Model save
Browse files- README.md +62 -0
- all_results.json +9 -0
- config.json +125 -0
- generation_config.json +6 -0
- model.safetensors +3 -0
- runs/Aug29_13-40-25_xe8545-a100-12/events.out.tfevents.1724932665.xe8545-a100-12.524617.0 +3 -0
- special_tokens_map.json +24 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +43 -0
- train_results.json +9 -0
- trainer_state.json +3241 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: other
|
| 4 |
+
base_model: apple/OpenELM-1_1B
|
| 5 |
+
tags:
|
| 6 |
+
- trl
|
| 7 |
+
- sft
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: OpenELM-1_1B-SFT-1
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# OpenELM-1_1B-SFT-1
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [apple/OpenELM-1_1B](https://huggingface.co/apple/OpenELM-1_1B) on the generator dataset.
|
| 22 |
+
|
| 23 |
+
## Model description
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Intended uses & limitations
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training and evaluation data
|
| 32 |
+
|
| 33 |
+
More information needed
|
| 34 |
+
|
| 35 |
+
## Training procedure
|
| 36 |
+
|
| 37 |
+
### Training hyperparameters
|
| 38 |
+
|
| 39 |
+
The following hyperparameters were used during training:
|
| 40 |
+
- learning_rate: 2e-05
|
| 41 |
+
- train_batch_size: 16
|
| 42 |
+
- eval_batch_size: 8
|
| 43 |
+
- seed: 42
|
| 44 |
+
- distributed_type: multi-GPU
|
| 45 |
+
- num_devices: 4
|
| 46 |
+
- total_train_batch_size: 64
|
| 47 |
+
- total_eval_batch_size: 32
|
| 48 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 49 |
+
- lr_scheduler_type: cosine
|
| 50 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 51 |
+
- num_epochs: 1
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
### Framework versions
|
| 58 |
+
|
| 59 |
+
- Transformers 4.44.2
|
| 60 |
+
- Pytorch 2.3.0
|
| 61 |
+
- Datasets 2.21.0
|
| 62 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 216025446481920.0,
|
| 4 |
+
"train_loss": 1.2513469975087603,
|
| 5 |
+
"train_runtime": 7749.9665,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 18.846,
|
| 8 |
+
"train_steps_per_second": 0.295
|
| 9 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,125 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "apple/OpenELM-1_1B",
|
| 3 |
+
"activation_fn_name": "swish",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"OpenELMForCausalLM"
|
| 6 |
+
],
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "apple/OpenELM-1_1B--configuration_openelm.OpenELMConfig",
|
| 9 |
+
"AutoModelForCausalLM": "apple/OpenELM-1_1B--modeling_openelm.OpenELMForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 1,
|
| 12 |
+
"eos_token_id": 2,
|
| 13 |
+
"ffn_dim_divisor": 256,
|
| 14 |
+
"ffn_multipliers": [
|
| 15 |
+
0.5,
|
| 16 |
+
0.63,
|
| 17 |
+
0.76,
|
| 18 |
+
0.89,
|
| 19 |
+
1.02,
|
| 20 |
+
1.15,
|
| 21 |
+
1.28,
|
| 22 |
+
1.41,
|
| 23 |
+
1.54,
|
| 24 |
+
1.67,
|
| 25 |
+
1.8,
|
| 26 |
+
1.93,
|
| 27 |
+
2.06,
|
| 28 |
+
2.19,
|
| 29 |
+
2.31,
|
| 30 |
+
2.44,
|
| 31 |
+
2.57,
|
| 32 |
+
2.7,
|
| 33 |
+
2.83,
|
| 34 |
+
2.96,
|
| 35 |
+
3.09,
|
| 36 |
+
3.22,
|
| 37 |
+
3.35,
|
| 38 |
+
3.48,
|
| 39 |
+
3.61,
|
| 40 |
+
3.74,
|
| 41 |
+
3.87,
|
| 42 |
+
4.0
|
| 43 |
+
],
|
| 44 |
+
"ffn_with_glu": true,
|
| 45 |
+
"head_dim": 64,
|
| 46 |
+
"initializer_range": 0.02,
|
| 47 |
+
"max_context_length": 2048,
|
| 48 |
+
"model_dim": 2048,
|
| 49 |
+
"model_type": "openelm",
|
| 50 |
+
"normalization_layer_name": "rms_norm",
|
| 51 |
+
"normalize_qk_projections": true,
|
| 52 |
+
"num_gqa_groups": 4,
|
| 53 |
+
"num_kv_heads": [
|
| 54 |
+
4,
|
| 55 |
+
4,
|
| 56 |
+
4,
|
| 57 |
+
5,
|
| 58 |
+
5,
|
| 59 |
+
5,
|
| 60 |
+
5,
|
| 61 |
+
5,
|
| 62 |
+
5,
|
| 63 |
+
5,
|
| 64 |
+
6,
|
| 65 |
+
6,
|
| 66 |
+
6,
|
| 67 |
+
6,
|
| 68 |
+
6,
|
| 69 |
+
6,
|
| 70 |
+
6,
|
| 71 |
+
6,
|
| 72 |
+
7,
|
| 73 |
+
7,
|
| 74 |
+
7,
|
| 75 |
+
7,
|
| 76 |
+
7,
|
| 77 |
+
7,
|
| 78 |
+
8,
|
| 79 |
+
8,
|
| 80 |
+
8,
|
| 81 |
+
8
|
| 82 |
+
],
|
| 83 |
+
"num_query_heads": [
|
| 84 |
+
16,
|
| 85 |
+
16,
|
| 86 |
+
16,
|
| 87 |
+
20,
|
| 88 |
+
20,
|
| 89 |
+
20,
|
| 90 |
+
20,
|
| 91 |
+
20,
|
| 92 |
+
20,
|
| 93 |
+
20,
|
| 94 |
+
24,
|
| 95 |
+
24,
|
| 96 |
+
24,
|
| 97 |
+
24,
|
| 98 |
+
24,
|
| 99 |
+
24,
|
| 100 |
+
24,
|
| 101 |
+
24,
|
| 102 |
+
28,
|
| 103 |
+
28,
|
| 104 |
+
28,
|
| 105 |
+
28,
|
| 106 |
+
28,
|
| 107 |
+
28,
|
| 108 |
+
32,
|
| 109 |
+
32,
|
| 110 |
+
32,
|
| 111 |
+
32
|
| 112 |
+
],
|
| 113 |
+
"num_transformer_layers": 28,
|
| 114 |
+
"qkv_multipliers": [
|
| 115 |
+
0.5,
|
| 116 |
+
1.0
|
| 117 |
+
],
|
| 118 |
+
"rope_freq_constant": 10000,
|
| 119 |
+
"rope_max_length": 4096,
|
| 120 |
+
"share_input_output_layers": true,
|
| 121 |
+
"torch_dtype": "bfloat16",
|
| 122 |
+
"transformers_version": "4.44.2",
|
| 123 |
+
"use_cache": false,
|
| 124 |
+
"vocab_size": 32000
|
| 125 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.44.2"
|
| 6 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0ff594c413384588fb87d2d6833960f2fce46099f0306d5c475dff4b1fef44a
|
| 3 |
+
size 2159808696
|
runs/Aug29_13-40-25_xe8545-a100-12/events.out.tfevents.1724932665.xe8545-a100-12.524617.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:68783fba98eedc0ec5545e5c0de113b8fd7bcb19aa29d2383e91533b2798d240
|
| 3 |
+
size 103184
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "</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.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"bos_token": "<s>",
|
| 32 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 33 |
+
"clean_up_tokenization_spaces": false,
|
| 34 |
+
"eos_token": "</s>",
|
| 35 |
+
"legacy": false,
|
| 36 |
+
"model_max_length": 2048,
|
| 37 |
+
"pad_token": "</s>",
|
| 38 |
+
"padding_side": "right",
|
| 39 |
+
"sp_model_kwargs": {},
|
| 40 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 41 |
+
"unk_token": "<unk>",
|
| 42 |
+
"use_default_system_prompt": false
|
| 43 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 216025446481920.0,
|
| 4 |
+
"train_loss": 1.2513469975087603,
|
| 5 |
+
"train_runtime": 7749.9665,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 18.846,
|
| 8 |
+
"train_steps_per_second": 0.295
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3241 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2283,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0004380201489268506,
|
| 13 |
+
"grad_norm": 244.0565657784503,
|
| 14 |
+
"learning_rate": 8.733624454148472e-08,
|
| 15 |
+
"loss": 8.7812,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.002190100744634253,
|
| 20 |
+
"grad_norm": 242.08745729404998,
|
| 21 |
+
"learning_rate": 4.366812227074236e-07,
|
| 22 |
+
"loss": 8.8047,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.004380201489268506,
|
| 27 |
+
"grad_norm": 215.31824760158125,
|
| 28 |
+
"learning_rate": 8.733624454148472e-07,
|
| 29 |
+
"loss": 8.6969,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.006570302233902759,
|
| 34 |
+
"grad_norm": 124.79925946857198,
|
| 35 |
+
"learning_rate": 1.3100436681222709e-06,
|
| 36 |
+
"loss": 8.3187,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.008760402978537012,
|
| 41 |
+
"grad_norm": 95.62100949931548,
|
| 42 |
+
"learning_rate": 1.7467248908296944e-06,
|
| 43 |
+
"loss": 7.6672,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.010950503723171266,
|
| 48 |
+
"grad_norm": 56.19522998010826,
|
| 49 |
+
"learning_rate": 2.183406113537118e-06,
|
| 50 |
+
"loss": 6.7922,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.013140604467805518,
|
| 55 |
+
"grad_norm": 34.7212601740852,
|
| 56 |
+
"learning_rate": 2.6200873362445417e-06,
|
| 57 |
+
"loss": 5.9031,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.015330705212439772,
|
| 62 |
+
"grad_norm": 30.849397382141074,
|
| 63 |
+
"learning_rate": 3.0567685589519653e-06,
|
| 64 |
+
"loss": 5.1156,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.017520805957074025,
|
| 69 |
+
"grad_norm": 16.184875412033424,
|
| 70 |
+
"learning_rate": 3.493449781659389e-06,
|
| 71 |
+
"loss": 3.9531,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.01971090670170828,
|
| 76 |
+
"grad_norm": 13.212043262995175,
|
| 77 |
+
"learning_rate": 3.930131004366812e-06,
|
| 78 |
+
"loss": 3.1414,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.021901007446342532,
|
| 83 |
+
"grad_norm": 6.575354808182119,
|
| 84 |
+
"learning_rate": 4.366812227074236e-06,
|
| 85 |
+
"loss": 2.5,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.024091108190976786,
|
| 90 |
+
"grad_norm": 2.5861843997481873,
|
| 91 |
+
"learning_rate": 4.80349344978166e-06,
|
| 92 |
+
"loss": 1.9785,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.026281208935611037,
|
| 97 |
+
"grad_norm": 1.566359722353574,
|
| 98 |
+
"learning_rate": 5.2401746724890834e-06,
|
| 99 |
+
"loss": 1.7488,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.02847130968024529,
|
| 104 |
+
"grad_norm": 0.9077919251278966,
|
| 105 |
+
"learning_rate": 5.676855895196507e-06,
|
| 106 |
+
"loss": 1.6184,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.030661410424879545,
|
| 111 |
+
"grad_norm": 0.6573333175178419,
|
| 112 |
+
"learning_rate": 6.1135371179039305e-06,
|
| 113 |
+
"loss": 1.5391,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0328515111695138,
|
| 118 |
+
"grad_norm": 0.5238887296216804,
|
| 119 |
+
"learning_rate": 6.550218340611354e-06,
|
| 120 |
+
"loss": 1.4633,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.03504161191414805,
|
| 125 |
+
"grad_norm": 0.4821823564791243,
|
| 126 |
+
"learning_rate": 6.986899563318778e-06,
|
| 127 |
+
"loss": 1.4199,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.03723171265878231,
|
| 132 |
+
"grad_norm": 0.4129490577402294,
|
| 133 |
+
