upload ckpt files
Browse files- config.json +75 -0
- merges.txt +0 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +65 -0
- trainer_state.json +1500 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,75 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "sshleifer/distilbart-cnn-12-3",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"activation_dropout": 0.0,
|
| 5 |
+
"activation_function": "gelu",
|
| 6 |
+
"add_bias_logits": false,
|
| 7 |
+
"add_final_layer_norm": false,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"NFBart"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"bos_token_id": 0,
|
| 13 |
+
"classif_dropout": 0.0,
|
| 14 |
+
"classifier_dropout": 0.0,
|
| 15 |
+
"d_model": 1024,
|
| 16 |
+
"decoder_attention_heads": 16,
|
| 17 |
+
"decoder_ffn_dim": 4096,
|
| 18 |
+
"decoder_layerdrop": 0.0,
|
| 19 |
+
"decoder_layers": 3,
|
| 20 |
+
"decoder_start_token_id": 2,
|
| 21 |
+
"dropout": 0.1,
|
| 22 |
+
"early_stopping": true,
|
| 23 |
+
"encoder_attention_heads": 16,
|
| 24 |
+
"encoder_ffn_dim": 4096,
|
| 25 |
+
"encoder_layerdrop": 0.0,
|
| 26 |
+
"encoder_layers": 12,
|
| 27 |
+
"eos_token_id": 2,
|
| 28 |
+
"extra_pos_embeddings": 2,
|
| 29 |
+
"force_bos_token_to_be_generated": true,
|
| 30 |
+
"forced_bos_token_id": 0,
|
| 31 |
+
"forced_eos_token_id": 2,
|
| 32 |
+
"gradient_checkpointing": false,
|
| 33 |
+
"id2label": {
|
| 34 |
+
"0": "LABEL_0",
|
| 35 |
+
"1": "LABEL_1",
|
| 36 |
+
"2": "LABEL_2"
|
| 37 |
+
},
|
| 38 |
+
"init_std": 0.02,
|
| 39 |
+
"is_encoder_decoder": true,
|
| 40 |
+
"label2id": {
|
| 41 |
+
"LABEL_0": 0,
|
| 42 |
+
"LABEL_1": 1,
|
| 43 |
+
"LABEL_2": 2
|
| 44 |
+
},
|
| 45 |
+
"length_penalty": 2.0,
|
| 46 |
+
"max_length": 142,
|
| 47 |
+
"max_position_embeddings": 1024,
|
| 48 |
+
"min_length": 56,
|
| 49 |
+
"model_type": "bart",
|
| 50 |
+
"no_repeat_ngram_size": 3,
|
| 51 |
+
"normalize_before": false,
|
| 52 |
+
"normalize_embedding": true,
|
| 53 |
+
"num_beams": 4,
|
| 54 |
+
"num_hidden_layers": 12,
|
| 55 |
+
"output_past": true,
|
| 56 |
+
"pad_token_id": 1,
|
| 57 |
+
"prefix": " ",
|
| 58 |
+
"save_step": 14,
|
| 59 |
+
"scale_embedding": false,
|
| 60 |
+
"static_position_embeddings": false,
|
| 61 |
+
"task_specific_params": {
|
| 62 |
+
"summarization": {
|
| 63 |
+
"early_stopping": true,
|
| 64 |
+
"length_penalty": 2.0,
|
| 65 |
+
"max_length": 142,
|
| 66 |
+
"min_length": 56,
|
| 67 |
+
"no_repeat_ngram_size": 3,
|
| 68 |
+
"num_beams": 4
|
| 69 |
+
}
|
| 70 |
+
},
|
| 71 |
+
"torch_dtype": "float32",
|
| 72 |
+
"transformers_version": "4.25.1",
|
| 73 |
+
"use_cache": true,
|
| 74 |
+
"vocab_size": 50265
|
| 75 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e72e0975dfbe0455970b9b0a0b404fa14de3dfedba3ae44792e9461cbae2dd9
|
| 3 |
+
size 1051630133
|
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": true,
|
| 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 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"cls_token": {
|
| 12 |
+
"__type": "AddedToken",
|
| 13 |
+
"content": "<s>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false
|
| 18 |
+
},
|
| 19 |
+
"eos_token": {
|
| 20 |
+
"__type": "AddedToken",
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false
|
| 26 |
+
},
|
| 27 |
+
"errors": "replace",
|
| 28 |
+
"mask_token": {
|
| 29 |
+
"__type": "AddedToken",
|
| 30 |
+
"content": "<mask>",
|
| 31 |
+
"lstrip": true,
|
| 32 |
+
"normalized": true,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false
|
| 35 |
+
},
|
| 36 |
+
"model_max_length": 1024,
|
| 37 |
+
"name_or_path": "sshleifer/distilbart-cnn-12-3",
|
| 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 |
+
"special_tokens_map_file": "/home/yang6367/.cache/huggingface/hub/models--sshleifer--distilbart-cnn-12-3/snapshots/e3a8f0dafad8b99df3209ff7cf33e1f8402619e1/special_tokens_map.json",
|
| 55 |
+
"tokenizer_class": "BartTokenizer",
|
| 56 |
+
"trim_offsets": true,
|
| 57 |
+
"unk_token": {
|
| 58 |
+
"__type": "AddedToken",
|
| 59 |
+
"content": "<unk>",
|
| 60 |
+
"lstrip": false,
|
| 61 |
+
"normalized": true,
|
| 62 |
+
"rstrip": false,
|
| 63 |
+
"single_word": false
|
| 64 |
+
}
|
| 65 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1500 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 5.177720546722412,
|
| 3 |
+
"best_model_checkpoint": "../checkpoints/nf-distilbart_12_3-rqnsf-lagtrain-augmented/checkpoint-212000",
|
| 4 |
+
"epoch": 2.953510079549729,
|
| 5 |
+
"global_step": 212000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"gate_score": 0.4928,
|
| 13 |
+
"learning_rate": 0.0,
|
| 14 |
+
"loss": 573.0775,
|
| 15 |
+
"nf_loss": 569.8872,
|
| 16 |
+
"ppl": 32.0588,
|
| 17 |
+
"step": 2000
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06,
|
| 21 |
+
"gate_score": 0.4835,
|
| 22 |
+
"learning_rate": 0.0,
|
| 23 |
+
"loss": 341.9544,
|
| 24 |
+
"nf_loss": 339.8841,
|
| 25 |
+
"ppl": 9.249,
|
| 26 |
+
"step": 4000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.06,
|
| 30 |
+
"eval_loss": 256.18603515625,
|
| 31 |
+
"eval_nf_loss": 253.6398468017578,
|
| 32 |
+
"eval_perplexity": 13.794677734375,
|
| 33 |
+
"eval_runtime": 646.8656,
|
| 34 |
+
"eval_samples_per_second": 20.666,
|
| 35 |
+
"eval_steps_per_second": 2.583,
|
| 36 |
+
"step": 4000
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.08,
|
| 40 |
+
"gate_score": 0.468,
|
| 41 |
+
"learning_rate": 0.0,
|
| 42 |
+
"loss": 216.1715,
|
| 43 |
+
"nf_loss": 214.5228,
|
| 44 |
+
"ppl": 5.9033,
|
| 45 |
+
"step": 6000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11,
|
| 49 |
+
"gate_score": 0.4488,
|
| 50 |
+
"learning_rate": 0.0,
|
| 51 |
+
"loss": 143.1113,
|
| 52 |
+
"nf_loss": 141.682,
|
| 53 |
+
"ppl": 4.666,
|
| 54 |
+
"step": 8000
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.11,
|
| 58 |
+
"eval_loss": 111.98719024658203,
|
| 59 |
+
"eval_nf_loss": 109.95372009277344,
|
| 60 |
+
"eval_perplexity": 8.187104225158691,
|
| 61 |
+
"eval_runtime": 836.6322,
|
| 62 |
+
"eval_samples_per_second": 15.978,
|
| 63 |
+
"eval_steps_per_second": 1.997,
|
| 64 |
+
"step": 8000
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.14,
|
| 68 |
+
"gate_score": 0.4331,
|
| 69 |
+
"learning_rate": 0.0,
|
| 70 |
+
"loss": 97.2206,
|
| 71 |
+
"nf_loss": 95.8896,
|
| 72 |
+
"ppl": 4.2098,
|
| 73 |
+
