Upload folder using huggingface_hub
Browse files- README.md +36 -0
- checkpoint-9360/config.json +55 -0
- checkpoint-9360/model.safetensors +3 -0
- checkpoint-9360/optimizer.pt +3 -0
- checkpoint-9360/rng_state.pth +3 -0
- checkpoint-9360/scheduler.pt +3 -0
- checkpoint-9360/trainer_state.json +2714 -0
- checkpoint-9360/training_args.bin +3 -0
- config.json +55 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- runs/Nov11_10-29-57_r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz/events.out.tfevents.1731321001.r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz.146.0 +2 -2
- runs/Nov11_10-29-57_r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz/events.out.tfevents.1731964149.r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz.146.1 +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +57 -0
- training_args.bin +3 -0
- training_params.json +30 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- text-classification
|
| 6 |
+
base_model: allenai/longformer-base-4096
|
| 7 |
+
widget:
|
| 8 |
+
- text: "I love AutoTrain"
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# Model Trained Using AutoTrain
|
| 12 |
+
|
| 13 |
+
- Problem type: Text Classification
|
| 14 |
+
|
| 15 |
+
## Validation Metrics
|
| 16 |
+
loss: 1.0986850261688232
|
| 17 |
+
|
| 18 |
+
f1_macro: 0.16666666666666666
|
| 19 |
+
|
| 20 |
+
f1_micro: 0.3333333333333333
|
| 21 |
+
|
| 22 |
+
f1_weighted: 0.16666666666666666
|
| 23 |
+
|
| 24 |
+
precision_macro: 0.1111111111111111
|
| 25 |
+
|
| 26 |
+
precision_micro: 0.3333333333333333
|
| 27 |
+
|
| 28 |
+
precision_weighted: 0.1111111111111111
|
| 29 |
+
|
| 30 |
+
recall_macro: 0.3333333333333333
|
| 31 |
+
|
| 32 |
+
recall_micro: 0.3333333333333333
|
| 33 |
+
|
| 34 |
+
recall_weighted: 0.3333333333333333
|
| 35 |
+
|
| 36 |
+
accuracy: 0.3333333333333333
|
checkpoint-9360/config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "allenai/longformer-base-4096",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"LongformerForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_mode": "longformer",
|
| 8 |
+
"attention_probs_dropout_prob": 0.1,
|
| 9 |
+
"attention_window": [
|
| 10 |
+
512,
|
| 11 |
+
512,
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512
|
| 22 |
+
],
|
| 23 |
+
"bos_token_id": 0,
|
| 24 |
+
"eos_token_id": 2,
|
| 25 |
+
"gradient_checkpointing": false,
|
| 26 |
+
"hidden_act": "gelu",
|
| 27 |
+
"hidden_dropout_prob": 0.1,
|
| 28 |
+
"hidden_size": 768,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "negative",
|
| 31 |
+
"1": "neutral",
|
| 32 |
+
"2": "positive"
|
| 33 |
+
},
|
| 34 |
+
"ignore_attention_mask": false,
|
| 35 |
+
"initializer_range": 0.02,
|
| 36 |
+
"intermediate_size": 3072,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"negative": 0,
|
| 39 |
+
"neutral": 1,
|
| 40 |
+
"positive": 2
|
| 41 |
+
},
|
| 42 |
+
"layer_norm_eps": 1e-05,
|
| 43 |
+
"max_position_embeddings": 4098,
|
| 44 |
+
"model_type": "longformer",
|
| 45 |
+
"num_attention_heads": 12,
|
| 46 |
+
"num_hidden_layers": 12,
|
| 47 |
+
"onnx_export": false,
|
| 48 |
+
"pad_token_id": 1,
|
| 49 |
+
"problem_type": "single_label_classification",
|
| 50 |
+
"sep_token_id": 2,
|
| 51 |
+
"torch_dtype": "float32",
|
| 52 |
+
"transformers_version": "4.46.2",
|
| 53 |
+
"type_vocab_size": 1,
|
| 54 |
+
"vocab_size": 50265
|
| 55 |
+
}
|
checkpoint-9360/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc18e806f34592b583e45db50ba2aa5982e9c1b7f599857ceae4def227334747
|
| 3 |
+
size 594681260
|
checkpoint-9360/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6028b1c652f6ed6b7a981140ac078674ad0a4a4d6f29a02f9f7bede464d1b4d
|
| 3 |
+
size 1189514810
|
checkpoint-9360/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9831b595be9fae244eb0e6b20f6405cc54c15e1119cbf0fefa64a3261ef4b20
|
| 3 |
+
size 13990
|
checkpoint-9360/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:54981e7e10dec68b1e0f291f9711f08e8a83ecff3509a3a468ab073c031b292f
|
| 3 |
+
size 1064
|
checkpoint-9360/trainer_state.json
ADDED
|
@@ -0,0 +1,2714 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.0986850261688232,
|
| 3 |
+
"best_model_checkpoint": "longfomer5balanced/checkpoint-9360",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 9360,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008012820512820512,
|
| 13 |
+
"grad_norm": 9.334623336791992,
|
| 14 |
+
"learning_rate": 1.3354700854700855e-06,
|
| 15 |
+
"loss": 1.119,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016025641025641024,
|
| 20 |
+
"grad_norm": 24.905296325683594,
|
| 21 |
+
"learning_rate": 2.670940170940171e-06,
|
| 22 |
+
"loss": 1.1108,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02403846153846154,
|
| 27 |
+
"grad_norm": 11.87872314453125,
|
| 28 |
+
"learning_rate": 4.006410256410257e-06,
|
| 29 |
+
"loss": 1.0617,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03205128205128205,
|
| 34 |
+
"grad_norm": 9.319697380065918,
|
| 35 |
+
"learning_rate": 5.341880341880342e-06,
|
| 36 |
+
"loss": 1.1114,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04006410256410257,
|
| 41 |
+
"grad_norm": 22.3123836517334,
|
| 42 |
+
"learning_rate": 6.677350427350427e-06,
|
| 43 |
+
"loss": 1.1116,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04807692307692308,
|
| 48 |
+
"grad_norm": 8.417306900024414,
|
| 49 |
+
"learning_rate": 8.012820512820515e-06,
|
| 50 |
+
"loss": 1.1279,
|
| 51 |
+
"step": 150
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05608974358974359,
|
| 55 |
+
"grad_norm": 8.819703102111816,
|
| 56 |
+
"learning_rate": 9.348290598290598e-06,
|
| 57 |
+
"loss": 1.092,
|
| 58 |
+
"step": 175
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0641025641025641,
|
| 62 |
+
"grad_norm": 8.240078926086426,
|
| 63 |
+
"learning_rate": 1.0683760683760684e-05,
|
| 64 |
+
"loss": 1.0826,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.07211538461538461,
|
| 69 |
+
"grad_norm": 4.286315441131592,
|
| 70 |
+
"learning_rate": 1.2019230769230771e-05,
|
| 71 |
+
"loss": 1.1173,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08012820512820513,
|
| 76 |
+
"grad_norm": 16.222150802612305,
|
| 77 |
+
"learning_rate": 1.3354700854700855e-05,
|
| 78 |
+
"loss": 1.1639,
|
| 79 |
+
"step": 250
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08814102564102565,
|
| 83 |
+
"grad_norm": 10.093961715698242,
|
| 84 |
+
"learning_rate": 1.4690170940170942e-05,
|
| 85 |
+
"loss": 1.1115,
|
| 86 |
+
"step": 275
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.09615384615384616,
|
| 90 |
+
"grad_norm": 8.11520767211914,
|
| 91 |
+
"learning_rate": 1.602564102564103e-05,
|
| 92 |
+
"loss": 1.0883,
|
| 93 |
+
"step": 300
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.10416666666666667,
|
| 97 |
+
"grad_norm": 9.468626976013184,
|
| 98 |
+
"learning_rate": 1.736111111111111e-05,
|
| 99 |
+
"loss": 1.133,
|
| 100 |
+
"step": 325
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.11217948717948718,
|
| 104 |
+
"grad_norm": 10.271284103393555,
|
| 105 |
+
"learning_rate": 1.8696581196581197e-05,
|
| 106 |
+
"loss": 1.1065,
|
| 107 |
+
"step": 350
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1201923076923077,
|
| 111 |
+
"grad_norm": 32.357994079589844,
|
| 112 |
+
"learning_rate": 2.0032051282051285e-05,
|
| 113 |
+
"loss": 1.1291,
|
| 114 |
+
"step": 375
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1282051282051282,
|
| 118 |
+
"grad_norm": 19.496496200561523,
|
| 119 |
+
"learning_rate": 2.1367521367521368e-05,
|
| 120 |
+
"loss": 1.1769,
|
| 121 |
+
"step": 400
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1362179487179487,
|
| 125 |
+
"grad_norm": 20.75379753112793,
|
| 126 |
+
"learning_rate": 2.2702991452991453e-05,
|
| 127 |
+
"loss": 1.0681,
|
| 128 |
+
"step": 425
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.14423076923076922,
|
| 132 |
+
"grad_norm": 1.7677823305130005,
|
| 133 |
+
"learning_rate": 2.4038461538461542e-05,
|
| 134 |
+
"loss": 0.9769,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.15224358974358973,
|
| 139 |
+
"grad_norm": 0.2692599594593048,
|
| 140 |
+
"learning_rate": 2.5373931623931624e-05,
|
| 141 |
+
"loss": 1.0149,
|
| 142 |
+
"step": 475
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.16025641025641027,
|
| 146 |
+
"grad_norm": 0.07610945403575897,
|
| 147 |
+
"learning_rate": 2.670940170940171e-05,
|
| 148 |
+
"loss": 0.9201,
|
| 149 |
+
"step": 500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.16826923076923078,
|
| 153 |
+
"grad_norm": 7.888437747955322,
|
| 154 |
+
"learning_rate": 2.8044871794871795e-05,
|
| 155 |
+
"loss": 2.0309,
|
| 156 |
+
"step": 525
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1762820512820513,
|
| 160 |
+
"grad_norm": 4.857154369354248,
|
| 161 |
+
"learning_rate": 2.9380341880341884e-05,
|
| 162 |
+
"loss": 1.2041,
|
| 163 |
+
"step": 550
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.1842948717948718,
|
| 167 |
+
"grad_norm": 3.4702775478363037,
|
| 168 |
+
"learning_rate": 3.071581196581197e-05,
|
| 169 |
+
"loss": 1.2659,
|
| 170 |
+
"step": 575
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.19230769230769232,
|
| 174 |
+
"grad_norm": 10.269780158996582,
|
| 175 |
+
"learning_rate": 3.205128205128206e-05,
|
| 176 |
+
"loss": 1.37,
|
| 177 |
+
"step": 600
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.20032051282051283,
|
| 181 |
+
"grad_norm": 14.360907554626465,
|
| 182 |
+
"learning_rate": 3.338675213675213e-05,
|
| 183 |
+
"loss": 1.0714,
|
| 184 |
+
"step": 625
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.20833333333333334,
|
| 188 |
+
"grad_norm": 3.105868101119995,
|
| 189 |
+
"learning_rate": 3.472222222222222e-05,
|
| 190 |
+
"loss": 1.3443,
|
| 191 |
+
"step": 650
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.21634615384615385,
|
| 195 |
+
"grad_norm": 5.522253036499023,
|
| 196 |
+
"learning_rate": 3.605769230769231e-05,
|
| 197 |
+
"loss": 1.2919,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.22435897435897437,
|
| 202 |
+
"grad_norm": 2.4862594604492188,
|
| 203 |
+
"learning_rate": 3.739316239316239e-05,
|
| 204 |
+
"loss": 1.1035,
|
| 205 |
+
"step": 700
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.23237179487179488,
|
| 209 |
+
"grad_norm": 4.2892022132873535,
|
| 210 |
+
"learning_rate": 3.872863247863248e-05,
|
| 211 |
+
"loss": 1.4755,
|
| 212 |
+
"step": 725
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2403846153846154,
|
| 216 |
+
"grad_norm": 2.7745046615600586,
|
| 217 |
+
"learning_rate": 4.006410256410257e-05,
|
| 218 |
+
"loss": 1.1173,
|
| 219 |
+
"step": 750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2483974358974359,
|
| 223 |
+
"grad_norm": 3.0554754734039307,
|
| 224 |
+
"learning_rate": 4.1399572649572646e-05,
|
| 225 |
+
"loss": 1.1771,
|
| 226 |
+
"step": 775
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2564102564102564,
|
| 230 |
+
"grad_norm": 25.598337173461914,
|
| 231 |
+
"learning_rate": 4.2735042735042735e-05,
|
| 232 |
+
"loss": 1.3468,
|
| 233 |
+
"step": 800
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2644230769230769,
|
| 237 |
+
"grad_norm": 2.8250620365142822,
|
| 238 |
+
"learning_rate": 4.4070512820512824e-05,
|
| 239 |
+
"loss": 1.3123,
|
| 240 |
+
"step": 825
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2724358974358974,
|
| 244 |
+
"grad_norm": 1.0886784791946411,
|
| 245 |
+
"learning_rate": 4.5405982905982906e-05,
|
| 246 |
+
"loss": 0.9359,
|
| 247 |
+
"step": 850
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.28044871794871795,
|
| 251 |
+
"grad_norm": 2.106900453567505,
|
| 252 |
+
"learning_rate": 4.6741452991452995e-05,
|
| 253 |
+
"loss": 1.1918,
|
| 254 |
+
"step": 875
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.28846153846153844,
|
| 258 |
+
"grad_norm": 21.711318969726562,
|
| 259 |
+
"learning_rate": 4.8076923076923084e-05,
|
| 260 |
+
"loss": 1.0437,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.296474358974359,
