Upload trained model from Colab
Browse files- config.json +40 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +59 -0
- trainer_state.json +636 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"HypencoderDualEncoder"
|
| 4 |
+
],
|
| 5 |
+
"loss_kwargs": [
|
| 6 |
+
{
|
| 7 |
+
"only_use_first_item": false,
|
| 8 |
+
"use_in_batch_negatives": true
|
| 9 |
+
}
|
| 10 |
+
],
|
| 11 |
+
"loss_type": [
|
| 12 |
+
"cross_entropy"
|
| 13 |
+
],
|
| 14 |
+
"passage_encoder_kwargs": {
|
| 15 |
+
"freeze_transformer": true,
|
| 16 |
+
"model_name_or_path": "cambridgeltl/SapBERT-from-PubMedBERT-fulltext",
|
| 17 |
+
"pooling_type": "cls"
|
| 18 |
+
},
|
| 19 |
+
"passage_encoder_type": "",
|
| 20 |
+
"query_encoder_kwargs": {
|
| 21 |
+
"base_encoder_output_dim": 768,
|
| 22 |
+
"converter_kwargs": {
|
| 23 |
+
"activation_type": "relu",
|
| 24 |
+
"do_residual_on_last": false,
|
| 25 |
+
"vector_dimensions": [
|
| 26 |
+
768,
|
| 27 |
+
768,
|
| 28 |
+
768,
|
| 29 |
+
1
|
| 30 |
+
]
|
| 31 |
+
},
|
| 32 |
+
"embedding_representation": null,
|
| 33 |
+
"freeze_transformer": true,
|
| 34 |
+
"model_name_or_path": "cambridgeltl/SapBERT-from-PubMedBERT-fulltext"
|
| 35 |
+
},
|
| 36 |
+
"query_encoder_type": "",
|
| 37 |
+
"shared_encoder": true,
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.50.0"
|
| 40 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3e9ce5b646eab1b080d83decbfef0c6acfa580ee686e473b1cfc2db60f6157d
|
| 3 |
+
size 89816779
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86be474515c14dd79b12662916b17b21a28675fdfde3d9c3fe28081fcf78a4b2
|
| 3 |
+
size 482938527
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ead4b63ccfe6bb0cd4044fb2c468922f48d0902fdfb976017847b7a6810c13d
|
| 3 |
+
size 14645
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32d1942f6b31b97eec2ecc7f6945f5159a4d117bbf4eb6f962d45753fe8b0378
|
| 3 |
+
size 1465
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"4": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_basic_tokenize": true,
|
| 47 |
+
"do_lower_case": true,
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"full_tokenizer_file": null,
|
| 50 |
+
"mask_token": "[MASK]",
|
| 51 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 52 |
+
"never_split": null,
|
| 53 |
+
"pad_token": "[PAD]",
|
| 54 |
+
"sep_token": "[SEP]",
|
| 55 |
+
"strip_accents": null,
|
| 56 |
+
"tokenize_chinese_chars": true,
|
| 57 |
+
"tokenizer_class": "BertTokenizer",
|
| 58 |
+
"unk_token": "[UNK]"
|
| 59 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,636 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 860,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.11627906976744186,
|
| 14 |
+
"grad_norm": 108.15798950195312,
|
| 15 |
+
"learning_rate": 2.3255813953488376e-06,
|
| 16 |
+
"loss": 4.5639,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.23255813953488372,
|
| 21 |
+
"grad_norm": 82.37198638916016,
|
| 22 |
+
"learning_rate": 4.651162790697675e-06,
|
| 23 |
+
"loss": 4.3868,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3488372093023256,
|
| 28 |
+
"grad_norm": 71.77903747558594,
|
| 29 |
+
"learning_rate": 6.976744186046513e-06,
|
| 30 |
+
"loss": 4.0951,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.46511627906976744,
|
| 35 |
+
"grad_norm": 66.50274658203125,
|
| 36 |
+
"learning_rate": 9.30232558139535e-06,
|
| 37 |
+
