": 3.6394,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 61.71,
|
| 480 |
+
"learning_rate": 7.900000000000001e-05,
|
| 481 |
+
"loss": 3.6262,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 62.5,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 3.5924,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 63.28,
|
| 492 |
+
"learning_rate": 8.1e-05,
|
| 493 |
+
"loss": 3.558,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 64.06,
|
| 498 |
+
"learning_rate": 8.2e-05,
|
| 499 |
+
"loss": 3.5255,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 64.84,
|
| 504 |
+
"learning_rate": 8.3e-05,
|
| 505 |
+
"loss": 3.4602,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 65.62,
|
| 510 |
+
"learning_rate": 8.4e-05,
|
| 511 |
+
"loss": 3.4641,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 66.4,
|
| 516 |
+
"learning_rate": 8.5e-05,
|
| 517 |
+
"loss": 3.435,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 67.19,
|
| 522 |
+
"learning_rate": 8.6e-05,
|
| 523 |
+
"loss": 3.408,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 67.96,
|
| 528 |
+
"learning_rate": 8.7e-05,
|
| 529 |
+
"loss": 3.3594,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 68.74,
|
| 534 |
+
"learning_rate": 8.800000000000001e-05,
|
| 535 |
+
"loss": 3.3593,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 69.53,
|
| 540 |
+
"learning_rate": 8.900000000000001e-05,
|
| 541 |
+
"loss": 3.3372,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 70.31,
|
| 546 |
+
"learning_rate": 9e-05,
|
| 547 |
+
"loss": 3.3217,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 71.09,
|
| 552 |
+
"learning_rate": 9.1e-05,
|
| 553 |
+
"loss": 3.2985,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 71.87,
|
| 558 |
+
"learning_rate": 9.200000000000001e-05,
|
| 559 |
+
"loss": 3.2509,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 72.65,
|
| 564 |
+
"learning_rate": 9.300000000000001e-05,
|
| 565 |
+
"loss": 3.2584,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 73.43,
|
| 570 |
+
"learning_rate": 9.4e-05,
|
| 571 |
+
"loss": 3.2386,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 74.22,
|
| 576 |
+
"learning_rate": 9.5e-05,
|
| 577 |
+
"loss": 3.2232,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 74.99,
|
| 582 |
+
"learning_rate": 9.6e-05,
|
| 583 |
+
"loss": 3.1786,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 75.78,
|
| 588 |
+
"learning_rate": 9.7e-05,
|
| 589 |
+
"loss": 3.1855,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 76.56,
|
| 594 |
+
"learning_rate": 9.8e-05,
|
| 595 |
+
"loss": 3.1737,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 77.34,
|
| 600 |
+
"learning_rate": 9.900000000000001e-05,
|
| 601 |
+
"loss": 3.1565,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 78.12,
|
| 606 |
+
"learning_rate": 0.0001,
|
| 607 |
+
"loss": 3.1442,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 78.9,
|
| 612 |
+
"learning_rate": 9.904580152671757e-05,
|
| 613 |
+
"loss": 3.1003,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 79.68,
|
| 618 |
+
"learning_rate": 9.809160305343512e-05,
|
| 619 |
+
"loss": 3.1137,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 80.47,
|
| 624 |
+
"learning_rate": 9.713740458015268e-05,
|
| 625 |
+
"loss": 3.0958,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 81.25,
|
| 630 |
+
"learning_rate": 9.618320610687024e-05,
|
| 631 |
+
"loss": 3.0853,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 82.03,
|
| 636 |
+
"learning_rate": 9.522900763358779e-05,
|
| 637 |
+
"loss": 3.0704,
|
| 638 |
+
"step": 10500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 82.81,
|
| 642 |
+
"learning_rate": 9.427480916030534e-05,
|
| 643 |
+
"loss": 3.03,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 83.59,
|
| 648 |
+
"learning_rate": 9.33206106870229e-05,
|
| 649 |
+
"loss": 3.0428,
|
| 650 |
+
"step": 10700
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 84.37,
|
| 654 |
+
"learning_rate": 9.236641221374047e-05,
|
| 655 |
+
"loss": 3.0299,
|
| 656 |
+
"step": 10800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 85.16,
|
| 660 |
+
"learning_rate": 9.141221374045802e-05,
|
| 661 |
+
"loss": 3.0239,
|
| 662 |
+
"step": 10900
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 85.93,
|
| 666 |
+
"learning_rate": 9.045801526717558e-05,
|
| 667 |
+
"loss": 2.9818,
|
| 668 |
+
"step": 11000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 86.71,
|
| 672 |
+
"learning_rate": 8.950381679389314e-05,
|
| 673 |
+
"loss": 2.9967,
|
| 674 |
+
"step": 11100
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 87.5,
|
| 678 |
+
"learning_rate": 8.854961832061069e-05,
|
| 679 |
+
"loss": 2.9866,
|
| 680 |
+
"step": 11200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 88.28,
|
| 684 |
+
"learning_rate": 8.759541984732825e-05,
|
| 685 |
+
"loss": 2.9758,
|
| 686 |
+
"step": 11300
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 89.06,
|
| 690 |
+
"learning_rate": 8.664122137404582e-05,
|
| 691 |
+
"loss": 2.968,
|
| 692 |
+
"step": 11400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 89.84,
|
| 696 |
+
"learning_rate": 8.568702290076335e-05,
|
| 697 |
+
"loss": 2.9276,
|
| 698 |
+
"step": 11500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 90.62,
|
| 702 |
+
"learning_rate": 8.473282442748092e-05,
|
| 703 |
+
"loss": 2.9464,
|
| 704 |
+
"step": 11600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 91.4,
|
| 708 |
+
"learning_rate": 8.377862595419848e-05,
|
| 709 |
+
"loss": 2.9342,
|
| 710 |
+
"step": 11700
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 92.19,
|
| 714 |
+
"learning_rate": 8.282442748091603e-05,
|
| 715 |
+
"loss": 2.9313,
|
| 716 |
+
"step": 11800
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 92.96,
|
| 720 |
+
"learning_rate": 8.187022900763359e-05,
|
| 721 |
+
"loss": 2.8932,
|
| 722 |
+
"step": 11900
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 93.74,
|
| 726 |
+
"learning_rate": 8.091603053435115e-05,
|
| 727 |
+
"loss": 2.9087,
|
| 728 |
+
"step": 12000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 94.53,
|
| 732 |
+
"learning_rate": 7.996183206106872e-05,
|
| 733 |
+
"loss": 2.9054,
|
| 734 |
+
"step": 12100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 95.31,
|
| 738 |
+
"learning_rate": 7.900763358778626e-05,
|
| 739 |
+
"loss": 2.8916,
|
| 740 |
+
"step": 12200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 96.09,
|
| 744 |
+
"learning_rate": 7.805343511450383e-05,
|
| 745 |
+
"loss": 2.8888,
|
| 746 |
+
"step": 12300
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 96.87,
|
| 750 |
+
"learning_rate": 7.709923664122138e-05,
|
| 751 |
+
"loss": 2.8505,
|
| 752 |
+
"step": 12400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 97.65,
|
| 756 |
+
"learning_rate": 7.614503816793893e-05,
|
| 757 |
+
"loss": 2.8729,
|
| 758 |
+
"step": 12500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 98.43,
|
| 762 |
+
"learning_rate": 7.519083969465649e-05,
|
| 763 |
+
"loss": 2.8576,
|
| 764 |
+
"step": 12600
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 99.22,
|
| 768 |
+
"learning_rate": 7.423664122137405e-05,
|
| 769 |
+
"loss": 2.8569,
|
| 770 |
+
"step": 12700
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 99.99,
|
| 774 |
+
"learning_rate": 7.32824427480916e-05,
|
| 775 |
+
"loss": 2.8276,
|
| 776 |
+
"step": 12800
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 100.78,
|
| 780 |
+
"learning_rate": 7.232824427480916e-05,
|
| 781 |
+
"loss": 2.8404,
|
| 782 |
+
"step": 12900
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 101.56,
|
| 786 |
+
"learning_rate": 7.137404580152673e-05,
|
| 787 |
+
"loss": 2.8337,
|
| 788 |
+
"step": 13000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 102.34,
|
| 792 |
+
"learning_rate": 7.041984732824428e-05,
|
| 793 |
+
"loss": 2.8285,
|
| 794 |
+
"step": 13100
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 103.12,
|
| 798 |
+
"learning_rate": 6.946564885496184e-05,
|
| 799 |
+
"loss": 2.8209,
|
| 800 |
+
"step": 13200
|
| 801 |
+
}
|
| 802 |
+
],
|
| 803 |
+
"max_steps": 20480,
|
| 804 |
+
"num_train_epochs": 160,
|
| 805 |
+
"total_flos": 8.948591590637568e+17,
|
| 806 |
+
"trial_name": null,
|
| 807 |
+
"trial_params": null
|
| 808 |
+
}
|
checkpoint-13200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6de0a5888f9534c475c0b5c0d6fa6d4920e1a50a89517b67872f8b6ca51ef166
|
| 3 |
+
size 3579
|
checkpoint-15600/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-15600/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a3e2c52e3ad73c5690c7fd3988703d0851fa21fb94faec91a09af579aac43b5
|
| 3 |
+
size 883771077
|
checkpoint-15600/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cebba5a25a952644988cd41329625765366452b9a418c608d86f9668c91ab9be
|
| 3 |
+
size 441897977
|
checkpoint-15600/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63eefde296117e60cfadb90b3252b400980366f430405f99c807b48bd3bd38d8
|
| 3 |
+
size 14511
|
checkpoint-15600/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53a2e3ddad874ea5cbb036b763892d0dc9047925ff82b6fb038844dd5f04a3d3
|
| 3 |
+
size 557
|
checkpoint-15600/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d702af4a167ac36439f4e82ab03ccb8e03403b3b51b549f5b1b11906a9342808
|
| 3 |
+
size 627
|
checkpoint-15600/trainer_state.json
ADDED
|
@@ -0,0 +1,952 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 121.86905916585839,
|
| 5 |
+
"global_step": 15600,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.78,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 10.1419,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.56,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 9.2788,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 2.34,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 8.8011,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 3.12,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 8.3846,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.9,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 7.8214,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.68,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 7.3094,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5.47,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 6.7911,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 6.25,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 6.4144,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 7.03,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 6.2168,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 7.81,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 6.0642,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 8.59,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 6.0511,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 9.37,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 5.995,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 10.16,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 5.9539,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 10.93,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 5.8675,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 11.71,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 