Commit
·
f3651eb
1
Parent(s):
c35ec4d
Upload v1 model (10k samples, 2k steps)
Browse files- config.json +70 -0
- generation_config.json +16 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +64 -0
- trainer_state.json +2480 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "philschmid/bart-large-cnn-samsum",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"activation_dropout": 0.0,
|
| 5 |
+
"activation_function": "gelu",
|
| 6 |
+
"add_final_layer_norm": false,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"BartForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"classif_dropout": 0.0,
|
| 13 |
+
"classifier_dropout": 0.0,
|
| 14 |
+
"d_model": 1024,
|
| 15 |
+
"decoder_attention_heads": 16,
|
| 16 |
+
"decoder_ffn_dim": 4096,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 12,
|
| 19 |
+
"decoder_start_token_id": 2,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"early_stopping": true,
|
| 22 |
+
"encoder_attention_heads": 16,
|
| 23 |
+
"encoder_ffn_dim": 4096,
|
| 24 |
+
"encoder_layerdrop": 0.0,
|
| 25 |
+
"encoder_layers": 12,
|
| 26 |
+
"eos_token_id": 2,
|
| 27 |
+
"force_bos_token_to_be_generated": true,
|
| 28 |
+
"forced_bos_token_id": 0,
|
| 29 |
+
"forced_eos_token_id": 2,
|
| 30 |
+
"gradient_checkpointing": false,
|
| 31 |
+
"id2label": {
|
| 32 |
+
"0": "LABEL_0",
|
| 33 |
+
"1": "LABEL_1",
|
| 34 |
+
"2": "LABEL_2"
|
| 35 |
+
},
|
| 36 |
+
"init_std": 0.02,
|
| 37 |
+
"is_encoder_decoder": true,
|
| 38 |
+
"label2id": {
|
| 39 |
+
"LABEL_0": 0,
|
| 40 |
+
"LABEL_1": 1,
|
| 41 |
+
"LABEL_2": 2
|
| 42 |
+
},
|
| 43 |
+
"length_penalty": 2.0,
|
| 44 |
+
"max_length": 142,
|
| 45 |
+
"max_position_embeddings": 1024,
|
| 46 |
+
"min_length": 56,
|
| 47 |
+
"model_type": "bart",
|
| 48 |
+
"no_repeat_ngram_size": 3,
|
| 49 |
+
"normalize_before": false,
|
| 50 |
+
"num_beams": 4,
|
| 51 |
+
"num_hidden_layers": 12,
|
| 52 |
+
"output_past": true,
|
| 53 |
+
"pad_token_id": 1,
|
| 54 |
+
"prefix": " ",
|
| 55 |
+
"scale_embedding": false,
|
| 56 |
+
"task_specific_params": {
|
| 57 |
+
"summarization": {
|
| 58 |
+
"early_stopping": true,
|
| 59 |
+
"length_penalty": 2.0,
|
| 60 |
+
"max_length": 142,
|
| 61 |
+
"min_length": 56,
|
| 62 |
+
"no_repeat_ngram_size": 3,
|
| 63 |
+
"num_beams": 4
|
| 64 |
+
}
|
| 65 |
+
},
|
| 66 |
+
"torch_dtype": "float32",
|
| 67 |
+
"transformers_version": "4.31.0.dev0",
|
| 68 |
+
"use_cache": true,
|
| 69 |
+
"vocab_size": 50264
|
| 70 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 2,
|
| 5 |
+
"early_stopping": true,
|
| 6 |
+
"eos_token_id": 2,
|
| 7 |
+
"forced_bos_token_id": 0,
|
| 8 |
+
"forced_eos_token_id": 2,
|
| 9 |
+
"length_penalty": 2.0,
|
| 10 |
+
"max_length": 142,
|
| 11 |
+
"min_length": 56,
|
| 12 |
+
"no_repeat_ngram_size": 3,
|
| 13 |
+
"num_beams": 4,
|
| 14 |
+
"pad_token_id": 1,
|
| 15 |
+
"transformers_version": "4.31.0.dev0"
|
| 16 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2269ba175c0ddb85ee39653cf80fdc8b608fadb676cf88ffda4a5da3ec8cd6c6
|
| 3 |
+
size 3250623963
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5bae1961e1eb654de469a18628bea3d1b9f358fe15078c50490ff9637751294
|
| 3 |
+
size 1625537293
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afaa7d2204af1b9ea9d5f56de32581dab41698f7e6d7cc921c4220c5540af0da
|
| 3 |
+
size 14575
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f2bb79ec8fdedf4ae57cb390e5769bf8e41e5a8c164fae4510e68640d13abcd
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 1024,
|
| 38 |
+
"pad_token": {
|
| 39 |
+
"__type": "AddedToken",
|
| 40 |
+
"content": "<pad>",
|
| 41 |
+
"lstrip": false,
|
| 42 |
+
"normalized": true,
|
| 43 |
+
"rstrip": false,
|
| 44 |
+
"single_word": false
|
| 45 |
+
},
|
| 46 |
+
"sep_token": {
|
| 47 |
+
"__type": "AddedToken",
|
| 48 |
+
"content": "</s>",
|
| 49 |
+
"lstrip": false,
|
| 50 |
+
"normalized": true,
|
| 51 |
+
"rstrip": false,
|
| 52 |
+
"single_word": false
|
| 53 |
+
},
|
| 54 |
+
"tokenizer_class": "BartTokenizer",
|
| 55 |
+
"trim_offsets": true,
|
| 56 |
+
"unk_token": {
|
| 57 |
+
"__type": "AddedToken",
|
| 58 |
+
"content": "<unk>",
|
| 59 |
+
"lstrip": false,
|
| 60 |
+
"normalized": true,
|
| 61 |
+
"rstrip": false,
|
| 62 |
+
"single_word": false
|
| 63 |
+
}
|
| 64 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2480 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.3787956237792969,
|
| 3 |
+
"best_model_checkpoint": "test-calls-summarization/checkpoint-1000",
|
| 4 |
+
"epoch": 25.579536370903277,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.06,
|
| 12 |
+
"learning_rate": 4.9999999999999996e-05,
|
| 13 |
+
"loss": 2.0533,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.13,
|
| 18 |
+
"learning_rate": 9.999999999999999e-05,
|
| 19 |
+
"loss": 1.0512,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.19,
|
| 24 |
+
"learning_rate": 0.00015,
|
| 25 |
+
"loss": 0.7919,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.26,
|
| 30 |
+
"learning_rate": 0.00019999999999999998,
|
| 31 |
+
"loss": 0.6876,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.32,
|
| 36 |
+
"learning_rate": 0.00025,
|
| 37 |
+
"loss": 0.6567,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.38,
|
| 42 |
+
"learning_rate": 0.0003,
|
| 43 |
+
"loss": 0.6271,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.45,
|
| 48 |
+
"learning_rate": 0.0002998495486459378,
|
| 49 |
+
"loss": 0.5943,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.51,
|
| 54 |
+
"learning_rate": 0.00029969909729187557,
|
| 55 |
+
"loss": 0.5666,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.58,
|
| 60 |
+
"learning_rate": 0.0002995486459378134,
|
| 61 |
+
"loss": 0.5603,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.64,
|
| 66 |
+
"learning_rate": 0.0002993981945837512,
|
| 67 |
+
"loss": 0.5282,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.7,
|
| 72 |
+
"learning_rate": 0.00029924774322968905,
|
| 73 |
+
"loss": 0.5173,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.77,
|
| 78 |
+
"learning_rate": 0.00029909729187562687,
|
| 79 |
+
"loss": 0.5011,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.83,
|
| 84 |
+
"learning_rate": 0.0002989468405215647,
|
| 85 |
+
"loss": 0.4926,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.9,
|
| 90 |
+
"learning_rate": 0.00029879638916750247,
|
| 91 |
+
"loss": 0.478,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.96,
|
| 96 |
+
"learning_rate": 0.0002986459378134403,
|
| 97 |
+
"loss": 0.4766,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.02,
|
| 102 |
+
"learning_rate": 0.0002984954864593781,
|
| 103 |
+
"loss": 0.4756,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.09,
|
| 108 |
+
"learning_rate": 0.0002983450351053159,
|
| 109 |
+
"loss": 0.4229,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.15,
|
| 114 |
+
"learning_rate": 0.0002981945837512537,
|
| 115 |
+
"loss": 0.4239,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.22,
|
| 120 |
+
"learning_rate": 0.00029804413239719154,
|
| 121 |
+
"loss": 0.431,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.28,
|
| 126 |
+
"learning_rate": 0.00029789368104312936,
|
| 127 |
+
"loss": 0.4333,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.34,
|
| 132 |
+
"learning_rate": 0.0002977432296890672,
|
| 133 |
+
"loss": 0.4218,
|
| 134 |
+
"step": 105
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.41,
|
| 138 |
+
"learning_rate": 0.000297592778335005,
|
| 139 |
+
"loss": 0.4357,
|
| 140 |
+
"step": 110
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.47,
|
| 144 |
+
"learning_rate": 0.0002974423269809428,
|
| 145 |
+
"loss": 0.4197,
|
| 146 |
+
"step": 115
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.53,
|
| 150 |
+
"learning_rate": 0.0002972918756268806,
|
| 151 |
+
"loss": 0.4213,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.6,
|
| 156 |
+
"learning_rate": 0.00029714142427281844,
|
| 157 |
+
"loss": 0.4087,
|
| 158 |
+
"step": 125
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.66,
|
| 162 |
+
"learning_rate": 0.0002969909729187562,
|
| 163 |
+
"loss": 0.4041,
|
| 164 |
+
"step": 130
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.73,
|
| 168 |
+
"learning_rate": 0.00029684052156469403,
|
| 169 |
+
"loss": 0.4103,
|
| 170 |
+
"step": 135
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.79,
|
| 174 |
+
"learning_rate": 0.00029669007021063186,
|
| 175 |
+
"loss": 0.4009,
|
| 176 |
+
"step": 140
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.85,
|
| 180 |
+
"learning_rate": 0.0002965396188565697,
|
| 181 |
+
"loss": 0.4077,
|
| 182 |
+
"step": 145
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.92,
|
| 186 |
+
"learning_rate": 0.0002963891675025075,
|
| 187 |
+
"loss": 0.4257,
|
| 188 |
+
"step": 150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.98,
|
| 192 |
+
"learning_rate": 0.00029623871614844533,
|
| 193 |
+
"loss": 0.4111,
|
| 194 |
+
"step": 155
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.05,
|
| 198 |
+
"learning_rate": 0.0002960882647943831,
|
| 199 |
+
"loss": 0.3634,
|
| 200 |
+
"step": 160
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.11,