"loss": 3.7553,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.5813953488372093,
|
| 42 |
+
"grad_norm": 73.22584533691406,
|
| 43 |
+
"learning_rate": 1.1627906976744187e-05,
|
| 44 |
+
"loss": 3.3389,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.6976744186046512,
|
| 49 |
+
"grad_norm": 78.85617065429688,
|
| 50 |
+
"learning_rate": 1.3953488372093025e-05,
|
| 51 |
+
"loss": 2.8809,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.813953488372093,
|
| 56 |
+
"grad_norm": 70.36491394042969,
|
| 57 |
+
"learning_rate": 1.6279069767441862e-05,
|
| 58 |
+
"loss": 2.4333,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.9302325581395349,
|
| 63 |
+
"grad_norm": 88.5320053100586,
|
| 64 |
+
"learning_rate": 1.86046511627907e-05,
|
| 65 |
+
"loss": 1.8066,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.0465116279069768,
|
| 70 |
+
"grad_norm": 90.84140014648438,
|
| 71 |
+
"learning_rate": 1.9896640826873385e-05,
|
| 72 |
+
"loss": 1.2644,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.1627906976744187,
|
| 77 |
+
"grad_norm": 107.49101257324219,
|
| 78 |
+
"learning_rate": 1.9638242894056848e-05,
|
| 79 |
+
"loss": 1.0383,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.2790697674418605,
|
| 84 |
+
"grad_norm": 101.91487884521484,
|
| 85 |
+
"learning_rate": 1.937984496124031e-05,
|
| 86 |
+
"loss": 1.0025,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.3953488372093024,
|
| 91 |
+
"grad_norm": 92.1374740600586,
|
| 92 |
+
"learning_rate": 1.9121447028423774e-05,
|
| 93 |
+
"loss": 0.8067,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.5116279069767442,
|
| 98 |
+
"grad_norm": 93.908447265625,
|
| 99 |
+
"learning_rate": 1.8863049095607237e-05,
|
| 100 |
+
"loss": 0.8832,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.627906976744186,
|
| 105 |
+
"grad_norm": 70.7513198852539,
|
| 106 |
+
"learning_rate": 1.86046511627907e-05,
|
| 107 |
+
"loss": 0.7466,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.744186046511628,
|
| 112 |
+
"grad_norm": 87.94328308105469,
|
| 113 |
+
"learning_rate": 1.8346253229974164e-05,
|
| 114 |
+
"loss": 0.7646,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.8604651162790697,
|
| 119 |
+
"grad_norm": 73.78392028808594,
|
| 120 |
+
"learning_rate": 1.8087855297157624e-05,
|
| 121 |
+
"loss": 0.6297,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.9767441860465116,
|
| 126 |
+
"grad_norm": 76.53567504882812,
|
| 127 |
+
"learning_rate": 1.7829457364341087e-05,
|
| 128 |
+
"loss": 0.5853,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 2.0930232558139537,
|
| 133 |
+
"grad_norm": 59.62578582763672,
|
| 134 |
+
"learning_rate": 1.757105943152455e-05,
|
| 135 |
+
"loss": 0.443,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.2093023255813953,
|
| 140 |
+
"grad_norm": 49.40946578979492,
|
| 141 |
+
"learning_rate": 1.7312661498708013e-05,
|
| 142 |
+
"loss": 0.4205,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 2.3255813953488373,
|
| 147 |
+
"grad_norm": 60.58675765991211,
|
| 148 |
+
"learning_rate": 1.7054263565891473e-05,
|
| 149 |
+
"loss": 0.3815,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 2.441860465116279,
|
| 154 |
+
"grad_norm": 84.80896759033203,
|
| 155 |
+
"learning_rate": 1.6795865633074936e-05,
|
| 156 |
+
"loss": 0.4841,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 2.558139534883721,
|