5.8874,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 12.5,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 5.8578,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 13.28,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 5.8386,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 14.06,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 5.8138,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 14.84,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 5.7399,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 15.62,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 5.7753,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 16.4,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 5.7564,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 17.19,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 5.738,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 17.96,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 5.6753,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 18.74,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 5.7082,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 19.53,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 5.6991,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 20.31,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 5.6801,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 21.09,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 5.6692,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 21.87,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 5.6063,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 22.65,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 5.6445,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 23.43,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 5.6328,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 24.22,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 5.6217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 24.99,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 5.5601,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 25.78,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 5.5976,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 26.56,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 5.5911,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 27.34,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 5.5738,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 28.12,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 5.566,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 28.9,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 5.5071,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 29.68,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 5.5438,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 30.47,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 5.5366,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 31.25,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 5.5268,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 32.03,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 5.517,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 32.81,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 5.4574,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 33.59,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 5.5002,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 34.37,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 5.4887,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 35.16,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 5.4805,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 35.93,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 5.4265,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 36.71,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 5.4615,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 37.5,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 5.4576,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 38.28,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 5.4421,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 39.06,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 5.4342,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 39.84,
|
| 312 |
+
"learning_rate": 5.1000000000000006e-05,
|
| 313 |
+
"loss": 5.3641,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 40.62,
|
| 318 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 319 |
+
"loss": 5.379,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 41.4,
|
| 324 |
+
"learning_rate": 5.300000000000001e-05,
|
| 325 |
+
"loss": 5.3638,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 42.19,
|
| 330 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 331 |
+
"loss": 5.3441,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 42.96,
|
| 336 |
+
"learning_rate": 5.500000000000001e-05,
|
| 337 |
+
"loss": 5.2759,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 43.74,
|
| 342 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 343 |
+
"loss": 5.3011,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 44.53,
|
| 348 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 349 |
+
"loss": 5.2758,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 45.31,
|
| 354 |
+
"learning_rate": 5.8e-05,
|
| 355 |
+
"loss": 5.2559,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 46.09,
|
| 360 |
+
"learning_rate": 5.9e-05,
|
| 361 |
+
"loss": 5.2326,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 46.87,
|
| 366 |
+
"learning_rate": 6e-05,
|
| 367 |
+
"loss": 5.1616,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 47.65,
|
| 372 |
+
"learning_rate": 6.1e-05,
|
| 373 |
+
"loss": 5.1753,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 48.43,
|
| 378 |
+
"learning_rate": 6.2e-05,
|
| 379 |
+
"loss": 5.1378,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 49.22,
|
| 384 |
+
"learning_rate": 6.3e-05,
|
| 385 |
+
"loss": 5.1,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 49.99,
|
| 390 |
+
"learning_rate": 6.400000000000001e-05,
|
| 391 |
+
"loss": 5.015,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 50.78,
|
| 396 |
+
"learning_rate": 6.500000000000001e-05,
|
| 397 |
+
"loss": 4.9758,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 51.56,
|
| 402 |
+
"learning_rate": 6.6e-05,
|
| 403 |
+
"loss": 4.8417,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 52.34,
|
| 408 |
+
"learning_rate": 6.7e-05,
|
| 409 |
+
"loss": 4.7116,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 53.12,
|
| 414 |
+
"learning_rate": 6.800000000000001e-05,
|
| 415 |
+
"loss": 4.5582,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 53.9,
|
| 420 |
+
"learning_rate": 6.9e-05,
|
| 421 |
+
"loss": 4.3437,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 54.68,
|
| 426 |
+
"learning_rate": 7e-05,
|
| 427 |
+
"loss": 4.2114,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 55.47,
|
| 432 |
+
"learning_rate": 7.1e-05,
|
| 433 |
+
"loss": 4.1021,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 56.25,
|
| 438 |
+
"learning_rate": 7.2e-05,
|
| 439 |
+
"loss": 4.0074,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 57.03,
|
| 444 |
+
"learning_rate": 7.3e-05,
|
| 445 |
+
"loss": 3.9346,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 57.81,
|
| 450 |
+
"learning_rate": 7.4e-05,
|
| 451 |
+
"loss": 3.8289,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 58.59,
|
| 456 |
+
"learning_rate": 7.500000000000001e-05,
|
| 457 |
+
"loss": 3.8105,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 59.37,
|
| 462 |
+
"learning_rate": 7.6e-05,
|
| 463 |
+
"loss": 3.755,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 60.16,
|
| 468 |
+
"learning_rate": 7.7e-05,
|
| 469 |
+
"loss": 3.7105,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 60.93,
|
| 474 |
+
"learning_rate": 7.800000000000001e-05,
|
| 475 |
+
"loss": 3.6394,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 61.71,
|
| 480 |
+
"learning_rate": 7.900000000000001e-05,
|
| 481 |
+
"loss": 3.6262,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 62.5,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 3.5924,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 63.28,
|
| 492 |
+
"learning_rate": 8.1e-05,
|
| 493 |
+
"loss": 3.558,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 64.06,
|
| 498 |
+
"learning_rate": 8.2e-05,
|
| 499 |
+
"loss": 3.5255,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 64.84,
|
| 504 |
+
"learning_rate": 8.3e-05,
|
| 505 |
+
"loss": 3.4602,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 65.62,
|
| 510 |
+
"learning_rate": 8.4e-05,
|
| 511 |
+
"loss": 3.4641,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 66.4,
|
| 516 |
+
"learning_rate": 8.5e-05,
|
| 517 |
+
"loss": 3.435,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 67.19,
|
| 522 |
+
"learning_rate": 8.6e-05,
|
| 523 |
+
"loss": 3.408,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 67.96,
|
| 528 |
+
"learning_rate": 8.7e-05,
|
| 529 |
+
"loss": 3.3594,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 68.74,
|
| 534 |
+
"learning_rate": 8.800000000000001e-05,
|
| 535 |
+
"loss": 3.3593,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 69.53,
|
| 540 |
+
"learning_rate": 8.900000000000001e-05,
|
| 541 |
+
"loss": 3.3372,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 70.31,
|
| 546 |
+
"learning_rate": 9e-05,
|
| 547 |
+
"loss": 3.3217,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 71.09,
|
| 552 |
+
"learning_rate": 9.1e-05,
|
| 553 |
+
"loss": 3.2985,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 71.87,
|
| 558 |
+
"learning_rate": 9.200000000000001e-05,
|
| 559 |
+
"loss": 3.2509,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 72.65,
|
| 564 |
+
"learning_rate": 9.300000000000001e-05,
|
| 565 |
+
"loss": 3.2584,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 73.43,
|
| 570 |
+
"learning_rate": 9.4e-05,
|
| 571 |
+
"loss": 3.2386,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 74.22,
|
| 576 |
+
"learning_rate": 9.5e-05,
|
| 577 |
+
"loss": 3.2232,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 74.99,
|
| 582 |
+
"learning_rate": 9.6e-05,
|
| 583 |
+
"loss": 3.1786,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 75.78,
|
| 588 |
+
"learning_rate": 9.7e-05,
|
| 589 |
+
"loss": 3.1855,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 76.56,
|
| 594 |
+
"learning_rate": 9.8e-05,
|
| 595 |
+
"loss": 3.1737,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 77.34,
|
| 600 |
+
"learning_rate": 9.900000000000001e-05,
|
| 601 |
+
"loss": 3.1565,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 78.12,
|
| 606 |
+
"learning_rate": 0.0001,
|
| 607 |
+
"loss": 3.1442,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 78.9,
|
| 612 |
+
"learning_rate": 9.904580152671757e-05,
|
| 613 |
+
"loss": 3.1003,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 79.68,
|
| 618 |
+
"learning_rate": 9.809160305343512e-05,
|
| 619 |
+
"loss": 3.1137,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 80.47,
|
| 624 |
+
"learning_rate": 9.713740458015268e-05,
|
| 625 |
+
"loss": 3.0958,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 81.25,
|
| 630 |