"step": 10000
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.17,
|
| 77 |
+
"gate_score": 0.4146,
|
| 78 |
+
"learning_rate": 0.0,
|
| 79 |
+
"loss": 72.4608,
|
| 80 |
+
"nf_loss": 71.1644,
|
| 81 |
+
"ppl": 4.0666,
|
| 82 |
+
"step": 12000
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.17,
|
| 86 |
+
"eval_loss": 60.48736572265625,
|
| 87 |
+
"eval_nf_loss": 58.55850601196289,
|
| 88 |
+
"eval_perplexity": 7.357144355773926,
|
| 89 |
+
"eval_runtime": 847.0188,
|
| 90 |
+
"eval_samples_per_second": 15.782,
|
| 91 |
+
"eval_steps_per_second": 1.973,
|
| 92 |
+
"step": 12000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.2,
|
| 96 |
+
"gate_score": 0.3943,
|
| 97 |
+
"learning_rate": 0.0,
|
| 98 |
+
"loss": 58.0258,
|
| 99 |
+
"nf_loss": 56.7432,
|
| 100 |
+
"ppl": 4.0134,
|
| 101 |
+
"step": 14000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.22,
|
| 105 |
+
"gate_score": 0.373,
|
| 106 |
+
"learning_rate": 0.0,
|
| 107 |
+
"loss": 45.4137,
|
| 108 |
+
"nf_loss": 44.1607,
|
| 109 |
+
"ppl": 3.8803,
|
| 110 |
+
"step": 16000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"eval_loss": 35.780086517333984,
|
| 115 |
+
"eval_nf_loss": 33.906742095947266,
|
| 116 |
+
"eval_perplexity": 6.943728923797607,
|
| 117 |
+
"eval_runtime": 829.3361,
|
| 118 |
+
"eval_samples_per_second": 16.119,
|
| 119 |
+
"eval_steps_per_second": 2.015,
|
| 120 |
+
"step": 16000
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.25,
|
| 124 |
+
"gate_score": 0.3552,
|
| 125 |
+
"learning_rate": 0.0,
|
| 126 |
+
"loss": 33.838,
|
| 127 |
+
"nf_loss": 32.5964,
|
| 128 |
+
"ppl": 3.8548,
|
| 129 |
+
"step": 18000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.28,
|
| 133 |
+
"gate_score": 0.3346,
|
| 134 |
+
"learning_rate": 0.0,
|
| 135 |
+
"loss": 28.827,
|
| 136 |
+
"nf_loss": 27.5837,
|
| 137 |
+
"ppl": 3.8361,
|
| 138 |
+
"step": 20000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.28,
|
| 142 |
+
"eval_loss": 24.475879669189453,
|
| 143 |
+
"eval_nf_loss": 22.578872680664062,
|
| 144 |
+
"eval_perplexity": 7.1263651847839355,
|
| 145 |
+
"eval_runtime": 878.8848,
|
| 146 |
+
"eval_samples_per_second": 15.21,
|
| 147 |
+
"eval_steps_per_second": 1.901,
|
| 148 |
+
"step": 20000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.31,
|
| 152 |
+
"gate_score": 0.3166,
|
| 153 |
+
"learning_rate": 0.0,
|
| 154 |
+
"loss": 25.3406,
|
| 155 |
+
"nf_loss": 24.0903,
|
| 156 |
+
"ppl": 3.8804,
|
| 157 |
+
"step": 22000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.33,
|
| 161 |
+
"gate_score": 0.2976,
|
| 162 |
+
"learning_rate": 0.0,
|
| 163 |
+
"loss": 22.6997,
|
| 164 |
+
"nf_loss": 21.451,
|
| 165 |
+
"ppl": 3.8523,
|
| 166 |
+
"step": 24000
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.33,
|
| 170 |
+
"eval_loss": 20.067764282226562,
|
| 171 |
+
"eval_nf_loss": 18.16155242919922,
|
| 172 |
+
"eval_perplexity": 7.192458152770996,
|
| 173 |
+
"eval_runtime": 909.546,
|
| 174 |
+
"eval_samples_per_second": 14.697,
|
| 175 |
+
"eval_steps_per_second": 1.837,
|
| 176 |
+
"step": 24000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.36,
|
| 180 |
+
"gate_score": 0.2793,
|
| 181 |
+
"learning_rate": 0.0,
|
| 182 |
+
"loss": 20.7264,
|
| 183 |
+
"nf_loss": 19.4539,
|
| 184 |
+
"ppl": 3.9698,
|
| 185 |
+
"step": 26000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.39,
|
| 189 |
+
"gate_score": 0.2666,
|
| 190 |
+
"learning_rate": 0.0,
|
| 191 |
+
"loss": 19.1146,
|
| 192 |
+
"nf_loss": 17.8629,
|
| 193 |
+
"ppl": 3.8907,
|
| 194 |
+
"step": 28000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.39,
|
| 198 |
+
"eval_loss": 16.49034309387207,
|
| 199 |
+
"eval_nf_loss": 14.588922500610352,
|
| 200 |
+
"eval_perplexity": 7.155718803405762,
|
| 201 |
+
"eval_runtime": 909.9257,
|
| 202 |
+
"eval_samples_per_second": 14.691,
|
| 203 |
+
"eval_steps_per_second": 1.836,
|
| 204 |
+
"step": 28000
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.42,
|
| 208 |
+
"gate_score": 0.2556,
|
| 209 |
+
"learning_rate": 0.0,
|
| 210 |
+
"loss": 17.6585,
|
| 211 |
+
"nf_loss": 16.365,
|
| 212 |
+
"ppl": 4.0564,
|
| 213 |
+
"step": 30000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.45,
|
| 217 |
+
"gate_score": 0.2421,
|
| 218 |
+
"learning_rate": 0.0,
|
| 219 |
+
"loss": 16.8685,
|
| 220 |
+
"nf_loss": 15.5951,
|
| 221 |
+
"ppl": 3.9944,
|
| 222 |
+
"step": 32000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.45,
|
| 226 |
+
"eval_loss": 15.092597961425781,
|
| 227 |
+
"eval_nf_loss": 13.178274154663086,
|
| 228 |
+
"eval_perplexity": 7.254034996032715,
|
| 229 |
+
"eval_runtime": 911.9245,
|
| 230 |
+
"eval_samples_per_second": 14.659,
|
| 231 |
+
"eval_steps_per_second": 1.832,
|
| 232 |
+
"step": 32000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.47,
|
| 236 |
+
"gate_score": 0.2294,
|
| 237 |
+
"learning_rate": 0.0,
|
| 238 |
+
"loss": 15.8886,
|
| 239 |
+
"nf_loss": 14.6156,
|
| 240 |
+
"ppl": 3.9955,
|
| 241 |
+
"step": 34000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5,
|
| 245 |
+
"gate_score": 0.218,
|
| 246 |
+
"learning_rate": 0.0,
|
| 247 |
+
"loss": 14.9949,
|
| 248 |
+
"nf_loss": 13.7324,
|
| 249 |
+
"ppl": 3.9536,
|
| 250 |
+
"step": 36000
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.5,
|
| 254 |
+
"eval_loss": 13.548192977905273,
|
| 255 |
+
"eval_nf_loss": 11.632102012634277,
|
| 256 |
+
"eval_perplexity": 7.279270172119141,
|
| 257 |
+
"eval_runtime": 927.089,
|
| 258 |
+
"eval_samples_per_second": 14.419,
|
| 259 |
+
"eval_steps_per_second": 1.802,
|
| 260 |
+
"step": 36000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.53,
|
| 264 |
+
"gate_score": 0.2095,
|
| 265 |
+
"learning_rate": 0.0,
|
| 266 |
+
"loss": 14.5153,
|
| 267 |
+
"nf_loss": 13.2641,
|
| 268 |
+
"ppl": 3.8886,
|
| 269 |
+
"step": 38000
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.56,
|
| 273 |
+
"gate_score": 0.2065,
|
| 274 |
+
"learning_rate": 0.0,
|
| 275 |
+
"loss": 13.8227,
|
| 276 |
+
"nf_loss": 12.5901,
|
| 277 |
+
"ppl": 3.8375,
|
| 278 |
+
"step": 40000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.56,
|
| 282 |
+
"eval_loss": 12.412004470825195,
|
| 283 |
+
"eval_nf_loss": 10.545845985412598,
|
| 284 |
+
"eval_perplexity": 6.89225435256958,
|
| 285 |
+
"eval_runtime": 905.0462,
|
| 286 |
+
