wangrongsheng commited on
Commit
c2d8633
1 Parent(s): 46ab33d
README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "../internlm-chat-20b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8692759d0e4907942aecd47602d501e3c0d99b8481710db9efa4c347ae2bbecb
3
+ size 39407885
all_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "train_loss": 1.4111853902105498,
4
+ "train_runtime": 51425.1791,
5
+ "train_samples_per_second": 2.684,
6
+ "train_steps_per_second": 0.042
7
+ }
checkpoint-1000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-1000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "../internlm-chat-20b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-1000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4998260bd526a9c3bee94abe2e54be3fec53559ce12bc659c21899480083d3e6
3
+ size 39407885
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5fa4225c75d79ec4376d38b077a7cdff57bb875e08cc33c5d567919695684878
3
+ size 78784709
checkpoint-1000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb09c0ab810491cec8dcbb357118e13006f5424ce1d217c048964a2052a80e9f
3
+ size 18679
checkpoint-1000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:addb671ff88a73231dd9b55b4517a019907c186cc25e81845c679685a9459723
3
+ size 18679
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3cf3e0db7ad9e9ba5dd577b3481b4ddbb58c5600860f207a1a04a3a4ee97a42
3
+ size 627
checkpoint-1000/special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<eoa>"
4
+ ],
5
+ "bos_token": "<s>",
6
+ "eos_token": "</s>",
7
+ "pad_token": "</s>",
8
+ "unk_token": "<unk>"
9
+ }
checkpoint-1000/tokenization_internlm.py ADDED
@@ -0,0 +1,242 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved.
3
+ #
4
+ # This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
5
+ # and OPT implementations in this library. It has been modified from its
6
+ # original forms to accommodate minor architectural differences compared
7
+ # to GPT-NeoX and OPT used by the Meta AI team that trained the model.
8
+ #
9
+ # Licensed under the Apache License, Version 2.0 (the "License");
10
+ # you may not use this file except in compliance with the License.
11
+ # You may obtain a copy of the License at
12
+ #
13
+ # http://www.apache.org/licenses/LICENSE-2.0
14
+ #
15
+ # Unless required by applicable law or agreed to in writing, software
16
+ # distributed under the License is distributed on an "AS IS" BASIS,
17
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
18
+ # See the License for the specific language governing permissions and
19
+ # limitations under the License.
20
+
21
+ """Tokenization classes for IntermLM."""
22
+ import os
23
+ from shutil import copyfile
24
+ from typing import Any, Dict, List, Optional, Tuple
25
+
26
+ import sentencepiece as spm
27
+
28
+ from transformers.tokenization_utils import PreTrainedTokenizer
29
+ from transformers.utils import logging
30
+
31
+
32
+ logger = logging.get_logger(__name__)
33
+
34
+ VOCAB_FILES_NAMES = {"vocab_file": "./tokenizer.model"}
35
+
36
+ PRETRAINED_VOCAB_FILES_MAP = {}
37
+
38
+
39
+ class InternLMTokenizer(PreTrainedTokenizer):
40
+ """
41
+ Construct a InternLM tokenizer. Based on byte-level Byte-Pair-Encoding.
42
+
43
+ Args:
44
+ vocab_file (`str`):
45
+ Path to the vocabulary file.
46
+ """
47
+
48
+ vocab_files_names = VOCAB_FILES_NAMES
49
+ pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
50
+ model_input_names = ["input_ids", "attention_mask"]
51
+ _auto_class = "AutoTokenizer"
52
+
53
+ def __init__(
54
+ self,
55
+ vocab_file,
56
+ unk_token="<unk>",
57
+ bos_token="<s>",
58
+ eos_token="</s>",
59
+ pad_token="</s>",
60
+ sp_model_kwargs: Optional[Dict[str, Any]] = None,
61
+ add_bos_token=True,
62
+ add_eos_token=False,
63
+ decode_with_prefix_space=False,
64
+ clean_up_tokenization_spaces=False,
65
+ **kwargs,
66
+ ):
67
+ self.sp_model_kwargs = {} if sp_model_kwargs is None else sp_model_kwargs
68
+ super().__init__(
69
+ bos_token=bos_token,
70
+ eos_token=eos_token,
71
+ unk_token=unk_token,
72
+ pad_token=pad_token,
73
+ clean_up_tokenization_spaces=clean_up_tokenization_spaces,
74
+ **kwargs,
75
+ )
76
+ self.vocab_file = vocab_file
77
+ self.add_bos_token = add_bos_token
78
+ self.add_eos_token = add_eos_token
79
+ self.decode_with_prefix_space = decode_with_prefix_space
80
+ self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
81
+ self.sp_model.Load(vocab_file)
82
+ self._no_prefix_space_tokens = None
83
+
84
+ """ Initialisation"""
85
+
86
+ @property
87
+ def no_prefix_space_tokens(self):
88
+ if self._no_prefix_space_tokens is None:
89
+ vocab = self.convert_ids_to_tokens(list(range(self.vocab_size)))
90
+ self._no_prefix_space_tokens = {i for i, tok in enumerate(vocab) if not tok.startswith("▁")}
91
+ return self._no_prefix_space_tokens
92
+
93
+ @property
94
+ def vocab_size(self):
95
+ """Returns vocab size"""
96
+ return self.sp_model.get_piece_size()
97
+
98
+ @property
99
+ def bos_token_id(self) -> Optional[int]:
100
+ return self.sp_model.bos_id()
101
+
102
+ @property
103
+ def eos_token_id(self) -> Optional[int]:
104
+ return self.sp_model.eos_id()
105
+
106
+ def get_vocab(self):
107
+ """Returns vocab as a dict"""
108
+ vocab = {self.convert_ids_to_tokens(i): i for i in range(self.vocab_size)}
109
+ vocab.update(self.added_tokens_encoder)
110
+ return vocab
111
+
112
+ def _tokenize(self, text):
113
+ """Returns a tokenized string."""
114
+ return self.sp_model.encode(text, out_type=str)
115
+
116
+ def _convert_token_to_id(self, token):
117
+ """Converts a token (str) in an id using the vocab."""
118
+ return self.sp_model.piece_to_id(token)
119
+
120
+ def _convert_id_to_token(self, index):
121
+ """Converts an index (integer) in a token (str) using the vocab."""
122
+ token = self.sp_model.IdToPiece(index)
123
+ return token
124
+
125
+ def _maybe_add_prefix_space(self, tokens, decoded):
126
+ if tokens and tokens[0] not in self.no_prefix_space_tokens:
127
+ return " " + decoded
128
+ else:
129
+ return decoded
130
+
131
+ def convert_tokens_to_string(self, tokens):
132
+ """Converts a sequence of tokens (string) in a single string."""
133
+ current_sub_tokens = []
134
+ out_string = ""
135
+ prev_is_special = False
136
+ for token in tokens:
137
+ # make sure that special tokens are not decoded using sentencepiece model
138
+ if token in self.all_special_tokens:
139
+ if not prev_is_special:
140
+ out_string += " "
141
+ out_string += self.sp_model.decode(current_sub_tokens) + token
142
+ prev_is_special = True
143
+ current_sub_tokens = []
144
+ else:
145
+ current_sub_tokens.append(token)
146
+ prev_is_special = False
147
+ out_string += self.sp_model.decode(current_sub_tokens)
148
+ out_string = self.clean_up_tokenization(out_string)
149
+ out_string = self._maybe_add_prefix_space(tokens=tokens, decoded=out_string)
150
+ return out_string[1:]
151
+
152
+ def save_vocabulary(self, save_directory, filename_prefix: Optional[str] = None) -> Tuple[str]:
153
+ """
154
+ Save the vocabulary and special tokens file to a directory.
155
+
156
+ Args:
157
+ save_directory (`str`):
158
+ The directory in which to save the vocabulary.
159
+
160
+ Returns:
161
+ `Tuple(str)`: Paths to the files saved.
162
+ """
163
+ if not os.path.isdir(save_directory):
164
+ logger.error(f"Vocabulary path ({save_directory}) should be a directory")
165
+ return
166
+ out_vocab_file = os.path.join(
167
+ save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
168
+ )
169
+
170
+ if os.path.abspath(self.vocab_file) != os.path.abspath(out_vocab_file) and os.path.isfile(self.vocab_file):
171
+ copyfile(self.vocab_file, out_vocab_file)
172
+ elif not os.path.isfile(self.vocab_file):
173
+ with open(out_vocab_file, "wb") as fi:
174
+ content_spiece_model = self.sp_model.serialized_model_proto()
175
+ fi.write(content_spiece_model)
176
+
177
+ return (out_vocab_file,)
178
+
179
+ def build_inputs_with_special_tokens(self, token_ids_0, token_ids_1=None):
180
+ if self.add_bos_token:
181
+ bos_token_ids = [self.bos_token_id]
182
+ else:
183
+ bos_token_ids = []
184
+
185
+ output = bos_token_ids + token_ids_0
186
+
187
+ if token_ids_1 is not None:
188
+ output = output + token_ids_1
189
+
190
+ if self.add_eos_token:
191
+ output = output + [self.eos_token_id]
192
+
193
+ return output
194
+
195
+ def get_special_tokens_mask(
196
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
197
+ ) -> List[int]:
198
+ """
199
+ Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
200
+ special tokens using the tokenizer `prepare_for_model` method.
201
+
202
+ Args:
203
+ token_ids_0 (`List[int]`):
204
+ List of IDs.
205
+ token_ids_1 (`List[int]`, *optional*):
206
+ Optional second list of IDs for sequence pairs.
207
+ already_has_special_tokens (`bool`, *optional*, defaults to `False`):
208
+ Whether or not the token list is already formatted with special tokens for the model.
209
+
210
+ Returns:
211
+ `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
212
+ """
213
+ if already_has_special_tokens:
214
+ return super().get_special_tokens_mask(
215
+ token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
216
+ )
217
+
218
+ if token_ids_1 is None:
219
+ return [1] + ([0] * len(token_ids_0)) + [1]
220
+ return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]
221
+
222
+ def create_token_type_ids_from_sequences(
223
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
224
+ ) -> List[int]:
225
+ """
226
+ Create a mask from the two sequences passed to be used in a sequence-pair classification task. T5 does not make
227
+ use of token type ids, therefore a list of zeros is returned.
228
+
229
+ Args:
230
+ token_ids_0 (`List[int]`):
231
+ List of IDs.
232
+ token_ids_1 (`List[int]`, *optional*):
233
+ Optional second list of IDs for sequence pairs.
234
+
235
+ Returns:
236
+ `List[int]`: List of zeros.
237
+ """
238
+ eos = [self.eos_token_id]
239
+
240
+ if token_ids_1 is None:
241
+ return len(token_ids_0 + eos) * [0]
242
+ return len(token_ids_0 + eos + token_ids_1 + eos) * [0]
checkpoint-1000/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aab622d98c98677a1a51f969e25765154487bf3e85c7819db105db2fcacba83f
3
+ size 1658691
checkpoint-1000/tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_map": {
3
+ "AutoTokenizer": [
4
+ "tokenization_internlm.InternLMTokenizer",
5
+ null
6
+ ]
7
+ },
8
+ "bos_token": "<s>",
9
+ "clean_up_tokenization_spaces": false,
10
+ "eos_token": "</s>",
11
+ "model_max_length": 1000000000000000019884624838656,
12
+ "pad_token": "</s>",
13
+ "padding_side": "right",
14
+ "tokenizer_class": "InternLMTokenizer",
15
+ "unk_token": "<unk>"
16
+ }
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,616 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9272137227630969,
5
+ "global_step": 1000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 4.999734597774032e-05,
13
+ "loss": 1.7592,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.998938447446803e-05,
19
+ "loss": 1.6546,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.997611718058365e-05,
25
+ "loss": 1.5632,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.9957546913022665e-05,
31
+ "loss": 1.5245,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.993367761465736e-05,
37
+ "loss": 1.5049,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 4.9904514353459654e-05,
43
+ "loss": 1.492,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "learning_rate": 4.9870063321425105e-05,
49
+ "loss": 1.4803,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.07,
54
+ "learning_rate": 4.983033183325818e-05,
55
+ "loss": 1.4827,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.08,
60
+ "learning_rate": 4.97853283248192e-05,
61
+ "loss": 1.4866,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.09,
66
+ "learning_rate": 4.973506235133323e-05,
67
+ "loss": 1.486,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.1,
72
+ "learning_rate": 4.967954458536126e-05,
73
+ "loss": 1.4822,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.11,
78
+ "learning_rate": 4.9618786814534226e-05,
79
+ "loss": 1.4801,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.12,
84
+ "learning_rate": 4.955280193905022e-05,
85
+ "loss": 1.4764,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.13,
90
+ "learning_rate": 4.948160396893553e-05,
91
+ "loss": 1.4662,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.14,
96
+ "learning_rate": 4.9405208021069946e-05,
97
+ "loss": 1.4627,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.15,
102
+ "learning_rate": 4.9323630315977156e-05,
103
+ "loss": 1.4543,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.16,
108
+ "learning_rate": 4.9236888174380784e-05,
109
+ "loss": 1.4508,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.17,
114
+ "learning_rate": 4.91450000135268e-05,
115
+ "loss": 1.464,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.18,
120
+ "learning_rate": 4.9047985343273154e-05,
121
+ "loss": 1.4705,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.19,
126
+ "learning_rate": 4.894586476194739e-05,
127
+ "loss": 1.4588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.19,
132
+ "learning_rate": 4.883865995197319e-05,
133
+ "loss": 1.4532,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.2,
138
+ "learning_rate": 4.8726393675266716e-05,
139
+ "loss": 1.4629,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.21,
144
+ "learning_rate": 4.860908976840376e-05,
145
+ "loss": 1.4429,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.22,
150
+ "learning_rate": 4.848677313755872e-05,
151
+ "loss": 1.441,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.23,
156
+ "learning_rate": 4.835946975321647e-05,
157
+ "loss": 1.4496,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.24,
162
+ "learning_rate": 4.822720664465827e-05,
163
+ "loss": 1.4459,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.25,
168
+ "learning_rate": 4.809001189422287e-05,
169
+ "loss": 1.4458,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.26,
174
+ "learning_rate": 4.794791463134399e-05,
175
+ "loss": 1.4484,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.27,
180
+ "learning_rate": 4.780094502636552e-05,
181
+ "loss": 1.4541,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.28,
186
+ "learning_rate": 4.764913428413572e-05,
187
+ "loss": 1.437,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.29,
192
+ "learning_rate": 4.7492514637381727e-05,
193
+ "loss": 1.4368,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.3,
198
+ "learning_rate": 4.733111933986583e-05,
199
+ "loss": 1.4449,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.31,
204
+ "learning_rate": 4.716498265932501e-05,
205
+ "loss": 1.4256,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.32,
210
+ "learning_rate": 4.699413987019512e-05,
211
+ "loss": 1.4378,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.32,
216
+ "learning_rate": 4.681862724612141e-05,
217
+ "loss": 1.454,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.33,
222
+ "learning_rate": 4.663848205225674e-05,
223
+ "loss": 1.4207,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.34,
228
+ "learning_rate": 4.645374253734949e-05,
229
+ "loss": 1.4286,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.35,
234
+ "learning_rate": 4.626444792562244e-05,
235
+ "loss": 1.4244,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.36,
240
+ "learning_rate": 4.607063840844463e-05,
241
+ "loss": 1.4342,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.37,
246
+ "learning_rate": 4.587235513579791e-05,
247
+ "loss": 1.4241,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.38,
252
+ "learning_rate": 4.5669640207539786e-05,
253
+ "loss": 1.443,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.39,
258
+ "learning_rate": 4.546253666446484e-05,
259
+ "loss": 1.4281,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.4,
264
+ "learning_rate": 4.525108847916614e-05,
265
+ "loss": 1.4193,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.41,
270
+ "learning_rate": 4.503534054669892e-05,
271
+ "loss": 1.4154,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.42,
276
+ "learning_rate": 4.481533867504841e-05,
277
+ "loss": 1.4273,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.43,
282
+ "learning_rate": 4.4591129575403765e-05,
283
+ "loss": 1.4329,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.44,
288