+
"learning_rate": 9.618320610687024e-05,
|
| 631 |
+
"loss": 3.0853,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 82.03,
|
| 636 |
+
"learning_rate": 9.522900763358779e-05,
|
| 637 |
+
"loss": 3.0704,
|
| 638 |
+
"step": 10500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 82.81,
|
| 642 |
+
"learning_rate": 9.427480916030534e-05,
|
| 643 |
+
"loss": 3.03,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 83.59,
|
| 648 |
+
"learning_rate": 9.33206106870229e-05,
|
| 649 |
+
"loss": 3.0428,
|
| 650 |
+
"step": 10700
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 84.37,
|
| 654 |
+
"learning_rate": 9.236641221374047e-05,
|
| 655 |
+
"loss": 3.0299,
|
| 656 |
+
"step": 10800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 85.16,
|
| 660 |
+
"learning_rate": 9.141221374045802e-05,
|
| 661 |
+
"loss": 3.0239,
|
| 662 |
+
"step": 10900
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 85.93,
|
| 666 |
+
"learning_rate": 9.045801526717558e-05,
|
| 667 |
+
"loss": 2.9818,
|
| 668 |
+
"step": 11000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 86.71,
|
| 672 |
+
"learning_rate": 8.950381679389314e-05,
|
| 673 |
+
"loss": 2.9967,
|
| 674 |
+
"step": 11100
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 87.5,
|
| 678 |
+
"learning_rate": 8.854961832061069e-05,
|
| 679 |
+
"loss": 2.9866,
|
| 680 |
+
"step": 11200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 88.28,
|
| 684 |
+
"learning_rate": 8.759541984732825e-05,
|
| 685 |
+
"loss": 2.9758,
|
| 686 |
+
"step": 11300
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 89.06,
|
| 690 |
+
"learning_rate": 8.664122137404582e-05,
|
| 691 |
+
"loss": 2.968,
|
| 692 |
+
"step": 11400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 89.84,
|
| 696 |
+
"learning_rate": 8.568702290076335e-05,
|
| 697 |
+
"loss": 2.9276,
|
| 698 |
+
"step": 11500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 90.62,
|
| 702 |
+
"learning_rate": 8.473282442748092e-05,
|
| 703 |
+
"loss": 2.9464,
|
| 704 |
+
"step": 11600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 91.4,
|
| 708 |
+
"learning_rate": 8.377862595419848e-05,
|
| 709 |
+
"loss": 2.9342,
|
| 710 |
+
"step": 11700
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 92.19,
|
| 714 |
+
"learning_rate": 8.282442748091603e-05,
|
| 715 |
+
"loss": 2.9313,
|
| 716 |
+
"step": 11800
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 92.96,
|
| 720 |
+
"learning_rate": 8.187022900763359e-05,
|
| 721 |
+
"loss": 2.8932,
|
| 722 |
+
"step": 11900
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 93.74,
|
| 726 |
+
"learning_rate": 8.091603053435115e-05,
|
| 727 |
+
"loss": 2.9087,
|
| 728 |
+
"step": 12000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 94.53,
|
| 732 |
+
"learning_rate": 7.996183206106872e-05,
|
| 733 |
+
"loss": 2.9054,
|
| 734 |
+
"step": 12100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 95.31,
|
| 738 |
+
"learning_rate": 7.900763358778626e-05,
|
| 739 |
+
"loss": 2.8916,
|
| 740 |
+
"step": 12200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 96.09,
|
| 744 |
+
"learning_rate": 7.805343511450383e-05,
|
| 745 |
+
"loss": 2.8888,
|
| 746 |
+
"step": 12300
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 96.87,
|
| 750 |
+
"learning_rate": 7.709923664122138e-05,
|
| 751 |
+
"loss": 2.8505,
|
| 752 |
+
"step": 12400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 97.65,
|
| 756 |
+
"learning_rate": 7.614503816793893e-05,
|
| 757 |
+
"loss": 2.8729,
|
| 758 |
+
"step": 12500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 98.43,
|
| 762 |
+
"learning_rate": 7.519083969465649e-05,
|
| 763 |
+
"loss": 2.8576,
|
| 764 |
+
"step": 12600
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 99.22,
|
| 768 |
+
"learning_rate": 7.423664122137405e-05,
|
| 769 |
+
"loss": 2.8569,
|
| 770 |
+
"step": 12700
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 99.99,
|
| 774 |
+
"learning_rate": 7.32824427480916e-05,
|
| 775 |
+
"loss": 2.8276,
|
| 776 |
+
"step": 12800
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 100.78,
|
| 780 |
+
"learning_rate": 7.232824427480916e-05,
|
| 781 |
+
"loss": 2.8404,
|
| 782 |
+
"step": 12900
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 101.56,
|
| 786 |
+
"learning_rate": 7.137404580152673e-05,
|
| 787 |
+
"loss": 2.8337,
|
| 788 |
+
"step": 13000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 102.34,
|
| 792 |
+
"learning_rate": 7.041984732824428e-05,
|
| 793 |
+
"loss": 2.8285,
|
| 794 |
+
"step": 13100
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 103.12,
|
| 798 |
+
"learning_rate": 6.946564885496184e-05,
|
| 799 |
+
"loss": 2.8209,
|
| 800 |
+
"step": 13200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 103.9,
|
| 804 |
+
"learning_rate": 6.851145038167939e-05,
|
| 805 |
+
"loss": 2.7906,
|
| 806 |
+
"step": 13300
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 104.68,
|
| 810 |
+
"learning_rate": 6.755725190839695e-05,
|
| 811 |
+
"loss": 2.8055,
|
| 812 |
+
"step": 13400
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 105.47,
|
| 816 |
+
"learning_rate": 6.66030534351145e-05,
|
| 817 |
+
"loss": 2.8034,
|
| 818 |
+
"step": 13500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 106.25,
|
| 822 |
+
"learning_rate": 6.564885496183206e-05,
|
| 823 |
+
"loss": 2.7957,
|
| 824 |
+
"step": 13600
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 107.03,
|
| 828 |
+
"learning_rate": 6.469465648854963e-05,
|
| 829 |
+
"loss": 2.7888,
|
| 830 |
+
"step": 13700
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 107.81,
|
| 834 |
+
"learning_rate": 6.374045801526718e-05,
|
| 835 |
+
"loss": 2.7596,
|
| 836 |
+
"step": 13800
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 108.59,
|
| 840 |
+
"learning_rate": 6.278625954198474e-05,
|
| 841 |
+
"loss": 2.7781,
|
| 842 |
+
"step": 13900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 109.37,
|
| 846 |
+
"learning_rate": 6.18320610687023e-05,
|
| 847 |
+
"loss": 2.7737,
|
| 848 |
+
"step": 14000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 110.16,
|
| 852 |
+
"learning_rate": 6.087786259541986e-05,
|
| 853 |
+
"loss": 2.7692,
|
| 854 |
+
"step": 14100
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 110.93,
|
| 858 |
+
"learning_rate": 5.992366412213741e-05,
|
| 859 |
+
"loss": 2.7379,
|
| 860 |
+
"step": 14200
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 111.71,
|
| 864 |
+
"learning_rate": 5.897900763358779e-05,
|
| 865 |
+
"loss": 2.754,
|
| 866 |
+
"step": 14300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 112.5,
|
| 870 |
+
"learning_rate": 5.802480916030535e-05,
|
| 871 |
+
"loss": 2.7503,
|
| 872 |
+
"step": 14400
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 113.28,
|
| 876 |
+
"learning_rate": 5.707061068702291e-05,
|
| 877 |
+
"loss": 2.7488,
|
| 878 |
+
"step": 14500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 114.06,
|
| 882 |
+
"learning_rate": 5.6116412213740466e-05,
|
| 883 |
+
"loss": 2.7461,
|
| 884 |
+
"step": 14600
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 114.84,
|
| 888 |
+
"learning_rate": 5.5162213740458016e-05,
|
| 889 |
+
"loss": 2.711,
|
| 890 |
+
"step": 14700
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 115.62,
|
| 894 |
+
"learning_rate": 5.420801526717557e-05,
|
| 895 |
+
"loss": 2.7314,
|
| 896 |
+
"step": 14800
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 116.4,
|
| 900 |
+
"learning_rate": 5.325381679389313e-05,
|
| 901 |
+
"loss": 2.7222,
|
| 902 |
+
"step": 14900
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 117.19,
|
| 906 |
+
"learning_rate": 5.229961832061069e-05,
|
| 907 |
+
"loss": 2.7267,
|
| 908 |
+
"step": 15000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 117.96,
|
| 912 |
+
"learning_rate": 5.134541984732825e-05,
|
| 913 |
+
"loss": 2.6951,
|
| 914 |
+
"step": 15100
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 118.74,
|
| 918 |
+
"learning_rate": 5.03912213740458e-05,
|
| 919 |
+
"loss": 2.7121,
|
| 920 |
+
"step": 15200
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 119.53,
|
| 924 |
+
"learning_rate": 4.9437022900763366e-05,
|
| 925 |
+
"loss": 2.7082,
|
| 926 |
+
"step": 15300
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 120.31,
|
| 930 |
+
"learning_rate": 4.8482824427480915e-05,
|
| 931 |
+
"loss": 2.7073,
|
| 932 |
+
"step": 15400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 121.09,
|
| 936 |
+
"learning_rate": 4.752862595419848e-05,
|
| 937 |
+
"loss": 2.7014,
|
| 938 |
+
"step": 15500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 121.87,
|
| 942 |
+
"learning_rate": 4.6574427480916034e-05,
|
| 943 |
+
"loss": 2.6753,
|
| 944 |
+
"step": 15600
|
| 945 |
+
}
|
| 946 |
+
],
|
| 947 |
+
"max_steps": 20480,
|
| 948 |
+
"num_train_epochs": 160,
|
| 949 |
+
"total_flos": 1.0575225393512448e+18,
|
| 950 |
+
"trial_name": null,
|
| 951 |
+
"trial_params": null
|
| 952 |
+
}
|
checkpoint-15600/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6de0a5888f9534c475c0b5c0d6fa6d4920e1a50a89517b67872f8b6ca51ef166
|
| 3 |
+
size 3579
|
checkpoint-18000/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-18000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6dfa9c28077c99d33b65f5ddfbc21d4cb5871a1286d2790a94dd3f7535345c55
|
| 3 |
+
size 883771077
|
checkpoint-18000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3cc69590fd21b64297c3aa0fa3a6c1911177fc720e054a4f263184ae15e5ea8f
|
| 3 |
+
size 441897977
|
checkpoint-18000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd2296218597355b0acd06d0a718241ed040d575334015384fa29a54ab36abac
|
| 3 |
+
size 14511
|
checkpoint-18000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31d90cd2d5e2b92e6c934619c0740e3cb9ddb00f99c21eae5a2b88210b3ba9a6
|
| 3 |
+
size 557
|
checkpoint-18000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8376a47017036f00282ed5a674372dba78d59d5b7cbf65e5098ea8802090de34
|
| 3 |
+
size 627
|
checkpoint-18000/trainer_state.json
ADDED
|
@@ -0,0 +1,1096 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 140.6207565470417,
|
| 5 |
+