|
| 265 |
+
"grad_norm": 2.498274803161621,
|
| 266 |
+
"learning_rate": 4.9412393162393166e-05,
|
| 267 |
+
"loss": 1.6057,
|
| 268 |
+
"step": 925
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.30448717948717946,
|
| 272 |
+
"grad_norm": 2.4843647480010986,
|
| 273 |
+
"learning_rate": 4.991690408357075e-05,
|
| 274 |
+
"loss": 1.3495,
|
| 275 |
+
"step": 950
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3125,
|
| 279 |
+
"grad_norm": 3.627005100250244,
|
| 280 |
+
"learning_rate": 4.976851851851852e-05,
|
| 281 |
+
"loss": 1.1726,
|
| 282 |
+
"step": 975
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.32051282051282054,
|
| 286 |
+
"grad_norm": 8.968929290771484,
|
| 287 |
+
"learning_rate": 4.962013295346629e-05,
|
| 288 |
+
"loss": 1.553,
|
| 289 |
+
"step": 1000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.328525641025641,
|
| 293 |
+
"grad_norm": 8.428389549255371,
|
| 294 |
+
"learning_rate": 4.947174738841406e-05,
|
| 295 |
+
"loss": 1.3413,
|
| 296 |
+
"step": 1025
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.33653846153846156,
|
| 300 |
+
"grad_norm": 2.397824287414551,
|
| 301 |
+
"learning_rate": 4.932336182336182e-05,
|
| 302 |
+
"loss": 1.4789,
|
| 303 |
+
"step": 1050
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.34455128205128205,
|
| 307 |
+
"grad_norm": 0.637091338634491,
|
| 308 |
+
"learning_rate": 4.91749762583096e-05,
|
| 309 |
+
"loss": 1.9015,
|
| 310 |
+
"step": 1075
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3525641025641026,
|
| 314 |
+
"grad_norm": 3.0233922004699707,
|
| 315 |
+
"learning_rate": 4.902659069325736e-05,
|
| 316 |
+
"loss": 1.6701,
|
| 317 |
+
"step": 1100
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3605769230769231,
|
| 321 |
+
"grad_norm": 0.6359624266624451,
|
| 322 |
+
"learning_rate": 4.887820512820513e-05,
|
| 323 |
+
"loss": 1.3547,
|
| 324 |
+
"step": 1125
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.3685897435897436,
|
| 328 |
+
"grad_norm": 2.4577717781066895,
|
| 329 |
+
"learning_rate": 4.87298195631529e-05,
|
| 330 |
+
"loss": 1.4376,
|
| 331 |
+
"step": 1150
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.3766025641025641,
|
| 335 |
+
"grad_norm": 2.6856417655944824,
|
| 336 |
+
"learning_rate": 4.858143399810067e-05,
|
| 337 |
+
"loss": 1.1159,
|
| 338 |
+
"step": 1175
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.38461538461538464,
|
| 342 |
+
"grad_norm": 2.470224618911743,
|
| 343 |
+
"learning_rate": 4.8433048433048433e-05,
|
| 344 |
+
"loss": 1.0919,
|
| 345 |
+
"step": 1200
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3926282051282051,
|
| 349 |
+
"grad_norm": 2.427523374557495,
|
| 350 |
+
"learning_rate": 4.8284662867996204e-05,
|
| 351 |
+
"loss": 1.1619,
|
| 352 |
+
"step": 1225
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.40064102564102566,
|
| 356 |
+
"grad_norm": 3.5226614475250244,
|
| 357 |
+
"learning_rate": 4.813627730294397e-05,
|
| 358 |
+
"loss": 1.2085,
|
| 359 |
+
"step": 1250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.40865384615384615,
|
| 363 |
+
"grad_norm": 3.534006357192993,
|
| 364 |
+
"learning_rate": 4.798789173789174e-05,
|
| 365 |
+
"loss": 1.1608,
|
| 366 |
+
"step": 1275
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4166666666666667,
|
| 370 |
+
"grad_norm": 3.317206621170044,
|
| 371 |
+
"learning_rate": 4.783950617283951e-05,
|
| 372 |
+
"loss": 1.1564,
|
| 373 |
+
"step": 1300
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.42467948717948717,
|
| 377 |
+
"grad_norm": 3.080186367034912,
|
| 378 |
+
"learning_rate": 4.7691120607787274e-05,
|
| 379 |
+
"loss": 1.3696,
|
| 380 |
+
"step": 1325
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.4326923076923077,
|
| 384 |
+
"grad_norm": 11.56621265411377,
|
| 385 |
+
"learning_rate": 4.7542735042735045e-05,
|
| 386 |
+
"loss": 1.0814,
|
| 387 |
+
"step": 1350
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.4407051282051282,
|
| 391 |
+
"grad_norm": 3.273176670074463,
|
| 392 |
+
"learning_rate": 4.7394349477682816e-05,
|
| 393 |
+
"loss": 1.1635,
|
| 394 |
+
"step": 1375
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.44871794871794873,
|
| 398 |
+
"grad_norm": 3.2494633197784424,
|
| 399 |
+
"learning_rate": 4.724596391263058e-05,
|
| 400 |
+
"loss": 1.1449,
|
| 401 |
+
"step": 1400
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4567307692307692,
|
| 405 |
+
"grad_norm": 3.049802303314209,
|
| 406 |
+
"learning_rate": 4.7097578347578345e-05,
|
| 407 |
+
"loss": 1.1183,
|
| 408 |
+
"step": 1425
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.46474358974358976,
|
| 412 |
+
"grad_norm": 2.630352735519409,
|
| 413 |
+
"learning_rate": 4.694919278252612e-05,
|
| 414 |
+
"loss": 1.1957,
|
| 415 |
+
"step": 1450
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.47275641025641024,
|
| 419 |
+
"grad_norm": 18.775863647460938,
|
| 420 |
+
"learning_rate": 4.6800807217473886e-05,
|
| 421 |
+
"loss": 1.1866,
|
| 422 |
+
"step": 1475
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4807692307692308,
|
| 426 |
+
"grad_norm": 3.0166356563568115,
|
| 427 |
+
"learning_rate": 4.665242165242166e-05,
|
| 428 |
+
"loss": 1.1525,
|
| 429 |
+
"step": 1500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.48878205128205127,
|
| 433 |
+
"grad_norm": 2.6879777908325195,
|
| 434 |
+
"learning_rate": 4.650403608736942e-05,
|
| 435 |
+
"loss": 1.1054,
|
| 436 |
+
"step": 1525
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.4967948717948718,
|
| 440 |
+
"grad_norm": 2.6179323196411133,
|
| 441 |
+
"learning_rate": 4.635565052231719e-05,
|
| 442 |
+
"loss": 1.1241,
|
| 443 |
+
"step": 1550
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5048076923076923,
|
| 447 |
+
"grad_norm": 2.5148842334747314,
|
| 448 |
+
"learning_rate": 4.620726495726496e-05,
|
| 449 |
+
"loss": 1.0893,
|
| 450 |
+
"step": 1575
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5128205128205128,
|
| 454 |
+
"grad_norm": 2.400773525238037,
|
| 455 |
+
"learning_rate": 4.605887939221273e-05,
|
| 456 |
+
"loss": 1.1956,
|
| 457 |
+
"step": 1600
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5208333333333334,
|
| 461 |
+
"grad_norm": 2.4545950889587402,
|
| 462 |
+
"learning_rate": 4.591049382716049e-05,
|
| 463 |
+
"loss": 1.0852,
|
| 464 |
+
"step": 1625
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5288461538461539,
|
| 468 |
+
"grad_norm": 14.888031959533691,
|
| 469 |
+
"learning_rate": 4.576210826210827e-05,
|
| 470 |
+
"loss": 1.0675,
|
| 471 |
+
"step": 1650
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5368589743589743,
|
| 475 |
+
"grad_norm": 2.9124226570129395,
|
| 476 |
+
"learning_rate": 4.5613722697056033e-05,
|
| 477 |
+
"loss": 1.2396,
|
| 478 |
+
"step": 1675
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5448717948717948,
|
| 482 |
+
"grad_norm": 2.9263124465942383,
|
| 483 |
+
"learning_rate": 4.54653371320038e-05,
|
| 484 |
+
"loss": 1.0151,
|
| 485 |
+
"step": 1700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5528846153846154,
|
| 489 |
+
"grad_norm": 19.248537063598633,
|
| 490 |
+
"learning_rate": 4.531695156695157e-05,
|
| 491 |
+
"loss": 1.0135,
|
| 492 |
+
"step": 1725
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5608974358974359,
|
| 496 |
+
"grad_norm": 3.1988131999969482,
|
| 497 |
+
"learning_rate": 4.516856600189934e-05,
|
| 498 |
+
"loss": 1.0969,
|
| 499 |
+
"step": 1750
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5689102564102564,
|
| 503 |
+
"grad_norm": 2.3869454860687256,
|
| 504 |
+
"learning_rate": 4.5020180436847104e-05,
|
| 505 |
+
"loss": 1.0116,
|
| 506 |
+
"step": 1775
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5769230769230769,
|
| 510 |
+
"grad_norm": 0.13602042198181152,
|
| 511 |
+
"learning_rate": 4.4871794871794874e-05,
|
| 512 |
+
"loss": 0.9412,
|
| 513 |
+
"step": 1800
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5849358974358975,
|
| 517 |
+
"grad_norm": 15.212441444396973,
|
| 518 |
+
"learning_rate": 4.472340930674264e-05,
|
| 519 |
+
"loss": 2.083,
|
| 520 |
+
"step": 1825
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.592948717948718,
|
| 524 |
+
"grad_norm": 2.62264084815979,
|
| 525 |
+
"learning_rate": 4.457502374169041e-05,
|
| 526 |
+
"loss": 1.1781,
|
| 527 |
+
"step": 1850
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6009615384615384,
|
| 531 |
+
"grad_norm": 4.369824409484863,
|
| 532 |
+
"learning_rate": 4.442663817663818e-05,
|
| 533 |
+
"loss": 1.0836,
|
| 534 |
+
"step": 1875
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.6089743589743589,
|
| 538 |
+
"grad_norm": 3.658947467803955,
|
| 539 |
+
"learning_rate": 4.4278252611585945e-05,
|
| 540 |
+
"loss": 1.1101,
|
| 541 |
+
"step": 1900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6169871794871795,
|
| 545 |
+
"grad_norm": 7.315481662750244,
|
| 546 |
+
"learning_rate": 4.4129867046533715e-05,
|
| 547 |
+
"loss": 1.0865,
|
| 548 |
+
"step": 1925
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.625,
|
| 552 |
+
"grad_norm": 3.12101411819458,
|
| 553 |
+
"learning_rate": 4.3981481481481486e-05,
|
| 554 |
+
"loss": 1.3192,
|
| 555 |
+
"step": 1950
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6330128205128205,
|
| 559 |
+
"grad_norm": 3.775122880935669,
|
| 560 |
+
"learning_rate": 4.383309591642925e-05,
|
| 561 |
+
"loss": 1.1183,
|
| 562 |
+
"step": 1975
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6410256410256411,
|
| 566 |
+
"grad_norm": 3.056248664855957,
|
| 567 |
+
"learning_rate": 4.368471035137702e-05,
|
| 568 |
+
"loss": 1.2241,
|
| 569 |
+
"step": 2000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.6490384615384616,
|
| 573 |
+
"grad_norm": 2.8912906646728516,
|
| 574 |
+
"learning_rate": 4.353632478632479e-05,
|
| 575 |
+
"loss": 1.1485,
|
| 576 |
+
"step": 2025
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.657051282051282,
|
| 580 |
+
"grad_norm": 2.820913076400757,
|
| 581 |
+
"learning_rate": 4.3387939221272557e-05,
|
| 582 |
+
"loss": 1.0915,
|
| 583 |
+
"step": 2050
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.6650641025641025,
|
| 587 |
+
"grad_norm": 2.897550344467163,
|
| 588 |
+
"learning_rate": 4.323955365622033e-05,
|
| 589 |
+
"loss": 1.2018,
|
| 590 |
+
"step": 2075
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6730769230769231,
|
| 594 |
+
"grad_norm": 2.613790273666382,
|
| 595 |
+
"learning_rate": 4.309116809116809e-05,
|
| 596 |
+
"loss": 1.1648,
|
| 597 |
+
"step": 2100
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.6810897435897436,
|
| 601 |
+
"grad_norm": 7.070204734802246,
|
| 602 |
+
"learning_rate": 4.294278252611586e-05,
|
| 603 |
+
"loss": 1.0465,
|
| 604 |
+
"step": 2125
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.6891025641025641,
|
| 608 |
+
"grad_norm": 2.2642040252685547,
|
| 609 |
+
"learning_rate": 4.2794396961063633e-05,
|
| 610 |
+
"loss": 1.2154,
|
| 611 |
+
"step": 2150
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.6971153846153846,
|
| 615 |
+
"grad_norm": 6.4770989418029785,
|
| 616 |
+
"learning_rate": 4.26460113960114e-05,
|
| 617 |
+
"loss": 1.3367,
|
| 618 |
+
"step": 2175
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7051282051282052,
|
| 622 |
+
"grad_norm": 2.463798999786377,
|
| 623 |
+
"learning_rate": 4.249762583095916e-05,
|
| 624 |
+
"loss": 1.1044,
|
| 625 |
+
"step": 2200
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7131410256410257,
|
| 629 |
+
"grad_norm": 2.3565423488616943,
|
| 630 |
+
"learning_rate": 4.234924026590694e-05,
|
| 631 |
+
"loss": 1.1145,
|
| 632 |
+
"step": 2225
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7211538461538461,
|
| 636 |
+
"grad_norm": 4.38250732421875,
|
| 637 |
+
"learning_rate": 4.2200854700854704e-05,
|
| 638 |
+
"loss": 1.3706,
|
| 639 |
+
"step": 2250
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7291666666666666,
|
| 643 |
