Upload folder using huggingface_hub
Browse files- README.md +143 -0
- config.json +29 -0
- generation_config.json +8 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +24 -0
- tokenizer.json +0 -0
- tokenizer_config.json +48 -0
- trainer_state.json +1582 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,143 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
## Upstream model config
|
| 5 |
+
```json
|
| 6 |
+
{
|
| 7 |
+
"_name_or_path": "output/hermes-llama2-4k/checkpoint-2259",
|
| 8 |
+
"architectures": [
|
| 9 |
+
"LlamaForCausalLM"
|
| 10 |
+
],
|
| 11 |
+
"bos_token_id": 1,
|
| 12 |
+
"eos_token_id": 2,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 4096,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 11008,
|
| 17 |
+
"max_position_embeddings": 4096,
|
| 18 |
+
"model_type": "llama",
|
| 19 |
+
"num_attention_heads": 32,
|
| 20 |
+
"num_hidden_layers": 32,
|
| 21 |
+
"num_key_value_heads": 32,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"pretraining_tp": 1,
|
| 24 |
+
"rms_norm_eps": 1e-05,
|
| 25 |
+
"rope_scaling": null,
|
| 26 |
+
"tie_word_embeddings": false,
|
| 27 |
+
"torch_dtype": "bfloat16",
|
| 28 |
+
"transformers_version": "4.32.0.dev0",
|
| 29 |
+
"use_cache": false,
|
| 30 |
+
"vocab_size": 32000
|
| 31 |
+
}
|
| 32 |
+
```
|
| 33 |
+
|
| 34 |
+
### Dataset
|
| 35 |
+
```
|
| 36 |
+
DATASET = "abideen/Cosmopedia-100k-pretrain" # @param
|
| 37 |
+
from datasets import load_dataset
|
| 38 |
+
# converted to BitLinear
|
| 39 |
+
|
| 40 |
+
class BitLinear(nn.Linear):
|
| 41 |
+
def forward(self, x):
|
| 42 |
+
w = self.weight # a weight tensor with shape [d, k]
|
| 43 |
+
x = x.to(w.device)
|
| 44 |
+
RMSNorm = LlamaRMSNorm(x.shape[-1]).to(w.device)
|
| 45 |
+
x_norm = RMSNorm(x)
|
| 46 |
+
# A trick for implementing Straight−Through−Estimator (STE) using detach()
|
| 47 |
+
x_quant = x_norm + (activation_quant(x_norm) - x_norm).detach()
|
| 48 |
+
w_quant = w + (weight_quant(w) - w).detach()
|
| 49 |
+
y = F.linear(x_quant, w_quant)
|
| 50 |
+
return y
|
| 51 |
+
|
| 52 |
+
### Create the llama model with our custom config. Convert it to bitnet.
|
| 53 |
+
model = LlamaForCausalLM(config)
|
| 54 |
+
convert_to_bitnet(model, copy_weights=False)
|
| 55 |
+
```
|
| 56 |
+
### Training
|
| 57 |
+
```python
|
| 58 |
+
args = TrainingArguments(
|
| 59 |
+
output_dir=output_path,
|
| 60 |
+
per_device_train_batch_size=BATCH_SIZE,
|
| 61 |
+
logging_steps=100,
|
| 62 |
+
gradient_accumulation_steps=2,
|
| 63 |
+
num_train_epochs=EPOCHS,
|
| 64 |
+
weight_decay=0.01,
|
| 65 |
+
warmup_steps=0.1,
|
| 66 |
+
lr_scheduler_type="cosine",
|
| 67 |
+
learning_rate=LEARNING_RATE,
|
| 68 |
+
# max_steps=5000,
|
| 69 |
+
save_steps=0.25,
|
| 70 |
+
fp16=True,
|
| 71 |
+
report_to="wandb"
|
| 72 |
+
)
|
| 73 |
+
|
| 74 |
+
trainer = Trainer(
|
| 75 |
+
model=model,
|
| 76 |
+
tokenizer=tokenizer,
|
| 77 |
+
args=args,
|
| 78 |
+
data_collator=data_collator,
|
| 79 |
+
train_dataset=tokenized_data["train"],
|
| 80 |
+
)
|
| 81 |
+
|
| 82 |
+
trainer.train()
|
| 83 |
+
```
|
| 84 |
+
### Inference
|
| 85 |
+
```python
|
| 86 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 87 |
+
from transformers.models.llama.modeling_llama import *
|
| 88 |
+
# Load a pretrained BitNet model
|
| 89 |
+
model = "saadnaeem/Llama2-70M-Cosmopedia-100k-Pretrain"
|
| 90 |
+
tokenizer = AutoTokenizer.from_pretrained(model)
|
| 91 |
+
model = AutoModelForCausalLM.from_pretrained(model)
|
| 92 |
+
|
| 93 |
+
|
| 94 |
+
def activation_quant(x):
|
| 95 |
+
scale = 127.0 / x.abs().max(dim=-1, keepdim=True).values.clamp_(min=1e-5)
|
| 96 |
+
y = (x * scale).round().clamp_(-128, 127)
|
| 97 |
+
y = y / scale
|
| 98 |
+
return y
|
| 99 |
+
def weight_quant(w):
|
| 100 |
+
scale = 1.0 / w.abs().mean().clamp_(min=1e-5)
|
| 101 |
+
u = (w * scale).round().clamp_(-1, 1)
|
| 102 |
+
u = u / scale
|
| 103 |
+
return u
|
| 104 |
+
|
| 105 |
+
class BitLinear(nn.Linear):
|
| 106 |
+
def forward(self, x):
|
| 107 |
+
w = self.weight # a weight tensor with shape [d, k]
|
| 108 |
+
x = x.to(w.device)
|
| 109 |
+
RMSNorm = LlamaRMSNorm(x.shape[-1]).to(w.device)
|
| 110 |
+
x_norm = RMSNorm(x)
|
| 111 |
+
# A trick for implementing Straight−Through−Estimator (STE) using detach()
|
| 112 |
+
x_quant = x_norm + (activation_quant(x_norm) - x_norm).detach()
|
| 113 |
+
w_quant = w + (weight_quant(w) - w).detach()
|
| 114 |
+
y = F.linear(x_quant, w_quant)
|
| 115 |
+
return y
|
| 116 |
+
|
| 117 |
+
def convert_to_bitnet(model, copy_weights):
|
| 118 |
+
for name, module in model.named_modules():
|
| 119 |
+
# Replace linear layers with BitNet
|
| 120 |
+
if isinstance(module, LlamaSdpaAttention) or isinstance(module, LlamaMLP):
|
| 121 |
+
for child_name, child_module in module.named_children():
|
| 122 |
+
if isinstance(child_module, nn.Linear):
|
| 123 |
+
bitlinear = BitLinear(child_module.in_features, child_module.out_features, child_module.bias is not None).to(device="cuda:0")
|
| 124 |
+
if copy_weights:
|
| 125 |
+
bitlinear.weight = child_module.weight
|
| 126 |
+
if child_module.bias is not None:
|
| 127 |
+
bitlinear.bias = child_module.bias
|
| 128 |
+
setattr(module, child_name, bitlinear)
|
| 129 |
+
# Remove redundant input_layernorms
|
| 130 |
+
elif isinstance(module, LlamaDecoderLayer):
|
| 131 |
+
for child_name, child_module in module.named_children():
|
| 132 |
+
if isinstance(child_module, LlamaRMSNorm) and child_name == "input_layernorm":
|
| 133 |
+
setattr(module, child_name, nn.Identity().to(device="cuda:0"))
|
| 134 |
+
|
| 135 |
+
|
| 136 |
+
convert_to_bitnet(model, copy_weights=True)
|
| 137 |
+
model.to(device="cuda:0")
|
| 138 |
+
|
| 139 |
+
prompt = "What is Machine Learning?"