"global_step": 18000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.78,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 10.1419,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.56,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 9.2788,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 2.34,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 8.8011,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 3.12,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 8.3846,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.9,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 7.8214,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.68,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 7.3094,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5.47,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 6.7911,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 6.25,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 6.4144,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 7.03,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 6.2168,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 7.81,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 6.0642,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 8.59,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 6.0511,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 9.37,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 5.995,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 10.16,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 5.9539,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 10.93,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 5.8675,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 11.71,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 5.8874,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 12.5,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 5.8578,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 13.28,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 5.8386,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 14.06,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 5.8138,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 14.84,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 5.7399,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 15.62,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 5.7753,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 16.4,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 5.7564,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 17.19,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 5.738,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 17.96,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 5.6753,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 18.74,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 5.7082,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 19.53,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 5.6991,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 20.31,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 5.6801,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 21.09,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 5.6692,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 21.87,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 5.6063,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 22.65,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 5.6445,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 23.43,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 5.6328,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 24.22,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 5.6217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 24.99,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 5.5601,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 25.78,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 5.5976,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 26.56,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 5.5911,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 27.34,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 5.5738,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 28.12,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 5.566,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 28.9,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 5.5071,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 29.68,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 5.5438,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 30.47,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 5.5366,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 31.25,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 5.5268,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 32.03,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 5.517,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 32.81,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 5.4574,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 33.59,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 5.5002,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 34.37,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 5.4887,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 35.16,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 5.4805,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 35.93,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 5.4265,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 36.71,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 5.4615,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 37.5,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 5.4576,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 38.28,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 5.4421,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 39.06,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 5.4342,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 39.84,
|
| 312 |
+
"learning_rate": 5.1000000000000006e-05,
|
| 313 |
+
"loss": 5.3641,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 40.62,
|
| 318 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 319 |
+
"loss": 5.379,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 41.4,
|
| 324 |
+
"learning_rate": 5.300000000000001e-05,
|
| 325 |
+
"loss": 5.3638,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 42.19,
|
| 330 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 331 |
+
"loss": 5.3441,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 42.96,
|
| 336 |
+
"learning_rate": 5.500000000000001e-05,
|
| 337 |
+
"loss": 5.2759,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 43.74,
|
| 342 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 343 |
+
"loss": 5.3011,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 44.53,
|
| 348 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 349 |
+
"loss": 5.2758,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 45.31,
|
| 354 |
+
"learning_rate": 5.8e-05,
|
| 355 |
+
"loss": 5.2559,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 46.09,
|
| 360 |
+
"learning_rate": 5.9e-05,
|
| 361 |
+
"loss": 5.2326,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 46.87,
|
| 366 |
+
"learning_rate": 6e-05,
|
| 367 |
+
"loss": 5.1616,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 47.65,
|
| 372 |
+
"learning_rate": 6.1e-05,
|
| 373 |
+
"loss": 5.1753,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 48.43,
|
| 378 |
+
"learning_rate": 6.2e-05,
|
| 379 |
+
"loss": 5.1378,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 49.22,
|
| 384 |
+
"learning_rate": 6.3e-05,
|
| 385 |
+
"loss": 5.1,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 49.99,
|
| 390 |
+
"learning_rate": 6.400000000000001e-05,
|
| 391 |
+
"loss": 5.015,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 50.78,
|
| 396 |
+
"learning_rate": 6.500000000000001e-05,
|
| 397 |
+
"loss": 4.9758,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 51.56,
|
| 402 |
+
"learning_rate": 6.6e-05,
|
| 403 |
+
"loss": 4.8417,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 52.34,
|
| 408 |
+
"learning_rate": 6.7e-05,
|
| 409 |
+
"loss": 4.7116,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 53.12,
|
| 414 |
+
"learning_rate": 6.800000000000001e-05,
|
| 415 |
+
"loss": 4.5582,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 53.9,
|
| 420 |
+
"learning_rate": 6.9e-05,
|
| 421 |
+
"loss": 4.3437,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 54.68,
|
| 426 |
+
"learning_rate": 7e-05,
|
| 427 |
+
"loss": 4.2114,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 55.47,
|
| 432 |
+
"learning_rate": 7.1e-05,
|
| 433 |
+
"loss": 4.1021,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 56.25,
|
| 438 |
+
"learning_rate": 7.2e-05,
|
| 439 |
+
"loss": 4.0074,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 57.03,
|
| 444 |
+
"learning_rate": 7.3e-05,
|
| 445 |
+
"loss": 3.9346,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 57.81,
|
| 450 |
+
"learning_rate": 7.4e-05,
|
| 451 |
+
"loss": 3.8289,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 58.59,
|
| 456 |
+
"learning_rate": 7.500000000000001e-05,
|
| 457 |
+
"loss": 3.8105,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 59.37,
|
| 462 |
+
"learning_rate": 7.6e-05,
|
| 463 |
+
"loss": 3.755,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 60.16,
|
| 468 |
+
"learning_rate": 7.7e-05,
|
| 469 |
+
"loss": 3.7105,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 60.93,
|
| 474 |
+
"learning_rate": 7.800000000000001e-05,
|
| 475 |
+
"loss": 3.6394,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 61.71,
|
| 480 |
+
"learning_rate": 7.900000000000001e-05,
|
| 481 |
+
"loss": 3.6262,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 62.5,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 3.5924,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 63.28,
|
| 492 |
+
"learning_rate": 8.1e-05,
|
| 493 |
+
"loss": 3.558,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 64.06,
|
| 498 |
+
"learning_rate": 8.2e-05,
|
| 499 |
+
"loss": 3.5255,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 64.84,
|
| 504 |
+
"learning_rate": 8.3e-05,
|
| 505 |
+
"loss": 3.4602,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 65.62,
|
| 510 |
+
"learning_rate": 8.4e-05,
|
| 511 |
+
"loss": 3.4641,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 66.4,
|
| 516 |
+
"learning_rate": 8.5e-05,
|
| 517 |
+
"loss": 3.435,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 67.19,
|
| 522 |
+
"learning_rate": 8.6e-05,
|
| 523 |
+
"loss": 3.408,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 67.96,
|
| 528 |
+
"learning_rate": 8.7e-05,
|
| 529 |
+
"loss": 3.3594,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 68.74,
|
| 534 |
+
"learning_rate": 8.800000000000001e-05,
|
| 535 |
+
"loss": 3.3593,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 69.53,
|
| 540 |
+
"learning_rate": 8.900000000000001e-05,
|
| 541 |
+
"loss": 3.3372,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 70.31,
|
| 546 |
+
"learning_rate": 9e-05,
|
| 547 |
+
"loss": 3.3217,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 71.09,
|
| 552 |
+
"learning_rate": 9.1e-05,
|
| 553 |
+
"loss": 3.2985,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 71.87,
|
| 558 |
+
"learning_rate": 9.200000000000001e-05,
|
| 559 |
+
"loss": 3.2509,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 72.65,
|
| 564 |
+
"learning_rate": 9.300000000000001e-05,
|
| 565 |
+
"loss": 3.2584,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 73.43,
|
| 570 |
+
"learning_rate": 9.4e-05,
|
| 571 |
+
"loss": 3.2386,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 74.22,
|
| 576 |
+
"learning_rate": 9.5e-05,
|
| 577 |
+
"loss": 3.2232,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 74.99,
|
| 582 |
+
"learning_rate": 9.6e-05,
|
| 583 |
+
"loss": 3.1786,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 75.78,
|
| 588 |
+
"learning_rate": 9.7e-05,
|
| 589 |
+
"loss": 3.1855,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 76.56,
|
| 594 |
+
"learning_rate": 9.8e-05,
|
| 595 |
+
"loss": 3.1737,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 77.34,
|
| 600 |
+
