Upload 41 files
Browse files- config.json +26 -0
- merges.txt +0 -0
- rng_state_0.pth +3 -0
- rng_state_1.pth +3 -0
- rng_state_10.pth +3 -0
- rng_state_11.pth +3 -0
- rng_state_12.pth +3 -0
- rng_state_13.pth +3 -0
- rng_state_14.pth +3 -0
- rng_state_15.pth +3 -0
- rng_state_16.pth +3 -0
- rng_state_17.pth +3 -0
- rng_state_18.pth +3 -0
- rng_state_19.pth +3 -0
- rng_state_2.pth +3 -0
- rng_state_20.pth +3 -0
- rng_state_21.pth +3 -0
- rng_state_22.pth +3 -0
- rng_state_23.pth +3 -0
- rng_state_24.pth +3 -0
- rng_state_25.pth +3 -0
- rng_state_26.pth +3 -0
- rng_state_27.pth +3 -0
- rng_state_28.pth +3 -0
- rng_state_29.pth +3 -0
- rng_state_3.pth +3 -0
- rng_state_30.pth +3 -0
- rng_state_31.pth +3 -0
- rng_state_4.pth +3 -0
- rng_state_5.pth +3 -0
- rng_state_6.pth +3 -0
- rng_state_7.pth +3 -0
- rng_state_8.pth +3 -0
- rng_state_9.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +51 -0
- tokenizer_config.json +66 -0
- trainer_state.json +1180 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"CamembertForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 514,
|
| 16 |
+
"model_type": "camembert",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 12,
|
| 19 |
+
"pad_token_id": 1,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.22.2",
|
| 23 |
+
"type_vocab_size": 1,
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 32000
|
| 26 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_10.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_11.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_12.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_13.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_14.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_15.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_16.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_17.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_18.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_19.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_20.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_21.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_22.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_23.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_24.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_25.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_26.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_27.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_28.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_29.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_30.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_31.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_8.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
rng_state_9.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14f1f81d5a0bb373e8ea473a062d6a4754d59b064d9961e81914c485138ca706
|
| 3 |
+
size 14503
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb815792218f10143365bc27840410b3471cd0fbf9ca581f7dd572fbce81a1bc
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:498cee63371d1ed48e2413cb778fc0507bf3a468e5010a210e5066bcf13c6af7
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": true,
|
| 26 |
+
"normalized": true,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": true,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": true,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": {
|
| 4 |
+
"__type": "AddedToken",
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false
|
| 10 |
+
},
|
| 11 |
+
"clean_up_tokenization_spaces": true,
|
| 12 |
+
"cls_token": {
|
| 13 |
+
"__type": "AddedToken",
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": true,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false
|
| 19 |
+
},
|
| 20 |
+
"eos_token": {
|
| 21 |
+
"__type": "AddedToken",
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": true,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false
|
| 27 |
+
},
|
| 28 |
+
"errors": "replace",
|
| 29 |
+
"mask_token": {
|
| 30 |
+
"__type": "AddedToken",
|
| 31 |
+
"content": "<mask>",
|
| 32 |
+
"lstrip": true,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 38 |
+
"name_or_path": "./tokenizers/BpeTokenizer_NACHOS_10M_lowercased_fixed_utf8-roberta/",
|
| 39 |
+
"pad_token": {
|
| 40 |
+
"__type": "AddedToken",
|
| 41 |
+
"content": "<pad>",
|
| 42 |
+
"lstrip": false,
|
| 43 |
+
"normalized": true,
|
| 44 |
+
"rstrip": false,
|
| 45 |
+
"single_word": false
|
| 46 |
+
},
|
| 47 |
+
"sep_token": {
|
| 48 |
+
"__type": "AddedToken",
|
| 49 |
+
"content": "</s>",
|
| 50 |
+
"lstrip": false,
|
| 51 |
+
"normalized": true,
|
| 52 |
+
"rstrip": false,
|
| 53 |
+
"single_word": false
|
| 54 |
+
},
|
