Delete last-checkpoint
Browse files- last-checkpoint/config.json +0 -26
- last-checkpoint/configuration_stlenc.py +0 -23
- last-checkpoint/model.safetensors +0 -3
- last-checkpoint/modeling_stlenc.py +0 -35
- last-checkpoint/optimizer.pt +0 -3
- last-checkpoint/rng_state.pth +0 -3
- last-checkpoint/scaler.pt +0 -3
- last-checkpoint/scheduler.pt +0 -3
- last-checkpoint/special_tokens_map.json +0 -30
- last-checkpoint/tokenizer_config.json +0 -50
- last-checkpoint/tokenizer_stlenc.py +0 -93
- last-checkpoint/trainer_state.json +0 -1234
- last-checkpoint/training_args.bin +0 -3
- last-checkpoint/vocab.json +0 -37
last-checkpoint/config.json
DELETED
|
@@ -1,26 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"STLEncoderModel"
|
| 4 |
-
],
|
| 5 |
-
"auto_map": {
|
| 6 |
-
"AutoConfig": "configuration_stlenc.STLEncoderConfig",
|
| 7 |
-
"AutoModel": "modeling_stlenc.STLEncoderModel",
|
| 8 |
-
"AutoTokenizer": [
|
| 9 |
-
"tokenizer_stlenc.STLTokenizer",
|
| 10 |
-
null
|
| 11 |
-
]
|
| 12 |
-
},
|
| 13 |
-
"bos_token_id": 2,
|
| 14 |
-
"dtype": "float32",
|
| 15 |
-
"embedding_dim_target": 1024,
|
| 16 |
-
"eos_token_id": 3,
|
| 17 |
-
"hidden_size": 1024,
|
| 18 |
-
"intermediate_size": 4096,
|
| 19 |
-
"max_position_embeddings": 512,
|
| 20 |
-
"model_type": "stl_encoder",
|
| 21 |
-
"num_attention_heads": 16,
|
| 22 |
-
"num_hidden_layers": 12,
|
| 23 |
-
"pad_token_id": 1,
|
| 24 |
-
"transformers_version": "4.57.3",
|
| 25 |
-
"vocab_size": 35
|
| 26 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/configuration_stlenc.py
DELETED
|
@@ -1,23 +0,0 @@
|
|
| 1 |
-
from transformers import PretrainedConfig
|
| 2 |
-
|
| 3 |
-
class STLEncoderConfig(PretrainedConfig):
|
| 4 |
-
model_type = "stl_encoder"
|
| 5 |
-
def __init__(
|
| 6 |
-
self,
|
| 7 |
-
vocab_size=35,
|
| 8 |
-
hidden_size=1024,
|
| 9 |
-
num_hidden_layers=12,
|
| 10 |
-
num_attention_heads=16,
|
| 11 |
-
intermediate_size=4096,
|
| 12 |
-
max_position_embeddings=512,
|
| 13 |
-
embedding_dim_target=1024,
|
| 14 |
-
**kwargs
|
| 15 |
-
):
|
| 16 |
-
super().__init__(**kwargs)
|
| 17 |
-
self.vocab_size = vocab_size
|
| 18 |
-
self.hidden_size = hidden_size
|
| 19 |
-
self.num_hidden_layers = num_hidden_layers
|
| 20 |
-
self.num_attention_heads = num_attention_heads
|
| 21 |
-
self.intermediate_size = intermediate_size
|
| 22 |
-
self.max_position_embeddings = max_position_embeddings
|
| 23 |
-
self.embedding_dim_target = embedding_dim_target
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/model.safetensors
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:4c14360134ba900d2dec38e5d1938ef436c781041120dd2b1646f4d2872f9d93
|
| 3 |
-
size 611073224
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/modeling_stlenc.py
DELETED
|
@@ -1,35 +0,0 @@
|
|
| 1 |
-
import torch
|
| 2 |
-
import torch.nn as nn
|
| 3 |
-
from transformers import PreTrainedModel
|
| 4 |
-
from .configuration_stlenc import STLEncoderConfig
|
| 5 |
-
|
| 6 |
-
class STLEncoderModel(PreTrainedModel):
|
| 7 |
-
config_class = STLEncoderConfig
|
| 8 |
-
|
| 9 |
-
def __init__(self, config):
|
| 10 |
-
super().__init__(config)
|
| 11 |
-
self.embeddings = nn.Embedding(config.vocab_size, config.hidden_size)
|
| 12 |
-
self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size)
|
| 13 |
-
|
| 14 |
-
encoder_layer = nn.TransformerEncoderLayer(
|
| 15 |
-
d_model=config.hidden_size,
|
| 16 |
-
nhead=config.num_attention_heads,
|
| 17 |
-
dim_feedforward=config.intermediate_size,
|
| 18 |
-
batch_first=True
|
| 19 |
-
)
|
| 20 |
-
self.encoder = nn.TransformerEncoder(encoder_layer, num_layers=config.num_hidden_layers)
|
| 21 |
-
self.pooler = nn.Linear(config.hidden_size, config.embedding_dim_target)
|
| 22 |
-
self.activation = nn.Tanh()
|
| 23 |
-
self.post_init()
|
| 24 |
-
|
| 25 |
-
def forward(self, input_ids, attention_mask=None, **kwargs):
|
| 26 |
-
batch_size, seq_length = input_ids.size()
|
| 27 |
-
position_ids = torch.arange(seq_length, dtype=torch.long, device=input_ids.device)
|
| 28 |
-
position_ids = position_ids.unsqueeze(0).expand(batch_size, seq_length)
|
| 29 |
-
|
| 30 |
-
x = self.embeddings(input_ids) + self.position_embeddings(position_ids)
|
| 31 |
-
padding_mask = (attention_mask == 0) if attention_mask is not None else None
|
| 32 |
-
x = self.encoder(x, src_key_padding_mask=padding_mask)
|
| 33 |
-
|
| 34 |
-
pooled_output = self.activation(self.pooler(x[:, 0, :]))
|
| 35 |
-
return pooled_output
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/optimizer.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:99060a481ee1e9c32b48e8e0ca4159f4449a57541b54802b60aeea2b6c77354f
|
| 3 |
-
size 1222241675
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/rng_state.pth
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:9301fdb63c8f7e117dacb34dd6b2675f07003e4703c56505ff6c5837f6209a98
|
| 3 |
-
size 14645
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/scaler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:ca372268f4fa9335030c0cb7aedb6cdba75f457da50e7a4034abb1a2d0843689
|
| 3 |
-
size 1383
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/scheduler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:314048210ad2604ec1b09b90e17a875d2f267f3c96d2b6a50754bb2f69863b15
|
| 3 |
-
size 1465
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/special_tokens_map.json
DELETED
|
@@ -1,30 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"bos_token": {
|
| 3 |
-
"content": "/s",
|
| 4 |
-
"lstrip": false,
|
| 5 |
-
"normalized": false,
|
| 6 |
-
"rstrip": false,
|
| 7 |
-
"single_word": false
|
| 8 |
-
},
|
| 9 |
-
"eos_token": {
|
| 10 |
-
"content": "s",
|
| 11 |
-
"lstrip": false,
|
| 12 |
-
"normalized": false,
|
| 13 |
-
"rstrip": false,
|
| 14 |
-
"single_word": false
|
| 15 |
-
},
|
| 16 |
-
"pad_token": {
|
| 17 |
-
"content": "pad",
|
| 18 |
-
"lstrip": false,
|
| 19 |
-
"normalized": false,
|
| 20 |
-
"rstrip": false,
|
| 21 |
-
"single_word": false
|
| 22 |
-
},
|
| 23 |
-
"unk_token": {
|
| 24 |
-
"content": "unk",
|
| 25 |
-
"lstrip": false,
|
| 26 |
-
"normalized": false,
|
| 27 |
-
"rstrip": false,
|
| 28 |
-
"single_word": false
|
| 29 |
-
}
|
| 30 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/tokenizer_config.json
DELETED
|
@@ -1,50 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"added_tokens_decoder": {
|
| 3 |
-
"0": {
|
| 4 |
-
"content": "unk",
|
| 5 |
-
"lstrip": false,
|
| 6 |
-
"normalized": false,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false,
|
| 9 |
-
"special": true
|
| 10 |
-
},
|
| 11 |
-