"learning_rate": 9.900000000000001e-05,
|
| 601 |
+
"loss": 3.1565,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 78.12,
|
| 606 |
+
"learning_rate": 0.0001,
|
| 607 |
+
"loss": 3.1442,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 78.9,
|
| 612 |
+
"learning_rate": 9.904580152671757e-05,
|
| 613 |
+
"loss": 3.1003,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 79.68,
|
| 618 |
+
"learning_rate": 9.809160305343512e-05,
|
| 619 |
+
"loss": 3.1137,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 80.47,
|
| 624 |
+
"learning_rate": 9.713740458015268e-05,
|
| 625 |
+
"loss": 3.0958,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 81.25,
|
| 630 |
+
"learning_rate": 9.618320610687024e-05,
|
| 631 |
+
"loss": 3.0853,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 82.03,
|
| 636 |
+
"learning_rate": 9.522900763358779e-05,
|
| 637 |
+
"loss": 3.0704,
|
| 638 |
+
"step": 10500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 82.81,
|
| 642 |
+
"learning_rate": 9.427480916030534e-05,
|
| 643 |
+
"loss": 3.03,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 83.59,
|
| 648 |
+
"learning_rate": 9.33206106870229e-05,
|
| 649 |
+
"loss": 3.0428,
|
| 650 |
+
"step": 10700
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 84.37,
|
| 654 |
+
"learning_rate": 9.236641221374047e-05,
|
| 655 |
+
"loss": 3.0299,
|
| 656 |
+
"step": 10800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 85.16,
|
| 660 |
+
"learning_rate": 9.141221374045802e-05,
|
| 661 |
+
"loss": 3.0239,
|
| 662 |
+
"step": 10900
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 85.93,
|
| 666 |
+
"learning_rate": 9.045801526717558e-05,
|
| 667 |
+
"loss": 2.9818,
|
| 668 |
+
"step": 11000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 86.71,
|
| 672 |
+
"learning_rate": 8.950381679389314e-05,
|
| 673 |
+
"loss": 2.9967,
|
| 674 |
+
"step": 11100
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 87.5,
|
| 678 |
+
"learning_rate": 8.854961832061069e-05,
|
| 679 |
+
"loss": 2.9866,
|
| 680 |
+
"step": 11200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 88.28,
|
| 684 |
+
"learning_rate": 8.759541984732825e-05,
|
| 685 |
+
"loss": 2.9758,
|
| 686 |
+
"step": 11300
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 89.06,
|
| 690 |
+
"learning_rate": 8.664122137404582e-05,
|
| 691 |
+
"loss": 2.968,
|
| 692 |
+
"step": 11400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 89.84,
|
| 696 |
+
"learning_rate": 8.568702290076335e-05,
|
| 697 |
+
"loss": 2.9276,
|
| 698 |
+
"step": 11500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 90.62,
|
| 702 |
+
"learning_rate": 8.473282442748092e-05,
|
| 703 |
+
"loss": 2.9464,
|
| 704 |
+
"step": 11600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 91.4,
|
| 708 |
+
"learning_rate": 8.377862595419848e-05,
|
| 709 |
+
"loss": 2.9342,
|
| 710 |
+
"step": 11700
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 92.19,
|
| 714 |
+
"learning_rate": 8.282442748091603e-05,
|
| 715 |
+
"loss": 2.9313,
|
| 716 |
+
"step": 11800
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 92.96,
|
| 720 |
+
"learning_rate": 8.187022900763359e-05,
|
| 721 |
+
"loss": 2.8932,
|
| 722 |
+
"step": 11900
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 93.74,
|
| 726 |
+
"learning_rate": 8.091603053435115e-05,
|
| 727 |
+
"loss": 2.9087,
|
| 728 |
+
"step": 12000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 94.53,
|
| 732 |
+
"learning_rate": 7.996183206106872e-05,
|
| 733 |
+
"loss": 2.9054,
|
| 734 |
+
"step": 12100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 95.31,
|
| 738 |
+
"learning_rate": 7.900763358778626e-05,
|
| 739 |
+
"loss": 2.8916,
|
| 740 |
+
"step": 12200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 96.09,
|
| 744 |
+
"learning_rate": 7.805343511450383e-05,
|
| 745 |
+
"loss": 2.8888,
|
| 746 |
+
"step": 12300
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 96.87,
|
| 750 |
+
"learning_rate": 7.709923664122138e-05,
|
| 751 |
+
"loss": 2.8505,
|
| 752 |
+
"step": 12400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 97.65,
|
| 756 |
+
"learning_rate": 7.614503816793893e-05,
|
| 757 |
+
"loss": 2.8729,
|
| 758 |
+
"step": 12500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 98.43,
|
| 762 |
+
"learning_rate": 7.519083969465649e-05,
|
| 763 |
+
"loss": 2.8576,
|
| 764 |
+
"step": 12600
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 99.22,
|
| 768 |
+
"learning_rate": 7.423664122137405e-05,
|
| 769 |
+
"loss": 2.8569,
|
| 770 |
+
"step": 12700
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 99.99,
|
| 774 |
+
"learning_rate": 7.32824427480916e-05,
|
| 775 |
+
"loss": 2.8276,
|
| 776 |
+
"step": 12800
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 100.78,
|
| 780 |
+
"learning_rate": 7.232824427480916e-05,
|
| 781 |
+
"loss": 2.8404,
|
| 782 |
+
"step": 12900
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 101.56,
|
| 786 |
+
"learning_rate": 7.137404580152673e-05,
|
| 787 |
+
"loss": 2.8337,
|
| 788 |
+
"step": 13000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 102.34,
|
| 792 |
+
"learning_rate": 7.041984732824428e-05,
|
| 793 |
+
"loss": 2.8285,
|
| 794 |
+
"step": 13100
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 103.12,
|
| 798 |
+
"learning_rate": 6.946564885496184e-05,
|
| 799 |
+
"loss": 2.8209,
|
| 800 |
+
"step": 13200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 103.9,
|
| 804 |
+
"learning_rate": 6.851145038167939e-05,
|
| 805 |
+
"loss": 2.7906,
|
| 806 |
+
"step": 13300
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 104.68,
|
| 810 |
+
"learning_rate": 6.755725190839695e-05,
|
| 811 |
+
"loss": 2.8055,
|
| 812 |
+
"step": 13400
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 105.47,
|
| 816 |
+
"learning_rate": 6.66030534351145e-05,
|
| 817 |
+
"loss": 2.8034,
|
| 818 |
+
"step": 13500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 106.25,
|
| 822 |
+
"learning_rate": 6.564885496183206e-05,
|
| 823 |
+
"loss": 2.7957,
|
| 824 |
+
"step": 13600
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 107.03,
|
| 828 |
+
"learning_rate": 6.469465648854963e-05,
|
| 829 |
+
"loss": 2.7888,
|
| 830 |
+
"step": 13700
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 107.81,
|
| 834 |
+
"learning_rate": 6.374045801526718e-05,
|
| 835 |
+
"loss": 2.7596,
|
| 836 |
+
"step": 13800
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 108.59,
|
| 840 |
+
"learning_rate": 6.278625954198474e-05,
|
| 841 |
+
"loss": 2.7781,
|
| 842 |
+
"step": 13900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 109.37,
|
| 846 |
+
"learning_rate": 6.18320610687023e-05,
|
| 847 |
+
"loss": 2.7737,
|
| 848 |
+
"step": 14000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 110.16,
|
| 852 |
+
"learning_rate": 6.087786259541986e-05,
|
| 853 |
+
"loss": 2.7692,
|
| 854 |
+
"step": 14100
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 110.93,
|
| 858 |
+
"learning_rate": 5.992366412213741e-05,
|
| 859 |
+
"loss": 2.7379,
|
| 860 |
+
"step": 14200
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 111.71,
|
| 864 |
+
"learning_rate": 5.897900763358779e-05,
|
| 865 |
+
"loss": 2.754,
|
| 866 |
+
"step": 14300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 112.5,
|
| 870 |
+
"learning_rate": 5.802480916030535e-05,
|
| 871 |
+
"loss": 2.7503,
|
| 872 |
+
"step": 14400
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 113.28,
|
| 876 |
+
"learning_rate": 5.707061068702291e-05,
|
| 877 |
+
"loss": 2.7488,
|
| 878 |
+
"step": 14500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 114.06,
|
| 882 |
+
"learning_rate": 5.6116412213740466e-05,
|
| 883 |
+
"loss": 2.7461,
|
| 884 |
+
"step": 14600
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 114.84,
|
| 888 |
+
"learning_rate": 5.5162213740458016e-05,
|
| 889 |
+
"loss": 2.711,
|
| 890 |
+
"step": 14700
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 115.62,
|
| 894 |
+
"learning_rate": 5.420801526717557e-05,
|
| 895 |
+
"loss": 2.7314,
|
| 896 |
+
"step": 14800
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 116.4,
|
| 900 |
+
"learning_rate": 5.325381679389313e-05,
|
| 901 |
+
"loss": 2.7222,
|
| 902 |
+
"step": 14900
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 117.19,
|
| 906 |
+
"learning_rate": 5.229961832061069e-05,
|
| 907 |
+
"loss": 2.7267,
|
| 908 |
+
"step": 15000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 117.96,
|
| 912 |
+
"learning_rate": 5.134541984732825e-05,
|
| 913 |
+
"loss": 2.6951,
|
| 914 |
+
"step": 15100
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 118.74,
|
| 918 |
+
"learning_rate": 5.03912213740458e-05,
|
| 919 |
+
"loss": 2.7121,
|
| 920 |
+
"step": 15200
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 119.53,
|
| 924 |
+
"learning_rate": 4.9437022900763366e-05,
|
| 925 |
+
"loss": 2.7082,
|
| 926 |
+
"step": 15300
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 120.31,
|
| 930 |
+
"learning_rate": 4.8482824427480915e-05,
|
| 931 |
+
"loss": 2.7073,
|
| 932 |
+
"step": 15400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 121.09,
|
| 936 |
+
"learning_rate": 4.752862595419848e-05,
|
| 937 |
+
"loss": 2.7014,
|
| 938 |
+
"step": 15500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 121.87,
|
| 942 |
+
"learning_rate": 4.6574427480916034e-05,
|
| 943 |
+
"loss": 2.6753,
|
| 944 |
+
"step": 15600
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 122.65,
|
| 948 |
+
"learning_rate": 4.562022900763359e-05,
|
| 949 |
+
"loss": 2.6895,
|
| 950 |
+
"step": 15700
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 123.43,
|
| 954 |
+
"learning_rate": 4.466603053435115e-05,
|
| 955 |
+
"loss": 2.693,
|
| 956 |
+
"step": 15800
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 124.22,
|
| 960 |
+
"learning_rate": 4.37118320610687e-05,
|
| 961 |
+
"loss": 2.6863,
|
| 962 |
+
"step": 15900
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 124.99,
|
| 966 |
+
"learning_rate": 4.275763358778626e-05,
|
| 967 |
+
"loss": 2.6588,
|
| 968 |
+
"step": 16000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 125.78,
|
| 972 |
+
"learning_rate": 4.180343511450382e-05,
|
| 973 |
+
"loss": 2.6755,
|
| 974 |
+
"step": 16100
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 126.56,
|
| 978 |
+
"learning_rate": 4.084923664122138e-05,
|
| 979 |
+
"loss": 2.6724,
|
| 980 |
+
"step": 16200
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 127.34,
|
| 984 |
+
"learning_rate": 3.989503816793893e-05,
|
| 985 |
+
"loss": 2.669,
|
| 986 |
+
"step": 16300
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 128.12,
|
| 990 |
+
"learning_rate": 3.894083969465649e-05,
|
| 991 |
+
"loss": 2.6671,
|
| 992 |
+
"step": 16400
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 128.9,
|
| 996 |
+
"learning_rate": 3.7986641221374045e-05,
|
| 997 |
+
"loss": 2.6429,
|
| 998 |
+
"step": 16500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 129.68,
|
| 1002 |
+
"learning_rate": 3.703244274809161e-05,
|
| 1003 |
+
"loss": 2.6615,
|
| 1004 |
+
"step": 16600
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 130.47,
|
| 1008 |
+
"learning_rate": 3.6078244274809164e-05,
|
| 1009 |
+
"loss": 2.6553,
|
| 1010 |
+
"step": 16700