+
"grad_norm": 4.341756343841553,
|
| 644 |
+
"learning_rate": 4.205246913580247e-05,
|
| 645 |
+
"loss": 1.3928,
|
| 646 |
+
"step": 2275
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.7371794871794872,
|
| 650 |
+
"grad_norm": 2.6549181938171387,
|
| 651 |
+
"learning_rate": 4.190408357075024e-05,
|
| 652 |
+
"loss": 1.1367,
|
| 653 |
+
"step": 2300
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.7451923076923077,
|
| 657 |
+
"grad_norm": 2.8435356616973877,
|
| 658 |
+
"learning_rate": 4.175569800569801e-05,
|
| 659 |
+
"loss": 1.1788,
|
| 660 |
+
"step": 2325
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7532051282051282,
|
| 664 |
+
"grad_norm": 2.6328985691070557,
|
| 665 |
+
"learning_rate": 4.1607312440645774e-05,
|
| 666 |
+
"loss": 1.0533,
|
| 667 |
+
"step": 2350
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.7612179487179487,
|
| 671 |
+
"grad_norm": 3.470773935317993,
|
| 672 |
+
"learning_rate": 4.1458926875593545e-05,
|
| 673 |
+
"loss": 1.3634,
|
| 674 |
+
"step": 2375
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7692307692307693,
|
| 678 |
+
"grad_norm": 4.499914646148682,
|
| 679 |
+
"learning_rate": 4.131054131054131e-05,
|
| 680 |
+
"loss": 1.1108,
|
| 681 |
+
"step": 2400
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7772435897435898,
|
| 685 |
+
"grad_norm": 3.694553852081299,
|
| 686 |
+
"learning_rate": 4.116215574548908e-05,
|
| 687 |
+
"loss": 1.1842,
|
| 688 |
+
"step": 2425
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.7852564102564102,
|
| 692 |
+
"grad_norm": 3.4035181999206543,
|
| 693 |
+
"learning_rate": 4.101377018043685e-05,
|
| 694 |
+
"loss": 1.1319,
|
| 695 |
+
"step": 2450
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.7932692307692307,
|
| 699 |
+
"grad_norm": 4.1596245765686035,
|
| 700 |
+
"learning_rate": 4.0865384615384615e-05,
|
| 701 |
+
"loss": 1.0582,
|
| 702 |
+
"step": 2475
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8012820512820513,
|
| 706 |
+
"grad_norm": 3.798130750656128,
|
| 707 |
+
"learning_rate": 4.0716999050332386e-05,
|
| 708 |
+
"loss": 1.409,
|
| 709 |
+
"step": 2500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8092948717948718,
|
| 713 |
+
"grad_norm": 2.8489089012145996,
|
| 714 |
+
"learning_rate": 4.0568613485280157e-05,
|
| 715 |
+
"loss": 1.2278,
|
| 716 |
+
"step": 2525
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8173076923076923,
|
| 720 |
+
"grad_norm": 10.12183952331543,
|
| 721 |
+
"learning_rate": 4.042022792022792e-05,
|
| 722 |
+
"loss": 1.1482,
|
| 723 |
+
"step": 2550
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8253205128205128,
|
| 727 |
+
"grad_norm": 3.5631895065307617,
|
| 728 |
+
"learning_rate": 4.027184235517569e-05,
|
| 729 |
+
"loss": 1.1261,
|
| 730 |
+
"step": 2575
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.8333333333333334,
|
| 734 |
+
"grad_norm": 3.193726062774658,
|
| 735 |
+
"learning_rate": 4.012345679012346e-05,
|
| 736 |
+
"loss": 1.1471,
|
| 737 |
+
"step": 2600
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.8413461538461539,
|
| 741 |
+
"grad_norm": 3.492716073989868,
|
| 742 |
+
"learning_rate": 3.997507122507123e-05,
|
| 743 |
+
"loss": 1.1515,
|
| 744 |
+
"step": 2625
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.8493589743589743,
|
| 748 |
+
"grad_norm": 3.2161166667938232,
|
| 749 |
+
"learning_rate": 3.9826685660019e-05,
|
| 750 |
+
"loss": 1.172,
|
| 751 |
+
"step": 2650
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.8573717948717948,
|
| 755 |
+
"grad_norm": 3.0341076850891113,
|
| 756 |
+
"learning_rate": 3.967830009496676e-05,
|
| 757 |
+
"loss": 1.0973,
|
| 758 |
+
"step": 2675
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.8653846153846154,
|
| 762 |
+
"grad_norm": 2.668952465057373,
|
| 763 |
+
"learning_rate": 3.952991452991453e-05,
|
| 764 |
+
"loss": 1.1266,
|
| 765 |
+
"step": 2700
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.8733974358974359,
|
| 769 |
+
"grad_norm": 2.9415502548217773,
|
| 770 |
+
"learning_rate": 3.9381528964862304e-05,
|
| 771 |
+
"loss": 1.1768,
|
| 772 |
+
"step": 2725
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.8814102564102564,
|
| 776 |
+
"grad_norm": 2.809612512588501,
|
| 777 |
+
"learning_rate": 3.923314339981007e-05,
|
| 778 |
+
"loss": 1.2288,
|
| 779 |
+
"step": 2750
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.8894230769230769,
|
| 783 |
+
"grad_norm": 3.0170650482177734,
|
| 784 |
+
"learning_rate": 3.908475783475783e-05,
|
| 785 |
+
"loss": 1.2015,
|
| 786 |
+
"step": 2775
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.8974358974358975,
|
| 790 |
+
"grad_norm": 3.068657159805298,
|
| 791 |
+
"learning_rate": 3.893637226970561e-05,
|
| 792 |
+
"loss": 1.1403,
|
| 793 |
+
"step": 2800
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.905448717948718,
|
| 797 |
+
"grad_norm": 3.0226268768310547,
|
| 798 |
+
"learning_rate": 3.8787986704653374e-05,
|
| 799 |
+
"loss": 1.0743,
|
| 800 |
+
"step": 2825
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9134615384615384,
|
| 804 |
+
"grad_norm": 3.1346497535705566,
|
| 805 |
+
"learning_rate": 3.863960113960114e-05,
|
| 806 |
+
"loss": 1.1489,
|
| 807 |
+
"step": 2850
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.9214743589743589,
|
| 811 |
+
"grad_norm": 2.6503517627716064,
|
| 812 |
+
"learning_rate": 3.849121557454891e-05,
|
| 813 |
+
"loss": 1.0438,
|
| 814 |
+
"step": 2875
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.9294871794871795,
|
| 818 |
+
"grad_norm": 3.2590255737304688,
|
| 819 |
+
"learning_rate": 3.834283000949668e-05,
|
| 820 |
+
"loss": 1.1691,
|
| 821 |
+
"step": 2900
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.9375,
|
| 825 |
+
"grad_norm": 2.9820985794067383,
|
| 826 |
+
"learning_rate": 3.8194444444444444e-05,
|
| 827 |
+
"loss": 1.2193,
|
| 828 |
+
"step": 2925
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.9455128205128205,
|
| 832 |
+
"grad_norm": 3.177725315093994,
|
| 833 |
+
"learning_rate": 3.8046058879392215e-05,
|
| 834 |
+
"loss": 1.1214,
|
| 835 |
+
"step": 2950
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9535256410256411,
|
| 839 |
+
"grad_norm": 3.0627474784851074,
|
| 840 |
+
"learning_rate": 3.789767331433998e-05,
|
| 841 |
+
"loss": 1.1154,
|
| 842 |
+
"step": 2975
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.9615384615384616,
|
| 846 |
+
"grad_norm": 3.4069268703460693,
|
| 847 |
+
"learning_rate": 3.774928774928775e-05,
|
| 848 |
+
"loss": 1.0663,
|
| 849 |
+
"step": 3000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.969551282051282,
|
| 853 |
+
"grad_norm": 3.2132928371429443,
|
| 854 |
+
"learning_rate": 3.760090218423552e-05,
|
| 855 |
+
"loss": 1.2517,
|
| 856 |
+
"step": 3025
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.9775641025641025,
|
| 860 |
+
"grad_norm": 2.8538827896118164,
|
| 861 |
+
"learning_rate": 3.7452516619183285e-05,
|
| 862 |
+
"loss": 1.1047,
|
| 863 |
+
"step": 3050
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.9855769230769231,
|
| 867 |
+
"grad_norm": 2.5981075763702393,
|
| 868 |
+
"learning_rate": 3.7304131054131056e-05,
|
| 869 |
+
"loss": 1.0797,
|
| 870 |
+
"step": 3075
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.9935897435897436,
|
| 874 |
+
"grad_norm": 3.4888546466827393,
|
| 875 |
+
"learning_rate": 3.7155745489078827e-05,
|
| 876 |
+
"loss": 1.2808,
|
| 877 |
+
"step": 3100
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.0,
|
| 881 |
+
"eval_accuracy": 0.3333333333333333,
|
| 882 |
+
"eval_f1_macro": 0.16666666666666666,
|
| 883 |
+
"eval_f1_micro": 0.3333333333333333,
|
| 884 |
+
"eval_f1_weighted": 0.16666666666666666,
|
| 885 |
+
"eval_loss": 1.1078214645385742,
|
| 886 |
+
"eval_precision_macro": 0.1111111111111111,
|
| 887 |
+
"eval_precision_micro": 0.3333333333333333,
|
| 888 |
+
"eval_precision_weighted": 0.1111111111111111,
|
| 889 |
+
"eval_recall_macro": 0.3333333333333333,
|
| 890 |
+
"eval_recall_micro": 0.3333333333333333,
|
| 891 |
+
"eval_recall_weighted": 0.3333333333333333,
|
| 892 |
+
"eval_runtime": 9174.8667,
|
| 893 |
+
"eval_samples_per_second": 0.085,
|
| 894 |
+
"eval_steps_per_second": 0.043,
|
| 895 |
+
"step": 3120
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 1.001602564102564,
|
| 899 |
+
"grad_norm": 3.344636917114258,
|
| 900 |
+
"learning_rate": 3.700735992402659e-05,
|
| 901 |
+
"loss": 1.07,
|
| 902 |
+
"step": 3125
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.0096153846153846,
|
| 906 |
+
"grad_norm": 3.0336191654205322,
|
| 907 |
+
"learning_rate": 3.685897435897436e-05,
|
| 908 |
+
"loss": 1.1197,
|
| 909 |
+
"step": 3150
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1.017628205128205,
|
| 913 |
+
"grad_norm": 2.9916834831237793,
|
| 914 |
+
"learning_rate": 3.671058879392213e-05,
|
| 915 |
+
"loss": 1.1369,
|
| 916 |
+
"step": 3175
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1.0256410256410255,
|
| 920 |
+
"grad_norm": 3.1131772994995117,
|
| 921 |
+
"learning_rate": 3.65622032288699e-05,
|
| 922 |
+
"loss": 1.1046,
|
| 923 |
+
"step": 3200
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.0336538461538463,
|
| 927 |
+
"grad_norm": 12.126420974731445,
|
| 928 |
+
"learning_rate": 3.641381766381767e-05,
|
| 929 |
+
"loss": 1.0638,
|
| 930 |
+
"step": 3225
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 1.0416666666666667,
|
| 934 |
+
"grad_norm": 2.9547293186187744,
|
| 935 |
+
"learning_rate": 3.626543209876543e-05,
|
| 936 |
+
"loss": 1.1266,
|
| 937 |
+
"step": 3250
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 1.0496794871794872,
|
| 941 |
+
"grad_norm": 2.816495895385742,
|
| 942 |
+
"learning_rate": 3.61170465337132e-05,
|
| 943 |
+
"loss": 1.1464,
|
| 944 |
+
"step": 3275
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.0576923076923077,
|
| 948 |
+
"grad_norm": 3.213977336883545,
|
| 949 |
+
"learning_rate": 3.5968660968660974e-05,
|
| 950 |
+
"loss": 1.0854,
|
| 951 |
+
"step": 3300
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 1.0657051282051282,
|
| 955 |
+
"grad_norm": 2.7554306983947754,
|
| 956 |
+
"learning_rate": 3.582027540360874e-05,
|
| 957 |
+
"loss": 1.0957,
|
| 958 |
+
"step": 3325
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.0737179487179487,
|
| 962 |
+
"grad_norm": 2.8204352855682373,
|
| 963 |
+
"learning_rate": 3.56718898385565e-05,
|
| 964 |
+
"loss": 1.1812,
|
| 965 |
+
"step": 3350
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.0817307692307692,
|
| 969 |
+
"grad_norm": 2.9325878620147705,
|
| 970 |
+
"learning_rate": 3.552350427350428e-05,
|
| 971 |
+
"loss": 1.1445,
|
| 972 |
+
"step": 3375
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 1.0897435897435896,
|
| 976 |
+
"grad_norm": 2.791325330734253,
|
| 977 |
+
"learning_rate": 3.5375118708452044e-05,
|
| 978 |
+
"loss": 1.1085,
|
| 979 |
+
"step": 3400
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 1.0977564102564104,
|
| 983 |
+
"grad_norm": 2.779751777648926,
|
| 984 |
+
"learning_rate": 3.522673314339981e-05,
|
| 985 |
+
"loss": 1.1322,
|
| 986 |
+
"step": 3425
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.1057692307692308,
|
| 990 |
+
"grad_norm": 2.7486491203308105,
|
| 991 |
+
"learning_rate": 3.507834757834758e-05,
|
| 992 |
+
"loss": 1.1122,
|
| 993 |
+
"step": 3450
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.1137820512820513,
|
| 997 |
+
"grad_norm": 2.7205097675323486,
|
| 998 |
+
"learning_rate": 3.492996201329535e-05,
|
| 999 |
+
"loss": 1.1151,
|
| 1000 |
+
"step": 3475
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.1217948717948718,
|
| 1004 |
+
"grad_norm": 2.7784721851348877,
|
| 1005 |
+
"learning_rate": 3.4781576448243114e-05,
|
| 1006 |
+
"loss": 1.149,
|
| 1007 |
+
"step": 3500
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 1.1298076923076923,
|
| 1011 |
+
"grad_norm": 2.705922842025757,
|
| 1012 |
+
"learning_rate": 3.4633190883190885e-05,