| 55 |
+
"special_tokens_map_file": "./tokenizers/BpeTokenizer_NACHOS_10M_lowercased_fixed_utf8-roberta/special_tokens_map.json",
|
| 56 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 57 |
+
"unk_token": {
|
| 58 |
+
"__type": "AddedToken",
|
| 59 |
+
"content": "<unk>",
|
| 60 |
+
"lstrip": false,
|
| 61 |
+
"normalized": true,
|
| 62 |
+
"rstrip": false,
|
| 63 |
+
"single_word": false
|
| 64 |
+
},
|
| 65 |
+
"use_fast": true
|
| 66 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1180 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 28.0,
|
| 5 |
+
"global_step": 96768,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 5e-09,
|
| 13 |
+
"loss": 10.5075,
|
| 14 |
+
"step": 1
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.14,
|
| 18 |
+
"learning_rate": 2.5e-06,
|
| 19 |
+
"loss": 9.5012,
|
| 20 |
+
"step": 500
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.29,
|
| 24 |
+
"learning_rate": 5e-06,
|
| 25 |
+
"loss": 7.9626,
|
| 26 |
+
"step": 1000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.43,
|
| 30 |
+
"learning_rate": 7.5e-06,
|
| 31 |
+
"loss": 6.9841,
|
| 32 |
+
"step": 1500
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.58,
|
| 36 |
+
"learning_rate": 1e-05,
|
| 37 |
+
"loss": 6.7031,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.72,
|
| 42 |
+
"learning_rate": 1.25e-05,
|
| 43 |
+
"loss": 6.5413,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.87,
|
| 48 |
+
"learning_rate": 1.5e-05,
|
| 49 |
+
"loss": 6.4291,
|
| 50 |
+
"step": 3000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.01,
|
| 54 |
+
"learning_rate": 1.75e-05,
|
| 55 |
+
"loss": 6.334,
|
| 56 |
+
"step": 3500
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.16,
|
| 60 |
+
"learning_rate": 2e-05,
|
| 61 |
+
"loss": 6.2554,
|
| 62 |
+
"step": 4000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.3,
|
| 66 |
+
"learning_rate": 2.25e-05,
|
| 67 |
+
"loss": 6.1905,
|
| 68 |
+
"step": 4500
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.45,
|
| 72 |
+
"learning_rate": 2.5e-05,
|
| 73 |
+
"loss": 6.1341,
|
| 74 |
+
"step": 5000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.59,
|
| 78 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 79 |
+
"loss": 6.0828,
|
| 80 |
+
"step": 5500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.74,
|
| 84 |
+
"learning_rate": 3e-05,
|
| 85 |
+
"loss": 6.0371,
|
| 86 |
+
"step": 6000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.88,
|
| 90 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 91 |
+
"loss": 6.0035,
|
| 92 |
+
"step": 6500
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.03,
|
| 96 |
+
"learning_rate": 3.5e-05,
|
| 97 |
+
"loss": 5.9683,
|
| 98 |
+
"step": 7000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.17,
|
| 102 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 103 |
+
"loss": 5.9306,
|
| 104 |
+
"step": 7500
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.31,
|
| 108 |
+
"learning_rate": 4e-05,
|
| 109 |
+
"loss": 5.9022,
|
| 110 |
+
"step": 8000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.46,
|
| 114 |
+
"learning_rate": 4.2495e-05,
|
| 115 |
+
"loss": 5.8759,
|
| 116 |
+
"step": 8500
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.6,
|
| 120 |
+
"learning_rate": 4.4995000000000005e-05,
|
| 121 |
+
"loss": 5.8494,
|
| 122 |
+
"step": 9000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 2.75,
|
| 126 |
+
"learning_rate": 4.7495e-05,
|
| 127 |
+
"loss": 5.8293,
|
| 128 |
+
"step": 9500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.89,
|
| 132 |
+
"learning_rate": 4.9995000000000005e-05,
|
| 133 |
+
"loss": 5.8098,
|
| 134 |
+
"step": 10000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.04,
|
| 138 |
+
"learning_rate": 4.998185660157389e-05,
|
| 139 |
+
"loss": 5.7938,
|
| 140 |
+
"step": 10500
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 3.18,
|
| 144 |
+
"learning_rate": 4.996364033809385e-05,
|
| 145 |
+
"loss": 5.7751,
|
| 146 |
+
"step": 11000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.33,
|
| 150 |
+
"learning_rate": 4.9945424074613815e-05,
|
| 151 |
+
"loss": 5.7603,
|
| 152 |
+
"step": 11500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.47,
|
| 156 |
+
"learning_rate": 4.992720781113378e-05,
|
| 157 |
+
"loss": 5.7461,
|
| 158 |
+
"step": 12000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 3.62,
|
| 162 |
+
"learning_rate": 4.9909027980180706e-05,
|
| 163 |
+
"loss": 5.7323,
|
| 164 |
+
"step": 12500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 3.76,
|
| 168 |
+
"learning_rate": 4.9890811716700676e-05,
|
| 169 |
+
"loss": 5.7245,
|
| 170 |
+
"step": 13000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.91,