| 161 |
+
"grad_norm": 63.476505279541016,
|
| 162 |
+
"learning_rate": 1.65374677002584e-05,
|
| 163 |
+
"loss": 0.4462,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 2.6744186046511627,
|
| 168 |
+
"grad_norm": 93.7468032836914,
|
| 169 |
+
"learning_rate": 1.6279069767441862e-05,
|
| 170 |
+
"loss": 0.458,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.7906976744186047,
|
| 175 |
+
"grad_norm": 72.0390625,
|
| 176 |
+
"learning_rate": 1.6020671834625325e-05,
|
| 177 |
+
"loss": 0.4498,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 2.9069767441860463,
|
| 182 |
+
"grad_norm": 57.512176513671875,
|
| 183 |
+
"learning_rate": 1.5762273901808785e-05,
|
| 184 |
+
"loss": 0.3927,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 3.0232558139534884,
|
| 189 |
+
"grad_norm": 51.03213882446289,
|
| 190 |
+
"learning_rate": 1.550387596899225e-05,
|
| 191 |
+
"loss": 0.3707,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 3.13953488372093,
|
| 196 |
+
"grad_norm": 60.818233489990234,
|
| 197 |
+
"learning_rate": 1.5245478036175711e-05,
|
| 198 |
+
"loss": 0.3091,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 3.255813953488372,
|
| 203 |
+
"grad_norm": 85.74930572509766,
|
| 204 |
+
"learning_rate": 1.4987080103359175e-05,
|
| 205 |
+
"loss": 0.3512,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.3720930232558137,
|
| 210 |
+
"grad_norm": 61.855953216552734,
|
| 211 |
+
"learning_rate": 1.4728682170542636e-05,
|
| 212 |
+
"loss": 0.3402,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 3.488372093023256,
|
| 217 |
+
"grad_norm": 73.42764282226562,
|
| 218 |
+
"learning_rate": 1.44702842377261e-05,
|
| 219 |
+
"loss": 0.2461,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 3.604651162790698,
|
| 224 |
+
"grad_norm": 58.168540954589844,
|
| 225 |
+
"learning_rate": 1.421188630490956e-05,
|
| 226 |
+
"loss": 0.2955,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 3.7209302325581395,
|
| 231 |
+
"grad_norm": 47.6633415222168,
|
| 232 |
+
"learning_rate": 1.3953488372093025e-05,
|
| 233 |
+
"loss": 0.336,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 3.8372093023255816,
|
| 238 |
+
"grad_norm": 63.79522705078125,
|
| 239 |
+
"learning_rate": 1.3695090439276487e-05,
|
| 240 |
+
"loss": 0.3029,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 3.953488372093023,
|
| 245 |
+
"grad_norm": 71.78118133544922,
|
| 246 |
+
"learning_rate": 1.343669250645995e-05,
|
| 247 |
+
"loss": 0.2938,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 4.069767441860465,
|
| 252 |
+
"grad_norm": 51.58977127075195,
|
| 253 |
+
"learning_rate": 1.3178294573643412e-05,
|
| 254 |
+
"loss": 0.2736,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.186046511627907,
|
| 259 |
+
"grad_norm": 62.82514572143555,
|
| 260 |
+
"learning_rate": 1.2919896640826875e-05,
|
| 261 |
+
"loss": 0.2936,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 4.3023255813953485,
|
| 266 |
+
"grad_norm": 61.00852584838867,
|
| 267 |
+
"learning_rate": 1.2661498708010338e-05,
|
| 268 |
+
"loss": 0.323,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 4.4186046511627906,
|
| 273 |
+
"grad_norm": 15.029560089111328,
|
| 274 |
+
"learning_rate": 1.24031007751938e-05,
|
| 275 |
+
"loss": 0.2644,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.534883720930233,
|
| 280 |
+
"grad_norm": 63.83937072753906,