|
| 1013 |
+
"loss": 1.1167,
|
| 1014 |
+
"step": 3525
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 1.1378205128205128,
|
| 1018 |
+
"grad_norm": 2.7289721965789795,
|
| 1019 |
+
"learning_rate": 3.448480531813865e-05,
|
| 1020 |
+
"loss": 1.1299,
|
| 1021 |
+
"step": 3550
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 1.1458333333333333,
|
| 1025 |
+
"grad_norm": 3.292160749435425,
|
| 1026 |
+
"learning_rate": 3.4336419753086427e-05,
|
| 1027 |
+
"loss": 1.1059,
|
| 1028 |
+
"step": 3575
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.1538461538461537,
|
| 1032 |
+
"grad_norm": 16.216575622558594,
|
| 1033 |
+
"learning_rate": 3.418803418803419e-05,
|
| 1034 |
+
"loss": 1.1736,
|
| 1035 |
+
"step": 3600
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 1.1618589743589745,
|
| 1039 |
+
"grad_norm": 2.9916632175445557,
|
| 1040 |
+
"learning_rate": 3.4039648622981955e-05,
|
| 1041 |
+
"loss": 1.1225,
|
| 1042 |
+
"step": 3625
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 1.169871794871795,
|
| 1046 |
+
"grad_norm": 2.8906617164611816,
|
| 1047 |
+
"learning_rate": 3.3891263057929726e-05,
|
| 1048 |
+
"loss": 1.1365,
|
| 1049 |
+
"step": 3650
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.1778846153846154,
|
| 1053 |
+
"grad_norm": 2.839803695678711,
|
| 1054 |
+
"learning_rate": 3.37428774928775e-05,
|
| 1055 |
+
"loss": 1.0881,
|
| 1056 |
+
"step": 3675
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 1.185897435897436,
|
| 1060 |
+
"grad_norm": 3.1688859462738037,
|
| 1061 |
+
"learning_rate": 3.359449192782526e-05,
|
| 1062 |
+
"loss": 1.1115,
|
| 1063 |
+
"step": 3700
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 1.1939102564102564,
|
| 1067 |
+
"grad_norm": 19.920875549316406,
|
| 1068 |
+
"learning_rate": 3.344610636277303e-05,
|
| 1069 |
+
"loss": 1.101,
|
| 1070 |
+
"step": 3725
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.2019230769230769,
|
| 1074 |
+
"grad_norm": 3.2455990314483643,
|
| 1075 |
+
"learning_rate": 3.32977207977208e-05,
|
| 1076 |
+
"loss": 1.2065,
|
| 1077 |
+
"step": 3750
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.2099358974358974,
|
| 1081 |
+
"grad_norm": 3.3635494709014893,
|
| 1082 |
+
"learning_rate": 3.314933523266857e-05,
|
| 1083 |
+
"loss": 1.2774,
|
| 1084 |
+
"step": 3775
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 1.217948717948718,
|
| 1088 |
+
"grad_norm": 2.943542718887329,
|
| 1089 |
+
"learning_rate": 3.300094966761634e-05,
|
| 1090 |
+
"loss": 1.1054,
|
| 1091 |
+
"step": 3800
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.2259615384615385,
|
| 1095 |
+
"grad_norm": 2.629531145095825,
|
| 1096 |
+
"learning_rate": 3.28525641025641e-05,
|
| 1097 |
+
"loss": 1.039,
|
| 1098 |
+
"step": 3825
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.233974358974359,
|
| 1102 |
+
"grad_norm": 3.587048292160034,
|
| 1103 |
+
"learning_rate": 3.270417853751187e-05,
|
| 1104 |
+
"loss": 1.2413,
|
| 1105 |
+
"step": 3850
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.2419871794871795,
|
| 1109 |
+
"grad_norm": 3.6215555667877197,
|
| 1110 |
+
"learning_rate": 3.2555792972459644e-05,
|
| 1111 |
+
"loss": 1.1113,
|
| 1112 |
+
"step": 3875
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.25,
|
| 1116 |
+
"grad_norm": 2.6239843368530273,
|
| 1117 |
+
"learning_rate": 3.240740740740741e-05,
|
| 1118 |
+
"loss": 1.129,
|
| 1119 |
+
"step": 3900
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.2580128205128205,
|
| 1123 |
+
"grad_norm": 3.372828960418701,
|
| 1124 |
+
"learning_rate": 3.225902184235517e-05,
|
| 1125 |
+
"loss": 1.2073,
|
| 1126 |
+
"step": 3925
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.266025641025641,
|
| 1130 |
+
"grad_norm": 10.59459114074707,
|
| 1131 |
+
"learning_rate": 3.211063627730295e-05,
|
| 1132 |
+
"loss": 1.1664,
|
| 1133 |
+
"step": 3950
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.2740384615384617,
|
| 1137 |
+
"grad_norm": 3.2925708293914795,
|
| 1138 |
+
"learning_rate": 3.1962250712250714e-05,
|
| 1139 |
+
"loss": 1.1196,
|
| 1140 |
+
"step": 3975
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 1.282051282051282,
|
| 1144 |
+
"grad_norm": 2.755282402038574,
|
| 1145 |
+
"learning_rate": 3.181386514719848e-05,
|
| 1146 |
+
"loss": 1.07,
|
| 1147 |
+
"step": 4000
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 1.2900641025641026,
|
| 1151 |
+
"grad_norm": 3.2294089794158936,
|
| 1152 |
+
"learning_rate": 3.166547958214625e-05,
|
| 1153 |
+
"loss": 1.233,
|
| 1154 |
+
"step": 4025
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.2980769230769231,
|
| 1158 |
+
"grad_norm": 2.9943435192108154,
|
| 1159 |
+
"learning_rate": 3.151709401709402e-05,
|
| 1160 |
+
"loss": 1.0481,
|
| 1161 |
+
"step": 4050
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.3060897435897436,
|
| 1165 |
+
"grad_norm": 3.735652446746826,
|
| 1166 |
+
"learning_rate": 3.136870845204179e-05,
|
| 1167 |
+
"loss": 1.1919,
|
| 1168 |
+
"step": 4075
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 1.314102564102564,
|
| 1172 |
+
"grad_norm": 3.579597234725952,
|
| 1173 |
+
"learning_rate": 3.1220322886989555e-05,
|
| 1174 |
+
"loss": 1.1711,
|
| 1175 |
+
"step": 4100
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.3221153846153846,
|
| 1179 |
+
"grad_norm": 2.677807331085205,
|
| 1180 |
+
"learning_rate": 3.107193732193732e-05,
|
| 1181 |
+
"loss": 1.0804,
|
| 1182 |
+
"step": 4125
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.330128205128205,
|
| 1186 |
+
"grad_norm": 3.562013864517212,
|
| 1187 |
+
"learning_rate": 3.09235517568851e-05,
|
| 1188 |
+
"loss": 1.136,
|
| 1189 |
+
"step": 4150
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 1.3381410256410255,
|
| 1193 |
+
"grad_norm": 3.4050207138061523,
|
| 1194 |
+
"learning_rate": 3.077516619183286e-05,
|
| 1195 |
+
"loss": 1.155,
|
| 1196 |
+
"step": 4175
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.3461538461538463,
|
| 1200 |
+
"grad_norm": 3.438920497894287,
|
| 1201 |
+
"learning_rate": 3.0626780626780625e-05,
|
| 1202 |
+
"loss": 1.0936,
|
| 1203 |
+
"step": 4200
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.3541666666666667,
|
| 1207 |
+
"grad_norm": 16.807218551635742,
|
| 1208 |
+
"learning_rate": 3.04783950617284e-05,
|
| 1209 |
+
"loss": 1.1692,
|
| 1210 |
+
"step": 4225
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.3621794871794872,
|
| 1214 |
+
"grad_norm": 3.7141730785369873,
|
| 1215 |
+
"learning_rate": 3.0330009496676163e-05,
|
| 1216 |
+
"loss": 1.1516,
|
| 1217 |
+
"step": 4250
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.3701923076923077,
|
| 1221 |
+
"grad_norm": 3.2258071899414062,
|
| 1222 |
+
"learning_rate": 3.018162393162393e-05,
|
| 1223 |
+
"loss": 1.1094,
|
| 1224 |
+
"step": 4275
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 1.3782051282051282,
|
| 1228 |
+
"grad_norm": 3.3547050952911377,
|
| 1229 |
+
"learning_rate": 3.0033238366571702e-05,
|
| 1230 |
+
"loss": 1.0942,
|
| 1231 |
+
"step": 4300
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 1.3862179487179487,
|
| 1235 |
+
"grad_norm": 3.3047409057617188,
|
| 1236 |
+
"learning_rate": 2.988485280151947e-05,
|
| 1237 |
+
"loss": 1.1491,
|
| 1238 |
+
"step": 4325
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.3942307692307692,
|
| 1242 |
+
"grad_norm": 3.3299431800842285,
|
| 1243 |
+
"learning_rate": 2.9736467236467237e-05,
|
| 1244 |
+
"loss": 1.1287,
|
| 1245 |
+
"step": 4350
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.4022435897435899,
|
| 1249 |
+
"grad_norm": 3.3957812786102295,
|
| 1250 |
+
"learning_rate": 2.9588081671415008e-05,
|
| 1251 |
+
"loss": 1.15,
|
| 1252 |
+
"step": 4375
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 1.4102564102564101,
|
| 1256 |
+
"grad_norm": 3.2437939643859863,
|
| 1257 |
+
"learning_rate": 2.9439696106362775e-05,
|
| 1258 |
+
"loss": 1.1289,
|
| 1259 |
+
"step": 4400
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.4182692307692308,
|
| 1263 |
+
"grad_norm": 2.974682092666626,
|
| 1264 |
+
"learning_rate": 2.929131054131054e-05,
|
| 1265 |
+
"loss": 1.0985,
|
| 1266 |
+
"step": 4425
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.4262820512820513,
|
| 1270 |
+
"grad_norm": 3.461221694946289,
|
| 1271 |
+
"learning_rate": 2.9142924976258314e-05,
|
| 1272 |
+
"loss": 1.059,
|
| 1273 |
+
"step": 4450
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 1.4342948717948718,
|
| 1277 |
+
"grad_norm": 3.822312355041504,
|
| 1278 |
+
"learning_rate": 2.8994539411206078e-05,
|
| 1279 |
+
"loss": 1.1434,
|
| 1280 |
+
"step": 4475
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.4423076923076923,
|
| 1284 |
+
"grad_norm": 3.6015384197235107,
|
| 1285 |
+
"learning_rate": 2.8846153846153845e-05,
|
| 1286 |
+
"loss": 1.1214,
|
| 1287 |
+
"step": 4500
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.4503205128205128,
|
| 1291 |
+
"grad_norm": 3.0729997158050537,
|
| 1292 |
+
"learning_rate": 2.8697768281101616e-05,
|
| 1293 |
+
"loss": 1.0615,
|
| 1294 |
+
"step": 4525
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 1.4583333333333333,
|
| 1298 |
+
"grad_norm": 3.443614959716797,
|
| 1299 |
+
"learning_rate": 2.8549382716049384e-05,
|
| 1300 |
+
"loss": 1.1941,
|
| 1301 |
+
"step": 4550
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.4663461538461537,
|
| 1305 |
+
"grad_norm": 3.0241081714630127,
|
| 1306 |
+
"learning_rate": 2.8400997150997155e-05,
|
| 1307 |
+
"loss": 1.1226,
|
| 1308 |
+
"step": 4575
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 1.4743589743589745,
|
| 1312 |
+
"grad_norm": 3.3162217140197754,
|
| 1313 |
+
"learning_rate": 2.8252611585944922e-05,
|
| 1314 |
+
"loss": 1.0887,
|
| 1315 |
+
"step": 4600
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.482371794871795,
|
| 1319 |
+
"grad_norm": 2.835750102996826,
|
| 1320 |
+
"learning_rate": 2.8104226020892686e-05,
|
| 1321 |
+
"loss": 1.1223,
|
| 1322 |
+
"step": 4625
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 1.4903846153846154,
|
| 1326 |
+
"grad_norm": 3.096282720565796,
|
| 1327 |
+
"learning_rate": 2.795584045584046e-05,
|
| 1328 |
+
"loss": 1.1135,
|
| 1329 |
+
"step": 4650
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 1.498397435897436,
|
| 1333 |
+
"grad_norm": 14.913049697875977,
|
| 1334 |
+
"learning_rate": 2.7807454890788225e-05,
|
| 1335 |
+
"loss": 1.1648,
|
| 1336 |
+
"step": 4675
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 1.5064102564102564,
|
| 1340 |
+
"grad_norm": 3.365718126296997,
|
| 1341 |
+
"learning_rate": 2.7659069325735992e-05,
|
| 1342 |
+
"loss": 1.1352,
|
| 1343 |
+
"step": 4700
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.5144230769230769,
|
| 1347 |
+
"grad_norm": 3.1023740768432617,
|
| 1348 |
+
"learning_rate": 2.7510683760683763e-05,
|
| 1349 |
+
"loss": 1.1149,
|
| 1350 |
+
"step": 4725
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.5224358974358974,
|
| 1354 |
+
"grad_norm": 2.9160704612731934,
|
| 1355 |
+
"learning_rate": 2.736229819563153e-05,
|
| 1356 |
+
"loss": 1.1208,
|
| 1357 |
+
"step": 4750
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 1.530448717948718,
|
| 1361 |
+
"grad_norm": 2.849268674850464,
|
| 1362 |
+
"learning_rate": 2.72139126305793e-05,
|
| 1363 |
+
"loss": 1.1122,
|
| 1364 |
+
"step": 4775
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 1.5384615384615383,
|
| 1368 |
+
"grad_norm": 3.153027057647705,
|
| 1369 |
+
"learning_rate": 2.706552706552707e-05,
|
| 1370 |
+
"loss": 1.1057,
|
| 1371 |
+
"step": 4800
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 1.546474358974359,
|
| 1375 |
+
"grad_norm": 3.0611486434936523,
|
| 1376 |
+
"learning_rate": 2.6917141500474833e-05,
|
| 1377 |
+
"loss": 1.1299,
|
| 1378 |
+
"step": 4825
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 1.5544871794871795,
|
| 1382 |
+
"grad_norm": 3.00612211227417,
|
| 1383 |
+