|
| 140 |
+
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
|
| 141 |
+
generate_ids = model.generate(inputs.input_ids, max_length=50)
|
| 142 |
+
tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
|
| 143 |
+
```
|
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "saadnaeem/Llama2-70M-Cosmopedia-100k-Pretrained",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 1,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 1024,
|
| 14 |
+
"max_position_embeddings": 768,
|
| 15 |
+
"model_type": "llama",
|
| 16 |
+
"num_attention_heads": 6,
|
| 17 |
+
"num_hidden_layers": 6,
|
| 18 |
+
"num_key_value_heads": 6,
|
| 19 |
+
"pad_token_id": 0,
|
| 20 |
+
"pretraining_tp": 1,
|
| 21 |
+
"rms_norm_eps": 1e-05,
|
| 22 |
+
"rope_scaling": null,
|
| 23 |
+
"rope_theta": 10000.0,
|
| 24 |
+
"tie_word_embeddings": false,
|
| 25 |
+
"torch_dtype": "float32",
|
| 26 |
+
"transformers_version": "4.39.0",
|
| 27 |
+
"use_cache": false,
|
| 28 |
+
"vocab_size": 32001
|
| 29 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.39.0",
|
| 7 |
+
"use_cache": false
|
| 8 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc27bd7963df45401a0ed885fb8c24506cc606041de2e09724981e64178d44e7
|
| 3 |
+
size 309887520
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c09424c1194a43809978b40c413c36ea407543160caf50bbbfe404166632080
|
| 3 |
+
size 619806725
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2278a87cdf86c3f9219223c847f6b27f6b7f15b8226b617f38936e8ff2cbcde
|
| 3 |
+
size 14575
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1dfedaca28676ad415c14a596160deaa0904b3d8cfa7d6515608ecc31a09c795
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": true,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": true,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"32000": {
|
| 30 |
+
"content": "<pad>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": true,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": false
|
| 36 |
+
}
|
| 37 |
+
},
|
| 38 |
+
"bos_token": "<s>",
|
| 39 |
+
"clean_up_tokenization_spaces": false,
|
| 40 |
+
"eos_token": "</s>",
|
| 41 |
+
"legacy": false,
|
| 42 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 43 |
+
"pad_token": "</s>",
|
| 44 |
+
"sp_model_kwargs": {},
|
| 45 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 46 |
+
"unk_token": "<unk>",
|
| 47 |
+
"use_default_system_prompt": false
|
| 48 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1582 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.5001006914143789,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 22347,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"grad_norm": 1.514354944229126,
|
| 14 |
+
"learning_rate": 0.0001499958389468596,
|
| 15 |
+
"loss": 6.8144,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01,
|
| 20 |
+
"grad_norm": 1.3240108489990234,
|
| 21 |
+
"learning_rate": 0.00014998333958519128,
|
| 22 |
+
"loss": 5.3488,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02,
|
| 27 |
+
"grad_norm": 1.3611284494400024,
|
| 28 |
+
"learning_rate": 0.00014996250330055047,
|
| 29 |
+
"loss": 4.9362,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03,
|
| 34 |
+
"grad_norm": 1.3592987060546875,
|
| 35 |
+
"learning_rate": 0.00014993333240959187,
|
| 36 |
+
"loss": 4.7356,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03,
|
| 41 |
+
"grad_norm": 1.3300018310546875,
|
| 42 |
+
"learning_rate": 0.00014989583015564216,
|
| 43 |
+
"loss": 4.5698,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04,
|
| 48 |
+
"grad_norm": 1.3178701400756836,
|
| 49 |
+
"learning_rate": 0.00014985000070833962,
|
| 50 |
+
"loss": 4.4608,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05,
|
| 55 |
+
"grad_norm": 1.2650636434555054,
|
| 56 |
+
"learning_rate": 0.00014979584916317048,
|
| 57 |
+
"loss": 4.3331,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05,
|
| 62 |
+
"grad_norm": 1.3434759378433228,
|
| 63 |
+
"learning_rate": 0.00014973338154090234,
|
| 64 |
+
"loss": 4.2745,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06,
|
| 69 |
+
"grad_norm": 1.3090848922729492,
|
| 70 |
+
"learning_rate": 0.00014966260478691473,
|
| 71 |
+
"loss": 4.1916,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07,
|
| 76 |
+
"grad_norm": 1.2796646356582642,
|
| 77 |
+
"learning_rate": 0.000149583526770427,
|
| 78 |
+
"loss": 4.1233,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07,
|
| 83 |
+
"grad_norm": 1.3859351873397827,
|
| 84 |
+
"learning_rate": 0.00014949615628362335,
|
| 85 |
+
"loss": 4.0486,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08,
|
| 90 |
+
"grad_norm": 1.3429782390594482,
|
| 91 |
+
"learning_rate": 0.00014940050304067528,
|
| 92 |
+
"loss": 4.0015,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09,
|
| 97 |
+
"grad_norm": 1.3343886137008667,
|
| 98 |
+
"learning_rate": 0.00014929657767666143,
|
| 99 |
+
"loss": 3.9459,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.09,
|
| 104 |
+
"grad_norm": 1.3510159254074097,
|
| 105 |
+
"learning_rate": 0.0001491843917463854,
|
| 106 |
+
"loss": 3.9364,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1,
|
| 111 |
+
"grad_norm": 1.3994427919387817,
|
| 112 |
+
"learning_rate": 0.00014906395772309067,
|
| 113 |
+
"loss": 3.8667,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.11,
|
| 118 |
+
"grad_norm": 1.3028327226638794,
|
| 119 |
+
"learning_rate": 0.00014893528899707415,
|
| 120 |
+
"loss": 3.8425,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.11,
|
| 125 |
+
"grad_norm": 1.3036125898361206,
|
| 126 |
+
"learning_rate": 0.00014879839987419713,
|
| 127 |
+
"loss": 3.8103,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.12,
|
| 132 |
+
"grad_norm": 1.250184178352356,
|
| 133 |
+
"learning_rate": 0.00014865330557429485,
|
| 134 |
+
"loss": 3.7818,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.13,
|
| 139 |
+
"grad_norm": 1.2782858610153198,
|
| 140 |
+
"learning_rate": 0.00014850002222948426,
|
| 141 |
+
"loss": 3.7404,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.13,
|
| 146 |
+
"grad_norm": 1.232730746269226,
|
| 147 |
+
"learning_rate": 0.00014833856688237034,
|
| 148 |
+
"loss": 3.7116,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.14,
|
| 153 |
+
"grad_norm": 1.2852957248687744,
|
| 154 |
+