|
| 174 |
+
"learning_rate": 4.987259545322064e-05,
|
| 175 |
+
"loss": 5.7123,
|
| 176 |
+
"step": 13500
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.05,
|
| 180 |
+
"learning_rate": 4.98543791897406e-05,
|
| 181 |
+
"loss": 5.7033,
|
| 182 |
+
"step": 14000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 4.2,
|
| 186 |
+
"learning_rate": 4.983619935878753e-05,
|
| 187 |
+
"loss": 5.6966,
|
| 188 |
+
"step": 14500
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.34,
|
| 192 |
+
"learning_rate": 4.981798309530749e-05,
|
| 193 |
+
"loss": 5.6868,
|
| 194 |
+
"step": 15000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.48,
|
| 198 |
+
"learning_rate": 4.979976683182746e-05,
|
| 199 |
+
"loss": 5.6794,
|
| 200 |
+
"step": 15500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 4.63,
|
| 204 |
+
"learning_rate": 4.9781550568347426e-05,
|
| 205 |
+
"loss": 5.6726,
|
| 206 |
+
"step": 16000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.77,
|
| 210 |
+
"learning_rate": 4.976337073739435e-05,
|
| 211 |
+
"loss": 5.6679,
|
| 212 |
+
"step": 16500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.92,
|
| 216 |
+
"learning_rate": 4.974515447391431e-05,
|
| 217 |
+
"loss": 5.6571,
|
| 218 |
+
"step": 17000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 5.06,
|
| 222 |
+
"learning_rate": 4.972697464296124e-05,
|
| 223 |
+
"loss": 5.6521,
|
| 224 |
+
"step": 17500
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.21,
|
| 228 |
+
"learning_rate": 4.97087583794812e-05,
|
| 229 |
+
"loss": 5.6461,
|
| 230 |
+
"step": 18000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.35,
|
| 234 |
+
"learning_rate": 4.969054211600117e-05,
|
| 235 |
+
"loss": 5.6394,
|
| 236 |
+
"step": 18500
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 5.5,
|
| 240 |
+
"learning_rate": 4.9672325852521135e-05,
|
| 241 |
+
"loss": 5.638,
|
| 242 |
+
"step": 19000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 5.64,
|
| 246 |
+
"learning_rate": 4.96541095890411e-05,
|
| 247 |
+
"loss": 5.6325,
|
| 248 |
+
"step": 19500
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 5.79,
|
| 252 |
+
"learning_rate": 4.9635929758088026e-05,
|
| 253 |
+
"loss": 5.6291,
|
| 254 |
+
"step": 20000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.93,
|
| 258 |
+
"learning_rate": 4.961771349460799e-05,
|
| 259 |
+
"loss": 5.6244,
|
| 260 |
+
"step": 20500
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.08,
|
| 264 |
+
"learning_rate": 4.959949723112795e-05,
|
| 265 |
+
"loss": 5.62,
|
| 266 |
+
"step": 21000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.22,
|
| 270 |
+
"learning_rate": 4.958128096764792e-05,
|
| 271 |
+
"loss": 5.6133,
|
| 272 |
+
"step": 21500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 6.37,
|
| 276 |
+
"learning_rate": 4.9563064704167886e-05,
|
| 277 |
+
"loss": 5.6107,
|
| 278 |
+
"step": 22000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 6.51,
|
| 282 |
+
"learning_rate": 4.954484844068785e-05,
|
| 283 |
+
"loss": 5.6072,
|
| 284 |
+
"step": 22500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.66,
|
| 288 |
+
"learning_rate": 4.952666860973477e-05,
|
| 289 |
+
"loss": 5.5987,
|
| 290 |
+
"step": 23000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 6.8,
|
| 294 |
+
"learning_rate": 4.950845234625474e-05,
|
| 295 |
+
"loss": 5.598,
|
| 296 |
+
"step": 23500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 6.94,
|
| 300 |
+
"learning_rate": 4.9490236082774703e-05,
|
| 301 |
+
"loss": 5.593,
|
| 302 |
+
"step": 24000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.09,
|
| 306 |
+
"learning_rate": 4.947201981929467e-05,
|
| 307 |
+
"loss": 5.5922,
|
| 308 |
+
"step": 24500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 7.23,
|
| 312 |
+
"learning_rate": 4.9453839988341595e-05,
|
| 313 |
+
"loss": 5.5881,
|
| 314 |
+
"step": 25000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 7.38,
|
| 318 |
+
"learning_rate": 4.943562372486156e-05,
|
| 319 |
+
"loss": 5.5834,
|
| 320 |
+
"step": 25500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 7.52,
|
| 324 |
+
"learning_rate": 4.941740746138153e-05,
|
| 325 |
+
"loss": 5.5838,
|
| 326 |
+
"step": 26000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 7.67,
|
| 330 |
+
"learning_rate": 4.939919119790149e-05,
|
| 331 |
+
"loss": 5.5782,
|
| 332 |
+
"step": 26500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 7.81,
|
| 336 |
+
"learning_rate": 4.9380974934421454e-05,
|
| 337 |
+
"loss": 5.58,
|
| 338 |
+
"step": 27000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 7.96,
|
| 342 |
+
"learning_rate": 4.936279510346838e-05,
|
| 343 |
+
"loss": 5.5751,