"learning_rate": 2.67687559354226e-05,
|
| 1384 |
+
"loss": 1.1102,
|
| 1385 |
+
"step": 4850
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.5625,
|
| 1389 |
+
"grad_norm": 3.0686211585998535,
|
| 1390 |
+
"learning_rate": 2.6620370370370372e-05,
|
| 1391 |
+
"loss": 1.0791,
|
| 1392 |
+
"step": 4875
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 1.5705128205128205,
|
| 1396 |
+
"grad_norm": 3.0146031379699707,
|
| 1397 |
+
"learning_rate": 2.647198480531814e-05,
|
| 1398 |
+
"loss": 1.0862,
|
| 1399 |
+
"step": 4900
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 1.578525641025641,
|
| 1403 |
+
"grad_norm": 3.0399394035339355,
|
| 1404 |
+
"learning_rate": 2.6323599240265907e-05,
|
| 1405 |
+
"loss": 1.1171,
|
| 1406 |
+
"step": 4925
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.5865384615384617,
|
| 1410 |
+
"grad_norm": 3.268636703491211,
|
| 1411 |
+
"learning_rate": 2.6175213675213678e-05,
|
| 1412 |
+
"loss": 1.1633,
|
| 1413 |
+
"step": 4950
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.594551282051282,
|
| 1417 |
+
"grad_norm": 3.0657496452331543,
|
| 1418 |
+
"learning_rate": 2.6026828110161445e-05,
|
| 1419 |
+
"loss": 1.0906,
|
| 1420 |
+
"step": 4975
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.6025641025641026,
|
| 1424 |
+
"grad_norm": 3.19344162940979,
|
| 1425 |
+
"learning_rate": 2.587844254510921e-05,
|
| 1426 |
+
"loss": 1.1373,
|
| 1427 |
+
"step": 5000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.6105769230769231,
|
| 1431 |
+
"grad_norm": 2.9674160480499268,
|
| 1432 |
+
"learning_rate": 2.5730056980056984e-05,
|
| 1433 |
+
"loss": 1.1195,
|
| 1434 |
+
"step": 5025
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1.6185897435897436,
|
| 1438 |
+
"grad_norm": 3.3577349185943604,
|
| 1439 |
+
"learning_rate": 2.5581671415004748e-05,
|
| 1440 |
+
"loss": 1.0626,
|
| 1441 |
+
"step": 5050
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 1.626602564102564,
|
| 1445 |
+
"grad_norm": 2.8500237464904785,
|
| 1446 |
+
"learning_rate": 2.5433285849952516e-05,
|
| 1447 |
+
"loss": 1.0467,
|
| 1448 |
+
"step": 5075
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.6346153846153846,
|
| 1452 |
+
"grad_norm": 2.666741371154785,
|
| 1453 |
+
"learning_rate": 2.5284900284900286e-05,
|
| 1454 |
+
"loss": 1.1259,
|
| 1455 |
+
"step": 5100
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.6426282051282053,
|
| 1459 |
+
"grad_norm": 3.875152587890625,
|
| 1460 |
+
"learning_rate": 2.5136514719848054e-05,
|
| 1461 |
+
"loss": 1.0974,
|
| 1462 |
+
"step": 5125
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 1.6506410256410255,
|
| 1466 |
+
"grad_norm": 3.22537899017334,
|
| 1467 |
+
"learning_rate": 2.498812915479582e-05,
|
| 1468 |
+
"loss": 1.1807,
|
| 1469 |
+
"step": 5150
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.6586538461538463,
|
| 1473 |
+
"grad_norm": 3.05743670463562,
|
| 1474 |
+
"learning_rate": 2.4839743589743592e-05,
|
| 1475 |
+
"loss": 1.1345,
|
| 1476 |
+
"step": 5175
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 1.6666666666666665,
|
| 1480 |
+
"grad_norm": 2.7266359329223633,
|
| 1481 |
+
"learning_rate": 2.4691358024691357e-05,
|
| 1482 |
+
"loss": 1.0556,
|
| 1483 |
+
"step": 5200
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 1.6746794871794872,
|
| 1487 |
+
"grad_norm": 2.699350118637085,
|
| 1488 |
+
"learning_rate": 2.4542972459639127e-05,
|
| 1489 |
+
"loss": 1.0807,
|
| 1490 |
+
"step": 5225
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.6826923076923077,
|
| 1494 |
+
"grad_norm": 3.3615918159484863,
|
| 1495 |
+
"learning_rate": 2.4394586894586895e-05,
|
| 1496 |
+
"loss": 1.1998,
|
| 1497 |
+
"step": 5250
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.6907051282051282,
|
| 1501 |
+
"grad_norm": 2.972320079803467,
|
| 1502 |
+
"learning_rate": 2.4246201329534666e-05,
|
| 1503 |
+
"loss": 1.1705,
|
| 1504 |
+
"step": 5275
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 1.6987179487179487,
|
| 1508 |
+
"grad_norm": 2.8785436153411865,
|
| 1509 |
+
"learning_rate": 2.409781576448243e-05,
|
| 1510 |
+
"loss": 1.1266,
|
| 1511 |
+
"step": 5300
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.7067307692307692,
|
| 1515 |
+
"grad_norm": 2.937675714492798,
|
| 1516 |
+
"learning_rate": 2.39494301994302e-05,
|
| 1517 |
+
"loss": 1.1328,
|
| 1518 |
+
"step": 5325
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 1.7147435897435899,
|
| 1522 |
+
"grad_norm": 2.808654546737671,
|
| 1523 |
+
"learning_rate": 2.380104463437797e-05,
|
| 1524 |
+
"loss": 1.1063,
|
| 1525 |
+
"step": 5350
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 1.7227564102564101,
|
| 1529 |
+
"grad_norm": 2.7585086822509766,
|
| 1530 |
+
"learning_rate": 2.3652659069325736e-05,
|
| 1531 |
+
"loss": 1.1229,
|
| 1532 |
+
"step": 5375
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 1.7307692307692308,
|
| 1536 |
+
"grad_norm": 3.1439080238342285,
|
| 1537 |
+
"learning_rate": 2.3504273504273504e-05,
|
| 1538 |
+
"loss": 1.1271,
|
| 1539 |
+
"step": 5400
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 1.7387820512820513,
|
| 1543 |
+
"grad_norm": 3.092245578765869,
|
| 1544 |
+
"learning_rate": 2.3355887939221274e-05,
|
| 1545 |
+
"loss": 1.1195,
|
| 1546 |
+
"step": 5425
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 1.7467948717948718,
|
| 1550 |
+
"grad_norm": 2.782118797302246,
|
| 1551 |
+
"learning_rate": 2.3207502374169042e-05,
|
| 1552 |
+
"loss": 1.0295,
|
| 1553 |
+
"step": 5450
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 1.7548076923076923,
|
| 1557 |
+
"grad_norm": 19.478776931762695,
|
| 1558 |
+
"learning_rate": 2.305911680911681e-05,
|
| 1559 |
+
"loss": 1.0735,
|
| 1560 |
+
"step": 5475
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.7628205128205128,
|
| 1564 |
+
"grad_norm": 3.266461133956909,
|
| 1565 |
+
"learning_rate": 2.291073124406458e-05,
|
| 1566 |
+
"loss": 1.3259,
|
| 1567 |
+
"step": 5500
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 1.7708333333333335,
|
| 1571 |
+
"grad_norm": 3.5363993644714355,
|
| 1572 |
+
"learning_rate": 2.2762345679012348e-05,
|
| 1573 |
+
"loss": 1.1531,
|
| 1574 |
+
"step": 5525
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 1.7788461538461537,
|
| 1578 |
+
"grad_norm": 16.577512741088867,
|
| 1579 |
+
"learning_rate": 2.2613960113960116e-05,
|
| 1580 |
+
"loss": 1.2163,
|
| 1581 |
+
"step": 5550
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.7868589743589745,
|
| 1585 |
+
"grad_norm": 3.019594669342041,
|
| 1586 |
+
"learning_rate": 2.2465574548907883e-05,
|
| 1587 |
+
"loss": 1.1258,
|
| 1588 |
+
"step": 5575
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 1.7948717948717947,
|
| 1592 |
+
"grad_norm": 3.258033275604248,
|
| 1593 |
+
"learning_rate": 2.2317188983855654e-05,
|
| 1594 |
+
"loss": 1.083,
|
| 1595 |
+
"step": 5600
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 1.8028846153846154,
|
| 1599 |
+
"grad_norm": 15.097603797912598,
|
| 1600 |
+
"learning_rate": 2.2168803418803418e-05,
|
| 1601 |
+
"loss": 1.1652,
|
| 1602 |
+
"step": 5625
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 1.810897435897436,
|
| 1606 |
+
"grad_norm": 13.604498863220215,
|
| 1607 |
+
"learning_rate": 2.202041785375119e-05,
|
| 1608 |
+
"loss": 1.1586,
|
| 1609 |
+
"step": 5650
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 1.8189102564102564,
|
| 1613 |
+
"grad_norm": 3.0593981742858887,
|
| 1614 |
+
"learning_rate": 2.1872032288698957e-05,
|
| 1615 |
+
"loss": 1.0843,
|
| 1616 |
+
"step": 5675
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.8269230769230769,
|
| 1620 |
+
"grad_norm": 2.907891035079956,
|
| 1621 |
+
"learning_rate": 2.1723646723646724e-05,
|
| 1622 |
+
"loss": 1.0899,
|
| 1623 |
+
"step": 5700
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.8349358974358974,
|
| 1627 |
+
"grad_norm": 3.4455509185791016,
|
| 1628 |
+
"learning_rate": 2.157526115859449e-05,
|
| 1629 |
+
"loss": 1.1503,
|
| 1630 |
+
"step": 5725
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.842948717948718,
|
| 1634 |
+
"grad_norm": 2.9597268104553223,
|
| 1635 |
+
"learning_rate": 2.1426875593542263e-05,
|
| 1636 |
+
"loss": 1.0963,
|
| 1637 |
+
"step": 5750
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 1.8509615384615383,
|
| 1641 |
+
"grad_norm": 2.6572976112365723,
|
| 1642 |
+
"learning_rate": 2.127849002849003e-05,
|
| 1643 |
+
"loss": 1.1049,
|
| 1644 |
+
"step": 5775
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 1.858974358974359,
|
| 1648 |
+
"grad_norm": 3.2196149826049805,
|
| 1649 |
+
"learning_rate": 2.1130104463437798e-05,
|
| 1650 |
+
"loss": 1.0864,
|
| 1651 |
+
"step": 5800
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 1.8669871794871795,
|
| 1655 |
+
"grad_norm": 2.79978346824646,
|
| 1656 |
+
"learning_rate": 2.0981718898385565e-05,
|
| 1657 |
+
"loss": 1.1215,
|
| 1658 |
+
"step": 5825
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 1.875,
|
| 1662 |
+
"grad_norm": 3.0026066303253174,
|
| 1663 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 1664 |
+
"loss": 1.1098,
|
| 1665 |
+
"step": 5850
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.8830128205128205,
|
| 1669 |
+
"grad_norm": 3.078766107559204,
|
| 1670 |
+
"learning_rate": 2.06849477682811e-05,
|
| 1671 |
+
"loss": 1.0868,
|
| 1672 |
+
"step": 5875
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 1.891025641025641,
|
| 1676 |
+
"grad_norm": 2.9932262897491455,
|
| 1677 |
+
"learning_rate": 2.053656220322887e-05,
|
| 1678 |
+
"loss": 1.0917,
|
| 1679 |
+
"step": 5900
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 1.8990384615384617,
|
| 1683 |
+
"grad_norm": 2.943328380584717,
|
| 1684 |
+
"learning_rate": 2.038817663817664e-05,
|
| 1685 |
+
"loss": 1.1566,
|
| 1686 |
+
"step": 5925
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.907051282051282,
|
| 1690 |
+
"grad_norm": 3.0116453170776367,
|
| 1691 |
+
"learning_rate": 2.0239791073124406e-05,
|
| 1692 |
+
"loss": 1.0899,
|
| 1693 |
+
"step": 5950
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 1.9150641025641026,
|
| 1697 |
+
"grad_norm": 2.7467472553253174,
|
| 1698 |
+
"learning_rate": 2.0091405508072174e-05,
|
| 1699 |
+
"loss": 1.1183,
|
| 1700 |
+
"step": 5975
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 1.9230769230769231,
|
| 1704 |
+
"grad_norm": 3.1416590213775635,
|
| 1705 |
+
"learning_rate": 1.9943019943019945e-05,
|
| 1706 |
+
"loss": 1.1246,
|
| 1707 |
+
"step": 6000
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.9310897435897436,
|
| 1711 |
+
"grad_norm": 3.230945110321045,
|
| 1712 |
+
"learning_rate": 1.9794634377967712e-05,
|
| 1713 |
+
"loss": 1.057,
|
| 1714 |
+
"step": 6025
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 1.939102564102564,
|
| 1718 |
+
"grad_norm": 3.211606979370117,
|
| 1719 |
+
"learning_rate": 1.964624881291548e-05,
|
| 1720 |
+
"loss": 1.0585,
|
| 1721 |
+
"step": 6050
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 1.9471153846153846,
|
| 1725 |
+
"grad_norm": 2.7880783081054688,
|
| 1726 |
+
"learning_rate": 1.949786324786325e-05,
|
| 1727 |
+
"loss": 1.0967,
|
| 1728 |
+
"step": 6075
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.9551282051282053,
|
| 1732 |
+
"grad_norm": 3.271465539932251,
|
| 1733 |
+
"learning_rate": 1.9349477682811018e-05,
|
| 1734 |
+
"loss": 1.1564,
|
| 1735 |
+
"step": 6100
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.9631410256410255,
|
| 1739 |
+
"grad_norm": 2.839891195297241,
|
| 1740 |
+
"learning_rate": 1.9201092117758786e-05,
|
| 1741 |
+
"loss": 1.1115,
|
| 1742 |
+
"step": 6125
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 1.9711538461538463,
|
| 1746 |
+
"grad_norm": 2.828284502029419,
|
| 1747 |
+
"learning_rate": 1.9052706552706553e-05,
|
| 1748 |
+
"loss": 1.1044,
|
| 1749 |
+
"step": 6150
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.9791666666666665,
|
| 1753 |
+