"learning_rate": 7.423580786026201e-06,
|
| 134 |
+
"loss": 1.4109,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.03942181340341656,
|
| 139 |
+
"grad_norm": 0.40472709691340175,
|
| 140 |
+
"learning_rate": 7.860262008733624e-06,
|
| 141 |
+
"loss": 1.4168,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.04161191414805081,
|
| 146 |
+
"grad_norm": 0.3808604144712055,
|
| 147 |
+
"learning_rate": 8.296943231441049e-06,
|
| 148 |
+
"loss": 1.3668,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.043802014892685065,
|
| 153 |
+
"grad_norm": 0.367625493325619,
|
| 154 |
+
"learning_rate": 8.733624454148473e-06,
|
| 155 |
+
"loss": 1.3066,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.045992115637319315,
|
| 160 |
+
"grad_norm": 0.359816415568165,
|
| 161 |
+
"learning_rate": 9.170305676855896e-06,
|
| 162 |
+
"loss": 1.3449,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.04818221638195357,
|
| 167 |
+
"grad_norm": 0.341837652413913,
|
| 168 |
+
"learning_rate": 9.60698689956332e-06,
|
| 169 |
+
"loss": 1.3273,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05037231712658782,
|
| 174 |
+
"grad_norm": 0.34301552589118167,
|
| 175 |
+
"learning_rate": 1.0043668122270742e-05,
|
| 176 |
+
"loss": 1.2949,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.052562417871222074,
|
| 181 |
+
"grad_norm": 0.3628435999185446,
|
| 182 |
+
"learning_rate": 1.0480349344978167e-05,
|
| 183 |
+
"loss": 1.2871,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.05475251861585633,
|
| 188 |
+
"grad_norm": 0.3679166427745723,
|
| 189 |
+
"learning_rate": 1.0917030567685592e-05,
|
| 190 |
+
"loss": 1.3207,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.05694261936049058,
|
| 195 |
+
"grad_norm": 0.34368277634691846,
|
| 196 |
+
"learning_rate": 1.1353711790393014e-05,
|
| 197 |
+
"loss": 1.2645,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.05913272010512484,
|
| 202 |
+
"grad_norm": 0.3551051401296533,
|
| 203 |
+
"learning_rate": 1.179039301310044e-05,
|
| 204 |
+
"loss": 1.2707,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.06132282084975909,
|
| 209 |
+
"grad_norm": 0.334086491935665,
|
| 210 |
+
"learning_rate": 1.2227074235807861e-05,
|
| 211 |
+
"loss": 1.2742,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.06351292159439334,
|
| 216 |
+
"grad_norm": 0.3344453192406718,
|
| 217 |
+
"learning_rate": 1.2663755458515286e-05,
|
| 218 |
+
"loss": 1.2441,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.0657030223390276,
|
| 223 |
+
"grad_norm": 0.3486113854746094,
|
| 224 |
+
"learning_rate": 1.3100436681222708e-05,
|
| 225 |
+
"loss": 1.2934,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.06789312308366185,
|
| 230 |
+
"grad_norm": 0.3450373622328735,
|
| 231 |
+
"learning_rate": 1.3537117903930132e-05,
|
| 232 |
+
"loss": 1.2473,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.0700832238282961,
|
| 237 |
+
"grad_norm": 0.3254863200637601,
|
| 238 |
+
"learning_rate": 1.3973799126637555e-05,
|
| 239 |
+
"loss": 1.2613,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.07227332457293036,
|
| 244 |
+
"grad_norm": 0.3206618735664889,
|
| 245 |
+
"learning_rate": 1.4410480349344979e-05,
|
| 246 |
+
"loss": 1.275,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.07446342531756461,
|
| 251 |
+
"grad_norm": 0.36155580225009337,
|
| 252 |
+
"learning_rate": 1.4847161572052402e-05,
|
| 253 |
+
"loss": 1.2379,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.07665352606219886,
|
| 258 |
+
"grad_norm": 0.331500878464258,
|
| 259 |
+
"learning_rate": 1.5283842794759826e-05,
|
| 260 |
+
"loss": 1.234,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.07884362680683311,
|
| 265 |
+
"grad_norm": 0.35558769100827337,
|
| 266 |
+
"learning_rate": 1.5720524017467248e-05,
|
| 267 |
+
"loss": 1.2641,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.08103372755146737,
|
| 272 |
+
"grad_norm": 0.3514292771304392,
|
| 273 |
+
"learning_rate": 1.6157205240174673e-05,
|
| 274 |
+
"loss": 1.2258,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.08322382829610162,
|
| 279 |
+
"grad_norm": 0.3158735252448049,
|
| 280 |
+
"learning_rate": 1.6593886462882098e-05,
|
| 281 |
+
"loss": 1.2414,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.08541392904073587,
|
| 286 |
+
"grad_norm": 0.3332987501140752,
|
| 287 |
+
"learning_rate": 1.703056768558952e-05,
|
| 288 |
+
"loss": 1.2238,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.08760402978537013,
|
| 293 |
+
"grad_norm": 0.3306118702879735,
|
| 294 |
+
"learning_rate": 1.7467248908296945e-05,
|
| 295 |
+
"loss": 1.2316,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.08979413053000437,
|
| 300 |
+
"grad_norm": 0.3189870404362557,
|
| 301 |
+
"learning_rate": 1.7903930131004367e-05,
|
| 302 |
+
"loss": 1.2355,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.09198423127463863,
|
| 307 |
+
"grad_norm": 0.3015408510885377,
|
| 308 |
+
"learning_rate": 1.8340611353711792e-05,
|
| 309 |
+
"loss": 1.2188,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.09417433201927289,
|
| 314 |
+
"grad_norm": 0.3227399112930752,
|
| 315 |
+
"learning_rate": 1.8777292576419214e-05,
|
| 316 |
+
"loss": 1.2164,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.09636443276390715,
|
| 321 |
+
"grad_norm": 0.3316376300779914,
|
| 322 |
+
"learning_rate": 1.921397379912664e-05,
|
| 323 |
+
"loss": 1.2457,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.09855453350854139,
|
| 328 |
+
"grad_norm": 0.31785017269226773,
|
| 329 |
+
"learning_rate": 1.965065502183406e-05,
|
| 330 |
+
"loss": 1.1949,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.10074463425317565,
|
| 335 |
+
"grad_norm": 0.3305460477093586,
|
| 336 |
+
"learning_rate": 1.9999988303153612e-05,
|
| 337 |
+
"loss": 1.2184,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1029347349978099,
|
| 342 |
+
"grad_norm": 0.32274702495374813,
|
| 343 |
+
"learning_rate": 1.9999578916403086e-05,
|
| 344 |
+
"loss": 1.2242,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.10512483574244415,
|
| 349 |
+
"grad_norm": 0.3309888084620984,
|
| 350 |
+
"learning_rate": 1.999858471469601e-05,
|
| 351 |
+
"loss": 1.1996,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.1073149364870784,
|
| 356 |
+
"grad_norm": 0.31589160373633207,
|
| 357 |
+
"learning_rate": 1.9997005756177228e-05,
|
| 358 |
+
"loss": 1.1754,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.10950503723171266,
|
| 363 |
+
"grad_norm": 0.3245071336888696,
|
| 364 |
+
"learning_rate": 1.9994842133190485e-05,
|
| 365 |
+
"loss": 1.2059,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.1116951379763469,
|
| 370 |
+
"grad_norm": 0.34753890904189133,
|
| 371 |
+
"learning_rate": 1.999209397227302e-05,
|
| 372 |
+
"loss": 1.2012,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.11388523872098116,
|
| 377 |
+
"grad_norm": 0.30057443535443235,
|
| 378 |
+
"learning_rate": 1.9988761434148164e-05,
|
| 379 |
+
"loss": 1.2281,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.11607533946561542,
|
| 384 |
+
"grad_norm": 0.30036747933816504,
|
| 385 |
+
"learning_rate": 1.998484471371593e-05,
|
| 386 |
+
"loss": 1.1785,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.11826544021024968,
|
| 391 |
+
"grad_norm": 0.32539575218814926,
|
| 392 |
+
"learning_rate": 1.9980344040041636e-05,
|
| 393 |
+
"loss": 1.1742,
|
| 394 |
+
"step": 270
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.12045554095488392,
|
| 398 |
+
"grad_norm": 0.31075948876414167,
|
| 399 |
+
"learning_rate": 1.99752596763425e-05,
|
| 400 |
+
"loss": 1.209,
|
| 401 |
+
"step": 275
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.12264564169951818,
|
| 405 |
+
"grad_norm": 0.2948383024933128,
|
| 406 |
+
"learning_rate": 1.9969591919972226e-05,
|
| 407 |
+
"loss": 1.2059,
|
| 408 |
+
"step": 280
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.12483574244415244,
|
| 412 |
+
"grad_norm": 0.30809860203285494,
|
| 413 |
+
"learning_rate": 1.9963341102403652e-05,
|
| 414 |
+
"loss": 1.1996,
|
| 415 |
+
"step": 285
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.12702584318878668,
|
| 419 |
+
"grad_norm": 0.3063101893237704,
|
| 420 |
+
"learning_rate": 1.995650758920932e-05,
|
| 421 |
+
"loss": 1.2,
|
| 422 |
+
"step": 290
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.12921594393342092,
|
| 426 |
+
"grad_norm": 0.29266235209581604,
|
| 427 |
+
"learning_rate": 1.9949091780040143e-05,
|
| 428 |
+
"loss": 1.173,
|
| 429 |
+
"step": 295
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.1314060446780552,
|
| 433 |
+
"grad_norm": 0.3236695292638355,
|
| 434 |
+
"learning_rate": 1.9941094108601985e-05,
|
| 435 |
+
"loss": 1.1832,
|
| 436 |
+
"step": 300
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.13359614542268944,
|
| 440 |
+
"grad_norm": 0.29386460049329405,
|
| 441 |
+
"learning_rate": 1.9932515042630335e-05,
|
| 442 |
+
"loss": 1.1746,
|
| 443 |
+
"step": 305
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.1357862461673237,
|
| 447 |
+
"grad_norm": 0.3304096067847152,
|
| 448 |
+
"learning_rate": 1.9923355083862933e-05,
|
| 449 |
+
"loss": 1.2004,
|
| 450 |
+
"step": 310
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.13797634691195795,
|
| 454 |
+
"grad_norm": 0.297618875850005,
|
| 455 |
+
"learning_rate": 1.9913614768010418e-05,
|
| 456 |
+
"loss": 1.1656,
|
| 457 |
+
"step": 315
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.1401664476565922,
|
| 461 |
+
"grad_norm": 0.3453392768634244,
|
| 462 |
+
"learning_rate": 1.9903294664725023e-05,
|
| 463 |
+
"loss": 1.1473,
|
| 464 |
+
"step": 320
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.14235654840122647,
|
| 468 |
+
"grad_norm": 0.2952934857480258,
|
| 469 |
+
"learning_rate": 1.989239537756723e-05,
|