"1": {
|
| 12 |
-
"content": "pad",
|
| 13 |
-
"lstrip": false,
|
| 14 |
-
"normalized": false,
|
| 15 |
-
"rstrip": false,
|
| 16 |
-
"single_word": false,
|
| 17 |
-
"special": true
|
| 18 |
-
},
|
| 19 |
-
"2": {
|
| 20 |
-
"content": "/s",
|
| 21 |
-
"lstrip": false,
|
| 22 |
-
"normalized": false,
|
| 23 |
-
"rstrip": false,
|
| 24 |
-
"single_word": false,
|
| 25 |
-
"special": true
|
| 26 |
-
},
|
| 27 |
-
"3": {
|
| 28 |
-
"content": "s",
|
| 29 |
-
"lstrip": false,
|
| 30 |
-
"normalized": false,
|
| 31 |
-
"rstrip": false,
|
| 32 |
-
"single_word": false,
|
| 33 |
-
"special": true
|
| 34 |
-
}
|
| 35 |
-
},
|
| 36 |
-
"auto_map": {
|
| 37 |
-
"AutoTokenizer": [
|
| 38 |
-
"tokenizer_stlenc.STLTokenizer",
|
| 39 |
-
null
|
| 40 |
-
]
|
| 41 |
-
},
|
| 42 |
-
"bos_token": "/s",
|
| 43 |
-
"clean_up_tokenization_spaces": false,
|
| 44 |
-
"eos_token": "s",
|
| 45 |
-
"extra_special_tokens": {},
|
| 46 |
-
"model_max_length": 512,
|
| 47 |
-
"pad_token": "pad",
|
| 48 |
-
"tokenizer_class": "STLTokenizer",
|
| 49 |
-
"unk_token": "unk"
|
| 50 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/tokenizer_stlenc.py
DELETED
|
@@ -1,93 +0,0 @@
|
|
| 1 |
-
import json
|
| 2 |
-
import os
|
| 3 |
-
import torch
|
| 4 |
-
from typing import Any, Dict, List, Optional, Tuple, Union
|
| 5 |
-
from transformers import PreTrainedTokenizer, AutoTokenizer
|
| 6 |
-
|
| 7 |
-
class STLTokenizer(PreTrainedTokenizer):
|
| 8 |
-
model_type = "stl_encoder"
|
| 9 |
-
|
| 10 |
-
def __init__(
|
| 11 |
-
self,
|
| 12 |
-
vocab_file="vocab.json",
|
| 13 |
-
unk_token="unk",
|
| 14 |
-
pad_token="pad",
|
| 15 |
-
bos_token="/s",
|
| 16 |
-
eos_token="s",
|
| 17 |
-
model_max_length=512,
|
| 18 |
-
**kwargs
|
| 19 |
-
):
|
| 20 |
-
current_dir = os.path.dirname(__file__)
|
| 21 |
-
full_vocab_path = os.path.join(current_dir, vocab_file)
|
| 22 |
-
|
| 23 |
-
if not os.path.exists(full_vocab_path):
|
| 24 |
-
from huggingface_hub import hf_hub_download
|
| 25 |
-
try:
|
| 26 |
-
full_vocab_path = hf_hub_download("saracandu/stlenc", vocab_file)
|
| 27 |
-
except:
|
| 28 |
-
full_vocab_path = vocab_file
|
| 29 |
-
|
| 30 |
-
with open(full_vocab_path, "r", encoding="utf-8") as f:
|
| 31 |
-
self.vocab = json.load(f)
|
| 32 |
-
|
| 33 |
-
self.id_to_token = {v: k for k, v in self.vocab.items()}
|
| 34 |
-
|
| 35 |
-
super().__init__(
|
| 36 |
-
unk_token=unk_token,
|
| 37 |
-
pad_token=pad_token,
|
| 38 |
-
bos_token=bos_token,
|
| 39 |
-
eos_token=eos_token,
|
| 40 |
-
model_max_length=model_max_length,
|
| 41 |
-
**kwargs
|
| 42 |
-
)
|
| 43 |
-
|
| 44 |
-
@property
|
| 45 |
-
def vocab_size(self) -> int:
|
| 46 |
-
return len(self.vocab)
|
| 47 |
-
|
| 48 |
-
def get_vocab(self) -> Dict[str, int]:
|
| 49 |
-
return dict(self.vocab)
|
| 50 |
-
|
| 51 |
-
def _tokenize(self, text: str) -> List[str]:
|
| 52 |
-
text = f'{self.bos_token} {text} {self.eos_token}'.replace(' ', '@')
|
| 53 |
-
|
| 54 |
-
tokens = []
|
| 55 |
-
i = 0
|
| 56 |
-
while i < len(text):
|
| 57 |
-
best_match = None
|
| 58 |
-
for j in range(min(i + 50, len(text)), i, -1):
|
| 59 |
-
subtoken = text[i:j]
|
| 60 |
-
if subtoken in self.vocab:
|
| 61 |
-
best_match = subtoken
|
| 62 |
-
break
|
| 63 |
-
|
| 64 |
-
if best_match:
|
| 65 |
-
tokens.append(best_match)
|
| 66 |
-
i += len(best_match)
|
| 67 |
-
else:
|
| 68 |
-
tokens.append(self.unk_token)
|
| 69 |
-
i += 1
|
| 70 |
-
return tokens
|
| 71 |
-
|
| 72 |
-
def _convert_token_to_id(self, token: str) -> int:
|
| 73 |
-
return self.vocab.get(token, self.vocab.get(self.unk_token))
|
| 74 |
-
|
| 75 |
-
def _convert_id_to_token(self, index: int) -> str:
|
| 76 |
-
return self.id_to_token.get(index, self.unk_token)
|
| 77 |
-
|
| 78 |
-
def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
|
| 79 |
-
if not os.path.isdir(save_directory):
|
| 80 |
-
os.makedirs(save_directory)
|
| 81 |
-
|
| 82 |
-
prefix = filename_prefix if filename_prefix is not None else ""
|
| 83 |
-
vocab_file = os.path.join(save_directory, prefix + "vocab.json")
|
| 84 |
-
|
| 85 |
-
with open(vocab_file, "w", encoding="utf-8") as f:
|
| 86 |
-
json.dump(self.vocab, f, indent=2, ensure_ascii=False)
|
| 87 |
-
|
| 88 |
-
return (vocab_file,)
|
| 89 |
-
|
| 90 |
-
try:
|
| 91 |
-
AutoTokenizer.register("stl_encoder", STLTokenizer)
|
| 92 |
-
except Exception:
|
| 93 |
-
pass
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/trainer_state.json
DELETED
|
@@ -1,1234 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": 1500,
|
| 3 |
-
"best_metric": 0.020201340437836564,
|
| 4 |
-
"best_model_checkpoint": "./stlenc-training/checkpoint-1500",
|
| 5 |
-
"epoch": 0.3048780487804878,
|
| 6 |
-
"eval_steps": 100,
|
| 7 |
-
"global_step": 1500,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0020325203252032522,
|
| 14 |
-
"grad_norm": 0.3073508143424988,
|
| 15 |
-
"learning_rate": 4.998170731707317e-05,
|
| 16 |
-
"loss": 0.1335,
|
| 17 |
-
"step": 10
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.0040650406504065045,
|
| 21 |
-
"grad_norm": 0.20867981016635895,
|
| 22 |
-
"learning_rate": 4.996138211382114e-05,
|
| 23 |
-
"loss": 0.0568,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.006097560975609756,
|
| 28 |
-
"grad_norm": 0.1618652641773224,
|
| 29 |
-
"learning_rate": 4.994105691056911e-05,
|
| 30 |
-
"loss": 0.0505,
|
| 31 |
-
"step": 30
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008130081300813009,
|
| 35 |
-
"grad_norm": 0.14075744152069092,
|
| 36 |
-
"learning_rate": 4.9920731707317074e-05,
|
| 37 |
-
"loss": 0.0459,
|
| 38 |
-
"step": 40
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.01016260162601626,
|
| 42 |
-
"grad_norm": 0.15348248183727264,
|
| 43 |
-
"learning_rate": 4.990040650406504e-05,
|
| 44 |
-
"loss": 0.0441,
|
| 45 |
-
"step": 50
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.012195121951219513,
|
| 49 |
-
"grad_norm": 0.1347082257270813,
|
| 50 |
-
"learning_rate": 4.988008130081301e-05,
|
| 51 |
-
"loss": 0.0426,
|
| 52 |
-
"step": 60
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.014227642276422764,
|
| 56 |
-
"grad_norm": 0.1535383015871048,
|
| 57 |
-
"learning_rate": 4.9859756097560977e-05,
|
| 58 |
-
"loss": 0.0428,
|
| 59 |
-
"step": 70
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.016260162601626018,
|
| 63 |
-
"grad_norm": 0.14122287929058075,
|
| 64 |
-
"learning_rate": 4.9839430894308944e-05,
|
| 65 |
-
"loss": 0.0395,
|
| 66 |
-
"step": 80
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.018292682926829267,
|
| 70 |
-
"grad_norm": 0.16186140477657318,