"grad_norm": 2.8832826614379883,
|
| 1754 |
+
"learning_rate": 1.8904320987654324e-05,
|
| 1755 |
+
"loss": 1.1098,
|
| 1756 |
+
"step": 6175
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 1.9871794871794872,
|
| 1760 |
+
"grad_norm": 3.1509151458740234,
|
| 1761 |
+
"learning_rate": 1.8755935422602088e-05,
|
| 1762 |
+
"loss": 1.0794,
|
| 1763 |
+
"step": 6200
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.9951923076923077,
|
| 1767 |
+
"grad_norm": 3.185197114944458,
|
| 1768 |
+
"learning_rate": 1.860754985754986e-05,
|
| 1769 |
+
"loss": 1.1081,
|
| 1770 |
+
"step": 6225
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 2.0,
|
| 1774 |
+
"eval_accuracy": 0.3333333333333333,
|
| 1775 |
+
"eval_f1_macro": 0.16666666666666666,
|
| 1776 |
+
"eval_f1_micro": 0.3333333333333333,
|
| 1777 |
+
"eval_f1_weighted": 0.16666666666666666,
|
| 1778 |
+
"eval_loss": 1.1047594547271729,
|
| 1779 |
+
"eval_precision_macro": 0.1111111111111111,
|
| 1780 |
+
"eval_precision_micro": 0.3333333333333333,
|
| 1781 |
+
"eval_precision_weighted": 0.1111111111111111,
|
| 1782 |
+
"eval_recall_macro": 0.3333333333333333,
|
| 1783 |
+
"eval_recall_micro": 0.3333333333333333,
|
| 1784 |
+
"eval_recall_weighted": 0.3333333333333333,
|
| 1785 |
+
"eval_runtime": 8884.3599,
|
| 1786 |
+
"eval_samples_per_second": 0.088,
|
| 1787 |
+
"eval_steps_per_second": 0.044,
|
| 1788 |
+
"step": 6240
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.003205128205128,
|
| 1792 |
+
"grad_norm": 2.789577007293701,
|
| 1793 |
+
"learning_rate": 1.8459164292497627e-05,
|
| 1794 |
+
"loss": 1.0774,
|
| 1795 |
+
"step": 6250
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.011217948717949,
|
| 1799 |
+
"grad_norm": 14.273537635803223,
|
| 1800 |
+
"learning_rate": 1.8310778727445398e-05,
|
| 1801 |
+
"loss": 1.0926,
|
| 1802 |
+
"step": 6275
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.019230769230769,
|
| 1806 |
+
"grad_norm": 9.026211738586426,
|
| 1807 |
+
"learning_rate": 1.8162393162393162e-05,
|
| 1808 |
+
"loss": 1.1515,
|
| 1809 |
+
"step": 6300
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.02724358974359,
|
| 1813 |
+
"grad_norm": 3.3749074935913086,
|
| 1814 |
+
"learning_rate": 1.8014007597340933e-05,
|
| 1815 |
+
"loss": 1.0652,
|
| 1816 |
+
"step": 6325
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.03525641025641,
|
| 1820 |
+
"grad_norm": 17.449108123779297,
|
| 1821 |
+
"learning_rate": 1.78656220322887e-05,
|
| 1822 |
+
"loss": 1.0352,
|
| 1823 |
+
"step": 6350
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.043269230769231,
|
| 1827 |
+
"grad_norm": 2.851402521133423,
|
| 1828 |
+
"learning_rate": 1.7717236467236468e-05,
|
| 1829 |
+
"loss": 1.2282,
|
| 1830 |
+
"step": 6375
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.051282051282051,
|
| 1834 |
+
"grad_norm": 3.254828691482544,
|
| 1835 |
+
"learning_rate": 1.7568850902184235e-05,
|
| 1836 |
+
"loss": 1.1782,
|
| 1837 |
+
"step": 6400
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.059294871794872,
|
| 1841 |
+
"grad_norm": 2.91908597946167,
|
| 1842 |
+
"learning_rate": 1.7420465337132006e-05,
|
| 1843 |
+
"loss": 1.1696,
|
| 1844 |
+
"step": 6425
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.0673076923076925,
|
| 1848 |
+
"grad_norm": 18.977935791015625,
|
| 1849 |
+
"learning_rate": 1.727207977207977e-05,
|
| 1850 |
+
"loss": 1.1896,
|
| 1851 |
+
"step": 6450
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.0753205128205128,
|
| 1855 |
+
"grad_norm": 2.8999760150909424,
|
| 1856 |
+
"learning_rate": 1.712369420702754e-05,
|
| 1857 |
+
"loss": 1.0776,
|
| 1858 |
+
"step": 6475
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.0833333333333335,
|
| 1862 |
+
"grad_norm": 3.6984691619873047,
|
| 1863 |
+
"learning_rate": 1.697530864197531e-05,
|
| 1864 |
+
"loss": 1.0885,
|
| 1865 |
+
"step": 6500
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.0913461538461537,
|
| 1869 |
+
"grad_norm": 3.422837495803833,
|
| 1870 |
+
"learning_rate": 1.682692307692308e-05,
|
| 1871 |
+
"loss": 1.2089,
|
| 1872 |
+
"step": 6525
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.0993589743589745,
|
| 1876 |
+
"grad_norm": 3.5225255489349365,
|
| 1877 |
+
"learning_rate": 1.6678537511870844e-05,
|
| 1878 |
+
"loss": 1.1512,
|
| 1879 |
+
"step": 6550
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.1073717948717947,
|
| 1883 |
+
"grad_norm": 2.9733846187591553,
|
| 1884 |
+
"learning_rate": 1.6530151946818615e-05,
|
| 1885 |
+
"loss": 1.1133,
|
| 1886 |
+
"step": 6575
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.1153846153846154,
|
| 1890 |
+
"grad_norm": 3.3690342903137207,
|
| 1891 |
+
"learning_rate": 1.6381766381766382e-05,
|
| 1892 |
+
"loss": 1.0982,
|
| 1893 |
+
"step": 6600
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.123397435897436,
|
| 1897 |
+
"grad_norm": 11.055793762207031,
|
| 1898 |
+
"learning_rate": 1.623338081671415e-05,
|
| 1899 |
+
"loss": 1.0777,
|
| 1900 |
+
"step": 6625
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.1314102564102564,
|
| 1904 |
+
"grad_norm": 2.8852550983428955,
|
| 1905 |
+
"learning_rate": 1.608499525166192e-05,
|
| 1906 |
+
"loss": 1.0926,
|
| 1907 |
+
"step": 6650
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.139423076923077,
|
| 1911 |
+
"grad_norm": 3.1558687686920166,
|
| 1912 |
+
"learning_rate": 1.5936609686609688e-05,
|
| 1913 |
+
"loss": 1.1717,
|
| 1914 |
+
"step": 6675
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.1474358974358974,
|
| 1918 |
+
"grad_norm": 3.21858811378479,
|
| 1919 |
+
"learning_rate": 1.5788224121557456e-05,
|
| 1920 |
+
"loss": 1.102,
|
| 1921 |
+
"step": 6700
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.155448717948718,
|
| 1925 |
+
"grad_norm": 2.9978301525115967,
|
| 1926 |
+
"learning_rate": 1.5639838556505223e-05,
|
| 1927 |
+
"loss": 1.1034,
|
| 1928 |
+
"step": 6725
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.1634615384615383,
|
| 1932 |
+
"grad_norm": 2.7126989364624023,
|
| 1933 |
+
"learning_rate": 1.5491452991452994e-05,
|
| 1934 |
+
"loss": 1.1177,
|
| 1935 |
+
"step": 6750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.171474358974359,
|
| 1939 |
+
"grad_norm": 3.0243561267852783,
|
| 1940 |
+
"learning_rate": 1.534306742640076e-05,
|
| 1941 |
+
"loss": 1.1268,
|
| 1942 |
+
"step": 6775
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.1794871794871793,
|
| 1946 |
+
"grad_norm": 2.9311728477478027,
|
| 1947 |
+
"learning_rate": 1.5194681861348528e-05,
|
| 1948 |
+
"loss": 1.1003,
|
| 1949 |
+
"step": 6800
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.1875,
|
| 1953 |
+
"grad_norm": 3.029709815979004,
|
| 1954 |
+
"learning_rate": 1.5046296296296297e-05,
|
| 1955 |
+
"loss": 1.1082,
|
| 1956 |
+
"step": 6825
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.1955128205128207,
|
| 1960 |
+
"grad_norm": 3.1518394947052,
|
| 1961 |
+
"learning_rate": 1.4897910731244066e-05,
|
| 1962 |
+
"loss": 1.1126,
|
| 1963 |
+
"step": 6850
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.203525641025641,
|
| 1967 |
+
"grad_norm": 3.3375720977783203,
|
| 1968 |
+
"learning_rate": 1.4749525166191833e-05,
|
| 1969 |
+
"loss": 1.102,
|
| 1970 |
+
"step": 6875
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.2115384615384617,
|
| 1974 |
+
"grad_norm": 13.11658000946045,
|
| 1975 |
+
"learning_rate": 1.4601139601139601e-05,
|
| 1976 |
+
"loss": 1.0884,
|
| 1977 |
+
"step": 6900
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.219551282051282,
|
| 1981 |
+
"grad_norm": 3.275269031524658,
|
| 1982 |
+
"learning_rate": 1.445275403608737e-05,
|
| 1983 |
+
"loss": 1.1101,
|
| 1984 |
+
"step": 6925
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.2275641025641026,
|
| 1988 |
+
"grad_norm": 3.0255274772644043,
|
| 1989 |
+
"learning_rate": 1.4304368471035138e-05,
|
| 1990 |
+
"loss": 1.1144,
|
| 1991 |
+
"step": 6950
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.235576923076923,
|
| 1995 |
+
"grad_norm": 3.155569314956665,
|
| 1996 |
+
"learning_rate": 1.4155982905982907e-05,
|
| 1997 |
+
"loss": 1.1013,
|
| 1998 |
+
"step": 6975
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.2435897435897436,
|
| 2002 |
+
"grad_norm": 3.1684317588806152,
|
| 2003 |
+
"learning_rate": 1.4007597340930676e-05,
|
| 2004 |
+
"loss": 1.112,
|
| 2005 |
+
"step": 7000
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.251602564102564,
|
| 2009 |
+
"grad_norm": 3.133126735687256,
|
| 2010 |
+
"learning_rate": 1.3859211775878444e-05,
|
| 2011 |
+
"loss": 1.0949,
|
| 2012 |
+
"step": 7025
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.2596153846153846,
|
| 2016 |
+
"grad_norm": 3.0783729553222656,
|
| 2017 |
+
"learning_rate": 1.3710826210826211e-05,
|
| 2018 |
+
"loss": 1.1002,
|
| 2019 |
+
"step": 7050
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.2676282051282053,
|
| 2023 |
+
"grad_norm": 3.0904178619384766,
|
| 2024 |
+
"learning_rate": 1.356244064577398e-05,
|
| 2025 |
+
"loss": 1.0911,
|
| 2026 |
+
"step": 7075
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.2756410256410255,
|
| 2030 |
+
"grad_norm": 3.140065908432007,
|
| 2031 |
+
"learning_rate": 1.341405508072175e-05,
|
| 2032 |
+
"loss": 1.1415,
|
| 2033 |
+
"step": 7100
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.2836538461538463,
|
| 2037 |
+
"grad_norm": 3.085331678390503,
|
| 2038 |
+
"learning_rate": 1.3265669515669516e-05,
|
| 2039 |
+
"loss": 1.1048,
|
| 2040 |
+
"step": 7125
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.2916666666666665,
|
| 2044 |
+
"grad_norm": 12.111364364624023,
|
| 2045 |
+
"learning_rate": 1.3117283950617285e-05,
|
| 2046 |
+
"loss": 1.0952,
|
| 2047 |
+
"step": 7150
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.2996794871794872,
|
| 2051 |
+
"grad_norm": 3.091158866882324,
|
| 2052 |
+
"learning_rate": 1.2968898385565054e-05,
|
| 2053 |
+
"loss": 1.0917,
|
| 2054 |
+
"step": 7175
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.3076923076923075,
|
| 2058 |
+
"grad_norm": 2.981398344039917,
|
| 2059 |
+
"learning_rate": 1.282051282051282e-05,
|
| 2060 |
+
"loss": 1.1104,
|
| 2061 |
+
"step": 7200
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.315705128205128,
|
| 2065 |
+
"grad_norm": 3.0888004302978516,
|
| 2066 |
+
"learning_rate": 1.2672127255460589e-05,
|
| 2067 |
+
"loss": 1.0745,
|
| 2068 |
+
"step": 7225
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.323717948717949,
|
| 2072 |
+
"grad_norm": 3.2106196880340576,
|
| 2073 |
+
"learning_rate": 1.2523741690408358e-05,
|
| 2074 |
+
"loss": 1.1027,
|
| 2075 |
+
"step": 7250
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.331730769230769,
|
| 2079 |
+
"grad_norm": 2.96028995513916,
|
| 2080 |
+
"learning_rate": 1.2375356125356126e-05,
|
| 2081 |
+
"loss": 1.07,
|
| 2082 |
+
"step": 7275
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.33974358974359,
|
| 2086 |
+
"grad_norm": 3.157464027404785,
|
| 2087 |
+
"learning_rate": 1.2226970560303893e-05,
|
| 2088 |
+
"loss": 1.1384,
|
| 2089 |
+
"step": 7300
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.34775641025641,
|
| 2093 |
+
"grad_norm": 3.1314988136291504,
|
| 2094 |
+
"learning_rate": 1.2078584995251663e-05,
|
| 2095 |
+
"loss": 1.0839,
|
| 2096 |
+
"step": 7325
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.355769230769231,
|
| 2100 |
+
"grad_norm": 3.3742153644561768,
|
| 2101 |
+
"learning_rate": 1.193019943019943e-05,
|
| 2102 |
+
"loss": 1.1234,
|
| 2103 |
+
"step": 7350
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.363782051282051,
|
| 2107 |
+
"grad_norm": 2.8460400104522705,
|
| 2108 |
+
"learning_rate": 1.17818138651472e-05,
|
| 2109 |
+
"loss": 1.1229,
|
| 2110 |
+
"step": 7375
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.371794871794872,
|
| 2114 |
+
"grad_norm": 2.7427093982696533,
|
| 2115 |
+
"learning_rate": 1.1633428300094967e-05,
|
| 2116 |
+
"loss": 1.1444,
|