"learning_rate": 0.00014816895748415133,
|
| 155 |
+
"loss": 3.7033,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.15,
|
| 160 |
+
"grad_norm": 1.3249268531799316,
|
| 161 |
+
"learning_rate": 0.0001479912128926228,
|
| 162 |
+
"loss": 3.6881,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.15,
|
| 167 |
+
"grad_norm": 1.222371220588684,
|
| 168 |
+
"learning_rate": 0.0001478053528700809,
|
| 169 |
+
"loss": 3.6735,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.16,
|
| 174 |
+
"grad_norm": 1.3207260370254517,
|
| 175 |
+
"learning_rate": 0.00014761139808112535,
|
| 176 |
+
"loss": 3.59,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17,
|
| 181 |
+
"grad_norm": 1.226083517074585,
|
| 182 |
+
"learning_rate": 0.00014740937009036158,
|
| 183 |
+
"loss": 3.6171,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.17,
|
| 188 |
+
"grad_norm": 1.2377468347549438,
|
| 189 |
+
"learning_rate": 0.0001471992913600033,
|
| 190 |
+
"loss": 3.5804,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.18,
|
| 195 |
+
"grad_norm": 1.2854132652282715,
|
| 196 |
+
"learning_rate": 0.00014698118524737497,
|
| 197 |
+
"loss": 3.5831,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.19,
|
| 202 |
+
"grad_norm": 1.2622026205062866,
|
| 203 |
+
"learning_rate": 0.00014675507600231494,
|
| 204 |
+
"loss": 3.5555,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.19,
|
| 209 |
+
"grad_norm": 1.2080059051513672,
|
| 210 |
+
"learning_rate": 0.00014652098876447912,
|
| 211 |
+
"loss": 3.5597,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2,
|
| 216 |
+
"grad_norm": 1.1679385900497437,
|
| 217 |
+
"learning_rate": 0.00014627894956054604,
|
| 218 |
+
"loss": 3.5245,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.21,
|
| 223 |
+
"grad_norm": 1.323114275932312,
|
| 224 |
+
"learning_rate": 0.0001460289853013229,
|
| 225 |
+
"loss": 3.5278,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.21,
|
| 230 |
+
"grad_norm": 1.1871461868286133,
|
| 231 |
+
"learning_rate": 0.00014577112377875374,
|
| 232 |
+
"loss": 3.5065,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.22,
|
| 237 |
+
"grad_norm": 1.2530100345611572,
|
| 238 |
+
"learning_rate": 0.0001455053936628293,
|
| 239 |
+
"loss": 3.4857,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.23,
|
| 244 |
+
"grad_norm": 1.2120153903961182,
|
| 245 |
+
"learning_rate": 0.00014523182449839938,
|
| 246 |
+
"loss": 3.469,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.23,
|
| 251 |
+
"grad_norm": 1.2787284851074219,
|
| 252 |
+
"learning_rate": 0.00014495044670188802,
|
| 253 |
+
"loss": 3.4693,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.24,
|
| 258 |
+
"grad_norm": 1.273194670677185,
|
| 259 |
+
"learning_rate": 0.00014466129155791166,
|
| 260 |
+
"loss": 3.4377,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.25,
|
| 265 |
+
"grad_norm": 1.2218822240829468,
|
| 266 |
+
"learning_rate": 0.00014436439121580068,
|
| 267 |
+
"loss": 3.4361,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.26,
|
| 272 |
+
"grad_norm": 1.241590142250061,
|
| 273 |
+
"learning_rate": 0.00014405977868602515,
|
| 274 |
+
"loss": 3.4371,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.26,
|
| 279 |
+
"grad_norm": 1.179774522781372,
|
| 280 |
+
"learning_rate": 0.0001437474878365244,
|
| 281 |
+
"loss": 3.3945,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.27,
|
| 286 |
+
"grad_norm": 1.269691824913025,
|
| 287 |
+
"learning_rate": 0.00014342755338894156,
|
| 288 |
+
"loss": 3.4382,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.28,
|
| 293 |
+
"grad_norm": 1.2015455961227417,
|
| 294 |
+
"learning_rate": 0.00014310001091476303,
|
| 295 |
+
"loss": 3.4006,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.28,
|
| 300 |
+
"grad_norm": 1.2061266899108887,
|
| 301 |
+
"learning_rate": 0.00014276489683136354,
|
| 302 |
+
"loss": 3.3891,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.29,
|
| 307 |
+
"grad_norm": 1.2331507205963135,
|
| 308 |
+
"learning_rate": 0.00014242224839795722,
|
| 309 |
+
"loss": 3.3673,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3,
|
| 314 |
+
"grad_norm": 1.173548936843872,
|
| 315 |
+
"learning_rate": 0.00014207210371145474,
|
| 316 |
+
"loss": 3.3796,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3,
|
| 321 |
+
"grad_norm": 1.564943790435791,
|
| 322 |
+
"learning_rate": 0.00014171450170222784,
|
| 323 |
+
"loss": 3.3584,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.31,
|
| 328 |
+
"grad_norm": 1.2429152727127075,
|
| 329 |
+
"learning_rate": 0.00014134948212978066,
|
| 330 |
+
"loss": 3.3383,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.32,
|
| 335 |
+
"grad_norm": 1.2660589218139648,
|
| 336 |
+
"learning_rate": 0.00014097708557832942,
|
| 337 |
+
"loss": 3.3562,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.32,
|
| 342 |
+
"grad_norm": 1.181443214416504,
|
| 343 |
+
"learning_rate": 0.00014059735345228977,
|
| 344 |
+
"loss": 3.3545,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.33,
|
| 349 |
+
"grad_norm": 1.172049641609192,
|
| 350 |
+
"learning_rate": 0.0001402103279716736,
|
| 351 |
+
"loss": 3.3377,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.34,
|
| 356 |
+
"grad_norm": 1.3030186891555786,
|
| 357 |
+
"learning_rate": 0.0001398160521673946,
|
| 358 |
+
"loss": 3.2984,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.34,
|
| 363 |
+
"grad_norm": 1.2264354228973389,
|
| 364 |
+
"learning_rate": 0.00013941456987648425,
|
| 365 |
+
"loss": 3.284,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.35,
|
| 370 |
+
"grad_norm": 1.2764644622802734,
|
| 371 |
+
"learning_rate": 0.0001390059257372175,
|
| 372 |
+
"loss": 3.3014,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.36,
|
| 377 |
+
"grad_norm": 1.2215057611465454,
|
| 378 |
+
"learning_rate": 0.00013859016518414998,
|
| 379 |
+
"loss": 3.2988,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.36,
|
| 384 |
+
"grad_norm": 1.2118747234344482,
|
| 385 |
+
"learning_rate": 0.00013816733444306633,
|
| 386 |
+
"loss": 3.295,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.37,
|
| 391 |
+
"grad_norm": 1.2118418216705322,
|
| 392 |
+
"learning_rate": 0.0001377374805258406,
|
| 393 |
+
"loss": 3.2743,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.38,
|