+ "learning_rate": 4.43627608522403e-05,
289
+ "loss": 1.4102,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.45,
294
+ "learning_rate": 4.4130280993211974e-05,
295
+ "loss": 1.4244,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.45,
300
+ "learning_rate": 4.389373935885646e-05,
301
+ "loss": 1.4407,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.46,
306
+ "learning_rate": 4.365318617211479e-05,
307
+ "loss": 1.4282,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.47,
312
+ "learning_rate": 4.340867250766794e-05,
313
+ "loss": 1.4269,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.48,
318
+ "learning_rate": 4.316025028109258e-05,
319
+ "loss": 1.43,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.49,
324
+ "learning_rate": 4.2907972237838225e-05,
325
+ "loss": 1.4271,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.5,
330
+ "learning_rate": 4.2651891942028274e-05,
331
+ "loss": 1.4321,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.51,
336
+ "learning_rate": 4.239206376508717e-05,
337
+ "loss": 1.4083,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.52,
342
+ "learning_rate": 4.212854287419611e-05,
343
+ "loss": 1.4168,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.53,
348
+ "learning_rate": 4.1861385220579934e-05,
349
+ "loss": 1.4144,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.54,
354
+ "learning_rate": 4.1590647527627404e-05,
355
+ "loss": 1.4106,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.55,
360
+ "learning_rate": 4.131638727884762e-05,
361
+ "loss": 1.4202,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.56,
366
+ "learning_rate": 4.103866270566498e-05,
367
+ "loss": 1.417,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.57,
372
+ "learning_rate": 4.075753277505544e-05,
373
+ "loss": 1.4324,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.57,
378
+ "learning_rate": 4.0473057177026484e-05,
379
+ "loss": 1.4147,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.58,
384
+ "learning_rate": 4.018529631194369e-05,
385
+ "loss": 1.4269,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.59,
390
+ "learning_rate": 3.989431127770635e-05,
391
+ "loss": 1.4133,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.6,
396
+ "learning_rate": 3.960016385677513e-05,
397
+ "loss": 1.4173,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.61,
402
+ "learning_rate": 3.9302916503054246e-05,
403
+ "loss": 1.4007,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.62,
408
+ "learning_rate": 3.9002632328631164e-05,
409
+ "loss": 1.4117,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.63,
414
+ "learning_rate": 3.8699375090376534e-05,
415
+ "loss": 1.4247,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.64,
420
+ "learning_rate": 3.8393209176407223e-05,
421
+ "loss": 1.4312,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.65,
426
+ "learning_rate": 3.8084199592415305e-05,
427
+ "loss": 1.4136,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.66,
432
+ "learning_rate": 3.777241194786591e-05,
433
+ "loss": 1.4132,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.67,
438
+ "learning_rate": 3.745791244206697e-05,
439
+ "loss": 1.4145,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.68,
444
+ "learning_rate": 3.714076785011359e-05,
445
+ "loss": 1.3951,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.69,
450
+ "learning_rate": 3.682104550871031e-05,
451
+ "loss": 1.4139,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.7,
456
+ "learning_rate": 3.649881330187401e-05,
457
+ "loss": 1.4078,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.7,
462
+ "learning_rate": 3.617413964652067e-05,
463
+ "loss": 1.4048,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.71,
468
+ "learning_rate": 3.5847093477938956e-05,
469
+ "loss": 1.4115,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.72,
474
+ "learning_rate": 3.551774423515378e-05,
475
+ "loss": 1.4018,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.73,
480
+ "learning_rate": 3.518616184618288e-05,
481
+ "loss": 1.4091,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.74,
486
+ "learning_rate": 3.4852416713189526e-05,
487
+ "loss": 1.4135,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.75,
492
+ "learning_rate": 3.4516579697534705e-05,
493
+ "loss": 1.3977,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.76,
498
+ "learning_rate": 3.417872210473162e-05,
499
+ "loss": 1.4123,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.77,
504
+ "learning_rate": 3.3838915669306034e-05,
505
+ "loss": 1.4096,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.78,
510
+ "learning_rate": 3.349723253956542e-05,
511
+ "loss": 1.4031,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.79,
516
+ "learning_rate": 3.315374526228036e-05,
517
+ "loss": 1.4104,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.8,
522
+ "learning_rate": 3.2808526767281225e-05,
523
+ "loss": 1.3972,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.81,
528
+ "learning_rate": 3.246165035197364e-05,
529
+ "loss": 1.3987,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.82,
534
+ "learning_rate": 3.211318966577581e-05,
535
+ "loss": 1.4197,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.83,
540
+ "learning_rate": 3.176321869448116e-05,
541
+ "loss": 1.4169,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.83,
546
+ "learning_rate": 3.1411811744549536e-05,
547
+ "loss": 1.4055,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.84,
552
+ "learning_rate": 3.105904342733032e-05,
553
+ "loss": 1.402,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.85,
558
+ "learning_rate": 3.070498864322081e-05,
559
+ "loss": 1.4148,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.86,
564
+ "learning_rate": 3.034972256576328e-05,
565
+ "loss": 1.4007,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.87,
570
+ "learning_rate": 2.999332062568395e-05,
571
+ "loss": 1.3926,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.88,
576
+ "learning_rate": 2.9635858494877384e-05,
577
+ "loss": 1.4099,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.89,
582
+ "learning_rate": 2.9277412070339782e-05,
583
+ "loss": 1.401,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.9,
588
+ "learning_rate": 2.891805745805429e-05,
589
+ "loss": 1.3995,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.91,
594
+ "learning_rate": 2.8557870956832132e-05,
595
+ "loss": 1.4108,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.92,
600
+ "learning_rate": 2.8196929042112652e-05,
601
+ "loss": 1.4006,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.93,
606
+ "learning_rate": 2.783530834972594e-05,
607
+ "loss": 1.4045,
608
+ "step": 1000
609
+ }
610
+ ],
611
+ "max_steps": 2156,
612
+ "num_train_epochs": 2,
613
+ "total_flos": 2.1503929922743173e+18,
614
+ "trial_name": null,
615
+ "trial_params": null
616
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:918a94b2b648985fdb03a840052eff79b0035343fd87ee2a0dc751e904fad0c4
3
+ size 4091
checkpoint-2000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-2000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "../internlm-chat-20b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-2000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62fc465988009c55add9449eaec20a52187ade9f6c05f11dd25215d9acc02f03
3
+ size 39407885
checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11a5b792ce3d6044eacf0edb6fe8dc6a5669b69aac9d1addf905381b5a71b29d
3
+ size 78784709
checkpoint-2000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:10784e74418dde082729bc4ccae91bab345cfe23991999f469fcec392df28c9c
3
+ size 18679
checkpoint-2000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:876ffe11f58c3eb52f532fa7b0a7081fa6782d9e569651792ca2ae30d57585f1
3
+ size 18679
checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd0b604a79d16282fd4dcd9803f6bbe0802175560169f2fdb3fe364043fcb56a
3
+ size 627
checkpoint-2000/special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<eoa>"
4
+ ],
5
+ "bos_token": "<s>",
6
+ "eos_token": "</s>",
7
+ "pad_token": "</s>",
8
+ "unk_token": "<unk>"
9
+ }
checkpoint-2000/tokenization_internlm.py ADDED
@@ -0,0 +1,242 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved.
3
+ #
4
+ # This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
5
+ # and OPT implementations in this library. It has been modified from its
6
+ # original forms to accommodate minor architectural differences compared
7
+ # to GPT-NeoX and OPT used by the Meta AI team that trained the model.
8
+ #
9
+ # Licensed under the Apache License, Version 2.0 (the "License");
10
+ # you may not use this file except in compliance with the License.
11
+ # You may obtain a copy of the License at
12
+ #
13
+ # http://www.apache.org/licenses/LICENSE-2.0
14
+ #
15
+ # Unless required by applicable law or agreed to in writing, software
16
+ # distributed under the License is distributed on an "AS IS" BASIS,
17
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
18
+ # See the License for the specific language governing permissions and
19
+ # limitations under the License.
20
+
21
+ """Tokenization classes for IntermLM."""
22
+ import os
23
+ from shutil import copyfile
24
+ from typing import Any, Dict, List, Optional, Tuple
25
+
26
+ import sentencepiece as spm
27
+
28
+ from transformers.tokenization_utils import PreTrainedTokenizer
29
+ from transformers.utils import logging
30
+
31
+
32
+ logger = logging.get_logger(__name__)
33
+
34
+ VOCAB_FILES_NAMES = {"vocab_file": "./tokenizer.model"}
35
+
36
+ PRETRAINED_VOCAB_FILES_MAP = {}
37
+
38
+
39
+ class InternLMTokenizer(PreTrainedTokenizer):
40
+ """
41
+ Construct a InternLM tokenizer. Based on byte-level Byte-Pair-Encoding.
42
+
43
+ Args:
44
+ vocab_file (`str`):
45
+ Path to the vocabulary file.
46
+ """
47
+
48
+ vocab_files_names = VOCAB_FILES_NAMES
49
+ pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
50
+ model_input_names = ["input_ids", "attention_mask"]
51
+ _auto_class = "AutoTokenizer"
52
+
53
+ def __init__(
54
+ self,
55
+ vocab_file,
56
+ unk_token="<unk>",
57
+ bos_token="<s>",
58
+ eos_token="</s>",
59
+ pad_token="</s>",
60
+ sp_model_kwargs: Optional[Dict[str, Any]] = None,
61
+ add_bos_token=True,
62
+ add_eos_token=False,
63
+ decode_with_prefix_space=False,
64
+ clean_up_tokenization_spaces=False,
65
+ **kwargs,
66
+ ):
67
+ self.sp_model_kwargs = {} if sp_model_kwargs is None else sp_model_kwargs
68
+ super().__init__(
69
+ bos_token=bos_token,
70
+ eos_token=eos_token,
71
+ unk_token=unk_token,
72
+ pad_token=pad_token,
73
+ clean_up_tokenization_spaces=clean_up_tokenization_spaces,
74
+ **kwargs,
75
+ )
76
+ self.vocab_file = vocab_file
77
+ self.add_bos_token = add_bos_token
78
+ self.add_eos_token = add_eos_token
79
+ self.decode_with_prefix_space = decode_with_prefix_space
80
+ self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
81
+ self.sp_model.Load(vocab_file)
82
+ self._no_prefix_space_tokens = None
83
+
84
+ """ Initialisation"""
85
+
86
+ @property
87
+ def no_prefix_space_tokens(self):
88
+ if self._no_prefix_space_tokens is None:
89
+ vocab = self.convert_ids_to_tokens(list(range(self.vocab_size)))
90
+ self._no_prefix_space_tokens = {i for i, tok in enumerate(vocab) if not tok.startswith("▁")}
91
+ return self._no_prefix_space_tokens
92
+
93
+ @property
94
+ def vocab_size(self):
95
+ """Returns vocab size"""
96
+ return self.sp_model.get_piece_size()
97
+
98
+ @property
99
+ def bos_token_id(self) -> Optional[int]:
100
+ return self.sp_model.bos_id()
101
+
102
+ @property
103
+ def eos_token_id(self) -> Optional[int]:
104
+ return self.sp_model.eos_id()
105
+
106
+ def get_vocab(self):
107
+ """Returns vocab as a dict"""
108
+ vocab = {self.convert_ids_to_tokens(i): i for i in range(self.vocab_size)}
109
+ vocab.update(self.added_tokens_encoder)
110
+ return vocab
111
+
112
+ def _tokenize(self, text):
113
+ """Returns a tokenized string."""
114
+ return self.sp_model.encode(text, out_type=str)
115
+
116
+ def _convert_token_to_id(self, token):
117
+ """Converts a token (str) in an id using the vocab."""
118
+ return self.sp_model.piece_to_id(token)
119
+
120
+ def _convert_id_to_token(self, index):
121
+ """Converts an index (integer) in a token (str) using the vocab."""
122
+ token = self.sp_model.IdToPiece(index)
123
+ return token
124
+
125
+ def _maybe_add_prefix_space(self, tokens, decoded):
126
+ if tokens and tokens[0] not in self.no_prefix_space_tokens:
127
+ return " " + decoded
128
+ else:
129
+ return decoded
130
+
131
+ def convert_tokens_to_string(self, tokens):
132
+ """Converts a sequence of tokens (string) in a single string."""
133
+ current_sub_tokens = []
134
+ out_string = ""
135
+ prev_is_special = False
136
+ for token in tokens:
137
+ # make sure that special tokens are not decoded using sentencepiece model
138
+ if token in self.all_special_tokens:
139
+ if not prev_is_special:
140
+ out_string += " "
141
+ out_string += self.sp_model.decode(current_sub_tokens) + token
142
+ prev_is_special = True
143
+ current_sub_tokens = []
144
+ else:
145
+ current_sub_tokens.append(token)
146
+ prev_is_special = False
147
+ out_string += self.sp_model.decode(current_sub_tokens)
148
+ out_string = self.clean_up_tokenization(out_string)
149
+ out_string = self._maybe_add_prefix_space(tokens=tokens, decoded=out_string)
150
+ return out_string[1:]
151
+
152
+ def save_vocabulary(self, save_directory, filename_prefix: Optional[str] = None) -> Tuple[str]:
153
+ """
154
+ Save the vocabulary and special tokens file to a directory.
155
+
156
+ Args:
157
+ save_directory (`str`):
158
+ The directory in which to save the vocabulary.
159
+
160
+ Returns:
161
+ `Tuple(str)`: Paths to the files saved.
162
+ """
163
+ if not os.path.isdir(save_directory):
164
+ logger.error(f"Vocabulary path ({save_directory}) should be a directory")
165
+ return
166
+ out_vocab_file = os.path.join(
167
+ save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
168
+ )
169
+
170
+ if os.path.abspath(self.vocab_file) != os.path.abspath(out_vocab_file) and os.path.isfile(self.vocab_file):
171
+ copyfile(self.vocab_file, out_vocab_file)
172
+ elif not os.path.isfile(self.vocab_file):
173
+ with open(out_vocab_file, "wb") as fi:
174
+ content_spiece_model = self.sp_model.serialized_model_proto()
175
+ fi.write(content_spiece_model)
176
+
177
+ return (out_vocab_file,)
178
+
179
+ def build_inputs_with_special_tokens(self, token_ids_0, token_ids_1=None):
180
+ if self.add_bos_token:
181
+ bos_token_ids = [self.bos_token_id]
182
+ else:
183
+ bos_token_ids = []
184
+
185
+ output = bos_token_ids + token_ids_0
186
+
187
+ if token_ids_1 is not None:
188
+ output = output + token_ids_1
189
+
190
+ if self.add_eos_token:
191
+ output = output + [self.eos_token_id]
192
+
193
+ return output
194
+
195
+ def get_special_tokens_mask(
196
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
197
+ ) -> List[int]:
198
+ """
199
+ Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
200
+ special tokens using the tokenizer `prepare_for_model` method.
201
+
202
+ Args:
203
+ token_ids_0 (`List[int]`):
204
+ List of IDs.
205
+ token_ids_1 (`List[int]`, *optional*):
206
+ Optional second list of IDs for sequence pairs.
207
+ already_has_special_tokens (`bool`, *optional*, defaults to `False`):
208
+ Whether or not the token list is already formatted with special tokens for the model.
209
+
210
+ Returns:
211
+ `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
212
+ """
213
+ if already_has_special_tokens:
214
+ return super().get_special_tokens_mask(
215
+ token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
216
+ )
217
+
218
+ if token_ids_1 is None:
219
+ return [1] + ([0] * len(token_ids_0)) + [1]
220
+ return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]
221
+
222
+ def create_token_type_ids_from_sequences(
223
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
224
+ ) -> List[int]:
225
+ """
226
+ Create a mask from the two sequences passed to be used in a sequence-pair classification task. T5 does not make
227
+ use of token type ids, therefore a list of zeros is returned.
228
+
229
+ Args:
230
+ token_ids_0 (`List[int]`):
231
+ List of IDs.
232
+ token_ids_1 (`List[int]`, *optional*):
233
+ Optional second list of IDs for sequence pairs.
234
+
235
+ Returns:
236
+ `List[int]`: List of zeros.