| 2117 |
+
"step": 7400
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.3798076923076925,
|
| 2121 |
+
"grad_norm": 2.87485933303833,
|
| 2122 |
+
"learning_rate": 1.1485042735042734e-05,
|
| 2123 |
+
"loss": 1.0885,
|
| 2124 |
+
"step": 7425
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.3878205128205128,
|
| 2128 |
+
"grad_norm": 2.758561611175537,
|
| 2129 |
+
"learning_rate": 1.1336657169990504e-05,
|
| 2130 |
+
"loss": 1.0802,
|
| 2131 |
+
"step": 7450
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.3958333333333335,
|
| 2135 |
+
"grad_norm": 2.8536176681518555,
|
| 2136 |
+
"learning_rate": 1.1188271604938271e-05,
|
| 2137 |
+
"loss": 1.1222,
|
| 2138 |
+
"step": 7475
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.4038461538461537,
|
| 2142 |
+
"grad_norm": 3.1613593101501465,
|
| 2143 |
+
"learning_rate": 1.103988603988604e-05,
|
| 2144 |
+
"loss": 1.1843,
|
| 2145 |
+
"step": 7500
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.4118589743589745,
|
| 2149 |
+
"grad_norm": 3.1586380004882812,
|
| 2150 |
+
"learning_rate": 1.0891500474833808e-05,
|
| 2151 |
+
"loss": 1.0965,
|
| 2152 |
+
"step": 7525
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.4198717948717947,
|
| 2156 |
+
"grad_norm": 3.20630145072937,
|
| 2157 |
+
"learning_rate": 1.0743114909781577e-05,
|
| 2158 |
+
"loss": 1.0952,
|
| 2159 |
+
"step": 7550
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.4278846153846154,
|
| 2163 |
+
"grad_norm": 14.078017234802246,
|
| 2164 |
+
"learning_rate": 1.0594729344729346e-05,
|
| 2165 |
+
"loss": 1.0659,
|
| 2166 |
+
"step": 7575
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.435897435897436,
|
| 2170 |
+
"grad_norm": 3.086122989654541,
|
| 2171 |
+
"learning_rate": 1.0446343779677114e-05,
|
| 2172 |
+
"loss": 1.1165,
|
| 2173 |
+
"step": 7600
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.4439102564102564,
|
| 2177 |
+
"grad_norm": 2.8791723251342773,
|
| 2178 |
+
"learning_rate": 1.0297958214624883e-05,
|
| 2179 |
+
"loss": 1.1071,
|
| 2180 |
+
"step": 7625
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.451923076923077,
|
| 2184 |
+
"grad_norm": 3.199122428894043,
|
| 2185 |
+
"learning_rate": 1.014957264957265e-05,
|
| 2186 |
+
"loss": 1.0854,
|
| 2187 |
+
"step": 7650
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.4599358974358974,
|
| 2191 |
+
"grad_norm": 2.861563205718994,
|
| 2192 |
+
"learning_rate": 1.0001187084520418e-05,
|
| 2193 |
+
"loss": 1.1338,
|
| 2194 |
+
"step": 7675
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.467948717948718,
|
| 2198 |
+
"grad_norm": 3.0902774333953857,
|
| 2199 |
+
"learning_rate": 9.852801519468187e-06,
|
| 2200 |
+
"loss": 1.1516,
|
| 2201 |
+
"step": 7700
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.4759615384615383,
|
| 2205 |
+
"grad_norm": 3.2896981239318848,
|
| 2206 |
+
"learning_rate": 9.704415954415955e-06,
|
| 2207 |
+
"loss": 1.0765,
|
| 2208 |
+
"step": 7725
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.483974358974359,
|
| 2212 |
+
"grad_norm": 2.9086813926696777,
|
| 2213 |
+
"learning_rate": 9.556030389363724e-06,
|
| 2214 |
+
"loss": 1.0936,
|
| 2215 |
+
"step": 7750
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.4919871794871793,
|
| 2219 |
+
"grad_norm": 3.230581045150757,
|
| 2220 |
+
"learning_rate": 9.407644824311492e-06,
|
| 2221 |
+
"loss": 1.1308,
|
| 2222 |
+
"step": 7775
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.5,
|
| 2226 |
+
"grad_norm": 3.3030433654785156,
|
| 2227 |
+
"learning_rate": 9.259259259259259e-06,
|
| 2228 |
+
"loss": 1.0452,
|
| 2229 |
+
"step": 7800
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.5080128205128203,
|
| 2233 |
+
"grad_norm": 3.3332784175872803,
|
| 2234 |
+
"learning_rate": 9.110873694207028e-06,
|
| 2235 |
+
"loss": 1.1698,
|
| 2236 |
+
"step": 7825
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.516025641025641,
|
| 2240 |
+
"grad_norm": 3.0965607166290283,
|
| 2241 |
+
"learning_rate": 8.962488129154796e-06,
|
| 2242 |
+
"loss": 1.0921,
|
| 2243 |
+
"step": 7850
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.5240384615384617,
|
| 2247 |
+
"grad_norm": 2.8891959190368652,
|
| 2248 |
+
"learning_rate": 8.814102564102565e-06,
|
| 2249 |
+
"loss": 1.0856,
|
| 2250 |
+
"step": 7875
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.532051282051282,
|
| 2254 |
+
"grad_norm": 3.045534372329712,
|
| 2255 |
+
"learning_rate": 8.665716999050333e-06,
|
| 2256 |
+
"loss": 1.086,
|
| 2257 |
+
"step": 7900
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.5400641025641026,
|
| 2261 |
+
"grad_norm": 15.829360961914062,
|
| 2262 |
+
"learning_rate": 8.5173314339981e-06,
|
| 2263 |
+
"loss": 1.1969,
|
| 2264 |
+
"step": 7925
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.5480769230769234,
|
| 2268 |
+
"grad_norm": 3.0161049365997314,
|
| 2269 |
+
"learning_rate": 8.36894586894587e-06,
|
| 2270 |
+
"loss": 1.1298,
|
| 2271 |
+
"step": 7950
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.5560897435897436,
|
| 2275 |
+
"grad_norm": 3.150747299194336,
|
| 2276 |
+
"learning_rate": 8.220560303893637e-06,
|
| 2277 |
+
"loss": 1.0871,
|
| 2278 |
+
"step": 7975
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.564102564102564,
|
| 2282 |
+
"grad_norm": 3.1695539951324463,
|
| 2283 |
+
"learning_rate": 8.072174738841406e-06,
|
| 2284 |
+
"loss": 1.0944,
|
| 2285 |
+
"step": 8000
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.5721153846153846,
|
| 2289 |
+
"grad_norm": 3.0507800579071045,
|
| 2290 |
+
"learning_rate": 7.923789173789174e-06,
|
| 2291 |
+
"loss": 1.0789,
|
| 2292 |
+
"step": 8025
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.5801282051282053,
|
| 2296 |
+
"grad_norm": 3.0250277519226074,
|
| 2297 |
+
"learning_rate": 7.775403608736941e-06,
|
| 2298 |
+
"loss": 1.1182,
|
| 2299 |
+
"step": 8050
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.5881410256410255,
|
| 2303 |
+
"grad_norm": 3.100142478942871,
|
| 2304 |
+
"learning_rate": 7.627018043684711e-06,
|
| 2305 |
+
"loss": 1.0961,
|
| 2306 |
+
"step": 8075
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.5961538461538463,
|
| 2310 |
+
"grad_norm": 2.901756763458252,
|
| 2311 |
+
"learning_rate": 7.478632478632479e-06,
|
| 2312 |
+
"loss": 1.1236,
|
| 2313 |
+
"step": 8100
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.6041666666666665,
|
| 2317 |
+
"grad_norm": 3.03731369972229,
|
| 2318 |
+
"learning_rate": 7.330246913580248e-06,
|
| 2319 |
+
"loss": 1.0897,
|
| 2320 |
+
"step": 8125
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.6121794871794872,
|
| 2324 |
+
"grad_norm": 3.0912184715270996,
|
| 2325 |
+
"learning_rate": 7.1818613485280155e-06,
|
| 2326 |
+
"loss": 1.1125,
|
| 2327 |
+
"step": 8150
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.6201923076923075,
|
| 2331 |
+
"grad_norm": 13.127511978149414,
|
| 2332 |
+
"learning_rate": 7.033475783475783e-06,
|
| 2333 |
+
"loss": 1.1167,
|
| 2334 |
+
"step": 8175
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.628205128205128,
|
| 2338 |
+
"grad_norm": 3.0535366535186768,
|
| 2339 |
+
"learning_rate": 6.885090218423552e-06,
|
| 2340 |
+
"loss": 1.1025,
|
| 2341 |
+
"step": 8200
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.636217948717949,
|
| 2345 |
+
"grad_norm": 3.029935359954834,
|
| 2346 |
+
"learning_rate": 6.73670465337132e-06,
|
| 2347 |
+
"loss": 1.1014,
|
| 2348 |
+
"step": 8225
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.644230769230769,
|
| 2352 |
+
"grad_norm": 14.514870643615723,
|
| 2353 |
+
"learning_rate": 6.588319088319089e-06,
|
| 2354 |
+
"loss": 1.1145,
|
| 2355 |
+
"step": 8250
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.65224358974359,
|
| 2359 |
+
"grad_norm": 2.957597494125366,
|
| 2360 |
+
"learning_rate": 6.439933523266857e-06,
|
| 2361 |
+
"loss": 1.0878,
|
| 2362 |
+
"step": 8275
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.66025641025641,
|
| 2366 |
+
"grad_norm": 3.116901159286499,
|
| 2367 |
+
"learning_rate": 6.291547958214625e-06,
|
| 2368 |
+
"loss": 1.1016,
|
| 2369 |
+
"step": 8300
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.668269230769231,
|
| 2373 |
+
"grad_norm": 3.118030309677124,
|
| 2374 |
+
"learning_rate": 6.143162393162393e-06,
|
| 2375 |
+
"loss": 1.0814,
|
| 2376 |
+
"step": 8325
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.676282051282051,
|
| 2380 |
+
"grad_norm": 3.0143322944641113,
|
| 2381 |
+
"learning_rate": 5.994776828110162e-06,
|
| 2382 |
+
"loss": 1.0755,
|
| 2383 |
+
"step": 8350
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.684294871794872,
|
| 2387 |
+
"grad_norm": 3.08807373046875,
|
| 2388 |
+
"learning_rate": 5.84639126305793e-06,
|
| 2389 |
+
"loss": 1.106,
|
| 2390 |
+
"step": 8375
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.6923076923076925,
|
| 2394 |
+
"grad_norm": 2.806126594543457,
|
| 2395 |
+
"learning_rate": 5.6980056980056985e-06,
|
| 2396 |
+
"loss": 1.0685,
|
| 2397 |
+
"step": 8400
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.7003205128205128,
|
| 2401 |
+
"grad_norm": 3.1736948490142822,
|
| 2402 |
+
"learning_rate": 5.549620132953467e-06,
|
| 2403 |
+
"loss": 1.0843,
|
| 2404 |
+
"step": 8425
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.7083333333333335,
|
| 2408 |
+
"grad_norm": 2.8049614429473877,
|
| 2409 |
+
"learning_rate": 5.401234567901234e-06,
|
| 2410 |
+
"loss": 1.0964,
|
| 2411 |
+
"step": 8450
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.7163461538461537,
|
| 2415 |
+
"grad_norm": 3.1841471195220947,
|
| 2416 |
+
"learning_rate": 5.252849002849003e-06,
|
| 2417 |
+
"loss": 1.1408,
|
| 2418 |
+
"step": 8475
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 2.7243589743589745,
|
| 2422 |
+
"grad_norm": 12.97529125213623,
|
| 2423 |
+
"learning_rate": 5.104463437796771e-06,
|
| 2424 |
+
"loss": 1.1415,
|
| 2425 |
+
"step": 8500
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 2.7323717948717947,
|
| 2429 |
+
"grad_norm": 2.944291830062866,
|
| 2430 |
+
"learning_rate": 4.9560778727445395e-06,
|
| 2431 |
+
"loss": 1.1071,
|
| 2432 |
+
"step": 8525
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 2.7403846153846154,
|
| 2436 |
+
"grad_norm": 2.8969268798828125,
|
| 2437 |
+
"learning_rate": 4.807692307692308e-06,
|
| 2438 |
+
"loss": 1.1047,
|
| 2439 |
+
"step": 8550
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.748397435897436,
|
| 2443 |
+
"grad_norm": 3.237588882446289,
|
| 2444 |
+
"learning_rate": 4.659306742640076e-06,
|
| 2445 |
+
"loss": 1.1015,
|
| 2446 |
+
"step": 8575
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.7564102564102564,
|
| 2450 |
+
"grad_norm": 2.9864237308502197,
|
| 2451 |
+
"learning_rate": 4.510921177587845e-06,
|
| 2452 |
+
"loss": 1.1155,
|
| 2453 |
+
"step": 8600
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.7644230769230766,
|
| 2457 |
+
"grad_norm": 3.131612777709961,
|
| 2458 |
+
"learning_rate": 4.362535612535613e-06,
|
| 2459 |
+
"loss": 1.1221,
|
| 2460 |
+
"step": 8625
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 2.7724358974358974,
|
| 2464 |
+
"grad_norm": 3.0235490798950195,
|
| 2465 |
+
"learning_rate": 4.214150047483381e-06,
|
| 2466 |
+
"loss": 1.0997,
|
| 2467 |
+
"step": 8650
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 2.780448717948718,
|
| 2471 |
+
"grad_norm": 3.132972240447998,
|
| 2472 |
+
"learning_rate": 4.06576448243115e-06,
|
| 2473 |
+
"loss": 1.1041,
|
| 2474 |
+
"step": 8675
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 2.7884615384615383,
|
| 2478 |
+
"grad_norm": 3.0675647258758545,
|
| 2479 |
+
"learning_rate": 3.917378917378917e-06,
|
| 2480 |
+
"loss": 1.1166,
|
| 2481 |
+
"step": 8700
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.796474358974359,
|
| 2485 |
+
"grad_norm": 14.459759712219238,
|
| 2486 |
+
"learning_rate": 3.7689933523266856e-06,
|
| 2487 |
+
"loss": 1.1164,
|
| 2488 |
+
"step": 8725