| 398 |
+
"grad_norm": 1.282964825630188,
|
| 399 |
+
"learning_rate": 0.00013730065122520942,
|
| 400 |
+
"loss": 3.2832,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.38,
|
| 405 |
+
"grad_norm": 1.1720709800720215,
|
| 406 |
+
"learning_rate": 0.00013685689510945802,
|
| 407 |
+
"loss": 3.2774,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.39,
|
| 412 |
+
"grad_norm": 1.1780056953430176,
|
| 413 |
+
"learning_rate": 0.0001364062615170206,
|
| 414 |
+
"loss": 3.2668,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.4,
|
| 419 |
+
"grad_norm": 1.1722214221954346,
|
| 420 |
+
"learning_rate": 0.0001359488005509942,
|
| 421 |
+
"loss": 3.2803,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4,
|
| 426 |
+
"grad_norm": 1.1970551013946533,
|
| 427 |
+
"learning_rate": 0.00013548456307356852,
|
| 428 |
+
"loss": 3.2319,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.41,
|
| 433 |
+
"grad_norm": 1.218648910522461,
|
| 434 |
+
"learning_rate": 0.00013501360070037057,
|
| 435 |
+
"loss": 3.2619,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.42,
|
| 440 |
+
"grad_norm": 1.1801834106445312,
|
| 441 |
+
"learning_rate": 0.00013453596579472596,
|
| 442 |
+
"loss": 3.2444,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.42,
|
| 447 |
+
"grad_norm": 1.1807547807693481,
|
| 448 |
+
"learning_rate": 0.00013405171146183706,
|
| 449 |
+
"loss": 3.2167,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.43,
|
| 454 |
+
"grad_norm": 1.1896024942398071,
|
| 455 |
+
"learning_rate": 0.00013356089154287833,
|
| 456 |
+
"loss": 3.201,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.44,
|
| 461 |
+
"grad_norm": 1.1902315616607666,
|
| 462 |
+
"learning_rate": 0.00013306356060901025,
|
| 463 |
+
"loss": 3.2061,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.44,
|
| 468 |
+
"grad_norm": 1.2040640115737915,
|
| 469 |
+
"learning_rate": 0.00013255977395531184,
|
| 470 |
+
"loss": 3.167,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.45,
|
| 475 |
+
"grad_norm": 1.3017876148223877,
|
| 476 |
+
"learning_rate": 0.00013204958759463278,
|
| 477 |
+
"loss": 3.212,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.46,
|
| 482 |
+
"grad_norm": 1.2039344310760498,
|
| 483 |
+
"learning_rate": 0.00013153305825136562,
|
| 484 |
+
"loss": 3.1882,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.46,
|
| 489 |
+
"grad_norm": 1.2072499990463257,
|
| 490 |
+
"learning_rate": 0.000131010243355139,
|
| 491 |
+
"loss": 3.215,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.47,
|
| 496 |
+
"grad_norm": 1.2576429843902588,
|
| 497 |
+
"learning_rate": 0.00013048120103443245,
|
| 498 |
+
"loss": 3.1899,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.48,
|
| 503 |
+
"grad_norm": 1.167734146118164,
|
| 504 |
+
"learning_rate": 0.00012994599011011335,
|
| 505 |
+
"loss": 3.1757,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.48,
|
| 510 |
+
"grad_norm": 1.2545503377914429,
|
| 511 |
+
"learning_rate": 0.00012940467008889722,
|
| 512 |
+
"loss": 3.1914,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.49,
|
| 517 |
+
"grad_norm": 1.1583225727081299,
|
| 518 |
+
"learning_rate": 0.00012885730115673122,
|
| 519 |
+
"loss": 3.1758,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5,
|
| 524 |
+
"grad_norm": 1.144148588180542,
|
| 525 |
+
"learning_rate": 0.00012830394417210275,
|
| 526 |
+
"loss": 3.1448,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5,
|
| 531 |
+
"grad_norm": 1.1552996635437012,
|
| 532 |
+
"learning_rate": 0.00012774466065927281,
|
| 533 |
+
"loss": 3.1527,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.51,
|
| 538 |
+
"grad_norm": 1.279171347618103,
|
| 539 |
+
"learning_rate": 0.00012717951280143562,
|
| 540 |
+
"loss": 3.1414,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.52,
|
| 545 |
+
"grad_norm": 1.1666594743728638,
|
| 546 |
+
"learning_rate": 0.00012660856343380474,
|
| 547 |
+
"loss": 3.159,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.52,
|
| 552 |
+
"grad_norm": 1.201820969581604,
|
| 553 |
+
"learning_rate": 0.00012603187603662695,
|
| 554 |
+
"loss": 3.129,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.53,
|
| 559 |
+
"grad_norm": 1.1756407022476196,
|
| 560 |
+
"learning_rate": 0.00012544951472812416,
|
| 561 |
+
"loss": 3.11,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.54,
|
| 566 |
+
"grad_norm": 1.1489793062210083,
|
| 567 |
+
"learning_rate": 0.00012486154425736464,
|
| 568 |
+
"loss": 3.1224,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.54,
|
| 573 |
+
"grad_norm": 1.154188632965088,
|
| 574 |
+
"learning_rate": 0.00012426802999706388,
|
| 575 |
+
"loss": 3.0954,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.55,
|
| 580 |
+
"grad_norm": 1.1527070999145508,
|
| 581 |
+
"learning_rate": 0.00012366903793631617,
|
| 582 |
+
"loss": 3.1256,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.56,
|
| 587 |
+
"grad_norm": 1.1733514070510864,
|
| 588 |
+
"learning_rate": 0.00012306463467325792,
|
| 589 |
+
"loss": 3.1132,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.56,
|
| 594 |
+
"grad_norm": 1.2460672855377197,
|
| 595 |
+
"learning_rate": 0.0001224548874076627,
|
| 596 |
+
"loss": 3.1227,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.57,
|
| 601 |
+
"grad_norm": 1.1254640817642212,
|
| 602 |
+
"learning_rate": 0.00012183986393347009,
|
| 603 |
+
"loss": 3.1007,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.58,
|
| 608 |
+
"grad_norm": 1.0967683792114258,
|
| 609 |
+
"learning_rate": 0.00012121963263124774,
|
| 610 |
+
"loss": 3.0978,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.58,
|
| 615 |
+
"grad_norm": 1.1684861183166504,
|
| 616 |
+
"learning_rate": 0.00012059426246058886,
|
| 617 |
+
"loss": 3.1022,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.59,
|
| 622 |
+
"grad_norm": 1.2061405181884766,
|
| 623 |
+
"learning_rate": 0.00011996382295244481,
|
| 624 |
+
"loss": 3.0694,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6,
|
| 629 |
+
"grad_norm": 1.1969140768051147,
|
| 630 |
+
"learning_rate": 0.00011932838420139461,
|
| 631 |
+
"loss": 3.0657,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6,
|
| 636 |
+
"grad_norm": 1.1569503545761108,
|
| 637 |
+
"learning_rate": 0.0001186880168578514,
|
| 638 |
+