"eval_samples_per_second": 14.771,
|
| 287 |
+
"eval_steps_per_second": 1.846,
|
| 288 |
+
"step": 40000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.59,
|
| 292 |
+
"gate_score": 0.2009,
|
| 293 |
+
"learning_rate": 0.0,
|
| 294 |
+
"loss": 13.3939,
|
| 295 |
+
"nf_loss": 12.1663,
|
| 296 |
+
"ppl": 3.7926,
|
| 297 |
+
"step": 42000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.61,
|
| 301 |
+
"gate_score": 0.1935,
|
| 302 |
+
"learning_rate": 0.0,
|
| 303 |
+
"loss": 13.0096,
|
| 304 |
+
"nf_loss": 11.7892,
|
| 305 |
+
"ppl": 3.758,
|
| 306 |
+
"step": 44000
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.61,
|
| 310 |
+
"eval_loss": 11.815178871154785,
|
| 311 |
+
"eval_nf_loss": 9.942044258117676,
|
| 312 |
+
"eval_perplexity": 6.960797309875488,
|
| 313 |
+
"eval_runtime": 909.5617,
|
| 314 |
+
"eval_samples_per_second": 14.697,
|
| 315 |
+
"eval_steps_per_second": 1.837,
|
| 316 |
+
"step": 44000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.64,
|
| 320 |
+
"gate_score": 0.1903,
|
| 321 |
+
"learning_rate": 0.0,
|
| 322 |
+
"loss": 12.4269,
|
| 323 |
+
"nf_loss": 11.2151,
|
| 324 |
+
"ppl": 3.7435,
|
| 325 |
+
"step": 46000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.67,
|
| 329 |
+
"gate_score": 0.1854,
|
| 330 |
+
"learning_rate": 0.0,
|
| 331 |
+
"loss": 11.9829,
|
| 332 |
+
"nf_loss": 10.7658,
|
| 333 |
+
"ppl": 3.7626,
|
| 334 |
+
"step": 48000
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.67,
|
| 338 |
+
"eval_loss": 10.730130195617676,
|
| 339 |
+
"eval_nf_loss": 8.879775047302246,
|
| 340 |
+
"eval_perplexity": 6.794580459594727,
|
| 341 |
+
"eval_runtime": 900.5692,
|
| 342 |
+
"eval_samples_per_second": 14.844,
|
| 343 |
+
"eval_steps_per_second": 1.855,
|
| 344 |
+
"step": 48000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.7,
|
| 348 |
+
"gate_score": 0.1842,
|
| 349 |
+
"learning_rate": 0.0,
|
| 350 |
+
"loss": 11.6604,
|
| 351 |
+
"nf_loss": 10.459,
|
| 352 |
+
"ppl": 3.6987,
|
| 353 |
+
"step": 50000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.72,
|
| 357 |
+
"gate_score": 0.1778,
|
| 358 |
+
"learning_rate": 0.0,
|
| 359 |
+
"loss": 11.3837,
|
| 360 |
+
"nf_loss": 10.1987,
|
| 361 |
+
"ppl": 3.6203,
|
| 362 |
+
"step": 52000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.72,
|
| 366 |
+
"eval_loss": 10.283514976501465,
|
| 367 |
+
"eval_nf_loss": 8.462200164794922,
|
| 368 |
+
"eval_perplexity": 6.589704513549805,
|
| 369 |
+
"eval_runtime": 897.7218,
|
| 370 |
+
"eval_samples_per_second": 14.891,
|
| 371 |
+
"eval_steps_per_second": 1.861,
|
| 372 |
+
"step": 52000
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.75,
|
| 376 |
+
"gate_score": 0.1743,
|
| 377 |
+
"learning_rate": 0.0,
|
| 378 |
+
"loss": 11.1345,
|
| 379 |
+
"nf_loss": 9.9505,
|
| 380 |
+
"ppl": 3.6226,
|
| 381 |
+
"step": 54000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.78,
|
| 385 |
+
"gate_score": 0.1681,
|
| 386 |
+
"learning_rate": 0.0,
|
| 387 |
+
"loss": 10.913,
|
| 388 |
+
"nf_loss": 9.7401,
|
| 389 |
+
"ppl": 3.5839,
|
| 390 |
+
"step": 56000
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.78,
|
| 394 |
+
"eval_loss": 9.798712730407715,
|
| 395 |
+
"eval_nf_loss": 7.988397598266602,
|
| 396 |
+
"eval_perplexity": 6.516669750213623,
|
| 397 |
+
"eval_runtime": 883.3578,
|
| 398 |
+
"eval_samples_per_second": 15.133,
|
| 399 |
+
"eval_steps_per_second": 1.892,
|
| 400 |
+
"step": 56000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.81,
|
| 404 |
+
"gate_score": 0.1714,
|
| 405 |
+
"learning_rate": 0.0,
|
| 406 |
+
"loss": 10.7764,
|
| 407 |
+
"nf_loss": 9.605,
|
| 408 |
+
"ppl": 3.5672,
|
| 409 |
+
"step": 58000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.84,
|
| 413 |
+
"gate_score": 0.1655,
|
| 414 |
+
"learning_rate": 0.0,
|
| 415 |
+
"loss": 10.5627,
|
| 416 |
+
"nf_loss": 9.4126,
|
| 417 |
+
"ppl": 3.5056,
|
| 418 |
+
"step": 60000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.84,
|
| 422 |
+
"eval_loss": 9.538662910461426,
|
| 423 |
+
"eval_nf_loss": 7.738859176635742,
|
| 424 |
+
"eval_perplexity": 6.450651168823242,
|
| 425 |
+
"eval_runtime": 876.1125,
|
| 426 |
+
"eval_samples_per_second": 15.258,
|
| 427 |
+
"eval_steps_per_second": 1.907,
|
| 428 |
+
"step": 60000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.86,
|
| 432 |
+
"gate_score": 0.1661,
|
| 433 |
+
"learning_rate": 0.0,
|
| 434 |
+
"loss": 10.4422,
|
| 435 |
+
"nf_loss": 9.2887,
|
| 436 |
+
"ppl": 3.523,
|
| 437 |
+
"step": 62000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.89,
|
| 441 |
+
"gate_score": 0.1643,
|
| 442 |
+
"learning_rate": 0.0,
|
| 443 |
+
"loss": 10.238,
|
| 444 |
+
"nf_loss": 9.092,
|
| 445 |
+
"ppl": 3.4769,
|
| 446 |
+
"step": 64000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.89,
|
| 450 |
+
"eval_loss": 9.2942533493042,
|
| 451 |
+
"eval_nf_loss": 7.510049819946289,
|
| 452 |
+
"eval_perplexity": 6.345398426055908,
|
| 453 |
+
"eval_runtime": 873.6279,
|
| 454 |
+
"eval_samples_per_second": 15.302,
|
| 455 |
+
"eval_steps_per_second": 1.913,
|
| 456 |
+
"step": 64000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.92,
|
| 460 |
+
"gate_score": 0.1625,
|
| 461 |
+
"learning_rate": 0.0,
|
| 462 |
+
"loss": 10.1516,
|
| 463 |
+
"nf_loss": 9.0024,
|
| 464 |
+
"ppl": 3.4898,
|
| 465 |
+
"step": 66000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.95,
|
| 469 |
+
"gate_score": 0.1609,
|
| 470 |
+
"learning_rate": 0.0,
|
| 471 |
+
"loss": 9.9819,
|
| 472 |
+
"nf_loss": 8.8616,
|
| 473 |
+
"ppl": 3.3834,
|
| 474 |
+
"step": 68000
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.95,
|
| 478 |
+
"eval_loss": 9.13103199005127,
|
| 479 |
+
"eval_nf_loss": 7.360151290893555,
|
| 480 |
+
"eval_perplexity": 6.2570929527282715,
|
| 481 |
+
"eval_runtime": 892.7167,
|
| 482 |
+
"eval_samples_per_second": 14.975,
|
| 483 |
+
"eval_steps_per_second": 1.872,
|
| 484 |
+
"step": 68000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.98,
|
| 488 |
+
"gate_score": 0.1605,
|
| 489 |
+
"learning_rate": 0.0,
|
| 490 |
+
"loss": 9.9759,
|
| 491 |
+
"nf_loss": 8.8506,
|
| 492 |
+
"ppl": 3.3913,
|
| 493 |
+
"step": 70000
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.0,
|
| 497 |
+
"gate_score": 0.161,
|
| 498 |
+
"learning_rate": 5.5250000000000005e-06,
|
| 499 |
+