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 2.8044871794871797,
|
| 2492 |
+
"grad_norm": 13.613649368286133,
|
| 2493 |
+
"learning_rate": 3.620607787274454e-06,
|
| 2494 |
+
"loss": 1.0888,
|
| 2495 |
+
"step": 8750
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 2.8125,
|
| 2499 |
+
"grad_norm": 3.0776546001434326,
|
| 2500 |
+
"learning_rate": 3.4722222222222224e-06,
|
| 2501 |
+
"loss": 1.094,
|
| 2502 |
+
"step": 8775
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.8205128205128203,
|
| 2506 |
+
"grad_norm": 3.1849546432495117,
|
| 2507 |
+
"learning_rate": 3.3238366571699908e-06,
|
| 2508 |
+
"loss": 1.0929,
|
| 2509 |
+
"step": 8800
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 2.828525641025641,
|
| 2513 |
+
"grad_norm": 2.8894577026367188,
|
| 2514 |
+
"learning_rate": 3.1754510921177587e-06,
|
| 2515 |
+
"loss": 1.0961,
|
| 2516 |
+
"step": 8825
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.8365384615384617,
|
| 2520 |
+
"grad_norm": 3.1093499660491943,
|
| 2521 |
+
"learning_rate": 3.027065527065527e-06,
|
| 2522 |
+
"loss": 1.0836,
|
| 2523 |
+
"step": 8850
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.844551282051282,
|
| 2527 |
+
"grad_norm": 3.0798821449279785,
|
| 2528 |
+
"learning_rate": 2.8786799620132955e-06,
|
| 2529 |
+
"loss": 1.1081,
|
| 2530 |
+
"step": 8875
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 2.8525641025641026,
|
| 2534 |
+
"grad_norm": 10.790550231933594,
|
| 2535 |
+
"learning_rate": 2.730294396961064e-06,
|
| 2536 |
+
"loss": 1.0906,
|
| 2537 |
+
"step": 8900
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 2.8605769230769234,
|
| 2541 |
+
"grad_norm": 3.195128917694092,
|
| 2542 |
+
"learning_rate": 2.581908831908832e-06,
|
| 2543 |
+
"loss": 1.1091,
|
| 2544 |
+
"step": 8925
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 2.8685897435897436,
|
| 2548 |
+
"grad_norm": 2.8264927864074707,
|
| 2549 |
+
"learning_rate": 2.4335232668566006e-06,
|
| 2550 |
+
"loss": 1.1043,
|
| 2551 |
+
"step": 8950
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 2.876602564102564,
|
| 2555 |
+
"grad_norm": 3.0284230709075928,
|
| 2556 |
+
"learning_rate": 2.2851377018043685e-06,
|
| 2557 |
+
"loss": 1.1308,
|
| 2558 |
+
"step": 8975
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 2.8846153846153846,
|
| 2562 |
+
"grad_norm": 2.7990450859069824,
|
| 2563 |
+
"learning_rate": 2.136752136752137e-06,
|
| 2564 |
+
"loss": 1.1158,
|
| 2565 |
+
"step": 9000
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 2.8926282051282053,
|
| 2569 |
+
"grad_norm": 2.9970715045928955,
|
| 2570 |
+
"learning_rate": 1.988366571699905e-06,
|
| 2571 |
+
"loss": 1.111,
|
| 2572 |
+
"step": 9025
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 2.9006410256410255,
|
| 2576 |
+
"grad_norm": 3.1771433353424072,
|
| 2577 |
+
"learning_rate": 1.8399810066476735e-06,
|
| 2578 |
+
"loss": 1.0942,
|
| 2579 |
+
"step": 9050
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 2.9086538461538463,
|
| 2583 |
+
"grad_norm": 12.921822547912598,
|
| 2584 |
+
"learning_rate": 1.6915954415954418e-06,
|
| 2585 |
+
"loss": 1.0977,
|
| 2586 |
+
"step": 9075
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 2.9166666666666665,
|
| 2590 |
+
"grad_norm": 2.956425428390503,
|
| 2591 |
+
"learning_rate": 1.5432098765432098e-06,
|
| 2592 |
+
"loss": 1.0975,
|
| 2593 |
+
"step": 9100
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 2.9246794871794872,
|
| 2597 |
+
"grad_norm": 3.100966691970825,
|
| 2598 |
+
"learning_rate": 1.3948243114909784e-06,
|
| 2599 |
+
"loss": 1.103,
|
| 2600 |
+
"step": 9125
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 2.9326923076923075,
|
| 2604 |
+
"grad_norm": 3.1469533443450928,
|
| 2605 |
+
"learning_rate": 1.2464387464387465e-06,
|
| 2606 |
+
"loss": 1.0842,
|
| 2607 |
+
"step": 9150
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.940705128205128,
|
| 2611 |
+
"grad_norm": 2.9754679203033447,
|
| 2612 |
+
"learning_rate": 1.0980531813865147e-06,
|
| 2613 |
+
"loss": 1.1,
|
| 2614 |
+
"step": 9175
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 2.948717948717949,
|
| 2618 |
+
"grad_norm": 3.2067482471466064,
|
| 2619 |
+
"learning_rate": 9.49667616334283e-07,
|
| 2620 |
+
"loss": 1.1009,
|
| 2621 |
+
"step": 9200
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.956730769230769,
|
| 2625 |
+
"grad_norm": 3.133969783782959,
|
| 2626 |
+
"learning_rate": 8.012820512820512e-07,
|
| 2627 |
+
"loss": 1.1021,
|
| 2628 |
+
"step": 9225
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 2.96474358974359,
|
| 2632 |
+
"grad_norm": 12.806807518005371,
|
| 2633 |
+
"learning_rate": 6.528964862298196e-07,
|
| 2634 |
+
"loss": 1.114,
|
| 2635 |
+
"step": 9250
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 2.97275641025641,
|
| 2639 |
+
"grad_norm": 2.973973274230957,
|
| 2640 |
+
"learning_rate": 5.045109211775879e-07,
|
| 2641 |
+
"loss": 1.0975,
|
| 2642 |
+
"step": 9275
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 2.980769230769231,
|
| 2646 |
+
"grad_norm": 3.126347303390503,
|
| 2647 |
+
"learning_rate": 3.5612535612535615e-07,
|
| 2648 |
+
"loss": 1.088,
|
| 2649 |
+
"step": 9300
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 2.988782051282051,
|
| 2653 |
+
"grad_norm": 13.450477600097656,
|
| 2654 |
+
"learning_rate": 2.0773979107312442e-07,
|
| 2655 |
+
"loss": 1.0843,
|
| 2656 |
+
"step": 9325
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 2.996794871794872,
|
| 2660 |
+
"grad_norm": 12.06689167022705,
|
| 2661 |
+
"learning_rate": 5.9354226020892686e-08,
|
| 2662 |
+
"loss": 1.0921,
|
| 2663 |
+
"step": 9350
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 3.0,
|
| 2667 |
+
"eval_accuracy": 0.3333333333333333,
|
| 2668 |
+
"eval_f1_macro": 0.16666666666666666,
|
| 2669 |
+
"eval_f1_micro": 0.3333333333333333,
|
| 2670 |
+
"eval_f1_weighted": 0.16666666666666666,
|
| 2671 |
+
"eval_loss": 1.0986850261688232,
|
| 2672 |
+
"eval_precision_macro": 0.1111111111111111,
|
| 2673 |
+
"eval_precision_micro": 0.3333333333333333,
|
| 2674 |
+
"eval_precision_weighted": 0.1111111111111111,
|
| 2675 |
+
"eval_recall_macro": 0.3333333333333333,
|
| 2676 |
+
"eval_recall_micro": 0.3333333333333333,
|
| 2677 |
+
"eval_recall_weighted": 0.3333333333333333,
|
| 2678 |
+
"eval_runtime": 10152.0725,
|
| 2679 |
+
"eval_samples_per_second": 0.077,
|
| 2680 |
+
"eval_steps_per_second": 0.038,
|
| 2681 |
+
"step": 9360
|
| 2682 |
+
}
|
| 2683 |
+
],
|
| 2684 |
+
"logging_steps": 25,
|
| 2685 |
+
"max_steps": 9360,
|
| 2686 |
+
"num_input_tokens_seen": 0,
|
| 2687 |
+
"num_train_epochs": 3,
|
| 2688 |
+
"save_steps": 500,
|
| 2689 |
+
"stateful_callbacks": {
|
| 2690 |
+
"EarlyStoppingCallback": {
|
| 2691 |
+
"args": {
|
| 2692 |
+
"early_stopping_patience": 5,
|
| 2693 |
+
"early_stopping_threshold": 0.01
|
| 2694 |
+
},
|
| 2695 |
+
"attributes": {
|
| 2696 |
+
"early_stopping_patience_counter": 2
|
| 2697 |
+
}
|
| 2698 |
+
},
|
| 2699 |
+
"TrainerControl": {
|
| 2700 |
+
"args": {
|
| 2701 |
+
"should_epoch_stop": false,
|
| 2702 |
+
"should_evaluate": false,
|
| 2703 |
+
"should_log": false,
|
| 2704 |
+
"should_save": true,
|
| 2705 |
+
"should_training_stop": true
|
| 2706 |
+
},
|
| 2707 |
+
"attributes": {}
|
| 2708 |
+
}
|
| 2709 |
+
},
|
| 2710 |
+
"total_flos": 2.459270123421696e+16,
|
| 2711 |
+
"train_batch_size": 1,
|
| 2712 |
+
"trial_name": null,
|
| 2713 |
+
"trial_params": null
|
| 2714 |
+
}
|
checkpoint-9360/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46cc5a8fc1ee3a80d1cf3a89e0365c8b31e18b3e81de6dc82c93116102a97d54
|
| 3 |
+
size 5304
|
config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "allenai/longformer-base-4096",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"LongformerForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_mode": "longformer",
|
| 8 |
+
"attention_probs_dropout_prob": 0.1,
|
| 9 |
+
"attention_window": [
|
| 10 |
+
512,
|
| 11 |
+
512,
|
| 12 |
+
512,
|
| 13 |
+
512,
|
| 14 |
+
512,
|
| 15 |
+
512,
|
| 16 |
+
512,
|
| 17 |
+
512,
|
| 18 |
+
512,
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512
|
| 22 |
+
],
|
| 23 |
+
"bos_token_id": 0,
|
| 24 |
+
"eos_token_id": 2,
|
| 25 |
+
"gradient_checkpointing": false,
|
| 26 |
+
"hidden_act": "gelu",
|
| 27 |
+
"hidden_dropout_prob": 0.1,
|
| 28 |
+
"hidden_size": 768,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "negative",
|
| 31 |
+
"1": "neutral",
|
| 32 |
+
"2": "positive"
|
| 33 |
+
},
|
| 34 |
+
"ignore_attention_mask": false,
|
| 35 |
+
"initializer_range": 0.02,
|
| 36 |
+
"intermediate_size": 3072,
|
| 37 |
+
"label2id": {
|
| 38 |
+
"negative": 0,
|
| 39 |
+
"neutral": 1,
|
| 40 |
+
"positive": 2
|
| 41 |
+
},
|
| 42 |
+
"layer_norm_eps": 1e-05,
|
| 43 |
+
"max_position_embeddings": 4098,
|
| 44 |
+
"model_type": "longformer",
|
| 45 |
+
"num_attention_heads": 12,
|
| 46 |
+
"num_hidden_layers": 12,
|
| 47 |
+
"onnx_export": false,
|
| 48 |
+
"pad_token_id": 1,
|
| 49 |
+
"problem_type": "single_label_classification",
|
| 50 |
+
"sep_token_id": 2,
|
| 51 |
+
"torch_dtype": "float32",
|
| 52 |
+
"transformers_version": "4.46.2",
|
| 53 |
+
"type_vocab_size": 1,
|
| 54 |
+
"vocab_size": 50265
|
| 55 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc18e806f34592b583e45db50ba2aa5982e9c1b7f599857ceae4def227334747
|
| 3 |
+
size 594681260
|
runs/Nov11_10-29-57_r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz/events.out.tfevents.1731321001.r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz.146.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2337940c409642b85edeb0a3b880bdcac8e88e6f05355a6afc144d9472dac638
|
| 3 |
+
size 87170
|
runs/Nov11_10-29-57_r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz/events.out.tfevents.1731964149.r-kafikani-autotrain-advanced-2-iz1ofkp6-94fef-9b4wz.146.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0147653a400283f40b223ecdc92bb1e2c7a6947a9916cb94748b5ab8dd0fe2c8
|
| 3 |
+
size 921
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<pad>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"50264": {
|
| 37 |
+
"content": "<mask>",
|
| 38 |
+
"lstrip": true,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"bos_token": "<s>",
|
| 46 |
+
"clean_up_tokenization_spaces": false,
|
| 47 |
+
"cls_token": "<s>",
|
| 48 |
+
"eos_token": "</s>",
|
| 49 |
+
"errors": "replace",
|
| 50 |
+
"mask_token": "<mask>",
|
| 51 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 52 |
+
"pad_token": "<pad>",
|
| 53 |
+
"sep_token": "</s>",
|
| 54 |
+
"tokenizer_class": "LongformerTokenizer",
|
| 55 |
+
"trim_offsets": true,
|
| 56 |
+
"unk_token": "<unk>"
|
| 57 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46cc5a8fc1ee3a80d1cf3a89e0365c8b31e18b3e81de6dc82c93116102a97d54
|
| 3 |
+
size 5304
|
training_params.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"data_path": "longfomer5balanced/autotrain-data",
|
| 3 |
+
"model": "allenai/longformer-base-4096",
|
| 4 |
+
"lr": 5e-05,
|
| 5 |
+
"epochs": 3,
|
| 6 |
+
"max_seq_length": 4096,
|
| 7 |
+
"batch_size": 1,
|
| 8 |
+
"warmup_ratio": 0.1,
|
| 9 |
+
"gradient_accumulation": 1,
|
| 10 |
+
"optimizer": "adamw_torch",
|
| 11 |
+
"scheduler": "linear",
|
| 12 |
+
"weight_decay": 0.0,
|
| 13 |
+
"max_grad_norm": 1.0,
|
| 14 |
+
"seed": 42,
|
| 15 |
+
"train_split": "train",
|
| 16 |
+
"valid_split": "validation",
|
| 17 |
+
"text_column": "autotrain_text",
|
| 18 |
+
"target_column": "autotrain_label",
|
| 19 |
+
"logging_steps": -1,
|
| 20 |
+
"project_name": "longfomer5balanced",
|
| 21 |
+
"auto_find_batch_size": false,
|
| 22 |
+
"mixed_precision": "fp16",
|
| 23 |
+
"save_total_limit": 1,
|
| 24 |
+
"push_to_hub": true,
|
| 25 |
+
"eval_strategy": "epoch",
|
| 26 |
+
"username": "kafikani",
|
| 27 |
+
"log": "tensorboard",
|
| 28 |
+
"early_stopping_patience": 5,
|
| 29 |
+
"early_stopping_threshold": 0.01
|
| 30 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|