|
| 71 |
-
"learning_rate": 4.981910569105691e-05,
|
| 72 |
-
"loss": 0.0397,
|
| 73 |
-
"step": 90
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.02032520325203252,
|
| 77 |
-
"grad_norm": 0.14353497326374054,
|
| 78 |
-
"learning_rate": 4.979878048780488e-05,
|
| 79 |
-
"loss": 0.0387,
|
| 80 |
-
"step": 100
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.02032520325203252,
|
| 84 |
-
"eval_cosine_similarity": 0.43845831272477015,
|
| 85 |
-
"eval_loss": 0.023171832785010338,
|
| 86 |
-
"eval_mse_sync": 0.023171832367424337,
|
| 87 |
-
"eval_runtime": 7.8562,
|
| 88 |
-
"eval_samples_per_second": 380.971,
|
| 89 |
-
"eval_steps_per_second": 23.93,
|
| 90 |
-
"step": 100
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"epoch": 0.022357723577235773,
|
| 94 |
-
"grad_norm": 0.16455022990703583,
|
| 95 |
-
"learning_rate": 4.977845528455285e-05,
|
| 96 |
-
"loss": 0.039,
|
| 97 |
-
"step": 110
|
| 98 |
-
},
|
| 99 |
-
{
|
| 100 |
-
"epoch": 0.024390243902439025,
|
| 101 |
-
"grad_norm": 0.14409448206424713,
|
| 102 |
-
"learning_rate": 4.9758130081300813e-05,
|
| 103 |
-
"loss": 0.0377,
|
| 104 |
-
"step": 120
|
| 105 |
-
},
|
| 106 |
-
{
|
| 107 |
-
"epoch": 0.026422764227642278,
|
| 108 |
-
"grad_norm": 0.14973576366901398,
|
| 109 |
-
"learning_rate": 4.973780487804878e-05,
|
| 110 |
-
"loss": 0.0367,
|
| 111 |
-
"step": 130
|
| 112 |
-
},
|
| 113 |
-
{
|
| 114 |
-
"epoch": 0.028455284552845527,
|
| 115 |
-
"grad_norm": 0.12309937924146652,
|
| 116 |
-
"learning_rate": 4.9717479674796755e-05,
|
| 117 |
-
"loss": 0.0358,
|
| 118 |
-
"step": 140
|
| 119 |
-
},
|
| 120 |
-
{
|
| 121 |
-
"epoch": 0.03048780487804878,
|
| 122 |
-
"grad_norm": 0.12204054743051529,
|
| 123 |
-
"learning_rate": 4.9697154471544716e-05,
|
| 124 |
-
"loss": 0.0339,
|
| 125 |
-
"step": 150
|
| 126 |
-
},
|
| 127 |
-
{
|
| 128 |
-
"epoch": 0.032520325203252036,
|
| 129 |
-
"grad_norm": 0.16795864701271057,
|
| 130 |
-
"learning_rate": 4.967682926829268e-05,
|
| 131 |
-
"loss": 0.0359,
|
| 132 |
-
"step": 160
|
| 133 |
-
},
|
| 134 |
-
{
|
| 135 |
-
"epoch": 0.034552845528455285,
|
| 136 |
-
"grad_norm": 0.14515043795108795,
|
| 137 |
-
"learning_rate": 4.965650406504066e-05,
|
| 138 |
-
"loss": 0.034,
|
| 139 |
-
"step": 170
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"epoch": 0.036585365853658534,
|
| 143 |
-
"grad_norm": 0.14003875851631165,
|
| 144 |
-
"learning_rate": 4.963617886178862e-05,
|
| 145 |
-
"loss": 0.0351,
|
| 146 |
-
"step": 180
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.03861788617886179,
|
| 150 |
-
"grad_norm": 0.1429203450679779,
|
| 151 |
-
"learning_rate": 4.9615853658536585e-05,
|
| 152 |
-
"loss": 0.035,
|
| 153 |
-
"step": 190
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.04065040650406504,
|
| 157 |
-
"grad_norm": 0.13865381479263306,
|
| 158 |
-
"learning_rate": 4.959552845528456e-05,
|
| 159 |
-
"loss": 0.0337,
|
| 160 |
-
"step": 200
|
| 161 |
-
},
|
| 162 |
-
{
|
| 163 |
-
"epoch": 0.04065040650406504,
|
| 164 |
-
"eval_cosine_similarity": 0.44644545833681326,
|
| 165 |
-
"eval_loss": 0.022488251328468323,
|
| 166 |
-
"eval_mse_sync": 0.02248825095382336,
|
| 167 |
-
"eval_runtime": 7.8597,
|
| 168 |
-
"eval_samples_per_second": 380.801,
|
| 169 |
-
"eval_steps_per_second": 23.919,
|
| 170 |
-
"step": 200
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.042682926829268296,
|
| 174 |
-
"grad_norm": 0.18544256687164307,
|
| 175 |
-
"learning_rate": 4.957520325203252e-05,
|
| 176 |
-
"loss": 0.0345,
|
| 177 |
-
"step": 210
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.044715447154471545,
|
| 181 |
-
"grad_norm": 0.1306702345609665,
|
| 182 |
-
"learning_rate": 4.955487804878049e-05,
|
| 183 |
-
"loss": 0.0351,
|
| 184 |
-
"step": 220
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.046747967479674794,
|
| 188 |
-
"grad_norm": 0.13073720037937164,
|
| 189 |
-
"learning_rate": 4.953455284552846e-05,
|
| 190 |
-
"loss": 0.0334,
|
| 191 |
-
"step": 230
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.04878048780487805,
|
| 195 |
-
"grad_norm": 0.13994655013084412,
|
| 196 |
-
"learning_rate": 4.951422764227642e-05,
|
| 197 |
-
"loss": 0.0329,
|
| 198 |
-
"step": 240
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.0508130081300813,
|
| 202 |
-
"grad_norm": 0.13009095191955566,
|
| 203 |
-
"learning_rate": 4.949390243902439e-05,
|
| 204 |
-
"loss": 0.0336,
|
| 205 |
-
"step": 250
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.052845528455284556,
|
| 209 |
-
"grad_norm": 0.12570971250534058,
|
| 210 |
-
"learning_rate": 4.9473577235772363e-05,
|
| 211 |
-
"loss": 0.033,
|
| 212 |
-
"step": 260
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.054878048780487805,
|
| 216 |
-
"grad_norm": 0.12797130644321442,
|
| 217 |
-
"learning_rate": 4.9453252032520324e-05,
|
| 218 |
-
"loss": 0.0323,
|
| 219 |
-
"step": 270
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.056910569105691054,
|
| 223 |
-
"grad_norm": 0.1270533800125122,
|
| 224 |
-
"learning_rate": 4.943292682926829e-05,
|
| 225 |
-
"loss": 0.0324,
|
| 226 |
-
"step": 280
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.05894308943089431,
|
| 230 |
-
"grad_norm": 0.13283374905586243,
|
| 231 |
-
"learning_rate": 4.9412601626016266e-05,
|
| 232 |
-
"loss": 0.0328,
|
| 233 |
-
"step": 290
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.06097560975609756,
|
| 237 |
-
"grad_norm": 0.14629143476486206,
|
| 238 |
-
"learning_rate": 4.9392276422764226e-05,
|
| 239 |
-
"loss": 0.0311,
|
| 240 |
-
"step": 300
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.06097560975609756,
|
| 244 |
-
"eval_cosine_similarity": 0.4329459800733757,
|
| 245 |
-
"eval_loss": 0.02308308333158493,
|
| 246 |
-
"eval_mse_sync": 0.02308308467053616,
|
| 247 |
-
"eval_runtime": 7.902,
|
| 248 |
-
"eval_samples_per_second": 378.765,
|
| 249 |
-
"eval_steps_per_second": 23.791,
|
| 250 |
-
"step": 300
|
| 251 |
-
},
|
| 252 |
-
{
|
| 253 |
-
"epoch": 0.06300813008130081,
|
| 254 |
-
"grad_norm": 0.15855424106121063,
|
| 255 |
-
"learning_rate": 4.93719512195122e-05,
|
| 256 |
-
"loss": 0.0324,
|
| 257 |
-
"step": 310
|
| 258 |
-
},
|
| 259 |
-
{
|
| 260 |
-
"epoch": 0.06504065040650407,
|
| 261 |
-
"grad_norm": 0.1387602835893631,
|
| 262 |
-
"learning_rate": 4.935162601626017e-05,
|
| 263 |
-
"loss": 0.0312,
|
| 264 |
-
"step": 320
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"epoch": 0.06707317073170732,
|
| 268 |
-
"grad_norm": 0.14994463324546814,
|
| 269 |
-
"learning_rate": 4.933130081300813e-05,