237
+ """
238
+ eos = [self.eos_token_id]
239
+
240
+ if token_ids_1 is None:
241
+ return len(token_ids_0 + eos) * [0]
242
+ return len(token_ids_0 + eos + token_ids_1 + eos) * [0]
checkpoint-2000/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aab622d98c98677a1a51f969e25765154487bf3e85c7819db105db2fcacba83f
3
+ size 1658691
checkpoint-2000/tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_map": {
3
+ "AutoTokenizer": [
4
+ "tokenization_internlm.InternLMTokenizer",
5
+ null
6
+ ]
7
+ },
8
+ "bos_token": "<s>",
9
+ "clean_up_tokenization_spaces": false,
10
+ "eos_token": "</s>",
11
+ "model_max_length": 1000000000000000019884624838656,
12
+ "pad_token": "</s>",
13
+ "padding_side": "right",
14
+ "tokenizer_class": "InternLMTokenizer",
15
+ "unk_token": "<unk>"
16
+ }
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,1216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.8544274455261938,
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": 0.01,
12
+ "learning_rate": 4.999734597774032e-05,
13
+ "loss": 1.7592,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.998938447446803e-05,
19
+ "loss": 1.6546,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.997611718058365e-05,
25
+ "loss": 1.5632,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.9957546913022665e-05,
31
+ "loss": 1.5245,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.993367761465736e-05,
37
+ "loss": 1.5049,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 4.9904514353459654e-05,
43
+ "loss": 1.492,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "learning_rate": 4.9870063321425105e-05,
49
+ "loss": 1.4803,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.07,
54
+ "learning_rate": 4.983033183325818e-05,
55
+ "loss": 1.4827,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.08,
60
+ "learning_rate": 4.97853283248192e-05,
61
+ "loss": 1.4866,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.09,
66
+ "learning_rate": 4.973506235133323e-05,
67
+ "loss": 1.486,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.1,
72
+ "learning_rate": 4.967954458536126e-05,
73
+ "loss": 1.4822,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.11,
78
+ "learning_rate": 4.9618786814534226e-05,
79
+ "loss": 1.4801,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.12,
84
+ "learning_rate": 4.955280193905022e-05,
85
+ "loss": 1.4764,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.13,
90
+ "learning_rate": 4.948160396893553e-05,
91
+ "loss": 1.4662,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.14,
96
+ "learning_rate": 4.9405208021069946e-05,
97
+ "loss": 1.4627,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.15,
102
+ "learning_rate": 4.9323630315977156e-05,
103
+ "loss": 1.4543,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.16,
108
+ "learning_rate": 4.9236888174380784e-05,
109
+ "loss": 1.4508,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.17,
114
+ "learning_rate": 4.91450000135268e-05,
115
+ "loss": 1.464,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.18,
120
+ "learning_rate": 4.9047985343273154e-05,
121
+ "loss": 1.4705,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.19,
126
+ "learning_rate": 4.894586476194739e-05,
127
+ "loss": 1.4588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.19,
132
+ "learning_rate": 4.883865995197319e-05,
133
+ "loss": 1.4532,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.2,
138
+ "learning_rate": 4.8726393675266716e-05,
139
+ "loss": 1.4629,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.21,
144
+ "learning_rate": 4.860908976840376e-05,
145
+ "loss": 1.4429,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.22,
150
+ "learning_rate": 4.848677313755872e-05,
151
+ "loss": 1.441,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.23,
156
+ "learning_rate": 4.835946975321647e-05,
157
+ "loss": 1.4496,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.24,
162
+ "learning_rate": 4.822720664465827e-05,
163
+ "loss": 1.4459,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.25,
168
+ "learning_rate": 4.809001189422287e-05,
169
+ "loss": 1.4458,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.26,
174
+ "learning_rate": 4.794791463134399e-05,
175
+ "loss": 1.4484,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.27,
180
+ "learning_rate": 4.780094502636552e-05,
181
+ "loss": 1.4541,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.28,
186
+ "learning_rate": 4.764913428413572e-05,
187
+ "loss": 1.437,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.29,
192
+ "learning_rate": 4.7492514637381727e-05,
193
+ "loss": 1.4368,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.3,
198
+ "learning_rate": 4.733111933986583e-05,
199
+ "loss": 1.4449,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.31,
204
+ "learning_rate": 4.716498265932501e-05,
205
+ "loss": 1.4256,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.32,
210
+ "learning_rate": 4.699413987019512e-05,
211
+ "loss": 1.4378,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.32,
216
+ "learning_rate": 4.681862724612141e-05,
217
+ "loss": 1.454,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.33,
222
+ "learning_rate": 4.663848205225674e-05,
223
+ "loss": 1.4207,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.34,
228
+ "learning_rate": 4.645374253734949e-05,
229
+ "loss": 1.4286,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.35,
234
+ "learning_rate": 4.626444792562244e-05,
235
+ "loss": 1.4244,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.36,
240
+ "learning_rate": 4.607063840844463e-05,
241
+ "loss": 1.4342,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.37,
246
+ "learning_rate": 4.587235513579791e-05,
247
+ "loss": 1.4241,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.38,
252
+ "learning_rate": 4.5669640207539786e-05,
253
+ "loss": 1.443,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.39,
258
+ "learning_rate": 4.546253666446484e-05,
259
+ "loss": 1.4281,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.4,
264
+ "learning_rate": 4.525108847916614e-05,
265
+ "loss": 1.4193,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.41,
270
+ "learning_rate": 4.503534054669892e-05,
271
+ "loss": 1.4154,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.42,
276
+ "learning_rate": 4.481533867504841e-05,
277
+ "loss": 1.4273,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.43,
282
+ "learning_rate": 4.4591129575403765e-05,
283
+ "loss": 1.4329,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.44,
288
+ "learning_rate": 4.43627608522403e-05,
289
+ "loss": 1.4102,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.45,
294
+ "learning_rate": 4.4130280993211974e-05,
295
+ "loss": 1.4244,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.45,
300
+ "learning_rate": 4.389373935885646e-05,
301
+ "loss": 1.4407,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.46,
306
+ "learning_rate": 4.365318617211479e-05,
307
+ "loss": 1.4282,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.47,
312
+ "learning_rate": 4.340867250766794e-05,
313
+ "loss": 1.4269,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.48,
318
+ "learning_rate": 4.316025028109258e-05,
319
+ "loss": 1.43,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.49,
324
+ "learning_rate": 4.2907972237838225e-05,
325
+ "loss": 1.4271,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.5,
330
+ "learning_rate": 4.2651891942028274e-05,
331
+ "loss": 1.4321,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.51,
336
+ "learning_rate": 4.239206376508717e-05,
337
+ "loss": 1.4083,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.52,
342
+ "learning_rate": 4.212854287419611e-05,
343
+ "loss": 1.4168,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.53,
348
+ "learning_rate": 4.1861385220579934e-05,
349
+ "loss": 1.4144,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.54,
354
+ "learning_rate": 4.1590647527627404e-05,
355
+ "loss": 1.4106,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.55,
360
+ "learning_rate": 4.131638727884762e-05,
361
+ "loss": 1.4202,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.56,
366
+ "learning_rate": 4.103866270566498e-05,
367
+ "loss": 1.417,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.57,
372
+ "learning_rate": 4.075753277505544e-05,
373
+ "loss": 1.4324,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.57,
378
+ "learning_rate": 4.0473057177026484e-05,
379
+ "loss": 1.4147,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.58,
384
+ "learning_rate": 4.018529631194369e-05,
385
+ "loss": 1.4269,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.59,
390
+ "learning_rate": 3.989431127770635e-05,
391
+ "loss": 1.4133,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.6,
396
+ "learning_rate": 3.960016385677513e-05,
397
+ "loss": 1.4173,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.61,
402
+ "learning_rate": 3.9302916503054246e-05,
403
+ "loss": 1.4007,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.62,
408
+ "learning_rate": 3.9002632328631164e-05,
409
+ "loss": 1.4117,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.63,
414
+ "learning_rate": 3.8699375090376534e-05,
415
+ "loss": 1.4247,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.64,
420
+ "learning_rate": 3.8393209176407223e-05,
421
+ "loss": 1.4312,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.65,
426
+ "learning_rate": 3.8084199592415305e-05,
427
+ "loss": 1.4136,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.66,
432
+ "learning_rate": 3.777241194786591e-05,
433
+ "loss": 1.4132,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.67,
438
+ "learning_rate": 3.745791244206697e-05,
439
+ "loss": 1.4145,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.68,
444
+ "learning_rate": 3.714076785011359e-05,
445
+ "loss": 1.3951,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.69,
450
+ "learning_rate": 3.682104550871031e-05,
451
+ "loss": 1.4139,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.7,
456
+ "learning_rate": 3.649881330187401e-05,
457
+ "loss": 1.4078,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.7,
462
+ "learning_rate": 3.617413964652067e-05,
463
+ "loss": 1.4048,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.71,
468
+ "learning_rate": 3.5847093477938956e-05,
469
+ "loss": 1.4115,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.72,
474
+ "learning_rate": 3.551774423515378e-05,
475
+ "loss": 1.4018,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.73,
480
+ "learning_rate": 3.518616184618288e-05,
481
+ "loss": 1.4091,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.74,
486
+ "learning_rate": 3.4852416713189526e-05,
487
+ "loss": 1.4135,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.75,
492
+ "learning_rate": 3.4516579697534705e-05,
493
+ "loss": 1.3977,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.76,
498
+ "learning_rate": 3.417872210473162e-05,
499
+ "loss": 1.4123,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.77,
504
+ "learning_rate": 3.3838915669306034e-05,
505
+ "loss": 1.4096,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.78,
510
+ "learning_rate": 3.349723253956542e-05,
511
+ "loss": 1.4031,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.79,
516
+ "learning_rate": 3.315374526228036e-05,
517
+ "loss": 1.4104,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.8,
522
+ "learning_rate": 3.2808526767281225e-05,
523
+ "loss": 1.3972,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.81,
528
+ "learning_rate": 3.246165035197364e-05,
529
+ "loss": 1.3987,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.82,
534
+ "learning_rate": 3.211318966577581e-05,
535
+ "loss": 1.4197,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.83,
540
+ "learning_rate": 3.176321869448116e-05,
541
+ "loss": 1.4169,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.83,
546
+ "learning_rate": 3.1411811744549536e-05,
547
+ "loss": 1.4055,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.84,
552
+ "learning_rate": 3.105904342733032e-05,
553
+ "loss": 1.402,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.85,
558
+ "learning_rate": 3.070498864322081e-05,
559
+ "loss": 1.4148,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.86,
564
+ "learning_rate": 3.034972256576328e-05,
565
+ "loss": 1.4007,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.87,
570
+ "learning_rate": 2.999332062568395e-05,
571
+ "loss": 1.3926,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.88,
576
+ "learning_rate": 2.9635858494877384e-05,
577
+ "loss": 1.4099,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.89,
582
+ "learning_rate": 2.9277412070339782e-05,
583
+ "loss": 1.401,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.9,
588
+ "learning_rate": 2.891805745805429e-05,
589
+ "loss": 1.3995,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.91,
594
+ "learning_rate": 2.8557870956832132e-05,
595
+ "loss": 1.4108,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.92,
600
+ "learning_rate": 2.8196929042112652e-05,
601
+ "loss": 1.4006,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.93,
606
+ "learning_rate": 2.783530834972594e-05,
607
+ "loss": 1.4045,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.94,
612
+ "learning_rate": 2.7473085659621377e-05,
613
+ "loss": 1.3907,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.95,
618
+ "learning_rate": 2.711033787956555e-05,
619
+ "loss": 1.3872,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.96,
624
+ "learning_rate": 2.6747142028813105e-05,
625
+ "loss": 1.405,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.96,
630
+ "learning_rate": 2.638357522175383e-05,
631
+ "loss": 1.3952,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.97,
636
+ "learning_rate": 2.6019714651539646e-05,
637
+ "loss": 1.3901,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.98,
642
+ "learning_rate": 2.565563757369475e-05,
643
+ "loss": 1.4027,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.99,
648
+ "learning_rate": 2.529142128971268e-05,
649
+ "loss": 1.3909,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 1.0,
654
+ "learning_rate": 2.492714313064342e-05,
655
+ "loss": 1.4039,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 1.01,
660
+ "learning_rate": 2.45628804406744e-05,
661
+ "loss": 1.3964,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 1.02,
666
+ "learning_rate": 2.419871056070862e-05,
667
+ "loss": 1.4076,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 1.03,
672
+ "learning_rate": 2.3834710811943514e-05,
673
+ "loss": 1.3891,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 1.04,
678
+ "learning_rate": 2.3470958479453938e-05,
679
+ "loss": 1.4004,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 1.05,
684
+ "learning_rate": 2.3107530795782877e-05,
685
+ "loss": 1.3879,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 1.06,
690
+ "learning_rate": 2.2744504924543313e-05,
691
+ "loss": 1.3879,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 1.07,
696
+ "learning_rate": 2.23819579440347e-05,
697
+ "loss": 1.3851,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 1.08,
702
+ "learning_rate": 2.2019966830877545e-05,
703
+ "loss": 1.3846,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 1.08,
708
+ "learning_rate": 2.1658608443669635e-05,
709
+ "loss": 1.3936,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 1.09,
714
+ "learning_rate": 2.1297959506667224e-05,
715
+ "loss": 1.3841,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 1.1,
720
+ "learning_rate": 2.0938096593494855e-05,
721
+ "loss": 1.3969,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 1.11,
726
+ "learning_rate": 2.057909611088709e-05,
727
+ "loss": 1.3953,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 1.12,
732
+ "learning_rate": 2.02210342824657e-05,
733
+ "loss": 1.3951,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 1.13,
738
+ "learning_rate": 1.9863987132555706e-05,
739
+ "loss": 1.4012,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 1.14,
744
+ "learning_rate": 1.9508030470043806e-05,
745
+ "loss": 1.3998,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 1.15,
750
+ "learning_rate": 1.915323987228247e-05,
751
+ "loss": 1.3933,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 1.16,
756
+ "learning_rate": 1.8799690669043212e-05,
757
+ "loss": 1.3977,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 1.17,
762
+ "learning_rate": 1.8447457926522454e-05,
763
+ "loss": 1.3818,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 1.18,
768
+ "learning_rate": 1.8096616431403325e-05,
769
+ "loss": 1.4032,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 1.19,
774
+ "learning_rate": 1.7747240674976857e-05,
775
+ "loss": 1.3853,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 1.2,
780
+ "learning_rate": 1.7399404837325796e-05,
781
+ "loss": 1.3988,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 1.21,
786
+ "learning_rate": 1.7053182771574633e-05,
787
+ "loss": 1.3963,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 1.21,
792
+ "learning_rate": 1.6708647988208887e-05,
793
+ "loss": 1.3881,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 1.22,
798
+ "learning_rate": 1.6365873639467315e-05,
799
+ "loss": 1.3804,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 1.23,
804
+ "learning_rate": 1.602493250381003e-05,
805
+ "loss": 1.3952,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 1.24,
810
+ "learning_rate": 1.5685896970466123e-05,
811
+ "loss": 1.4015,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 1.25,
816
+ "learning_rate": 1.534883902406375e-05,
817
+ "loss": 1.3797,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 1.26,
822
+ "learning_rate": 1.5013830229346326e-05,
823
+ "loss": 1.3931,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 1.27,
828
+ "learning_rate": 1.4680941715977722e-05,
829
+ "loss": 1.3946,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 1.28,
834
+ "learning_rate": 1.4350244163439892e-05,
835
+ "loss": 1.3851,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 1.29,
840
+ "learning_rate": 1.4021807786026108e-05,
841
+ "loss": 1.3891,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 1.3,
846
+ "learning_rate": 1.3695702317932862e-05,
847
+ "loss": 1.3958,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 1.31,
852
+ "learning_rate": 1.337199699845387e-05,
853
+ "loss": 1.3909,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 1.32,
858
+ "learning_rate": 1.3050760557279023e-05,
859
+ "loss": 1.3785,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 1.33,
864
+ "learning_rate": 1.2732061199901562e-05,
865
+ "loss": 1.3882,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 1.34,
870
+ "learning_rate": 1.2415966593136547e-05,
871
+ "loss": 1.3872,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 1.34,
876
+ "learning_rate": 1.2102543850753808e-05,
877
+ "loss": 1.4032,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 1.35,
882
+ "learning_rate": 1.1791859519228138e-05,
883
+ "loss": 1.3928,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 1.36,
888
+ "learning_rate": 1.148397956361007e-05,
889
+ "loss": 1.3785,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 1.37,
894
+ "learning_rate": 1.1178969353520018e-05,
895
+ "loss": 1.3797,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 1.38,
900
+ "learning_rate": 1.087689364926897e-05,
901
+ "loss": 1.3966,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 1.39,
906
+ "learning_rate": 1.0577816588108378e-05,
907
+ "loss": 1.3849,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 1.4,
912
+ "learning_rate": 1.0281801670612449e-05,
913
+ "loss": 1.3732,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 1.41,
918
+ "learning_rate": 9.988911747195603e-06,
919
+ "loss": 1.4018,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 1.42,
924
+ "learning_rate": 9.699209004767953e-06,
925
+ "loss": 1.3902,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 1.43,
930
+ "learning_rate": 9.412754953531663e-06,
931
+ "loss": 1.3991,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 1.44,
936
+ "learning_rate": 9.129610413921005e-06,
937
+ "loss": 1.3824,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 1.45,
942
+ "learning_rate": 8.849835503688847e-06,
943
+ "loss": 1.3904,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 1.46,
948
+ "learning_rate": 8.57348962514234e-06,
949
+ "loss": 1.3937,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 1.46,
954
+ "learning_rate": 8.30063145253053e-06,
955
+ "loss": 1.39,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 1.47,
960
+ "learning_rate": 8.031318919586523e-06,
961
+ "loss": 1.3773,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 1.48,
966
+ "learning_rate": 7.7656092072269e-06,
967
+ "loss": 1.3875,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 1.49,
972
+ "learning_rate": 7.503558731410959e-06,
973
+ "loss": 1.3859,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 1.5,
978
+ "learning_rate": 7.245223131162376e-06,
979
+ "loss": 1.3716,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 1.51,
984
+ "learning_rate": 6.9906572567558285e-06,
985
+ "loss": 1.3846,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 1.52,
990
+ "learning_rate": 6.739915158071106e-06,
991
+ "loss": 1.3741,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 1.53,
996
+ "learning_rate": 6.493050073117116e-06,
997
+ "loss": 1.3827,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 1.54,
1002
+ "learning_rate": 6.250114416728298e-06,
1003
+ "loss": 1.3951,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 1.55,
1008
+ "learning_rate": 6.011159769435823e-06,
1009
+ "loss": 1.3875,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 1.56,
1014
+ "learning_rate": 5.776236866515947e-06,
1015
+ "loss": 1.3783,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 1.57,
1020
+ "learning_rate": 5.545395587217763e-06,
1021
+ "loss": 1.3888,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 1.58,
1026
+ "learning_rate": 5.318684944172752e-06,
1027
+ "loss": 1.4006,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 1.59,
1032
+ "learning_rate": 5.096153072988366e-06,
1033
+ "loss": 1.4041,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 1.59,
1038
+ "learning_rate": 4.8778472220277874e-06,
1039
+ "loss": 1.3865,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 1.6,
1044
+ "learning_rate": 4.6638137423780395e-06,
1045
+ "loss": 1.3789,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 1.61,
1050
+ "learning_rate": 4.454098078008667e-06,
1051
+ "loss": 1.3861,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 1.62,
1056
+ "learning_rate": 4.248744756122986e-06,
1057
+ "loss": 1.3803,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 1.63,
1062
+ "learning_rate": 4.047797377703985e-06,
1063
+ "loss": 1.3695,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 1.64,
1068
+ "learning_rate": 3.851298608256892e-06,
1069
+ "loss": 1.3937,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 1.65,
1074
+ "learning_rate": 3.6592901687503566e-06,
1075
+ "loss": 1.3865,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 1.66,
1080
+ "learning_rate": 3.471812826758178e-06,
1081
+ "loss": 1.3799,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 1.67,
1086
+ "learning_rate": 3.288906387803464e-06,
1087
+ "loss": 1.3911,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 1.68,
1092
+ "learning_rate": 3.1106096869070483e-06,
1093
+ "loss": 1.3801,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 1.69,
1098
+ "learning_rate": 2.9369605803419715e-06,
1099
+ "loss": 1.382,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 1.7,
1104
+ "learning_rate": 2.767995937595766e-06,
1105
+ "loss": 1.3898,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 1.71,
1110
+ "learning_rate": 2.6037516335422728e-06,
1111
+ "loss": 1.3752,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 1.72,
1116
+ "learning_rate": 2.4442625408246074e-06,
1117
+ "loss": 1.3855,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 1.72,
1122
+ "learning_rate": 2.289562522450947e-06,
1123
+ "loss": 1.3665,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 1.73,
1128
+ "learning_rate": 2.1396844246046903e-06,
1129
+ "loss": 1.3747,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 1.74,
1134
+ "learning_rate": 1.9946600696704592e-06,
1135
+ "loss": 1.3861,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 1.75,
1140
+ "learning_rate": 1.8545202494775509e-06,
1141
+ "loss": 1.3844,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 1.76,
1146
+ "learning_rate": 1.7192947187621434e-06,
1147
+ "loss": 1.3796,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 1.77,
1152
+ "learning_rate": 1.5890121888497366e-06,
1153
+ "loss": 1.3963,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 1.78,
1158
+ "learning_rate": 1.463700321559075e-06,
1159
+ "loss": 1.385,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 1.79,
1164
+ "learning_rate": 1.3433857233289714e-06,
1165
+ "loss": 1.3811,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 1.8,
1170
+ "learning_rate": 1.2280939395691859e-06,
1171
+ "loss": 1.384,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 1.81,
1176
+ "learning_rate": 1.1178494492365465e-06,
1177
+ "loss": 1.3821,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 1.82,
1182
+ "learning_rate": 1.0126756596375686e-06,
1183
+ "loss": 1.3978,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 1.83,
1188
+ "learning_rate": 9.125949014585383e-07,
1189
+ "loss": 1.3724,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 1.84,
1194
+ "learning_rate": 8.176284240242638e-07,
1195
+ "loss": 1.3851,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 1.85,
1200
+ "learning_rate": 7.277963907863478e-07,
1201
+ "loss": 1.3793,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 1.85,
1206
+ "learning_rate": 6.431178750420513e-07,
1207
+ "loss": 1.3946,
1208
+ "step": 2000
1209
+ }
1210
+ ],
1211
+ "max_steps": 2156,
1212
+ "num_train_epochs": 2,
1213
+ "total_flos": 4.3007034481621074e+18,
1214
+ "trial_name": null,
1215
+ "trial_params": null
1216
+ }
checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:918a94b2b648985fdb03a840052eff79b0035343fd87ee2a0dc751e904fad0c4
3
+ size 4091
special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<eoa>"
4
+ ],
5
+ "bos_token": "<s>",
6
+ "eos_token": "</s>",
7
+ "pad_token": "</s>",
8
+ "unk_token": "<unk>"
9
+ }
tokenization_internlm.py ADDED
@@ -0,0 +1,242 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved.