|
| 204 |
+
"learning_rate": 0.00029593781344032093,
|
| 205 |
+
"loss": 0.3532,
|
| 206 |
+
"step": 165
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.17,
|
| 210 |
+
"learning_rate": 0.00029578736208625875,
|
| 211 |
+
"loss": 0.3591,
|
| 212 |
+
"step": 170
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.24,
|
| 216 |
+
"learning_rate": 0.0002956369107321965,
|
| 217 |
+
"loss": 0.3506,
|
| 218 |
+
"step": 175
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.3,
|
| 222 |
+
"learning_rate": 0.00029548645937813435,
|
| 223 |
+
"loss": 0.3494,
|
| 224 |
+
"step": 180
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.37,
|
| 228 |
+
"learning_rate": 0.0002953360080240722,
|
| 229 |
+
"loss": 0.3598,
|
| 230 |
+
"step": 185
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.43,
|
| 234 |
+
"learning_rate": 0.00029518555667001,
|
| 235 |
+
"loss": 0.3656,
|
| 236 |
+
"step": 190
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.49,
|
| 240 |
+
"learning_rate": 0.0002950351053159478,
|
| 241 |
+
"loss": 0.351,
|
| 242 |
+
"step": 195
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.56,
|
| 246 |
+
"learning_rate": 0.00029488465396188565,
|
| 247 |
+
"loss": 0.3589,
|
| 248 |
+
"step": 200
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.62,
|
| 252 |
+
"learning_rate": 0.0002947342026078234,
|
| 253 |
+
"loss": 0.3564,
|
| 254 |
+
"step": 205
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.69,
|
| 258 |
+
"learning_rate": 0.00029458375125376125,
|
| 259 |
+
"loss": 0.3665,
|
| 260 |
+
"step": 210
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.75,
|
| 264 |
+
"learning_rate": 0.0002944332998996991,
|
| 265 |
+
"loss": 0.3472,
|
| 266 |
+
"step": 215
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.81,
|
| 270 |
+
"learning_rate": 0.0002942828485456369,
|
| 271 |
+
"loss": 0.3876,
|
| 272 |
+
"step": 220
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.88,
|
| 276 |
+
"learning_rate": 0.0002941323971915747,
|
| 277 |
+
"loss": 0.3831,
|
| 278 |
+
"step": 225
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.94,
|
| 282 |
+
"learning_rate": 0.00029398194583751255,
|
| 283 |
+
"loss": 0.3597,
|
| 284 |
+
"step": 230
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 3.01,
|
| 288 |
+
"learning_rate": 0.0002938314944834503,
|
| 289 |
+
"loss": 0.3781,
|
| 290 |
+
"step": 235
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.07,
|
| 294 |
+
"learning_rate": 0.00029368104312938815,
|
| 295 |
+
"loss": 0.3078,
|
| 296 |
+
"step": 240
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 3.13,
|
| 300 |
+
"learning_rate": 0.00029353059177532597,
|
| 301 |
+
"loss": 0.2986,
|
| 302 |
+
"step": 245
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.2,
|
| 306 |
+
"learning_rate": 0.00029338014042126374,
|
| 307 |
+
"loss": 0.3031,
|
| 308 |
+
"step": 250
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.2,
|
| 312 |
+
"eval_loss": 0.9930399656295776,
|
| 313 |
+
"eval_runtime": 0.3176,
|
| 314 |
+
"eval_samples_per_second": 31.491,
|
| 315 |
+
"eval_steps_per_second": 3.149,
|
| 316 |
+
"step": 250
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.26,
|
| 320 |
+
"learning_rate": 0.00029322968906720157,
|
| 321 |
+
"loss": 0.3132,
|
| 322 |
+
"step": 255
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 3.33,
|
| 326 |
+
"learning_rate": 0.0002930792377131394,
|
| 327 |
+
"loss": 0.3044,
|
| 328 |
+
"step": 260
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 3.39,
|
| 332 |
+
"learning_rate": 0.0002929287863590772,
|
| 333 |
+
"loss": 0.3154,
|
| 334 |
+
"step": 265
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 3.45,
|
| 338 |
+
"learning_rate": 0.00029277833500501504,
|
| 339 |
+
"loss": 0.3143,
|
| 340 |
+
"step": 270
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 3.52,
|
| 344 |
+
"learning_rate": 0.00029262788365095287,
|
| 345 |
+
"loss": 0.3169,
|
| 346 |
+
"step": 275
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 3.58,
|
| 350 |
+
"learning_rate": 0.00029247743229689064,
|
| 351 |
+
"loss": 0.3181,
|
| 352 |
+
"step": 280
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.65,
|
| 356 |
+
"learning_rate": 0.00029232698094282846,
|
| 357 |
+
"loss": 0.3227,
|
| 358 |
+
"step": 285
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.71,
|
| 362 |
+
"learning_rate": 0.0002921765295887663,
|
| 363 |
+
"loss": 0.3181,
|
| 364 |
+
"step": 290
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 3.77,
|
| 368 |
+
"learning_rate": 0.00029202607823470406,
|
| 369 |
+
"loss": 0.3528,
|
| 370 |
+
"step": 295
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 3.84,
|
| 374 |
+
"learning_rate": 0.0002918756268806419,
|
| 375 |
+
"loss": 0.3341,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 3.9,
|
| 380 |
+
"learning_rate": 0.0002917251755265797,
|
| 381 |
+
"loss": 0.3274,
|
| 382 |
+
"step": 305
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 3.96,
|
| 386 |
+
"learning_rate": 0.00029157472417251754,
|
| 387 |
+
"loss": 0.3227,
|
| 388 |
+
"step": 310
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.03,
|
| 392 |
+
"learning_rate": 0.00029142427281845536,
|
| 393 |
+
"loss": 0.3069,
|
| 394 |
+
"step": 315
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 4.09,
|
| 398 |
+
"learning_rate": 0.0002912738214643932,
|
| 399 |
+
"loss": 0.2687,
|
| 400 |
+
"step": 320
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 4.16,
|
| 404 |
+
"learning_rate": 0.00029112337011033096,
|
| 405 |
+
"loss": 0.2736,
|
| 406 |
+
"step": 325
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 4.22,
|
| 410 |
+
"learning_rate": 0.0002909729187562688,
|
| 411 |
+
"loss": 0.2736,
|
| 412 |
+
"step": 330
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 4.28,
|
| 416 |
+
"learning_rate": 0.0002908224674022066,
|
| 417 |
+
"loss": 0.2684,
|
| 418 |
+
"step": 335
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 4.35,
|
| 422 |
+
"learning_rate": 0.0002906720160481444,
|
| 423 |
+
"loss": 0.2608,
|
| 424 |
+
"step": 340
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 4.41,
|
| 428 |
+
"learning_rate": 0.0002905215646940822,
|
| 429 |
+
"loss": 0.2688,
|
| 430 |
+
"step": 345
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.48,
|
| 434 |
+
"learning_rate": 0.00029037111334002003,
|
| 435 |
+
"loss": 0.3065,
|
| 436 |
+
"step": 350
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 4.54,
|
| 440 |
+
"learning_rate": 0.00029022066198595785,
|
| 441 |
+
"loss": 0.2872,
|
| 442 |
+
"step": 355
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.6,
|
| 446 |
+
"learning_rate": 0.0002900702106318957,
|
| 447 |
+
"loss": 0.285,
|
| 448 |
+
"step": 360
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.67,
|
| 452 |
+
"learning_rate": 0.0002899197592778335,
|
| 453 |
+
"loss": 0.4004,
|
| 454 |
+
"step": 365
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 4.73,
|
| 458 |
+
"learning_rate": 0.0002897693079237713,
|
| 459 |
+
"loss": 0.2869,
|
| 460 |
+
"step": 370
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.8,
|
| 464 |
+
"learning_rate": 0.0002896188565697091,
|
| 465 |
+
"loss": 0.288,
|
| 466 |
+
"step": 375
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.86,
|
| 470 |
+
"learning_rate": 0.0002894684052156469,
|
| 471 |
+
"loss": 0.2779,
|
| 472 |
+
"step": 380
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 4.92,
|
| 476 |
+
"learning_rate": 0.0002893179538615847,
|
| 477 |
+
"loss": 0.2856,
|
| 478 |
+
"step": 385
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 4.99,
|
| 482 |
+
"learning_rate": 0.0002891675025075225,
|
| 483 |
+
"loss": 0.3069,
|
| 484 |
+
"step": 390
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 5.05,
|
| 488 |
+
"learning_rate": 0.00028901705115346035,
|
| 489 |
+
"loss": 0.2484,
|
| 490 |
+
"step": 395
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 5.12,
|
| 494 |
+
"learning_rate": 0.0002888665997993982,
|
| 495 |
+
"loss": 0.2466,
|
| 496 |
+
"step": 400
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 5.18,
|
| 500 |
+
"learning_rate": 0.000288716148445336,
|
| 501 |
+
"loss": 0.2498,
|
| 502 |
+
"step": 405
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 5.24,
|
| 506 |
+
"learning_rate": 0.0002885656970912738,
|
| 507 |
+
"loss": 0.2475,
|
| 508 |
+
"step": 410
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 5.31,
|
| 512 |
+
"learning_rate": 0.0002884152457372116,
|
| 513 |
+
"loss": 0.2406,
|
| 514 |
+
"step": 415
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 5.37,
|
| 518 |
+
"learning_rate": 0.0002882647943831494,
|
| 519 |
+
"loss": 0.2436,
|
| 520 |
+
"step": 420
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 5.44,
|
| 524 |
+
"learning_rate": 0.00028811434302908724,
|
| 525 |
+
"loss": 0.2492,
|
| 526 |
+
"step": 425
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 5.5,
|
| 530 |
+
"learning_rate": 0.000287963891675025,
|
| 531 |
+
"loss": 0.2553,
|
| 532 |
+
"step": 430
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 5.56,
|
| 536 |
+
"learning_rate": 0.00028781344032096284,
|
| 537 |
+
"loss": 0.2551,