"loss": 3.0282,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.61,
|
| 643 |
+
"grad_norm": 1.1696115732192993,
|
| 644 |
+
"learning_rate": 0.00011804279212020734,
|
| 645 |
+
"loss": 3.092,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.62,
|
| 650 |
+
"grad_norm": 1.1510956287384033,
|
| 651 |
+
"learning_rate": 0.0001173927817269175,
|
| 652 |
+
"loss": 3.0625,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.62,
|
| 657 |
+
"grad_norm": 1.2039140462875366,
|
| 658 |
+
"learning_rate": 0.00011673805794852368,
|
| 659 |
+
"loss": 3.092,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.63,
|
| 664 |
+
"grad_norm": 1.1726746559143066,
|
| 665 |
+
"learning_rate": 0.00011607869357961926,
|
| 666 |
+
"loss": 3.0508,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.64,
|
| 671 |
+
"grad_norm": 1.1399447917938232,
|
| 672 |
+
"learning_rate": 0.0001154147619307553,
|
| 673 |
+
"loss": 3.0551,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.64,
|
| 678 |
+
"grad_norm": 1.1841970682144165,
|
| 679 |
+
"learning_rate": 0.00011474633682029,
|
| 680 |
+
"loss": 3.0403,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.65,
|
| 685 |
+
"grad_norm": 1.1455024480819702,
|
| 686 |
+
"learning_rate": 0.00011407349256618108,
|
| 687 |
+
"loss": 3.0581,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.66,
|
| 692 |
+
"grad_norm": 1.1628191471099854,
|
| 693 |
+
"learning_rate": 0.00011339630397772279,
|
| 694 |
+
"loss": 3.0535,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.66,
|
| 699 |
+
"grad_norm": 1.1066713333129883,
|
| 700 |
+
"learning_rate": 0.0001127148463472285,
|
| 701 |
+
"loss": 3.0366,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.67,
|
| 706 |
+
"grad_norm": 1.1257274150848389,
|
| 707 |
+
"learning_rate": 0.0001120291954416593,
|
| 708 |
+
"loss": 3.0328,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.68,
|
| 713 |
+
"grad_norm": 1.1711469888687134,
|
| 714 |
+
"learning_rate": 0.00011134634530159263,
|
| 715 |
+
"loss": 3.0334,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.68,
|
| 720 |
+
"grad_norm": 1.1563602685928345,
|
| 721 |
+
"learning_rate": 0.0001106525770255833,
|
| 722 |
+
"loss": 3.0456,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.69,
|
| 727 |
+
"grad_norm": 1.1844093799591064,
|
| 728 |
+
"learning_rate": 0.00010995484476517324,
|
| 729 |
+
"loss": 3.0022,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.7,
|
| 734 |
+
"grad_norm": 1.182077169418335,
|
| 735 |
+
"learning_rate": 0.00010925322609679667,
|
| 736 |
+
"loss": 3.0217,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7,
|
| 741 |
+
"grad_norm": 1.0532621145248413,
|
| 742 |
+
"learning_rate": 0.00010854779902899285,
|
| 743 |
+
"loss": 3.0144,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.71,
|
| 748 |
+
"grad_norm": 1.1866600513458252,
|
| 749 |
+
"learning_rate": 0.00010783864199373296,
|
| 750 |
+
"loss": 3.0324,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.72,
|
| 755 |
+
"grad_norm": 1.1243449449539185,
|
| 756 |
+
"learning_rate": 0.00010712583383769964,
|
| 757 |
+
"loss": 3.0104,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.72,
|
| 762 |
+
"grad_norm": 1.1526159048080444,
|
| 763 |
+
"learning_rate": 0.0001064094538135205,
|
| 764 |
+
"loss": 3.0088,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.73,
|
| 769 |
+
"grad_norm": 1.1495548486709595,
|
| 770 |
+
"learning_rate": 0.00010568958157095663,
|
| 771 |
+
"loss": 3.0232,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.74,
|
| 776 |
+
"grad_norm": 1.157415747642517,
|
| 777 |
+
"learning_rate": 0.00010496629714804666,
|
| 778 |
+
"loss": 3.0066,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.75,
|
| 783 |
+
"grad_norm": 1.1491670608520508,
|
| 784 |
+
"learning_rate": 0.00010423968096220812,
|
| 785 |
+
"loss": 3.0175,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.75,
|
| 790 |
+
"grad_norm": 1.1492184400558472,
|
| 791 |
+
"learning_rate": 0.00010350981380129603,
|
| 792 |
+
"loss": 3.006,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.76,
|
| 797 |
+
"grad_norm": 1.1413304805755615,
|
| 798 |
+
"learning_rate": 0.00010277677681462085,
|
| 799 |
+
"loss": 2.9945,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.77,
|
| 804 |
+
"grad_norm": 1.2421963214874268,
|
| 805 |
+
"learning_rate": 0.00010204065150392585,
|
| 806 |
+
"loss": 2.9847,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.77,
|
| 811 |
+
"grad_norm": 1.1723310947418213,
|
| 812 |
+
"learning_rate": 0.00010130892564472877,
|
| 813 |
+
"loss": 2.9935,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.78,
|
| 818 |
+
"grad_norm": 1.222485065460205,
|
| 819 |
+
"learning_rate": 0.00010056689839074104,
|
| 820 |
+
"loss": 2.9807,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.79,
|
| 825 |
+
"grad_norm": 1.1612520217895508,
|
| 826 |
+
"learning_rate": 9.982202851512956e-05,
|
| 827 |
+
"loss": 2.981,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.79,
|
| 832 |
+
"grad_norm": 1.2819604873657227,
|
| 833 |
+
"learning_rate": 9.907439883526162e-05,
|
| 834 |
+
"loss": 2.982,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.8,
|
| 839 |
+
"grad_norm": 1.139885663986206,
|
| 840 |
+
"learning_rate": 9.832409247534969e-05,
|
| 841 |
+
"loss": 2.9772,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.81,
|
| 846 |
+
"grad_norm": 1.188217282295227,
|
| 847 |
+
"learning_rate": 9.757119285720941e-05,
|
| 848 |
+
"loss": 2.9569,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.81,
|
| 853 |
+
"grad_norm": 1.1877431869506836,
|
| 854 |
+
"learning_rate": 9.681578369098434e-05,
|
| 855 |
+
"loss": 2.9621,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.82,
|
| 860 |
+
"grad_norm": 1.1754024028778076,
|
| 861 |
+
"learning_rate": 9.605794896583886e-05,
|
| 862 |
+
"loss": 2.9724,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.83,
|
| 867 |
+
"grad_norm": 1.1438241004943848,
|
| 868 |
+
"learning_rate": 9.52977729406198e-05,
|
| 869 |
+
"loss": 2.9519,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.83,
|
| 874 |
+
"grad_norm": 1.1510008573532104,
|
| 875 |
+
"learning_rate": 9.45353401344884e-05,
|
| 876 |
+
"loss": 2.953,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.84,
|
| 881 |
+
"grad_norm": 1.1256269216537476,
