Commit ·
d388fac
1
Parent(s): d107c0f
Upload . with huggingface_hub
Browse files- config.json +40 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +616 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/bert_uncased_L-4_H-512_A-8",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 512,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "negative",
|
| 13 |
+
"1": "somewhat negative",
|
| 14 |
+
"2": "neutral",
|
| 15 |
+
"3": "somewhat positive",
|
| 16 |
+
"4": "positive"
|
| 17 |
+
},
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 2048,
|
| 20 |
+
"label2id": {
|
| 21 |
+
"negative": 0,
|
| 22 |
+
"neutral": 2,
|
| 23 |
+
"positive": 4,
|
| 24 |
+
"somewhat negative": 1,
|
| 25 |
+
"somewhat positive": 3
|
| 26 |
+
},
|
| 27 |
+
"layer_norm_eps": 1e-12,
|
| 28 |
+
"max_position_embeddings": 512,
|
| 29 |
+
"model_type": "bert",
|
| 30 |
+
"num_attention_heads": 8,
|
| 31 |
+
"num_hidden_layers": 4,
|
| 32 |
+
"pad_token_id": 0,
|
| 33 |
+
"position_embedding_type": "absolute",
|
| 34 |
+
"problem_type": "single_label_classification",
|
| 35 |
+
"torch_dtype": "float32",
|
| 36 |
+
"transformers_version": "4.20.1",
|
| 37 |
+
"type_vocab_size": 2,
|
| 38 |
+
"use_cache": true,
|
| 39 |
+
"vocab_size": 30522
|
| 40 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3461e5fd9b147b7e41fb708cf4a5215ea745b4b5a2911d849c950d388416062
|
| 3 |
+
size 230171549
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:235e025b68b2e2dfe0bcf29d7f0e124d17d3305f21106a0c3ba9403a9b8cdc8e
|
| 3 |
+
size 115092807
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e5b50f01f500865f544669c4bcaee728ed289916f9e66b97cbcc129ccc6cffbb
|
| 3 |
+
size 14503
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4fa4c7be44c959599b8b43bb9bc3371e9e4e5bbc5758b3ab5afcccfda3e72e67
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dbf813382ccdaa822c4dea5e4d8e232bb9a90b87b1dfdb5cb285670917cb5faf
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,616 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5467842252751008,
|
| 5 |
+
"global_step": 500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 1.989059080962801e-05,
|
| 13 |
+
"loss": 1.8081,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 1.978118161925602e-05,
|
| 19 |
+
"loss": 1.5823,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.02,
|
| 24 |
+
"learning_rate": 1.9671772428884027e-05,
|
| 25 |
+
"loss": 1.4255,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 1.9562363238512036e-05,
|
| 31 |
+
"loss": 1.3195,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.03,
|
| 36 |
+
"learning_rate": 1.9452954048140046e-05,
|
| 37 |
+
"loss": 1.3249,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03,
|
| 42 |
+
"learning_rate": 1.9343544857768056e-05,
|
| 43 |
+
"loss": 1.2716,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04,
|
| 48 |
+
"learning_rate": 1.9234135667396062e-05,
|
| 49 |
+
"loss": 1.1967,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.04,
|
| 54 |
+
"learning_rate": 1.912472647702407e-05,
|
| 55 |
+
"loss": 1.2207,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.05,
|
| 60 |
+
"learning_rate": 1.901531728665208e-05,
|
| 61 |
+
"loss": 1.1743,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.05,
|
| 66 |
+
"learning_rate": 1.890590809628009e-05,
|
| 67 |
+
"loss": 1.1534,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.06,
|
| 72 |
+
"learning_rate": 1.87964989059081e-05,
|
| 73 |
+
"loss": 1.1279,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.07,
|
| 78 |
+
"learning_rate": 1.8687089715536106e-05,
|
| 79 |
+