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 131.25,
|
| 1014 |
+
"learning_rate": 3.512404580152672e-05,
|
| 1015 |
+
"loss": 2.6541,
|
| 1016 |
+
"step": 16800
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 132.03,
|
| 1020 |
+
"learning_rate": 3.4169847328244276e-05,
|
| 1021 |
+
"loss": 2.6543,
|
| 1022 |
+
"step": 16900
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 132.81,
|
| 1026 |
+
"learning_rate": 3.321564885496183e-05,
|
| 1027 |
+
"loss": 2.6242,
|
| 1028 |
+
"step": 17000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 133.59,
|
| 1032 |
+
"learning_rate": 3.2261450381679395e-05,
|
| 1033 |
+
"loss": 2.6412,
|
| 1034 |
+
"step": 17100
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 134.37,
|
| 1038 |
+
"learning_rate": 3.130725190839695e-05,
|
| 1039 |
+
"loss": 2.6423,
|
| 1040 |
+
"step": 17200
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 135.16,
|
| 1044 |
+
"learning_rate": 3.0353053435114503e-05,
|
| 1045 |
+
"loss": 2.6372,
|
| 1046 |
+
"step": 17300
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 135.93,
|
| 1050 |
+
"learning_rate": 2.9398854961832063e-05,
|
| 1051 |
+
"loss": 2.6107,
|
| 1052 |
+
"step": 17400
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 136.71,
|
| 1056 |
+
"learning_rate": 2.844465648854962e-05,
|
| 1057 |
+
"loss": 2.6298,
|
| 1058 |
+
"step": 17500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 137.5,
|
| 1062 |
+
"learning_rate": 2.7490458015267178e-05,
|
| 1063 |
+
"loss": 2.6265,
|
| 1064 |
+
"step": 17600
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 138.28,
|
| 1068 |
+
"learning_rate": 2.6545801526717557e-05,
|
| 1069 |
+
"loss": 2.6274,
|
| 1070 |
+
"step": 17700
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 139.06,
|
| 1074 |
+
"learning_rate": 2.5591603053435116e-05,
|
| 1075 |
+
"loss": 2.626,
|
| 1076 |
+
"step": 17800
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 139.84,
|
| 1080 |
+
"learning_rate": 2.4637404580152672e-05,
|
| 1081 |
+
"loss": 2.6,
|
| 1082 |
+
"step": 17900
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 140.62,
|
| 1086 |
+
"learning_rate": 2.368320610687023e-05,
|
| 1087 |
+
"loss": 2.6194,
|
| 1088 |
+
"step": 18000
|
| 1089 |
+
}
|
| 1090 |
+
],
|
| 1091 |
+
"max_steps": 20480,
|
| 1092 |
+
"num_train_epochs": 160,
|
| 1093 |
+
"total_flos": 1.220238561509376e+18,
|
| 1094 |
+
"trial_name": null,
|
| 1095 |
+
"trial_params": null
|
| 1096 |
+
}
|
checkpoint-18000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6de0a5888f9534c475c0b5c0d6fa6d4920e1a50a89517b67872f8b6ca51ef166
|
| 3 |
+
size 3579
|
checkpoint-20000/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72056f8aab17063872be49b83d736ffdf2934e4b3af70af70f5d2114e10d94f3
|
| 3 |
+
size 883771077
|
checkpoint-20000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e03675839d8acc7188b3941d8a6a5da29db20e7d5e3d76c0c1e8891a1630e5b
|
| 3 |
+
size 441897977
|
checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4335a8157d5cdbb8e82c8726feef433b662b7e4743daba3cf3e93980c91f88ef
|
| 3 |
+
size 14511
|
checkpoint-20000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca93c9b6da6a63f8168493043a49578a48495785a70fd3d11703a732fa0f8e96
|
| 3 |
+
size 557
|
checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e815534a0b7ece4c95dc1b5e50be186ea45fccf76c290e2008d4dbc2b8aea43
|
| 3 |
+
size 627
|
checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,1216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 156.2483026188167,
|
| 5 |
+
"global_step": 20000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.78,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 10.1419,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.56,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 9.2788,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 2.34,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 8.8011,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 3.12,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 8.3846,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.9,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 7.8214,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 4.68,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 7.3094,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 5.47,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 6.7911,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 6.25,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 6.4144,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 7.03,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 6.2168,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 7.81,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 6.0642,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 8.59,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 6.0511,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 9.37,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 5.995,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 10.16,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 5.9539,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 10.93,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 5.8675,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 11.71,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 5.8874,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 12.5,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 5.8578,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 13.28,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 5.8386,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 14.06,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 5.8138,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 14.84,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 5.7399,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 15.62,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 5.7753,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 16.4,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 5.7564,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 17.19,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 5.738,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 17.96,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 5.6753,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 18.74,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 5.7082,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 19.53,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 5.6991,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 20.31,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 5.6801,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 21.09,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 5.6692,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 21.87,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 5.6063,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 22.65,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 5.6445,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 23.43,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 5.6328,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 24.22,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 5.6217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 24.99,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 5.5601,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 25.78,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 5.5976,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 26.56,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 5.5911,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 27.34,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 5.5738,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 28.12,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 5.566,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 28.9,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 5.5071,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 29.68,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 5.5438,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 30.47,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 5.5366,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 31.25,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 5.5268,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 32.03,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 5.517,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 32.81,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 5.4574,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 33.59,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 5.5002,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 34.37,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 5.4887,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 35.16,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 5.4805,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 35.93,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 5.4265,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 36.71,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 5.4615,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 37.5,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 5.4576,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 38.28,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 5.4421,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 39.06,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 5.4342,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 39.84,
|
| 312 |
+
"learning_rate": 5.1000000000000006e-05,
|
| 313 |
+
"loss": 5.3641,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 40.62,
|
| 318 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 319 |
+
"loss": 5.379,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 41.4,
|
| 324 |
+
"learning_rate": 5.300000000000001e-05,
|
| 325 |
+
"loss": 5.3638,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 42.19,
|
| 330 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 331 |
+
"loss": 5.3441,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 42.96,
|
| 336 |
+
"learning_rate": 5.500000000000001e-05,
|
| 337 |
+
"loss": 5.2759,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 43.74,
|
| 342 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 343 |
+
"loss": 5.3011,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 44.53,
|
| 348 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 349 |
+
"loss": 5.2758,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 45.31,
|
| 354 |
+
"learning_rate": 5.8e-05,
|
| 355 |
+
"loss": 5.2559,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 46.09,
|
| 360 |
+
"learning_rate": 5.9e-05,
|
| 361 |
+
"loss": 5.2326,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 46.87,
|
| 366 |
+
"learning_rate": 6e-05,
|
| 367 |
+
"loss": 5.1616,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 47.65,
|
| 372 |
+
"learning_rate": 6.1e-05,
|
| 373 |
+
"loss": 5.1753,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 48.43,
|
| 378 |
+
"learning_rate": 6.2e-05,
|
| 379 |
+
"loss": 5.1378,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 49.22,