3
+ #
4
+ # This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
5
+ # and OPT implementations in this library. It has been modified from its
6
+ # original forms to accommodate minor architectural differences compared
7
+ # to GPT-NeoX and OPT used by the Meta AI team that trained the model.
8
+ #
9
+ # Licensed under the Apache License, Version 2.0 (the "License");
10
+ # you may not use this file except in compliance with the License.
11
+ # You may obtain a copy of the License at
12
+ #
13
+ # http://www.apache.org/licenses/LICENSE-2.0
14
+ #
15
+ # Unless required by applicable law or agreed to in writing, software
16
+ # distributed under the License is distributed on an "AS IS" BASIS,
17
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
18
+ # See the License for the specific language governing permissions and
19
+ # limitations under the License.
20
+
21
+ """Tokenization classes for IntermLM."""
22
+ import os
23
+ from shutil import copyfile
24
+ from typing import Any, Dict, List, Optional, Tuple
25
+
26
+ import sentencepiece as spm
27
+
28
+ from transformers.tokenization_utils import PreTrainedTokenizer
29
+ from transformers.utils import logging
30
+
31
+
32
+ logger = logging.get_logger(__name__)
33
+
34
+ VOCAB_FILES_NAMES = {"vocab_file": "./tokenizer.model"}
35
+
36
+ PRETRAINED_VOCAB_FILES_MAP = {}
37
+
38
+
39
+ class InternLMTokenizer(PreTrainedTokenizer):
40
+ """
41
+ Construct a InternLM tokenizer. Based on byte-level Byte-Pair-Encoding.
42
+
43
+ Args:
44
+ vocab_file (`str`):
45
+ Path to the vocabulary file.
46
+ """
47
+
48
+ vocab_files_names = VOCAB_FILES_NAMES
49
+ pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
50
+ model_input_names = ["input_ids", "attention_mask"]
51
+ _auto_class = "AutoTokenizer"
52
+
53
+ def __init__(
54
+ self,
55
+ vocab_file,
56
+ unk_token="<unk>",
57
+ bos_token="<s>",
58
+ eos_token="</s>",
59
+ pad_token="</s>",
60
+ sp_model_kwargs: Optional[Dict[str, Any]] = None,
61
+ add_bos_token=True,
62
+ add_eos_token=False,
63
+ decode_with_prefix_space=False,
64
+ clean_up_tokenization_spaces=False,
65
+ **kwargs,
66
+ ):
67
+ self.sp_model_kwargs = {} if sp_model_kwargs is None else sp_model_kwargs
68
+ super().__init__(
69
+ bos_token=bos_token,
70
+ eos_token=eos_token,
71
+ unk_token=unk_token,
72
+ pad_token=pad_token,
73
+ clean_up_tokenization_spaces=clean_up_tokenization_spaces,
74
+ **kwargs,
75
+ )
76
+ self.vocab_file = vocab_file
77
+ self.add_bos_token = add_bos_token
78
+ self.add_eos_token = add_eos_token
79
+ self.decode_with_prefix_space = decode_with_prefix_space
80
+ self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
81
+ self.sp_model.Load(vocab_file)
82
+ self._no_prefix_space_tokens = None
83
+
84
+ """ Initialisation"""
85
+
86
+ @property
87
+ def no_prefix_space_tokens(self):
88
+ if self._no_prefix_space_tokens is None:
89
+ vocab = self.convert_ids_to_tokens(list(range(self.vocab_size)))
90
+ self._no_prefix_space_tokens = {i for i, tok in enumerate(vocab) if not tok.startswith("▁")}
91
+ return self._no_prefix_space_tokens
92
+
93
+ @property
94
+ def vocab_size(self):
95
+ """Returns vocab size"""
96
+ return self.sp_model.get_piece_size()
97
+
98
+ @property
99
+ def bos_token_id(self) -> Optional[int]:
100
+ return self.sp_model.bos_id()
101
+
102
+ @property
103
+ def eos_token_id(self) -> Optional[int]:
104
+ return self.sp_model.eos_id()
105
+
106
+ def get_vocab(self):
107
+ """Returns vocab as a dict"""
108
+ vocab = {self.convert_ids_to_tokens(i): i for i in range(self.vocab_size)}
109
+ vocab.update(self.added_tokens_encoder)
110
+ return vocab
111
+
112
+ def _tokenize(self, text):
113
+ """Returns a tokenized string."""
114
+ return self.sp_model.encode(text, out_type=str)
115
+
116
+ def _convert_token_to_id(self, token):
117
+ """Converts a token (str) in an id using the vocab."""
118
+ return self.sp_model.piece_to_id(token)
119
+
120
+ def _convert_id_to_token(self, index):
121
+ """Converts an index (integer) in a token (str) using the vocab."""
122
+ token = self.sp_model.IdToPiece(index)
123
+ return token
124
+
125
+ def _maybe_add_prefix_space(self, tokens, decoded):
126
+ if tokens and tokens[0] not in self.no_prefix_space_tokens:
127
+ return " " + decoded
128
+ else:
129
+ return decoded
130
+
131
+ def convert_tokens_to_string(self, tokens):
132
+ """Converts a sequence of tokens (string) in a single string."""
133
+ current_sub_tokens = []
134
+ out_string = ""
135
+ prev_is_special = False
136
+ for token in tokens:
137
+ # make sure that special tokens are not decoded using sentencepiece model
138
+ if token in self.all_special_tokens:
139
+ if not prev_is_special:
140
+ out_string += " "
141
+ out_string += self.sp_model.decode(current_sub_tokens) + token
142
+ prev_is_special = True
143
+ current_sub_tokens = []
144
+ else:
145
+ current_sub_tokens.append(token)
146
+ prev_is_special = False
147
+ out_string += self.sp_model.decode(current_sub_tokens)
148
+ out_string = self.clean_up_tokenization(out_string)
149
+ out_string = self._maybe_add_prefix_space(tokens=tokens, decoded=out_string)
150
+ return out_string[1:]
151
+
152
+ def save_vocabulary(self, save_directory, filename_prefix: Optional[str] = None) -> Tuple[str]:
153
+ """
154
+ Save the vocabulary and special tokens file to a directory.
155
+
156
+ Args:
157
+ save_directory (`str`):
158
+ The directory in which to save the vocabulary.
159
+
160
+ Returns:
161
+ `Tuple(str)`: Paths to the files saved.
162
+ """
163
+ if not os.path.isdir(save_directory):
164
+ logger.error(f"Vocabulary path ({save_directory}) should be a directory")
165
+ return
166
+ out_vocab_file = os.path.join(
167
+ save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
168
+ )
169
+
170
+ if os.path.abspath(self.vocab_file) != os.path.abspath(out_vocab_file) and os.path.isfile(self.vocab_file):
171
+ copyfile(self.vocab_file, out_vocab_file)
172
+ elif not os.path.isfile(self.vocab_file):
173
+ with open(out_vocab_file, "wb") as fi:
174
+ content_spiece_model = self.sp_model.serialized_model_proto()
175
+ fi.write(content_spiece_model)
176
+
177
+ return (out_vocab_file,)
178
+
179
+ def build_inputs_with_special_tokens(self, token_ids_0, token_ids_1=None):
180
+ if self.add_bos_token:
181
+ bos_token_ids = [self.bos_token_id]
182
+ else:
183
+ bos_token_ids = []
184
+
185
+ output = bos_token_ids + token_ids_0
186
+
187
+ if token_ids_1 is not None:
188
+ output = output + token_ids_1
189
+
190
+ if self.add_eos_token:
191
+ output = output + [self.eos_token_id]
192
+
193
+ return output
194
+
195
+ def get_special_tokens_mask(
196
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
197
+ ) -> List[int]:
198
+ """
199
+ Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
200
+ special tokens using the tokenizer `prepare_for_model` method.
201
+
202
+ Args:
203
+ token_ids_0 (`List[int]`):
204
+ List of IDs.
205
+ token_ids_1 (`List[int]`, *optional*):
206
+ Optional second list of IDs for sequence pairs.
207
+ already_has_special_tokens (`bool`, *optional*, defaults to `False`):
208
+ Whether or not the token list is already formatted with special tokens for the model.
209
+
210
+ Returns:
211
+ `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
212
+ """
213
+ if already_has_special_tokens:
214
+ return super().get_special_tokens_mask(
215
+ token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
216
+ )
217
+
218
+ if token_ids_1 is None:
219
+ return [1] + ([0] * len(token_ids_0)) + [1]
220
+ return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]
221
+
222
+ def create_token_type_ids_from_sequences(
223
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
224
+ ) -> List[int]:
225
+ """
226
+ Create a mask from the two sequences passed to be used in a sequence-pair classification task. T5 does not make
227
+ use of token type ids, therefore a list of zeros is returned.
228
+
229
+ Args:
230
+ token_ids_0 (`List[int]`):
231
+ List of IDs.
232
+ token_ids_1 (`List[int]`, *optional*):
233
+ Optional second list of IDs for sequence pairs.
234
+
235
+ Returns:
236
+ `List[int]`: List of zeros.
237
+ """
238
+ eos = [self.eos_token_id]
239
+
240
+ if token_ids_1 is None:
241
+ return len(token_ids_0 + eos) * [0]
242
+ return len(token_ids_0 + eos + token_ids_1 + eos) * [0]
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aab622d98c98677a1a51f969e25765154487bf3e85c7819db105db2fcacba83f
3
+ size 1658691
tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_map": {
3
+ "AutoTokenizer": [
4
+ "tokenization_internlm.InternLMTokenizer",
5
+ null
6
+ ]
7
+ },
8
+ "bos_token": "<s>",
9
+ "clean_up_tokenization_spaces": false,
10
+ "eos_token": "</s>",
11
+ "model_max_length": 1000000000000000019884624838656,
12
+ "pad_token": "</s>",
13
+ "padding_side": "right",
14
+ "tokenizer_class": "InternLMTokenizer",
15
+ "unk_token": "<unk>"
16
+ }
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "train_loss": 1.4111853902105498,
4
+ "train_runtime": 51425.1791,
5
+ "train_samples_per_second": 2.684,
6
+ "train_steps_per_second": 0.042
7
+ }
trainer_log.jsonl ADDED
@@ -0,0 +1,216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"current_steps": 10, "total_steps": 2156, "loss": 1.7592, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.999734597774032e-05, "epoch": 0.01, "percentage": 0.46, "elapsed_time": "0:03:56", "remaining_time": "14:04:54"}
2
+ {"current_steps": 20, "total_steps": 2156, "loss": 1.6546, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.998938447446803e-05, "epoch": 0.02, "percentage": 0.93, "elapsed_time": "0:07:53", "remaining_time": "14:03:23"}
3
+ {"current_steps": 30, "total_steps": 2156, "loss": 1.5632, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.997611718058365e-05, "epoch": 0.03, "percentage": 1.39, "elapsed_time": "0:11:45", "remaining_time": "13:53:05"}
4
+ {"current_steps": 40, "total_steps": 2156, "loss": 1.5245, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9957546913022665e-05, "epoch": 0.04, "percentage": 1.86, "elapsed_time": "0:15:44", "remaining_time": "13:52:29"}
5
+ {"current_steps": 50, "total_steps": 2156, "loss": 1.5049, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.993367761465736e-05, "epoch": 0.05, "percentage": 2.32, "elapsed_time": "0:19:48", "remaining_time": "13:54:22"}
6
+ {"current_steps": 60, "total_steps": 2156, "loss": 1.492, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9904514353459654e-05, "epoch": 0.06, "percentage": 2.78, "elapsed_time": "0:23:41", "remaining_time": "13:47:41"}
7
+ {"current_steps": 70, "total_steps": 2156, "loss": 1.4803, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9870063321425105e-05, "epoch": 0.06, "percentage": 3.25, "elapsed_time": "0:27:43", "remaining_time": "13:46:02"}
8
+ {"current_steps": 80, "total_steps": 2156, "loss": 1.4827, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.983033183325818e-05, "epoch": 0.07, "percentage": 3.71, "elapsed_time": "0:31:38", "remaining_time": "13:40:57"}
9
+ {"current_steps": 90, "total_steps": 2156, "loss": 1.4866, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.97853283248192e-05, "epoch": 0.08, "percentage": 4.17, "elapsed_time": "0:35:30", "remaining_time": "13:35:03"}
10
+ {"current_steps": 100, "total_steps": 2156, "loss": 1.486, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.973506235133323e-05, "epoch": 0.09, "percentage": 4.64, "elapsed_time": "0:39:25", "remaining_time": "13:30:27"}
11
+ {"current_steps": 110, "total_steps": 2156, "loss": 1.4822, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.967954458536126e-05, "epoch": 0.1, "percentage": 5.1, "elapsed_time": "0:43:22", "remaining_time": "13:26:52"}
12
+ {"current_steps": 120, "total_steps": 2156, "loss": 1.4801, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9618786814534226e-05, "epoch": 0.11, "percentage": 5.57, "elapsed_time": "0:47:19", "remaining_time": "13:22:52"}
13
+ {"current_steps": 130, "total_steps": 2156, "loss": 1.4764, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.955280193905022e-05, "epoch": 0.12, "percentage": 6.03, "elapsed_time": "0:51:16", "remaining_time": "13:19:07"}
14
+ {"current_steps": 140, "total_steps": 2156, "loss": 1.4662, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.948160396893553e-05, "epoch": 0.13, "percentage": 6.49, "elapsed_time": "0:55:09", "remaining_time": "13:14:11"}
15
+ {"current_steps": 150, "total_steps": 2156, "loss": 1.4627, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9405208021069946e-05, "epoch": 0.14, "percentage": 6.96, "elapsed_time": "0:59:02", "remaining_time": "13:09:34"}
16
+ {"current_steps": 160, "total_steps": 2156, "loss": 1.4543, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9323630315977156e-05, "epoch": 0.15, "percentage": 7.42, "elapsed_time": "1:03:04", "remaining_time": "13:06:48"}
17
+ {"current_steps": 170, "total_steps": 2156, "loss": 1.4508, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9236888174380784e-05, "epoch": 0.16, "percentage": 7.88, "elapsed_time": "1:06:57", "remaining_time": "13:02:15"}
18
+ {"current_steps": 180, "total_steps": 2156, "loss": 1.464, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.91450000135268e-05, "epoch": 0.17, "percentage": 8.35, "elapsed_time": "1:10:51", "remaining_time": "12:57:47"}
19
+ {"current_steps": 190, "total_steps": 2156, "loss": 1.4705, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9047985343273154e-05, "epoch": 0.18, "percentage": 8.81, "elapsed_time": "1:14:48", "remaining_time": "12:54:02"}
20
+ {"current_steps": 200, "total_steps": 2156, "loss": 1.4588, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.894586476194739e-05, "epoch": 0.19, "percentage": 9.28, "elapsed_time": "1:18:40", "remaining_time": "12:49:30"}
21
+ {"current_steps": 210, "total_steps": 2156, "loss": 1.4532, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.883865995197319e-05, "epoch": 0.19, "percentage": 9.74, "elapsed_time": "1:22:38", "remaining_time": "12:45:48"}
22
+ {"current_steps": 220, "total_steps": 2156, "loss": 1.4629, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.8726393675266716e-05, "epoch": 0.2, "percentage": 10.2, "elapsed_time": "1:26:32", "remaining_time": "12:41:37"}
23
+ {"current_steps": 230, "total_steps": 2156, "loss": 1.4429, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.860908976840376e-05, "epoch": 0.21, "percentage": 10.67, "elapsed_time": "1:30:23", "remaining_time": "12:36:52"}
24
+ {"current_steps": 240, "total_steps": 2156, "loss": 1.441, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.848677313755872e-05, "epoch": 0.22, "percentage": 11.13, "elapsed_time": "1:34:21", "remaining_time": "12:33:20"}
25
+ {"current_steps": 250, "total_steps": 2156, "loss": 1.4496, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.835946975321647e-05, "epoch": 0.23, "percentage": 11.6, "elapsed_time": "1:38:16", "remaining_time": "12:29:16"}
26
+ {"current_steps": 260, "total_steps": 2156, "loss": 1.4459, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.822720664465827e-05, "epoch": 0.24, "percentage": 12.06, "elapsed_time": "1:42:13", "remaining_time": "12:25:28"}
27
+ {"current_steps": 270, "total_steps": 2156, "loss": 1.4458, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.809001189422287e-05, "epoch": 0.25, "percentage": 12.52, "elapsed_time": "1:46:20", "remaining_time": "12:22:48"}
28
+ {"current_steps": 280, "total_steps": 2156, "loss": 1.4484, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.794791463134399e-05, "epoch": 0.26, "percentage": 12.99, "elapsed_time": "1:50:17", "remaining_time": "12:18:59"}
29
+ {"current_steps": 290, "total_steps": 2156, "loss": 1.4541, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.780094502636552e-05, "epoch": 0.27, "percentage": 13.45, "elapsed_time": "1:54:05", "remaining_time": "12:14:08"}
30
+ {"current_steps": 300, "total_steps": 2156, "loss": 1.437, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.764913428413572e-05, "epoch": 0.28, "percentage": 13.91, "elapsed_time": "1:58:02", "remaining_time": "12:10:16"}
31
+ {"current_steps": 310, "total_steps": 2156, "loss": 1.4368, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.7492514637381727e-05, "epoch": 0.29, "percentage": 14.38, "elapsed_time": "2:01:56", "remaining_time": "12:06:09"}
32
+ {"current_steps": 320, "total_steps": 2156, "loss": 1.4449, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.733111933986583e-05, "epoch": 0.3, "percentage": 14.84, "elapsed_time": "2:05:52", "remaining_time": "12:02:14"}
33
+ {"current_steps": 330, "total_steps": 2156, "loss": 1.4256, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.716498265932501e-05, "epoch": 0.31, "percentage": 15.31, "elapsed_time": "2:09:46", "remaining_time": "11:58:04"}
34
+ {"current_steps": 340, "total_steps": 2156, "loss": 1.4378, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.699413987019512e-05, "epoch": 0.32, "percentage": 15.77, "elapsed_time": "2:13:49", "remaining_time": "11:54:45"}
35
+ {"current_steps": 350, "total_steps": 2156, "loss": 1.454, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.681862724612141e-05, "epoch": 0.32, "percentage": 16.23, "elapsed_time": "2:17:55", "remaining_time": "11:51:41"}
36
+ {"current_steps": 360, "total_steps": 2156, "loss": 1.4207, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.663848205225674e-05, "epoch": 0.33, "percentage": 16.7, "elapsed_time": "2:21:56", "remaining_time": "11:48:08"}