| 470 |
+
"loss": 1.1898,
|
| 471 |
+
"step": 325
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.1445466491458607,
|
| 475 |
+
"grad_norm": 0.3096490330878614,
|
| 476 |
+
"learning_rate": 1.9880917543970503e-05,
|
| 477 |
+
"loss": 1.1727,
|
| 478 |
+
"step": 330
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.14673674989049496,
|
| 482 |
+
"grad_norm": 0.305522689793203,
|
| 483 |
+
"learning_rate": 1.986886183520398e-05,
|
| 484 |
+
"loss": 1.166,
|
| 485 |
+
"step": 335
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.14892685063512923,
|
| 489 |
+
"grad_norm": 0.2977399891862773,
|
| 490 |
+
"learning_rate": 1.985622895633323e-05,
|
| 491 |
+
"loss": 1.1605,
|
| 492 |
+
"step": 340
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.15111695137976347,
|
| 496 |
+
"grad_norm": 0.3080937705338333,
|
| 497 |
+
"learning_rate": 1.9843019646179014e-05,
|
| 498 |
+
"loss": 1.1785,
|
| 499 |
+
"step": 345
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.1533070521243977,
|
| 503 |
+
"grad_norm": 0.28864764946025717,
|
| 504 |
+
"learning_rate": 1.982923467727408e-05,
|
| 505 |
+
"loss": 1.1797,
|
| 506 |
+
"step": 350
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.15549715286903198,
|
| 510 |
+
"grad_norm": 0.29002073593801136,
|
| 511 |
+
"learning_rate": 1.981487485581797e-05,
|
| 512 |
+
"loss": 1.1676,
|
| 513 |
+
"step": 355
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.15768725361366623,
|
| 517 |
+
"grad_norm": 0.3018457130707611,
|
| 518 |
+
"learning_rate": 1.9799941021629886e-05,
|
| 519 |
+
"loss": 1.1738,
|
| 520 |
+
"step": 360
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.15987735435830047,
|
| 524 |
+
"grad_norm": 0.30507162972432783,
|
| 525 |
+
"learning_rate": 1.9784434048099565e-05,
|
| 526 |
+
"loss": 1.1668,
|
| 527 |
+
"step": 365
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.16206745510293474,
|
| 531 |
+
"grad_norm": 0.3049773910908397,
|
| 532 |
+
"learning_rate": 1.97683548421362e-05,
|
| 533 |
+
"loss": 1.1781,
|
| 534 |
+
"step": 370
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.164257555847569,
|
| 538 |
+
"grad_norm": 0.3024971125857499,
|
| 539 |
+
"learning_rate": 1.9751704344115402e-05,
|
| 540 |
+
"loss": 1.1625,
|
| 541 |
+
"step": 375
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.16644765659220323,
|
| 545 |
+
"grad_norm": 0.28148210956054226,
|
| 546 |
+
"learning_rate": 1.97344835278242e-05,
|
| 547 |
+
"loss": 1.1687,
|
| 548 |
+
"step": 380
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.1686377573368375,
|
| 552 |
+
"grad_norm": 0.2996144816775727,
|
| 553 |
+
"learning_rate": 1.97166934004041e-05,
|
| 554 |
+
"loss": 1.207,
|
| 555 |
+
"step": 385
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.17082785808147175,
|
| 559 |
+
"grad_norm": 0.29532940992965984,
|
| 560 |
+
"learning_rate": 1.9698335002292176e-05,
|
| 561 |
+
"loss": 1.1848,
|
| 562 |
+
"step": 390
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.173017958826106,
|
| 566 |
+
"grad_norm": 0.3001673093160351,
|
| 567 |
+
"learning_rate": 1.967940940716021e-05,
|
| 568 |
+
"loss": 1.1828,
|
| 569 |
+
"step": 395
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.17520805957074026,
|
| 573 |
+
"grad_norm": 0.30610198280704054,
|
| 574 |
+
"learning_rate": 1.9659917721851924e-05,
|
| 575 |
+
"loss": 1.1867,
|
| 576 |
+
"step": 400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.1773981603153745,
|
| 580 |
+
"grad_norm": 0.2925881036239608,
|
| 581 |
+
"learning_rate": 1.963986108631823e-05,
|
| 582 |
+
"loss": 1.1418,
|
| 583 |
+
"step": 405
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.17958826106000875,
|
| 587 |
+
"grad_norm": 0.29339092688050755,
|
| 588 |
+
"learning_rate": 1.961924067355056e-05,
|
| 589 |
+
"loss": 1.1637,
|
| 590 |
+
"step": 410
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.18177836180464302,
|
| 594 |
+
"grad_norm": 0.30407476923887133,
|
| 595 |
+
"learning_rate": 1.9598057689512277e-05,
|
| 596 |
+
"loss": 1.1508,
|
| 597 |
+
"step": 415
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.18396846254927726,
|
| 601 |
+
"grad_norm": 0.29998056902468595,
|
| 602 |
+
"learning_rate": 1.957631337306814e-05,
|
| 603 |
+
"loss": 1.1578,
|
| 604 |
+
"step": 420
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.18615856329391153,
|
| 608 |
+
"grad_norm": 0.3025021503047764,
|
| 609 |
+
"learning_rate": 1.9554008995911837e-05,
|
| 610 |
+
"loss": 1.1812,
|
| 611 |
+
"step": 425
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.18834866403854578,
|
| 615 |
+
"grad_norm": 0.2960918602914393,
|
| 616 |
+
"learning_rate": 1.953114586249164e-05,
|
| 617 |
+
"loss": 1.1793,
|
| 618 |
+
"step": 430
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.19053876478318002,
|
| 622 |
+
"grad_norm": 0.3196259778914729,
|
| 623 |
+
"learning_rate": 1.950772530993409e-05,
|
| 624 |
+
"loss": 1.1738,
|
| 625 |
+
"step": 435
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.1927288655278143,
|
| 629 |
+
"grad_norm": 0.2993685894942767,
|
| 630 |
+
"learning_rate": 1.9483748707965803e-05,
|
| 631 |
+
"loss": 1.1512,
|
| 632 |
+
"step": 440
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.19491896627244854,
|
| 636 |
+
"grad_norm": 0.3002942784736228,
|
| 637 |
+
"learning_rate": 1.945921745883337e-05,
|
| 638 |
+
"loss": 1.1625,
|
| 639 |
+
"step": 445
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.19710906701708278,
|
| 643 |
+
"grad_norm": 0.30887780802399506,
|
| 644 |
+
"learning_rate": 1.9434132997221347e-05,
|
| 645 |
+
"loss": 1.1719,
|
| 646 |
+
"step": 450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.19929916776171705,
|
| 650 |
+
"grad_norm": 0.29738335551869177,
|
| 651 |
+
"learning_rate": 1.9408496790168337e-05,
|
| 652 |
+
"loss": 1.1641,
|
| 653 |
+
"step": 455
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.2014892685063513,
|
| 657 |
+
"grad_norm": 0.30233929664046716,
|
| 658 |
+
"learning_rate": 1.9382310336981206e-05,
|
| 659 |
+
"loss": 1.1512,
|
| 660 |
+
"step": 460
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.20367936925098554,
|
| 664 |
+
"grad_norm": 0.29527032498064754,
|
| 665 |
+
"learning_rate": 1.935557516914739e-05,
|
| 666 |
+
"loss": 1.148,
|
| 667 |
+
"step": 465
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.2058694699956198,
|
| 671 |
+
"grad_norm": 0.29787746244856167,
|
| 672 |
+
"learning_rate": 1.932829285024533e-05,
|
| 673 |
+
"loss": 1.1516,
|
| 674 |
+
"step": 470
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.20805957074025405,
|
| 678 |
+
"grad_norm": 0.2913683096092213,
|
| 679 |
+
"learning_rate": 1.9300464975853032e-05,
|
| 680 |
+
"loss": 1.1344,
|
| 681 |
+
"step": 475
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.2102496714848883,
|
| 685 |
+
"grad_norm": 0.27500340802613793,
|
| 686 |
+
"learning_rate": 1.9272093173454735e-05,
|
| 687 |
+
"loss": 1.1574,
|
| 688 |
+
"step": 480
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.21243977222952257,
|
| 692 |
+
"grad_norm": 0.2812766184977934,
|
| 693 |
+
"learning_rate": 1.9243179102345753e-05,
|
| 694 |
+
"loss": 1.148,
|
| 695 |
+
"step": 485
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.2146298729741568,
|
| 699 |
+
"grad_norm": 0.2865109750359857,
|
| 700 |
+
"learning_rate": 1.9213724453535417e-05,
|
| 701 |
+
"loss": 1.1578,
|
| 702 |
+
"step": 490
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.21681997371879105,
|
| 706 |
+
"grad_norm": 0.29189523569878567,
|
| 707 |
+
"learning_rate": 1.9183730949648173e-05,
|
| 708 |
+
"loss": 1.1602,
|
| 709 |
+
"step": 495
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.21901007446342532,
|
| 713 |
+
"grad_norm": 0.29979676151962115,
|
| 714 |
+
"learning_rate": 1.9153200344822855e-05,
|
| 715 |
+
"loss": 1.1355,
|
| 716 |
+
"step": 500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.22120017520805957,
|
| 720 |
+
"grad_norm": 0.300831896633153,
|
| 721 |
+
"learning_rate": 1.912213442461009e-05,
|
| 722 |
+
"loss": 1.1391,
|
| 723 |
+
"step": 505
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.2233902759526938,
|
| 727 |
+
"grad_norm": 0.29049722290498076,
|
| 728 |
+
"learning_rate": 1.9090535005867853e-05,
|
| 729 |
+
"loss": 1.1727,
|
| 730 |
+
"step": 510
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.22558037669732808,
|
| 734 |
+
"grad_norm": 0.3104171548245787,
|
| 735 |
+
"learning_rate": 1.9058403936655235e-05,
|
| 736 |
+
"loss": 1.1551,
|
| 737 |
+
"step": 515
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.22777047744196233,
|
| 741 |
+
"grad_norm": 0.2944155570969255,
|
| 742 |
+
"learning_rate": 1.9025743096124346e-05,
|
| 743 |
+
"loss": 1.1527,
|
| 744 |
+
"step": 520
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.22996057818659657,
|
| 748 |
+
"grad_norm": 0.3169970998158994,
|
| 749 |
+
"learning_rate": 1.899255439441043e-05,
|
| 750 |
+
"loss": 1.158,
|
| 751 |
+
"step": 525
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.23215067893123084,
|
| 755 |
+
"grad_norm": 0.2842476129923102,
|
| 756 |
+
"learning_rate": 1.8958839772520125e-05,
|
| 757 |
+
"loss": 1.1758,
|
| 758 |
+
"step": 530
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.23434077967586509,
|
| 762 |
+
"grad_norm": 0.2855513642633395,
|
| 763 |
+
"learning_rate": 1.8924601202217977e-05,
|
| 764 |
+
"loss": 1.1357,
|
| 765 |
+
"step": 535
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.23653088042049936,
|
| 769 |
+
"grad_norm": 0.2848004802653552,
|
| 770 |
+
"learning_rate": 1.888984068591111e-05,
|
| 771 |
+
"loss": 1.1391,
|
| 772 |
+
"step": 540
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.2387209811651336,
|
| 776 |
+
"grad_norm": 0.28099993296693176,
|
| 777 |
+
"learning_rate": 1.8854560256532098e-05,
|
| 778 |