|
| 882 |
+
"learning_rate": 9.377073531752312e-05,
|
| 883 |
+
"loss": 2.9507,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.85,
|
| 888 |
+
"grad_norm": 1.1786478757858276,
|
| 889 |
+
"learning_rate": 9.300404350129445e-05,
|
| 890 |
+
"loss": 2.9556,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.85,
|
| 895 |
+
"grad_norm": 1.0819822549819946,
|
| 896 |
+
"learning_rate": 9.223534992941322e-05,
|
| 897 |
+
"loss": 2.9454,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.86,
|
| 902 |
+
"grad_norm": 1.151418685913086,
|
| 903 |
+
"learning_rate": 9.146474006805276e-05,
|
| 904 |
+
"loss": 2.9253,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.87,
|
| 909 |
+
"grad_norm": 1.1259241104125977,
|
| 910 |
+
"learning_rate": 9.069229959644649e-05,
|
| 911 |
+
"loss": 2.9155,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.87,
|
| 916 |
+
"grad_norm": 1.1931371688842773,
|
| 917 |
+
"learning_rate": 8.991811439736184e-05,
|
| 918 |
+
"loss": 2.9481,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.88,
|
| 923 |
+
"grad_norm": 1.1462775468826294,
|
| 924 |
+
"learning_rate": 8.91422705475514e-05,
|
| 925 |
+
"loss": 2.9172,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.89,
|
| 930 |
+
"grad_norm": 1.101654052734375,
|
| 931 |
+
"learning_rate": 8.836485430818266e-05,
|
| 932 |
+
"loss": 2.9058,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.89,
|
| 937 |
+
"grad_norm": 1.175592303276062,
|
| 938 |
+
"learning_rate": 8.758595211524713e-05,
|
| 939 |
+
"loss": 2.9418,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.9,
|
| 944 |
+
"grad_norm": 1.1647474765777588,
|
| 945 |
+
"learning_rate": 8.680565056995012e-05,
|
| 946 |
+
"loss": 2.9435,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.91,
|
| 951 |
+
"grad_norm": 1.0977836847305298,
|
| 952 |
+
"learning_rate": 8.602403642908208e-05,
|
| 953 |
+
"loss": 2.9332,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.91,
|
| 958 |
+
"grad_norm": 1.2897889614105225,
|
| 959 |
+
"learning_rate": 8.524119659537264e-05,
|
| 960 |
+
"loss": 2.9269,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.92,
|
| 965 |
+
"grad_norm": 1.1218163967132568,
|
| 966 |
+
"learning_rate": 8.445721810782842e-05,
|
| 967 |
+
"loss": 2.9158,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.93,
|
| 972 |
+
"grad_norm": 1.1555371284484863,
|
| 973 |
+
"learning_rate": 8.367218813205584e-05,
|
| 974 |
+
"loss": 2.9385,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.93,
|
| 979 |
+
"grad_norm": 1.2027695178985596,
|
| 980 |
+
"learning_rate": 8.288619395056957e-05,
|
| 981 |
+
"loss": 2.9362,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.94,
|
| 986 |
+
"grad_norm": 1.1329079866409302,
|
| 987 |
+
"learning_rate": 8.209932295308822e-05,
|
| 988 |
+
"loss": 2.8918,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.95,
|
| 993 |
+
"grad_norm": 1.1126582622528076,
|
| 994 |
+
"learning_rate": 8.131166262681805e-05,
|
| 995 |
+
"loss": 2.8756,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.95,
|
| 1000 |
+
"grad_norm": 1.1206066608428955,
|
| 1001 |
+
"learning_rate": 8.052330054672577e-05,
|
| 1002 |
+
"loss": 2.9013,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.96,
|
| 1007 |
+
"grad_norm": 1.149262547492981,
|
| 1008 |
+
"learning_rate": 7.973432436580172e-05,
|
| 1009 |
+
"loss": 2.9273,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.97,
|
| 1014 |
+
"grad_norm": 1.135891318321228,
|
| 1015 |
+
"learning_rate": 7.894482180531406e-05,
|
| 1016 |
+
"loss": 2.9141,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.97,
|
| 1021 |
+
"grad_norm": 1.128321886062622,
|
| 1022 |
+
"learning_rate": 7.815488064505597e-05,
|
| 1023 |
+
"loss": 2.9142,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.98,
|
| 1028 |
+
"grad_norm": 1.1105716228485107,
|
| 1029 |
+
"learning_rate": 7.736458871358563e-05,
|
| 1030 |
+
"loss": 2.8954,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.99,
|
| 1035 |
+
"grad_norm": 1.16962730884552,
|
| 1036 |
+
"learning_rate": 7.657403387846125e-05,
|
| 1037 |
+
"loss": 2.9179,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.99,
|
| 1042 |
+
"grad_norm": 1.1613777875900269,
|
| 1043 |
+
"learning_rate": 7.578330403647171e-05,
|
| 1044 |
+
"loss": 2.8691,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.0,
|
| 1049 |
+
"grad_norm": 1.148231029510498,
|
| 1050 |
+
"learning_rate": 7.499248710386376e-05,
|
| 1051 |
+
"loss": 2.8779,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.01,
|
| 1056 |
+
"grad_norm": 1.1060874462127686,
|
| 1057 |
+
"learning_rate": 7.420167100656715e-05,
|
| 1058 |
+
"loss": 2.8672,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.01,
|
| 1063 |
+
"grad_norm": 1.2024569511413574,
|
| 1064 |
+
"learning_rate": 7.341094367041887e-05,
|
| 1065 |
+
"loss": 2.8258,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.02,
|
| 1070 |
+
"grad_norm": 1.178033709526062,
|
| 1071 |
+
"learning_rate": 7.262039301138704e-05,
|
| 1072 |
+
"loss": 2.8431,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.03,
|
| 1077 |
+
"grad_norm": 1.1503459215164185,
|
| 1078 |
+
"learning_rate": 7.183010692579615e-05,
|
| 1079 |
+
"loss": 2.857,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.03,
|
| 1084 |
+
"grad_norm": 1.1336513757705688,
|
| 1085 |
+
"learning_rate": 7.10401732805545e-05,
|
| 1086 |
+
"loss": 2.8337,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.04,
|
| 1091 |
+
"grad_norm": 1.1450071334838867,
|
| 1092 |
+
"learning_rate": 7.025067990338475e-05,
|
| 1093 |
+
"loss": 2.8487,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.05,
|
| 1098 |
+
"grad_norm": 1.1344424486160278,
|
| 1099 |
+
"learning_rate": 6.946171457305895e-05,
|
| 1100 |
+
"loss": 2.8621,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.05,
|
| 1105 |
+
"grad_norm": 1.1332283020019531,
|
| 1106 |
+
"learning_rate": 6.867336500963898e-05,
|
| 1107 |
+
"loss": 2.8434,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.06,
|
| 1112 |
+
"grad_norm": 1.1330145597457886,
|
| 1113 |
+
"learning_rate": 6.788571886472344e-05,
|
| 1114 |
+
"loss": 2.8555,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.07,
|
| 1119 |
+
"grad_norm": 1.0932788848876953,
|
| 1120 |
+
"learning_rate": 6.709886371170234e-05,
|
| 1121 |
+