37
+ {"current_steps": 370, "total_steps": 2156, "loss": 1.4286, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.645374253734949e-05, "epoch": 0.34, "percentage": 17.16, "elapsed_time": "2:25:51", "remaining_time": "11:44:02"}
38
+ {"current_steps": 380, "total_steps": 2156, "loss": 1.4244, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.626444792562244e-05, "epoch": 0.35, "percentage": 17.63, "elapsed_time": "2:29:47", "remaining_time": "11:40:03"}
39
+ {"current_steps": 390, "total_steps": 2156, "loss": 1.4342, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.607063840844463e-05, "epoch": 0.36, "percentage": 18.09, "elapsed_time": "2:33:48", "remaining_time": "11:36:28"}
40
+ {"current_steps": 400, "total_steps": 2156, "loss": 1.4241, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.587235513579791e-05, "epoch": 0.37, "percentage": 18.55, "elapsed_time": "2:37:54", "remaining_time": "11:33:12"}
41
+ {"current_steps": 410, "total_steps": 2156, "loss": 1.443, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.5669640207539786e-05, "epoch": 0.38, "percentage": 19.02, "elapsed_time": "2:41:50", "remaining_time": "11:29:13"}
42
+ {"current_steps": 420, "total_steps": 2156, "loss": 1.4281, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.546253666446484e-05, "epoch": 0.39, "percentage": 19.48, "elapsed_time": "2:45:55", "remaining_time": "11:25:47"}
43
+ {"current_steps": 430, "total_steps": 2156, "loss": 1.4193, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.525108847916614e-05, "epoch": 0.4, "percentage": 19.94, "elapsed_time": "2:49:54", "remaining_time": "11:21:58"}
44
+ {"current_steps": 440, "total_steps": 2156, "loss": 1.4154, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.503534054669892e-05, "epoch": 0.41, "percentage": 20.41, "elapsed_time": "2:53:52", "remaining_time": "11:18:07"}
45
+ {"current_steps": 450, "total_steps": 2156, "loss": 1.4273, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.481533867504841e-05, "epoch": 0.42, "percentage": 20.87, "elapsed_time": "2:58:00", "remaining_time": "11:14:50"}
46
+ {"current_steps": 460, "total_steps": 2156, "loss": 1.4329, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.4591129575403765e-05, "epoch": 0.43, "percentage": 21.34, "elapsed_time": "3:01:55", "remaining_time": "11:10:43"}
47
+ {"current_steps": 470, "total_steps": 2156, "loss": 1.4102, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.43627608522403e-05, "epoch": 0.44, "percentage": 21.8, "elapsed_time": "3:05:54", "remaining_time": "11:06:53"}
48
+ {"current_steps": 480, "total_steps": 2156, "loss": 1.4244, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.4130280993211974e-05, "epoch": 0.45, "percentage": 22.26, "elapsed_time": "3:09:56", "remaining_time": "11:03:12"}
49
+ {"current_steps": 490, "total_steps": 2156, "loss": 1.4407, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.389373935885646e-05, "epoch": 0.45, "percentage": 22.73, "elapsed_time": "3:13:57", "remaining_time": "10:59:27"}
50
+ {"current_steps": 500, "total_steps": 2156, "loss": 1.4282, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.365318617211479e-05, "epoch": 0.46, "percentage": 23.19, "elapsed_time": "3:17:54", "remaining_time": "10:55:29"}
51
+ {"current_steps": 510, "total_steps": 2156, "loss": 1.4269, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.340867250766794e-05, "epoch": 0.47, "percentage": 23.65, "elapsed_time": "3:21:49", "remaining_time": "10:51:22"}
52
+ {"current_steps": 520, "total_steps": 2156, "loss": 1.43, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.316025028109258e-05, "epoch": 0.48, "percentage": 24.12, "elapsed_time": "3:25:47", "remaining_time": "10:47:25"}
53
+ {"current_steps": 530, "total_steps": 2156, "loss": 1.4271, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.2907972237838225e-05, "epoch": 0.49, "percentage": 24.58, "elapsed_time": "3:29:54", "remaining_time": "10:43:58"}
54
+ {"current_steps": 540, "total_steps": 2156, "loss": 1.4321, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.2651891942028274e-05, "epoch": 0.5, "percentage": 25.05, "elapsed_time": "3:33:53", "remaining_time": "10:40:04"}
55
+ {"current_steps": 550, "total_steps": 2156, "loss": 1.4083, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.239206376508717e-05, "epoch": 0.51, "percentage": 25.51, "elapsed_time": "3:37:48", "remaining_time": "10:36:00"}
56
+ {"current_steps": 560, "total_steps": 2156, "loss": 1.4168, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.212854287419611e-05, "epoch": 0.52, "percentage": 25.97, "elapsed_time": "3:41:42", "remaining_time": "10:31:50"}
57
+ {"current_steps": 570, "total_steps": 2156, "loss": 1.4144, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.1861385220579934e-05, "epoch": 0.53, "percentage": 26.44, "elapsed_time": "3:45:37", "remaining_time": "10:27:47"}
58
+ {"current_steps": 580, "total_steps": 2156, "loss": 1.4106, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.1590647527627404e-05, "epoch": 0.54, "percentage": 26.9, "elapsed_time": "3:49:26", "remaining_time": "10:23:26"}
59
+ {"current_steps": 590, "total_steps": 2156, "loss": 1.4202, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.131638727884762e-05, "epoch": 0.55, "percentage": 27.37, "elapsed_time": "3:53:19", "remaining_time": "10:19:19"}
60
+ {"current_steps": 600, "total_steps": 2156, "loss": 1.417, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.103866270566498e-05, "epoch": 0.56, "percentage": 27.83, "elapsed_time": "3:57:15", "remaining_time": "10:15:18"}
61
+ {"current_steps": 610, "total_steps": 2156, "loss": 1.4324, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.075753277505544e-05, "epoch": 0.57, "percentage": 28.29, "elapsed_time": "4:01:13", "remaining_time": "10:11:23"}
62
+ {"current_steps": 620, "total_steps": 2156, "loss": 1.4147, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.0473057177026484e-05, "epoch": 0.57, "percentage": 28.76, "elapsed_time": "4:05:23", "remaining_time": "10:07:57"}
63
+ {"current_steps": 630, "total_steps": 2156, "loss": 1.4269, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.018529631194369e-05, "epoch": 0.58, "percentage": 29.22, "elapsed_time": "4:09:20", "remaining_time": "10:03:58"}
64
+ {"current_steps": 640, "total_steps": 2156, "loss": 1.4133, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.989431127770635e-05, "epoch": 0.59, "percentage": 29.68, "elapsed_time": "4:13:09", "remaining_time": "9:59:40"}
65
+ {"current_steps": 650, "total_steps": 2156, "loss": 1.4173, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.960016385677513e-05, "epoch": 0.6, "percentage": 30.15, "elapsed_time": "4:17:06", "remaining_time": "9:55:41"}
66
+ {"current_steps": 660, "total_steps": 2156, "loss": 1.4007, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.9302916503054246e-05, "epoch": 0.61, "percentage": 30.61, "elapsed_time": "4:21:12", "remaining_time": "9:52:04"}
67
+ {"current_steps": 670, "total_steps": 2156, "loss": 1.4117, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.9002632328631164e-05, "epoch": 0.62, "percentage": 31.08, "elapsed_time": "4:25:07", "remaining_time": "9:48:00"}
68
+ {"current_steps": 680, "total_steps": 2156, "loss": 1.4247, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.8699375090376534e-05, "epoch": 0.63, "percentage": 31.54, "elapsed_time": "4:29:03", "remaining_time": "9:44:01"}
69
+ {"current_steps": 690, "total_steps": 2156, "loss": 1.4312, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.8393209176407223e-05, "epoch": 0.64, "percentage": 32.0, "elapsed_time": "4:32:59", "remaining_time": "9:40:00"}
70
+ {"current_steps": 700, "total_steps": 2156, "loss": 1.4136, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.8084199592415305e-05, "epoch": 0.65, "percentage": 32.47, "elapsed_time": "4:37:02", "remaining_time": "9:36:13"}
71
+ {"current_steps": 710, "total_steps": 2156, "loss": 1.4132, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.777241194786591e-05, "epoch": 0.66, "percentage": 32.93, "elapsed_time": "4:40:58", "remaining_time": "9:32:13"}
72
+ {"current_steps": 720, "total_steps": 2156, "loss": 1.4145, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.745791244206697e-05, "epoch": 0.67, "percentage": 33.4, "elapsed_time": "4:44:52", "remaining_time": "9:28:10"}
73
+ {"current_steps": 730, "total_steps": 2156, "loss": 1.3951, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.714076785011359e-05, "epoch": 0.68, "percentage": 33.86, "elapsed_time": "4:48:57", "remaining_time": "9:24:26"}
74
+ {"current_steps": 740, "total_steps": 2156, "loss": 1.4139, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.682104550871031e-05, "epoch": 0.69, "percentage": 34.32, "elapsed_time": "4:52:56", "remaining_time": "9:20:32"}
75
+ {"current_steps": 750, "total_steps": 2156, "loss": 1.4078, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.649881330187401e-05, "epoch": 0.7, "percentage": 34.79, "elapsed_time": "4:56:50", "remaining_time": "9:16:28"}
76
+ {"current_steps": 760, "total_steps": 2156, "loss": 1.4048, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.617413964652067e-05, "epoch": 0.7, "percentage": 35.25, "elapsed_time": "5:00:52", "remaining_time": "9:12:40"}
77
+ {"current_steps": 770, "total_steps": 2156, "loss": 1.4115, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.5847093477938956e-05, "epoch": 0.71, "percentage": 35.71, "elapsed_time": "5:04:55", "remaining_time": "9:08:51"}
78
+ {"current_steps": 780, "total_steps": 2156, "loss": 1.4018, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.551774423515378e-05, "epoch": 0.72, "percentage": 36.18, "elapsed_time": "5:08:45", "remaining_time": "9:04:40"}
79
+ {"current_steps": 790, "total_steps": 2156, "loss": 1.4091, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.518616184618288e-05, "epoch": 0.73, "percentage": 36.64, "elapsed_time": "5:12:43", "remaining_time": "9:00:44"}
80
+ {"current_steps": 800, "total_steps": 2156, "loss": 1.4135, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.4852416713189526e-05, "epoch": 0.74, "percentage": 37.11, "elapsed_time": "5:16:40", "remaining_time": "8:56:45"}
81
+ {"current_steps": 810, "total_steps": 2156, "loss": 1.3977, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.4516579697534705e-05, "epoch": 0.75, "percentage": 37.57, "elapsed_time": "5:20:49", "remaining_time": "8:53:07"}
82
+ {"current_steps": 820, "total_steps": 2156, "loss": 1.4123, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.417872210473162e-05, "epoch": 0.76, "percentage": 38.03, "elapsed_time": "5:24:55", "remaining_time": "8:49:23"}
83
+ {"current_steps": 830, "total_steps": 2156, "loss": 1.4096, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.3838915669306034e-05, "epoch": 0.77, "percentage": 38.5, "elapsed_time": "5:28:51", "remaining_time": "8:45:22"}
84
+ {"current_steps": 840, "total_steps": 2156, "loss": 1.4031, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.349723253956542e-05, "epoch": 0.78, "percentage": 38.96, "elapsed_time": "5:32:48", "remaining_time": "8:41:24"}
85
+ {"current_steps": 850, "total_steps": 2156, "loss": 1.4104, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.315374526228036e-05, "epoch": 0.79, "percentage": 39.42, "elapsed_time": "5:36:46", "remaining_time": "8:37:26"}
86
+ {"current_steps": 860, "total_steps": 2156, "loss": 1.3972, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.2808526767281225e-05, "epoch": 0.8, "percentage": 39.89, "elapsed_time": "5:40:47", "remaining_time": "8:33:33"}
87
+ {"current_steps": 870, "total_steps": 2156, "loss": 1.3987, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.246165035197364e-05, "epoch": 0.81, "percentage": 40.35, "elapsed_time": "5:44:47", "remaining_time": "8:29:38"}
88
+ {"current_steps": 880, "total_steps": 2156, "loss": 1.4197, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.211318966577581e-05, "epoch": 0.82, "percentage": 40.82, "elapsed_time": "5:48:49", "remaining_time": "8:25:47"}
89
+ {"current_steps": 890, "total_steps": 2156, "loss": 1.4169, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.176321869448116e-05, "epoch": 0.83, "percentage": 41.28, "elapsed_time": "5:52:46", "remaining_time": "8:21:48"}
90
+ {"current_steps": 900, "total_steps": 2156, "loss": 1.4055, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.1411811744549536e-05, "epoch": 0.83, "percentage": 41.74, "elapsed_time": "5:56:38", "remaining_time": "8:17:42"}
91
+ {"current_steps": 910, "total_steps": 2156, "loss": 1.402, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.105904342733032e-05, "epoch": 0.84, "percentage": 42.21, "elapsed_time": "6:00:35", "remaining_time": "8:13:44"}
92
+ {"current_steps": 920, "total_steps": 2156, "loss": 1.4148, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.070498864322081e-05, "epoch": 0.85, "percentage": 42.67, "elapsed_time": "6:04:30", "remaining_time": "8:09:42"}
93
+ {"current_steps": 930, "total_steps": 2156, "loss": 1.4007, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.034972256576328e-05, "epoch": 0.86, "percentage": 43.14, "elapsed_time": "6:08:25", "remaining_time": "8:05:41"}
94
+ {"current_steps": 940, "total_steps": 2156, "loss": 1.3926, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.999332062568395e-05, "epoch": 0.87, "percentage": 43.6, "elapsed_time": "6:12:19", "remaining_time": "8:01:39"}
95
+ {"current_steps": 950, "total_steps": 2156, "loss": 1.4099, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9635858494877384e-05, "epoch": 0.88, "percentage": 44.06, "elapsed_time": "6:16:19", "remaining_time": "7:57:44"}
96
+ {"current_steps": 960, "total_steps": 2156, "loss": 1.401, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9277412070339782e-05, "epoch": 0.89, "percentage": 44.53, "elapsed_time": "6:20:23", "remaining_time": "7:53:53"}
97
+ {"current_steps": 970, "total_steps": 2156, "loss": 1.3995, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.891805745805429e-05, "epoch": 0.9, "percentage": 44.99, "elapsed_time": "6:24:32", "remaining_time": "7:50:09"}
98
+ {"current_steps": 980, "total_steps": 2156, "loss": 1.4108, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.8557870956832132e-05, "epoch": 0.91, "percentage": 45.45, "elapsed_time": "6:28:27", "remaining_time": "7:46:09"}
99
+ {"current_steps": 990, "total_steps": 2156, "loss": 1.4006, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.8196929042112652e-05, "epoch": 0.92, "percentage": 45.92, "elapsed_time": "6:32:21", "remaining_time": "7:42:07"}
100
+ {"current_steps": 1000, "total_steps": 2156, "loss": 1.4045, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.783530834972594e-05, "epoch": 0.93, "percentage": 46.38, "elapsed_time": "6:36:20", "remaining_time": "7:38:09"}
101
+ {"current_steps": 1010, "total_steps": 2156, "loss": 1.3907, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.7473085659621377e-05, "epoch": 0.94, "percentage": 46.85, "elapsed_time": "6:40:20", "remaining_time": "7:34:14"}
102
+ {"current_steps": 1020, "total_steps": 2156, "loss": 1.3872, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.711033787956555e-05, "epoch": 0.95, "percentage": 47.31, "elapsed_time": "6:44:22", "remaining_time": "7:30:21"}
103
+ {"current_steps": 1030, "total_steps": 2156, "loss": 1.405, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.6747142028813105e-05, "epoch": 0.96, "percentage": 47.77, "elapsed_time": "6:48:20", "remaining_time": "7:26:24"}
104
+ {"current_steps": 1040, "total_steps": 2156, "loss": 1.3952, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.638357522175383e-05, "epoch": 0.96, "percentage": 48.24, "elapsed_time": "6:52:13", "remaining_time": "7:22:20"}
105
+ {"current_steps": 1050, "total_steps": 2156, "loss": 1.3901, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.6019714651539646e-05, "epoch": 0.97, "percentage": 48.7, "elapsed_time": "6:56:14", "remaining_time": "7:18:26"}
106
+ {"current_steps": 1060, "total_steps": 2156, "loss": 1.4027, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.565563757369475e-05, "epoch": 0.98, "percentage": 49.17, "elapsed_time": "7:00:12", "remaining_time": "7:14:29"}
107
+ {"current_steps": 1070, "total_steps": 2156, "loss": 1.3909, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.529142128971268e-05, "epoch": 0.99, "percentage": 49.63, "elapsed_time": "7:04:17", "remaining_time": "7:10:38"}
108
+ {"current_steps": 1080, "total_steps": 2156, "loss": 1.4039, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.492714313064342e-05, "epoch": 1.0, "percentage": 50.09, "elapsed_time": "7:08:25", "remaining_time": "7:06:50"}
109
+ {"current_steps": 1090, "total_steps": 2156, "loss": 1.3964, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.45628804406744e-05, "epoch": 1.01, "percentage": 50.56, "elapsed_time": "7:12:23", "remaining_time": "7:02:52"}
110
+ {"current_steps": 1100, "total_steps": 2156, "loss": 1.4076, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.419871056070862e-05, "epoch": 1.02, "percentage": 51.02, "elapsed_time": "7:16:17", "remaining_time": "6:58:50"}
111
+ {"current_steps": 1110, "total_steps": 2156, "loss": 1.3891, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.3834710811943514e-05, "epoch": 1.03, "percentage": 51.48, "elapsed_time": "7:20:12", "remaining_time": "6:54:50"}
