Upload 7 files
Browse files- config.json +38 -0
- pytorch_model.bin +3 -0
- sentencepiece.bpe.model +3 -0
- special_tokens_map.json +15 -0
- tokenizer_config.json +55 -0
- trainer_state.json +621 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "../model/cino-small-v2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"XLMRobertaForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"gradient_checkpointing": false,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": "Negative",
|
| 16 |
+
"1": "Positive"
|
| 17 |
+
},
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 3072,
|
| 20 |
+
"label2id": {
|
| 21 |
+
"Negative": 0,
|
| 22 |
+
"Positive": 1
|
| 23 |
+
},
|
| 24 |
+
"layer_norm_eps": 1e-05,
|
| 25 |
+
"max_position_embeddings": 514,
|
| 26 |
+
"model_type": "xlm-roberta",
|
| 27 |
+
"num_attention_heads": 12,
|
| 28 |
+
"num_hidden_layers": 6,
|
| 29 |
+
"output_past": true,
|
| 30 |
+
"pad_token_id": 1,
|
| 31 |
+
"position_embedding_type": "absolute",
|
| 32 |
+
"problem_type": "single_label_classification",
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.36.2",
|
| 35 |
+
"type_vocab_size": 1,
|
| 36 |
+
"use_cache": true,
|
| 37 |
+
"vocab_size": 135359
|
| 38 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bea49f6ed12dafc11177d0b643de8ce50949404d43609f645988610edfb94a12
|
| 3 |
+
size 589922991
|
sentencepiece.bpe.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abec8706178924453be115cd2da858ef32de70ba60d0c10300822a732a868cf7
|
| 3 |
+
size 2814898
|
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_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"135358": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": true,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"mask_token": "<mask>",
|
| 49 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 50 |
+
"pad_token": "<pad>",
|
| 51 |
+
"sep_token": "</s>",
|
| 52 |
+
"sp_model_kwargs": {},
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,621 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.781834372217275,
|
| 3 |
+
"best_model_checkpoint": "../saved_model/cino-small-v2_tusa/checkpoint-500",
|
| 4 |
+
"epoch": 40.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"eval_accuracy": 0.759,
|
| 14 |
+
"eval_f1": 0.7449735449735448,
|
| 15 |
+
"eval_loss": 0.538531482219696,
|
| 16 |
+
"eval_precision": 0.7910112359550562,
|
| 17 |
+
"eval_recall": 0.704,
|
| 18 |
+
"eval_runtime": 3.843,
|
| 19 |
+
"eval_samples_per_second": 260.21,
|
| 20 |
+
"eval_steps_per_second": 8.327,
|
| 21 |
+
"step": 250
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 2.0,
|
| 25 |
+
"learning_rate": 2.5e-05,
|
| 26 |
+
"loss": 0.4062,
|
| 27 |
+
"step": 500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 2.0,
|
| 31 |
+
"eval_accuracy": 0.755,
|
| 32 |
+
"eval_f1": 0.781834372217275,
|
| 33 |
+
"eval_loss": 0.5724849700927734,
|
| 34 |
+
"eval_precision": 0.7046548956661316,
|
| 35 |
+
"eval_recall": 0.878,
|
| 36 |
+
"eval_runtime": 3.8652,
|
| 37 |
+
"eval_samples_per_second": 258.722,
|
| 38 |
+
"eval_steps_per_second": 8.279,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 3.0,
|
| 43 |
+
"eval_accuracy": 0.763,
|
| 44 |
+
"eval_f1": 0.7718960538979788,