|
| 538 |
+
"step": 435
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 5.63,
|
| 542 |
+
"learning_rate": 0.00028766298896690067,
|
| 543 |
+
"loss": 0.2585,
|
| 544 |
+
"step": 440
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 5.69,
|
| 548 |
+
"learning_rate": 0.0002875125376128385,
|
| 549 |
+
"loss": 0.2648,
|
| 550 |
+
"step": 445
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 5.76,
|
| 554 |
+
"learning_rate": 0.0002873620862587763,
|
| 555 |
+
"loss": 0.2755,
|
| 556 |
+
"step": 450
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 5.82,
|
| 560 |
+
"learning_rate": 0.00028721163490471414,
|
| 561 |
+
"loss": 0.2614,
|
| 562 |
+
"step": 455
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.88,
|
| 566 |
+
"learning_rate": 0.0002870611835506519,
|
| 567 |
+
"loss": 0.2678,
|
| 568 |
+
"step": 460
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 5.95,
|
| 572 |
+
"learning_rate": 0.00028691073219658974,
|
| 573 |
+
"loss": 0.2726,
|
| 574 |
+
"step": 465
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 6.01,
|
| 578 |
+
"learning_rate": 0.00028676028084252756,
|
| 579 |
+
"loss": 0.2561,
|
| 580 |
+
"step": 470
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 6.08,
|
| 584 |
+
"learning_rate": 0.00028660982948846533,
|
| 585 |
+
"loss": 0.214,
|
| 586 |
+
"step": 475
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 6.14,
|
| 590 |
+
"learning_rate": 0.00028645937813440316,
|
| 591 |
+
"loss": 0.2213,
|
| 592 |
+
"step": 480
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 6.2,
|
| 596 |
+
"learning_rate": 0.000286308926780341,
|
| 597 |
+
"loss": 0.2217,
|
| 598 |
+
"step": 485
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 6.27,
|
| 602 |
+
"learning_rate": 0.0002861584754262788,
|
| 603 |
+
"loss": 0.2213,
|
| 604 |
+
"step": 490
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 6.33,
|
| 608 |
+
"learning_rate": 0.00028600802407221664,
|
| 609 |
+
"loss": 0.2344,
|
| 610 |
+
"step": 495
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 6.39,
|
| 614 |
+
"learning_rate": 0.00028585757271815446,
|
| 615 |
+
"loss": 0.4603,
|
| 616 |
+
"step": 500
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 6.39,
|
| 620 |
+
"eval_loss": 1.0373685359954834,
|
| 621 |
+
"eval_runtime": 0.3183,
|
| 622 |
+
"eval_samples_per_second": 31.418,
|
| 623 |
+
"eval_steps_per_second": 3.142,
|
| 624 |
+
"step": 500
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 6.46,
|
| 628 |
+
"learning_rate": 0.00028570712136409223,
|
| 629 |
+
"loss": 0.2353,
|
| 630 |
+
"step": 505
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 6.52,
|
| 634 |
+
"learning_rate": 0.00028555667001003006,
|
| 635 |
+
"loss": 0.2269,
|
| 636 |
+
"step": 510
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 6.59,
|
| 640 |
+
"learning_rate": 0.0002854062186559679,
|
| 641 |
+
"loss": 0.2318,
|
| 642 |
+
"step": 515
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 6.65,
|
| 646 |
+
"learning_rate": 0.00028525576730190565,
|
| 647 |
+
"loss": 0.2338,
|
| 648 |
+
"step": 520
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 6.71,
|
| 652 |
+
"learning_rate": 0.0002851053159478435,
|
| 653 |
+
"loss": 0.2336,
|
| 654 |
+
"step": 525
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 6.78,
|
| 658 |
+
"learning_rate": 0.0002849548645937813,
|
| 659 |
+
"loss": 0.2385,
|
| 660 |
+
"step": 530
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.84,
|
| 664 |
+
"learning_rate": 0.00028480441323971913,
|
| 665 |
+
"loss": 0.2356,
|
| 666 |
+
"step": 535
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 6.91,
|
| 670 |
+
"learning_rate": 0.00028465396188565695,
|
| 671 |
+
"loss": 0.2437,
|
| 672 |
+
"step": 540
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 6.97,
|
| 676 |
+
"learning_rate": 0.0002845035105315948,
|
| 677 |
+
"loss": 0.2417,
|
| 678 |
+
"step": 545
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 7.03,
|
| 682 |
+
"learning_rate": 0.00028435305917753255,
|
| 683 |
+
"loss": 0.2167,
|
| 684 |
+
"step": 550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 7.1,
|
| 688 |
+
"learning_rate": 0.0002842026078234704,
|
| 689 |
+
"loss": 0.1974,
|
| 690 |
+
"step": 555
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 7.16,
|
| 694 |
+
"learning_rate": 0.0002840521564694082,
|
| 695 |
+
"loss": 0.2092,
|
| 696 |
+
"step": 560
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 7.23,
|
| 700 |
+
"learning_rate": 0.000283901705115346,
|
| 701 |
+
"loss": 0.2009,
|
| 702 |
+
"step": 565
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 7.29,
|
| 706 |
+
"learning_rate": 0.00028375125376128385,
|
| 707 |
+
"loss": 0.5423,
|
| 708 |
+
"step": 570
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 7.35,
|
| 712 |
+
"learning_rate": 0.0002836008024072217,
|
| 713 |
+
"loss": 0.7044,
|
| 714 |
+
"step": 575
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 7.42,
|
| 718 |
+
"learning_rate": 0.00028345035105315945,
|
| 719 |
+
"loss": 0.3172,
|
| 720 |
+
"step": 580
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 7.48,
|
| 724 |
+
"learning_rate": 0.00028329989969909727,
|
| 725 |
+
"loss": 0.2165,
|
| 726 |
+
"step": 585
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 7.55,
|
| 730 |
+
"learning_rate": 0.0002831494483450351,
|
| 731 |
+
"loss": 0.2017,
|
| 732 |
+
"step": 590
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 7.61,
|
| 736 |
+
"learning_rate": 0.00028299899699097287,
|
| 737 |
+
"loss": 0.1976,
|
| 738 |
+
"step": 595
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 7.67,
|
| 742 |
+
"learning_rate": 0.0002828485456369107,
|
| 743 |
+
"loss": 0.2036,
|
| 744 |
+
"step": 600
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 7.74,
|
| 748 |
+
"learning_rate": 0.0002826980942828485,
|
| 749 |
+
"loss": 0.2082,
|
| 750 |
+
"step": 605
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 7.8,
|
| 754 |
+
"learning_rate": 0.00028254764292878634,
|
| 755 |
+
"loss": 0.2104,
|
| 756 |
+
"step": 610
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 7.87,
|
| 760 |
+
"learning_rate": 0.00028239719157472417,
|
| 761 |
+
"loss": 0.2194,
|
| 762 |
+
"step": 615
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 7.93,
|
| 766 |
+
"learning_rate": 0.000282246740220662,
|
| 767 |
+
"loss": 0.2104,
|
| 768 |
+
"step": 620
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 7.99,
|
| 772 |
+
"learning_rate": 0.00028209628886659977,
|
| 773 |
+
"loss": 0.2183,
|
| 774 |
+
"step": 625
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 8.06,
|
| 778 |
+
"learning_rate": 0.0002819458375125376,
|
| 779 |
+
"loss": 0.1773,
|
| 780 |
+
"step": 630
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 8.12,
|
| 784 |
+
"learning_rate": 0.0002817953861584754,
|
| 785 |
+
"loss": 0.1753,
|
| 786 |
+
"step": 635
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 8.19,
|
| 790 |
+
"learning_rate": 0.0002816449348044132,
|
| 791 |
+
"loss": 0.1737,
|
| 792 |
+
"step": 640
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 8.25,
|
| 796 |
+
"learning_rate": 0.000281494483450351,
|
| 797 |
+
"loss": 0.1807,
|
| 798 |
+
"step": 645
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 8.31,
|
| 802 |
+
"learning_rate": 0.00028134403209628884,
|
| 803 |
+
"loss": 0.1858,
|
| 804 |
+
"step": 650
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 8.38,
|
| 808 |
+
"learning_rate": 0.00028119358074222666,
|
| 809 |
+
"loss": 0.1836,
|
| 810 |
+
"step": 655
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 8.44,
|
| 814 |
+
"learning_rate": 0.0002810431293881645,
|
| 815 |
+
"loss": 0.1887,
|
| 816 |
+
"step": 660
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 8.51,
|
| 820 |
+
"learning_rate": 0.0002808926780341023,
|
| 821 |
+
"loss": 0.1908,
|
| 822 |
+
"step": 665
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 8.57,
|
| 826 |
+
"learning_rate": 0.0002807422266800401,
|
| 827 |
+
"loss": 0.1933,
|
| 828 |
+
"step": 670
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 8.63,
|
| 832 |
+
"learning_rate": 0.0002805917753259779,
|
| 833 |
+
"loss": 0.1969,
|
| 834 |
+
"step": 675
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 8.7,
|
| 838 |
+
"learning_rate": 0.00028044132397191574,
|
| 839 |
+
"loss": 0.194,
|
| 840 |
+
"step": 680
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 8.76,
|
| 844 |
+
"learning_rate": 0.0002802908726178535,
|
| 845 |
+
"loss": 0.1895,
|
| 846 |
+
"step": 685
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 8.82,
|
| 850 |
+
"learning_rate": 0.00028014042126379133,
|
| 851 |
+
"loss": 0.1957,
|
| 852 |
+
"step": 690
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 8.89,
|
| 856 |
+
"learning_rate": 0.00027998996990972916,
|
| 857 |
+
"loss": 0.1927,
|
| 858 |
+
"step": 695
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 8.95,
|
| 862 |
+
"learning_rate": 0.000279839518555667,
|
| 863 |
+
"loss": 0.1891,
|
| 864 |
+
"step": 700
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 9.02,
|
| 868 |
+
"learning_rate": 0.0002796890672016048,
|
| 869 |
+
"loss": 0.1809,
|
| 870 |
+