112
+ {"current_steps": 1120, "total_steps": 2156, "loss": 1.4004, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.3470958479453938e-05, "epoch": 1.04, "percentage": 51.95, "elapsed_time": "7:24:12", "remaining_time": "6:50:53"}
113
+ {"current_steps": 1130, "total_steps": 2156, "loss": 1.3879, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.3107530795782877e-05, "epoch": 1.05, "percentage": 52.41, "elapsed_time": "7:28:14", "remaining_time": "6:46:58"}
114
+ {"current_steps": 1140, "total_steps": 2156, "loss": 1.3879, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.2744504924543313e-05, "epoch": 1.06, "percentage": 52.88, "elapsed_time": "7:32:09", "remaining_time": "6:42:58"}
115
+ {"current_steps": 1150, "total_steps": 2156, "loss": 1.3851, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.23819579440347e-05, "epoch": 1.07, "percentage": 53.34, "elapsed_time": "7:36:09", "remaining_time": "6:39:02"}
116
+ {"current_steps": 1160, "total_steps": 2156, "loss": 1.3846, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.2019966830877545e-05, "epoch": 1.08, "percentage": 53.8, "elapsed_time": "7:40:18", "remaining_time": "6:35:13"}
117
+ {"current_steps": 1170, "total_steps": 2156, "loss": 1.3936, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.1658608443669635e-05, "epoch": 1.08, "percentage": 54.27, "elapsed_time": "7:44:16", "remaining_time": "6:31:15"}
118
+ {"current_steps": 1180, "total_steps": 2156, "loss": 1.3841, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.1297959506667224e-05, "epoch": 1.09, "percentage": 54.73, "elapsed_time": "7:48:20", "remaining_time": "6:27:22"}
119
+ {"current_steps": 1190, "total_steps": 2156, "loss": 1.3969, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.0938096593494855e-05, "epoch": 1.1, "percentage": 55.19, "elapsed_time": "7:52:27", "remaining_time": "6:23:31"}
120
+ {"current_steps": 1200, "total_steps": 2156, "loss": 1.3953, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.057909611088709e-05, "epoch": 1.11, "percentage": 55.66, "elapsed_time": "7:56:26", "remaining_time": "6:19:34"}
121
+ {"current_steps": 1210, "total_steps": 2156, "loss": 1.3951, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.02210342824657e-05, "epoch": 1.12, "percentage": 56.12, "elapsed_time": "8:00:22", "remaining_time": "6:15:33"}
122
+ {"current_steps": 1220, "total_steps": 2156, "loss": 1.4012, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.9863987132555706e-05, "epoch": 1.13, "percentage": 56.59, "elapsed_time": "8:04:18", "remaining_time": "6:11:34"}
123
+ {"current_steps": 1230, "total_steps": 2156, "loss": 1.3998, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.9508030470043806e-05, "epoch": 1.14, "percentage": 57.05, "elapsed_time": "8:08:21", "remaining_time": "6:07:39"}
124
+ {"current_steps": 1240, "total_steps": 2156, "loss": 1.3933, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.915323987228247e-05, "epoch": 1.15, "percentage": 57.51, "elapsed_time": "8:12:22", "remaining_time": "6:03:43"}
125
+ {"current_steps": 1250, "total_steps": 2156, "loss": 1.3977, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.8799690669043212e-05, "epoch": 1.16, "percentage": 57.98, "elapsed_time": "8:16:18", "remaining_time": "5:59:43"}
126
+ {"current_steps": 1260, "total_steps": 2156, "loss": 1.3818, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.8447457926522454e-05, "epoch": 1.17, "percentage": 58.44, "elapsed_time": "8:20:15", "remaining_time": "5:55:44"}
127
+ {"current_steps": 1270, "total_steps": 2156, "loss": 1.4032, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.8096616431403325e-05, "epoch": 1.18, "percentage": 58.91, "elapsed_time": "8:24:17", "remaining_time": "5:51:48"}
128
+ {"current_steps": 1280, "total_steps": 2156, "loss": 1.3853, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.7747240674976857e-05, "epoch": 1.19, "percentage": 59.37, "elapsed_time": "8:28:14", "remaining_time": "5:47:49"}
129
+ {"current_steps": 1290, "total_steps": 2156, "loss": 1.3988, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.7399404837325796e-05, "epoch": 1.2, "percentage": 59.83, "elapsed_time": "8:32:08", "remaining_time": "5:43:48"}
130
+ {"current_steps": 1300, "total_steps": 2156, "loss": 1.3963, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.7053182771574633e-05, "epoch": 1.21, "percentage": 60.3, "elapsed_time": "8:35:59", "remaining_time": "5:39:45"}
131
+ {"current_steps": 1310, "total_steps": 2156, "loss": 1.3881, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.6708647988208887e-05, "epoch": 1.21, "percentage": 60.76, "elapsed_time": "8:39:58", "remaining_time": "5:35:47"}
132
+ {"current_steps": 1320, "total_steps": 2156, "loss": 1.3804, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.6365873639467315e-05, "epoch": 1.22, "percentage": 61.22, "elapsed_time": "8:43:59", "remaining_time": "5:31:51"}
133
+ {"current_steps": 1330, "total_steps": 2156, "loss": 1.3952, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.602493250381003e-05, "epoch": 1.23, "percentage": 61.69, "elapsed_time": "8:48:00", "remaining_time": "5:27:55"}
134
+ {"current_steps": 1340, "total_steps": 2156, "loss": 1.4015, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.5685896970466123e-05, "epoch": 1.24, "percentage": 62.15, "elapsed_time": "8:51:53", "remaining_time": "5:23:53"}
135
+ {"current_steps": 1350, "total_steps": 2156, "loss": 1.3797, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.534883902406375e-05, "epoch": 1.25, "percentage": 62.62, "elapsed_time": "8:55:45", "remaining_time": "5:19:51"}
136
+ {"current_steps": 1360, "total_steps": 2156, "loss": 1.3931, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.5013830229346326e-05, "epoch": 1.26, "percentage": 63.08, "elapsed_time": "8:59:40", "remaining_time": "5:15:52"}
137
+ {"current_steps": 1370, "total_steps": 2156, "loss": 1.3946, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.4680941715977722e-05, "epoch": 1.27, "percentage": 63.54, "elapsed_time": "9:03:37", "remaining_time": "5:11:53"}
138
+ {"current_steps": 1380, "total_steps": 2156, "loss": 1.3851, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.4350244163439892e-05, "epoch": 1.28, "percentage": 64.01, "elapsed_time": "9:07:33", "remaining_time": "5:07:54"}
139
+ {"current_steps": 1390, "total_steps": 2156, "loss": 1.3891, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.4021807786026108e-05, "epoch": 1.29, "percentage": 64.47, "elapsed_time": "9:11:38", "remaining_time": "5:03:59"}
140
+ {"current_steps": 1400, "total_steps": 2156, "loss": 1.3958, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.3695702317932862e-05, "epoch": 1.3, "percentage": 64.94, "elapsed_time": "9:15:33", "remaining_time": "4:59:59"}
141
+ {"current_steps": 1410, "total_steps": 2156, "loss": 1.3909, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.337199699845387e-05, "epoch": 1.31, "percentage": 65.4, "elapsed_time": "9:19:20", "remaining_time": "4:55:56"}
142
+ {"current_steps": 1420, "total_steps": 2156, "loss": 1.3785, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.3050760557279023e-05, "epoch": 1.32, "percentage": 65.86, "elapsed_time": "9:23:32", "remaining_time": "4:52:05"}
143
+ {"current_steps": 1430, "total_steps": 2156, "loss": 1.3882, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.2732061199901562e-05, "epoch": 1.33, "percentage": 66.33, "elapsed_time": "9:27:34", "remaining_time": "4:48:09"}
144
+ {"current_steps": 1440, "total_steps": 2156, "loss": 1.3872, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.2415966593136547e-05, "epoch": 1.34, "percentage": 66.79, "elapsed_time": "9:31:33", "remaining_time": "4:44:11"}
145
+ {"current_steps": 1450, "total_steps": 2156, "loss": 1.4032, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.2102543850753808e-05, "epoch": 1.34, "percentage": 67.25, "elapsed_time": "9:35:31", "remaining_time": "4:40:13"}
146
+ {"current_steps": 1460, "total_steps": 2156, "loss": 1.3928, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1791859519228138e-05, "epoch": 1.35, "percentage": 67.72, "elapsed_time": "9:39:28", "remaining_time": "4:36:14"}
147
+ {"current_steps": 1470, "total_steps": 2156, "loss": 1.3785, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.148397956361007e-05, "epoch": 1.36, "percentage": 68.18, "elapsed_time": "9:43:33", "remaining_time": "4:32:19"}
148
+ {"current_steps": 1480, "total_steps": 2156, "loss": 1.3797, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1178969353520018e-05, "epoch": 1.37, "percentage": 68.65, "elapsed_time": "9:47:27", "remaining_time": "4:28:19"}
149
+ {"current_steps": 1490, "total_steps": 2156, "loss": 1.3966, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.087689364926897e-05, "epoch": 1.38, "percentage": 69.11, "elapsed_time": "9:51:30", "remaining_time": "4:24:23"}
150
+ {"current_steps": 1500, "total_steps": 2156, "loss": 1.3849, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.0577816588108378e-05, "epoch": 1.39, "percentage": 69.57, "elapsed_time": "9:55:21", "remaining_time": "4:20:22"}
151
+ {"current_steps": 1510, "total_steps": 2156, "loss": 1.3732, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.0281801670612449e-05, "epoch": 1.4, "percentage": 70.04, "elapsed_time": "9:59:19", "remaining_time": "4:16:23"}
152
+ {"current_steps": 1520, "total_steps": 2156, "loss": 1.4018, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.988911747195603e-06, "epoch": 1.41, "percentage": 70.5, "elapsed_time": "10:03:22", "remaining_time": "4:12:27"}
153
+ {"current_steps": 1530, "total_steps": 2156, "loss": 1.3902, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.699209004767953e-06, "epoch": 1.42, "percentage": 70.96, "elapsed_time": "10:07:13", "remaining_time": "4:08:26"}
154
+ {"current_steps": 1540, "total_steps": 2156, "loss": 1.3991, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.412754953531663e-06, "epoch": 1.43, "percentage": 71.43, "elapsed_time": "10:11:04", "remaining_time": "4:04:25"}
155
+ {"current_steps": 1550, "total_steps": 2156, "loss": 1.3824, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.129610413921005e-06, "epoch": 1.44, "percentage": 71.89, "elapsed_time": "10:14:57", "remaining_time": "4:00:25"}
156
+ {"current_steps": 1560, "total_steps": 2156, "loss": 1.3904, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.849835503688847e-06, "epoch": 1.45, "percentage": 72.36, "elapsed_time": "10:18:49", "remaining_time": "3:56:25"}
157
+ {"current_steps": 1570, "total_steps": 2156, "loss": 1.3937, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.57348962514234e-06, "epoch": 1.46, "percentage": 72.82, "elapsed_time": "10:22:47", "remaining_time": "3:52:27"}
158
+ {"current_steps": 1580, "total_steps": 2156, "loss": 1.39, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.30063145253053e-06, "epoch": 1.46, "percentage": 73.28, "elapsed_time": "10:26:57", "remaining_time": "3:48:33"}
159
+ {"current_steps": 1590, "total_steps": 2156, "loss": 1.3773, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.031318919586523e-06, "epoch": 1.47, "percentage": 73.75, "elapsed_time": "10:30:54", "remaining_time": "3:44:35"}
160
+ {"current_steps": 1600, "total_steps": 2156, "loss": 1.3875, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.7656092072269e-06, "epoch": 1.48, "percentage": 74.21, "elapsed_time": "10:34:53", "remaining_time": "3:40:37"}
161
+ {"current_steps": 1610, "total_steps": 2156, "loss": 1.3859, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.503558731410959e-06, "epoch": 1.49, "percentage": 74.68, "elapsed_time": "10:38:44", "remaining_time": "3:36:37"}
162
+ {"current_steps": 1620, "total_steps": 2156, "loss": 1.3716, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.245223131162376e-06, "epoch": 1.5, "percentage": 75.14, "elapsed_time": "10:42:38", "remaining_time": "3:32:37"}
163
+ {"current_steps": 1630, "total_steps": 2156, "loss": 1.3846, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.9906572567558285e-06, "epoch": 1.51, "percentage": 75.6, "elapsed_time": "10:46:30", "remaining_time": "3:28:37"}
164
+ {"current_steps": 1640, "total_steps": 2156, "loss": 1.3741, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.739915158071106e-06, "epoch": 1.52, "percentage": 76.07, "elapsed_time": "10:50:24", "remaining_time": "3:24:38"}
165
+ {"current_steps": 1650, "total_steps": 2156, "loss": 1.3827, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.493050073117116e-06, "epoch": 1.53, "percentage": 76.53, "elapsed_time": "10:54:22", "remaining_time": "3:20:40"}
166
+ {"current_steps": 1660, "total_steps": 2156, "loss": 1.3951, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.250114416728298e-06, "epoch": 1.54, "percentage": 76.99, "elapsed_time": "10:58:20", "remaining_time": "3:16:42"}
167
+ {"current_steps": 1670, "total_steps": 2156, "loss": 1.3875, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.011159769435823e-06, "epoch": 1.55, "percentage": 77.46, "elapsed_time": "11:02:18", "remaining_time": "3:12:44"}
168
+ {"current_steps": 1680, "total_steps": 2156, "loss": 1.3783, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.776236866515947e-06, "epoch": 1.56, "percentage": 77.92, "elapsed_time": "11:06:17", "remaining_time": "3:08:46"}
169
+ {"current_steps": 1690, "total_steps": 2156, "loss": 1.3888, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.545395587217763e-06, "epoch": 1.57, "percentage": 78.39, "elapsed_time": "11:10:20", "remaining_time": "3:04:50"}
170
+ {"current_steps": 1700, "total_steps": 2156, "loss": 1.4006, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.318684944172752e-06, "epoch": 1.58, "percentage": 78.85, "elapsed_time": "11:14:18", "remaining_time": "3:00:52"}
171
+ {"current_steps": 1710, "total_steps": 2156, "loss": 1.4041, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.096153072988366e-06, "epoch": 1.59, "percentage": 79.31, "elapsed_time": "11:18:08", "remaining_time": "2:56:52"}
172
+ {"current_steps": 1720, "total_steps": 2156, "loss": 1.3865, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.8778472220277874e-06, "epoch": 1.59, "percentage": 79.78, "elapsed_time": "11:21:59", "remaining_time": "2:52:52"}
173
+ {"current_steps": 1730, "total_steps": 2156, "loss": 1.3789, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.6638137423780395e-06, "epoch": 1.6, "percentage": 80.24, "elapsed_time": "11:26:02", "remaining_time": "2:48:55"}
174
+ {"current_steps": 1740, "total_steps": 2156, "loss": 1.3861, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.454098078008667e-06, "epoch": 1.61, "percentage": 80.71, "elapsed_time": "11:30:04", "remaining_time": "2:44:59"}
175
+ {"current_steps": 1750, "total_steps": 2156, "loss": 1.3803, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.248744756122986e-06, "epoch": 1.62, "percentage": 81.17, "elapsed_time": "11:34:05", "remaining_time": "2:41:01"}
176
+ {"current_steps": 1760, "total_steps": 2156, "loss": 1.3695, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.047797377703985e-06, "epoch": 1.63, "percentage": 81.63, "elapsed_time": "11:38:04", "remaining_time": "2:37:04"}
177
+ {"current_steps": 1770, "total_steps": 2156, "loss": 1.3937, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.851298608256892e-06, "epoch": 1.64, "percentage": 82.1, "elapsed_time": "11:42:06", "remaining_time": "2:33:06"}
178
+ {"current_steps": 1780, "total_steps": 2156, "loss": 1.3865, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.6592901687503566e-06, "epoch": 1.65, "percentage": 82.56, "elapsed_time": "11:46:05", "remaining_time": "2:29:09"}
179
+ {"current_steps": 1790, "total_steps": 2156, "loss": 1.3799, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.471812826758178e-06, "epoch": 1.66, "percentage": 83.02, "elapsed_time": "11:50:08", "remaining_time": "2:25:12"}
180
+ {"current_steps": 1800, "total_steps": 2156, "loss": 1.3911, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.288906387803464e-06, "epoch": 1.67, "percentage": 83.49, "elapsed_time": "11:54:06", "remaining_time": "2:21:14"}
181
+ {"current_steps": 1810, "total_steps": 2156, "loss": 1.3801, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.1106096869070483e-06, "epoch": 1.68, "percentage": 83.95, "elapsed_time": "11:58:01", "remaining_time": "2:17:15"}
182
+ {"current_steps": 1820, "total_steps": 2156, "loss": 1.382, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9369605803419715e-06, "epoch": 1.69, "percentage": 84.42, "elapsed_time": "12:02:04", "remaining_time": "2:13:18"}
183
+ {"current_steps": 1830, "total_steps": 2156, "loss": 1.3898, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.767995937595766e-06, "epoch": 1.7, "percentage": 84.88, "elapsed_time": "12:06:04", "remaining_time": "2:09:20"}
184
+ {"current_steps": 1840, "total_steps": 2156, "loss": 1.3752, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.6037516335422728e-06, "epoch": 1.71, "percentage": 85.34, "elapsed_time": "12:10:12", "remaining_time": "2:05:24"}
185
+ {"current_steps": 1850, "total_steps": 2156, "loss": 1.3855, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.4442625408246074e-06, "epoch": 1.72, "percentage": 85.81, "elapsed_time": "12:14:13", "remaining_time": "2:01:26"}
186
+ {"current_steps": 1860, "total_steps": 2156, "loss": 1.3665, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.289562522450947e-06, "epoch": 1.72, "percentage": 86.27, "elapsed_time": "12:18:04", "remaining_time": "1:57:27"}