|
| 281 |
+
"learning_rate": 1.2144702842377262e-05,
|
| 282 |
+
"loss": 0.2595,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 4.651162790697675,
|
| 287 |
+
"grad_norm": 50.91780090332031,
|
| 288 |
+
"learning_rate": 1.1886304909560724e-05,
|
| 289 |
+
"loss": 0.2734,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.767441860465116,
|
| 294 |
+
"grad_norm": 49.75490951538086,
|
| 295 |
+
"learning_rate": 1.1627906976744187e-05,
|
| 296 |
+
"loss": 0.2704,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.883720930232558,
|
| 301 |
+
"grad_norm": 49.518131256103516,
|
| 302 |
+
"learning_rate": 1.1369509043927648e-05,
|
| 303 |
+
"loss": 0.299,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.0,
|
| 308 |
+
"grad_norm": 30.25467872619629,
|
| 309 |
+
"learning_rate": 1.1111111111111113e-05,
|
| 310 |
+
"loss": 0.2598,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 5.116279069767442,
|
| 315 |
+
"grad_norm": 40.72892379760742,
|
| 316 |
+
"learning_rate": 1.0852713178294573e-05,
|
| 317 |
+
"loss": 0.21,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 5.232558139534884,
|
| 322 |
+
"grad_norm": 41.70243835449219,
|
| 323 |
+
"learning_rate": 1.0594315245478038e-05,
|
| 324 |
+
"loss": 0.2307,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 5.348837209302325,
|
| 329 |
+
"grad_norm": 66.0749740600586,
|
| 330 |
+
"learning_rate": 1.03359173126615e-05,
|
| 331 |
+
"loss": 0.1893,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 5.465116279069767,
|
| 336 |
+
"grad_norm": 55.968231201171875,
|
| 337 |
+
"learning_rate": 1.0077519379844963e-05,
|
| 338 |
+
"loss": 0.2739,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 5.5813953488372094,
|
| 343 |
+
"grad_norm": 44.32405090332031,
|
| 344 |
+
"learning_rate": 9.819121447028424e-06,
|
| 345 |
+
"loss": 0.2613,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 5.6976744186046515,
|
| 350 |
+
"grad_norm": 80.5435791015625,
|
| 351 |
+
"learning_rate": 9.560723514211887e-06,
|
| 352 |
+
"loss": 0.2298,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 5.813953488372093,
|
| 357 |
+
"grad_norm": 51.336830139160156,
|
| 358 |
+
"learning_rate": 9.30232558139535e-06,
|
| 359 |
+
"loss": 0.2614,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 5.930232558139535,
|
| 364 |
+
"grad_norm": 24.42147445678711,
|
| 365 |
+
"learning_rate": 9.043927648578812e-06,
|
| 366 |
+
"loss": 0.2813,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 6.046511627906977,
|
| 371 |
+
"grad_norm": 43.1801872253418,
|
| 372 |
+
"learning_rate": 8.785529715762275e-06,
|
| 373 |
+
"loss": 0.1733,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.162790697674419,
|
| 378 |
+
"grad_norm": 46.86786651611328,
|
| 379 |
+
"learning_rate": 8.527131782945736e-06,
|
| 380 |
+
"loss": 0.2382,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 6.27906976744186,
|
| 385 |
+
"grad_norm": 33.578487396240234,
|
| 386 |
+
"learning_rate": 8.2687338501292e-06,
|
| 387 |
+
"loss": 0.1822,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 6.395348837209302,
|
| 392 |
+
"grad_norm": 39.639198303222656,
|
| 393 |
+
"learning_rate": 8.010335917312663e-06,
|
| 394 |
+
"loss": 0.202,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 6.511627906976744,
|
| 399 |
+
"grad_norm": 27.308820724487305,
|
| 400 |
+
"learning_rate": 7.751937984496126e-06,
|
| 401 |