|
| 45 |
+
"eval_loss": 0.6156216263771057,
|
| 46 |
+
"eval_precision": 0.7439703153988868,
|
| 47 |
+
"eval_recall": 0.802,
|
| 48 |
+
"eval_runtime": 3.8674,
|
| 49 |
+
"eval_samples_per_second": 258.57,
|
| 50 |
+
"eval_steps_per_second": 8.274,
|
| 51 |
+
"step": 750
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4.0,
|
| 55 |
+
"learning_rate": 5e-05,
|
| 56 |
+
"loss": 0.1767,
|
| 57 |
+
"step": 1000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 4.0,
|
| 61 |
+
"eval_accuracy": 0.763,
|
| 62 |
+
"eval_f1": 0.7351955307262571,
|
| 63 |
+
"eval_loss": 0.7199212908744812,
|
| 64 |
+
"eval_precision": 0.8329113924050633,
|
| 65 |
+
"eval_recall": 0.658,
|
| 66 |
+
"eval_runtime": 3.8486,
|
| 67 |
+
"eval_samples_per_second": 259.833,
|
| 68 |
+
"eval_steps_per_second": 8.315,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 5.0,
|
| 73 |
+
"eval_accuracy": 0.72,
|
| 74 |
+
"eval_f1": 0.7188755020080321,
|
| 75 |
+
"eval_loss": 1.1636953353881836,
|
| 76 |
+
"eval_precision": 0.7217741935483871,
|
| 77 |
+
"eval_recall": 0.716,
|
| 78 |
+
"eval_runtime": 3.8679,
|
| 79 |
+
"eval_samples_per_second": 258.536,
|
| 80 |
+
"eval_steps_per_second": 8.273,
|
| 81 |
+
"step": 1250
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 6.0,
|
| 85 |
+
"learning_rate": 4.722222222222222e-05,
|
| 86 |
+
"loss": 0.0851,
|
| 87 |
+
"step": 1500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 6.0,
|
| 91 |
+
"eval_accuracy": 0.668,
|
| 92 |
+
"eval_f1": 0.7167235494880547,
|
| 93 |
+
"eval_loss": 1.6249794960021973,
|
| 94 |
+
"eval_precision": 0.625,
|
| 95 |
+
"eval_recall": 0.84,
|
| 96 |
+
"eval_runtime": 3.8509,
|
| 97 |
+
"eval_samples_per_second": 259.682,
|
| 98 |
+
"eval_steps_per_second": 8.31,
|
| 99 |
+
"step": 1500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 7.0,
|
| 103 |
+
"eval_accuracy": 0.66,
|
| 104 |
+
"eval_f1": 0.6903460837887068,
|
| 105 |
+
"eval_loss": 1.97330904006958,
|
| 106 |
+
"eval_precision": 0.6337792642140468,
|
| 107 |
+
"eval_recall": 0.758,
|
| 108 |
+
"eval_runtime": 3.8611,
|
| 109 |
+
"eval_samples_per_second": 258.997,
|
| 110 |
+
"eval_steps_per_second": 8.288,
|
| 111 |
+
"step": 1750
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 8.0,
|
| 115 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 116 |
+
"loss": 0.0374,
|
| 117 |
+
"step": 2000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 8.0,
|
| 121 |
+
"eval_accuracy": 0.672,
|
| 122 |
+
"eval_f1": 0.6371681415929205,
|
| 123 |
+
"eval_loss": 2.0722925662994385,
|
| 124 |
+
"eval_precision": 0.7128712871287128,
|
| 125 |
+
"eval_recall": 0.576,
|
| 126 |
+
"eval_runtime": 3.8504,
|
| 127 |
+
"eval_samples_per_second": 259.712,
|
| 128 |
+
"eval_steps_per_second": 8.311,
|
| 129 |
+
"step": 2000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 9.0,
|
| 133 |
+
"eval_accuracy": 0.712,
|
| 134 |
+
"eval_f1": 0.7037037037037037,
|
| 135 |
+
"eval_loss": 2.0254833698272705,
|
| 136 |
+
"eval_precision": 0.7245762711864406,
|
| 137 |
+
"eval_recall": 0.684,
|
| 138 |
+
"eval_runtime": 3.8436,
|
| 139 |
+
"eval_samples_per_second": 260.171,
|
| 140 |
+
"eval_steps_per_second": 8.325,
|
| 141 |
+
"step": 2250