187
+ {"current_steps": 1870, "total_steps": 2156, "loss": 1.3747, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.1396844246046903e-06, "epoch": 1.73, "percentage": 86.73, "elapsed_time": "12:22:07", "remaining_time": "1:53:30"}
188
+ {"current_steps": 1880, "total_steps": 2156, "loss": 1.3861, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.9946600696704592e-06, "epoch": 1.74, "percentage": 87.2, "elapsed_time": "12:25:58", "remaining_time": "1:49:30"}
189
+ {"current_steps": 1890, "total_steps": 2156, "loss": 1.3844, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.8545202494775509e-06, "epoch": 1.75, "percentage": 87.66, "elapsed_time": "12:29:57", "remaining_time": "1:45:32"}
190
+ {"current_steps": 1900, "total_steps": 2156, "loss": 1.3796, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.7192947187621434e-06, "epoch": 1.76, "percentage": 88.13, "elapsed_time": "12:33:59", "remaining_time": "1:41:35"}
191
+ {"current_steps": 1910, "total_steps": 2156, "loss": 1.3963, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.5890121888497366e-06, "epoch": 1.77, "percentage": 88.59, "elapsed_time": "12:37:55", "remaining_time": "1:37:37"}
192
+ {"current_steps": 1920, "total_steps": 2156, "loss": 1.385, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.463700321559075e-06, "epoch": 1.78, "percentage": 89.05, "elapsed_time": "12:41:56", "remaining_time": "1:33:39"}
193
+ {"current_steps": 1930, "total_steps": 2156, "loss": 1.3811, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.3433857233289714e-06, "epoch": 1.79, "percentage": 89.52, "elapsed_time": "12:45:53", "remaining_time": "1:29:41"}
194
+ {"current_steps": 1940, "total_steps": 2156, "loss": 1.384, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.2280939395691859e-06, "epoch": 1.8, "percentage": 89.98, "elapsed_time": "12:49:48", "remaining_time": "1:25:42"}
195
+ {"current_steps": 1950, "total_steps": 2156, "loss": 1.3821, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1178494492365465e-06, "epoch": 1.81, "percentage": 90.45, "elapsed_time": "12:53:43", "remaining_time": "1:21:44"}
196
+ {"current_steps": 1960, "total_steps": 2156, "loss": 1.3978, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.0126756596375686e-06, "epoch": 1.82, "percentage": 90.91, "elapsed_time": "12:57:51", "remaining_time": "1:17:47"}
197
+ {"current_steps": 1970, "total_steps": 2156, "loss": 1.3724, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.125949014585383e-07, "epoch": 1.83, "percentage": 91.37, "elapsed_time": "13:02:04", "remaining_time": "1:13:50"}
198
+ {"current_steps": 1980, "total_steps": 2156, "loss": 1.3851, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.176284240242638e-07, "epoch": 1.84, "percentage": 91.84, "elapsed_time": "13:06:07", "remaining_time": "1:09:52"}
199
+ {"current_steps": 1990, "total_steps": 2156, "loss": 1.3793, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.277963907863478e-07, "epoch": 1.85, "percentage": 92.3, "elapsed_time": "13:10:10", "remaining_time": "1:05:54"}
200
+ {"current_steps": 2000, "total_steps": 2156, "loss": 1.3946, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.431178750420513e-07, "epoch": 1.85, "percentage": 92.76, "elapsed_time": "13:14:06", "remaining_time": "1:01:56"}
201
+ {"current_steps": 2010, "total_steps": 2156, "loss": 1.3794, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.636108558846359e-07, "epoch": 1.86, "percentage": 93.23, "elapsed_time": "13:18:10", "remaining_time": "0:57:58"}
202
+ {"current_steps": 2020, "total_steps": 2156, "loss": 1.3886, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.892922143859918e-07, "epoch": 1.87, "percentage": 93.69, "elapsed_time": "13:22:06", "remaining_time": "0:54:00"}
203
+ {"current_steps": 2030, "total_steps": 2156, "loss": 1.3794, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.201777300124249e-07, "epoch": 1.88, "percentage": 94.16, "elapsed_time": "13:26:06", "remaining_time": "0:50:02"}
204
+ {"current_steps": 2040, "total_steps": 2156, "loss": 1.3828, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.562820772743392e-07, "epoch": 1.89, "percentage": 94.62, "elapsed_time": "13:30:14", "remaining_time": "0:46:04"}
205
+ {"current_steps": 2050, "total_steps": 2156, "loss": 1.374, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9761882261050143e-07, "epoch": 1.9, "percentage": 95.08, "elapsed_time": "13:34:17", "remaining_time": "0:42:06"}
206
+ {"current_steps": 2060, "total_steps": 2156, "loss": 1.3804, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.4420042150761215e-07, "epoch": 1.91, "percentage": 95.55, "elapsed_time": "13:38:15", "remaining_time": "0:38:07"}
207
+ {"current_steps": 2070, "total_steps": 2156, "loss": 1.3772, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.9603821585572058e-07, "epoch": 1.92, "percentage": 96.01, "elapsed_time": "13:42:17", "remaining_time": "0:34:09"}
208
+ {"current_steps": 2080, "total_steps": 2156, "loss": 1.3913, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.531424315400931e-07, "epoch": 1.93, "percentage": 96.47, "elapsed_time": "13:46:29", "remaining_time": "0:30:11"}
209
+ {"current_steps": 2090, "total_steps": 2156, "loss": 1.3859, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1552217627004425e-07, "epoch": 1.94, "percentage": 96.94, "elapsed_time": "13:50:32", "remaining_time": "0:26:13"}
210
+ {"current_steps": 2100, "total_steps": 2156, "loss": 1.4035, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.318543764516961e-08, "epoch": 1.95, "percentage": 97.4, "elapsed_time": "13:54:40", "remaining_time": "0:22:15"}
211
+ {"current_steps": 2110, "total_steps": 2156, "loss": 1.4047, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.613908145939428e-08, "epoch": 1.96, "percentage": 97.87, "elapsed_time": "13:58:37", "remaining_time": "0:18:16"}
212
+ {"current_steps": 2120, "total_steps": 2156, "loss": 1.3886, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.438885024322769e-08, "epoch": 1.97, "percentage": 98.33, "elapsed_time": "14:02:39", "remaining_time": "0:14:18"}
213
+ {"current_steps": 2130, "total_steps": 2156, "loss": 1.3874, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.7939362044494534e-08, "epoch": 1.97, "percentage": 98.79, "elapsed_time": "14:06:37", "remaining_time": "0:10:20"}
214
+ {"current_steps": 2140, "total_steps": 2156, "loss": 1.3852, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.794109447824659e-09, "epoch": 1.98, "percentage": 99.26, "elapsed_time": "14:10:36", "remaining_time": "0:06:21"}
215
+ {"current_steps": 2150, "total_steps": 2156, "loss": 1.3859, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.554588330934478e-10, "epoch": 1.99, "percentage": 99.72, "elapsed_time": "14:14:42", "remaining_time": "0:02:23"}
216
+ {"current_steps": 2156, "total_steps": 2156, "loss": null, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "14:17:05", "remaining_time": "0:00:00"}
trainer_state.json ADDED
@@ -0,0 +1,1315 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.999072786277237,
5
+ "global_step": 2156,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 4.999734597774032e-05,
13
+ "loss": 1.7592,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.998938447446803e-05,
19
+ "loss": 1.6546,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.997611718058365e-05,
25
+ "loss": 1.5632,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.9957546913022665e-05,
31
+ "loss": 1.5245,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.993367761465736e-05,
37
+ "loss": 1.5049,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 4.9904514353459654e-05,
43
+ "loss": 1.492,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.06,
48
+ "learning_rate": 4.9870063321425105e-05,
49
+ "loss": 1.4803,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.07,
54
+ "learning_rate": 4.983033183325818e-05,
55
+ "loss": 1.4827,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.08,
60
+ "learning_rate": 4.97853283248192e-05,
61
+ "loss": 1.4866,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.09,
66
+ "learning_rate": 4.973506235133323e-05,
67
+ "loss": 1.486,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.1,
72
+ "learning_rate": 4.967954458536126e-05,
73
+ "loss": 1.4822,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.11,
78
+ "learning_rate": 4.9618786814534226e-05,
79
+ "loss": 1.4801,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.12,
84
+ "learning_rate": 4.955280193905022e-05,
85
+ "loss": 1.4764,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.13,
90
+ "learning_rate": 4.948160396893553e-05,
91
+ "loss": 1.4662,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.14,
96
+ "learning_rate": 4.9405208021069946e-05,
97
+ "loss": 1.4627,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.15,
102
+ "learning_rate": 4.9323630315977156e-05,
103
+ "loss": 1.4543,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.16,
108
+ "learning_rate": 4.9236888174380784e-05,
109
+ "loss": 1.4508,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.17,
114
+ "learning_rate": 4.91450000135268e-05,
115
+ "loss": 1.464,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.18,
120
+ "learning_rate": 4.9047985343273154e-05,
121
+ "loss": 1.4705,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.19,
126
+ "learning_rate": 4.894586476194739e-05,
127
+ "loss": 1.4588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.19,
132
+ "learning_rate": 4.883865995197319e-05,
133
+ "loss": 1.4532,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.2,
138
+ "learning_rate": 4.8726393675266716e-05,
139
+ "loss": 1.4629,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.21,
144
+ "learning_rate": 4.860908976840376e-05,
145
+ "loss": 1.4429,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.22,
150
+ "learning_rate": 4.848677313755872e-05,
151
+ "loss": 1.441,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.23,
156
+ "learning_rate": 4.835946975321647e-05,
157
+ "loss": 1.4496,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.24,
162
+ "learning_rate": 4.822720664465827e-05,
163
+ "loss": 1.4459,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.25,
168
+ "learning_rate": 4.809001189422287e-05,
169
+ "loss": 1.4458,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.26,
174
+ "learning_rate": 4.794791463134399e-05,
175
+ "loss": 1.4484,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.27,
180
+ "learning_rate": 4.780094502636552e-05,
181
+ "loss": 1.4541,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.28,
186
+ "learning_rate": 4.764913428413572e-05,
187
+ "loss": 1.437,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.29,
192
+ "learning_rate": 4.7492514637381727e-05,
193
+ "loss": 1.4368,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.3,
198
+ "learning_rate": 4.733111933986583e-05,
199
+ "loss": 1.4449,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.31,
204
+ "learning_rate": 4.716498265932501e-05,
205
+ "loss": 1.4256,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.32,
210
+ "learning_rate": 4.699413987019512e-05,
211
+ "loss": 1.4378,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.32,
216
+ "learning_rate": 4.681862724612141e-05,
217
+ "loss": 1.454,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.33,
222
+ "learning_rate": 4.663848205225674e-05,
223
+ "loss": 1.4207,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.34,
228
+ "learning_rate": 4.645374253734949e-05,
229
+ "loss": 1.4286,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.35,
234
+ "learning_rate": 4.626444792562244e-05,
235
+ "loss": 1.4244,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.36,
240
+ "learning_rate": 4.607063840844463e-05,
241
+ "loss": 1.4342,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.37,
246
+ "learning_rate": 4.587235513579791e-05,
247
+ "loss": 1.4241,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.38,
252
+ "learning_rate": 4.5669640207539786e-05,
253
+ "loss": 1.443,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.39,
258
+ "learning_rate": 4.546253666446484e-05,
259
+ "loss": 1.4281,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.4,
264
+ "learning_rate": 4.525108847916614e-05,
265
+ "loss": 1.4193,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.41,
270
+ "learning_rate": 4.503534054669892e-05,
271
+ "loss": 1.4154,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.42,
276
+ "learning_rate": 4.481533867504841e-05,
277
+ "loss": 1.4273,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.43,
282
+ "learning_rate": 4.4591129575403765e-05,
283
+ "loss": 1.4329,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.44,
288
+ "learning_rate": 4.43627608522403e-05,
289
+ "loss": 1.4102,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.45,
294
+ "learning_rate": 4.4130280993211974e-05,
295
+ "loss": 1.4244,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.45,
300
+ "learning_rate": 4.389373935885646e-05,
301
+ "loss": 1.4407,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.46,
306
+ "learning_rate": 4.365318617211479e-05,
307
+ "loss": 1.4282,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.47,
312
+ "learning_rate": 4.340867250766794e-05,
313
+ "loss": 1.4269,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.48,
318
+ "learning_rate": 4.316025028109258e-05,
319
+ "loss": 1.43,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.49,
324
+ "learning_rate": 4.2907972237838225e-05,
325
+ "loss": 1.4271,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.5,
330
+ "learning_rate": 4.2651891942028274e-05,
331
+ "loss": 1.4321,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.51,
336
+ "learning_rate": 4.239206376508717e-05,
337
+ "loss": 1.4083,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.52,
342
+ "learning_rate": 4.212854287419611e-05,
343
+ "loss": 1.4168,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.53,
348
+ "learning_rate": 4.1861385220579934e-05,
349
+ "loss": 1.4144,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.54,
354
+ "learning_rate": 4.1590647527627404e-05,
355
+ "loss": 1.4106,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.55,
360
+ "learning_rate": 4.131638727884762e-05,
361
+ "loss": 1.4202,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.56,
366
+ "learning_rate": 4.103866270566498e-05,
367
+ "loss": 1.417,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.57,
372
+ "learning_rate": 4.075753277505544e-05,
373
+ "loss": 1.4324,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.57,
378
+ "learning_rate": 4.0473057177026484e-05,
379
+ "loss": 1.4147,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.58,
384
+ "learning_rate": 4.018529631194369e-05,
385
+ "loss": 1.4269,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.59,
390
+ "learning_rate": 3.989431127770635e-05,
391
+ "loss": 1.4133,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.6,
396
+ "learning_rate": 3.960016385677513e-05,
397
+ "loss": 1.4173,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.61,
402
+ "learning_rate": 3.9302916503054246e-05,
403
+ "loss": 1.4007,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.62,
408
+ "learning_rate": 3.9002632328631164e-05,
409
+ "loss": 1.4117,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.63,
414
+ "learning_rate": 3.8699375090376534e-05,
415
+ "loss": 1.4247,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.64,
420
+ "learning_rate": 3.8393209176407223e-05,
421
+ "loss": 1.4312,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.65,
426
+ "learning_rate": 3.8084199592415305e-05,
427
+ "loss": 1.4136,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.66,
432
+ "learning_rate": 3.777241194786591e-05,
433
+ "loss": 1.4132,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.67,
438
+ "learning_rate": 3.745791244206697e-05,
439
+ "loss": 1.4145,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.68,
444
+ "learning_rate": 3.714076785011359e-05,
445
+ "loss": 1.3951,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.69,
450
+ "learning_rate": 3.682104550871031e-05,
451
+ "loss": 1.4139,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.7,
456
+ "learning_rate": 3.649881330187401e-05,
457
+ "loss": 1.4078,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.7,
462
+ "learning_rate": 3.617413964652067e-05,
463
+ "loss": 1.4048,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.71,
468
+ "learning_rate": 3.5847093477938956e-05,
469
+ "loss": 1.4115,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.72,
474
+ "learning_rate": 3.551774423515378e-05,
475
+ "loss": 1.4018,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.73,
480
+ "learning_rate": 3.518616184618288e-05,
481
+ "loss": 1.4091,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.74,
486
+ "learning_rate": 3.4852416713189526e-05,
487
+ "loss": 1.4135,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.75,
492
+ "learning_rate": 3.4516579697534705e-05,
493
+ "loss": 1.3977,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.76,
498
+ "learning_rate": 3.417872210473162e-05,
499
+ "loss": 1.4123,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.77,
504
+ "learning_rate": 3.3838915669306034e-05,
505
+ "loss": 1.4096,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.78,