+
"loss": 0.2184,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 6.627906976744186,
|
| 406 |
+
"grad_norm": 30.545543670654297,
|
| 407 |
+
"learning_rate": 7.493540051679587e-06,
|
| 408 |
+
"loss": 0.185,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 6.7441860465116275,
|
| 413 |
+
"grad_norm": 47.775875091552734,
|
| 414 |
+
"learning_rate": 7.23514211886305e-06,
|
| 415 |
+
"loss": 0.2075,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 6.8604651162790695,
|
| 420 |
+
"grad_norm": 50.34706115722656,
|
| 421 |
+
"learning_rate": 6.976744186046513e-06,
|
| 422 |
+
"loss": 0.2519,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 6.976744186046512,
|
| 427 |
+
"grad_norm": 23.37942886352539,
|
| 428 |
+
"learning_rate": 6.718346253229975e-06,
|
| 429 |
+
"loss": 0.2196,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.093023255813954,
|
| 434 |
+
"grad_norm": 29.51810646057129,
|
| 435 |
+
"learning_rate": 6.459948320413437e-06,
|
| 436 |
+
"loss": 0.1596,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.209302325581396,
|
| 441 |
+
"grad_norm": 44.744171142578125,
|
| 442 |
+
"learning_rate": 6.2015503875969e-06,
|
| 443 |
+
"loss": 0.1719,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.325581395348837,
|
| 448 |
+
"grad_norm": 64.436767578125,
|
| 449 |
+
"learning_rate": 5.943152454780362e-06,
|
| 450 |
+
"loss": 0.2198,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.441860465116279,
|
| 455 |
+
"grad_norm": 19.253662109375,
|
| 456 |
+
"learning_rate": 5.684754521963824e-06,
|
| 457 |
+
"loss": 0.1481,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.558139534883721,
|
| 462 |
+
"grad_norm": 33.91790771484375,
|
| 463 |
+
"learning_rate": 5.4263565891472865e-06,
|
| 464 |
+
"loss": 0.2362,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.674418604651163,
|
| 469 |
+
"grad_norm": 27.47454071044922,
|
| 470 |
+
"learning_rate": 5.16795865633075e-06,
|
| 471 |
+
"loss": 0.1741,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.790697674418604,
|
| 476 |
+
"grad_norm": 33.89091491699219,
|
| 477 |
+
"learning_rate": 4.909560723514212e-06,
|
| 478 |
+
"loss": 0.1759,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.906976744186046,
|
| 483 |
+
"grad_norm": 50.72035217285156,
|
| 484 |
+
"learning_rate": 4.651162790697675e-06,
|
| 485 |
+
"loss": 0.1852,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 8.023255813953488,
|
| 490 |
+
"grad_norm": 15.040735244750977,
|
| 491 |
+
"learning_rate": 4.3927648578811375e-06,
|
| 492 |
+
"loss": 0.1454,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.13953488372093,
|
| 497 |
+
"grad_norm": 45.22159194946289,
|
| 498 |
+
"learning_rate": 4.1343669250646e-06,
|
| 499 |
+
"loss": 0.2117,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 8.255813953488373,
|
| 504 |
+
"grad_norm": 18.63289451599121,
|
| 505 |
+
"learning_rate": 3.875968992248063e-06,
|
| 506 |
+
"loss": 0.1718,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 8.372093023255815,
|
| 511 |
+
"grad_norm": 16.854759216308594,
|
| 512 |
+
"learning_rate": 3.617571059431525e-06,
|
| 513 |
+
"loss": 0.176,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 8.488372093023255,
|
| 518 |
+
"grad_norm": 49.81332015991211,
|
| 519 |
+
"learning_rate": 3.3591731266149875e-06,
|
| 520 |
+
"loss": 0.1962,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 8.604651162790697,