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 10.0,
|
| 145 |
+
"learning_rate": 4.166666666666667e-05,
|
| 146 |
+
"loss": 0.0196,
|
| 147 |
+
"step": 2500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 10.0,
|
| 151 |
+
"eval_accuracy": 0.72,
|
| 152 |
+
"eval_f1": 0.7421731123388583,
|
| 153 |
+
"eval_loss": 2.1418700218200684,
|
| 154 |
+
"eval_precision": 0.6877133105802048,
|
| 155 |
+
"eval_recall": 0.806,
|
| 156 |
+
"eval_runtime": 3.8516,
|
| 157 |
+
"eval_samples_per_second": 259.63,
|
| 158 |
+
"eval_steps_per_second": 8.308,
|
| 159 |
+
"step": 2500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 11.0,
|
| 163 |
+
"eval_accuracy": 0.732,
|
| 164 |
+
"eval_f1": 0.7709401709401711,
|
| 165 |
+
"eval_loss": 2.037372350692749,
|
| 166 |
+
"eval_precision": 0.673134328358209,
|
| 167 |
+
"eval_recall": 0.902,
|
| 168 |
+
"eval_runtime": 3.8641,
|
| 169 |
+
"eval_samples_per_second": 258.791,
|
| 170 |
+
"eval_steps_per_second": 8.281,
|
| 171 |
+
"step": 2750
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 12.0,
|
| 175 |
+
"learning_rate": 3.888888888888889e-05,
|
| 176 |
+
"loss": 0.0178,
|
| 177 |
+
"step": 3000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 12.0,
|
| 181 |
+
"eval_accuracy": 0.734,
|
| 182 |
+
"eval_f1": 0.7387033398821218,
|
| 183 |
+
"eval_loss": 2.172621488571167,
|
| 184 |
+
"eval_precision": 0.7258687258687259,
|
| 185 |
+
"eval_recall": 0.752,
|
| 186 |
+
"eval_runtime": 3.853,
|
| 187 |
+
"eval_samples_per_second": 259.541,
|
| 188 |
+
"eval_steps_per_second": 8.305,
|
| 189 |
+
"step": 3000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 13.0,
|
| 193 |
+
"eval_accuracy": 0.694,
|
| 194 |
+
"eval_f1": 0.66,
|
| 195 |
+
"eval_loss": 2.3637471199035645,
|
| 196 |
+
"eval_precision": 0.7425,
|
| 197 |
+
"eval_recall": 0.594,
|
| 198 |
+
"eval_runtime": 3.8612,
|
| 199 |
+
"eval_samples_per_second": 258.984,
|
| 200 |
+
"eval_steps_per_second": 8.288,
|
| 201 |
+
"step": 3250
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 14.0,
|
| 205 |
+
"learning_rate": 3.611111111111111e-05,
|
| 206 |
+
"loss": 0.0166,
|
| 207 |
+
"step": 3500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 14.0,
|
| 211 |
+
"eval_accuracy": 0.679,
|
| 212 |
+
"eval_f1": 0.664576802507837,
|
| 213 |
+
"eval_loss": 2.3070590496063232,
|
| 214 |
+
"eval_precision": 0.6958424507658644,
|
| 215 |
+
"eval_recall": 0.636,
|
| 216 |
+
"eval_runtime": 3.8463,
|
| 217 |
+
"eval_samples_per_second": 259.991,
|
| 218 |
+
"eval_steps_per_second": 8.32,
|
| 219 |
+
"step": 3500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 15.0,
|
| 223 |
+
"eval_accuracy": 0.72,
|
| 224 |
+
"eval_f1": 0.7265625,
|
| 225 |
+
"eval_loss": 2.2964775562286377,
|
| 226 |
+
"eval_precision": 0.7099236641221374,
|
| 227 |
+
"eval_recall": 0.744,
|
| 228 |
+
"eval_runtime": 3.8729,
|
| 229 |
+
"eval_samples_per_second": 258.207,
|
| 230 |
+
"eval_steps_per_second": 8.263,
|
| 231 |
+
"step": 3750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 16.0,
|
| 235 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 236 |
+
"loss": 0.0074,
|
| 237 |
+
"step": 4000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 16.0,
|
| 241 |
+