510
+ "learning_rate": 3.349723253956542e-05,
511
+ "loss": 1.4031,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.79,
516
+ "learning_rate": 3.315374526228036e-05,
517
+ "loss": 1.4104,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.8,
522
+ "learning_rate": 3.2808526767281225e-05,
523
+ "loss": 1.3972,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.81,
528
+ "learning_rate": 3.246165035197364e-05,
529
+ "loss": 1.3987,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.82,
534
+ "learning_rate": 3.211318966577581e-05,
535
+ "loss": 1.4197,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.83,
540
+ "learning_rate": 3.176321869448116e-05,
541
+ "loss": 1.4169,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.83,
546
+ "learning_rate": 3.1411811744549536e-05,
547
+ "loss": 1.4055,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.84,
552
+ "learning_rate": 3.105904342733032e-05,
553
+ "loss": 1.402,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.85,
558
+ "learning_rate": 3.070498864322081e-05,
559
+ "loss": 1.4148,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.86,
564
+ "learning_rate": 3.034972256576328e-05,
565
+ "loss": 1.4007,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.87,
570
+ "learning_rate": 2.999332062568395e-05,
571
+ "loss": 1.3926,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.88,
576
+ "learning_rate": 2.9635858494877384e-05,
577
+ "loss": 1.4099,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.89,
582
+ "learning_rate": 2.9277412070339782e-05,
583
+ "loss": 1.401,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.9,
588
+ "learning_rate": 2.891805745805429e-05,
589
+ "loss": 1.3995,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.91,
594
+ "learning_rate": 2.8557870956832132e-05,
595
+ "loss": 1.4108,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.92,
600
+ "learning_rate": 2.8196929042112652e-05,
601
+ "loss": 1.4006,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.93,
606
+ "learning_rate": 2.783530834972594e-05,
607
+ "loss": 1.4045,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.94,
612
+ "learning_rate": 2.7473085659621377e-05,
613
+ "loss": 1.3907,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.95,
618
+ "learning_rate": 2.711033787956555e-05,
619
+ "loss": 1.3872,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.96,
624
+ "learning_rate": 2.6747142028813105e-05,
625
+ "loss": 1.405,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.96,
630
+ "learning_rate": 2.638357522175383e-05,
631
+ "loss": 1.3952,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.97,
636
+ "learning_rate": 2.6019714651539646e-05,
637
+ "loss": 1.3901,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.98,
642
+ "learning_rate": 2.565563757369475e-05,
643
+ "loss": 1.4027,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.99,
648
+ "learning_rate": 2.529142128971268e-05,
649
+ "loss": 1.3909,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 1.0,
654
+ "learning_rate": 2.492714313064342e-05,
655
+ "loss": 1.4039,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 1.01,
660
+ "learning_rate": 2.45628804406744e-05,
661
+ "loss": 1.3964,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 1.02,
666
+ "learning_rate": 2.419871056070862e-05,
667
+ "loss": 1.4076,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 1.03,
672
+ "learning_rate": 2.3834710811943514e-05,
673
+ "loss": 1.3891,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 1.04,
678
+ "learning_rate": 2.3470958479453938e-05,
679
+ "loss": 1.4004,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 1.05,
684
+ "learning_rate": 2.3107530795782877e-05,
685
+ "loss": 1.3879,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 1.06,
690
+ "learning_rate": 2.2744504924543313e-05,
691
+ "loss": 1.3879,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 1.07,
696
+ "learning_rate": 2.23819579440347e-05,
697
+ "loss": 1.3851,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 1.08,
702
+ "learning_rate": 2.2019966830877545e-05,
703
+ "loss": 1.3846,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 1.08,
708
+ "learning_rate": 2.1658608443669635e-05,
709
+ "loss": 1.3936,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 1.09,
714
+ "learning_rate": 2.1297959506667224e-05,
715
+ "loss": 1.3841,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 1.1,
720
+ "learning_rate": 2.0938096593494855e-05,
721
+ "loss": 1.3969,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 1.11,
726
+ "learning_rate": 2.057909611088709e-05,
727
+ "loss": 1.3953,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 1.12,
732
+ "learning_rate": 2.02210342824657e-05,
733
+ "loss": 1.3951,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 1.13,
738
+ "learning_rate": 1.9863987132555706e-05,
739
+ "loss": 1.4012,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 1.14,
744
+ "learning_rate": 1.9508030470043806e-05,
745
+ "loss": 1.3998,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 1.15,
750
+ "learning_rate": 1.915323987228247e-05,
751
+ "loss": 1.3933,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 1.16,
756
+ "learning_rate": 1.8799690669043212e-05,
757
+ "loss": 1.3977,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 1.17,
762
+ "learning_rate": 1.8447457926522454e-05,
763
+ "loss": 1.3818,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 1.18,
768
+ "learning_rate": 1.8096616431403325e-05,
769
+ "loss": 1.4032,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 1.19,
774
+ "learning_rate": 1.7747240674976857e-05,
775
+ "loss": 1.3853,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 1.2,
780
+ "learning_rate": 1.7399404837325796e-05,
781
+ "loss": 1.3988,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 1.21,
786
+ "learning_rate": 1.7053182771574633e-05,
787
+ "loss": 1.3963,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 1.21,
792
+ "learning_rate": 1.6708647988208887e-05,
793
+ "loss": 1.3881,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 1.22,
798
+ "learning_rate": 1.6365873639467315e-05,
799
+ "loss": 1.3804,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 1.23,
804
+ "learning_rate": 1.602493250381003e-05,
805
+ "loss": 1.3952,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 1.24,
810
+ "learning_rate": 1.5685896970466123e-05,
811
+ "loss": 1.4015,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 1.25,
816
+ "learning_rate": 1.534883902406375e-05,
817
+ "loss": 1.3797,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 1.26,
822
+ "learning_rate": 1.5013830229346326e-05,
823
+ "loss": 1.3931,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 1.27,
828
+ "learning_rate": 1.4680941715977722e-05,
829
+ "loss": 1.3946,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 1.28,
834
+ "learning_rate": 1.4350244163439892e-05,
835
+ "loss": 1.3851,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 1.29,
840
+ "learning_rate": 1.4021807786026108e-05,
841
+ "loss": 1.3891,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 1.3,
846
+ "learning_rate": 1.3695702317932862e-05,
847
+ "loss": 1.3958,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 1.31,
852
+ "learning_rate": 1.337199699845387e-05,
853
+ "loss": 1.3909,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 1.32,
858
+ "learning_rate": 1.3050760557279023e-05,
859
+ "loss": 1.3785,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 1.33,
864
+ "learning_rate": 1.2732061199901562e-05,
865
+ "loss": 1.3882,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 1.34,
870
+ "learning_rate": 1.2415966593136547e-05,
871
+ "loss": 1.3872,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 1.34,
876
+ "learning_rate": 1.2102543850753808e-05,
877
+ "loss": 1.4032,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 1.35,
882
+ "learning_rate": 1.1791859519228138e-05,
883
+ "loss": 1.3928,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 1.36,
888
+ "learning_rate": 1.148397956361007e-05,
889
+ "loss": 1.3785,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 1.37,
894
+ "learning_rate": 1.1178969353520018e-05,
895
+ "loss": 1.3797,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 1.38,
900
+ "learning_rate": 1.087689364926897e-05,
901
+ "loss": 1.3966,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 1.39,
906
+ "learning_rate": 1.0577816588108378e-05,
907
+ "loss": 1.3849,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 1.4,
912
+ "learning_rate": 1.0281801670612449e-05,
913
+ "loss": 1.3732,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 1.41,
918
+ "learning_rate": 9.988911747195603e-06,
919
+ "loss": 1.4018,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 1.42,
924
+ "learning_rate": 9.699209004767953e-06,
925
+ "loss": 1.3902,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 1.43,
930
+ "learning_rate": 9.412754953531663e-06,
931
+ "loss": 1.3991,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 1.44,
936
+ "learning_rate": 9.129610413921005e-06,
937
+ "loss": 1.3824,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 1.45,
942
+ "learning_rate": 8.849835503688847e-06,
943
+ "loss": 1.3904,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 1.46,
948
+ "learning_rate": 8.57348962514234e-06,
949
+ "loss": 1.3937,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 1.46,
954
+ "learning_rate": 8.30063145253053e-06,
955
+ "loss": 1.39,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 1.47,
960
+ "learning_rate": 8.031318919586523e-06,
961
+ "loss": 1.3773,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 1.48,
966
+ "learning_rate": 7.7656092072269e-06,
967
+ "loss": 1.3875,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 1.49,
972
+ "learning_rate": 7.503558731410959e-06,
973
+ "loss": 1.3859,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 1.5,
978
+ "learning_rate": 7.245223131162376e-06,
979
+ "loss": 1.3716,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 1.51,
984
+ "learning_rate": 6.9906572567558285e-06,
985
+ "loss": 1.3846,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 1.52,
990
+ "learning_rate": 6.739915158071106e-06,
991
+ "loss": 1.3741,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 1.53,
996
+ "learning_rate": 6.493050073117116e-06,
997
+ "loss": 1.3827,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 1.54,
1002
+ "learning_rate": 6.250114416728298e-06,
1003
+ "loss": 1.3951,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 1.55,
1008
+ "learning_rate": 6.011159769435823e-06,
1009
+ "loss": 1.3875,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 1.56,
1014
+ "learning_rate": 5.776236866515947e-06,
1015
+ "loss": 1.3783,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 1.57,
1020
+ "learning_rate": 5.545395587217763e-06,
1021
+ "loss": 1.3888,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 1.58,
1026
+ "learning_rate": 5.318684944172752e-06,
1027
+ "loss": 1.4006,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 1.59,
1032
+ "learning_rate": 5.096153072988366e-06,
1033
+ "loss": 1.4041,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 1.59,
1038
+ "learning_rate": 4.8778472220277874e-06,
1039
+ "loss": 1.3865,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 1.6,
1044
+ "learning_rate": 4.6638137423780395e-06,
1045
+ "loss": 1.3789,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 1.61,
1050
+ "learning_rate": 4.454098078008667e-06,
1051
+ "loss": 1.3861,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 1.62,
1056
+ "learning_rate": 4.248744756122986e-06,
1057
+ "loss": 1.3803,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 1.63,
1062
+ "learning_rate": 4.047797377703985e-06,
1063
+ "loss": 1.3695,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 1.64,
1068
+ "learning_rate": 3.851298608256892e-06,
1069
+ "loss": 1.3937,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 1.65,
1074
+ "learning_rate": 3.6592901687503566e-06,
1075
+ "loss": 1.3865,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 1.66,
1080
+ "learning_rate": 3.471812826758178e-06,
1081
+ "loss": 1.3799,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 1.67,
1086
+ "learning_rate": 3.288906387803464e-06,
1087
+ "loss": 1.3911,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 1.68,
1092
+ "learning_rate": 3.1106096869070483e-06,
1093
+ "loss": 1.3801,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 1.69,
1098
+ "learning_rate": 2.9369605803419715e-06,
1099
+ "loss": 1.382,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 1.7,
1104
+ "learning_rate": 2.767995937595766e-06,
1105
+ "loss": 1.3898,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 1.71,
1110
+ "learning_rate": 2.6037516335422728e-06,
1111
+ "loss": 1.3752,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 1.72,
1116
+ "learning_rate": 2.4442625408246074e-06,
1117
+ "loss": 1.3855,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 1.72,
1122
+ "learning_rate": 2.289562522450947e-06,
1123
+ "loss": 1.3665,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 1.73,
1128
+ "learning_rate": 2.1396844246046903e-06,
1129
+ "loss": 1.3747,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 1.74,
1134
+ "learning_rate": 1.9946600696704592e-06,
1135
+ "loss": 1.3861,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 1.75,
1140
+ "learning_rate": 1.8545202494775509e-06,
1141
+ "loss": 1.3844,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 1.76,
1146
+ "learning_rate": 1.7192947187621434e-06,
1147
+ "loss": 1.3796,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 1.77,
1152
+ "learning_rate": 1.5890121888497366e-06,
1153
+ "loss": 1.3963,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 1.78,
1158
+ "learning_rate": 1.463700321559075e-06,
1159
+ "loss": 1.385,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 1.79,
1164
+ "learning_rate": 1.3433857233289714e-06,
1165
+ "loss": 1.3811,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 1.8,
1170
+ "learning_rate": 1.2280939395691859e-06,
1171
+ "loss": 1.384,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 1.81,
1176
+ "learning_rate": 1.1178494492365465e-06,
1177
+ "loss": 1.3821,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 1.82,
1182
+ "learning_rate": 1.0126756596375686e-06,
1183
+ "loss": 1.3978,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 1.83,
1188
+ "learning_rate": 9.125949014585383e-07,
1189
+ "loss": 1.3724,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 1.84,
1194
+ "learning_rate": 8.176284240242638e-07,
1195
+ "loss": 1.3851,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 1.85,
1200
+ "learning_rate": 7.277963907863478e-07,
1201
+ "loss": 1.3793,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 1.85,
1206
+ "learning_rate": 6.431178750420513e-07,
1207
+ "loss": 1.3946,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 1.86,
1212
+ "learning_rate": 5.636108558846359e-07,
1213
+ "loss": 1.3794,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 1.87,
1218
+ "learning_rate": 4.892922143859918e-07,
1219
+ "loss": 1.3886,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 1.88,
1224
+ "learning_rate": 4.201777300124249e-07,
1225
+ "loss": 1.3794,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 1.89,
1230
+ "learning_rate": 3.562820772743392e-07,
1231
+ "loss": 1.3828,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 1.9,
1236
+ "learning_rate": 2.9761882261050143e-07,
1237
+ "loss": 1.374,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 1.91,
1242
+ "learning_rate": 2.4420042150761215e-07,
1243
+ "loss": 1.3804,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 1.92,
1248
+ "learning_rate": 1.9603821585572058e-07,
1249
+ "loss": 1.3772,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 1.93,
1254
+ "learning_rate": 1.531424315400931e-07,
1255
+ "loss": 1.3913,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 1.94,
1260
+ "learning_rate": 1.1552217627004425e-07,
1261
+ "loss": 1.3859,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 1.95,
1266
+ "learning_rate": 8.318543764516961e-08,
1267
+ "loss": 1.4035,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 1.96,
1272
+ "learning_rate": 5.613908145939428e-08,
1273
+ "loss": 1.4047,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 1.97,
1278
+ "learning_rate": 3.438885024322769e-08,
1279
+ "loss": 1.3886,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 1.97,
1284
+ "learning_rate": 1.7939362044494534e-08,
1285
+ "loss": 1.3874,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 1.98,
1290
+ "learning_rate": 6.794109447824659e-09,
1291
+ "loss": 1.3852,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 1.99,
1296
+ "learning_rate": 9.554588330934478e-10,
1297
+ "loss": 1.3859,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 2.0,
1302
+ "step": 2156,
1303
+ "total_flos": 4.637442392914395e+18,
1304
+ "train_loss": 1.4111853902105498,
1305
+ "train_runtime": 51425.1791,
1306
+ "train_samples_per_second": 2.684,
1307
+ "train_steps_per_second": 0.042
1308
+ }
1309
+ ],
1310
+ "max_steps": 2156,
1311
+ "num_train_epochs": 2,
1312
+ "total_flos": 4.637442392914395e+18,
1313
+ "trial_name": null,
1314
+ "trial_params": null
1315
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:918a94b2b648985fdb03a840052eff79b0035343fd87ee2a0dc751e904fad0c4
3
+ size 4091
training_loss.png ADDED