|
| 270 |
-
"loss": 0.0342,
|
| 271 |
-
"step": 330
|
| 272 |
-
},
|
| 273 |
-
{
|
| 274 |
-
"epoch": 0.06910569105691057,
|
| 275 |
-
"grad_norm": 0.1439775824546814,
|
| 276 |
-
"learning_rate": 4.93109756097561e-05,
|
| 277 |
-
"loss": 0.032,
|
| 278 |
-
"step": 340
|
| 279 |
-
},
|
| 280 |
-
{
|
| 281 |
-
"epoch": 0.07113821138211382,
|
| 282 |
-
"grad_norm": 0.11875730007886887,
|
| 283 |
-
"learning_rate": 4.929065040650407e-05,
|
| 284 |
-
"loss": 0.0325,
|
| 285 |
-
"step": 350
|
| 286 |
-
},
|
| 287 |
-
{
|
| 288 |
-
"epoch": 0.07317073170731707,
|
| 289 |
-
"grad_norm": 0.12371476739645004,
|
| 290 |
-
"learning_rate": 4.927032520325203e-05,
|
| 291 |
-
"loss": 0.0319,
|
| 292 |
-
"step": 360
|
| 293 |
-
},
|
| 294 |
-
{
|
| 295 |
-
"epoch": 0.07520325203252033,
|
| 296 |
-
"grad_norm": 0.12192820757627487,
|
| 297 |
-
"learning_rate": 4.9250000000000004e-05,
|
| 298 |
-
"loss": 0.0312,
|
| 299 |
-
"step": 370
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 0.07723577235772358,
|
| 303 |
-
"grad_norm": 0.14944523572921753,
|
| 304 |
-
"learning_rate": 4.922967479674797e-05,
|
| 305 |
-
"loss": 0.0325,
|
| 306 |
-
"step": 380
|
| 307 |
-
},
|
| 308 |
-
{
|
| 309 |
-
"epoch": 0.07926829268292683,
|
| 310 |
-
"grad_norm": 0.13014467060565948,
|
| 311 |
-
"learning_rate": 4.920934959349593e-05,
|
| 312 |
-
"loss": 0.0296,
|
| 313 |
-
"step": 390
|
| 314 |
-
},
|
| 315 |
-
{
|
| 316 |
-
"epoch": 0.08130081300813008,
|
| 317 |
-
"grad_norm": 0.12468204647302628,
|
| 318 |
-
"learning_rate": 4.9189024390243907e-05,
|
| 319 |
-
"loss": 0.0336,
|
| 320 |
-
"step": 400
|
| 321 |
-
},
|
| 322 |
-
{
|
| 323 |
-
"epoch": 0.08130081300813008,
|
| 324 |
-
"eval_cosine_similarity": 0.45511097012655893,
|
| 325 |
-
"eval_loss": 0.023316312581300735,
|
| 326 |
-
"eval_mse_sync": 0.02331631100243981,
|
| 327 |
-
"eval_runtime": 7.896,
|
| 328 |
-
"eval_samples_per_second": 379.053,
|
| 329 |
-
"eval_steps_per_second": 23.81,
|
| 330 |
-
"step": 400
|
| 331 |
-
},
|
| 332 |
-
{
|
| 333 |
-
"epoch": 0.08333333333333333,
|
| 334 |
-
"grad_norm": 0.12918700277805328,
|
| 335 |
-
"learning_rate": 4.9168699186991874e-05,
|
| 336 |
-
"loss": 0.0307,
|
| 337 |
-
"step": 410
|
| 338 |
-
},
|
| 339 |
-
{
|
| 340 |
-
"epoch": 0.08536585365853659,
|
| 341 |
-
"grad_norm": 0.16114428639411926,
|
| 342 |
-
"learning_rate": 4.9148373983739835e-05,
|
| 343 |
-
"loss": 0.0302,
|
| 344 |
-
"step": 420
|
| 345 |
-
},
|
| 346 |
-
{
|
| 347 |
-
"epoch": 0.08739837398373984,
|
| 348 |
-
"grad_norm": 0.13520212471485138,
|
| 349 |
-
"learning_rate": 4.912804878048781e-05,
|
| 350 |
-
"loss": 0.0311,
|
| 351 |
-
"step": 430
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 0.08943089430894309,
|
| 355 |
-
"grad_norm": 0.12596647441387177,
|
| 356 |
-
"learning_rate": 4.9107723577235776e-05,
|
| 357 |
-
"loss": 0.0297,
|
| 358 |
-
"step": 440
|
| 359 |
-
},
|
| 360 |
-
{
|
| 361 |
-
"epoch": 0.09146341463414634,
|
| 362 |
-
"grad_norm": 0.10777679830789566,
|
| 363 |
-
"learning_rate": 4.908739837398374e-05,
|
| 364 |
-
"loss": 0.0304,
|
| 365 |
-
"step": 450
|
| 366 |
-
},
|
| 367 |
-
{
|
| 368 |
-
"epoch": 0.09349593495934959,
|
| 369 |
-
"grad_norm": 0.127786323428154,
|
| 370 |
-
"learning_rate": 4.906707317073171e-05,
|
| 371 |
-
"loss": 0.0308,
|
| 372 |
-
"step": 460
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"epoch": 0.09552845528455285,
|
| 376 |
-
"grad_norm": 0.165998637676239,
|
| 377 |
-
"learning_rate": 4.904674796747968e-05,
|
| 378 |
-
"loss": 0.0294,
|
| 379 |
-
"step": 470
|
| 380 |
-
},
|
| 381 |
-
{
|
| 382 |
-
"epoch": 0.0975609756097561,
|
| 383 |
-
"grad_norm": 0.1122402474284172,
|
| 384 |
-
"learning_rate": 4.902642276422764e-05,
|
| 385 |
-
"loss": 0.0305,
|
| 386 |
-
"step": 480
|
| 387 |
-
},
|
| 388 |
-
{
|
| 389 |
-
"epoch": 0.09959349593495935,
|
| 390 |
-
"grad_norm": 0.1091734766960144,
|
| 391 |
-
"learning_rate": 4.900609756097561e-05,
|
| 392 |
-
"loss": 0.0323,
|
| 393 |
-
"step": 490
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 0.1016260162601626,
|
| 397 |
-
"grad_norm": 0.12982085347175598,
|
| 398 |
-
"learning_rate": 4.898577235772358e-05,
|
| 399 |
-
"loss": 0.0297,
|
| 400 |
-
"step": 500
|
| 401 |
-
},
|
| 402 |
-
{
|
| 403 |
-
"epoch": 0.1016260162601626,
|
| 404 |
-
"eval_cosine_similarity": 0.44197734325098587,
|
| 405 |
-
"eval_loss": 0.022700216621160507,
|
| 406 |
-
"eval_mse_sync": 0.02270021719246296,
|
| 407 |
-
"eval_runtime": 7.917,
|
| 408 |
-
"eval_samples_per_second": 378.046,
|
| 409 |
-
"eval_steps_per_second": 23.746,
|
| 410 |
-
"step": 500
|
| 411 |
-
},
|
| 412 |
-
{
|
| 413 |
-
"epoch": 0.10365853658536585,
|
| 414 |
-
"grad_norm": 0.14651747047901154,
|
| 415 |
-
"learning_rate": 4.896544715447154e-05,
|
| 416 |
-
"loss": 0.0291,
|
| 417 |
-
"step": 510
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 0.10569105691056911,
|
| 421 |
-
"grad_norm": 0.13730144500732422,
|
| 422 |
-
"learning_rate": 4.8945121951219515e-05,
|
| 423 |
-
"loss": 0.0313,
|
| 424 |
-
"step": 520
|
| 425 |
-
},
|
| 426 |
-
{
|
| 427 |
-
"epoch": 0.10772357723577236,
|
| 428 |
-
"grad_norm": 0.1328081637620926,
|
| 429 |
-
"learning_rate": 4.892479674796748e-05,
|
| 430 |
-
"loss": 0.0305,
|
| 431 |
-
"step": 530
|
| 432 |
-
},
|
| 433 |
-
{
|
| 434 |
-
"epoch": 0.10975609756097561,
|
| 435 |
-
"grad_norm": 0.15137328207492828,
|
| 436 |
-
"learning_rate": 4.890447154471545e-05,
|
| 437 |
-
"loss": 0.0299,
|
| 438 |
-
"step": 540
|
| 439 |
-
},
|
| 440 |
-
{
|
| 441 |
-
"epoch": 0.11178861788617886,
|
| 442 |
-
"grad_norm": 0.15069301426410675,
|
| 443 |
-
"learning_rate": 4.888414634146342e-05,
|
| 444 |
-
"loss": 0.0302,
|
| 445 |
-
"step": 550
|
| 446 |
-
},
|
| 447 |
-
{
|
| 448 |
-
"epoch": 0.11382113821138211,
|
| 449 |
-
"grad_norm": 0.13555215299129486,
|
| 450 |
-
"learning_rate": 4.8863821138211385e-05,
|
| 451 |
-
"loss": 0.031,
|
| 452 |
-
"step": 560
|
| 453 |
-
},
|
| 454 |
-
{
|
| 455 |
-
"epoch": 0.11585365853658537,
|
| 456 |
-
"grad_norm": 0.11980469524860382,
|
| 457 |
-
"learning_rate": 4.884349593495935e-05,
|
| 458 |
-
"loss": 0.0302,
|
| 459 |
-
"step": 570
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 0.11788617886178862,
|
| 463 |
-
"grad_norm": 0.11329913884401321,
|
| 464 |
-
"learning_rate": 4.882317073170732e-05,
|
| 465 |
-
"loss": 0.0302,
|
| 466 |
-
"step": 580
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 0.11991869918699187,
|
| 470 |
-
"grad_norm": 0.11942901462316513,
|
| 471 |
-
"learning_rate": 4.880284552845529e-05,