+
"loss": 1.1355,
|
| 779 |
+
"step": 545
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.24091108190976784,
|
| 783 |
+
"grad_norm": 0.3070333803041777,
|
| 784 |
+
"learning_rate": 1.8818761977420114e-05,
|
| 785 |
+
"loss": 1.143,
|
| 786 |
+
"step": 550
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.24310118265440211,
|
| 790 |
+
"grad_norm": 0.28305196192266907,
|
| 791 |
+
"learning_rate": 1.878244794220022e-05,
|
| 792 |
+
"loss": 1.1625,
|
| 793 |
+
"step": 555
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.24529128339903636,
|
| 797 |
+
"grad_norm": 0.29533684816936556,
|
| 798 |
+
"learning_rate": 1.874562027466093e-05,
|
| 799 |
+
"loss": 1.1695,
|
| 800 |
+
"step": 560
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2474813841436706,
|
| 804 |
+
"grad_norm": 0.29819831907768873,
|
| 805 |
+
"learning_rate": 1.8708281128630023e-05,
|
| 806 |
+
"loss": 1.1469,
|
| 807 |
+
"step": 565
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.24967148488830487,
|
| 811 |
+
"grad_norm": 0.28827778580342983,
|
| 812 |
+
"learning_rate": 1.8670432687848562e-05,
|
| 813 |
+
"loss": 1.149,
|
| 814 |
+
"step": 570
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.2518615856329391,
|
| 818 |
+
"grad_norm": 0.2865593620968661,
|
| 819 |
+
"learning_rate": 1.8632077165843174e-05,
|
| 820 |
+
"loss": 1.1246,
|
| 821 |
+
"step": 575
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.25405168637757336,
|
| 825 |
+
"grad_norm": 0.2873333172793282,
|
| 826 |
+
"learning_rate": 1.8593216805796612e-05,
|
| 827 |
+
"loss": 1.1402,
|
| 828 |
+
"step": 580
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.25624178712220763,
|
| 832 |
+
"grad_norm": 0.28696704010138063,
|
| 833 |
+
"learning_rate": 1.8553853880416555e-05,
|
| 834 |
+
"loss": 1.1293,
|
| 835 |
+
"step": 585
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.25843188786684185,
|
| 839 |
+
"grad_norm": 0.27933859864997074,
|
| 840 |
+
"learning_rate": 1.8513990691802686e-05,
|
| 841 |
+
"loss": 1.1691,
|
| 842 |
+
"step": 590
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.2606219886114761,
|
| 846 |
+
"grad_norm": 0.2870969603657859,
|
| 847 |
+
"learning_rate": 1.8473629571312073e-05,
|
| 848 |
+
"loss": 1.1617,
|
| 849 |
+
"step": 595
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.2628120893561104,
|
| 853 |
+
"grad_norm": 0.2843186093708674,
|
| 854 |
+
"learning_rate": 1.8432772879422797e-05,
|
| 855 |
+
"loss": 1.1355,
|
| 856 |
+
"step": 600
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.26500219010074466,
|
| 860 |
+
"grad_norm": 0.31588736128898504,
|
| 861 |
+
"learning_rate": 1.8391423005595928e-05,
|
| 862 |
+
"loss": 1.1359,
|
| 863 |
+
"step": 605
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.2671922908453789,
|
| 867 |
+
"grad_norm": 0.29315028625446876,
|
| 868 |
+
"learning_rate": 1.8349582368135764e-05,
|
| 869 |
+
"loss": 1.1551,
|
| 870 |
+
"step": 610
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.26938239159001315,
|
| 874 |
+
"grad_norm": 0.2882074502213313,
|
| 875 |
+
"learning_rate": 1.8307253414048395e-05,
|
| 876 |
+
"loss": 1.1363,
|
| 877 |
+
"step": 615
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.2715724923346474,
|
| 881 |
+
"grad_norm": 0.2932578089430528,
|
| 882 |
+
"learning_rate": 1.8264438618898617e-05,
|
| 883 |
+
"loss": 1.1455,
|
| 884 |
+
"step": 620
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.27376259307928164,
|
| 888 |
+
"grad_norm": 0.27801162707284016,
|
| 889 |
+
"learning_rate": 1.8221140486665125e-05,
|
| 890 |
+
"loss": 1.1621,
|
| 891 |
+
"step": 625
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.2759526938239159,
|
| 895 |
+
"grad_norm": 0.29375240288760845,
|
| 896 |
+
"learning_rate": 1.8177361549594068e-05,
|
| 897 |
+
"loss": 1.1492,
|
| 898 |
+
"step": 630
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.2781427945685502,
|
| 902 |
+
"grad_norm": 0.276903179408844,
|
| 903 |
+
"learning_rate": 1.813310436805099e-05,
|
| 904 |
+
"loss": 1.1445,
|
| 905 |
+
"step": 635
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.2803328953131844,
|
| 909 |
+
"grad_norm": 0.27955633362670035,
|
| 910 |
+
"learning_rate": 1.8088371530371044e-05,
|
| 911 |
+
"loss": 1.1531,
|
| 912 |
+
"step": 640
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.28252299605781866,
|
| 916 |
+
"grad_norm": 0.2799230772005429,
|
| 917 |
+
"learning_rate": 1.804316565270765e-05,
|
| 918 |
+
"loss": 1.1512,
|
| 919 |
+
"step": 645
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.28471309680245294,
|
| 923 |
+
"grad_norm": 0.27521037714056357,
|
| 924 |
+
"learning_rate": 1.7997489378879465e-05,
|
| 925 |
+
"loss": 1.1438,
|
| 926 |
+
"step": 650
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.28690319754708715,
|
| 930 |
+
"grad_norm": 0.29389670124687384,
|
| 931 |
+
"learning_rate": 1.7951345380215795e-05,
|
| 932 |
+
"loss": 1.152,
|
| 933 |
+
"step": 655
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.2890932982917214,
|
| 937 |
+
"grad_norm": 0.28364632855748595,
|
| 938 |
+
"learning_rate": 1.7904736355400324e-05,
|
| 939 |
+
"loss": 1.1576,
|
| 940 |
+
"step": 660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.2912833990363557,
|
| 944 |
+
"grad_norm": 0.2767305609560192,
|
| 945 |
+
"learning_rate": 1.785766503031332e-05,
|
| 946 |
+
"loss": 1.1414,
|
| 947 |
+
"step": 665
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.2934734997809899,
|
| 951 |
+
"grad_norm": 0.28006968803895216,
|
| 952 |
+
"learning_rate": 1.78101341578722e-05,
|
| 953 |
+
"loss": 1.1367,
|
| 954 |
+
"step": 670
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.2956636005256242,
|
| 958 |
+
"grad_norm": 0.2938177814391214,
|
| 959 |
+
"learning_rate": 1.7762146517870526e-05,
|
| 960 |
+
"loss": 1.132,
|
| 961 |
+
"step": 675
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.29785370127025845,
|
| 965 |
+
"grad_norm": 0.2763427209869464,
|
| 966 |
+
"learning_rate": 1.7713704916815432e-05,
|
| 967 |
+
"loss": 1.1461,
|
| 968 |
+
"step": 680
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.30004380201489267,
|
| 972 |
+
"grad_norm": 0.3006381369806697,
|
| 973 |
+
"learning_rate": 1.76648121877635e-05,
|
| 974 |
+
"loss": 1.134,
|
| 975 |
+
"step": 685
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.30223390275952694,
|
| 979 |
+
"grad_norm": 0.2910687567455647,
|
| 980 |
+
"learning_rate": 1.7615471190155056e-05,
|
| 981 |
+
"loss": 1.1176,
|
| 982 |
+
"step": 690
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.3044240035041612,
|
| 986 |
+
"grad_norm": 0.2678904446910071,
|
| 987 |
+
"learning_rate": 1.7565684809646946e-05,
|
| 988 |
+
"loss": 1.1223,
|
| 989 |
+
"step": 695
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.3066141042487954,
|
| 993 |
+
"grad_norm": 0.2753813506939354,
|
| 994 |
+
"learning_rate": 1.7515455957943772e-05,
|
| 995 |
+
"loss": 1.1574,
|
| 996 |
+
"step": 700
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.3088042049934297,
|
| 1000 |
+
"grad_norm": 0.28669942485252364,
|
| 1001 |
+
"learning_rate": 1.746478757262761e-05,
|
| 1002 |
+
"loss": 1.1496,
|
| 1003 |
+
"step": 705
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.31099430573806397,
|
| 1007 |
+
"grad_norm": 0.2843980836766018,
|
| 1008 |
+
"learning_rate": 1.7413682616986185e-05,
|
| 1009 |
+
"loss": 1.1363,
|
| 1010 |
+
"step": 710
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.3131844064826982,
|
| 1014 |
+
"grad_norm": 0.2808223481563682,
|
| 1015 |
+
"learning_rate": 1.73621440798396e-05,
|
| 1016 |
+
"loss": 1.1305,
|
| 1017 |
+
"step": 715
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.31537450722733246,
|
| 1021 |
+
"grad_norm": 0.2922348240666733,
|
| 1022 |
+
"learning_rate": 1.7310174975365508e-05,
|
| 1023 |
+
"loss": 1.1252,
|
| 1024 |
+
"step": 720
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.31756460797196673,
|
| 1028 |
+
"grad_norm": 0.28067686922560126,
|
| 1029 |
+
"learning_rate": 1.7257778342922853e-05,
|
| 1030 |
+
"loss": 1.1348,
|
| 1031 |
+
"step": 725
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.31975470871660094,
|
| 1035 |
+
"grad_norm": 0.2945526878670907,
|
| 1036 |
+
"learning_rate": 1.7204957246874103e-05,
|
| 1037 |
+
"loss": 1.1527,
|
| 1038 |
+
"step": 730
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.3219448094612352,
|
| 1042 |
+
"grad_norm": 0.276941149759669,
|
| 1043 |
+
"learning_rate": 1.7151714776406034e-05,
|
| 1044 |
+
"loss": 1.1102,
|
| 1045 |
+
"step": 735
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.3241349102058695,
|
| 1049 |
+
"grad_norm": 0.289439650523264,
|
| 1050 |
+
"learning_rate": 1.7098054045349076e-05,
|
| 1051 |
+
"loss": 1.125,
|
| 1052 |
+
"step": 740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.3263250109505037,
|
| 1056 |
+
"grad_norm": 0.2801285692740154,
|
| 1057 |
+
"learning_rate": 1.7043978191995177e-05,
|
| 1058 |
+
"loss": 1.127,
|
| 1059 |
+
"step": 745
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.328515111695138,
|
| 1063 |
+
"grad_norm": 0.3350495049060471,
|
| 1064 |
+
"learning_rate": 1.6989490378914304e-05,
|
| 1065 |
+
"loss": 1.1227,
|
| 1066 |
+
"step": 750
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.33070521243977224,
|
| 1070 |
+
"grad_norm": 0.2808592002786063,
|
| 1071 |
+
"learning_rate": 1.6934593792769435e-05,
|
| 1072 |
+
"loss": 1.152,
|
| 1073 |
+
"step": 755
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.33289531318440646,
|
| 1077 |
+
"grad_norm": 0.27808585554127363,
|
| 1078 |
+
"learning_rate": 1.6879291644130235e-05,
|
| 1079 |
+
"loss": 1.1309,
|
| 1080 |
+
"step": 760
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.33508541392904073,
|
| 1084 |
+
"grad_norm": 0.27069745729859374,
|
| 1085 |
+