|
| 384 |
+
"learning_rate": 6.3e-05,
|
| 385 |
+
"loss": 5.1,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 49.99,
|
| 390 |
+
"learning_rate": 6.400000000000001e-05,
|
| 391 |
+
"loss": 5.015,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 50.78,
|
| 396 |
+
"learning_rate": 6.500000000000001e-05,
|
| 397 |
+
"loss": 4.9758,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 51.56,
|
| 402 |
+
"learning_rate": 6.6e-05,
|
| 403 |
+
"loss": 4.8417,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 52.34,
|
| 408 |
+
"learning_rate": 6.7e-05,
|
| 409 |
+
"loss": 4.7116,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 53.12,
|
| 414 |
+
"learning_rate": 6.800000000000001e-05,
|
| 415 |
+
"loss": 4.5582,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 53.9,
|
| 420 |
+
"learning_rate": 6.9e-05,
|
| 421 |
+
"loss": 4.3437,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 54.68,
|
| 426 |
+
"learning_rate": 7e-05,
|
| 427 |
+
"loss": 4.2114,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 55.47,
|
| 432 |
+
"learning_rate": 7.1e-05,
|
| 433 |
+
"loss": 4.1021,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 56.25,
|
| 438 |
+
"learning_rate": 7.2e-05,
|
| 439 |
+
"loss": 4.0074,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 57.03,
|
| 444 |
+
"learning_rate": 7.3e-05,
|
| 445 |
+
"loss": 3.9346,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 57.81,
|
| 450 |
+
"learning_rate": 7.4e-05,
|
| 451 |
+
"loss": 3.8289,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 58.59,
|
| 456 |
+
"learning_rate": 7.500000000000001e-05,
|
| 457 |
+
"loss": 3.8105,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 59.37,
|
| 462 |
+
"learning_rate": 7.6e-05,
|
| 463 |
+
"loss": 3.755,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 60.16,
|
| 468 |
+
"learning_rate": 7.7e-05,
|
| 469 |
+
"loss": 3.7105,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 60.93,
|
| 474 |
+
"learning_rate": 7.800000000000001e-05,
|
| 475 |
+
"loss": 3.6394,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 61.71,
|
| 480 |
+
"learning_rate": 7.900000000000001e-05,
|
| 481 |
+
"loss": 3.6262,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 62.5,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 3.5924,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 63.28,
|
| 492 |
+
"learning_rate": 8.1e-05,
|
| 493 |
+
"loss": 3.558,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 64.06,
|
| 498 |
+
"learning_rate": 8.2e-05,
|
| 499 |
+
"loss": 3.5255,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 64.84,
|
| 504 |
+
"learning_rate": 8.3e-05,
|
| 505 |
+
"loss": 3.4602,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 65.62,
|
| 510 |
+
"learning_rate": 8.4e-05,
|
| 511 |
+
"loss": 3.4641,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 66.4,
|
| 516 |
+
"learning_rate": 8.5e-05,
|
| 517 |
+
"loss": 3.435,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 67.19,
|
| 522 |
+
"learning_rate": 8.6e-05,
|
| 523 |
+
"loss": 3.408,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 67.96,
|
| 528 |
+
"learning_rate": 8.7e-05,
|
| 529 |
+
"loss": 3.3594,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 68.74,
|
| 534 |
+
"learning_rate": 8.800000000000001e-05,
|
| 535 |
+
"loss": 3.3593,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 69.53,
|
| 540 |
+
"learning_rate": 8.900000000000001e-05,
|
| 541 |
+
"loss": 3.3372,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 70.31,
|
| 546 |
+
"learning_rate": 9e-05,
|
| 547 |
+
"loss": 3.3217,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 71.09,
|
| 552 |
+
"learning_rate": 9.1e-05,
|
| 553 |
+
"loss": 3.2985,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 71.87,
|
| 558 |
+
"learning_rate": 9.200000000000001e-05,
|
| 559 |
+
"loss": 3.2509,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 72.65,
|
| 564 |
+
"learning_rate": 9.300000000000001e-05,
|
| 565 |
+
"loss": 3.2584,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 73.43,
|
| 570 |
+
"learning_rate": 9.4e-05,
|
| 571 |
+
"loss": 3.2386,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 74.22,
|
| 576 |
+
"learning_rate": 9.5e-05,
|
| 577 |
+
"loss": 3.2232,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 74.99,
|
| 582 |
+
"learning_rate": 9.6e-05,
|
| 583 |
+
"loss": 3.1786,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 75.78,
|
| 588 |
+
"learning_rate": 9.7e-05,
|
| 589 |
+
"loss": 3.1855,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 76.56,
|
| 594 |
+
"learning_rate": 9.8e-05,
|
| 595 |
+
"loss": 3.1737,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 77.34,
|
| 600 |
+
"learning_rate": 9.900000000000001e-05,
|
| 601 |
+
"loss": 3.1565,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 78.12,
|
| 606 |
+
"learning_rate": 0.0001,
|
| 607 |
+
"loss": 3.1442,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 78.9,
|
| 612 |
+
"learning_rate": 9.904580152671757e-05,
|
| 613 |
+
"loss": 3.1003,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 79.68,
|
| 618 |
+
"learning_rate": 9.809160305343512e-05,
|
| 619 |
+
"loss": 3.1137,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 80.47,
|
| 624 |
+
"learning_rate": 9.713740458015268e-05,
|
| 625 |
+
"loss": 3.0958,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 81.25,
|
| 630 |
+
"learning_rate": 9.618320610687024e-05,
|
| 631 |
+
"loss": 3.0853,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 82.03,
|
| 636 |
+
"learning_rate": 9.522900763358779e-05,
|
| 637 |
+
"loss": 3.0704,
|
| 638 |
+
"step": 10500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 82.81,
|
| 642 |
+
"learning_rate": 9.427480916030534e-05,
|
| 643 |
+
"loss": 3.03,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 83.59,
|
| 648 |
+
"learning_rate": 9.33206106870229e-05,
|
| 649 |
+
"loss": 3.0428,
|
| 650 |
+
"step": 10700
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 84.37,
|
| 654 |
+
"learning_rate": 9.236641221374047e-05,
|
| 655 |
+
"loss": 3.0299,
|
| 656 |
+
"step": 10800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 85.16,
|
| 660 |
+
"learning_rate": 9.141221374045802e-05,
|
| 661 |
+
"loss": 3.0239,
|
| 662 |
+
"step": 10900
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 85.93,
|
| 666 |
+
"learning_rate": 9.045801526717558e-05,
|
| 667 |
+
"loss": 2.9818,
|
| 668 |
+
"step": 11000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 86.71,
|
| 672 |
+
"learning_rate": 8.950381679389314e-05,
|
| 673 |
+
"loss": 2.9967,
|
| 674 |
+
"step": 11100
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 87.5,
|
| 678 |
+
"learning_rate": 8.854961832061069e-05,
|
| 679 |
+
"loss": 2.9866,
|
| 680 |
+
"step": 11200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 88.28,
|
| 684 |
+
"learning_rate": 8.759541984732825e-05,
|
| 685 |
+
"loss": 2.9758,
|
| 686 |
+
"step": 11300
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 89.06,
|
| 690 |
+
"learning_rate": 8.664122137404582e-05,
|
| 691 |
+
"loss": 2.968,
|
| 692 |
+
"step": 11400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 89.84,
|
| 696 |
+
"learning_rate": 8.568702290076335e-05,
|
| 697 |
+
"loss": 2.9276,
|
| 698 |
+
"step": 11500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 90.62,
|
| 702 |
+
"learning_rate": 8.473282442748092e-05,
|
| 703 |
+
"loss": 2.9464,
|
| 704 |
+
"step": 11600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 91.4,
|
| 708 |
+
"learning_rate": 8.377862595419848e-05,
|
| 709 |
+
"loss": 2.9342,
|
| 710 |
+
"step": 11700
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 92.19,
|
| 714 |
+
"learning_rate": 8.282442748091603e-05,
|
| 715 |
+
"loss": 2.9313,
|
| 716 |
+
"step": 11800
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 92.96,
|
| 720 |
+
"learning_rate": 8.187022900763359e-05,
|
| 721 |
+
"loss": 2.8932,
|
| 722 |
+
"step": 11900
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 93.74,
|
| 726 |
+
"learning_rate": 8.091603053435115e-05,
|
| 727 |
+
"loss": 2.9087,
|
| 728 |
+
"step": 12000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 94.53,
|
| 732 |
+
"learning_rate": 7.996183206106872e-05,
|
| 733 |
+
"loss": 2.9054,
|
| 734 |
+
"step": 12100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 95.31,
|
| 738 |
+
"learning_rate": 7.900763358778626e-05,
|
| 739 |
+
"loss": 2.8916,
|
| 740 |
+
"step": 12200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 96.09,
|
| 744 |
+
"learning_rate": 7.805343511450383e-05,
|
| 745 |
+
"loss": 2.8888,
|
| 746 |
+
"step": 12300
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 96.87,
|
| 750 |
+
"learning_rate": 7.709923664122138e-05,
|
| 751 |
+
"loss": 2.8505,
|
| 752 |
+
"step": 12400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 97.65,
|
| 756 |
+
"learning_rate": 7.614503816793893e-05,
|
| 757 |
+
"loss": 2.8729,
|
| 758 |
+
"step": 12500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 98.43,
|
| 762 |
+
"learning_rate": 7.519083969465649e-05,
|
| 763 |
+
"loss": 2.8576,
|
| 764 |
+
"step": 12600
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 99.22,
|
| 768 |
+
"learning_rate": 7.423664122137405e-05,
|
| 769 |
+
"loss": 2.8569,
|
| 770 |
+
"step": 12700
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 99.99,
|
| 774 |
+
"learning_rate": 7.32824427480916e-05,
|
| 775 |
+
"loss": 2.8276,
|
| 776 |
+
"step": 12800
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 100.78,
|
| 780 |
+
"learning_rate": 7.232824427480916e-05,
|
| 781 |
+
"loss": 2.8404,
|
| 782 |
+
"step": 12900
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 101.56,
|
| 786 |
+
"learning_rate": 7.137404580152673e-05,
|
| 787 |
+
"loss": 2.8337,
|
| 788 |
+
"step": 13000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 102.34,
|
| 792 |
+
"learning_rate": 7.041984732824428e-05,
|
| 793 |
+
"loss": 2.8285,
|
| 794 |
+
"step": 13100
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 103.12,
|
| 798 |
+
"learning_rate": 6.946564885496184e-05,
|
| 799 |
+
"loss": 2.8209,
|
| 800 |
+
"step": 13200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 103.9,
|
| 804 |
+
"learning_rate": 6.851145038167939e-05,
|
| 805 |
+
"loss": 2.7906,
|
| 806 |
+
"step": 13300
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 104.68,
|
| 810 |
+
"learning_rate": 6.755725190839695e-05,
|
| 811 |
+
"loss": 2.8055,
|
| 812 |
+
"step": 13400
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 105.47,
|
| 816 |
+
"learning_rate": 6.66030534351145e-05,
|
| 817 |
+
"loss": 2.8034,
|
| 818 |
+
"step": 13500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 106.25,
|
| 822 |
+
"learning_rate": 6.564885496183206e-05,
|
| 823 |
+
"loss": 2.7957,
|
| 824 |
+
"step": 13600
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 107.03,
|
| 828 |
+
"learning_rate": 6.469465648854963e-05,
|
| 829 |
+
"loss": 2.7888,
|
| 830 |
+
"step": 13700
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 107.81,
|
| 834 |
+
"learning_rate": 6.374045801526718e-05,
|
| 835 |
+
"loss": 2.7596,
|
| 836 |
+
"step": 13800
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 108.59,
|
| 840 |
+
"learning_rate": 6.278625954198474e-05,
|
| 841 |
+
"loss": 2.7781,
|
| 842 |
+