"step": 705
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 9.08,
|
| 874 |
+
"learning_rate": 0.00027953861584754263,
|
| 875 |
+
"loss": 0.1443,
|
| 876 |
+
"step": 710
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 9.14,
|
| 880 |
+
"learning_rate": 0.0002793881644934804,
|
| 881 |
+
"loss": 0.1487,
|
| 882 |
+
"step": 715
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 9.21,
|
| 886 |
+
"learning_rate": 0.00027923771313941823,
|
| 887 |
+
"loss": 0.1521,
|
| 888 |
+
"step": 720
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 9.27,
|
| 892 |
+
"learning_rate": 0.00027908726178535605,
|
| 893 |
+
"loss": 0.1544,
|
| 894 |
+
"step": 725
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 9.34,
|
| 898 |
+
"learning_rate": 0.0002789368104312938,
|
| 899 |
+
"loss": 0.1564,
|
| 900 |
+
"step": 730
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 9.4,
|
| 904 |
+
"learning_rate": 0.00027878635907723165,
|
| 905 |
+
"loss": 0.1634,
|
| 906 |
+
"step": 735
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 9.46,
|
| 910 |
+
"learning_rate": 0.0002786359077231695,
|
| 911 |
+
"loss": 0.1644,
|
| 912 |
+
"step": 740
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 9.53,
|
| 916 |
+
"learning_rate": 0.0002784854563691073,
|
| 917 |
+
"loss": 0.1643,
|
| 918 |
+
"step": 745
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 9.59,
|
| 922 |
+
"learning_rate": 0.0002783350050150451,
|
| 923 |
+
"loss": 0.1721,
|
| 924 |
+
"step": 750
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 9.59,
|
| 928 |
+
"eval_loss": 1.1369589567184448,
|
| 929 |
+
"eval_runtime": 0.3185,
|
| 930 |
+
"eval_samples_per_second": 31.399,
|
| 931 |
+
"eval_steps_per_second": 3.14,
|
| 932 |
+
"step": 750
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 9.66,
|
| 936 |
+
"learning_rate": 0.00027818455366098295,
|
| 937 |
+
"loss": 0.1755,
|
| 938 |
+
"step": 755
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 9.72,
|
| 942 |
+
"learning_rate": 0.0002780341023069207,
|
| 943 |
+
"loss": 0.1733,
|
| 944 |
+
"step": 760
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 9.78,
|
| 948 |
+
"learning_rate": 0.00027788365095285855,
|
| 949 |
+
"loss": 0.1689,
|
| 950 |
+
"step": 765
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 9.85,
|
| 954 |
+
"learning_rate": 0.00027773319959879637,
|
| 955 |
+
"loss": 0.1668,
|
| 956 |
+
"step": 770
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 9.91,
|
| 960 |
+
"learning_rate": 0.00027758274824473414,
|
| 961 |
+
"loss": 0.1774,
|
| 962 |
+
"step": 775
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 9.98,
|
| 966 |
+
"learning_rate": 0.00027743229689067197,
|
| 967 |
+
"loss": 0.1742,
|
| 968 |
+
"step": 780
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 10.04,
|
| 972 |
+
"learning_rate": 0.0002772818455366098,
|
| 973 |
+
"loss": 0.1513,
|
| 974 |
+
"step": 785
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 10.1,
|
| 978 |
+
"learning_rate": 0.0002771313941825476,
|
| 979 |
+
"loss": 0.1361,
|
| 980 |
+
"step": 790
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 10.17,
|
| 984 |
+
"learning_rate": 0.00027698094282848544,
|
| 985 |
+
"loss": 0.1387,
|
| 986 |
+
"step": 795
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 10.23,
|
| 990 |
+
"learning_rate": 0.00027683049147442327,
|
| 991 |
+
"loss": 0.1472,
|
| 992 |
+
"step": 800
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 10.3,
|
| 996 |
+
"learning_rate": 0.00027668004012036104,
|
| 997 |
+
"loss": 0.1439,
|
| 998 |
+
"step": 805
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 10.36,
|
| 1002 |
+
"learning_rate": 0.00027652958876629887,
|
| 1003 |
+
"loss": 0.1389,
|
| 1004 |
+
"step": 810
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 10.42,
|
| 1008 |
+
"learning_rate": 0.0002763791374122367,
|
| 1009 |
+
"loss": 0.1454,
|
| 1010 |
+
"step": 815
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 10.49,
|
| 1014 |
+
"learning_rate": 0.00027622868605817446,
|
| 1015 |
+
"loss": 0.1504,
|
| 1016 |
+
"step": 820
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 10.55,
|
| 1020 |
+
"learning_rate": 0.0002760782347041123,
|
| 1021 |
+
"loss": 0.1475,
|
| 1022 |
+
"step": 825
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 10.62,
|
| 1026 |
+
"learning_rate": 0.0002759277833500501,
|
| 1027 |
+
"loss": 0.1478,
|
| 1028 |
+
"step": 830
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 10.68,
|
| 1032 |
+
"learning_rate": 0.00027577733199598794,
|
| 1033 |
+
"loss": 0.1496,
|
| 1034 |
+
"step": 835
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 10.74,
|
| 1038 |
+
"learning_rate": 0.00027562688064192576,
|
| 1039 |
+
"loss": 0.1475,
|
| 1040 |
+
"step": 840
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 10.81,
|
| 1044 |
+
"learning_rate": 0.0002754764292878636,
|
| 1045 |
+
"loss": 0.1499,
|
| 1046 |
+
"step": 845
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 10.87,
|
| 1050 |
+
"learning_rate": 0.00027532597793380136,
|
| 1051 |
+
"loss": 0.152,
|
| 1052 |
+
"step": 850
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 10.94,
|
| 1056 |
+
"learning_rate": 0.0002751755265797392,
|
| 1057 |
+
"loss": 0.1514,
|
| 1058 |
+
"step": 855
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 11.0,
|
| 1062 |
+
"learning_rate": 0.000275025075225677,
|
| 1063 |
+
"loss": 0.153,
|
| 1064 |
+
"step": 860
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 11.06,
|
| 1068 |
+
"learning_rate": 0.0002748746238716148,
|
| 1069 |
+
"loss": 0.1153,
|
| 1070 |
+
"step": 865
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 11.13,
|
| 1074 |
+
"learning_rate": 0.0002747241725175526,
|
| 1075 |
+
"loss": 0.1181,
|
| 1076 |
+
"step": 870
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 11.19,
|
| 1080 |
+
"learning_rate": 0.00027457372116349043,
|
| 1081 |
+
"loss": 0.1201,
|
| 1082 |
+
"step": 875
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 11.25,
|
| 1086 |
+
"learning_rate": 0.00027442326980942826,
|
| 1087 |
+
"loss": 0.1164,
|
| 1088 |
+
"step": 880
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 11.32,
|
| 1092 |
+
"learning_rate": 0.0002742728184553661,
|
| 1093 |
+
"loss": 0.1177,
|
| 1094 |
+
"step": 885
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 11.38,
|
| 1098 |
+
"learning_rate": 0.0002741223671013039,
|
| 1099 |
+
"loss": 0.1205,
|
| 1100 |
+
"step": 890
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 11.45,
|
| 1104 |
+
"learning_rate": 0.0002739719157472417,
|
| 1105 |
+
"loss": 0.1257,
|
| 1106 |
+
"step": 895
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 11.51,
|
| 1110 |
+
"learning_rate": 0.0002738214643931795,
|
| 1111 |
+
"loss": 0.1291,
|
| 1112 |
+
"step": 900
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 11.57,
|
| 1116 |
+
"learning_rate": 0.00027367101303911733,
|
| 1117 |
+
"loss": 0.1277,
|
| 1118 |
+
"step": 905
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 11.64,
|
| 1122 |
+
"learning_rate": 0.00027352056168505515,
|
| 1123 |
+
"loss": 0.1343,
|
| 1124 |
+
"step": 910
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 11.7,
|
| 1128 |
+
"learning_rate": 0.000273370110330993,
|
| 1129 |
+
"loss": 0.1312,
|
| 1130 |
+
"step": 915
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 11.77,
|
| 1134 |
+
"learning_rate": 0.0002732196589769308,
|
| 1135 |
+
"loss": 0.1314,
|
| 1136 |
+
"step": 920
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 11.83,
|
| 1140 |
+
"learning_rate": 0.0002730692076228686,
|
| 1141 |
+
"loss": 0.1367,
|
| 1142 |
+
"step": 925
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 11.89,
|
| 1146 |
+
"learning_rate": 0.0002729187562688064,
|
| 1147 |
+
"loss": 0.1336,
|
| 1148 |
+
"step": 930
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 11.96,
|
| 1152 |
+
"learning_rate": 0.0002727683049147442,
|
| 1153 |
+
"loss": 0.1367,
|
| 1154 |
+
"step": 935
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 12.02,
|
| 1158 |
+
"learning_rate": 0.000272617853560682,
|
| 1159 |
+
"loss": 0.1266,
|
| 1160 |
+
"step": 940
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 12.09,
|
| 1164 |
+
"learning_rate": 0.0002724674022066198,
|
| 1165 |
+
"loss": 0.1001,
|
| 1166 |
+
"step": 945
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 12.15,
|
| 1170 |
+
"learning_rate": 0.00027231695085255765,
|
| 1171 |
+
"loss": 0.106,
|
| 1172 |
+
"step": 950
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 12.21,
|
| 1176 |
+
"learning_rate": 0.00027216649949849547,
|
| 1177 |
+
"loss": 0.1078,
|
| 1178 |
+
"step": 955
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 12.28,
|
| 1182 |
+
"learning_rate": 0.0002720160481444333,
|
| 1183 |
+
"loss": 0.1065,
|
| 1184 |
+
"step": 960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 12.34,
|
| 1188 |
+
"learning_rate": 0.0002718655967903711,
|
| 1189 |
+
"loss": 0.1077,
|
| 1190 |
+
"step": 965
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 12.41,
|
| 1194 |
+
"learning_rate": 0.0002717151454363089,
|
| 1195 |
+
"loss": 0.1089,
|
| 1196 |
+
"step": 970
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 12.47,