"loss": 2.8157,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.07,
|
| 1126 |
+
"grad_norm": 1.1656392812728882,
|
| 1127 |
+
"learning_rate": 6.631288703602026e-05,
|
| 1128 |
+
"loss": 2.81,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.08,
|
| 1133 |
+
"grad_norm": 1.1445406675338745,
|
| 1134 |
+
"learning_rate": 6.552787622544951e-05,
|
| 1135 |
+
"loss": 2.8489,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.09,
|
| 1140 |
+
"grad_norm": 1.141064167022705,
|
| 1141 |
+
"learning_rate": 6.474391856037385e-05,
|
| 1142 |
+
"loss": 2.8315,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.09,
|
| 1147 |
+
"grad_norm": 1.1442021131515503,
|
| 1148 |
+
"learning_rate": 6.39611012040846e-05,
|
| 1149 |
+
"loss": 2.8549,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.1,
|
| 1154 |
+
"grad_norm": 1.1235743761062622,
|
| 1155 |
+
"learning_rate": 6.317951119308926e-05,
|
| 1156 |
+
"loss": 2.8072,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.11,
|
| 1161 |
+
"grad_norm": 1.164588212966919,
|
| 1162 |
+
"learning_rate": 6.23992354274346e-05,
|
| 1163 |
+
"loss": 2.821,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.11,
|
| 1168 |
+
"grad_norm": 1.1956677436828613,
|
| 1169 |
+
"learning_rate": 6.162036066104482e-05,
|
| 1170 |
+
"loss": 2.8299,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.12,
|
| 1175 |
+
"grad_norm": 1.1507422924041748,
|
| 1176 |
+
"learning_rate": 6.08429734920758e-05,
|
| 1177 |
+
"loss": 2.8082,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.13,
|
| 1182 |
+
"grad_norm": 1.1700375080108643,
|
| 1183 |
+
"learning_rate": 6.006716035328694e-05,
|
| 1184 |
+
"loss": 2.8142,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.13,
|
| 1189 |
+
"grad_norm": 1.1259818077087402,
|
| 1190 |
+
"learning_rate": 5.9293007502431174e-05,
|
| 1191 |
+
"loss": 2.8122,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.14,
|
| 1196 |
+
"grad_norm": 1.1975033283233643,
|
| 1197 |
+
"learning_rate": 5.8520601012664463e-05,
|
| 1198 |
+
"loss": 2.819,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.15,
|
| 1203 |
+
"grad_norm": 1.1672072410583496,
|
| 1204 |
+
"learning_rate": 5.775002676297601e-05,
|
| 1205 |
+
"loss": 2.8085,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.15,
|
| 1210 |
+
"grad_norm": 1.1098130941390991,
|
| 1211 |
+
"learning_rate": 5.698137042863983e-05,
|
| 1212 |
+
"loss": 2.839,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.16,
|
| 1217 |
+
"grad_norm": 1.1094157695770264,
|
| 1218 |
+
"learning_rate": 5.621471747168903e-05,
|
| 1219 |
+
"loss": 2.7956,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.17,
|
| 1224 |
+
"grad_norm": 1.0807161331176758,
|
| 1225 |
+
"learning_rate": 5.545015313141388e-05,
|
| 1226 |
+
"loss": 2.8237,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.17,
|
| 1231 |
+
"grad_norm": 1.1589038372039795,
|
| 1232 |
+
"learning_rate": 5.468776241488462e-05,
|
| 1233 |
+
"loss": 2.8045,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.18,
|
| 1238 |
+
"grad_norm": 1.176377773284912,
|
| 1239 |
+
"learning_rate": 5.393521995403198e-05,
|
| 1240 |
+
"loss": 2.828,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.19,
|
| 1245 |
+
"grad_norm": 1.1362498998641968,
|
| 1246 |
+
"learning_rate": 5.3177406683561924e-05,
|
| 1247 |
+
"loss": 2.8048,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.19,
|
| 1252 |
+
"grad_norm": 1.1269686222076416,
|
| 1253 |
+
"learning_rate": 5.2422019729131476e-05,
|
| 1254 |
+
"loss": 2.8244,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.2,
|
| 1259 |
+
"grad_norm": 1.1583458185195923,
|
| 1260 |
+
"learning_rate": 5.166914307743539e-05,
|
| 1261 |
+
"loss": 2.7744,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.21,
|
| 1266 |
+
"grad_norm": 1.1909713745117188,
|
| 1267 |
+
"learning_rate": 5.091886043606376e-05,
|
| 1268 |
+
"loss": 2.8013,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.22,
|
| 1273 |
+
"grad_norm": 1.162377119064331,
|
| 1274 |
+
"learning_rate": 5.0171255224195034e-05,
|
| 1275 |
+
"loss": 2.7849,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.22,
|
| 1280 |
+
"grad_norm": 1.1313903331756592,
|
| 1281 |
+
"learning_rate": 4.943384507302191e-05,
|
| 1282 |
+
"loss": 2.7909,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.23,
|
| 1287 |
+
"grad_norm": 1.144729495048523,
|
| 1288 |
+
"learning_rate": 4.8691814935149756e-05,
|
| 1289 |
+
"loss": 2.8118,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.24,
|
| 1294 |
+
"grad_norm": 1.1973861455917358,
|
| 1295 |
+
"learning_rate": 4.795270983787518e-05,
|
| 1296 |
+
"loss": 2.7996,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.24,
|
| 1301 |
+
"grad_norm": 1.1937367916107178,
|
| 1302 |
+
"learning_rate": 4.721661195761631e-05,
|
| 1303 |
+
"loss": 2.8093,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.25,
|
| 1308 |
+
"grad_norm": 1.1672101020812988,
|
| 1309 |
+
"learning_rate": 4.648360313643783e-05,
|
| 1310 |
+
"loss": 2.7831,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.26,
|
| 1315 |
+
"grad_norm": 1.172425627708435,
|
| 1316 |
+
"learning_rate": 4.5753764872951475e-05,
|
| 1317 |
+
"loss": 2.8096,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.26,
|
| 1322 |
+
"grad_norm": 1.1739559173583984,
|
| 1323 |
+
"learning_rate": 4.502717831325476e-05,
|
| 1324 |
+
"loss": 2.7785,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.27,
|
| 1329 |
+
"grad_norm": 1.1311064958572388,
|
| 1330 |
+
"learning_rate": 4.430392424190881e-05,
|
| 1331 |
+
"loss": 2.7669,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.28,
|
| 1336 |
+
"grad_norm": 1.1691126823425293,
|
| 1337 |
+
"learning_rate": 4.35840830729565e-05,
|
| 1338 |
+
"loss": 2.777,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.28,
|
| 1343 |
+
"grad_norm": 1.1401257514953613,
|
| 1344 |
+
"learning_rate": 4.2867734840981604e-05,
|
| 1345 |
+
"loss": 2.7905,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.29,
|
| 1350 |
+
"grad_norm": 1.1707675457000732,
|
| 1351 |
+
"learning_rate": 4.215495919221047e-05,
|
| 1352 |
+
"loss": 2.7827,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.3,
|
| 1357 |
+
"grad_norm": 1.1320890188217163,
|
| 1358 |
+
"learning_rate": 4.144583537565638e-05,
|
| 1359 |
+