"loss": 9.9267,
|
| 500 |
+
"nf_loss": 8.805,
|
| 501 |
+
"ppl": 3.3955,
|
| 502 |
+
"step": 72000
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.0,
|
| 506 |
+
"eval_loss": 9.130549430847168,
|
| 507 |
+
"eval_nf_loss": 7.364365577697754,
|
| 508 |
+
"eval_perplexity": 6.226567268371582,
|
| 509 |
+
"eval_runtime": 865.1706,
|
| 510 |
+
"eval_samples_per_second": 15.451,
|
| 511 |
+
"eval_steps_per_second": 1.931,
|
| 512 |
+
"step": 72000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.03,
|
| 516 |
+
"gate_score": 0.1525,
|
| 517 |
+
"learning_rate": 4.9922293335593894e-05,
|
| 518 |
+
"loss": 10.6431,
|
| 519 |
+
"nf_loss": 9.4961,
|
| 520 |
+
"ppl": 3.4863,
|
| 521 |
+
"step": 74000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.06,
|
| 525 |
+
"gate_score": 0.1281,
|
| 526 |
+
"learning_rate": 4.921622232583111e-05,
|
| 527 |
+
"loss": 11.018,
|
| 528 |
+
"nf_loss": 9.7948,
|
| 529 |
+
"ppl": 3.8142,
|
| 530 |
+
"step": 76000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.06,
|
| 534 |
+
"eval_loss": 10.00072193145752,
|
| 535 |
+
"eval_nf_loss": 8.153450965881348,
|
| 536 |
+
"eval_perplexity": 6.768190383911133,
|
| 537 |
+
"eval_runtime": 618.9198,
|
| 538 |
+
"eval_samples_per_second": 21.599,
|
| 539 |
+
"eval_steps_per_second": 2.7,
|
| 540 |
+
"step": 76000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.09,
|
| 544 |
+
"gate_score": 0.1074,
|
| 545 |
+
"learning_rate": 4.850979810395739e-05,
|
| 546 |
+
"loss": 10.5531,
|
| 547 |
+
"nf_loss": 9.3486,
|
| 548 |
+
"ppl": 3.7186,
|
| 549 |
+
"step": 78000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.11,
|
| 553 |
+
"gate_score": 0.0924,
|
| 554 |
+
"learning_rate": 4.780408030630554e-05,
|
| 555 |
+
"loss": 10.0575,
|
| 556 |
+
"nf_loss": 8.8613,
|
| 557 |
+
"ppl": 3.7072,
|
| 558 |
+
"step": 80000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 1.11,
|
| 562 |
+
"eval_loss": 9.039822578430176,
|
| 563 |
+
"eval_nf_loss": 7.206754684448242,
|
| 564 |
+
"eval_perplexity": 6.669988632202148,
|
| 565 |
+
"eval_runtime": 639.9491,
|
| 566 |
+
"eval_samples_per_second": 20.889,
|
| 567 |
+
"eval_steps_per_second": 2.611,
|
| 568 |
+
"step": 80000
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.14,
|
| 572 |
+
"gate_score": 0.0874,
|
| 573 |
+
"learning_rate": 4.7097656084431825e-05,
|
| 574 |
+
"loss": 9.7084,
|
| 575 |
+
"nf_loss": 8.5025,
|
| 576 |
+
"ppl": 3.7371,
|
| 577 |
+
"step": 82000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.17,
|
| 581 |
+
"gate_score": 0.0759,
|
| 582 |
+
"learning_rate": 4.639158507466904e-05,
|
| 583 |
+
"loss": 9.3613,
|
| 584 |
+
"nf_loss": 8.1756,
|
| 585 |
+
"ppl": 3.6444,
|
| 586 |
+
"step": 84000
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.17,
|
| 590 |
+
"eval_loss": 8.67480182647705,
|
| 591 |
+
"eval_nf_loss": 6.853387832641602,
|
| 592 |
+
"eval_perplexity": 6.5951128005981445,
|
| 593 |
+
"eval_runtime": 655.8198,
|
| 594 |
+
"eval_samples_per_second": 20.384,
|
| 595 |
+
"eval_steps_per_second": 2.548,
|
| 596 |
+
"step": 84000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.2,
|
| 600 |
+
"gate_score": 0.0725,
|
| 601 |
+
"learning_rate": 4.5685160852795324e-05,
|
| 602 |
+
"loss": 8.9712,
|
| 603 |
+
"nf_loss": 7.7979,
|
| 604 |
+
"ppl": 3.5998,
|
| 605 |
+
"step": 86000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.23,
|
| 609 |
+
"gate_score": 0.0666,
|
| 610 |
+
"learning_rate": 4.497908984303254e-05,
|
| 611 |
+
"loss": 8.7453,
|
| 612 |
+
"nf_loss": 7.5777,
|
| 613 |
+
"ppl": 3.5764,
|
| 614 |
+
"step": 88000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.23,
|
| 618 |
+
"eval_loss": 7.947096347808838,
|
| 619 |
+
"eval_nf_loss": 6.137327671051025,
|
| 620 |
+
"eval_perplexity": 6.509429454803467,
|
| 621 |
+
"eval_runtime": 654.1896,
|
| 622 |
+
"eval_samples_per_second": 20.434,
|
| 623 |
+
"eval_steps_per_second": 2.554,
|
| 624 |
+
"step": 88000
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 1.25,
|
| 628 |
+
"gate_score": 0.0657,
|
| 629 |
+
"learning_rate": 4.4273018833269757e-05,
|
| 630 |
+
"loss": 8.4051,
|
| 631 |
+
"nf_loss": 7.2351,
|
| 632 |
+
"ppl": 3.5928,
|
| 633 |
+
"step": 90000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.28,
|
| 637 |
+
"gate_score": 0.0632,
|
| 638 |
+
"learning_rate": 4.356694782350697e-05,
|
| 639 |
+
"loss": 8.2083,
|
| 640 |
+
"nf_loss": 7.0471,
|
| 641 |
+
"ppl": 3.5522,
|
| 642 |
+
"step": 92000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.28,
|
| 646 |
+
"eval_loss": 7.776934623718262,
|
| 647 |
+
"eval_nf_loss": 5.991891384124756,
|
| 648 |
+
"eval_perplexity": 6.346434593200684,
|
| 649 |
+
"eval_runtime": 664.9292,
|
| 650 |
+
"eval_samples_per_second": 20.104,
|
| 651 |
+
"eval_steps_per_second": 2.513,
|
| 652 |
+
"step": 92000
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 1.31,
|
| 656 |
+
"gate_score": 0.0614,
|
| 657 |
+
"learning_rate": 4.2860523601633256e-05,
|
| 658 |
+
"loss": 7.8299,
|
| 659 |
+
"nf_loss": 6.6969,
|
| 660 |
+
"ppl": 3.4547,
|
| 661 |
+
"step": 94000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.34,
|
| 665 |
+
"gate_score": 0.0596,
|
| 666 |
+
"learning_rate": 4.215445259187047e-05,
|
| 667 |
+
"loss": 7.6361,
|
| 668 |
+
"nf_loss": 6.4891,
|
| 669 |
+
"ppl": 3.5083,
|
| 670 |
+
"step": 96000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.34,
|
| 674 |
+
"eval_loss": 7.583176612854004,
|
| 675 |
+
"eval_nf_loss": 5.804019927978516,
|
| 676 |
+
"eval_perplexity": 6.316826343536377,
|
| 677 |
+
"eval_runtime": 649.9826,
|
| 678 |
+
"eval_samples_per_second": 20.567,
|
| 679 |
+
"eval_steps_per_second": 2.571,
|
| 680 |
+
"step": 96000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.37,
|
| 684 |
+
"gate_score": 0.0567,
|
| 685 |
+
"learning_rate": 4.144838158210769e-05,
|
| 686 |
+
"loss": 7.4853,
|
| 687 |
+
"nf_loss": 6.36,
|
| 688 |
+
"ppl": 3.4233,
|
| 689 |
+
"step": 98000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.39,
|
| 693 |
+
"gate_score": 0.0557,
|
| 694 |
+
"learning_rate": 4.074195736023397e-05,
|
| 695 |
+
"loss": 7.2881,
|
| 696 |
+
"nf_loss": 6.1511,
|
| 697 |
+
"ppl": 3.4723,
|
| 698 |
+
"step": 100000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.39,
|
| 702 |
+
"eval_loss": 7.400701999664307,
|
| 703 |
+
"eval_nf_loss": 5.625098705291748,