"loss": 1.1936,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.07,
|
| 84 |
+
"learning_rate": 1.8577680525164116e-05,
|
| 85 |
+
"loss": 1.1124,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08,
|
| 90 |
+
"learning_rate": 1.8468271334792125e-05,
|
| 91 |
+
"loss": 1.1489,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.08,
|
| 96 |
+
"learning_rate": 1.8358862144420135e-05,
|
| 97 |
+
"loss": 1.1339,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.09,
|
| 102 |
+
"learning_rate": 1.824945295404814e-05,
|
| 103 |
+
"loss": 1.0732,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.09,
|
| 108 |
+
"learning_rate": 1.814004376367615e-05,
|
| 109 |
+
"loss": 1.0833,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.1,
|
| 114 |
+
"learning_rate": 1.803063457330416e-05,
|
| 115 |
+
"loss": 1.1498,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.1,
|
| 120 |
+
"learning_rate": 1.792122538293217e-05,
|
| 121 |
+
"loss": 1.0865,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.11,
|
| 126 |
+
"learning_rate": 1.7811816192560176e-05,
|
| 127 |
+
"loss": 1.0837,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.11,
|
| 132 |
+
"learning_rate": 1.7702407002188185e-05,
|
| 133 |
+
"loss": 1.0382,
|
| 134 |
+
"step": 105
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.12,
|
| 138 |
+
"learning_rate": 1.7592997811816195e-05,
|
| 139 |
+
"loss": 1.038,
|
| 140 |
+
"step": 110
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.13,
|
| 144 |
+
"learning_rate": 1.7483588621444204e-05,
|
| 145 |
+
"loss": 1.0549,
|
| 146 |
+
"step": 115
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.13,
|
| 150 |
+
"learning_rate": 1.737417943107221e-05,
|
| 151 |
+
"loss": 1.1154,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.14,
|
| 156 |
+
"learning_rate": 1.726477024070022e-05,
|
| 157 |
+
"loss": 1.0393,
|
| 158 |
+
"step": 125
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.14,
|
| 162 |
+
"learning_rate": 1.715536105032823e-05,
|
| 163 |
+
"loss": 1.0386,
|
| 164 |
+
"step": 130
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.15,
|
| 168 |
+
"learning_rate": 1.704595185995624e-05,
|
| 169 |
+
"loss": 1.0751,
|
| 170 |
+
"step": 135
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.15,
|
| 174 |
+
"learning_rate": 1.6936542669584245e-05,
|
| 175 |
+
"loss": 1.0139,
|
| 176 |
+
"step": 140
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.16,
|
| 180 |
+
"learning_rate": 1.6827133479212255e-05,
|
| 181 |
+
"loss": 1.0316,
|
| 182 |
+
"step": 145
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.16,
|
| 186 |
+
"learning_rate": 1.6717724288840264e-05,
|
| 187 |
+
"loss": 1.063,
|
| 188 |
+
"step": 150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.17,
|
| 192 |
+
"learning_rate": 1.6608315098468274e-05,
|
| 193 |
+
"loss": 1.0042,
|
| 194 |
+
"step": 155
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.17,
|
| 198 |
+
"learning_rate": 1.649890590809628e-05,
|
| 199 |
+
"loss": 1.0592,
|
| 200 |
+
"step": 160
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.18,
|
| 204 |
+
"learning_rate": 1.638949671772429e-05,
|
| 205 |
+
"loss": 1.0097,
|
| 206 |
+
"step": 165
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.19,
|
| 210 |
+
"learning_rate": 1.62800875273523e-05,
|
| 211 |
+
"loss": 1.0066,
|
| 212 |
+
"step": 170
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.19,
|
| 216 |
+
"learning_rate": 1.617067833698031e-05,
|
| 217 |
+
"loss": 1.0446,
|
| 218 |
+
"step": 175
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.2,
|
| 222 |
+
"learning_rate": 1.6061269146608315e-05,