|
| 472 |
-
"loss": 0.0283,
|
| 473 |
-
"step": 590
|
| 474 |
-
},
|
| 475 |
-
{
|
| 476 |
-
"epoch": 0.12195121951219512,
|
| 477 |
-
"grad_norm": 0.17668181657791138,
|
| 478 |
-
"learning_rate": 4.8782520325203254e-05,
|
| 479 |
-
"loss": 0.0298,
|
| 480 |
-
"step": 600
|
| 481 |
-
},
|
| 482 |
-
{
|
| 483 |
-
"epoch": 0.12195121951219512,
|
| 484 |
-
"eval_cosine_similarity": 0.45993702271452247,
|
| 485 |
-
"eval_loss": 0.022243835031986237,
|
| 486 |
-
"eval_mse_sync": 0.02224383312391072,
|
| 487 |
-
"eval_runtime": 7.95,
|
| 488 |
-
"eval_samples_per_second": 376.478,
|
| 489 |
-
"eval_steps_per_second": 23.648,
|
| 490 |
-
"step": 600
|
| 491 |
-
},
|
| 492 |
-
{
|
| 493 |
-
"epoch": 0.12398373983739837,
|
| 494 |
-
"grad_norm": 0.14265325665473938,
|
| 495 |
-
"learning_rate": 4.876219512195122e-05,
|
| 496 |
-
"loss": 0.0298,
|
| 497 |
-
"step": 610
|
| 498 |
-
},
|
| 499 |
-
{
|
| 500 |
-
"epoch": 0.12601626016260162,
|
| 501 |
-
"grad_norm": 0.1118614450097084,
|
| 502 |
-
"learning_rate": 4.874186991869919e-05,
|
| 503 |
-
"loss": 0.0303,
|
| 504 |
-
"step": 620
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 0.12804878048780488,
|
| 508 |
-
"grad_norm": 0.11675341427326202,
|
| 509 |
-
"learning_rate": 4.8721544715447156e-05,
|
| 510 |
-
"loss": 0.0304,
|
| 511 |
-
"step": 630
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 0.13008130081300814,
|
| 515 |
-
"grad_norm": 0.12549124658107758,
|
| 516 |
-
"learning_rate": 4.8701219512195124e-05,
|
| 517 |
-
"loss": 0.0297,
|
| 518 |
-
"step": 640
|
| 519 |
-
},
|
| 520 |
-
{
|
| 521 |
-
"epoch": 0.13211382113821138,
|
| 522 |
-
"grad_norm": 0.13423743844032288,
|
| 523 |
-
"learning_rate": 4.868089430894309e-05,
|
| 524 |
-
"loss": 0.0289,
|
| 525 |
-
"step": 650
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 0.13414634146341464,
|
| 529 |
-
"grad_norm": 0.1256653517484665,
|
| 530 |
-
"learning_rate": 4.8660569105691065e-05,
|
| 531 |
-
"loss": 0.0291,
|
| 532 |
-
"step": 660
|
| 533 |
-
},
|
| 534 |
-
{
|
| 535 |
-
"epoch": 0.13617886178861788,
|
| 536 |
-
"grad_norm": 0.1281777024269104,
|
| 537 |
-
"learning_rate": 4.8640243902439026e-05,
|
| 538 |
-
"loss": 0.0302,
|
| 539 |
-
"step": 670
|
| 540 |
-
},
|
| 541 |
-
{
|
| 542 |
-
"epoch": 0.13821138211382114,
|
| 543 |
-
"grad_norm": 0.1304328590631485,
|
| 544 |
-
"learning_rate": 4.861991869918699e-05,
|
| 545 |
-
"loss": 0.0282,
|
| 546 |
-
"step": 680
|
| 547 |
-
},
|
| 548 |
-
{
|
| 549 |
-
"epoch": 0.1402439024390244,
|
| 550 |
-
"grad_norm": 0.1297086775302887,
|
| 551 |
-
"learning_rate": 4.859959349593497e-05,
|
| 552 |
-
"loss": 0.0299,
|
| 553 |
-
"step": 690
|
| 554 |
-
},
|
| 555 |
-
{
|
| 556 |
-
"epoch": 0.14227642276422764,
|
| 557 |
-
"grad_norm": 0.13382111489772797,
|
| 558 |
-
"learning_rate": 4.857926829268293e-05,
|
| 559 |
-
"loss": 0.0303,
|
| 560 |
-
"step": 700
|
| 561 |
-
},
|
| 562 |
-
{
|
| 563 |
-
"epoch": 0.14227642276422764,
|
| 564 |
-
"eval_cosine_similarity": 0.4609147342917037,
|
| 565 |
-
"eval_loss": 0.02211085520684719,
|
| 566 |
-
"eval_mse_sync": 0.022110854542816995,
|
| 567 |
-
"eval_runtime": 7.95,
|
| 568 |
-
"eval_samples_per_second": 376.479,
|
| 569 |
-
"eval_steps_per_second": 23.648,
|
| 570 |
-
"step": 700
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.1443089430894309,
|
| 574 |
-
"grad_norm": 0.14310456812381744,
|
| 575 |
-
"learning_rate": 4.8558943089430895e-05,
|
| 576 |
-
"loss": 0.0282,
|
| 577 |
-
"step": 710
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.14634146341463414,
|
| 581 |
-
"grad_norm": 0.13597296178340912,
|
| 582 |
-
"learning_rate": 4.853861788617887e-05,
|
| 583 |
-
"loss": 0.0282,
|
| 584 |
-
"step": 720
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.1483739837398374,
|
| 588 |
-
"grad_norm": 0.21410208940505981,
|
| 589 |
-
"learning_rate": 4.851829268292683e-05,
|
| 590 |
-
"loss": 0.03,
|
| 591 |
-
"step": 730
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.15040650406504066,
|
| 595 |
-
"grad_norm": 0.14297862350940704,
|
| 596 |
-
"learning_rate": 4.84979674796748e-05,
|
| 597 |
-
"loss": 0.0288,
|
| 598 |
-
"step": 740
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.1524390243902439,
|
| 602 |
-
"grad_norm": 0.13763341307640076,
|
| 603 |
-
"learning_rate": 4.847764227642277e-05,
|
| 604 |
-
"loss": 0.0293,
|
| 605 |
-
"step": 750
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.15447154471544716,
|
| 609 |
-
"grad_norm": 0.14061112701892853,
|
| 610 |
-
"learning_rate": 4.845731707317073e-05,
|
| 611 |
-
"loss": 0.0294,
|
| 612 |
-
"step": 760
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.1565040650406504,
|
| 616 |
-
"grad_norm": 0.1257963925600052,
|
| 617 |
-
"learning_rate": 4.84369918699187e-05,
|
| 618 |
-
"loss": 0.0292,
|
| 619 |
-
"step": 770
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 0.15853658536585366,
|
| 623 |
-
"grad_norm": 0.10741665959358215,
|
| 624 |
-
"learning_rate": 4.8416666666666673e-05,
|
| 625 |
-
"loss": 0.0284,
|
| 626 |
-
"step": 780
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 0.16056910569105692,
|
| 630 |
-
"grad_norm": 0.15043672919273376,
|
| 631 |
-
"learning_rate": 4.8396341463414634e-05,
|
| 632 |
-
"loss": 0.029,
|
| 633 |
-
"step": 790
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.16260162601626016,
|
| 637 |
-
"grad_norm": 0.15317371487617493,
|
| 638 |
-
"learning_rate": 4.83760162601626e-05,
|
| 639 |
-
"loss": 0.0298,
|
| 640 |
-
"step": 800
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.16260162601626016,
|
| 644 |
-
"eval_cosine_similarity": 0.4601094912733874,
|
| 645 |
-
"eval_loss": 0.022229857742786407,
|
| 646 |
-
"eval_mse_sync": 0.022229857477672196,
|
| 647 |
-
"eval_runtime": 7.9894,
|
| 648 |
-
"eval_samples_per_second": 374.619,
|
| 649 |
-
"eval_steps_per_second": 23.531,
|
| 650 |
-
"step": 800
|
| 651 |
-
},
|
| 652 |
-
{
|
| 653 |
-
"epoch": 0.16463414634146342,
|
| 654 |
-
"grad_norm": 0.11747121810913086,
|
| 655 |
-
"learning_rate": 4.8355691056910576e-05,
|
| 656 |
-
"loss": 0.029,
|
| 657 |
-
"step": 810
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"epoch": 0.16666666666666666,
|
| 661 |
-
"grad_norm": 0.12080563604831696,
|
| 662 |
-
"learning_rate": 4.8335365853658536e-05,
|
| 663 |
-
"loss": 0.029,
|
| 664 |
-
"step": 820
|
| 665 |
-
},
|
| 666 |
-
{
|
| 667 |
-
"epoch": 0.16869918699186992,
|
| 668 |
-
"grad_norm": 0.11985825002193451,
|
| 669 |
-
"learning_rate": 4.8315040650406504e-05,
|
| 670 |
-
"loss": 0.0289,
|
| 671 |
-
"step": 830
|
| 672 |
-