"eval_accuracy": 0.708,
|
| 242 |
+
"eval_f1": 0.7176015473887815,
|
| 243 |
+
"eval_loss": 2.449232339859009,
|
| 244 |
+
"eval_precision": 0.6947565543071161,
|
| 245 |
+
"eval_recall": 0.742,
|
| 246 |
+
"eval_runtime": 3.861,
|
| 247 |
+
"eval_samples_per_second": 258.998,
|
| 248 |
+
"eval_steps_per_second": 8.288,
|
| 249 |
+
"step": 4000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 17.0,
|
| 253 |
+
"eval_accuracy": 0.726,
|
| 254 |
+
"eval_f1": 0.727634194831014,
|
| 255 |
+
"eval_loss": 2.2294154167175293,
|
| 256 |
+
"eval_precision": 0.7233201581027668,
|
| 257 |
+
"eval_recall": 0.732,
|
| 258 |
+
"eval_runtime": 3.8559,
|
| 259 |
+
"eval_samples_per_second": 259.344,
|
| 260 |
+
"eval_steps_per_second": 8.299,
|
| 261 |
+
"step": 4250
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 18.0,
|
| 265 |
+
"learning_rate": 3.055555555555556e-05,
|
| 266 |
+
"loss": 0.0127,
|
| 267 |
+
"step": 4500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 18.0,
|
| 271 |
+
"eval_accuracy": 0.736,
|
| 272 |
+
"eval_f1": 0.7333333333333334,
|
| 273 |
+
"eval_loss": 2.354280471801758,
|
| 274 |
+
"eval_precision": 0.7408163265306122,
|
| 275 |
+
"eval_recall": 0.726,
|
| 276 |
+
"eval_runtime": 3.8511,
|
| 277 |
+
"eval_samples_per_second": 259.668,
|
| 278 |
+
"eval_steps_per_second": 8.309,
|
| 279 |
+
"step": 4500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 19.0,
|
| 283 |
+
"eval_accuracy": 0.72,
|
| 284 |
+
"eval_f1": 0.688195991091314,
|
| 285 |
+
"eval_loss": 2.646545886993408,
|
| 286 |
+
"eval_precision": 0.7763819095477387,
|
| 287 |
+
"eval_recall": 0.618,
|
| 288 |
+
"eval_runtime": 3.8559,
|
| 289 |
+
"eval_samples_per_second": 259.341,
|
| 290 |
+
"eval_steps_per_second": 8.299,
|
| 291 |
+
"step": 4750
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 20.0,
|
| 295 |
+
"learning_rate": 2.777777777777778e-05,
|
| 296 |
+
"loss": 0.0071,
|
| 297 |
+
"step": 5000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 20.0,
|
| 301 |
+
"eval_accuracy": 0.728,
|
| 302 |
+
"eval_f1": 0.7190082644628099,
|
| 303 |
+
"eval_loss": 2.392427682876587,
|
| 304 |
+
"eval_precision": 0.7435897435897436,
|
| 305 |
+
"eval_recall": 0.696,
|
| 306 |
+
"eval_runtime": 3.841,
|
| 307 |
+
"eval_samples_per_second": 260.348,
|
| 308 |
+
"eval_steps_per_second": 8.331,
|
| 309 |
+
"step": 5000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 21.0,
|
| 313 |
+
"eval_accuracy": 0.712,
|
| 314 |
+
"eval_f1": 0.7181996086105676,
|
| 315 |
+
"eval_loss": 2.35839581489563,
|
| 316 |
+
"eval_precision": 0.7030651340996169,
|
| 317 |
+
"eval_recall": 0.734,
|
| 318 |
+
"eval_runtime": 3.8576,
|
| 319 |
+
"eval_samples_per_second": 259.231,
|
| 320 |
+
"eval_steps_per_second": 8.295,
|
| 321 |
+
"step": 5250
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 22.0,
|
| 325 |
+
"learning_rate": 2.5e-05,
|
| 326 |
+
"loss": 0.0085,
|
| 327 |
+
"step": 5500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 22.0,
|
| 331 |
+
"eval_accuracy": 0.739,
|
| 332 |
+
"eval_f1": 0.7116022099447514,
|
| 333 |
+
"eval_loss": 1.9896154403686523,
|
| 334 |
+
"eval_precision": 0.7950617283950617,
|
| 335 |
+
"eval_recall": 0.644,
|
| 336 |
+