"step": 13900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 109.37,
|
| 846 |
+
"learning_rate": 6.18320610687023e-05,
|
| 847 |
+
"loss": 2.7737,
|
| 848 |
+
"step": 14000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 110.16,
|
| 852 |
+
"learning_rate": 6.087786259541986e-05,
|
| 853 |
+
"loss": 2.7692,
|
| 854 |
+
"step": 14100
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 110.93,
|
| 858 |
+
"learning_rate": 5.992366412213741e-05,
|
| 859 |
+
"loss": 2.7379,
|
| 860 |
+
"step": 14200
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 111.71,
|
| 864 |
+
"learning_rate": 5.897900763358779e-05,
|
| 865 |
+
"loss": 2.754,
|
| 866 |
+
"step": 14300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 112.5,
|
| 870 |
+
"learning_rate": 5.802480916030535e-05,
|
| 871 |
+
"loss": 2.7503,
|
| 872 |
+
"step": 14400
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 113.28,
|
| 876 |
+
"learning_rate": 5.707061068702291e-05,
|
| 877 |
+
"loss": 2.7488,
|
| 878 |
+
"step": 14500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 114.06,
|
| 882 |
+
"learning_rate": 5.6116412213740466e-05,
|
| 883 |
+
"loss": 2.7461,
|
| 884 |
+
"step": 14600
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 114.84,
|
| 888 |
+
"learning_rate": 5.5162213740458016e-05,
|
| 889 |
+
"loss": 2.711,
|
| 890 |
+
"step": 14700
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 115.62,
|
| 894 |
+
"learning_rate": 5.420801526717557e-05,
|
| 895 |
+
"loss": 2.7314,
|
| 896 |
+
"step": 14800
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 116.4,
|
| 900 |
+
"learning_rate": 5.325381679389313e-05,
|
| 901 |
+
"loss": 2.7222,
|
| 902 |
+
"step": 14900
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 117.19,
|
| 906 |
+
"learning_rate": 5.229961832061069e-05,
|
| 907 |
+
"loss": 2.7267,
|
| 908 |
+
"step": 15000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 117.96,
|
| 912 |
+
"learning_rate": 5.134541984732825e-05,
|
| 913 |
+
"loss": 2.6951,
|
| 914 |
+
"step": 15100
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 118.74,
|
| 918 |
+
"learning_rate": 5.03912213740458e-05,
|
| 919 |
+
"loss": 2.7121,
|
| 920 |
+
"step": 15200
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 119.53,
|
| 924 |
+
"learning_rate": 4.9437022900763366e-05,
|
| 925 |
+
"loss": 2.7082,
|
| 926 |
+
"step": 15300
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 120.31,
|
| 930 |
+
"learning_rate": 4.8482824427480915e-05,
|
| 931 |
+
"loss": 2.7073,
|
| 932 |
+
"step": 15400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 121.09,
|
| 936 |
+
"learning_rate": 4.752862595419848e-05,
|
| 937 |
+
"loss": 2.7014,
|
| 938 |
+
"step": 15500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 121.87,
|
| 942 |
+
"learning_rate": 4.6574427480916034e-05,
|
| 943 |
+
"loss": 2.6753,
|
| 944 |
+
"step": 15600
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 122.65,
|
| 948 |
+
"learning_rate": 4.562022900763359e-05,
|
| 949 |
+
"loss": 2.6895,
|
| 950 |
+
"step": 15700
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 123.43,
|
| 954 |
+
"learning_rate": 4.466603053435115e-05,
|
| 955 |
+
"loss": 2.693,
|
| 956 |
+
"step": 15800
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 124.22,
|
| 960 |
+
"learning_rate": 4.37118320610687e-05,
|
| 961 |
+
"loss": 2.6863,
|
| 962 |
+
"step": 15900
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 124.99,
|
| 966 |
+
"learning_rate": 4.275763358778626e-05,
|
| 967 |
+
"loss": 2.6588,
|
| 968 |
+
"step": 16000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 125.78,
|
| 972 |
+
"learning_rate": 4.180343511450382e-05,
|
| 973 |
+
"loss": 2.6755,
|
| 974 |
+
"step": 16100
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 126.56,
|
| 978 |
+
"learning_rate": 4.084923664122138e-05,
|
| 979 |
+
"loss": 2.6724,
|
| 980 |
+
"step": 16200
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 127.34,
|
| 984 |
+
"learning_rate": 3.989503816793893e-05,
|
| 985 |
+
"loss": 2.669,
|
| 986 |
+
"step": 16300
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 128.12,
|
| 990 |
+
"learning_rate": 3.894083969465649e-05,
|
| 991 |
+
"loss": 2.6671,
|
| 992 |
+
"step": 16400
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 128.9,
|
| 996 |
+
"learning_rate": 3.7986641221374045e-05,
|
| 997 |
+
"loss": 2.6429,
|
| 998 |
+
"step": 16500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 129.68,
|
| 1002 |
+
"learning_rate": 3.703244274809161e-05,
|
| 1003 |
+
"loss": 2.6615,
|
| 1004 |
+
"step": 16600
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 130.47,
|
| 1008 |
+
"learning_rate": 3.6078244274809164e-05,
|
| 1009 |
+
"loss": 2.6553,
|
| 1010 |
+
"step": 16700
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 131.25,
|
| 1014 |
+
"learning_rate": 3.512404580152672e-05,
|
| 1015 |
+
"loss": 2.6541,
|
| 1016 |
+
"step": 16800
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 132.03,
|
| 1020 |
+
"learning_rate": 3.4169847328244276e-05,
|
| 1021 |
+
"loss": 2.6543,
|
| 1022 |
+
"step": 16900
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 132.81,
|
| 1026 |
+
"learning_rate": 3.321564885496183e-05,
|
| 1027 |
+
"loss": 2.6242,
|
| 1028 |
+
"step": 17000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 133.59,
|
| 1032 |
+
"learning_rate": 3.2261450381679395e-05,
|
| 1033 |
+
"loss": 2.6412,
|
| 1034 |
+
"step": 17100
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 134.37,
|
| 1038 |
+
"learning_rate": 3.130725190839695e-05,
|
| 1039 |
+
"loss": 2.6423,
|
| 1040 |
+
"step": 17200
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 135.16,
|
| 1044 |
+
"learning_rate": 3.0353053435114503e-05,
|
| 1045 |
+
"loss": 2.6372,
|
| 1046 |
+
"step": 17300
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 135.93,
|
| 1050 |
+
"learning_rate": 2.9398854961832063e-05,
|
| 1051 |
+
"loss": 2.6107,
|
| 1052 |
+
"step": 17400
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 136.71,
|
| 1056 |
+
"learning_rate": 2.844465648854962e-05,
|
| 1057 |
+
"loss": 2.6298,
|
| 1058 |
+
"step": 17500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 137.5,
|
| 1062 |
+
"learning_rate": 2.7490458015267178e-05,
|
| 1063 |
+
"loss": 2.6265,
|
| 1064 |
+
"step": 17600
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 138.28,
|
| 1068 |
+
"learning_rate": 2.6545801526717557e-05,
|
| 1069 |
+
"loss": 2.6274,
|
| 1070 |
+
"step": 17700
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 139.06,
|
| 1074 |
+
"learning_rate": 2.5591603053435116e-05,
|
| 1075 |
+
"loss": 2.626,
|
| 1076 |
+
"step": 17800
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 139.84,
|
| 1080 |
+
"learning_rate": 2.4637404580152672e-05,
|
| 1081 |
+
"loss": 2.6,
|
| 1082 |
+
"step": 17900
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 140.62,
|
| 1086 |
+
"learning_rate": 2.368320610687023e-05,
|
| 1087 |
+
"loss": 2.6194,
|
| 1088 |
+
"step": 18000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 141.4,
|
| 1092 |
+
"learning_rate": 2.2729007633587788e-05,
|
| 1093 |
+
"loss": 2.6174,
|
| 1094 |
+
"step": 18100
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 142.19,
|
| 1098 |
+
"learning_rate": 2.1774809160305344e-05,
|
| 1099 |
+
"loss": 2.6168,
|
| 1100 |
+
"step": 18200
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 142.96,
|
| 1104 |
+
"learning_rate": 2.08206106870229e-05,
|
| 1105 |
+
"loss": 2.5903,
|
| 1106 |
+
"step": 18300
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 143.74,
|
| 1110 |
+
"learning_rate": 1.986641221374046e-05,
|
| 1111 |
+
"loss": 2.6097,
|
| 1112 |
+
"step": 18400
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 144.53,
|
| 1116 |
+
"learning_rate": 1.8912213740458016e-05,
|
| 1117 |
+
"loss": 2.6107,
|
| 1118 |
+
"step": 18500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 145.31,
|
| 1122 |
+
"learning_rate": 1.7958015267175575e-05,
|
| 1123 |
+
"loss": 2.6078,
|
| 1124 |
+
"step": 18600
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 146.09,
|
| 1128 |
+
"learning_rate": 1.7003816793893128e-05,
|
| 1129 |
+
"loss": 2.6033,
|
| 1130 |
+
"step": 18700
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 146.87,
|
| 1134 |
+
"learning_rate": 1.6049618320610687e-05,
|
| 1135 |
+
"loss": 2.5791,
|
| 1136 |
+
"step": 18800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 147.65,
|
| 1140 |
+
"learning_rate": 1.5095419847328246e-05,
|
| 1141 |
+
"loss": 2.6027,
|
| 1142 |
+
"step": 18900
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 148.43,
|
| 1146 |
+
"learning_rate": 1.41412213740458e-05,
|
| 1147 |
+
"loss": 2.597,
|
| 1148 |
+
"step": 19000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 149.22,
|
| 1152 |
+
"learning_rate": 1.318702290076336e-05,
|
| 1153 |
+
"loss": 2.5998,
|
| 1154 |
+
"step": 19100
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 149.99,
|
| 1158 |
+
"learning_rate": 1.2232824427480916e-05,
|
| 1159 |
+
"loss": 2.5747,
|
| 1160 |
+
"step": 19200
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 150.78,
|
| 1164 |
+
"learning_rate": 1.1278625954198474e-05,
|
| 1165 |
+
"loss": 2.5939,
|
| 1166 |
+
"step": 19300
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 151.56,
|
| 1170 |
+
"learning_rate": 1.0324427480916032e-05,
|
| 1171 |
+
"loss": 2.5941,
|
| 1172 |
+
"step": 19400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 152.34,
|
| 1176 |
+
"learning_rate": 9.370229007633588e-06,
|
| 1177 |
+
"loss": 2.5906,
|
| 1178 |
+
"step": 19500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 153.12,
|
| 1182 |
+
"learning_rate": 8.416030534351146e-06,
|
| 1183 |
+
"loss": 2.593,
|
| 1184 |
+
"step": 19600
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 153.9,
|
| 1188 |
+
"learning_rate": 7.4618320610687024e-06,
|
| 1189 |
+
"loss": 2.5669,
|
| 1190 |
+
"step": 19700
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 154.68,
|
| 1194 |
+
"learning_rate": 6.50763358778626e-06,
|
| 1195 |
+
"loss": 2.5849,
|
| 1196 |
+
"step": 19800
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 155.47,
|
| 1200 |
+
"learning_rate": 5.553435114503817e-06,
|
| 1201 |
+
"loss": 2.584,
|
| 1202 |
+
"step": 19900
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 156.25,
|
| 1206 |
+
"learning_rate": 4.599236641221375e-06,
|
| 1207 |
+
"loss": 2.5826,
|
| 1208 |
+
"step": 20000
|
| 1209 |
+
}
|
| 1210 |
+
],
|
| 1211 |
+
"max_steps": 20480,
|
| 1212 |
+
"num_train_epochs": 160,
|
| 1213 |
+
"total_flos": 1.3558440202862592e+18,
|
| 1214 |
+
"trial_name": null,
|
| 1215 |
+
"trial_params": null
|
| 1216 |
+
}
|
checkpoint-20000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6de0a5888f9534c475c0b5c0d6fa6d4920e1a50a89517b67872f8b6ca51ef166
|
| 3 |
+
size 3579
|
checkpoint-3100/config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 260,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 3,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"sep_token_id": 2,
|
| 23 |
+
"torch_dtype": "float32",
|
| 24 |
+
"transformers_version": "4.27.0.dev0",
|
| 25 |
+
"type_vocab_size": 1,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
checkpoint-3100/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad434181a50ada83a588c3c46b63f3dfca29a5b657d01b7a28d713fc6433602c
|
| 3 |
+
size 883771077
|