|
| 1200 |
+
"learning_rate": 0.0002715646940822467,
|
| 1201 |
+
"loss": 0.1094,
|
| 1202 |
+
"step": 975
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 12.53,
|
| 1206 |
+
"learning_rate": 0.00027141424272818454,
|
| 1207 |
+
"loss": 0.1205,
|
| 1208 |
+
"step": 980
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 12.6,
|
| 1212 |
+
"learning_rate": 0.0002712637913741223,
|
| 1213 |
+
"loss": 0.1231,
|
| 1214 |
+
"step": 985
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 12.66,
|
| 1218 |
+
"learning_rate": 0.00027111334002006014,
|
| 1219 |
+
"loss": 0.1169,
|
| 1220 |
+
"step": 990
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 12.73,
|
| 1224 |
+
"learning_rate": 0.00027096288866599797,
|
| 1225 |
+
"loss": 0.1154,
|
| 1226 |
+
"step": 995
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 12.79,
|
| 1230 |
+
"learning_rate": 0.0002708124373119358,
|
| 1231 |
+
"loss": 0.113,
|
| 1232 |
+
"step": 1000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 12.79,
|
| 1236 |
+
"eval_loss": 1.3787956237792969,
|
| 1237 |
+
"eval_runtime": 0.319,
|
| 1238 |
+
"eval_samples_per_second": 31.345,
|
| 1239 |
+
"eval_steps_per_second": 3.135,
|
| 1240 |
+
"step": 1000
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 12.85,
|
| 1244 |
+
"learning_rate": 0.0002706619859578736,
|
| 1245 |
+
"loss": 0.1123,
|
| 1246 |
+
"step": 1005
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 12.92,
|
| 1250 |
+
"learning_rate": 0.00027051153460381144,
|
| 1251 |
+
"loss": 0.117,
|
| 1252 |
+
"step": 1010
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 12.98,
|
| 1256 |
+
"learning_rate": 0.0002703610832497492,
|
| 1257 |
+
"loss": 0.1199,
|
| 1258 |
+
"step": 1015
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 13.05,
|
| 1262 |
+
"learning_rate": 0.00027021063189568704,
|
| 1263 |
+
"loss": 0.0943,
|
| 1264 |
+
"step": 1020
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 13.11,
|
| 1268 |
+
"learning_rate": 0.00027006018054162486,
|
| 1269 |
+
"loss": 0.0851,
|
| 1270 |
+
"step": 1025
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 13.17,
|
| 1274 |
+
"learning_rate": 0.00026990972918756263,
|
| 1275 |
+
"loss": 0.0906,
|
| 1276 |
+
"step": 1030
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 13.24,
|
| 1280 |
+
"learning_rate": 0.00026975927783350046,
|
| 1281 |
+
"loss": 0.0911,
|
| 1282 |
+
"step": 1035
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 13.3,
|
| 1286 |
+
"learning_rate": 0.0002696088264794383,
|
| 1287 |
+
"loss": 0.0925,
|
| 1288 |
+
"step": 1040
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 13.37,
|
| 1292 |
+
"learning_rate": 0.0002694583751253761,
|
| 1293 |
+
"loss": 0.095,
|
| 1294 |
+
"step": 1045
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 13.43,
|
| 1298 |
+
"learning_rate": 0.00026930792377131393,
|
| 1299 |
+
"loss": 0.0921,
|
| 1300 |
+
"step": 1050
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 13.49,
|
| 1304 |
+
"learning_rate": 0.00026915747241725176,
|
| 1305 |
+
"loss": 0.0958,
|
| 1306 |
+
"step": 1055
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 13.56,
|
| 1310 |
+
"learning_rate": 0.00026900702106318953,
|
| 1311 |
+
"loss": 0.0969,
|
| 1312 |
+
"step": 1060
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 13.62,
|
| 1316 |
+
"learning_rate": 0.00026885656970912736,
|
| 1317 |
+
"loss": 0.0973,
|
| 1318 |
+
"step": 1065
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 13.69,
|
| 1322 |
+
"learning_rate": 0.0002687061183550652,
|
| 1323 |
+
"loss": 0.1008,
|
| 1324 |
+
"step": 1070
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 13.75,
|
| 1328 |
+
"learning_rate": 0.00026855566700100295,
|
| 1329 |
+
"loss": 0.1021,
|
| 1330 |
+
"step": 1075
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 13.81,
|
| 1334 |
+
"learning_rate": 0.0002684052156469408,
|
| 1335 |
+
"loss": 0.1018,
|
| 1336 |
+
"step": 1080
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 13.88,
|
| 1340 |
+
"learning_rate": 0.0002682547642928786,
|
| 1341 |
+
"loss": 0.1019,
|
| 1342 |
+
"step": 1085
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 13.94,
|
| 1346 |
+
"learning_rate": 0.00026810431293881643,
|
| 1347 |
+
"loss": 0.1035,
|
| 1348 |
+
"step": 1090
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 14.0,
|
| 1352 |
+
"learning_rate": 0.00026795386158475425,
|
| 1353 |
+
"loss": 0.1081,
|
| 1354 |
+
"step": 1095
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 14.07,
|
| 1358 |
+
"learning_rate": 0.0002678034102306921,
|
| 1359 |
+
"loss": 0.0823,
|
| 1360 |
+
"step": 1100
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 14.13,
|
| 1364 |
+
"learning_rate": 0.00026765295887662985,
|
| 1365 |
+
"loss": 0.0767,
|
| 1366 |
+
"step": 1105
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 14.2,
|
| 1370 |
+
"learning_rate": 0.0002675025075225677,
|
| 1371 |
+
"loss": 0.0808,
|
| 1372 |
+
"step": 1110
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 14.26,
|
| 1376 |
+
"learning_rate": 0.0002673520561685055,
|
| 1377 |
+
"loss": 0.0827,
|
| 1378 |
+
"step": 1115
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 14.32,
|
| 1382 |
+
"learning_rate": 0.00026720160481444327,
|
| 1383 |
+
"loss": 0.0831,
|
| 1384 |
+
"step": 1120
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 14.39,
|
| 1388 |
+
"learning_rate": 0.0002670511534603811,
|
| 1389 |
+
"loss": 0.0853,
|
| 1390 |
+
"step": 1125
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 14.45,
|
| 1394 |
+
"learning_rate": 0.0002669007021063189,
|
| 1395 |
+
"loss": 0.084,
|
| 1396 |
+
"step": 1130
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 14.52,
|
| 1400 |
+
"learning_rate": 0.00026675025075225675,
|
| 1401 |
+
"loss": 0.0867,
|
| 1402 |
+
"step": 1135
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 14.58,
|
| 1406 |
+
"learning_rate": 0.00026659979939819457,
|
| 1407 |
+
"loss": 0.0856,
|
| 1408 |
+
"step": 1140
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 14.64,
|
| 1412 |
+
"learning_rate": 0.0002664493480441324,
|
| 1413 |
+
"loss": 0.0864,
|
| 1414 |
+
"step": 1145
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 14.71,
|
| 1418 |
+
"learning_rate": 0.00026629889669007017,
|
| 1419 |
+
"loss": 0.0867,
|
| 1420 |
+
"step": 1150
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 14.77,
|
| 1424 |
+
"learning_rate": 0.000266148445336008,
|
| 1425 |
+
"loss": 0.0862,
|
| 1426 |
+
"step": 1155
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 14.84,
|
| 1430 |
+
"learning_rate": 0.0002659979939819458,
|
| 1431 |
+
"loss": 0.0902,
|
| 1432 |
+
"step": 1160
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 14.9,
|
| 1436 |
+
"learning_rate": 0.0002658475426278836,
|
| 1437 |
+
"loss": 0.0924,
|
| 1438 |
+
"step": 1165
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 14.96,
|
| 1442 |
+
"learning_rate": 0.0002656970912738214,
|
| 1443 |
+
"loss": 0.0939,
|
| 1444 |
+
"step": 1170
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 15.03,
|
| 1448 |
+
"learning_rate": 0.00026554663991975924,
|
| 1449 |
+
"loss": 0.082,
|
| 1450 |
+
"step": 1175
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 15.09,
|
| 1454 |
+
"learning_rate": 0.00026539618856569707,
|
| 1455 |
+
"loss": 0.0703,
|
| 1456 |
+
"step": 1180
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 15.16,
|
| 1460 |
+
"learning_rate": 0.0002652457372116349,
|
| 1461 |
+
"loss": 0.0695,
|
| 1462 |
+
"step": 1185
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 15.22,
|
| 1466 |
+
"learning_rate": 0.0002650952858575727,
|
| 1467 |
+
"loss": 0.0703,
|
| 1468 |
+
"step": 1190
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 15.28,
|
| 1472 |
+
"learning_rate": 0.0002649448345035105,
|
| 1473 |
+
"loss": 0.0707,
|
| 1474 |
+
"step": 1195
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 15.35,
|
| 1478 |
+
"learning_rate": 0.0002647943831494483,
|
| 1479 |
+
"loss": 0.071,
|
| 1480 |
+
"step": 1200
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 15.41,
|
| 1484 |
+
"learning_rate": 0.00026464393179538614,
|
| 1485 |
+
"loss": 0.0727,
|
| 1486 |
+
"step": 1205
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 15.48,
|
| 1490 |
+
"learning_rate": 0.0002644934804413239,
|
| 1491 |
+
"loss": 0.0771,
|
| 1492 |
+
"step": 1210
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 15.54,
|
| 1496 |
+
"learning_rate": 0.00026434302908726173,
|
| 1497 |
+
"loss": 0.0793,
|
| 1498 |
+
"step": 1215
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 15.6,
|
| 1502 |
+
"learning_rate": 0.00026419257773319956,
|
| 1503 |
+
"loss": 0.076,
|
| 1504 |
+
"step": 1220
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 15.67,
|
| 1508 |
+
"learning_rate": 0.0002640421263791374,
|
| 1509 |
+
"loss": 0.0744,
|
| 1510 |
+
"step": 1225
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 15.73,
|
| 1514 |
+
"learning_rate": 0.0002638916750250752,
|
| 1515 |
+
"loss": 0.0779,
|
| 1516 |
+
"step": 1230
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 15.8,
|
| 1520 |
+
"learning_rate": 0.00026374122367101303,
|