|
| 704 |
+
"eval_perplexity": 6.289206504821777,
|
| 705 |
+
"eval_runtime": 667.0031,
|
| 706 |
+
"eval_samples_per_second": 20.042,
|
| 707 |
+
"eval_steps_per_second": 2.505,
|
| 708 |
+
"step": 100000
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.42,
|
| 712 |
+
"gate_score": 0.054,
|
| 713 |
+
"learning_rate": 4.003588635047119e-05,
|
| 714 |
+
"loss": 7.165,
|
| 715 |
+
"nf_loss": 6.0472,
|
| 716 |
+
"ppl": 3.3942,
|
| 717 |
+
"step": 102000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.45,
|
| 721 |
+
"gate_score": 0.0532,
|
| 722 |
+
"learning_rate": 3.9329815340708403e-05,
|
| 723 |
+
"loss": 6.9929,
|
| 724 |
+
"nf_loss": 5.864,
|
| 725 |
+
"ppl": 3.45,
|
| 726 |
+
"step": 104000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.45,
|
| 730 |
+
"eval_loss": 6.6446003913879395,
|
| 731 |
+
"eval_nf_loss": 4.897329330444336,
|
| 732 |
+
"eval_perplexity": 6.099964141845703,
|
| 733 |
+
"eval_runtime": 665.1582,
|
| 734 |
+
"eval_samples_per_second": 20.097,
|
| 735 |
+
"eval_steps_per_second": 2.512,
|
| 736 |
+
"step": 104000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 1.48,
|
| 740 |
+
"gate_score": 0.0508,
|
| 741 |
+
"learning_rate": 3.8623391118834686e-05,
|
| 742 |
+
"loss": 6.8638,
|
| 743 |
+
"nf_loss": 5.7412,
|
| 744 |
+
"ppl": 3.4016,
|
| 745 |
+
"step": 106000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.5,
|
| 749 |
+
"gate_score": 0.048,
|
| 750 |
+
"learning_rate": 3.79173201090719e-05,
|
| 751 |
+
"loss": 6.7439,
|
| 752 |
+
"nf_loss": 5.6285,
|
| 753 |
+
"ppl": 3.3851,
|
| 754 |
+
"step": 108000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 1.5,
|
| 758 |
+
"eval_loss": 6.440381050109863,
|
| 759 |
+
"eval_nf_loss": 4.687047004699707,
|
| 760 |
+
"eval_perplexity": 6.147556304931641,
|
| 761 |
+
"eval_runtime": 671.2658,
|
| 762 |
+
"eval_samples_per_second": 19.915,
|
| 763 |
+
"eval_steps_per_second": 2.489,
|
| 764 |
+
"step": 108000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.53,
|
| 768 |
+
"gate_score": 0.0489,
|
| 769 |
+
"learning_rate": 3.721124909930912e-05,
|
| 770 |
+
"loss": 6.6258,
|
| 771 |
+
"nf_loss": 5.5227,
|
| 772 |
+
"ppl": 3.34,
|
| 773 |
+
"step": 110000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.56,
|
| 777 |
+
"gate_score": 0.0476,
|
| 778 |
+
"learning_rate": 3.65048248774354e-05,
|
| 779 |
+
"loss": 6.5119,
|
| 780 |
+
"nf_loss": 5.4155,
|
| 781 |
+
"ppl": 3.3128,
|
| 782 |
+
"step": 112000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.56,
|
| 786 |
+
"eval_loss": 6.296238422393799,
|
| 787 |
+
"eval_nf_loss": 4.570157527923584,
|
| 788 |
+
"eval_perplexity": 5.967574119567871,
|
| 789 |
+
"eval_runtime": 658.6675,
|
| 790 |
+
"eval_samples_per_second": 20.296,
|
| 791 |
+
"eval_steps_per_second": 2.537,
|
| 792 |
+
"step": 112000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.59,
|
| 796 |
+
"gate_score": 0.0482,
|
| 797 |
+
"learning_rate": 3.579910707978356e-05,
|
| 798 |
+
"loss": 6.3285,
|
| 799 |
+
"nf_loss": 5.2244,
|
| 800 |
+
"ppl": 3.3628,
|
| 801 |
+
"step": 114000
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.62,
|
| 805 |
+
"gate_score": 0.0482,
|
| 806 |
+
"learning_rate": 3.5092682857909834e-05,
|
| 807 |
+
"loss": 6.2206,
|
| 808 |
+
"nf_loss": 5.1234,
|
| 809 |
+
"ppl": 3.3327,
|
| 810 |
+
"step": 116000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.62,
|
| 814 |
+
"eval_loss": 6.046633720397949,
|
| 815 |
+
"eval_nf_loss": 4.321578025817871,
|
| 816 |
+
"eval_perplexity": 5.9654130935668945,
|
| 817 |
+
"eval_runtime": 653.7138,
|
| 818 |
+
"eval_samples_per_second": 20.449,
|
| 819 |
+
"eval_steps_per_second": 2.556,
|
| 820 |
+
"step": 116000
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 1.64,
|
| 824 |
+
"gate_score": 0.0489,
|
| 825 |
+
"learning_rate": 3.438625863603611e-05,
|
| 826 |
+
"loss": 6.0978,
|
| 827 |
+
"nf_loss": 5.0037,
|
| 828 |
+
"ppl": 3.3053,
|
| 829 |
+
"step": 118000
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.67,
|
| 833 |
+
"gate_score": 0.0468,
|
| 834 |
+
"learning_rate": 3.3680540838384266e-05,
|
| 835 |
+
"loss": 6.0198,
|
| 836 |
+
"nf_loss": 4.9354,
|
| 837 |
+
"ppl": 3.2737,
|
| 838 |
+
"step": 120000
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.67,
|
| 842 |
+
"eval_loss": 5.779796123504639,
|
| 843 |
+
"eval_nf_loss": 4.067526817321777,
|
| 844 |
+
"eval_perplexity": 5.880522727966309,
|
| 845 |
+
"eval_runtime": 641.0411,
|
| 846 |
+
"eval_samples_per_second": 20.854,
|
| 847 |
+
"eval_steps_per_second": 2.607,
|
| 848 |
+
"step": 120000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.7,
|
| 852 |
+
"gate_score": 0.0425,
|
| 853 |
+
"learning_rate": 3.297446982862148e-05,
|
| 854 |
+
"loss": 5.907,
|
| 855 |
+
"nf_loss": 4.8253,
|
| 856 |
+
"ppl": 3.2734,
|
| 857 |
+
"step": 122000
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.73,
|
| 861 |
+
"gate_score": 0.0432,
|
| 862 |
+
"learning_rate": 3.2268045606747765e-05,
|
| 863 |
+
"loss": 5.7937,
|
| 864 |
+
"nf_loss": 4.7167,
|
| 865 |
+
"ppl": 3.2463,
|
| 866 |
+
"step": 124000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.73,
|
| 870 |
+
"eval_loss": 5.693485260009766,
|
| 871 |
+
"eval_nf_loss": 3.997667074203491,
|
| 872 |
+
"eval_perplexity": 5.778055667877197,
|
| 873 |
+
"eval_runtime": 629.6888,
|
| 874 |
+
"eval_samples_per_second": 21.23,
|
| 875 |
+
"eval_steps_per_second": 2.654,
|
| 876 |
+
"step": 124000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 1.76,
|
| 880 |
+
"gate_score": 0.0413,
|
| 881 |
+
"learning_rate": 3.156197459698498e-05,
|
| 882 |
+
"loss": 5.6708,
|
| 883 |
+
"nf_loss": 4.5839,
|
| 884 |
+
"ppl": 3.2695,
|
| 885 |
+
"step": 126000
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.78,
|
| 889 |
+
"gate_score": 0.0423,
|
| 890 |
+
"learning_rate": 3.08559035872222e-05,
|
| 891 |
+
"loss": 5.5759,
|
| 892 |
+
"nf_loss": 4.5071,
|
| 893 |
+
"ppl": 3.2127,
|
| 894 |
+
"step": 128000
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.78,
|
| 898 |
+
"eval_loss": 5.489774227142334,
|
| 899 |
+
"eval_nf_loss": 3.8035385608673096,
|
| 900 |
+
"eval_perplexity": 5.725496292114258,
|
| 901 |
+
"eval_runtime": 629.3114,
|
| 902 |
+
"eval_samples_per_second": 21.242,
|
| 903 |
+
"eval_steps_per_second": 2.655,
|
| 904 |
+
"step": 128000