"learning_rate": 1.682358716728525e-05,
|
| 1086 |
+
"loss": 1.1416,
|
| 1087 |
+
"step": 765
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.337275514673675,
|
| 1091 |
+
"grad_norm": 0.29843211507422185,
|
| 1092 |
+
"learning_rate": 1.676748362005279e-05,
|
| 1093 |
+
"loss": 1.1449,
|
| 1094 |
+
"step": 770
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.3394656154183092,
|
| 1098 |
+
"grad_norm": 0.28556892181225596,
|
| 1099 |
+
"learning_rate": 1.671098428359037e-05,
|
| 1100 |
+
"loss": 1.143,
|
| 1101 |
+
"step": 775
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.3416557161629435,
|
| 1105 |
+
"grad_norm": 0.30065398619371225,
|
| 1106 |
+
"learning_rate": 1.6654092462202828e-05,
|
| 1107 |
+
"loss": 1.1227,
|
| 1108 |
+
"step": 780
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.34384581690757776,
|
| 1112 |
+
"grad_norm": 0.2681674526100603,
|
| 1113 |
+
"learning_rate": 1.6596811483149077e-05,
|
| 1114 |
+
"loss": 1.1273,
|
| 1115 |
+
"step": 785
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.346035917652212,
|
| 1119 |
+
"grad_norm": 0.29230660631987243,
|
| 1120 |
+
"learning_rate": 1.6539144696447504e-05,
|
| 1121 |
+
"loss": 1.1277,
|
| 1122 |
+
"step": 790
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.34822601839684625,
|
| 1126 |
+
"grad_norm": 0.27613144345526425,
|
| 1127 |
+
"learning_rate": 1.6481095474680062e-05,
|
| 1128 |
+
"loss": 1.1285,
|
| 1129 |
+
"step": 795
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.3504161191414805,
|
| 1133 |
+
"grad_norm": 0.28721223205254864,
|
| 1134 |
+
"learning_rate": 1.642266721279502e-05,
|
| 1135 |
+
"loss": 1.1215,
|
| 1136 |
+
"step": 800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.35260621988611474,
|
| 1140 |
+
"grad_norm": 0.2839922006300779,
|
| 1141 |
+
"learning_rate": 1.6363863327908405e-05,
|
| 1142 |
+
"loss": 1.1219,
|
| 1143 |
+
"step": 805
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.354796320630749,
|
| 1147 |
+
"grad_norm": 0.2721319526014653,
|
| 1148 |
+
"learning_rate": 1.630468725910417e-05,
|
| 1149 |
+
"loss": 1.1297,
|
| 1150 |
+
"step": 810
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.3569864213753833,
|
| 1154 |
+
"grad_norm": 0.2666725910793476,
|
| 1155 |
+
"learning_rate": 1.6245142467233067e-05,
|
| 1156 |
+
"loss": 1.1525,
|
| 1157 |
+
"step": 815
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.3591765221200175,
|
| 1161 |
+
"grad_norm": 0.2749768948580958,
|
| 1162 |
+
"learning_rate": 1.618523243471021e-05,
|
| 1163 |
+
"loss": 1.1252,
|
| 1164 |
+
"step": 820
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.36136662286465177,
|
| 1168 |
+
"grad_norm": 0.27001888357231557,
|
| 1169 |
+
"learning_rate": 1.6124960665311447e-05,
|
| 1170 |
+
"loss": 1.1227,
|
| 1171 |
+
"step": 825
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.36355672360928604,
|
| 1175 |
+
"grad_norm": 0.27097536893189883,
|
| 1176 |
+
"learning_rate": 1.606433068396843e-05,
|
| 1177 |
+
"loss": 1.1273,
|
| 1178 |
+
"step": 830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.3657468243539203,
|
| 1182 |
+
"grad_norm": 0.28488649062206617,
|
| 1183 |
+
"learning_rate": 1.6003346036562457e-05,
|
| 1184 |
+
"loss": 1.1186,
|
| 1185 |
+
"step": 835
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.3679369250985545,
|
| 1189 |
+
"grad_norm": 0.27688500142196154,
|
| 1190 |
+
"learning_rate": 1.5942010289717108e-05,
|
| 1191 |
+
"loss": 1.1402,
|
| 1192 |
+
"step": 840
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.3701270258431888,
|
| 1196 |
+
"grad_norm": 0.2674674925612205,
|
| 1197 |
+
"learning_rate": 1.588032703058964e-05,
|
| 1198 |
+
"loss": 1.0869,
|
| 1199 |
+
"step": 845
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.37231712658782307,
|
| 1203 |
+
"grad_norm": 0.2856317902597983,
|
| 1204 |
+
"learning_rate": 1.5818299866661216e-05,
|
| 1205 |
+
"loss": 1.1398,
|
| 1206 |
+
"step": 850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.3745072273324573,
|
| 1210 |
+
"grad_norm": 0.27673240509708347,
|
| 1211 |
+
"learning_rate": 1.5755932425525907e-05,
|
| 1212 |
+
"loss": 1.1215,
|
| 1213 |
+
"step": 855
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.37669732807709155,
|
| 1217 |
+
"grad_norm": 0.2736444678975509,
|
| 1218 |
+
"learning_rate": 1.569322835467853e-05,
|
| 1219 |
+
"loss": 1.1176,
|
| 1220 |
+
"step": 860
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.3788874288217258,
|
| 1224 |
+
"grad_norm": 0.2768113382591734,
|
| 1225 |
+
"learning_rate": 1.563019132130136e-05,
|
| 1226 |
+
"loss": 1.1383,
|
| 1227 |
+
"step": 865
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.38107752956636004,
|
| 1231 |
+
"grad_norm": 0.28162020446683744,
|
| 1232 |
+
"learning_rate": 1.5566825012049623e-05,
|
| 1233 |
+
"loss": 1.1363,
|
| 1234 |
+
"step": 870
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.3832676303109943,
|
| 1238 |
+
"grad_norm": 0.27091660063236284,
|
| 1239 |
+
"learning_rate": 1.5503133132835916e-05,
|
| 1240 |
+
"loss": 1.1215,
|
| 1241 |
+
"step": 875
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.3854577310556286,
|
| 1245 |
+
"grad_norm": 0.2666248763577975,
|
| 1246 |
+
"learning_rate": 1.5439119408613442e-05,
|
| 1247 |
+
"loss": 1.1254,
|
| 1248 |
+
"step": 880
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.3876478318002628,
|
| 1252 |
+
"grad_norm": 0.2703004250271796,
|
| 1253 |
+
"learning_rate": 1.5374787583158188e-05,
|
| 1254 |
+
"loss": 1.1582,
|
| 1255 |
+
"step": 885
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.38983793254489707,
|
| 1259 |
+
"grad_norm": 0.28795897475828547,
|
| 1260 |
+
"learning_rate": 1.531014141884994e-05,
|
| 1261 |
+
"loss": 1.1301,
|
| 1262 |
+
"step": 890
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.39202803328953134,
|
| 1266 |
+
"grad_norm": 0.26937126154119884,
|
| 1267 |
+
"learning_rate": 1.5245184696452286e-05,
|
| 1268 |
+
"loss": 1.0959,
|
| 1269 |
+
"step": 895
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.39421813403416556,
|
| 1273 |
+
"grad_norm": 0.2708023087895912,
|
| 1274 |
+
"learning_rate": 1.5179921214891469e-05,
|
| 1275 |
+
"loss": 1.1313,
|
| 1276 |
+
"step": 900
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.39640823477879983,
|
| 1280 |
+
"grad_norm": 0.2851242116128433,
|
| 1281 |
+
"learning_rate": 1.5114354791034225e-05,
|
| 1282 |
+
"loss": 1.1316,
|
| 1283 |
+
"step": 905
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.3985983355234341,
|
| 1287 |
+
"grad_norm": 0.270999449215019,
|
| 1288 |
+
"learning_rate": 1.5048489259464552e-05,
|
| 1289 |
+
"loss": 1.1297,
|
| 1290 |
+
"step": 910
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.4007884362680683,
|
| 1294 |
+
"grad_norm": 0.2815986051423743,
|
| 1295 |
+
"learning_rate": 1.4982328472259453e-05,
|
| 1296 |
+
"loss": 1.1414,
|
| 1297 |
+
"step": 915
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.4029785370127026,
|
| 1301 |
+
"grad_norm": 0.25822461525801005,
|
| 1302 |
+
"learning_rate": 1.4915876298763654e-05,
|
| 1303 |
+
"loss": 1.1287,
|
| 1304 |
+
"step": 920
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.40516863775733686,
|
| 1308 |
+
"grad_norm": 0.2747332123131356,
|
| 1309 |
+
"learning_rate": 1.4849136625363297e-05,
|
| 1310 |
+
"loss": 1.1141,
|
| 1311 |
+
"step": 925
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.4073587385019711,
|
| 1315 |
+
"grad_norm": 0.2712357004324246,
|
| 1316 |
+
"learning_rate": 1.4782113355258656e-05,
|
| 1317 |
+
"loss": 1.1074,
|
| 1318 |
+
"step": 930
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.40954883924660535,
|
| 1322 |
+
"grad_norm": 0.27085881871204187,
|
| 1323 |
+
"learning_rate": 1.471481040823587e-05,
|
| 1324 |
+
"loss": 1.1105,
|
| 1325 |
+
"step": 935
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.4117389399912396,
|
| 1329 |
+
"grad_norm": 0.27599882086276023,
|
| 1330 |
+
"learning_rate": 1.4647231720437687e-05,
|
| 1331 |
+
"loss": 1.123,
|
| 1332 |
+
"step": 940
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.41392904073587383,
|
| 1336 |
+
"grad_norm": 0.27922116769860744,
|
| 1337 |
+
"learning_rate": 1.4579381244133265e-05,
|
| 1338 |
+
"loss": 1.1469,
|
| 1339 |
+
"step": 945
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.4161191414805081,
|
| 1343 |
+
"grad_norm": 0.2735506939973002,
|
| 1344 |
+
"learning_rate": 1.4511262947487037e-05,
|
| 1345 |
+
"loss": 1.134,
|
| 1346 |
+
"step": 950
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.4183092422251424,
|
| 1350 |
+
"grad_norm": 0.26886344993062933,
|
| 1351 |
+
"learning_rate": 1.444288081432662e-05,
|
| 1352 |
+
"loss": 1.1246,
|
| 1353 |
+
"step": 955
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.4204993429697766,
|
| 1357 |
+
"grad_norm": 0.26964142533259994,
|
| 1358 |
+
"learning_rate": 1.4374238843909845e-05,
|
| 1359 |
+
"loss": 1.1275,
|
| 1360 |
+
"step": 960
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.42268944371441086,
|
| 1364 |
+
"grad_norm": 0.2745245268675833,
|
| 1365 |
+
"learning_rate": 1.4305341050690845e-05,
|
| 1366 |
+
"loss": 1.1457,
|
| 1367 |
+
"step": 965
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.42487954445904513,
|
| 1371 |
+
"grad_norm": 0.27343512900772715,
|
| 1372 |
+
"learning_rate": 1.4236191464085286e-05,
|
| 1373 |
+
"loss": 1.1109,
|
| 1374 |
+
"step": 970
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.42706964520367935,
|
| 1378 |
+
"grad_norm": 0.2743571452940947,
|
| 1379 |
+
"learning_rate": 1.4166794128234705e-05,
|
| 1380 |
+
"loss": 1.1187,
|
| 1381 |
+
"step": 975
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.4292597459483136,
|
| 1385 |
+
"grad_norm": 0.26544274644395,
|
| 1386 |
+
"learning_rate": 1.4097153101770008e-05,
|
| 1387 |
+
"loss": 1.1469,
|