|
| 344 |
+
"step": 27500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 8.1,
|
| 348 |
+
"learning_rate": 4.9344578839988345e-05,
|
| 349 |
+
"loss": 5.5704,
|
| 350 |
+
"step": 28000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 8.25,
|
| 354 |
+
"learning_rate": 4.932636257650831e-05,
|
| 355 |
+
"loss": 5.5679,
|
| 356 |
+
"step": 28500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 8.39,
|
| 360 |
+
"learning_rate": 4.930814631302827e-05,
|
| 361 |
+
"loss": 5.5643,
|
| 362 |
+
"step": 29000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 8.54,
|
| 366 |
+
"learning_rate": 4.928993004954824e-05,
|
| 367 |
+
"loss": 5.5602,
|
| 368 |
+
"step": 29500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 8.68,
|
| 372 |
+
"learning_rate": 4.9271713786068205e-05,
|
| 373 |
+
"loss": 5.5599,
|
| 374 |
+
"step": 30000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.83,
|
| 378 |
+
"learning_rate": 4.9253533955115126e-05,
|
| 379 |
+
"loss": 5.5597,
|
| 380 |
+
"step": 30500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 8.97,
|
| 384 |
+
"learning_rate": 4.9235317691635096e-05,
|
| 385 |
+
"loss": 5.5553,
|
| 386 |
+
"step": 31000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 9.11,
|
| 390 |
+
"learning_rate": 4.921710142815506e-05,
|
| 391 |
+
"loss": 5.554,
|
| 392 |
+
"step": 31500
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 9.26,
|
| 396 |
+
"learning_rate": 4.919888516467502e-05,
|
| 397 |
+
"loss": 5.5466,
|
| 398 |
+
"step": 32000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 9.4,
|
| 402 |
+
"learning_rate": 4.918070533372195e-05,
|
| 403 |
+
"loss": 5.5485,
|
| 404 |
+
"step": 32500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 9.55,
|
| 408 |
+
"learning_rate": 4.9162489070241914e-05,
|
| 409 |
+
"loss": 5.5428,
|
| 410 |
+
"step": 33000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 9.69,
|
| 414 |
+
"learning_rate": 4.9144272806761884e-05,
|
| 415 |
+
"loss": 5.5481,
|
| 416 |
+
"step": 33500
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 9.84,
|
| 420 |
+
"learning_rate": 4.912605654328185e-05,
|
| 421 |
+
"loss": 5.5396,
|
| 422 |
+
"step": 34000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 9.98,
|
| 426 |
+
"learning_rate": 4.910784027980181e-05,
|
| 427 |
+
"loss": 5.5398,
|
| 428 |
+
"step": 34500
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 10.13,
|
| 432 |
+
"learning_rate": 4.908966044884873e-05,
|
| 433 |
+
"loss": 5.5342,
|
| 434 |
+
"step": 35000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 10.27,
|
| 438 |
+
"learning_rate": 4.90714441853687e-05,
|
| 439 |
+
"loss": 5.5344,
|
| 440 |
+
"step": 35500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 10.42,
|
| 444 |
+
"learning_rate": 4.9053227921888664e-05,
|
| 445 |
+
"loss": 5.5328,
|
| 446 |
+
"step": 36000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 10.56,
|
| 450 |
+
"learning_rate": 4.903501165840863e-05,
|
| 451 |
+
"loss": 5.5305,
|
| 452 |
+
"step": 36500
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 10.71,
|
| 456 |
+
"learning_rate": 4.9016831827455556e-05,
|
| 457 |
+
"loss": 5.5289,
|
| 458 |
+
"step": 37000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 10.85,
|
| 462 |
+
"learning_rate": 4.899861556397552e-05,
|
| 463 |
+
"loss": 5.5273,
|
| 464 |
+
"step": 37500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.0,
|
| 468 |
+
"learning_rate": 4.898039930049548e-05,
|
| 469 |
+
"loss": 5.5278,
|
| 470 |
+
"step": 38000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 11.14,
|
| 474 |
+
"learning_rate": 4.896218303701545e-05,
|
| 475 |
+
"loss": 5.5238,
|
| 476 |
+
"step": 38500
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 11.28,
|
| 480 |
+
"learning_rate": 4.894400320606237e-05,
|
| 481 |
+
"loss": 5.5243,
|
| 482 |
+
"step": 39000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 11.43,
|
| 486 |
+
"learning_rate": 4.892578694258234e-05,
|
| 487 |
+
"loss": 5.5204,
|
| 488 |
+
"step": 39500
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 11.57,
|
| 492 |
+
"learning_rate": 4.8907570679102306e-05,
|
| 493 |
+
"loss": 5.4293,
|
| 494 |
+
"step": 40000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 11.72,
|
| 498 |
+
"learning_rate": 4.888935441562227e-05,
|
| 499 |
+
"loss": 5.2003,
|
| 500 |
+
"step": 40500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 11.86,
|
| 504 |
+
"learning_rate": 4.887113815214224e-05,
|
| 505 |
+
"loss": 5.0337,
|
| 506 |
+
"step": 41000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 12.01,
|
| 510 |
+
"learning_rate": 4.88529218886622e-05,
|
| 511 |
+
"loss": 4.8789,
|
| 512 |
+
"step": 41500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 12.15,