| 1521 |
+
"loss": 0.0785,
|
| 1522 |
+
"step": 1235
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 15.86,
|
| 1526 |
+
"learning_rate": 0.0002635907723169508,
|
| 1527 |
+
"loss": 0.0782,
|
| 1528 |
+
"step": 1240
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 15.92,
|
| 1532 |
+
"learning_rate": 0.00026344032096288863,
|
| 1533 |
+
"loss": 0.079,
|
| 1534 |
+
"step": 1245
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 15.99,
|
| 1538 |
+
"learning_rate": 0.00026328986960882646,
|
| 1539 |
+
"loss": 0.0815,
|
| 1540 |
+
"step": 1250
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 15.99,
|
| 1544 |
+
"eval_loss": 1.4458354711532593,
|
| 1545 |
+
"eval_runtime": 0.3198,
|
| 1546 |
+
"eval_samples_per_second": 31.27,
|
| 1547 |
+
"eval_steps_per_second": 3.127,
|
| 1548 |
+
"step": 1250
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 16.05,
|
| 1552 |
+
"learning_rate": 0.0002631394182547643,
|
| 1553 |
+
"loss": 0.0598,
|
| 1554 |
+
"step": 1255
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 16.12,
|
| 1558 |
+
"learning_rate": 0.0002629889669007021,
|
| 1559 |
+
"loss": 0.0578,
|
| 1560 |
+
"step": 1260
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 16.18,
|
| 1564 |
+
"learning_rate": 0.00026283851554663993,
|
| 1565 |
+
"loss": 0.0596,
|
| 1566 |
+
"step": 1265
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 16.24,
|
| 1570 |
+
"learning_rate": 0.0002626880641925777,
|
| 1571 |
+
"loss": 0.0588,
|
| 1572 |
+
"step": 1270
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 16.31,
|
| 1576 |
+
"learning_rate": 0.00026253761283851553,
|
| 1577 |
+
"loss": 0.0592,
|
| 1578 |
+
"step": 1275
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 16.37,
|
| 1582 |
+
"learning_rate": 0.00026238716148445335,
|
| 1583 |
+
"loss": 0.0614,
|
| 1584 |
+
"step": 1280
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 16.43,
|
| 1588 |
+
"learning_rate": 0.0002622367101303911,
|
| 1589 |
+
"loss": 0.0641,
|
| 1590 |
+
"step": 1285
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 16.5,
|
| 1594 |
+
"learning_rate": 0.00026208625877632895,
|
| 1595 |
+
"loss": 0.0661,
|
| 1596 |
+
"step": 1290
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 16.56,
|
| 1600 |
+
"learning_rate": 0.0002619358074222668,
|
| 1601 |
+
"loss": 0.0685,
|
| 1602 |
+
"step": 1295
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 16.63,
|
| 1606 |
+
"learning_rate": 0.0002617853560682046,
|
| 1607 |
+
"loss": 0.0646,
|
| 1608 |
+
"step": 1300
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 16.69,
|
| 1612 |
+
"learning_rate": 0.0002616349047141424,
|
| 1613 |
+
"loss": 0.0662,
|
| 1614 |
+
"step": 1305
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 16.75,
|
| 1618 |
+
"learning_rate": 0.00026148445336008025,
|
| 1619 |
+
"loss": 0.0699,
|
| 1620 |
+
"step": 1310
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 16.82,
|
| 1624 |
+
"learning_rate": 0.000261334002006018,
|
| 1625 |
+
"loss": 0.069,
|
| 1626 |
+
"step": 1315
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 16.88,
|
| 1630 |
+
"learning_rate": 0.00026118355065195585,
|
| 1631 |
+
"loss": 0.0683,
|
| 1632 |
+
"step": 1320
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 16.95,
|
| 1636 |
+
"learning_rate": 0.00026103309929789367,
|
| 1637 |
+
"loss": 0.0714,
|
| 1638 |
+
"step": 1325
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 17.01,
|
| 1642 |
+
"learning_rate": 0.00026088264794383144,
|
| 1643 |
+
"loss": 0.0682,
|
| 1644 |
+
"step": 1330
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 17.07,
|
| 1648 |
+
"learning_rate": 0.00026073219658976927,
|
| 1649 |
+
"loss": 0.0537,
|
| 1650 |
+
"step": 1335
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 17.14,
|
| 1654 |
+
"learning_rate": 0.0002605817452357071,
|
| 1655 |
+
"loss": 0.055,
|
| 1656 |
+
"step": 1340
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 17.2,
|
| 1660 |
+
"learning_rate": 0.0002604312938816449,
|
| 1661 |
+
"loss": 0.0557,
|
| 1662 |
+
"step": 1345
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 17.27,
|
| 1666 |
+
"learning_rate": 0.00026028084252758274,
|
| 1667 |
+
"loss": 0.0544,
|
| 1668 |
+
"step": 1350
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 17.33,
|
| 1672 |
+
"learning_rate": 0.00026013039117352057,
|
| 1673 |
+
"loss": 0.0545,
|
| 1674 |
+
"step": 1355
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 17.39,
|
| 1678 |
+
"learning_rate": 0.00025997993981945834,
|
| 1679 |
+
"loss": 0.0589,
|
| 1680 |
+
"step": 1360
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 17.46,
|
| 1684 |
+
"learning_rate": 0.00025982948846539616,
|
| 1685 |
+
"loss": 0.0598,
|
| 1686 |
+
"step": 1365
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 17.52,
|
| 1690 |
+
"learning_rate": 0.000259679037111334,
|
| 1691 |
+
"loss": 0.0608,
|
| 1692 |
+
"step": 1370
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 17.59,
|
| 1696 |
+
"learning_rate": 0.00025952858575727176,
|
| 1697 |
+
"loss": 0.0628,
|
| 1698 |
+
"step": 1375
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 17.65,
|
| 1702 |
+
"learning_rate": 0.0002593781344032096,
|
| 1703 |
+
"loss": 0.0603,
|
| 1704 |
+
"step": 1380
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 17.71,
|
| 1708 |
+
"learning_rate": 0.0002592276830491474,
|
| 1709 |
+
"loss": 0.065,
|
| 1710 |
+
"step": 1385
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 17.78,
|
| 1714 |
+
"learning_rate": 0.00025907723169508524,
|
| 1715 |
+
"loss": 0.0648,
|
| 1716 |
+
"step": 1390
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 17.84,
|
| 1720 |
+
"learning_rate": 0.00025892678034102306,
|
| 1721 |
+
"loss": 0.0649,
|
| 1722 |
+
"step": 1395
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 17.91,
|
| 1726 |
+
"learning_rate": 0.0002587763289869609,
|
| 1727 |
+
"loss": 0.0647,
|
| 1728 |
+
"step": 1400
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 17.97,
|
| 1732 |
+
"learning_rate": 0.00025862587763289866,
|
| 1733 |
+
"loss": 0.0687,
|
| 1734 |
+
"step": 1405
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 18.03,
|
| 1738 |
+
"learning_rate": 0.0002584754262788365,
|
| 1739 |
+
"loss": 0.0592,
|
| 1740 |
+
"step": 1410
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 18.1,
|
| 1744 |
+
"learning_rate": 0.0002583249749247743,
|
| 1745 |
+
"loss": 0.047,
|
| 1746 |
+
"step": 1415
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 18.16,
|
| 1750 |
+
"learning_rate": 0.0002581745235707121,
|
| 1751 |
+
"loss": 0.0486,
|
| 1752 |
+
"step": 1420
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 18.23,
|
| 1756 |
+
"learning_rate": 0.0002580240722166499,
|
| 1757 |
+
"loss": 0.0526,
|
| 1758 |
+
"step": 1425
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 18.29,
|
| 1762 |
+
"learning_rate": 0.00025787362086258773,
|
| 1763 |
+
"loss": 0.05,
|
| 1764 |
+
"step": 1430
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 18.35,
|
| 1768 |
+
"learning_rate": 0.00025772316950852556,
|
| 1769 |
+
"loss": 0.0646,
|
| 1770 |
+
"step": 1435
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 18.42,
|
| 1774 |
+
"learning_rate": 0.0002575727181544634,
|
| 1775 |
+
"loss": 0.0557,
|
| 1776 |
+
"step": 1440
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 18.48,
|
| 1780 |
+
"learning_rate": 0.0002574222668004012,
|
| 1781 |
+
"loss": 0.0534,
|
| 1782 |
+
"step": 1445
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 18.55,
|
| 1786 |
+
"learning_rate": 0.000257271815446339,
|
| 1787 |
+
"loss": 0.056,
|
| 1788 |
+
"step": 1450
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 18.61,
|
| 1792 |
+
"learning_rate": 0.0002571213640922768,
|
| 1793 |
+
"loss": 0.0553,
|
| 1794 |
+
"step": 1455
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 18.67,
|
| 1798 |
+
"learning_rate": 0.00025697091273821463,
|
| 1799 |
+
"loss": 0.057,
|
| 1800 |
+
"step": 1460
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 18.74,
|
| 1804 |
+
"learning_rate": 0.0002568204613841524,
|
| 1805 |
+
"loss": 0.0545,
|
| 1806 |
+
"step": 1465
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 18.8,
|
| 1810 |
+
"learning_rate": 0.0002566700100300902,
|
| 1811 |
+
"loss": 0.0557,
|
| 1812 |
+
"step": 1470
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 18.86,
|
| 1816 |
+
"learning_rate": 0.00025651955867602805,
|
| 1817 |
+
"loss": 0.0587,
|
| 1818 |
+
"step": 1475
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 18.93,
|
| 1822 |
+
"learning_rate": 0.0002563691073219659,
|
| 1823 |
+
"loss": 0.0574,
|
| 1824 |
+
"step": 1480
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 18.99,
|
| 1828 |
+
"learning_rate": 0.0002562186559679037,
|
| 1829 |
+
"loss": 0.0625,
|
| 1830 |
+
"step": 1485
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 19.06,
|
| 1834 |
+
"learning_rate": 0.0002560682046138415,
|
| 1835 |
+
"loss": 0.1149,
|
| 1836 |
+
"step": 1490
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 19.12,
|
| 1840 |
+
"learning_rate": 0.0002559177532597793,
|
| 1841 |
+
"loss": 0.059,
|
| 1842 |
+
"step": 1495