|
| 223 |
+
"loss": 1.0535,
|
| 224 |
+
"step": 180
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.2,
|
| 228 |
+
"learning_rate": 1.5951859956236324e-05,
|
| 229 |
+
"loss": 1.0232,
|
| 230 |
+
"step": 185
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.21,
|
| 234 |
+
"learning_rate": 1.5842450765864334e-05,
|
| 235 |
+
"loss": 0.9909,
|
| 236 |
+
"step": 190
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.21,
|
| 240 |
+
"learning_rate": 1.5733041575492344e-05,
|
| 241 |
+
"loss": 0.9984,
|
| 242 |
+
"step": 195
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.22,
|
| 246 |
+
"learning_rate": 1.562363238512035e-05,
|
| 247 |
+
"loss": 1.016,
|
| 248 |
+
"step": 200
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.22,
|
| 252 |
+
"learning_rate": 1.551422319474836e-05,
|
| 253 |
+
"loss": 1.0308,
|
| 254 |
+
"step": 205
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.23,
|
| 258 |
+
"learning_rate": 1.540481400437637e-05,
|
| 259 |
+
"loss": 1.0138,
|
| 260 |
+
"step": 210
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.24,
|
| 264 |
+
"learning_rate": 1.529540481400438e-05,
|
| 265 |
+
"loss": 0.9115,
|
| 266 |
+
"step": 215
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.24,
|
| 270 |
+
"learning_rate": 1.5185995623632386e-05,
|
| 271 |
+
"loss": 0.9768,
|
| 272 |
+
"step": 220
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.25,
|
| 276 |
+
"learning_rate": 1.5076586433260396e-05,
|
| 277 |
+
"loss": 0.9658,
|
| 278 |
+
"step": 225
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.25,
|
| 282 |
+
"learning_rate": 1.4967177242888404e-05,
|
| 283 |
+
"loss": 1.0273,
|
| 284 |
+
"step": 230
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.26,
|
| 288 |
+
"learning_rate": 1.4857768052516413e-05,
|
| 289 |
+
"loss": 1.0152,
|
| 290 |
+
"step": 235
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.26,
|
| 294 |
+
"learning_rate": 1.4748358862144421e-05,
|
| 295 |
+
"loss": 0.9777,
|
| 296 |
+
"step": 240
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.27,
|
| 300 |
+
"learning_rate": 1.463894967177243e-05,
|
| 301 |
+
"loss": 0.9664,
|
| 302 |
+
"step": 245
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.27,
|
| 306 |
+
"learning_rate": 1.452954048140044e-05,
|
| 307 |
+
"loss": 1.0219,
|
| 308 |
+
"step": 250
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.28,
|
| 312 |
+
"learning_rate": 1.4420131291028448e-05,
|
| 313 |
+
"loss": 0.9662,
|
| 314 |
+
"step": 255
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.28,
|
| 318 |
+
"learning_rate": 1.4310722100656456e-05,
|
| 319 |
+
"loss": 0.9881,
|
| 320 |
+
"step": 260
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.29,
|
| 324 |
+
"learning_rate": 1.4201312910284465e-05,
|
| 325 |
+
"loss": 0.9458,
|
| 326 |
+
"step": 265
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.3,
|
| 330 |
+
"learning_rate": 1.4091903719912475e-05,
|
| 331 |
+
"loss": 0.9368,
|
| 332 |
+
"step": 270
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3,
|
| 336 |
+
"learning_rate": 1.3982494529540483e-05,
|
| 337 |
+
"loss": 0.983,
|
| 338 |
+
"step": 275
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.31,
|
| 342 |
+
"learning_rate": 1.387308533916849e-05,
|
| 343 |
+
"loss": 0.9804,
|
| 344 |
+
"step": 280
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.31,
|
| 348 |
+
"learning_rate": 1.37636761487965e-05,
|
| 349 |
+
"loss": 0.9812,
|
| 350 |
+
"step": 285
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.32,
|
| 354 |
+
"learning_rate": 1.365426695842451e-05,
|
| 355 |
+
"loss": 0.937,