"eval_runtime": 3.8513,
|
| 337 |
+
"eval_samples_per_second": 259.655,
|
| 338 |
+
"eval_steps_per_second": 8.309,
|
| 339 |
+
"step": 5500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 23.0,
|
| 343 |
+
"eval_accuracy": 0.751,
|
| 344 |
+
"eval_f1": 0.7278688524590164,
|
| 345 |
+
"eval_loss": 2.147411584854126,
|
| 346 |
+
"eval_precision": 0.8024096385542169,
|
| 347 |
+
"eval_recall": 0.666,
|
| 348 |
+
"eval_runtime": 3.9041,
|
| 349 |
+
"eval_samples_per_second": 256.144,
|
| 350 |
+
"eval_steps_per_second": 8.197,
|
| 351 |
+
"step": 5750
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 24.0,
|
| 355 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 356 |
+
"loss": 0.0069,
|
| 357 |
+
"step": 6000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 24.0,
|
| 361 |
+
"eval_accuracy": 0.75,
|
| 362 |
+
"eval_f1": 0.7306034482758621,
|
| 363 |
+
"eval_loss": 2.172492265701294,
|
| 364 |
+
"eval_precision": 0.7920560747663551,
|
| 365 |
+
"eval_recall": 0.678,
|
| 366 |
+
"eval_runtime": 3.8555,
|
| 367 |
+
"eval_samples_per_second": 259.368,
|
| 368 |
+
"eval_steps_per_second": 8.3,
|
| 369 |
+
"step": 6000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 25.0,
|
| 373 |
+
"eval_accuracy": 0.732,
|
| 374 |
+
"eval_f1": 0.7112068965517241,
|
| 375 |
+
"eval_loss": 2.3147952556610107,
|
| 376 |
+
"eval_precision": 0.7710280373831776,
|
| 377 |
+
"eval_recall": 0.66,
|
| 378 |
+
"eval_runtime": 3.8522,
|
| 379 |
+
"eval_samples_per_second": 259.59,
|
| 380 |
+
"eval_steps_per_second": 8.307,
|
| 381 |
+
"step": 6250
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 26.0,
|
| 385 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 386 |
+
"loss": 0.0067,
|
| 387 |
+
"step": 6500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 26.0,
|
| 391 |
+
"eval_accuracy": 0.726,
|
| 392 |
+
"eval_f1": 0.7385496183206107,
|
| 393 |
+
"eval_loss": 2.2775561809539795,
|
| 394 |
+
"eval_precision": 0.7062043795620438,
|
| 395 |
+
"eval_recall": 0.774,
|
| 396 |
+
"eval_runtime": 3.8474,
|
| 397 |
+
"eval_samples_per_second": 259.914,
|
| 398 |
+
"eval_steps_per_second": 8.317,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 27.0,
|
| 403 |
+
"eval_accuracy": 0.718,
|
| 404 |
+
"eval_f1": 0.699360341151386,
|
| 405 |
+
"eval_loss": 2.492180109024048,
|
| 406 |
+
"eval_precision": 0.7488584474885844,
|
| 407 |
+
"eval_recall": 0.656,
|
| 408 |
+
"eval_runtime": 3.8591,
|
| 409 |
+
"eval_samples_per_second": 259.131,
|
| 410 |
+
"eval_steps_per_second": 8.292,
|
| 411 |
+
"step": 6750
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 28.0,
|
| 415 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 416 |
+
"loss": 0.0032,
|
| 417 |
+
"step": 7000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 28.0,
|
| 421 |
+
"eval_accuracy": 0.726,
|
| 422 |
+
"eval_f1": 0.7133891213389122,
|
| 423 |
+
"eval_loss": 2.4056484699249268,
|
| 424 |
+
"eval_precision": 0.7478070175438597,
|
| 425 |
+
"eval_recall": 0.682,
|
| 426 |
+
"eval_runtime": 3.8454,
|
| 427 |
+
"eval_samples_per_second": 260.05,
|
| 428 |
+
"eval_steps_per_second": 8.322,
|
| 429 |
+
"step": 7000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 29.0,
|
| 433 |
+