|
| 525 |
+
"grad_norm": 31.597118377685547,
|
| 526 |
+
"learning_rate": 3.10077519379845e-06,
|
| 527 |
+
"loss": 0.1685,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 8.720930232558139,
|
| 532 |
+
"grad_norm": 45.57378387451172,
|
| 533 |
+
"learning_rate": 2.842377260981912e-06,
|
| 534 |
+
"loss": 0.1693,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 8.837209302325581,
|
| 539 |
+
"grad_norm": 33.74851989746094,
|
| 540 |
+
"learning_rate": 2.583979328165375e-06,
|
| 541 |
+
"loss": 0.161,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 8.953488372093023,
|
| 546 |
+
"grad_norm": 32.361839294433594,
|
| 547 |
+
"learning_rate": 2.3255813953488376e-06,
|
| 548 |
+
"loss": 0.1661,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 9.069767441860465,
|
| 553 |
+
"grad_norm": 25.80358123779297,
|
| 554 |
+
"learning_rate": 2.0671834625323e-06,
|
| 555 |
+
"loss": 0.1393,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 9.186046511627907,
|
| 560 |
+
"grad_norm": 37.717708587646484,
|
| 561 |
+
"learning_rate": 1.8087855297157624e-06,
|
| 562 |
+
"loss": 0.1522,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 9.30232558139535,
|
| 567 |
+
"grad_norm": 36.026527404785156,
|
| 568 |
+
"learning_rate": 1.550387596899225e-06,
|
| 569 |
+
"loss": 0.1683,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 9.418604651162791,
|
| 574 |
+
"grad_norm": 41.67938232421875,
|
| 575 |
+
"learning_rate": 1.2919896640826874e-06,
|
| 576 |
+
"loss": 0.146,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 9.534883720930232,
|
| 581 |
+
"grad_norm": 26.428848266601562,
|
| 582 |
+
"learning_rate": 1.03359173126615e-06,
|
| 583 |
+
"loss": 0.1653,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 9.651162790697674,
|
| 588 |
+
"grad_norm": 20.74589729309082,
|
| 589 |
+
"learning_rate": 7.751937984496125e-07,
|
| 590 |
+
"loss": 0.1543,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 9.767441860465116,
|
| 595 |
+
"grad_norm": 25.37066078186035,
|
| 596 |
+
"learning_rate": 5.16795865633075e-07,
|
| 597 |
+
"loss": 0.1644,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 9.883720930232558,
|
| 602 |
+
"grad_norm": 56.05203628540039,
|
| 603 |
+
"learning_rate": 2.583979328165375e-07,
|
| 604 |
+
"loss": 0.1921,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 10.0,
|
| 609 |
+
"grad_norm": 0.19254037737846375,
|
| 610 |
+
"learning_rate": 0.0,
|
| 611 |
+
"loss": 0.1691,
|
| 612 |
+
"step": 860
|
| 613 |
+
}
|
| 614 |
+
],
|
| 615 |
+
"logging_steps": 10,
|
| 616 |
+
"max_steps": 860,
|
| 617 |
+
"num_input_tokens_seen": 0,
|
| 618 |
+
"num_train_epochs": 10,
|
| 619 |
+
"save_steps": 2500,
|
| 620 |
+
"stateful_callbacks": {
|
| 621 |
+
"TrainerControl": {
|
| 622 |
+
"args": {
|
| 623 |
+
"should_epoch_stop": false,
|
| 624 |
+
"should_evaluate": false,
|
| 625 |
+
"should_log": false,
|
| 626 |
+
"should_save": true,
|
| 627 |
+
"should_training_stop": true
|
| 628 |
+
},
|
| 629 |
+
"attributes": {}
|
| 630 |
+
}
|
| 631 |
+
},
|
| 632 |
+
"total_flos": 0.0,
|
| 633 |
+
"train_batch_size": 64,
|
| 634 |
+
"trial_name": null,
|
| 635 |
+
"trial_params": null
|
| 636 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28dccd6852282dbe630c1844fcc84e1777fd602b981895a0af7dbc24b4de493a
|
| 3 |
+
size 5841
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|