|
| 356 |
+
"step": 290
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.32,
|
| 360 |
+
"learning_rate": 1.3544857768052518e-05,
|
| 361 |
+
"loss": 0.8969,
|
| 362 |
+
"step": 295
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.33,
|
| 366 |
+
"learning_rate": 1.3435448577680525e-05,
|
| 367 |
+
"loss": 0.9726,
|
| 368 |
+
"step": 300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.33,
|
| 372 |
+
"learning_rate": 1.3326039387308537e-05,
|
| 373 |
+
"loss": 0.992,
|
| 374 |
+
"step": 305
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.34,
|
| 378 |
+
"learning_rate": 1.3216630196936544e-05,
|
| 379 |
+
"loss": 0.9751,
|
| 380 |
+
"step": 310
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.34,
|
| 384 |
+
"learning_rate": 1.3107221006564552e-05,
|
| 385 |
+
"loss": 0.9933,
|
| 386 |
+
"step": 315
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.35,
|
| 390 |
+
"learning_rate": 1.299781181619256e-05,
|
| 391 |
+
"loss": 0.9336,
|
| 392 |
+
"step": 320
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.36,
|
| 396 |
+
"learning_rate": 1.2888402625820571e-05,
|
| 397 |
+
"loss": 0.9506,
|
| 398 |
+
"step": 325
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.36,
|
| 402 |
+
"learning_rate": 1.277899343544858e-05,
|
| 403 |
+
"loss": 0.9428,
|
| 404 |
+
"step": 330
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.37,
|
| 408 |
+
"learning_rate": 1.2669584245076587e-05,
|
| 409 |
+
"loss": 0.9445,
|
| 410 |
+
"step": 335
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.37,
|
| 414 |
+
"learning_rate": 1.2560175054704595e-05,
|
| 415 |
+
"loss": 0.9749,
|
| 416 |
+
"step": 340
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.38,
|
| 420 |
+
"learning_rate": 1.2450765864332606e-05,
|
| 421 |
+
"loss": 0.9063,
|
| 422 |
+
"step": 345
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.38,
|
| 426 |
+
"learning_rate": 1.2341356673960614e-05,
|
| 427 |
+
"loss": 0.9826,
|
| 428 |
+
"step": 350
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.39,
|
| 432 |
+
"learning_rate": 1.2231947483588622e-05,
|
| 433 |
+
"loss": 1.0168,
|
| 434 |
+
"step": 355
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.39,
|
| 438 |
+
"learning_rate": 1.212253829321663e-05,
|
| 439 |
+
"loss": 0.9563,
|
| 440 |
+
"step": 360
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.4,
|
| 444 |
+
"learning_rate": 1.2013129102844641e-05,
|
| 445 |
+
"loss": 0.9916,
|
| 446 |
+
"step": 365
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.4,
|
| 450 |
+
"learning_rate": 1.1903719912472649e-05,
|
| 451 |
+
"loss": 0.9757,
|
| 452 |
+
"step": 370
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.41,
|
| 456 |
+
"learning_rate": 1.1794310722100657e-05,
|
| 457 |
+
"loss": 0.9735,
|
| 458 |
+
"step": 375
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.42,
|
| 462 |
+
"learning_rate": 1.1684901531728665e-05,
|
| 463 |
+
"loss": 0.9303,
|
| 464 |
+
"step": 380
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.42,
|
| 468 |
+
"learning_rate": 1.1575492341356676e-05,
|
| 469 |
+
"loss": 0.9511,
|
| 470 |
+
"step": 385
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.43,
|
| 474 |
+
"learning_rate": 1.1466083150984684e-05,
|
| 475 |
+
"loss": 0.9352,
|
| 476 |
+
"step": 390
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.43,
|
| 480 |
+
"learning_rate": 1.1356673960612692e-05,
|
| 481 |
+
"loss": 0.9395,
|
| 482 |
+
"step": 395
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.44,
|
| 486 |
+
"learning_rate": 1.12472647702407e-05,
|
| 487 |
+
"loss": 0.9469,
|
| 488 |
+