"eval_accuracy": 0.724,
|
| 434 |
+
"eval_f1": 0.7177914110429447,
|
| 435 |
+
"eval_loss": 2.5566771030426025,
|
| 436 |
+
"eval_precision": 0.7343096234309623,
|
| 437 |
+
"eval_recall": 0.702,
|
| 438 |
+
"eval_runtime": 3.8506,
|
| 439 |
+
"eval_samples_per_second": 259.7,
|
| 440 |
+
"eval_steps_per_second": 8.31,
|
| 441 |
+
"step": 7250
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 30.0,
|
| 445 |
+
"learning_rate": 1.388888888888889e-05,
|
| 446 |
+
"loss": 0.001,
|
| 447 |
+
"step": 7500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 30.0,
|
| 451 |
+
"eval_accuracy": 0.733,
|
| 452 |
+
"eval_f1": 0.7516279069767442,
|
| 453 |
+
"eval_loss": 2.586094617843628,
|
| 454 |
+
"eval_precision": 0.7026086956521739,
|
| 455 |
+
"eval_recall": 0.808,
|
| 456 |
+
"eval_runtime": 3.8404,
|
| 457 |
+
"eval_samples_per_second": 260.391,
|
| 458 |
+
"eval_steps_per_second": 8.333,
|
| 459 |
+
"step": 7500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 31.0,
|
| 463 |
+
"eval_accuracy": 0.738,
|
| 464 |
+
"eval_f1": 0.7282157676348547,
|
| 465 |
+
"eval_loss": 2.566093683242798,
|
| 466 |
+
"eval_precision": 0.7564655172413793,
|
| 467 |
+
"eval_recall": 0.702,
|
| 468 |
+
"eval_runtime": 3.8517,
|
| 469 |
+
"eval_samples_per_second": 259.629,
|
| 470 |
+
"eval_steps_per_second": 8.308,
|
| 471 |
+
"step": 7750
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 32.0,
|
| 475 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 476 |
+
"loss": 0.0016,
|
| 477 |
+
"step": 8000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 32.0,
|
| 481 |
+
"eval_accuracy": 0.74,
|
| 482 |
+
"eval_f1": 0.7192224622030238,
|
| 483 |
+
"eval_loss": 2.6535303592681885,
|
| 484 |
+
"eval_precision": 0.7816901408450704,
|
| 485 |
+
"eval_recall": 0.666,
|
| 486 |
+
"eval_runtime": 3.9723,
|
| 487 |
+
"eval_samples_per_second": 251.74,
|
| 488 |
+
"eval_steps_per_second": 8.056,
|
| 489 |
+
"step": 8000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 33.0,
|
| 493 |
+
"eval_accuracy": 0.74,
|
| 494 |
+
"eval_f1": 0.717391304347826,
|
| 495 |
+
"eval_loss": 2.7205991744995117,
|
| 496 |
+
"eval_precision": 0.7857142857142857,
|
| 497 |
+
"eval_recall": 0.66,
|
| 498 |
+
"eval_runtime": 3.9432,
|
| 499 |
+
"eval_samples_per_second": 253.602,
|
| 500 |
+
"eval_steps_per_second": 8.115,
|
| 501 |
+
"step": 8250
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 34.0,
|
| 505 |
+
"learning_rate": 8.333333333333334e-06,
|
| 506 |
+
"loss": 0.0014,
|
| 507 |
+
"step": 8500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 34.0,
|
| 511 |
+
"eval_accuracy": 0.725,
|
| 512 |
+
"eval_f1": 0.7422680412371134,
|
| 513 |
+
"eval_loss": 2.6660046577453613,
|
| 514 |
+
"eval_precision": 0.6984126984126984,
|
| 515 |
+
"eval_recall": 0.792,
|
| 516 |
+
"eval_runtime": 3.9996,
|
| 517 |
+
"eval_samples_per_second": 250.027,
|
| 518 |
+
"eval_steps_per_second": 8.001,
|
| 519 |
+
"step": 8500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 35.0,
|
| 523 |
+
"eval_accuracy": 0.742,
|
| 524 |
+
"eval_f1": 0.7361963190184049,
|
| 525 |
+
"eval_loss": 2.5981457233428955,
|
| 526 |
+
"eval_precision": 0.7531380753138075,
|
| 527 |
+
"eval_recall": 0.72,
|
| 528 |
+