|
| 516 |
+
"learning_rate": 4.8834742057709124e-05,
|
| 517 |
+
"loss": 4.7277,
|
| 518 |
+
"step": 42000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 12.3,
|
| 522 |
+
"learning_rate": 4.881652579422909e-05,
|
| 523 |
+
"loss": 4.5847,
|
| 524 |
+
"step": 42500
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 12.44,
|
| 528 |
+
"learning_rate": 4.879830953074905e-05,
|
| 529 |
+
"loss": 4.4549,
|
| 530 |
+
"step": 43000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 12.59,
|
| 534 |
+
"learning_rate": 4.878009326726902e-05,
|
| 535 |
+
"loss": 4.3158,
|
| 536 |
+
"step": 43500
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 12.73,
|
| 540 |
+
"learning_rate": 4.8761877003788984e-05,
|
| 541 |
+
"loss": 4.1826,
|
| 542 |
+
"step": 44000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 12.88,
|
| 546 |
+
"learning_rate": 4.874366074030895e-05,
|
| 547 |
+
"loss": 4.0502,
|
| 548 |
+
"step": 44500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 13.02,
|
| 552 |
+
"learning_rate": 4.872544447682891e-05,
|
| 553 |
+
"loss": 3.8966,
|
| 554 |
+
"step": 45000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 13.17,
|
| 558 |
+
"learning_rate": 4.870722821334888e-05,
|
| 559 |
+
"loss": 3.645,
|
| 560 |
+
"step": 45500
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 13.31,
|
| 564 |
+
"learning_rate": 4.868904838239581e-05,
|
| 565 |
+
"loss": 3.1092,
|
| 566 |
+
"step": 46000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 13.45,
|
| 570 |
+
"learning_rate": 4.8670868551442736e-05,
|
| 571 |
+
"loss": 2.6286,
|
| 572 |
+
"step": 46500
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 13.6,
|
| 576 |
+
"learning_rate": 4.86526522879627e-05,
|
| 577 |
+
"loss": 2.419,
|
| 578 |
+
"step": 47000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 13.74,
|
| 582 |
+
"learning_rate": 4.863443602448266e-05,
|
| 583 |
+
"loss": 2.2904,
|
| 584 |
+
"step": 47500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 13.89,
|
| 588 |
+
"learning_rate": 4.8616219761002625e-05,
|
| 589 |
+
"loss": 2.1957,
|
| 590 |
+
"step": 48000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 14.03,
|
| 594 |
+
"learning_rate": 4.8598003497522595e-05,
|
| 595 |
+
"loss": 2.1272,
|
| 596 |
+
"step": 48500
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 14.18,
|
| 600 |
+
"learning_rate": 4.8579823666569517e-05,
|
| 601 |
+
"loss": 2.0666,
|
| 602 |
+
"step": 49000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 14.32,
|
| 606 |
+
"learning_rate": 4.856160740308948e-05,
|
| 607 |
+
"loss": 2.0147,
|
| 608 |
+
"step": 49500
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 14.47,
|
| 612 |
+
"learning_rate": 4.854339113960944e-05,
|
| 613 |
+
"loss": 1.9653,
|
| 614 |
+
"step": 50000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 14.61,
|
| 618 |
+
"learning_rate": 4.8525174876129406e-05,
|
| 619 |
+
"loss": 1.9253,
|
| 620 |
+
"step": 50500
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 14.76,
|
| 624 |
+
"learning_rate": 4.8506958612649376e-05,
|
| 625 |
+
"loss": 1.8919,
|
| 626 |
+
"step": 51000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 14.9,
|
| 630 |
+
"learning_rate": 4.8488778781696304e-05,
|
| 631 |
+
"loss": 1.8565,
|
| 632 |
+
"step": 51500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 15.05,
|
| 636 |
+
"learning_rate": 4.847056251821627e-05,
|
| 637 |
+
"loss": 1.8254,
|
| 638 |
+
"step": 52000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 15.19,
|
| 642 |
+
"learning_rate": 4.845234625473623e-05,
|
| 643 |
+
"loss": 1.7994,
|
| 644 |
+
"step": 52500
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 15.34,
|
| 648 |
+
"learning_rate": 4.843416642378316e-05,
|
| 649 |
+
"loss": 1.7745,
|
| 650 |
+
"step": 53000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 15.48,
|
| 654 |
+
"learning_rate": 4.841595016030312e-05,
|
| 655 |
+
"loss": 1.7497,
|
| 656 |
+
"step": 53500
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 15.62,
|
| 660 |
+
"learning_rate": 4.839773389682309e-05,
|
| 661 |
+
"loss": 1.7284,
|
| 662 |
+
"step": 54000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 15.77,
|
| 666 |
+
"learning_rate": 4.8379517633343055e-05,
|
| 667 |
+
"loss": 1.7083,
|
| 668 |
+
"step": 54500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 15.91,
|
| 672 |
+
"learning_rate": 4.836130136986302e-05,
|
| 673 |
+
"loss": 1.6876,
|
| 674 |
+
"step": 55000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 16.06,
|
| 678 |
+
"learning_rate": 4.834308510638298e-05,
|
| 679 |
+
"loss": 1.6699,
|
| 680 |
+
"step": 55500
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 16.2,
|