},
|
| 673 |
-
{
|
| 674 |
-
"epoch": 0.17073170731707318,
|
| 675 |
-
"grad_norm": 0.139726921916008,
|
| 676 |
-
"learning_rate": 4.829471544715448e-05,
|
| 677 |
-
"loss": 0.0289,
|
| 678 |
-
"step": 840
|
| 679 |
-
},
|
| 680 |
-
{
|
| 681 |
-
"epoch": 0.17276422764227642,
|
| 682 |
-
"grad_norm": 0.12052454799413681,
|
| 683 |
-
"learning_rate": 4.827439024390244e-05,
|
| 684 |
-
"loss": 0.0278,
|
| 685 |
-
"step": 850
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 0.17479674796747968,
|
| 689 |
-
"grad_norm": 0.1230531707406044,
|
| 690 |
-
"learning_rate": 4.825406504065041e-05,
|
| 691 |
-
"loss": 0.0282,
|
| 692 |
-
"step": 860
|
| 693 |
-
},
|
| 694 |
-
{
|
| 695 |
-
"epoch": 0.17682926829268292,
|
| 696 |
-
"grad_norm": 0.12765666842460632,
|
| 697 |
-
"learning_rate": 4.823373983739838e-05,
|
| 698 |
-
"loss": 0.0277,
|
| 699 |
-
"step": 870
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 0.17886178861788618,
|
| 703 |
-
"grad_norm": 0.11474256962537766,
|
| 704 |
-
"learning_rate": 4.821341463414634e-05,
|
| 705 |
-
"loss": 0.0259,
|
| 706 |
-
"step": 880
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 0.18089430894308944,
|
| 710 |
-
"grad_norm": 0.12510469555854797,
|
| 711 |
-
"learning_rate": 4.8193089430894315e-05,
|
| 712 |
-
"loss": 0.0295,
|
| 713 |
-
"step": 890
|
| 714 |
-
},
|
| 715 |
-
{
|
| 716 |
-
"epoch": 0.18292682926829268,
|
| 717 |
-
"grad_norm": 0.13411709666252136,
|
| 718 |
-
"learning_rate": 4.817276422764228e-05,
|
| 719 |
-
"loss": 0.0286,
|
| 720 |
-
"step": 900
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 0.18292682926829268,
|
| 724 |
-
"eval_cosine_similarity": 0.47567485046354857,
|
| 725 |
-
"eval_loss": 0.021220851689577103,
|
| 726 |
-
"eval_mse_sync": 0.021220852660106707,
|
| 727 |
-
"eval_runtime": 7.9662,
|
| 728 |
-
"eval_samples_per_second": 375.714,
|
| 729 |
-
"eval_steps_per_second": 23.6,
|
| 730 |
-
"step": 900
|
| 731 |
-
},
|
| 732 |
-
{
|
| 733 |
-
"epoch": 0.18495934959349594,
|
| 734 |
-
"grad_norm": 0.12140695750713348,
|
| 735 |
-
"learning_rate": 4.815243902439024e-05,
|
| 736 |
-
"loss": 0.0271,
|
| 737 |
-
"step": 910
|
| 738 |
-
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 0.18699186991869918,
|
| 741 |
-
"grad_norm": 0.1151667982339859,
|
| 742 |
-
"learning_rate": 4.813211382113822e-05,
|
| 743 |
-
"loss": 0.0261,
|
| 744 |
-
"step": 920
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 0.18902439024390244,
|
| 748 |
-
"grad_norm": 0.13169187307357788,
|
| 749 |
-
"learning_rate": 4.8111788617886184e-05,
|
| 750 |
-
"loss": 0.0284,
|
| 751 |
-
"step": 930
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.1910569105691057,
|
| 755 |
-
"grad_norm": 0.11413775384426117,
|
| 756 |
-
"learning_rate": 4.8091463414634145e-05,
|
| 757 |
-
"loss": 0.0278,
|
| 758 |
-
"step": 940
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.19308943089430894,
|
| 762 |
-
"grad_norm": 0.13462452590465546,
|
| 763 |
-
"learning_rate": 4.807113821138212e-05,
|
| 764 |
-
"loss": 0.0269,
|
| 765 |
-
"step": 950
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 0.1951219512195122,
|
| 769 |
-
"grad_norm": 0.0989966094493866,
|
| 770 |
-
"learning_rate": 4.8050813008130086e-05,
|
| 771 |
-
"loss": 0.0269,
|
| 772 |
-
"step": 960
|
| 773 |
-
},
|
| 774 |
-
{
|
| 775 |
-
"epoch": 0.19715447154471544,
|
| 776 |
-
"grad_norm": 0.11530207097530365,
|
| 777 |
-
"learning_rate": 4.803048780487805e-05,
|
| 778 |
-
"loss": 0.0273,
|
| 779 |
-
"step": 970
|
| 780 |
-
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 0.1991869918699187,
|
| 783 |
-
"grad_norm": 0.11543365567922592,
|
| 784 |
-
"learning_rate": 4.801016260162602e-05,
|
| 785 |
-
"loss": 0.0265,
|
| 786 |
-
"step": 980
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 0.20121951219512196,
|
| 790 |
-
"grad_norm": 0.12322687357664108,
|
| 791 |
-
"learning_rate": 4.798983739837399e-05,
|
| 792 |
-
"loss": 0.0277,
|
| 793 |
-
"step": 990
|
| 794 |
-
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 0.2032520325203252,
|
| 797 |
-
"grad_norm": 0.12608106434345245,
|
| 798 |
-
"learning_rate": 4.796951219512195e-05,
|
| 799 |
-
"loss": 0.0254,
|
| 800 |
-
"step": 1000
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 0.2032520325203252,
|
| 804 |
-
"eval_cosine_similarity": 0.4821661613555485,
|
| 805 |
-
"eval_loss": 0.021430717781186104,
|
| 806 |
-
"eval_mse_sync": 0.021430718056257656,
|
| 807 |
-
"eval_runtime": 8.0256,
|
| 808 |
-
"eval_samples_per_second": 372.931,
|
| 809 |
-
"eval_steps_per_second": 23.425,
|
| 810 |
-
"step": 1000
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 0.20528455284552846,
|
| 814 |
-
"grad_norm": 0.1113312616944313,
|
| 815 |
-
"learning_rate": 4.794918699186992e-05,
|
| 816 |
-
"loss": 0.0277,
|
| 817 |
-
"step": 1010
|
| 818 |
-
},
|
| 819 |
-
{
|
| 820 |
-
"epoch": 0.2073170731707317,
|
| 821 |
-
"grad_norm": 0.1282844841480255,
|
| 822 |
-
"learning_rate": 4.792886178861789e-05,
|
| 823 |
-
"loss": 0.0275,
|
| 824 |
-
"step": 1020
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 0.20934959349593496,
|
| 828 |
-
"grad_norm": 0.10981626808643341,
|
| 829 |
-
"learning_rate": 4.790853658536585e-05,
|
| 830 |
-
"loss": 0.0282,
|
| 831 |
-
"step": 1030
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 0.21138211382113822,
|
| 835 |
-
"grad_norm": 0.1443098783493042,
|
| 836 |
-
"learning_rate": 4.7888211382113825e-05,
|
| 837 |
-
"loss": 0.0268,
|
| 838 |
-
"step": 1040
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 0.21341463414634146,
|
| 842 |
-
"grad_norm": 0.10926985740661621,
|
| 843 |
-
"learning_rate": 4.786788617886179e-05,
|
| 844 |
-
"loss": 0.0265,
|
| 845 |
-
"step": 1050
|
| 846 |
-
},
|
| 847 |
-
{
|
| 848 |
-
"epoch": 0.21544715447154472,
|
| 849 |
-
"grad_norm": 0.12235318124294281,
|
| 850 |
-
"learning_rate": 4.784756097560975e-05,
|
| 851 |
-
"loss": 0.0271,
|
| 852 |
-
"step": 1060
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 0.21747967479674796,
|
| 856 |
-
"grad_norm": 0.12022686749696732,
|
| 857 |
-
"learning_rate": 4.782723577235773e-05,
|
| 858 |
-
"loss": 0.0269,
|
| 859 |
-
"step": 1070
|
| 860 |
-
},
|
| 861 |
-
{
|
| 862 |
-
"epoch": 0.21951219512195122,
|
| 863 |
-
"grad_norm": 0.13947460055351257,
|
| 864 |
-
"learning_rate": 4.7806910569105695e-05,
|
| 865 |
-
"loss": 0.0272,
|
| 866 |
-
"step": 1080
|
| 867 |
-
},
|
| 868 |
-
{
|
| 869 |
-
"epoch": 0.22154471544715448,
|
| 870 |
-
"grad_norm": 0.10289262980222702,
|
| 871 |
-
"learning_rate": 4.778658536585366e-05,