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 1.81,
|
| 908 |
+
"gate_score": 0.0415,
|
| 909 |
+
"learning_rate": 3.0149832577459418e-05,
|
| 910 |
+
"loss": 5.4762,
|
| 911 |
+
"nf_loss": 4.4182,
|
| 912 |
+
"ppl": 3.189,
|
| 913 |
+
"step": 130000
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.84,
|
| 917 |
+
"gate_score": 0.043,
|
| 918 |
+
"learning_rate": 2.9443761567696637e-05,
|
| 919 |
+
"loss": 5.4304,
|
| 920 |
+
"nf_loss": 4.3627,
|
| 921 |
+
"ppl": 3.2132,
|
| 922 |
+
"step": 132000
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.84,
|
| 926 |
+
"eval_loss": 5.458424091339111,
|
| 927 |
+
"eval_nf_loss": 3.7697036266326904,
|
| 928 |
+
"eval_perplexity": 5.744439125061035,
|
| 929 |
+
"eval_runtime": 641.9111,
|
| 930 |
+
"eval_samples_per_second": 20.825,
|
| 931 |
+
"eval_steps_per_second": 2.603,
|
| 932 |
+
"step": 132000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 1.87,
|
| 936 |
+
"gate_score": 0.0438,
|
| 937 |
+
"learning_rate": 2.8737337345822913e-05,
|
| 938 |
+
"loss": 5.3504,
|
| 939 |
+
"nf_loss": 4.2998,
|
| 940 |
+
"ppl": 3.1531,
|
| 941 |
+
"step": 134000
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.89,
|
| 945 |
+
"gate_score": 0.0429,
|
| 946 |
+
"learning_rate": 2.803126633606013e-05,
|
| 947 |
+
"loss": 5.2695,
|
| 948 |
+
"nf_loss": 4.2383,
|
| 949 |
+
"ppl": 3.0881,
|
| 950 |
+
"step": 136000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.89,
|
| 954 |
+
"eval_loss": 5.256904125213623,
|
| 955 |
+
"eval_nf_loss": 3.5766844749450684,
|
| 956 |
+
"eval_perplexity": 5.6950812339782715,
|
| 957 |
+
"eval_runtime": 660.429,
|
| 958 |
+
"eval_samples_per_second": 20.241,
|
| 959 |
+
"eval_steps_per_second": 2.53,
|
| 960 |
+
"step": 136000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 1.92,
|
| 964 |
+
"gate_score": 0.0421,
|
| 965 |
+
"learning_rate": 2.7324842114186412e-05,
|
| 966 |
+
"loss": 5.2219,
|
| 967 |
+
"nf_loss": 4.1712,
|
| 968 |
+
"ppl": 3.1497,
|
| 969 |
+
"step": 138000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.95,
|
| 973 |
+
"gate_score": 0.0417,
|
| 974 |
+
"learning_rate": 2.661877110442363e-05,
|
| 975 |
+
"loss": 5.1817,
|
| 976 |
+
"nf_loss": 4.1256,
|
| 977 |
+
"ppl": 3.1777,
|
| 978 |
+
"step": 140000
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.95,
|
| 982 |
+
"eval_loss": 5.158201217651367,
|
| 983 |
+
"eval_nf_loss": 3.487734794616699,
|
| 984 |
+
"eval_perplexity": 5.636016845703125,
|
| 985 |
+
"eval_runtime": 644.0563,
|
| 986 |
+
"eval_samples_per_second": 20.756,
|
| 987 |
+
"eval_steps_per_second": 2.594,
|
| 988 |
+
"step": 140000
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.98,
|
| 992 |
+
"gate_score": 0.0422,
|
| 993 |
+
"learning_rate": 2.5913053306771785e-05,
|
| 994 |
+
"loss": 5.1108,
|
| 995 |
+
"nf_loss": 4.0728,
|
| 996 |
+
"ppl": 3.1163,
|
| 997 |
+
"step": 142000
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.01,
|
| 1001 |
+
"gate_score": 0.0402,
|
| 1002 |
+
"learning_rate": 2.5206982297009e-05,
|
| 1003 |
+
"loss": 5.0219,
|
| 1004 |
+
"nf_loss": 4.0038,
|
| 1005 |
+
"ppl": 3.0436,
|
| 1006 |
+
"step": 144000
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 2.01,
|
| 1010 |
+
"eval_loss": 5.062300205230713,
|
| 1011 |
+
"eval_nf_loss": 3.3892769813537598,
|
| 1012 |
+
"eval_perplexity": 5.655203819274902,
|
| 1013 |
+
"eval_runtime": 654.7044,
|
| 1014 |
+
"eval_samples_per_second": 20.418,
|
| 1015 |
+
"eval_steps_per_second": 2.552,
|
| 1016 |
+
"step": 144000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.03,
|
| 1020 |
+
"gate_score": 0.0407,
|
| 1021 |
+
"learning_rate": 2.4500558075135284e-05,
|
| 1022 |
+
"loss": 4.8922,
|
| 1023 |
+
"nf_loss": 3.9548,
|
| 1024 |
+
"ppl": 2.775,
|
| 1025 |
+
"step": 146000
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.06,
|
| 1029 |
+
"gate_score": 0.0409,
|
| 1030 |
+
"learning_rate": 2.379413385326156e-05,
|
| 1031 |
+
"loss": 4.8179,
|
| 1032 |
+
"nf_loss": 3.8903,
|
| 1033 |
+
"ppl": 2.7374,
|
| 1034 |
+
"step": 148000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.06,
|
| 1038 |
+
"eval_loss": 4.891648769378662,
|
| 1039 |
+
"eval_nf_loss": 3.21927547454834,
|
| 1040 |
+
"eval_perplexity": 5.654395580291748,
|
| 1041 |
+
"eval_runtime": 649.5705,
|
| 1042 |
+
"eval_samples_per_second": 20.58,
|
| 1043 |
+
"eval_steps_per_second": 2.572,
|
| 1044 |
+
"step": 148000
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 2.09,
|
| 1048 |
+
"gate_score": 0.0431,
|
| 1049 |
+
"learning_rate": 2.308806284349878e-05,
|
| 1050 |
+
"loss": 4.7494,
|
| 1051 |
+
"nf_loss": 3.8238,
|
| 1052 |
+
"ppl": 2.7455,
|
| 1053 |
+
"step": 150000
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.12,
|
| 1057 |
+
"gate_score": 0.0424,
|
| 1058 |
+
"learning_rate": 2.2381991833735996e-05,
|
| 1059 |
+
"loss": 4.7155,
|
| 1060 |
+
"nf_loss": 3.7946,
|
| 1061 |
+
"ppl": 2.7302,
|
| 1062 |
+
"step": 152000
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 2.12,
|
| 1066 |
+
"eval_loss": 4.806872844696045,
|
| 1067 |
+
"eval_nf_loss": 3.148994207382202,
|
| 1068 |
+
"eval_perplexity": 5.567600250244141,
|
| 1069 |
+
"eval_runtime": 630.6434,
|
| 1070 |
+
"eval_samples_per_second": 21.197,
|
| 1071 |
+
"eval_steps_per_second": 2.65,
|
| 1072 |
+
"step": 152000
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 2.15,
|
| 1076 |
+
"gate_score": 0.0408,
|
| 1077 |
+
"learning_rate": 2.1675567611862275e-05,
|
| 1078 |
+
"loss": 4.6944,
|
| 1079 |
+
"nf_loss": 3.7643,
|
| 1080 |
+
"ppl": 2.7512,
|
| 1081 |
+
"step": 154000
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.17,
|
| 1085 |
+
"gate_score": 0.0426,
|
| 1086 |
+
"learning_rate": 2.0969496602099495e-05,
|
| 1087 |
+
"loss": 4.6805,
|
| 1088 |
+
"nf_loss": 3.7557,
|
| 1089 |
+
"ppl": 2.7415,
|
| 1090 |
+
"step": 156000
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.17,
|
| 1094 |
+
"eval_loss": 4.722836494445801,
|
| 1095 |
+
"eval_nf_loss": 3.066758632659912,
|
| 1096 |
+
"eval_perplexity": 5.557853698730469,
|
| 1097 |
+
"eval_runtime": 604.3797,
|
| 1098 |
+
"eval_samples_per_second": 22.119,
|
| 1099 |
+
"eval_steps_per_second": 2.765,
|
| 1100 |
+
"step": 156000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.2,
|
| 1104 |
+
"gate_score": 0.0438,
|
| 1105 |
+
"learning_rate": 2.026342559233671e-05,
|
| 1106 |