| 684 |
+
"learning_rate": 4.832486884290295e-05,
|
| 685 |
+
"loss": 1.6482,
|
| 686 |
+
"step": 56000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 16.35,
|
| 690 |
+
"learning_rate": 4.8306652579422915e-05,
|
| 691 |
+
"loss": 1.6352,
|
| 692 |
+
"step": 56500
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 16.49,
|
| 696 |
+
"learning_rate": 4.828843631594288e-05,
|
| 697 |
+
"loss": 1.6164,
|
| 698 |
+
"step": 57000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 16.64,
|
| 702 |
+
"learning_rate": 4.827022005246284e-05,
|
| 703 |
+
"loss": 1.601,
|
| 704 |
+
"step": 57500
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 16.78,
|
| 708 |
+
"learning_rate": 4.825204022150976e-05,
|
| 709 |
+
"loss": 1.5858,
|
| 710 |
+
"step": 58000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 16.93,
|
| 714 |
+
"learning_rate": 4.823382395802973e-05,
|
| 715 |
+
"loss": 1.5718,
|
| 716 |
+
"step": 58500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 17.07,
|
| 720 |
+
"learning_rate": 4.8215607694549695e-05,
|
| 721 |
+
"loss": 1.5598,
|
| 722 |
+
"step": 59000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 17.22,
|
| 726 |
+
"learning_rate": 4.819739143106966e-05,
|
| 727 |
+
"loss": 1.5424,
|
| 728 |
+
"step": 59500
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 17.36,
|
| 732 |
+
"learning_rate": 4.817917516758962e-05,
|
| 733 |
+
"loss": 1.5323,
|
| 734 |
+
"step": 60000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 17.51,
|
| 738 |
+
"learning_rate": 4.816099533663655e-05,
|
| 739 |
+
"loss": 1.5216,
|
| 740 |
+
"step": 60500
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 17.65,
|
| 744 |
+
"learning_rate": 4.814277907315652e-05,
|
| 745 |
+
"loss": 1.511,
|
| 746 |
+
"step": 61000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 17.8,
|
| 750 |
+
"learning_rate": 4.812456280967648e-05,
|
| 751 |
+
"loss": 1.5005,
|
| 752 |
+
"step": 61500
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 17.94,
|
| 756 |
+
"learning_rate": 4.8106346546196446e-05,
|
| 757 |
+
"loss": 1.488,
|
| 758 |
+
"step": 62000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 18.08,
|
| 762 |
+
"learning_rate": 4.8088166715243374e-05,
|
| 763 |
+
"loss": 1.4788,
|
| 764 |
+
"step": 62500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 18.23,
|
| 768 |
+
"learning_rate": 4.806995045176334e-05,
|
| 769 |
+
"loss": 1.4691,
|
| 770 |
+
"step": 63000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 18.37,
|
| 774 |
+
"learning_rate": 4.80517341882833e-05,
|
| 775 |
+
"loss": 1.4588,
|
| 776 |
+
"step": 63500
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 18.52,
|
| 780 |
+
"learning_rate": 4.803355435733023e-05,
|
| 781 |
+
"loss": 1.4515,
|
| 782 |
+
"step": 64000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 18.66,
|
| 786 |
+
"learning_rate": 4.801533809385019e-05,
|
| 787 |
+
"loss": 1.4426,
|
| 788 |
+
"step": 64500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 18.81,
|
| 792 |
+
"learning_rate": 4.7997121830370155e-05,
|
| 793 |
+
"loss": 1.4344,
|
| 794 |
+
"step": 65000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 18.95,
|
| 798 |
+
"learning_rate": 4.797890556689012e-05,
|
| 799 |
+
"loss": 1.4257,
|
| 800 |
+
"step": 65500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 19.1,
|
| 804 |
+
"learning_rate": 4.796068930341009e-05,
|
| 805 |
+
"loss": 1.4187,
|
| 806 |
+
"step": 66000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 19.24,
|
| 810 |
+
"learning_rate": 4.794247303993005e-05,
|
| 811 |
+
"loss": 1.4092,
|
| 812 |
+
"step": 66500
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 19.39,
|
| 816 |
+
"learning_rate": 4.7924256776450015e-05,
|
| 817 |
+
"loss": 1.4034,
|
| 818 |
+
"step": 67000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 19.53,
|
| 822 |
+
"learning_rate": 4.790604051296998e-05,
|
| 823 |
+
"loss": 1.3929,
|
| 824 |
+
"step": 67500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 19.68,
|
| 828 |
+
"learning_rate": 4.7887860682016906e-05,
|
| 829 |
+
"loss": 1.3877,
|
| 830 |
+
"step": 68000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 19.82,
|
| 834 |
+
"learning_rate": 4.7869680851063833e-05,
|
| 835 |
+
"loss": 1.3808,
|
| 836 |
+
"step": 68500
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 19.97,
|
| 840 |
+
"learning_rate": 4.78514645875838e-05,
|
| 841 |
+
"loss": 1.375,
|
| 842 |
+
"step": 69000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 20.11,
|
| 846 |
+
"learning_rate": 4.783324832410376e-05,
|
| 847 |
+
"loss": 1.3667,
|
| 848 |
+
"step": 69500
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 20.25,
|
| 852 |
+
"learning_rate": 4.781503206062373e-05,