| 1388 |
+
"step": 980
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.4314498466929479,
|
| 1392 |
+
"grad_norm": 0.5158719980721467,
|
| 1393 |
+
"learning_rate": 1.4027272457574082e-05,
|
| 1394 |
+
"loss": 1.125,
|
| 1395 |
+
"step": 985
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.4336399474375821,
|
| 1399 |
+
"grad_norm": 0.27681551273825566,
|
| 1400 |
+
"learning_rate": 1.3957156282543605e-05,
|
| 1401 |
+
"loss": 1.1363,
|
| 1402 |
+
"step": 990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.4358300481822164,
|
| 1406 |
+
"grad_norm": 0.27949169274878094,
|
| 1407 |
+
"learning_rate": 1.388680867735004e-05,
|
| 1408 |
+
"loss": 1.1254,
|
| 1409 |
+
"step": 995
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.43802014892685065,
|
| 1413 |
+
"grad_norm": 0.2693597774135224,
|
| 1414 |
+
"learning_rate": 1.381623375619979e-05,
|
| 1415 |
+
"loss": 1.1104,
|
| 1416 |
+
"step": 1000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.44021024967148487,
|
| 1420 |
+
"grad_norm": 0.27336241345507317,
|
| 1421 |
+
"learning_rate": 1.3745435646593613e-05,
|
| 1422 |
+
"loss": 1.1346,
|
| 1423 |
+
"step": 1005
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.44240035041611914,
|
| 1427 |
+
"grad_norm": 0.2697587637867333,
|
| 1428 |
+
"learning_rate": 1.3674418489085192e-05,
|
| 1429 |
+
"loss": 1.1313,
|
| 1430 |
+
"step": 1010
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.4445904511607534,
|
| 1434 |
+
"grad_norm": 0.2637933472764974,
|
| 1435 |
+
"learning_rate": 1.360318643703901e-05,
|
| 1436 |
+
"loss": 1.1174,
|
| 1437 |
+
"step": 1015
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.4467805519053876,
|
| 1441 |
+
"grad_norm": 0.2625274872058839,
|
| 1442 |
+
"learning_rate": 1.3531743656387426e-05,
|
| 1443 |
+
"loss": 1.1314,
|
| 1444 |
+
"step": 1020
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.4489706526500219,
|
| 1448 |
+
"grad_norm": 0.2612762476528872,
|
| 1449 |
+
"learning_rate": 1.346009432538705e-05,
|
| 1450 |
+
"loss": 1.1205,
|
| 1451 |
+
"step": 1025
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.45116075339465617,
|
| 1455 |
+
"grad_norm": 0.2804760409295694,
|
| 1456 |
+
"learning_rate": 1.3388242634374367e-05,
|
| 1457 |
+
"loss": 1.1547,
|
| 1458 |
+
"step": 1030
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.4533508541392904,
|
| 1462 |
+
"grad_norm": 0.2730844934530705,
|
| 1463 |
+
"learning_rate": 1.331619278552068e-05,
|
| 1464 |
+
"loss": 1.1301,
|
| 1465 |
+
"step": 1035
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.45554095488392465,
|
| 1469 |
+
"grad_norm": 0.2761820674316818,
|
| 1470 |
+
"learning_rate": 1.3243948992586347e-05,
|
| 1471 |
+
"loss": 1.1383,
|
| 1472 |
+
"step": 1040
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.4577310556285589,
|
| 1476 |
+
"grad_norm": 0.2687484002119245,
|
| 1477 |
+
"learning_rate": 1.3171515480674342e-05,
|
| 1478 |
+
"loss": 1.1248,
|
| 1479 |
+
"step": 1045
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.45992115637319314,
|
| 1483 |
+
"grad_norm": 0.27893259508437857,
|
| 1484 |
+
"learning_rate": 1.309889648598316e-05,
|
| 1485 |
+
"loss": 1.1535,
|
| 1486 |
+
"step": 1050
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.4621112571178274,
|
| 1490 |
+
"grad_norm": 0.2741964520386731,
|
| 1491 |
+
"learning_rate": 1.3026096255559055e-05,
|
| 1492 |
+
"loss": 1.1066,
|
| 1493 |
+
"step": 1055
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.4643013578624617,
|
| 1497 |
+
"grad_norm": 0.2766533566738994,
|
| 1498 |
+
"learning_rate": 1.295311904704767e-05,
|
| 1499 |
+
"loss": 1.1641,
|
| 1500 |
+
"step": 1060
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.4664914586070959,
|
| 1504 |
+
"grad_norm": 0.2715162801769272,
|
| 1505 |
+
"learning_rate": 1.2879969128445025e-05,
|
| 1506 |
+
"loss": 1.125,
|
| 1507 |
+
"step": 1065
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.46868155935173017,
|
| 1511 |
+
"grad_norm": 0.26726252131908934,
|
| 1512 |
+
"learning_rate": 1.2806650777847913e-05,
|
| 1513 |
+
"loss": 1.116,
|
| 1514 |
+
"step": 1070
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.47087166009636444,
|
| 1518 |
+
"grad_norm": 0.2840924656915755,
|
| 1519 |
+
"learning_rate": 1.2733168283203692e-05,
|
| 1520 |
+
"loss": 1.1318,
|
| 1521 |
+
"step": 1075
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.4730617608409987,
|
| 1525 |
+
"grad_norm": 0.26702580899914813,
|
| 1526 |
+
"learning_rate": 1.265952594205952e-05,
|
| 1527 |
+
"loss": 1.1156,
|
| 1528 |
+
"step": 1080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.47525186158563293,
|
| 1532 |
+
"grad_norm": 0.2773818319183054,
|
| 1533 |
+
"learning_rate": 1.2585728061311003e-05,
|
| 1534 |
+
"loss": 1.1184,
|
| 1535 |
+
"step": 1085
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.4774419623302672,
|
| 1539 |
+
"grad_norm": 0.27175516307591485,
|
| 1540 |
+
"learning_rate": 1.2511778956950325e-05,
|
| 1541 |
+
"loss": 1.1418,
|
| 1542 |
+
"step": 1090
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.47963206307490147,
|
| 1546 |
+
"grad_norm": 0.2689155932559786,
|
| 1547 |
+
"learning_rate": 1.243768295381382e-05,
|
| 1548 |
+
"loss": 1.1006,
|
| 1549 |
+
"step": 1095
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.4818221638195357,
|
| 1553 |
+
"grad_norm": 0.26134780287986276,
|
| 1554 |
+
"learning_rate": 1.2363444385329052e-05,
|
| 1555 |
+
"loss": 1.108,
|
| 1556 |
+
"step": 1100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.48401226456416996,
|
| 1560 |
+
"grad_norm": 0.2742997540168624,
|
| 1561 |
+
"learning_rate": 1.2289067593261358e-05,
|
| 1562 |
+
"loss": 1.1141,
|
| 1563 |
+
"step": 1105
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.48620236530880423,
|
| 1567 |
+
"grad_norm": 0.29141512596729113,
|
| 1568 |
+
"learning_rate": 1.2214556927459955e-05,
|
| 1569 |
+
"loss": 1.1176,
|
| 1570 |
+
"step": 1110
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.48839246605343845,
|
| 1574 |
+
"grad_norm": 0.27353849576189837,
|
| 1575 |
+
"learning_rate": 1.2139916745603509e-05,
|
| 1576 |
+
"loss": 1.1258,
|
| 1577 |
+
"step": 1115
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.4905825667980727,
|
| 1581 |
+
"grad_norm": 0.27107693520298354,
|
| 1582 |
+
"learning_rate": 1.2065151412945308e-05,
|
| 1583 |
+
"loss": 1.1098,
|
| 1584 |
+
"step": 1120
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.492772667542707,
|
| 1588 |
+
"grad_norm": 0.2819238585953994,
|
| 1589 |
+
"learning_rate": 1.1990265302057948e-05,
|
| 1590 |
+
"loss": 1.1225,
|
| 1591 |
+
"step": 1125
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.4949627682873412,
|
| 1595 |
+
"grad_norm": 0.266552343582286,
|
| 1596 |
+
"learning_rate": 1.191526279257762e-05,
|
| 1597 |
+
"loss": 1.1016,
|
| 1598 |
+
"step": 1130
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.4971528690319755,
|
| 1602 |
+
"grad_norm": 0.278057928098842,
|
| 1603 |
+
"learning_rate": 1.1840148270947962e-05,
|
| 1604 |
+
"loss": 1.1252,
|
| 1605 |
+
"step": 1135
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.49934296977660975,
|
| 1609 |
+
"grad_norm": 0.2936883558777508,
|
| 1610 |
+
"learning_rate": 1.1764926130163525e-05,
|
| 1611 |
+
"loss": 1.1359,
|
| 1612 |
+
"step": 1140
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.501533070521244,
|
| 1616 |
+
"grad_norm": 0.2834365935036193,
|
| 1617 |
+
"learning_rate": 1.1689600769512855e-05,
|
| 1618 |
+
"loss": 1.0994,
|
| 1619 |
+
"step": 1145
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.5037231712658782,
|
| 1623 |
+
"grad_norm": 0.274648980748947,
|
| 1624 |
+
"learning_rate": 1.1614176594321202e-05,
|
| 1625 |
+
"loss": 1.092,
|
| 1626 |
+
"step": 1150
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.5059132720105125,
|
| 1630 |
+
"grad_norm": 0.28409682473235875,
|
| 1631 |
+
"learning_rate": 1.1538658015692892e-05,
|
| 1632 |
+
"loss": 1.1346,
|
| 1633 |
+
"step": 1155
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.5081033727551467,
|
| 1637 |
+
"grad_norm": 0.2850776782098923,
|
| 1638 |
+
"learning_rate": 1.146304945025332e-05,
|
| 1639 |
+
"loss": 1.1227,
|
| 1640 |
+
"step": 1160
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.5102934734997809,
|
| 1644 |
+
"grad_norm": 0.2610428092893358,
|
| 1645 |
+
"learning_rate": 1.1387355319890685e-05,
|
| 1646 |
+
"loss": 1.0969,
|
| 1647 |
+
"step": 1165
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.5124835742444153,
|
| 1651 |
+
"grad_norm": 0.2740958199191816,
|
| 1652 |
+
"learning_rate": 1.1311580051497344e-05,
|
| 1653 |
+
"loss": 1.1246,
|
| 1654 |
+
"step": 1170
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.5146736749890495,
|
| 1658 |
+
"grad_norm": 0.2848288147278702,
|
| 1659 |
+
"learning_rate": 1.123572807671094e-05,
|
| 1660 |
+
"loss": 1.1109,
|
| 1661 |
+
"step": 1175
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.5168637757336837,
|
| 1665 |
+
"grad_norm": 0.26817858926421695,
|
| 1666 |
+
"learning_rate": 1.1159803831655206e-05,
|
| 1667 |
+
"loss": 1.1273,
|
| 1668 |
+
"step": 1180
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.519053876478318,
|
| 1672 |
+
"grad_norm": 0.2697805969038893,
|
| 1673 |
+
"learning_rate": 1.1083811756680523e-05,
|
| 1674 |
+
"loss": 1.1289,
|
| 1675 |
+
"step": 1185
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.5212439772229522,
|
| 1679 |
+
"grad_norm": 0.2705343306376679,
|
| 1680 |
+
"learning_rate": 1.1007756296104237e-05,
|
| 1681 |
+
"loss": 1.1469,
|
| 1682 |
+
"step": 1190
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.5234340779675866,
|
| 1686 |
+
"grad_norm": 0.27282726875245455,
|
| 1687 |
+
"learning_rate": 1.0931641897950733e-05,
|
| 1688 |
+
"loss": 1.1332,
|
| 1689 |
+
"step": 1195