|
| 872 |
-
"loss": 0.027,
|
| 873 |
-
"step": 1090
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 0.22357723577235772,
|
| 877 |
-
"grad_norm": 0.10644431412220001,
|
| 878 |
-
"learning_rate": 4.776626016260163e-05,
|
| 879 |
-
"loss": 0.026,
|
| 880 |
-
"step": 1100
|
| 881 |
-
},
|
| 882 |
-
{
|
| 883 |
-
"epoch": 0.22357723577235772,
|
| 884 |
-
"eval_cosine_similarity": 0.490080117004355,
|
| 885 |
-
"eval_loss": 0.020727790892124176,
|
| 886 |
-
"eval_mse_sync": 0.020727790696711355,
|
| 887 |
-
"eval_runtime": 7.9802,
|
| 888 |
-
"eval_samples_per_second": 375.055,
|
| 889 |
-
"eval_steps_per_second": 23.558,
|
| 890 |
-
"step": 1100
|
| 891 |
-
},
|
| 892 |
-
{
|
| 893 |
-
"epoch": 0.22560975609756098,
|
| 894 |
-
"grad_norm": 0.1217174306511879,
|
| 895 |
-
"learning_rate": 4.77459349593496e-05,
|
| 896 |
-
"loss": 0.0255,
|
| 897 |
-
"step": 1110
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 0.22764227642276422,
|
| 901 |
-
"grad_norm": 0.13680125772953033,
|
| 902 |
-
"learning_rate": 4.7725609756097564e-05,
|
| 903 |
-
"loss": 0.0261,
|
| 904 |
-
"step": 1120
|
| 905 |
-
},
|
| 906 |
-
{
|
| 907 |
-
"epoch": 0.22967479674796748,
|
| 908 |
-
"grad_norm": 0.10098998993635178,
|
| 909 |
-
"learning_rate": 4.770528455284553e-05,
|
| 910 |
-
"loss": 0.0266,
|
| 911 |
-
"step": 1130
|
| 912 |
-
},
|
| 913 |
-
{
|
| 914 |
-
"epoch": 0.23170731707317074,
|
| 915 |
-
"grad_norm": 0.10652109980583191,
|
| 916 |
-
"learning_rate": 4.76849593495935e-05,
|
| 917 |
-
"loss": 0.0254,
|
| 918 |
-
"step": 1140
|
| 919 |
-
},
|
| 920 |
-
{
|
| 921 |
-
"epoch": 0.23373983739837398,
|
| 922 |
-
"grad_norm": 0.14970383048057556,
|
| 923 |
-
"learning_rate": 4.7664634146341466e-05,
|
| 924 |
-
"loss": 0.0255,
|
| 925 |
-
"step": 1150
|
| 926 |
-
},
|
| 927 |
-
{
|
| 928 |
-
"epoch": 0.23577235772357724,
|
| 929 |
-
"grad_norm": 0.10969521850347519,
|
| 930 |
-
"learning_rate": 4.7644308943089434e-05,
|
| 931 |
-
"loss": 0.0262,
|
| 932 |
-
"step": 1160
|
| 933 |
-
},
|
| 934 |
-
{
|
| 935 |
-
"epoch": 0.23780487804878048,
|
| 936 |
-
"grad_norm": 0.18681780993938446,
|
| 937 |
-
"learning_rate": 4.76239837398374e-05,
|
| 938 |
-
"loss": 0.0267,
|
| 939 |
-
"step": 1170
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 0.23983739837398374,
|
| 943 |
-
"grad_norm": 0.13186466693878174,
|
| 944 |
-
"learning_rate": 4.760365853658537e-05,
|
| 945 |
-
"loss": 0.0279,
|
| 946 |
-
"step": 1180
|
| 947 |
-
},
|
| 948 |
-
{
|
| 949 |
-
"epoch": 0.241869918699187,
|
| 950 |
-
"grad_norm": 0.09688113629817963,
|
| 951 |
-
"learning_rate": 4.7583333333333336e-05,
|
| 952 |
-
"loss": 0.0259,
|
| 953 |
-
"step": 1190
|
| 954 |
-
},
|
| 955 |
-
{
|
| 956 |
-
"epoch": 0.24390243902439024,
|
| 957 |
-
"grad_norm": 0.1350603997707367,
|
| 958 |
-
"learning_rate": 4.75630081300813e-05,
|
| 959 |
-
"loss": 0.0274,
|
| 960 |
-
"step": 1200
|
| 961 |
-
},
|
| 962 |
-
{
|
| 963 |
-
"epoch": 0.24390243902439024,
|
| 964 |
-
"eval_cosine_similarity": 0.46065077879657484,
|
| 965 |
-
"eval_loss": 0.02211129106581211,
|
| 966 |
-
"eval_mse_sync": 0.0221112903003415,
|
| 967 |
-
"eval_runtime": 8.0094,
|
| 968 |
-
"eval_samples_per_second": 373.686,
|
| 969 |
-
"eval_steps_per_second": 23.472,
|
| 970 |
-
"step": 1200
|
| 971 |
-
},
|
| 972 |
-
{
|
| 973 |
-
"epoch": 0.2459349593495935,
|
| 974 |
-
"grad_norm": 0.13613583147525787,
|
| 975 |
-
"learning_rate": 4.754268292682927e-05,
|
| 976 |
-
"loss": 0.0275,
|
| 977 |
-
"step": 1210
|
| 978 |
-
},
|
| 979 |
-
{
|
| 980 |
-
"epoch": 0.24796747967479674,
|
| 981 |
-
"grad_norm": 0.1278214156627655,
|
| 982 |
-
"learning_rate": 4.752235772357724e-05,
|
| 983 |
-
"loss": 0.0257,
|
| 984 |
-
"step": 1220
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 0.25,
|
| 988 |
-
"grad_norm": 0.1331510692834854,
|
| 989 |
-
"learning_rate": 4.7502032520325205e-05,
|
| 990 |
-
"loss": 0.0251,
|
| 991 |
-
"step": 1230
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 0.25203252032520324,
|
| 995 |
-
"grad_norm": 0.15280725061893463,
|
| 996 |
-
"learning_rate": 4.748170731707317e-05,
|
| 997 |
-
"loss": 0.0268,
|
| 998 |
-
"step": 1240
|
| 999 |
-
},
|
| 1000 |
-
{
|
| 1001 |
-
"epoch": 0.2540650406504065,
|
| 1002 |
-
"grad_norm": 0.11740286648273468,
|
| 1003 |
-
"learning_rate": 4.746138211382114e-05,
|
| 1004 |
-
"loss": 0.0258,
|
| 1005 |
-
"step": 1250
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 0.25609756097560976,
|
| 1009 |
-
"grad_norm": 0.09965524077415466,
|
| 1010 |
-
"learning_rate": 4.744105691056911e-05,
|
| 1011 |
-
"loss": 0.0274,
|
| 1012 |
-
"step": 1260
|
| 1013 |
-
},
|
| 1014 |
-
{
|
| 1015 |
-
"epoch": 0.258130081300813,
|
| 1016 |
-
"grad_norm": 0.12852798402309418,
|
| 1017 |
-
"learning_rate": 4.7420731707317075e-05,
|
| 1018 |
-
"loss": 0.0272,
|
| 1019 |
-
"step": 1270
|
| 1020 |
-
},
|
| 1021 |
-
{
|
| 1022 |
-
"epoch": 0.2601626016260163,
|
| 1023 |
-
"grad_norm": 0.11415109783411026,
|
| 1024 |
-
"learning_rate": 4.740040650406504e-05,
|
| 1025 |
-
"loss": 0.026,
|
| 1026 |
-
"step": 1280
|
| 1027 |
-
},
|
| 1028 |
-
{
|
| 1029 |
-
"epoch": 0.2621951219512195,
|
| 1030 |
-
"grad_norm": 0.1628946214914322,
|
| 1031 |
-
"learning_rate": 4.738008130081301e-05,
|
| 1032 |
-
"loss": 0.0279,
|
| 1033 |
-
"step": 1290
|
| 1034 |
-
},
|
| 1035 |
-
{
|
| 1036 |
-
"epoch": 0.26422764227642276,
|
| 1037 |
-
"grad_norm": 0.11203841865062714,
|
| 1038 |
-
"learning_rate": 4.735975609756098e-05,
|
| 1039 |
-
"loss": 0.0253,
|
| 1040 |
-
"step": 1300
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 0.26422764227642276,
|
| 1044 |
-
"eval_cosine_similarity": 0.477201213187053,
|
| 1045 |
-
"eval_loss": 0.021068334579467773,
|
| 1046 |
-
"eval_mse_sync": 0.02106833454150541,
|
| 1047 |
-
"eval_runtime": 7.9772,
|
| 1048 |
-
"eval_samples_per_second": 375.194,
|
| 1049 |
-
"eval_steps_per_second": 23.567,
|
| 1050 |
-
"step": 1300
|
| 1051 |
-
},
|
| 1052 |
-
{
|
| 1053 |
-
"epoch": 0.266260162601626,
|
| 1054 |
-
"grad_norm": 0.10183481872081757,
|
| 1055 |
-
"learning_rate": 4.7339430894308944e-05,
|
| 1056 |
-
"loss": 0.0266,
|
| 1057 |
-
"step": 1310
|
| 1058 |
-
},
|
| 1059 |
-
{
|
| 1060 |
-
"epoch": 0.2682926829268293,
|
| 1061 |
-
"grad_norm": 0.09649122506380081,
|
| 1062 |
-
"learning_rate": 4.731910569105691e-05,
|
| 1063 |
-
"loss": 0.0254,
|
| 1064 |
-
"step": 1320
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 0.2703252032520325,
|
| 1068 |
-
"grad_norm": 0.11728145182132721,
|
| 1069 |
-
"learning_rate": 4.729878048780488e-05,
|
| 1070 |
-