|
| 853 |
+
"loss": 1.3623,
|
| 854 |
+
"step": 70000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 20.4,
|
| 858 |
+
"learning_rate": 4.779681579714369e-05,
|
| 859 |
+
"loss": 1.3533,
|
| 860 |
+
"step": 70500
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 20.54,
|
| 864 |
+
"learning_rate": 4.7778599533663656e-05,
|
| 865 |
+
"loss": 1.3478,
|
| 866 |
+
"step": 71000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 20.69,
|
| 870 |
+
"learning_rate": 4.7760419702710584e-05,
|
| 871 |
+
"loss": 1.3446,
|
| 872 |
+
"step": 71500
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 20.83,
|
| 876 |
+
"learning_rate": 4.774220343923055e-05,
|
| 877 |
+
"loss": 1.338,
|
| 878 |
+
"step": 72000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 20.98,
|
| 882 |
+
"learning_rate": 4.772398717575051e-05,
|
| 883 |
+
"loss": 1.333,
|
| 884 |
+
"step": 72500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 21.12,
|
| 888 |
+
"learning_rate": 4.7705770912270474e-05,
|
| 889 |
+
"loss": 1.3274,
|
| 890 |
+
"step": 73000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 21.27,
|
| 894 |
+
"learning_rate": 4.7687554648790444e-05,
|
| 895 |
+
"loss": 1.3204,
|
| 896 |
+
"step": 73500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 21.41,
|
| 900 |
+
"learning_rate": 4.766933838531041e-05,
|
| 901 |
+
"loss": 1.3178,
|
| 902 |
+
"step": 74000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 21.56,
|
| 906 |
+
"learning_rate": 4.765112212183037e-05,
|
| 907 |
+
"loss": 1.3109,
|
| 908 |
+
"step": 74500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 21.7,
|
| 912 |
+
"learning_rate": 4.76329422908773e-05,
|
| 913 |
+
"loss": 1.3055,
|
| 914 |
+
"step": 75000
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 21.85,
|
| 918 |
+
"learning_rate": 4.761472602739726e-05,
|
| 919 |
+
"loss": 1.2995,
|
| 920 |
+
"step": 75500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 21.99,
|
| 924 |
+
"learning_rate": 4.759654619644419e-05,
|
| 925 |
+
"loss": 1.2955,
|
| 926 |
+
"step": 76000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 22.14,
|
| 930 |
+
"learning_rate": 4.757832993296415e-05,
|
| 931 |
+
"loss": 1.2903,
|
| 932 |
+
"step": 76500
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 22.28,
|
| 936 |
+
"learning_rate": 4.7560113669484116e-05,
|
| 937 |
+
"loss": 1.2848,
|
| 938 |
+
"step": 77000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 22.42,
|
| 942 |
+
"learning_rate": 4.754189740600408e-05,
|
| 943 |
+
"loss": 1.2824,
|
| 944 |
+
"step": 77500
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 22.57,
|
| 948 |
+
"learning_rate": 4.752368114252405e-05,
|
| 949 |
+
"loss": 1.2784,
|
| 950 |
+
"step": 78000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 22.71,
|
| 954 |
+
"learning_rate": 4.750546487904401e-05,
|
| 955 |
+
"loss": 1.2747,
|
| 956 |
+
"step": 78500
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 22.86,
|
| 960 |
+
"learning_rate": 4.748728504809094e-05,
|
| 961 |
+
"loss": 1.2678,
|
| 962 |
+
"step": 79000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 23.0,
|
| 966 |
+
"learning_rate": 4.74690687846109e-05,
|
| 967 |
+
"loss": 1.2636,
|
| 968 |
+
"step": 79500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 23.15,
|
| 972 |
+
"learning_rate": 4.7450852521130867e-05,
|
| 973 |
+
"loss": 1.2593,
|
| 974 |
+
"step": 80000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 23.29,
|
| 978 |
+
"learning_rate": 4.743263625765083e-05,
|
| 979 |
+
"loss": 1.2539,
|
| 980 |
+
"step": 80500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 23.44,
|
| 984 |
+
"learning_rate": 4.74144199941708e-05,
|
| 985 |
+
"loss": 1.2525,
|
| 986 |
+
"step": 81000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 23.58,
|
| 990 |
+
"learning_rate": 4.739620373069076e-05,
|
| 991 |
+
"loss": 1.2484,
|
| 992 |
+
"step": 81500
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 23.73,
|
| 996 |
+
"learning_rate": 4.7377987467210726e-05,
|
| 997 |
+
"loss": 1.2431,
|
| 998 |
+
"step": 82000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 23.87,
|
| 1002 |
+
"learning_rate": 4.735977120373069e-05,
|
| 1003 |
+
"loss": 1.2402,
|
| 1004 |
+
"step": 82500
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 24.02,
|
| 1008 |
+
"learning_rate": 4.734159137277762e-05,
|
| 1009 |
+
"loss": 1.2358,
|
| 1010 |
+
"step": 83000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 24.16,
|
| 1014 |
+
"learning_rate": 4.732337510929759e-05,
|
| 1015 |
+
"loss": 1.2315,
|
| 1016 |
+
"step": 83500
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 24.31,
|
| 1020 |
+
"learning_rate": 4.730515884581755e-05,
|
| 1021 |
+