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 19.18,
|
| 1846 |
+
"learning_rate": 0.0002557673019057171,
|
| 1847 |
+
"loss": 0.0472,
|
| 1848 |
+
"step": 1500
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 19.18,
|
| 1852 |
+
"eval_loss": 1.7803338766098022,
|
| 1853 |
+
"eval_runtime": 0.3184,
|
| 1854 |
+
"eval_samples_per_second": 31.405,
|
| 1855 |
+
"eval_steps_per_second": 3.14,
|
| 1856 |
+
"step": 1500
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 19.25,
|
| 1860 |
+
"learning_rate": 0.00025561685055165495,
|
| 1861 |
+
"loss": 0.0432,
|
| 1862 |
+
"step": 1505
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 19.31,
|
| 1866 |
+
"learning_rate": 0.0002554663991975927,
|
| 1867 |
+
"loss": 0.0415,
|
| 1868 |
+
"step": 1510
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 19.38,
|
| 1872 |
+
"learning_rate": 0.00025531594784353054,
|
| 1873 |
+
"loss": 0.047,
|
| 1874 |
+
"step": 1515
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 19.44,
|
| 1878 |
+
"learning_rate": 0.00025516549648946837,
|
| 1879 |
+
"loss": 0.0473,
|
| 1880 |
+
"step": 1520
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 19.5,
|
| 1884 |
+
"learning_rate": 0.0002550150451354062,
|
| 1885 |
+
"loss": 0.0482,
|
| 1886 |
+
"step": 1525
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 19.57,
|
| 1890 |
+
"learning_rate": 0.000254864593781344,
|
| 1891 |
+
"loss": 0.0493,
|
| 1892 |
+
"step": 1530
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 19.63,
|
| 1896 |
+
"learning_rate": 0.00025471414242728184,
|
| 1897 |
+
"loss": 0.0513,
|
| 1898 |
+
"step": 1535
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 19.7,
|
| 1902 |
+
"learning_rate": 0.0002545636910732196,
|
| 1903 |
+
"loss": 0.0514,
|
| 1904 |
+
"step": 1540
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 19.76,
|
| 1908 |
+
"learning_rate": 0.00025441323971915744,
|
| 1909 |
+
"loss": 0.0509,
|
| 1910 |
+
"step": 1545
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 19.82,
|
| 1914 |
+
"learning_rate": 0.00025426278836509526,
|
| 1915 |
+
"loss": 0.0512,
|
| 1916 |
+
"step": 1550
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 19.89,
|
| 1920 |
+
"learning_rate": 0.0002541123370110331,
|
| 1921 |
+
"loss": 0.0535,
|
| 1922 |
+
"step": 1555
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 19.95,
|
| 1926 |
+
"learning_rate": 0.00025396188565697086,
|
| 1927 |
+
"loss": 0.0522,
|
| 1928 |
+
"step": 1560
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 20.02,
|
| 1932 |
+
"learning_rate": 0.0002538114343029087,
|
| 1933 |
+
"loss": 0.0491,
|
| 1934 |
+
"step": 1565
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 20.08,
|
| 1938 |
+
"learning_rate": 0.0002536609829488465,
|
| 1939 |
+
"loss": 0.0364,
|
| 1940 |
+
"step": 1570
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 20.14,
|
| 1944 |
+
"learning_rate": 0.00025351053159478434,
|
| 1945 |
+
"loss": 0.0398,
|
| 1946 |
+
"step": 1575
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 20.21,
|
| 1950 |
+
"learning_rate": 0.00025336008024072216,
|
| 1951 |
+
"loss": 0.0393,
|
| 1952 |
+
"step": 1580
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 20.27,
|
| 1956 |
+
"learning_rate": 0.00025320962888665993,
|
| 1957 |
+
"loss": 0.0418,
|
| 1958 |
+
"step": 1585
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 20.34,
|
| 1962 |
+
"learning_rate": 0.00025305917753259776,
|
| 1963 |
+
"loss": 0.0436,
|
| 1964 |
+
"step": 1590
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 20.4,
|
| 1968 |
+
"learning_rate": 0.0002529087261785356,
|
| 1969 |
+
"loss": 0.0439,
|
| 1970 |
+
"step": 1595
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 20.46,
|
| 1974 |
+
"learning_rate": 0.0002527582748244734,
|
| 1975 |
+
"loss": 0.0458,
|
| 1976 |
+
"step": 1600
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 20.53,
|
| 1980 |
+
"learning_rate": 0.00025260782347041123,
|
| 1981 |
+
"loss": 0.0443,
|
| 1982 |
+
"step": 1605
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 20.59,
|
| 1986 |
+
"learning_rate": 0.00025245737211634906,
|
| 1987 |
+
"loss": 0.0443,
|
| 1988 |
+
"step": 1610
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 20.66,
|
| 1992 |
+
"learning_rate": 0.00025230692076228683,
|
| 1993 |
+
"loss": 0.0439,
|
| 1994 |
+
"step": 1615
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 20.72,
|
| 1998 |
+
"learning_rate": 0.00025215646940822465,
|
| 1999 |
+
"loss": 0.0431,
|
| 2000 |
+
"step": 1620
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 20.78,
|
| 2004 |
+
"learning_rate": 0.0002520060180541625,
|
| 2005 |
+
"loss": 0.0451,
|
| 2006 |
+
"step": 1625
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 20.85,
|
| 2010 |
+
"learning_rate": 0.00025185556670010025,
|
| 2011 |
+
"loss": 0.0474,
|
| 2012 |
+
"step": 1630
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 20.91,
|
| 2016 |
+
"learning_rate": 0.0002517051153460381,
|
| 2017 |
+
"loss": 0.0479,
|
| 2018 |
+
"step": 1635
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 20.98,
|
| 2022 |
+
"learning_rate": 0.0002515546639919759,
|
| 2023 |
+
"loss": 0.0483,
|
| 2024 |
+
"step": 1640
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 21.04,
|
| 2028 |
+
"learning_rate": 0.0002514042126379137,
|
| 2029 |
+
"loss": 0.0368,
|
| 2030 |
+
"step": 1645
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 21.1,
|
| 2034 |
+
"learning_rate": 0.00025125376128385155,
|
| 2035 |
+
"loss": 0.032,
|
| 2036 |
+
"step": 1650
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 21.17,
|
| 2040 |
+
"learning_rate": 0.0002511033099297894,
|
| 2041 |
+
"loss": 0.0343,
|
| 2042 |
+
"step": 1655
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 21.23,
|
| 2046 |
+
"learning_rate": 0.00025095285857572715,
|
| 2047 |
+
"loss": 0.0366,
|
| 2048 |
+
"step": 1660
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 21.29,
|
| 2052 |
+
"learning_rate": 0.000250802407221665,
|
| 2053 |
+
"loss": 0.0377,
|
| 2054 |
+
"step": 1665
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 21.36,
|
| 2058 |
+
"learning_rate": 0.0002506519558676028,
|
| 2059 |
+
"loss": 0.0378,
|
| 2060 |
+
"step": 1670
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 21.42,
|
| 2064 |
+
"learning_rate": 0.00025050150451354057,
|
| 2065 |
+
"loss": 0.0374,
|
| 2066 |
+
"step": 1675
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 21.49,
|
| 2070 |
+
"learning_rate": 0.0002503510531594784,
|
| 2071 |
+
"loss": 0.0421,
|
| 2072 |
+
"step": 1680
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 21.55,
|
| 2076 |
+
"learning_rate": 0.0002502006018054162,
|
| 2077 |
+
"loss": 0.037,
|
| 2078 |
+
"step": 1685
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 21.61,
|
| 2082 |
+
"learning_rate": 0.00025005015045135405,
|
| 2083 |
+
"loss": 0.0392,
|
| 2084 |
+
"step": 1690
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 21.68,
|
| 2088 |
+
"learning_rate": 0.00024989969909729187,
|
| 2089 |
+
"loss": 0.0421,
|
| 2090 |
+
"step": 1695
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 21.74,
|
| 2094 |
+
"learning_rate": 0.0002497492477432297,
|
| 2095 |
+
"loss": 0.0425,
|
| 2096 |
+
"step": 1700
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 21.81,
|
| 2100 |
+
"learning_rate": 0.00024959879638916747,
|
| 2101 |
+
"loss": 0.0442,
|
| 2102 |
+
"step": 1705
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 21.87,
|
| 2106 |
+
"learning_rate": 0.0002494483450351053,
|
| 2107 |
+
"loss": 0.042,
|
| 2108 |
+
"step": 1710
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 21.93,
|
| 2112 |
+
"learning_rate": 0.0002492978936810431,
|
| 2113 |
+
"loss": 0.0425,
|
| 2114 |
+
"step": 1715
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 22.0,
|
| 2118 |
+
"learning_rate": 0.0002491474423269809,
|
| 2119 |
+
"loss": 0.0432,
|
| 2120 |
+
"step": 1720
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 22.06,
|
| 2124 |
+
"learning_rate": 0.0002489969909729187,
|
| 2125 |
+
"loss": 0.029,
|
| 2126 |
+
"step": 1725
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 22.13,
|
| 2130 |
+
"learning_rate": 0.00024884653961885654,
|
| 2131 |
+
"loss": 0.0303,
|
| 2132 |
+
"step": 1730
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 22.19,
|
| 2136 |
+
"learning_rate": 0.00024869608826479436,
|
| 2137 |
+
"loss": 0.0324,
|
| 2138 |
+
"step": 1735
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 22.25,
|
| 2142 |
+
"learning_rate": 0.0002485456369107322,
|
| 2143 |
+
"loss": 0.0319,
|
| 2144 |
+
"step": 1740
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 22.32,
|
| 2148 |
+
"learning_rate": 0.00024839518555667,
|
| 2149 |
+
"loss": 0.0325,
|
| 2150 |
+
"step": 1745
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 22.38,
|
| 2154 |
+
"learning_rate": 0.0002482447342026078,
|
| 2155 |
+
"loss": 0.033,
|
| 2156 |
+
"step": 1750
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 22.38,
|
| 2160 |
+
"eval_loss": 1.8502811193466187,
|
| 2161 |
+
"eval_runtime": 0.319,
|
| 2162 |
+
"eval_samples_per_second": 31.348,
|
| 2163 |
+
"eval_steps_per_second": 3.135,
|
| 2164 |
+
"step": 1750