"step": 400
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.44,
|
| 492 |
+
"learning_rate": 1.113785557986871e-05,
|
| 493 |
+
"loss": 0.9058,
|
| 494 |
+
"step": 405
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.45,
|
| 498 |
+
"learning_rate": 1.1028446389496719e-05,
|
| 499 |
+
"loss": 0.9149,
|
| 500 |
+
"step": 410
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.45,
|
| 504 |
+
"learning_rate": 1.0919037199124726e-05,
|
| 505 |
+
"loss": 0.9393,
|
| 506 |
+
"step": 415
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.46,
|
| 510 |
+
"learning_rate": 1.0809628008752738e-05,
|
| 511 |
+
"loss": 0.9517,
|
| 512 |
+
"step": 420
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.46,
|
| 516 |
+
"learning_rate": 1.0700218818380745e-05,
|
| 517 |
+
"loss": 0.9113,
|
| 518 |
+
"step": 425
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.47,
|
| 522 |
+
"learning_rate": 1.0590809628008753e-05,
|
| 523 |
+
"loss": 0.8842,
|
| 524 |
+
"step": 430
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.48,
|
| 528 |
+
"learning_rate": 1.0481400437636761e-05,
|
| 529 |
+
"loss": 0.953,
|
| 530 |
+
"step": 435
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.48,
|
| 534 |
+
"learning_rate": 1.0371991247264772e-05,
|
| 535 |
+
"loss": 1.0312,
|
| 536 |
+
"step": 440
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.49,
|
| 540 |
+
"learning_rate": 1.026258205689278e-05,
|
| 541 |
+
"loss": 0.9126,
|
| 542 |
+
"step": 445
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.49,
|
| 546 |
+
"learning_rate": 1.0153172866520788e-05,
|
| 547 |
+
"loss": 0.886,
|
| 548 |
+
"step": 450
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5,
|
| 552 |
+
"learning_rate": 1.0043763676148796e-05,
|
| 553 |
+
"loss": 0.9477,
|
| 554 |
+
"step": 455
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5,
|
| 558 |
+
"learning_rate": 9.934354485776806e-06,
|
| 559 |
+
"loss": 0.9368,
|
| 560 |
+
"step": 460
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.51,
|
| 564 |
+
"learning_rate": 9.824945295404815e-06,
|
| 565 |
+
"loss": 0.9132,
|
| 566 |
+
"step": 465
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.51,
|
| 570 |
+
"learning_rate": 9.715536105032823e-06,
|
| 571 |
+
"loss": 0.9229,
|
| 572 |
+
"step": 470
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.52,
|
| 576 |
+
"learning_rate": 9.606126914660832e-06,
|
| 577 |
+
"loss": 0.9743,
|
| 578 |
+
"step": 475
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.52,
|
| 582 |
+
"learning_rate": 9.49671772428884e-06,
|
| 583 |
+
"loss": 0.9515,
|
| 584 |
+
"step": 480
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.53,
|
| 588 |
+
"learning_rate": 9.38730853391685e-06,
|
| 589 |
+
"loss": 0.9278,
|
| 590 |
+
"step": 485
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.54,
|
| 594 |
+
"learning_rate": 9.277899343544858e-06,
|
| 595 |
+
"loss": 0.9402,
|
| 596 |
+
"step": 490
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.54,
|
| 600 |
+
"learning_rate": 9.168490153172867e-06,
|
| 601 |
+
"loss": 0.9358,
|
| 602 |
+
"step": 495
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.55,
|
| 606 |
+
"learning_rate": 9.059080962800875e-06,
|
| 607 |
+
"loss": 0.9695,
|
| 608 |
+
"step": 500
|
| 609 |
+
}
|
| 610 |
+
],
|
| 611 |
+
"max_steps": 914,
|
| 612 |
+
"num_train_epochs": 1,
|
| 613 |
+
"total_flos": 2531481550848000.0,
|
| 614 |
+
"trial_name": null,
|
| 615 |
+
"trial_params": null
|
| 616 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94557e383bfadf91f87decc055477ab5e2ad0487c3b578e5283617b494933aa6
|
| 3 |
+
size 3247
|