"loss": 1.2275,
|
| 1022 |
+
"step": 84000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 24.45,
|
| 1026 |
+
"learning_rate": 4.7286942582337514e-05,
|
| 1027 |
+
"loss": 1.2223,
|
| 1028 |
+
"step": 84500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 24.59,
|
| 1032 |
+
"learning_rate": 4.726872631885748e-05,
|
| 1033 |
+
"loss": 1.2205,
|
| 1034 |
+
"step": 85000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 24.74,
|
| 1038 |
+
"learning_rate": 4.725051005537745e-05,
|
| 1039 |
+
"loss": 1.2182,
|
| 1040 |
+
"step": 85500
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 24.88,
|
| 1044 |
+
"learning_rate": 4.723233022442437e-05,
|
| 1045 |
+
"loss": 1.216,
|
| 1046 |
+
"step": 86000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 25.03,
|
| 1050 |
+
"learning_rate": 4.721411396094433e-05,
|
| 1051 |
+
"loss": 1.2103,
|
| 1052 |
+
"step": 86500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 25.17,
|
| 1056 |
+
"learning_rate": 4.7195897697464295e-05,
|
| 1057 |
+
"loss": 1.206,
|
| 1058 |
+
"step": 87000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 25.32,
|
| 1062 |
+
"learning_rate": 4.717768143398426e-05,
|
| 1063 |
+
"loss": 1.206,
|
| 1064 |
+
"step": 87500
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 25.46,
|
| 1068 |
+
"learning_rate": 4.715946517050423e-05,
|
| 1069 |
+
"loss": 1.2014,
|
| 1070 |
+
"step": 88000
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 25.61,
|
| 1074 |
+
"learning_rate": 4.7141285339551156e-05,
|
| 1075 |
+
"loss": 1.199,
|
| 1076 |
+
"step": 88500
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 25.75,
|
| 1080 |
+
"learning_rate": 4.712306907607112e-05,
|
| 1081 |
+
"loss": 1.195,
|
| 1082 |
+
"step": 89000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 25.9,
|
| 1086 |
+
"learning_rate": 4.710485281259108e-05,
|
| 1087 |
+
"loss": 1.1918,
|
| 1088 |
+
"step": 89500
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 26.04,
|
| 1092 |
+
"learning_rate": 4.7086636549111045e-05,
|
| 1093 |
+
"loss": 1.1878,
|
| 1094 |
+
"step": 90000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 26.19,
|
| 1098 |
+
"learning_rate": 4.7068420285631015e-05,
|
| 1099 |
+
"loss": 1.1847,
|
| 1100 |
+
"step": 90500
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 26.33,
|
| 1104 |
+
"learning_rate": 4.705020402215098e-05,
|
| 1105 |
+
"loss": 1.1833,
|
| 1106 |
+
"step": 91000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 26.48,
|
| 1110 |
+
"learning_rate": 4.703198775867094e-05,
|
| 1111 |
+
"loss": 1.1797,
|
| 1112 |
+
"step": 91500
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 26.62,
|
| 1116 |
+
"learning_rate": 4.701377149519091e-05,
|
| 1117 |
+
"loss": 1.178,
|
| 1118 |
+
"step": 92000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 26.77,
|
| 1122 |
+
"learning_rate": 4.6995555231710875e-05,
|
| 1123 |
+
"loss": 1.1754,
|
| 1124 |
+
"step": 92500
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 26.91,
|
| 1128 |
+
"learning_rate": 4.6977411833284754e-05,
|
| 1129 |
+
"loss": 1.1705,
|
| 1130 |
+
"step": 93000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 27.05,
|
| 1134 |
+
"learning_rate": 4.6959195569804724e-05,
|
| 1135 |
+
"loss": 1.168,
|
| 1136 |
+
"step": 93500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 27.2,
|
| 1140 |
+
"learning_rate": 4.694101573885165e-05,
|
| 1141 |
+
"loss": 1.1667,
|
| 1142 |
+
"step": 94000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 27.34,
|
| 1146 |
+
"learning_rate": 4.6922799475371615e-05,
|
| 1147 |
+
"loss": 1.1635,
|
| 1148 |
+
"step": 94500
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 27.49,
|
| 1152 |
+
"learning_rate": 4.690458321189158e-05,
|
| 1153 |
+
"loss": 1.1615,
|
| 1154 |
+
"step": 95000
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 27.63,
|
| 1158 |
+
"learning_rate": 4.688636694841154e-05,
|
| 1159 |
+
"loss": 1.1577,
|
| 1160 |
+
"step": 95500
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 27.78,
|
| 1164 |
+
"learning_rate": 4.686815068493151e-05,
|
| 1165 |
+
"loss": 1.1564,
|
| 1166 |
+
"step": 96000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 27.92,
|
| 1170 |
+
"learning_rate": 4.6849934421451475e-05,
|
| 1171 |
+
"loss": 1.1531,
|
| 1172 |
+
"step": 96500
|
| 1173 |
+
}
|
| 1174 |
+
],
|
| 1175 |
+
"max_steps": 1382400,
|
| 1176 |
+
"num_train_epochs": 400,
|
| 1177 |
+
"total_flos": 2.607846263447198e+19,
|
| 1178 |
+
"trial_name": null,
|
| 1179 |
+
"trial_params": null
|
| 1180 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26e6f7e774db1e0cf987e0bb96d32f23b81111f0e9d9bfeee5adb21624a70cab
|
| 3 |
+
size 3503
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|