"eval_runtime": 3.9964,
|
| 529 |
+
"eval_samples_per_second": 250.226,
|
| 530 |
+
"eval_steps_per_second": 8.007,
|
| 531 |
+
"step": 8750
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 36.0,
|
| 535 |
+
"learning_rate": 5.555555555555556e-06,
|
| 536 |
+
"loss": 0.0003,
|
| 537 |
+
"step": 9000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 36.0,
|
| 541 |
+
"eval_accuracy": 0.745,
|
| 542 |
+
"eval_f1": 0.7384615384615385,
|
| 543 |
+
"eval_loss": 2.6297309398651123,
|
| 544 |
+
"eval_precision": 0.7578947368421053,
|
| 545 |
+
"eval_recall": 0.72,
|
| 546 |
+
"eval_runtime": 3.9747,
|
| 547 |
+
"eval_samples_per_second": 251.591,
|
| 548 |
+
"eval_steps_per_second": 8.051,
|
| 549 |
+
"step": 9000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 37.0,
|
| 553 |
+
"eval_accuracy": 0.738,
|
| 554 |
+
"eval_f1": 0.7358870967741935,
|
| 555 |
+
"eval_loss": 2.6337215900421143,
|
| 556 |
+
"eval_precision": 0.741869918699187,
|
| 557 |
+
"eval_recall": 0.73,
|
| 558 |
+
"eval_runtime": 3.9752,
|
| 559 |
+
"eval_samples_per_second": 251.558,
|
| 560 |
+
"eval_steps_per_second": 8.05,
|
| 561 |
+
"step": 9250
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 38.0,
|
| 565 |
+
"learning_rate": 2.777777777777778e-06,
|
| 566 |
+
"loss": 0.0007,
|
| 567 |
+
"step": 9500
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 38.0,
|
| 571 |
+
"eval_accuracy": 0.739,
|
| 572 |
+
"eval_f1": 0.7300930713547054,
|
| 573 |
+
"eval_loss": 2.645658493041992,
|
| 574 |
+
"eval_precision": 0.7558886509635975,
|
| 575 |
+
"eval_recall": 0.706,
|
| 576 |
+
"eval_runtime": 3.9919,
|
| 577 |
+
"eval_samples_per_second": 250.508,
|
| 578 |
+
"eval_steps_per_second": 8.016,
|
| 579 |
+
"step": 9500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 39.0,
|
| 583 |
+
"eval_accuracy": 0.741,
|
| 584 |
+
"eval_f1": 0.7304890738813736,
|
| 585 |
+
"eval_loss": 2.6524713039398193,
|
| 586 |
+
"eval_precision": 0.7613882863340564,
|
| 587 |
+
"eval_recall": 0.702,
|
| 588 |
+
"eval_runtime": 3.9627,
|
| 589 |
+
"eval_samples_per_second": 252.356,
|
| 590 |
+
"eval_steps_per_second": 8.075,
|
| 591 |
+
"step": 9750
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 40.0,
|
| 595 |
+
"learning_rate": 0.0,
|
| 596 |
+
"loss": 0.0,
|
| 597 |
+
"step": 10000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 40.0,
|
| 601 |
+
"eval_accuracy": 0.742,
|
| 602 |
+
"eval_f1": 0.734020618556701,
|
| 603 |
+
"eval_loss": 2.6484949588775635,
|
| 604 |
+
"eval_precision": 0.7574468085106383,
|
| 605 |
+
"eval_recall": 0.712,
|
| 606 |
+
"eval_runtime": 3.9449,
|
| 607 |
+
"eval_samples_per_second": 253.495,
|
| 608 |
+
"eval_steps_per_second": 8.112,
|
| 609 |
+
"step": 10000
|
| 610 |
+
}
|
| 611 |
+
],
|
| 612 |
+
"logging_steps": 500,
|
| 613 |
+
"max_steps": 10000,
|
| 614 |
+
"num_input_tokens_seen": 0,
|
| 615 |
+
"num_train_epochs": 40,
|
| 616 |
+
"save_steps": 500,
|
| 617 |
+
"total_flos": 4.238956756992e+16,
|
| 618 |
+
"train_batch_size": 32,
|
| 619 |
+
"trial_name": null,
|
| 620 |
+
"trial_params": null
|
| 621 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c9d296093f6d5970881323e5267bcb2c9595693ac4cc65611c24f2eaf6012de0
|
| 3 |
+
size 4271
|