"loss": 2.7917,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.3,
|
| 1364 |
+
"grad_norm": 1.1413663625717163,
|
| 1365 |
+
"learning_rate": 4.074044223430865e-05,
|
| 1366 |
+
"loss": 2.8054,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.31,
|
| 1371 |
+
"grad_norm": 1.1374040842056274,
|
| 1372 |
+
"learning_rate": 4.0038858196366326e-05,
|
| 1373 |
+
"loss": 2.7656,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.32,
|
| 1378 |
+
"grad_norm": 1.132629632949829,
|
| 1379 |
+
"learning_rate": 3.934116126651836e-05,
|
| 1380 |
+
"loss": 2.7975,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.32,
|
| 1385 |
+
"grad_norm": 1.1647270917892456,
|
| 1386 |
+
"learning_rate": 3.864742901727086e-05,
|
| 1387 |
+
"loss": 2.7712,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.33,
|
| 1392 |
+
"grad_norm": 1.1258630752563477,
|
| 1393 |
+
"learning_rate": 3.7957738580322054e-05,
|
| 1394 |
+
"loss": 2.7728,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.34,
|
| 1399 |
+
"grad_norm": 1.1994359493255615,
|
| 1400 |
+
"learning_rate": 3.727216663798669e-05,
|
| 1401 |
+
"loss": 2.7834,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.34,
|
| 1406 |
+
"grad_norm": 1.1636048555374146,
|
| 1407 |
+
"learning_rate": 3.659078941467023e-05,
|
| 1408 |
+
"loss": 2.7697,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.35,
|
| 1413 |
+
"grad_norm": 1.183493733406067,
|
| 1414 |
+
"learning_rate": 3.5913682668393765e-05,
|
| 1415 |
+
"loss": 2.7718,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.36,
|
| 1420 |
+
"grad_norm": 1.0956398248672485,
|
| 1421 |
+
"learning_rate": 3.524092168237118e-05,
|
| 1422 |
+
"loss": 2.7493,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.36,
|
| 1427 |
+
"grad_norm": 1.1194324493408203,
|
| 1428 |
+
"learning_rate": 3.4572581256638674e-05,
|
| 1429 |
+
"loss": 2.7813,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.37,
|
| 1434 |
+
"grad_norm": 1.1716376543045044,
|
| 1435 |
+
"learning_rate": 3.3908735699738456e-05,
|
| 1436 |
+
"loss": 2.7666,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.38,
|
| 1441 |
+
"grad_norm": 1.111285924911499,
|
| 1442 |
+
"learning_rate": 3.3249458820456536e-05,
|
| 1443 |
+
"loss": 2.7619,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.38,
|
| 1448 |
+
"grad_norm": 1.1338977813720703,
|
| 1449 |
+
"learning_rate": 3.2601347051411326e-05,
|
| 1450 |
+
"loss": 2.7741,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.39,
|
| 1455 |
+
"grad_norm": 1.1371550559997559,
|
| 1456 |
+
"learning_rate": 3.1951379407523905e-05,
|
| 1457 |
+
"loss": 2.778,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.4,
|
| 1462 |
+
"grad_norm": 1.1465747356414795,
|
| 1463 |
+
"learning_rate": 3.130619806731321e-05,
|
| 1464 |
+
"loss": 2.7631,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.4,
|
| 1469 |
+
"grad_norm": 1.1635748147964478,
|
| 1470 |
+
"learning_rate": 3.066587476440921e-05,
|
| 1471 |
+
"loss": 2.7729,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.41,
|
| 1476 |
+
"grad_norm": 1.1626834869384766,
|
| 1477 |
+
"learning_rate": 3.0030480692307497e-05,
|
| 1478 |
+
"loss": 2.7743,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.42,
|
| 1483 |
+
"grad_norm": 1.1853559017181396,
|
| 1484 |
+
"learning_rate": 2.9400086496453632e-05,
|
| 1485 |
+
"loss": 2.7283,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.42,
|
| 1490 |
+
"grad_norm": 1.1449042558670044,
|
| 1491 |
+
"learning_rate": 2.877476226638858e-05,
|
| 1492 |
+
"loss": 2.7604,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.43,
|
| 1497 |
+
"grad_norm": 1.1443166732788086,
|
| 1498 |
+
"learning_rate": 2.815457752795596e-05,
|
| 1499 |
+
"loss": 2.7578,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.44,
|
| 1504 |
+
"grad_norm": 1.1765707731246948,
|
| 1505 |
+
"learning_rate": 2.7539601235571768e-05,
|
| 1506 |
+
"loss": 2.7546,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.44,
|
| 1511 |
+
"grad_norm": 1.1438168287277222,
|
| 1512 |
+
"learning_rate": 2.6929901764557916e-05,
|
| 1513 |
+
"loss": 2.7438,
|
| 1514 |
+
"step": 21500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.45,
|
| 1518 |
+
"grad_norm": 1.1334246397018433,
|
| 1519 |
+
"learning_rate": 2.6325546903539884e-05,
|
| 1520 |
+
"loss": 2.7395,
|
| 1521 |
+
"step": 21600
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.46,
|
| 1525 |
+
"grad_norm": 1.2056312561035156,
|
| 1526 |
+
"learning_rate": 2.5726603846909792e-05,
|
| 1527 |
+
"loss": 2.7291,
|
| 1528 |
+
"step": 21700
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.46,
|
| 1532 |
+
"grad_norm": 1.1193504333496094,
|
| 1533 |
+
"learning_rate": 2.5133139187355565e-05,
|
| 1534 |
+
"loss": 2.7419,
|
| 1535 |
+
"step": 21800
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.47,
|
| 1539 |
+
"grad_norm": 1.1401234865188599,
|
| 1540 |
+
"learning_rate": 2.4545218908456703e-05,
|
| 1541 |
+
"loss": 2.7505,
|
| 1542 |
+
"step": 21900
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.48,
|
| 1546 |
+
"grad_norm": 1.1264578104019165,
|
| 1547 |
+
"learning_rate": 2.3962908377348237e-05,
|
| 1548 |
+
"loss": 2.7381,
|
| 1549 |
+
"step": 22000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.48,
|
| 1553 |
+
"grad_norm": 1.132291316986084,
|
| 1554 |
+
"learning_rate": 2.3386272337452726e-05,
|
| 1555 |
+
"loss": 2.7349,
|
| 1556 |
+
"step": 22100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.49,
|
| 1560 |
+
"grad_norm": 1.1569644212722778,
|
| 1561 |
+
"learning_rate": 2.2815374901282095e-05,
|
| 1562 |
+
"loss": 2.7248,
|
| 1563 |
+
"step": 22200
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.5,
|
| 1567 |
+
"grad_norm": 1.141642451286316,
|
| 1568 |
+
"learning_rate": 2.2250279543309062e-05,
|
| 1569 |
+
"loss": 2.7441,
|
| 1570 |
+
"step": 22300
|
| 1571 |
+
}
|
| 1572 |
+
],
|
| 1573 |
+
"logging_steps": 100,
|
| 1574 |
+
"max_steps": 29794,
|
| 1575 |
+
"num_input_tokens_seen": 0,
|
| 1576 |
+
"num_train_epochs": 2,
|
| 1577 |
+
"save_steps": 7449,
|
| 1578 |
+
"total_flos": 5.809825965617971e+16,
|
| 1579 |
+
"train_batch_size": 16,
|
| 1580 |
+
"trial_name": null,
|
| 1581 |
+
"trial_params": null
|
| 1582 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cfaa872c80b43aebdda330927cb4746739fb3aa28bb0f42bc733d38ca7c5209
|
| 3 |
+
size 4475
|