+
"loss": 4.5943,
|
| 1107 |
+
"nf_loss": 3.6721,
|
| 1108 |
+
"ppl": 2.722,
|
| 1109 |
+
"step": 158000
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.23,
|
| 1113 |
+
"gate_score": 0.0447,
|
| 1114 |
+
"learning_rate": 1.9557354582573927e-05,
|
| 1115 |
+
"loss": 4.5781,
|
| 1116 |
+
"nf_loss": 3.6564,
|
| 1117 |
+
"ppl": 2.728,
|
| 1118 |
+
"step": 160000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.23,
|
| 1122 |
+
"eval_loss": 4.671295166015625,
|
| 1123 |
+
"eval_nf_loss": 3.0203938484191895,
|
| 1124 |
+
"eval_perplexity": 5.526300430297852,
|
| 1125 |
+
"eval_runtime": 604.9582,
|
| 1126 |
+
"eval_samples_per_second": 22.097,
|
| 1127 |
+
"eval_steps_per_second": 2.762,
|
| 1128 |
+
"step": 160000
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 2.26,
|
| 1132 |
+
"gate_score": 0.0436,
|
| 1133 |
+
"learning_rate": 1.8851283572811147e-05,
|
| 1134 |
+
"loss": 4.5642,
|
| 1135 |
+
"nf_loss": 3.6404,
|
| 1136 |
+
"ppl": 2.7413,
|
| 1137 |
+
"step": 162000
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.28,
|
| 1141 |
+
"gate_score": 0.0453,
|
| 1142 |
+
"learning_rate": 1.8144859350937426e-05,
|
| 1143 |
+
"loss": 4.4834,
|
| 1144 |
+
"nf_loss": 3.5645,
|
| 1145 |
+
"ppl": 2.7145,
|
| 1146 |
+
"step": 164000
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.28,
|
| 1150 |
+
"eval_loss": 4.710845947265625,
|
| 1151 |
+
"eval_nf_loss": 3.063795566558838,
|
| 1152 |
+
"eval_perplexity": 5.505098342895508,
|
| 1153 |
+
"eval_runtime": 601.1865,
|
| 1154 |
+
"eval_samples_per_second": 22.236,
|
| 1155 |
+
"eval_steps_per_second": 2.78,
|
| 1156 |
+
"step": 164000
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 2.31,
|
| 1160 |
+
"gate_score": 0.0442,
|
| 1161 |
+
"learning_rate": 1.7438788341174643e-05,
|
| 1162 |
+
"loss": 4.4998,
|
| 1163 |
+
"nf_loss": 3.5735,
|
| 1164 |
+
"ppl": 2.7361,
|
| 1165 |
+
"step": 166000
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.34,
|
| 1169 |
+
"gate_score": 0.0442,
|
| 1170 |
+
"learning_rate": 1.6733070543522796e-05,
|
| 1171 |
+
"loss": 4.4172,
|
| 1172 |
+
"nf_loss": 3.4989,
|
| 1173 |
+
"ppl": 2.7118,
|
| 1174 |
+
"step": 168000
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 2.34,
|
| 1178 |
+
"eval_loss": 4.5222930908203125,
|
| 1179 |
+
"eval_nf_loss": 2.8794257640838623,
|
| 1180 |
+
"eval_perplexity": 5.477485179901123,
|
| 1181 |
+
"eval_runtime": 609.6448,
|
| 1182 |
+
"eval_samples_per_second": 21.928,
|
| 1183 |
+
"eval_steps_per_second": 2.741,
|
| 1184 |
+
"step": 168000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 2.37,
|
| 1188 |
+
"gate_score": 0.0464,
|
| 1189 |
+
"learning_rate": 1.602664632164908e-05,
|
| 1190 |
+
"loss": 4.3817,
|
| 1191 |
+
"nf_loss": 3.4671,
|
| 1192 |
+
"ppl": 2.7031,
|
| 1193 |
+
"step": 170000
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.4,
|
| 1197 |
+
"gate_score": 0.0447,
|
| 1198 |
+
"learning_rate": 1.5320575311886295e-05,
|
| 1199 |
+
"loss": 4.3776,
|
| 1200 |
+
"nf_loss": 3.4701,
|
| 1201 |
+
"ppl": 2.6834,
|
| 1202 |
+
"step": 172000
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.4,
|
| 1206 |
+
"eval_loss": 4.446962356567383,
|
| 1207 |
+
"eval_nf_loss": 2.8145503997802734,
|
| 1208 |
+
"eval_perplexity": 5.419012546539307,
|
| 1209 |
+
"eval_runtime": 591.7923,
|
| 1210 |
+
"eval_samples_per_second": 22.589,
|
| 1211 |
+
"eval_steps_per_second": 2.824,
|
| 1212 |
+
"step": 172000
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 2.42,
|
| 1216 |
+
"gate_score": 0.0438,
|
| 1217 |
+
"learning_rate": 1.4614151090012576e-05,
|
| 1218 |
+
"loss": 4.34,
|
| 1219 |
+
"nf_loss": 3.4303,
|
| 1220 |
+
"ppl": 2.6917,
|
| 1221 |
+
"step": 174000
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.45,
|
| 1225 |
+
"gate_score": 0.0433,
|
| 1226 |
+
"learning_rate": 1.390808008024979e-05,
|
| 1227 |
+
"loss": 4.3096,
|
| 1228 |
+
"nf_loss": 3.4,
|
| 1229 |
+
"ppl": 2.6915,
|
| 1230 |
+
"step": 176000
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 2.45,
|
| 1234 |
+
"eval_loss": 4.412790298461914,
|
| 1235 |
+
"eval_nf_loss": 2.7833149433135986,
|
| 1236 |
+
"eval_perplexity": 5.404773712158203,
|
| 1237 |
+
"eval_runtime": 593.2765,
|
| 1238 |
+
"eval_samples_per_second": 22.532,
|
| 1239 |
+
"eval_steps_per_second": 2.817,
|
| 1240 |
+
"step": 176000
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 2.48,
|
| 1244 |
+
"gate_score": 0.0449,
|
| 1245 |
+
"learning_rate": 1.3201655858376072e-05,
|
| 1246 |
+
"loss": 4.289,
|
| 1247 |
+
"nf_loss": 3.3752,
|
| 1248 |
+
"ppl": 2.7063,
|
| 1249 |
+
"step": 178000
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.51,
|
| 1253 |
+
"gate_score": 0.0431,
|
| 1254 |
+
"learning_rate": 1.2495938060724226e-05,
|
| 1255 |
+
"loss": 4.2453,
|
| 1256 |
+
"nf_loss": 3.3433,
|
| 1257 |
+
"ppl": 2.6679,
|
| 1258 |
+
"step": 180000
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 2.51,
|
| 1262 |
+
"eval_loss": 4.363792896270752,
|
| 1263 |
+
"eval_nf_loss": 2.7436203956604004,
|
| 1264 |
+
"eval_perplexity": 5.350350856781006,
|
| 1265 |
+
"eval_runtime": 592.6812,
|
| 1266 |
+
"eval_samples_per_second": 22.555,
|
| 1267 |
+
"eval_steps_per_second": 2.819,
|
| 1268 |
+
"step": 180000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.54,
|
| 1272 |
+
"gate_score": 0.0444,
|
| 1273 |
+
"learning_rate": 1.1789867050961444e-05,
|
| 1274 |
+
"loss": 4.1812,
|
| 1275 |
+
"nf_loss": 3.2773,
|
| 1276 |
+
"ppl": 2.6769,
|
| 1277 |
+
"step": 182000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.56,
|
| 1281 |
+
"gate_score": 0.0439,
|
| 1282 |
+
"learning_rate": 1.1083442829087724e-05,
|
| 1283 |
+
"loss": 4.2063,
|
| 1284 |
+
"nf_loss": 3.3076,
|
| 1285 |
+
"ppl": 2.6592,
|
| 1286 |
+
"step": 184000
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.56,
|
| 1290 |
+
"eval_loss": 4.401766777038574,
|
| 1291 |
+
"eval_nf_loss": 2.7845916748046875,
|
| 1292 |
+
"eval_perplexity": 5.334909915924072,
|
| 1293 |
+
"eval_runtime": 593.569,
|
| 1294 |
+
"eval_samples_per_second": 22.521,
|
| 1295 |
+
"eval_steps_per_second": 2.815,
|
| 1296 |
+
"step": 184000
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 2.59,
|
| 1300 |
+
"gate_score": 0.0435,
|
| 1301 |
+
"learning_rate": 1.0377371819324942e-05,
|
| 1302 |
+
"loss": 4.1697,
|
| 1303 |
+
"nf_loss": 3.2627,
|
| 1304 |