"loss": 0.0256,
|
| 1071 |
-
"step": 1330
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 0.27235772357723576,
|
| 1075 |
-
"grad_norm": 0.11149100959300995,
|
| 1076 |
-
"learning_rate": 4.7278455284552846e-05,
|
| 1077 |
-
"loss": 0.0247,
|
| 1078 |
-
"step": 1340
|
| 1079 |
-
},
|
| 1080 |
-
{
|
| 1081 |
-
"epoch": 0.27439024390243905,
|
| 1082 |
-
"grad_norm": 0.12775219976902008,
|
| 1083 |
-
"learning_rate": 4.7258130081300814e-05,
|
| 1084 |
-
"loss": 0.0263,
|
| 1085 |
-
"step": 1350
|
| 1086 |
-
},
|
| 1087 |
-
{
|
| 1088 |
-
"epoch": 0.2764227642276423,
|
| 1089 |
-
"grad_norm": 0.1255025416612625,
|
| 1090 |
-
"learning_rate": 4.723780487804878e-05,
|
| 1091 |
-
"loss": 0.0248,
|
| 1092 |
-
"step": 1360
|
| 1093 |
-
},
|
| 1094 |
-
{
|
| 1095 |
-
"epoch": 0.2784552845528455,
|
| 1096 |
-
"grad_norm": 0.11553100496530533,
|
| 1097 |
-
"learning_rate": 4.721747967479675e-05,
|
| 1098 |
-
"loss": 0.0261,
|
| 1099 |
-
"step": 1370
|
| 1100 |
-
},
|
| 1101 |
-
{
|
| 1102 |
-
"epoch": 0.2804878048780488,
|
| 1103 |
-
"grad_norm": 0.115130715072155,
|
| 1104 |
-
"learning_rate": 4.7197154471544716e-05,
|
| 1105 |
-
"loss": 0.0266,
|
| 1106 |
-
"step": 1380
|
| 1107 |
-
},
|
| 1108 |
-
{
|
| 1109 |
-
"epoch": 0.28252032520325204,
|
| 1110 |
-
"grad_norm": 0.10078372061252594,
|
| 1111 |
-
"learning_rate": 4.717682926829268e-05,
|
| 1112 |
-
"loss": 0.0249,
|
| 1113 |
-
"step": 1390
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 0.2845528455284553,
|
| 1117 |
-
"grad_norm": 0.11324010044336319,
|
| 1118 |
-
"learning_rate": 4.715650406504065e-05,
|
| 1119 |
-
"loss": 0.0257,
|
| 1120 |
-
"step": 1400
|
| 1121 |
-
},
|
| 1122 |
-
{
|
| 1123 |
-
"epoch": 0.2845528455284553,
|
| 1124 |
-
"eval_cosine_similarity": 0.490024376186367,
|
| 1125 |
-
"eval_loss": 0.02175692841410637,
|
| 1126 |
-
"eval_mse_sync": 0.021756927129920514,
|
| 1127 |
-
"eval_runtime": 8.0149,
|
| 1128 |
-
"eval_samples_per_second": 373.432,
|
| 1129 |
-
"eval_steps_per_second": 23.456,
|
| 1130 |
-
"step": 1400
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.2865853658536585,
|
| 1134 |
-
"grad_norm": 0.1149812862277031,
|
| 1135 |
-
"learning_rate": 4.7136178861788625e-05,
|
| 1136 |
-
"loss": 0.0254,
|
| 1137 |
-
"step": 1410
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 0.2886178861788618,
|
| 1141 |
-
"grad_norm": 0.11180515587329865,
|
| 1142 |
-
"learning_rate": 4.7115853658536585e-05,
|
| 1143 |
-
"loss": 0.0266,
|
| 1144 |
-
"step": 1420
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 0.29065040650406504,
|
| 1148 |
-
"grad_norm": 0.11090870201587677,
|
| 1149 |
-
"learning_rate": 4.709552845528455e-05,
|
| 1150 |
-
"loss": 0.0252,
|
| 1151 |
-
"step": 1430
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 0.2926829268292683,
|
| 1155 |
-
"grad_norm": 0.15381398797035217,
|
| 1156 |
-
"learning_rate": 4.707520325203253e-05,
|
| 1157 |
-
"loss": 0.0267,
|
| 1158 |
-
"step": 1440
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 0.29471544715447157,
|
| 1162 |
-
"grad_norm": 0.11547625809907913,
|
| 1163 |
-
"learning_rate": 4.705487804878049e-05,
|
| 1164 |
-
"loss": 0.0259,
|
| 1165 |
-
"step": 1450
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 0.2967479674796748,
|
| 1169 |
-
"grad_norm": 0.12333870679140091,
|
| 1170 |
-
"learning_rate": 4.7034552845528455e-05,
|
| 1171 |
-
"loss": 0.0273,
|
| 1172 |
-
"step": 1460
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.29878048780487804,
|
| 1176 |
-
"grad_norm": 0.13967566192150116,
|
| 1177 |
-
"learning_rate": 4.701422764227643e-05,
|
| 1178 |
-
"loss": 0.0255,
|
| 1179 |
-
"step": 1470
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.3008130081300813,
|
| 1183 |
-
"grad_norm": 0.12606772780418396,
|
| 1184 |
-
"learning_rate": 4.699390243902439e-05,
|
| 1185 |
-
"loss": 0.0245,
|
| 1186 |
-
"step": 1480
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 0.30284552845528456,
|
| 1190 |
-
"grad_norm": 0.11195407062768936,
|
| 1191 |
-
"learning_rate": 4.697357723577236e-05,
|
| 1192 |
-
"loss": 0.0258,
|
| 1193 |
-
"step": 1490
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 0.3048780487804878,
|
| 1197 |
-
"grad_norm": 0.1050952821969986,
|
| 1198 |
-
"learning_rate": 4.695325203252033e-05,
|
| 1199 |
-
"loss": 0.0256,
|
| 1200 |
-
"step": 1500
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 0.3048780487804878,
|
| 1204 |
-
"eval_cosine_similarity": 0.5019669328052943,
|
| 1205 |
-
"eval_loss": 0.020201340317726135,
|
| 1206 |
-
"eval_mse_sync": 0.020201340437836564,
|
| 1207 |
-
"eval_runtime": 8.0028,
|
| 1208 |
-
"eval_samples_per_second": 373.996,
|
| 1209 |
-
"eval_steps_per_second": 23.492,
|
| 1210 |
-
"step": 1500
|
| 1211 |
-
}
|
| 1212 |
-
],
|
| 1213 |
-
"logging_steps": 10,
|
| 1214 |
-
"max_steps": 24600,
|
| 1215 |
-
"num_input_tokens_seen": 0,
|
| 1216 |
-
"num_train_epochs": 5,
|
| 1217 |
-
"save_steps": 100,
|
| 1218 |
-
"stateful_callbacks": {
|
| 1219 |
-
"TrainerControl": {
|
| 1220 |
-
"args": {
|
| 1221 |
-
"should_epoch_stop": false,
|
| 1222 |
-
"should_evaluate": false,
|
| 1223 |
-
"should_log": false,
|
| 1224 |
-
"should_save": true,
|
| 1225 |
-
"should_training_stop": false
|
| 1226 |
-
},
|
| 1227 |
-
"attributes": {}
|
| 1228 |
-
}
|
| 1229 |
-
},
|
| 1230 |
-
"total_flos": 1.1221717745664e+16,
|
| 1231 |
-
"train_batch_size": 16,
|
| 1232 |
-
"trial_name": null,
|
| 1233 |
-
"trial_params": null
|
| 1234 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/training_args.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:cb70f48c7fa1622aca3a115af51cf270a88d8053efe253371c58188a6a763ba4
|
| 3 |
-
size 5841
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/vocab.json
DELETED
|
@@ -1,37 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"unk": 0,
|
| 3 |
-
"pad": 1,
|
| 4 |
-
"/s": 2,
|
| 5 |
-
"s": 3,
|
| 6 |
-
"(": 4,
|
| 7 |
-
")": 5,
|
| 8 |
-
"always": 6,
|
| 9 |
-
"eventually": 7,
|
| 10 |
-
"until": 8,
|
| 11 |
-
"and": 9,
|
| 12 |
-
"or": 10,
|
| 13 |
-
"not": 11,
|
| 14 |
-
">=": 12,
|
| 15 |
-
"<=": 13,
|
| 16 |
-
">": 14,
|
| 17 |
-
"<": 15,
|
| 18 |
-
"=": 16,
|
| 19 |
-
"x_": 17,
|
| 20 |
-
"[": 18,
|
| 21 |
-
"]": 19,
|
| 22 |
-
",": 20,
|
| 23 |
-
"inf": 21,
|
| 24 |
-
"-": 22,
|
| 25 |
-
".": 23,
|
| 26 |
-
"0": 24,
|
| 27 |
-
"1": 25,
|
| 28 |
-
"2": 26,
|
| 29 |
-
"3": 27,
|
| 30 |
-
"4": 28,
|
| 31 |
-
"5": 29,
|
| 32 |
-
"6": 30,
|
| 33 |
-
"7": 31,
|
| 34 |
-
"8": 32,
|
| 35 |
-
"9": 33,
|
| 36 |
-
"@": 34
|
| 37 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|