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.5256241787122208,
|
| 1693 |
+
"grad_norm": 0.28059459041359747,
|
| 1694 |
+
"learning_rate": 1.08554730136913e-05,
|
| 1695 |
+
"loss": 1.0916,
|
| 1696 |
+
"step": 1200
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.527814279456855,
|
| 1700 |
+
"grad_norm": 0.2638181805357147,
|
| 1701 |
+
"learning_rate": 1.0779254097983788e-05,
|
| 1702 |
+
"loss": 1.1057,
|
| 1703 |
+
"step": 1205
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.5300043802014893,
|
| 1707 |
+
"grad_norm": 0.26522994514380227,
|
| 1708 |
+
"learning_rate": 1.0702989608412096e-05,
|
| 1709 |
+
"loss": 1.1027,
|
| 1710 |
+
"step": 1210
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.5321944809461235,
|
| 1714 |
+
"grad_norm": 0.25855768118273403,
|
| 1715 |
+
"learning_rate": 1.0626684005225443e-05,
|
| 1716 |
+
"loss": 1.1332,
|
| 1717 |
+
"step": 1215
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.5343845816907578,
|
| 1721 |
+
"grad_norm": 0.27413179764811996,
|
| 1722 |
+
"learning_rate": 1.0550341751077553e-05,
|
| 1723 |
+
"loss": 1.1176,
|
| 1724 |
+
"step": 1220
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.5365746824353921,
|
| 1728 |
+
"grad_norm": 0.2642903693875254,
|
| 1729 |
+
"learning_rate": 1.0473967310765629e-05,
|
| 1730 |
+
"loss": 1.0857,
|
| 1731 |
+
"step": 1225
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.5387647831800263,
|
| 1735 |
+
"grad_norm": 0.2628203067700955,
|
| 1736 |
+
"learning_rate": 1.039756515096926e-05,
|
| 1737 |
+
"loss": 1.1195,
|
| 1738 |
+
"step": 1230
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.5409548839246605,
|
| 1742 |
+
"grad_norm": 0.2783963215365502,
|
| 1743 |
+
"learning_rate": 1.0321139739989167e-05,
|
| 1744 |
+
"loss": 1.1215,
|
| 1745 |
+
"step": 1235
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.5431449846692948,
|
| 1749 |
+
"grad_norm": 0.2772449600445937,
|
| 1750 |
+
"learning_rate": 1.0244695547485912e-05,
|
| 1751 |
+
"loss": 1.1037,
|
| 1752 |
+
"step": 1240
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.545335085413929,
|
| 1756 |
+
"grad_norm": 0.2660438451116181,
|
| 1757 |
+
"learning_rate": 1.0168237044218452e-05,
|
| 1758 |
+
"loss": 1.1262,
|
| 1759 |
+
"step": 1245
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.5475251861585633,
|
| 1763 |
+
"grad_norm": 0.2598928720700204,
|
| 1764 |
+
"learning_rate": 1.0091768701782714e-05,
|
| 1765 |
+
"loss": 1.1246,
|
| 1766 |
+
"step": 1250
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.5497152869031976,
|
| 1770 |
+
"grad_norm": 0.2777074874692869,
|
| 1771 |
+
"learning_rate": 1.0015294992350044e-05,
|
| 1772 |
+
"loss": 1.1055,
|
| 1773 |
+
"step": 1255
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.5519053876478318,
|
| 1777 |
+
"grad_norm": 0.28449089911369874,
|
| 1778 |
+
"learning_rate": 9.93882038840568e-06,
|
| 1779 |
+
"loss": 1.1223,
|
| 1780 |
+
"step": 1260
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.554095488392466,
|
| 1784 |
+
"grad_norm": 0.2768874782485583,
|
| 1785 |
+
"learning_rate": 9.862349362487172e-06,
|
| 1786 |
+
"loss": 1.0961,
|
| 1787 |
+
"step": 1265
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.5562855891371004,
|
| 1791 |
+
"grad_norm": 0.26460753865275577,
|
| 1792 |
+
"learning_rate": 9.785886386922806e-06,
|
| 1793 |
+
"loss": 1.1076,
|
| 1794 |
+
"step": 1270
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.5584756898817346,
|
| 1798 |
+
"grad_norm": 0.28161562628451337,
|
| 1799 |
+
"learning_rate": 9.709435933570068e-06,
|
| 1800 |
+
"loss": 1.1029,
|
| 1801 |
+
"step": 1275
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.5606657906263688,
|
| 1805 |
+
"grad_norm": 0.2563062588428807,
|
| 1806 |
+
"learning_rate": 9.633002473554077e-06,
|
| 1807 |
+
"loss": 1.1096,
|
| 1808 |
+
"step": 1280
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.5628558913710031,
|
| 1812 |
+
"grad_norm": 0.26175322953452584,
|
| 1813 |
+
"learning_rate": 9.556590477006123e-06,
|
| 1814 |
+
"loss": 1.1229,
|
| 1815 |
+
"step": 1285
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.5650459921156373,
|
| 1819 |
+
"grad_norm": 0.26468057620238034,
|
| 1820 |
+
"learning_rate": 9.480204412802236e-06,
|
| 1821 |
+
"loss": 1.0994,
|
| 1822 |
+
"step": 1290
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.5672360928602715,
|
| 1826 |
+
"grad_norm": 0.25809069797603174,
|
| 1827 |
+
"learning_rate": 9.403848748301802e-06,
|
| 1828 |
+
"loss": 1.1059,
|
| 1829 |
+
"step": 1295
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.5694261936049059,
|
| 1833 |
+
"grad_norm": 0.27252102620530055,
|
| 1834 |
+
"learning_rate": 9.327527949086327e-06,
|
| 1835 |
+
"loss": 1.1092,
|
| 1836 |
+
"step": 1300
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.5716162943495401,
|
| 1840 |
+
"grad_norm": 0.2715402906362093,
|
| 1841 |
+
"learning_rate": 9.251246478698242e-06,
|
| 1842 |
+
"loss": 1.1367,
|
| 1843 |
+
"step": 1305
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.5738063950941743,
|
| 1847 |
+
"grad_norm": 0.26828019230229255,
|
| 1848 |
+
"learning_rate": 9.175008798379893e-06,
|
| 1849 |
+
"loss": 1.1043,
|
| 1850 |
+
"step": 1310
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.5759964958388086,
|
| 1854 |
+
"grad_norm": 0.2770878196295294,
|
| 1855 |
+
"learning_rate": 9.098819366812594e-06,
|
| 1856 |
+
"loss": 1.1213,
|
| 1857 |
+
"step": 1315
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.5781865965834428,
|
| 1861 |
+
"grad_norm": 0.281651051703252,
|
| 1862 |
+
"learning_rate": 9.022682639855883e-06,
|
| 1863 |
+
"loss": 1.1367,
|
| 1864 |
+
"step": 1320
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.5803766973280771,
|
| 1868 |
+
"grad_norm": 0.27314753294139044,
|
| 1869 |
+
"learning_rate": 8.946603070286926e-06,
|
| 1870 |
+
"loss": 1.107,
|
| 1871 |
+
"step": 1325
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.5825667980727114,
|
| 1875 |
+
"grad_norm": 0.26226614181355684,
|
| 1876 |
+
"learning_rate": 8.870585107540101e-06,
|
| 1877 |
+
"loss": 1.1281,
|
| 1878 |
+
"step": 1330
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.5847568988173456,
|
| 1882 |
+
"grad_norm": 0.27293558156801573,
|
| 1883 |
+
"learning_rate": 8.79463319744677e-06,
|
| 1884 |
+
"loss": 1.0809,
|
| 1885 |
+
"step": 1335
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.5869469995619798,
|
| 1889 |
+
"grad_norm": 0.26602352956316966,
|
| 1890 |
+
"learning_rate": 8.718751781975284e-06,
|
| 1891 |
+
"loss": 1.1385,
|
| 1892 |
+
"step": 1340
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.5891371003066141,
|
| 1896 |
+
"grad_norm": 0.2746959499141311,
|
| 1897 |
+
"learning_rate": 8.642945298971168e-06,
|
| 1898 |
+
"loss": 1.1172,
|
| 1899 |
+
"step": 1345
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.5913272010512484,
|
| 1903 |
+
"grad_norm": 0.2677554902377095,
|
| 1904 |
+
"learning_rate": 8.56721818189763e-06,
|
| 1905 |
+
"loss": 1.1133,
|
| 1906 |
+
"step": 1350
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.5935173017958826,
|
| 1910 |
+
"grad_norm": 0.2662798900188826,
|
| 1911 |
+
"learning_rate": 8.491574859576222e-06,
|
| 1912 |
+
"loss": 1.1104,
|
| 1913 |
+
"step": 1355
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.5957074025405169,
|
| 1917 |
+
"grad_norm": 0.26469719917634077,
|
| 1918 |
+
"learning_rate": 8.416019755927851e-06,
|
| 1919 |
+
"loss": 1.1086,
|
| 1920 |
+
"step": 1360
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.5978975032851511,
|
| 1924 |
+
"grad_norm": 0.26386029178727227,
|
| 1925 |
+
"learning_rate": 8.340557289714055e-06,
|
| 1926 |
+
"loss": 1.1168,
|
| 1927 |
+
"step": 1365
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.6000876040297853,
|
| 1931 |
+
"grad_norm": 0.253515604893805,
|
| 1932 |
+
"learning_rate": 8.26519187427856e-06,
|
| 1933 |
+
"loss": 1.1187,
|
| 1934 |
+
"step": 1370
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.6022777047744197,
|
| 1938 |
+
"grad_norm": 0.277605351754751,
|
| 1939 |
+
"learning_rate": 8.189927917289182e-06,
|
| 1940 |
+
"loss": 1.1059,
|
| 1941 |
+
"step": 1375
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.6044678055190539,
|
| 1945 |
+
"grad_norm": 0.26442216190859963,
|
| 1946 |
+
"learning_rate": 8.114769820480039e-06,
|
| 1947 |
+
"loss": 1.1176,
|
| 1948 |
+
"step": 1380
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.6066579062636881,
|
| 1952 |
+
"grad_norm": 0.2734826868231617,
|
| 1953 |
+
"learning_rate": 8.03972197939414e-06,
|
| 1954 |
+
"loss": 1.1215,
|
| 1955 |
+
"step": 1385
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.6088480070083224,
|
| 1959 |
+
"grad_norm": 0.2663060487289671,
|
| 1960 |
+
"learning_rate": 7.964788783126294e-06,
|
| 1961 |
+
"loss": 1.1037,
|
| 1962 |
+
"step": 1390
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.6110381077529566,
|
| 1966 |
+
"grad_norm": 0.26118107033447696,
|
| 1967 |
+
"learning_rate": 7.889974614066425e-06,
|
| 1968 |
+
"loss": 1.0967,
|
| 1969 |
+
"step": 1395
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.6132282084975909,
|
| 1973 |
+
"grad_norm": 0.27155741223634616,
|
| 1974 |
+
"learning_rate": 7.815283847643277e-06,
|
| 1975 |
+
"loss": 1.1203,
|
| 1976 |
+
"step": 1400
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.6154183092422252,
|
| 1980 |
+
"grad_norm": 0.27035981345201915,
|
| 1981 |
+
"learning_rate": 7.740720852068524e-06,
|
| 1982 |
+
"loss": 1.1428,
|
| 1983 |
+
"step": 1405
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.6176084099868594,
|
| 1987 |
+
"grad_norm": 0.2683573304766547,
|
| 1988 |
+
"learning_rate": 7.66628998808129e-06,
|
| 1989 |
+
"loss": 1.1301,
|
| 1990 |
+
"step": 1410
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.6197985107314936,