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 22.45,
|
| 2168 |
+
"learning_rate": 0.0002480942828485456,
|
| 2169 |
+
"loss": 0.032,
|
| 2170 |
+
"step": 1755
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 22.51,
|
| 2174 |
+
"learning_rate": 0.00024794383149448344,
|
| 2175 |
+
"loss": 0.0352,
|
| 2176 |
+
"step": 1760
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 22.57,
|
| 2180 |
+
"learning_rate": 0.0002477933801404212,
|
| 2181 |
+
"loss": 0.0367,
|
| 2182 |
+
"step": 1765
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 22.64,
|
| 2186 |
+
"learning_rate": 0.00024764292878635903,
|
| 2187 |
+
"loss": 0.0382,
|
| 2188 |
+
"step": 1770
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 22.7,
|
| 2192 |
+
"learning_rate": 0.00024749247743229686,
|
| 2193 |
+
"loss": 0.039,
|
| 2194 |
+
"step": 1775
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 22.77,
|
| 2198 |
+
"learning_rate": 0.0002473420260782347,
|
| 2199 |
+
"loss": 0.0387,
|
| 2200 |
+
"step": 1780
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 22.83,
|
| 2204 |
+
"learning_rate": 0.0002471915747241725,
|
| 2205 |
+
"loss": 0.0411,
|
| 2206 |
+
"step": 1785
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 22.89,
|
| 2210 |
+
"learning_rate": 0.00024704112337011033,
|
| 2211 |
+
"loss": 0.04,
|
| 2212 |
+
"step": 1790
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 22.96,
|
| 2216 |
+
"learning_rate": 0.0002468906720160481,
|
| 2217 |
+
"loss": 0.0405,
|
| 2218 |
+
"step": 1795
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 23.02,
|
| 2222 |
+
"learning_rate": 0.00024674022066198593,
|
| 2223 |
+
"loss": 0.0347,
|
| 2224 |
+
"step": 1800
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 23.09,
|
| 2228 |
+
"learning_rate": 0.00024658976930792375,
|
| 2229 |
+
"loss": 0.0295,
|
| 2230 |
+
"step": 1805
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 23.15,
|
| 2234 |
+
"learning_rate": 0.0002464393179538615,
|
| 2235 |
+
"loss": 0.0298,
|
| 2236 |
+
"step": 1810
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 23.21,
|
| 2240 |
+
"learning_rate": 0.00024628886659979935,
|
| 2241 |
+
"loss": 0.0314,
|
| 2242 |
+
"step": 1815
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 23.28,
|
| 2246 |
+
"learning_rate": 0.0002461384152457372,
|
| 2247 |
+
"loss": 0.0334,
|
| 2248 |
+
"step": 1820
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 23.34,
|
| 2252 |
+
"learning_rate": 0.000245987963891675,
|
| 2253 |
+
"loss": 0.035,
|
| 2254 |
+
"step": 1825
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 23.41,
|
| 2258 |
+
"learning_rate": 0.0002458375125376128,
|
| 2259 |
+
"loss": 0.033,
|
| 2260 |
+
"step": 1830
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 23.47,
|
| 2264 |
+
"learning_rate": 0.00024568706118355065,
|
| 2265 |
+
"loss": 0.0336,
|
| 2266 |
+
"step": 1835
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 23.53,
|
| 2270 |
+
"learning_rate": 0.0002455366098294884,
|
| 2271 |
+
"loss": 0.0338,
|
| 2272 |
+
"step": 1840
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 23.6,
|
| 2276 |
+
"learning_rate": 0.00024538615847542625,
|
| 2277 |
+
"loss": 0.0371,
|
| 2278 |
+
"step": 1845
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 23.66,
|
| 2282 |
+
"learning_rate": 0.0002452357071213641,
|
| 2283 |
+
"loss": 0.0372,
|
| 2284 |
+
"step": 1850
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 23.73,
|
| 2288 |
+
"learning_rate": 0.00024508525576730184,
|
| 2289 |
+
"loss": 0.0365,
|
| 2290 |
+
"step": 1855
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 23.79,
|
| 2294 |
+
"learning_rate": 0.00024493480441323967,
|
| 2295 |
+
"loss": 0.0368,
|
| 2296 |
+
"step": 1860
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 23.85,
|
| 2300 |
+
"learning_rate": 0.0002447843530591775,
|
| 2301 |
+
"loss": 0.0347,
|
| 2302 |
+
"step": 1865
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 23.92,
|
| 2306 |
+
"learning_rate": 0.0002446339017051153,
|
| 2307 |
+
"loss": 0.0362,
|
| 2308 |
+
"step": 1870
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 23.98,
|
| 2312 |
+
"learning_rate": 0.00024448345035105315,
|
| 2313 |
+
"loss": 0.0379,
|
| 2314 |
+
"step": 1875
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 24.04,
|
| 2318 |
+
"learning_rate": 0.00024433299899699097,
|
| 2319 |
+
"loss": 0.0398,
|
| 2320 |
+
"step": 1880
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 24.11,
|
| 2324 |
+
"learning_rate": 0.00024418254764292874,
|
| 2325 |
+
"loss": 0.0288,
|
| 2326 |
+
"step": 1885
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 24.17,
|
| 2330 |
+
"learning_rate": 0.00024403209628886657,
|
| 2331 |
+
"loss": 0.0305,
|
| 2332 |
+
"step": 1890
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 24.24,
|
| 2336 |
+
"learning_rate": 0.0002438816449348044,
|
| 2337 |
+
"loss": 0.0304,
|
| 2338 |
+
"step": 1895
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 24.3,
|
| 2342 |
+
"learning_rate": 0.0002437311935807422,
|
| 2343 |
+
"loss": 0.0282,
|
| 2344 |
+
"step": 1900
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 24.36,
|
| 2348 |
+
"learning_rate": 0.00024358074222668002,
|
| 2349 |
+
"loss": 0.0289,
|
| 2350 |
+
"step": 1905
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 24.43,
|
| 2354 |
+
"learning_rate": 0.00024343029087261784,
|
| 2355 |
+
"loss": 0.0302,
|
| 2356 |
+
"step": 1910
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 24.49,
|
| 2360 |
+
"learning_rate": 0.00024327983951855564,
|
| 2361 |
+
"loss": 0.0258,
|
| 2362 |
+
"step": 1915
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 24.56,
|
| 2366 |
+
"learning_rate": 0.00024312938816449346,
|
| 2367 |
+
"loss": 0.0266,
|
| 2368 |
+
"step": 1920
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 24.62,
|
| 2372 |
+
"learning_rate": 0.0002429789368104313,
|
| 2373 |
+
"loss": 0.0278,
|
| 2374 |
+
"step": 1925
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 24.68,
|
| 2378 |
+
"learning_rate": 0.00024282848545636906,
|
| 2379 |
+
"loss": 0.0316,
|
| 2380 |
+
"step": 1930
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 24.75,
|
| 2384 |
+
"learning_rate": 0.00024267803410230689,
|
| 2385 |
+
"loss": 0.0328,
|
| 2386 |
+
"step": 1935
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 24.81,
|
| 2390 |
+
"learning_rate": 0.0002425275827482447,
|
| 2391 |
+
"loss": 0.0336,
|
| 2392 |
+
"step": 1940
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 24.88,
|
| 2396 |
+
"learning_rate": 0.0002423771313941825,
|
| 2397 |
+
"loss": 0.0311,
|
| 2398 |
+
"step": 1945
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 24.94,
|
| 2402 |
+
"learning_rate": 0.00024222668004012033,
|
| 2403 |
+
"loss": 0.0324,
|
| 2404 |
+
"step": 1950
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 25.0,
|
| 2408 |
+
"learning_rate": 0.00024207622868605816,
|
| 2409 |
+
"loss": 0.0339,
|
| 2410 |
+
"step": 1955
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 25.07,
|
| 2414 |
+
"learning_rate": 0.00024192577733199596,
|
| 2415 |
+
"loss": 0.0235,
|
| 2416 |
+
"step": 1960
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 25.13,
|
| 2420 |
+
"learning_rate": 0.00024177532597793378,
|
| 2421 |
+
"loss": 0.029,
|
| 2422 |
+
"step": 1965
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 25.2,
|
| 2426 |
+
"learning_rate": 0.0002416248746238716,
|
| 2427 |
+
"loss": 0.0281,
|
| 2428 |
+
"step": 1970
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 25.26,
|
| 2432 |
+
"learning_rate": 0.00024147442326980938,
|
| 2433 |
+
"loss": 0.0276,
|
| 2434 |
+
"step": 1975
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 25.32,
|
| 2438 |
+
"learning_rate": 0.0002413239719157472,
|
| 2439 |
+
"loss": 0.0282,
|
| 2440 |
+
"step": 1980
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 25.39,
|
| 2444 |
+
"learning_rate": 0.00024117352056168503,
|
| 2445 |
+
"loss": 0.0308,
|
| 2446 |
+
"step": 1985
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 25.45,
|
| 2450 |
+
"learning_rate": 0.00024102306920762283,
|
| 2451 |
+
"loss": 0.0313,
|
| 2452 |
+
"step": 1990
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 25.52,
|
| 2456 |
+
"learning_rate": 0.00024087261785356065,
|
| 2457 |
+
"loss": 0.0311,
|
| 2458 |
+
"step": 1995
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 25.58,
|
| 2462 |
+
"learning_rate": 0.00024072216649949848,
|
| 2463 |
+
"loss": 0.0307,
|
| 2464 |
+
"step": 2000
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 25.58,
|
| 2468 |
+
"eval_loss": 1.802398920059204,
|
| 2469 |
+
"eval_runtime": 0.3194,
|
| 2470 |
+
"eval_samples_per_second": 31.309,
|
| 2471 |
+
"eval_steps_per_second": 3.131,
|
| 2472 |
+
"step": 2000
|
| 2473 |
+
}
|
| 2474 |
+
],
|
| 2475 |
+
"max_steps": 10000,
|
| 2476 |
+
"num_train_epochs": 129,
|
| 2477 |
+
"total_flos": 4.2458758196763034e+17,
|
| 2478 |
+
"trial_name": null,
|
| 2479 |
+
"trial_params": null
|
| 2480 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60971362b90305a243e89870fdbd1e58c6051b6dc7d4bf488197c29f83972e9a
|
| 3 |
+
size 4091
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|