+
"ppl": 2.6781,
|
| 1305 |
+
"step": 186000
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.62,
|
| 1309 |
+
"gate_score": 0.0455,
|
| 1310 |
+
"learning_rate": 9.67130080956216e-06,
|
| 1311 |
+
"loss": 4.1413,
|
| 1312 |
+
"nf_loss": 3.2508,
|
| 1313 |
+
"ppl": 2.6383,
|
| 1314 |
+
"step": 188000
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 2.62,
|
| 1318 |
+
"eval_loss": 4.287137031555176,
|
| 1319 |
+
"eval_nf_loss": 2.6753435134887695,
|
| 1320 |
+
"eval_perplexity": 5.3036627769470215,
|
| 1321 |
+
"eval_runtime": 592.6474,
|
| 1322 |
+
"eval_samples_per_second": 22.556,
|
| 1323 |
+
"eval_steps_per_second": 2.82,
|
| 1324 |
+
"step": 188000
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 2.65,
|
| 1328 |
+
"gate_score": 0.0438,
|
| 1329 |
+
"learning_rate": 8.964876587688439e-06,
|
| 1330 |
+
"loss": 4.1244,
|
| 1331 |
+
"nf_loss": 3.2269,
|
| 1332 |
+
"ppl": 2.6537,
|
| 1333 |
+
"step": 190000
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.67,
|
| 1337 |
+
"gate_score": 0.0446,
|
| 1338 |
+
"learning_rate": 8.258805577925655e-06,
|
| 1339 |
+
"loss": 4.0845,
|
| 1340 |
+
"nf_loss": 3.1966,
|
| 1341 |
+
"ppl": 2.6201,
|
| 1342 |
+
"step": 192000
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 2.67,
|
| 1346 |
+
"eval_loss": 4.247885704040527,
|
| 1347 |
+
"eval_nf_loss": 2.640995502471924,
|
| 1348 |
+
"eval_perplexity": 5.275094985961914,
|
| 1349 |
+
"eval_runtime": 591.9491,
|
| 1350 |
+
"eval_samples_per_second": 22.583,
|
| 1351 |
+
"eval_steps_per_second": 2.823,
|
| 1352 |
+
"step": 192000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 2.7,
|
| 1356 |
+
"gate_score": 0.0445,
|
| 1357 |
+
"learning_rate": 7.552734568162874e-06,
|
| 1358 |
+
"loss": 4.0604,
|
| 1359 |
+
"nf_loss": 3.1769,
|
| 1360 |
+
"ppl": 2.6091,
|
| 1361 |
+
"step": 194000
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.73,
|
| 1365 |
+
"gate_score": 0.0433,
|
| 1366 |
+
"learning_rate": 6.846310346289154e-06,
|
| 1367 |
+
"loss": 4.06,
|
| 1368 |
+
"nf_loss": 3.1758,
|
| 1369 |
+
"ppl": 2.6114,
|
| 1370 |
+
"step": 196000
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 2.73,
|
| 1374 |
+
"eval_loss": 4.171822547912598,
|
| 1375 |
+
"eval_nf_loss": 2.567200183868408,
|
| 1376 |
+
"eval_perplexity": 5.264888763427734,
|
| 1377 |
+
"eval_runtime": 589.4312,
|
| 1378 |
+
"eval_samples_per_second": 22.679,
|
| 1379 |
+
"eval_steps_per_second": 2.835,
|
| 1380 |
+
"step": 196000
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 2.76,
|
| 1384 |
+
"gate_score": 0.0436,
|
| 1385 |
+
"learning_rate": 6.140239336526371e-06,
|
| 1386 |
+
"loss": 4.0426,
|
| 1387 |
+
"nf_loss": 3.1603,
|
| 1388 |
+
"ppl": 2.6136,
|
| 1389 |
+
"step": 198000
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.79,
|
| 1393 |
+
"gate_score": 0.0444,
|
| 1394 |
+
"learning_rate": 5.4338151146526515e-06,
|
| 1395 |
+
"loss": 4.0352,
|
| 1396 |
+
"nf_loss": 3.1654,
|
| 1397 |
+
"ppl": 2.5709,
|
| 1398 |
+
"step": 200000
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 2.79,
|
| 1402 |
+
"eval_loss": 4.141200542449951,
|
| 1403 |
+
"eval_nf_loss": 2.539963483810425,
|
| 1404 |
+
"eval_perplexity": 5.245651721954346,
|
| 1405 |
+
"eval_runtime": 622.9626,
|
| 1406 |
+
"eval_samples_per_second": 21.459,
|
| 1407 |
+
"eval_steps_per_second": 2.682,
|
| 1408 |
+
"step": 200000
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 2.81,
|
| 1412 |
+
"gate_score": 0.0442,
|
| 1413 |
+
"learning_rate": 4.727744104889869e-06,
|
| 1414 |
+
"loss": 3.9924,
|
| 1415 |
+
"nf_loss": 3.1183,
|
| 1416 |
+
"ppl": 2.5982,
|
| 1417 |
+
"step": 202000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.84,
|
| 1421 |
+
"gate_score": 0.0437,
|
| 1422 |
+
"learning_rate": 4.021673095127087e-06,
|
| 1423 |
+
"loss": 3.982,
|
| 1424 |
+
"nf_loss": 3.0969,
|
| 1425 |
+
"ppl": 2.6167,
|
| 1426 |
+
"step": 204000
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.84,
|
| 1430 |
+
"eval_loss": 4.099726676940918,
|
| 1431 |
+
"eval_nf_loss": 2.5044403076171875,
|
| 1432 |
+
"eval_perplexity": 5.212295055389404,
|
| 1433 |
+
"eval_runtime": 628.0641,
|
| 1434 |
+
"eval_samples_per_second": 21.284,
|
| 1435 |
+
"eval_steps_per_second": 2.661,
|
| 1436 |
+
"step": 204000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 2.87,
|
| 1440 |
+
"gate_score": 0.044,
|
| 1441 |
+
"learning_rate": 3.315248873253366e-06,
|
| 1442 |
+
"loss": 3.9519,
|
| 1443 |
+
"nf_loss": 3.0778,
|
| 1444 |
+
"ppl": 2.5849,
|
| 1445 |
+
"step": 206000
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.9,
|
| 1449 |
+
"gate_score": 0.0427,
|
| 1450 |
+
"learning_rate": 2.6091778634905835e-06,
|
| 1451 |
+
"loss": 3.9487,
|
| 1452 |
+
"nf_loss": 3.082,
|
| 1453 |
+
"ppl": 2.5664,
|
| 1454 |
+
"step": 208000
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 2.9,
|
| 1458 |
+
"eval_loss": 4.066439151763916,
|
| 1459 |
+
"eval_nf_loss": 2.474716901779175,
|
| 1460 |
+
"eval_perplexity": 5.1935272216796875,
|
| 1461 |
+
"eval_runtime": 625.0657,
|
| 1462 |
+
"eval_samples_per_second": 21.387,
|
| 1463 |
+
"eval_steps_per_second": 2.673,
|
| 1464 |
+
"step": 208000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 2.93,
|
| 1468 |
+
"gate_score": 0.0425,
|
| 1469 |
+
"learning_rate": 1.9031068537278006e-06,
|
| 1470 |
+
"loss": 3.9389,
|
| 1471 |
+
"nf_loss": 3.0634,
|
| 1472 |
+
"ppl": 2.5914,
|
| 1473 |
+
"step": 210000
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.95,
|
| 1477 |
+
"gate_score": 0.0424,
|
| 1478 |
+
"learning_rate": 1.196682631854081e-06,
|
| 1479 |
+
"loss": 3.9603,
|
| 1480 |
+
"nf_loss": 3.0926,
|
| 1481 |
+
"ppl": 2.5703,
|
| 1482 |
+
"step": 212000
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 2.95,
|
| 1486 |
+
"eval_loss": 4.057208061218262,
|
| 1487 |
+
"eval_nf_loss": 2.4683539867401123,
|
| 1488 |
+
"eval_perplexity": 5.177720546722412,
|
| 1489 |
+
"eval_runtime": 618.2401,
|
| 1490 |
+
"eval_samples_per_second": 21.623,
|
| 1491 |
+
"eval_steps_per_second": 2.703,
|
| 1492 |
+
"step": 212000
|
| 1493 |
+
}
|
| 1494 |
+
],
|
| 1495 |
+
"max_steps": 215337,
|
| 1496 |
+
"num_train_epochs": 3,
|
| 1497 |
+
"total_flos": 2.1693560999113897e+18,
|
| 1498 |
+
"trial_name": null,
|
| 1499 |
+
"trial_params": null
|
| 1500 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|