|
| 1994 |
+
"grad_norm": 0.26324766469257666,
|
| 1995 |
+
"learning_rate": 7.591995608693118e-06,
|
| 1996 |
+
"loss": 1.1115,
|
| 1997 |
+
"step": 1415
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.6219886114761279,
|
| 2001 |
+
"grad_norm": 0.26863159797075176,
|
| 2002 |
+
"learning_rate": 7.5178420589334025e-06,
|
| 2003 |
+
"loss": 1.1055,
|
| 2004 |
+
"step": 1420
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.6241787122207622,
|
| 2008 |
+
"grad_norm": 0.25941294580030216,
|
| 2009 |
+
"learning_rate": 7.443833675595254e-06,
|
| 2010 |
+
"loss": 1.1293,
|
| 2011 |
+
"step": 1425
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.6263688129653964,
|
| 2015 |
+
"grad_norm": 0.2623305114567609,
|
| 2016 |
+
"learning_rate": 7.369974786981879e-06,
|
| 2017 |
+
"loss": 1.0961,
|
| 2018 |
+
"step": 1430
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.6285589137100307,
|
| 2022 |
+
"grad_norm": 0.2857793789171064,
|
| 2023 |
+
"learning_rate": 7.296269712653436e-06,
|
| 2024 |
+
"loss": 1.1051,
|
| 2025 |
+
"step": 1435
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.6307490144546649,
|
| 2029 |
+
"grad_norm": 0.2738731727686292,
|
| 2030 |
+
"learning_rate": 7.222722763174428e-06,
|
| 2031 |
+
"loss": 1.0965,
|
| 2032 |
+
"step": 1440
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.6329391151992991,
|
| 2036 |
+
"grad_norm": 0.2780751677609743,
|
| 2037 |
+
"learning_rate": 7.149338239861579e-06,
|
| 2038 |
+
"loss": 1.1125,
|
| 2039 |
+
"step": 1445
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.6351292159439335,
|
| 2043 |
+
"grad_norm": 0.271978624523703,
|
| 2044 |
+
"learning_rate": 7.076120434532292e-06,
|
| 2045 |
+
"loss": 1.0795,
|
| 2046 |
+
"step": 1450
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.6373193166885677,
|
| 2050 |
+
"grad_norm": 0.2889878149869347,
|
| 2051 |
+
"learning_rate": 7.003073629253638e-06,
|
| 2052 |
+
"loss": 1.1094,
|
| 2053 |
+
"step": 1455
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.6395094174332019,
|
| 2057 |
+
"grad_norm": 0.26177418700227856,
|
| 2058 |
+
"learning_rate": 6.9302020960919405e-06,
|
| 2059 |
+
"loss": 1.132,
|
| 2060 |
+
"step": 1460
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.6416995181778362,
|
| 2064 |
+
"grad_norm": 0.27487300616889226,
|
| 2065 |
+
"learning_rate": 6.857510096862901e-06,
|
| 2066 |
+
"loss": 1.1131,
|
| 2067 |
+
"step": 1465
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.6438896189224704,
|
| 2071 |
+
"grad_norm": 0.2673004044565668,
|
| 2072 |
+
"learning_rate": 6.785001882882376e-06,
|
| 2073 |
+
"loss": 1.1047,
|
| 2074 |
+
"step": 1470
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.6460797196671046,
|
| 2078 |
+
"grad_norm": 0.2637312152468036,
|
| 2079 |
+
"learning_rate": 6.712681694717723e-06,
|
| 2080 |
+
"loss": 1.1096,
|
| 2081 |
+
"step": 1475
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.648269820411739,
|
| 2085 |
+
"grad_norm": 0.2630882968696908,
|
| 2086 |
+
"learning_rate": 6.640553761939822e-06,
|
| 2087 |
+
"loss": 1.0973,
|
| 2088 |
+
"step": 1480
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.6504599211563732,
|
| 2092 |
+
"grad_norm": 0.2638392419100043,
|
| 2093 |
+
"learning_rate": 6.568622302875682e-06,
|
| 2094 |
+
"loss": 1.0982,
|
| 2095 |
+
"step": 1485
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.6526500219010074,
|
| 2099 |
+
"grad_norm": 0.2626309470206596,
|
| 2100 |
+
"learning_rate": 6.496891524361757e-06,
|
| 2101 |
+
"loss": 1.1172,
|
| 2102 |
+
"step": 1490
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.6548401226456417,
|
| 2106 |
+
"grad_norm": 0.265160449047166,
|
| 2107 |
+
"learning_rate": 6.42536562149791e-06,
|
| 2108 |
+
"loss": 1.1125,
|
| 2109 |
+
"step": 1495
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.657030223390276,
|
| 2113 |
+
"grad_norm": 0.26619884121137627,
|
| 2114 |
+
"learning_rate": 6.354048777402061e-06,
|
| 2115 |
+
"loss": 1.1328,
|
| 2116 |
+
"step": 1500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.6592203241349102,
|
| 2120 |
+
"grad_norm": 0.2600814085765291,
|
| 2121 |
+
"learning_rate": 6.282945162965548e-06,
|
| 2122 |
+
"loss": 1.117,
|
| 2123 |
+
"step": 1505
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.6614104248795445,
|
| 2127 |
+
"grad_norm": 0.272185670744957,
|
| 2128 |
+
"learning_rate": 6.212058936609189e-06,
|
| 2129 |
+
"loss": 1.1209,
|
| 2130 |
+
"step": 1510
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.6636005256241787,
|
| 2134 |
+
"grad_norm": 0.2641333188928149,
|
| 2135 |
+
"learning_rate": 6.1413942440400994e-06,
|
| 2136 |
+
"loss": 1.1168,
|
| 2137 |
+
"step": 1515
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.6657906263688129,
|
| 2141 |
+
"grad_norm": 0.26009676604807275,
|
| 2142 |
+
"learning_rate": 6.070955218009213e-06,
|
| 2143 |
+
"loss": 1.0936,
|
| 2144 |
+
"step": 1520
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.6679807271134472,
|
| 2148 |
+
"grad_norm": 0.2615413153711293,
|
| 2149 |
+
"learning_rate": 6.0007459780695885e-06,
|
| 2150 |
+
"loss": 1.1061,
|
| 2151 |
+
"step": 1525
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.6701708278580815,
|
| 2155 |
+
"grad_norm": 0.2892089912864181,
|
| 2156 |
+
"learning_rate": 5.93077063033549e-06,
|
| 2157 |
+
"loss": 1.1055,
|
| 2158 |
+
"step": 1530
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.6723609286027157,
|
| 2162 |
+
"grad_norm": 0.2629578247792853,
|
| 2163 |
+
"learning_rate": 5.861033267242238e-06,
|
| 2164 |
+
"loss": 1.1146,
|
| 2165 |
+
"step": 1535
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.67455102934735,
|
| 2169 |
+
"grad_norm": 0.2653857163412152,
|
| 2170 |
+
"learning_rate": 5.791537967306869e-06,
|
| 2171 |
+
"loss": 1.1107,
|
| 2172 |
+
"step": 1540
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.6767411300919842,
|
| 2176 |
+
"grad_norm": 0.2641301458887951,
|
| 2177 |
+
"learning_rate": 5.722288794889603e-06,
|
| 2178 |
+
"loss": 1.0953,
|
| 2179 |
+
"step": 1545
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.6789312308366184,
|
| 2183 |
+
"grad_norm": 0.2627115214160706,
|
| 2184 |
+
"learning_rate": 5.653289799956161e-06,
|
| 2185 |
+
"loss": 1.0957,
|
| 2186 |
+
"step": 1550
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.6811213315812528,
|
| 2190 |
+
"grad_norm": 0.2744248525372843,
|
| 2191 |
+
"learning_rate": 5.584545017840886e-06,
|
| 2192 |
+
"loss": 1.0973,
|
| 2193 |
+
"step": 1555
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.683311432325887,
|
| 2197 |
+
"grad_norm": 0.26271843375673565,
|
| 2198 |
+
"learning_rate": 5.516058469010741e-06,
|
| 2199 |
+
"loss": 1.1066,
|
| 2200 |
+
"step": 1560
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.6855015330705212,
|
| 2204 |
+
"grad_norm": 0.28239608054669124,
|
| 2205 |
+
"learning_rate": 5.447834158830202e-06,
|
| 2206 |
+
"loss": 1.1457,
|
| 2207 |
+
"step": 1565
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.6876916338151555,
|
| 2211 |
+
"grad_norm": 0.2682726187574874,
|
| 2212 |
+
"learning_rate": 5.379876077326979e-06,
|
| 2213 |
+
"loss": 1.0771,
|
| 2214 |
+
"step": 1570
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.6898817345597897,
|
| 2218 |
+
"grad_norm": 0.2693944920009028,
|
| 2219 |
+
"learning_rate": 5.312188198958681e-06,
|
| 2220 |
+
"loss": 1.1191,
|
| 2221 |
+
"step": 1575
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.692071835304424,
|
| 2225 |
+
"grad_norm": 0.27066235636281594,
|
| 2226 |
+
"learning_rate": 5.244774482380361e-06,
|
| 2227 |
+
"loss": 1.1191,
|
| 2228 |
+
"step": 1580
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.6942619360490583,
|
| 2232 |
+
"grad_norm": 0.2607089341186109,
|
| 2233 |
+
"learning_rate": 5.177638870213008e-06,
|
| 2234 |
+
"loss": 1.0949,
|
| 2235 |
+
"step": 1585
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.6964520367936925,
|
| 2239 |
+
"grad_norm": 0.26730049681200124,
|
| 2240 |
+
"learning_rate": 5.1107852888129715e-06,
|
| 2241 |
+
"loss": 1.159,
|
| 2242 |
+
"step": 1590
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.6986421375383267,
|
| 2246 |
+
"grad_norm": 0.2561679895991625,
|
| 2247 |
+
"learning_rate": 5.04421764804232e-06,
|
| 2248 |
+
"loss": 1.0938,
|
| 2249 |
+
"step": 1595
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.700832238282961,
|
| 2253 |
+
"grad_norm": 0.25827915348141745,
|
| 2254 |
+
"learning_rate": 4.977939841040175e-06,
|
| 2255 |
+
"loss": 1.1178,
|
| 2256 |
+
"step": 1600
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.7030223390275953,
|
| 2260 |
+
"grad_norm": 0.2672625917790982,
|
| 2261 |
+
"learning_rate": 4.911955743995042e-06,
|
| 2262 |
+
"loss": 1.1195,
|
| 2263 |
+
"step": 1605
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.7052124397722295,
|
| 2267 |
+
"grad_norm": 0.2712982534329863,
|
| 2268 |
+
"learning_rate": 4.846269215918106e-06,
|
| 2269 |
+
"loss": 1.1215,
|
| 2270 |
+
"step": 1610
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.7074025405168638,
|
| 2274 |
+
"grad_norm": 0.2637027906298979,
|
| 2275 |
+
"learning_rate": 4.780884098417531e-06,
|
| 2276 |
+
"loss": 1.098,
|
| 2277 |
+
"step": 1615
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.709592641261498,
|
| 2281 |
+
"grad_norm": 0.26810920883601624,
|
| 2282 |
+
"learning_rate": 4.7158042154738094e-06,
|
| 2283 |
+
"loss": 1.1176,
|
| 2284 |
+
"step": 1620
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.7117827420061322,
|
| 2288 |
+
"grad_norm": 0.26084193501541225,
|
| 2289 |
+
"learning_rate": 4.6510333732160915e-06,
|
| 2290 |
+
"loss": 1.1262,
|
| 2291 |
+
"step": 1625
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.7139728427507666,
|
| 2295 |
+
"grad_norm": 0.2700256165770299,
|
| 2296 |
+
"learning_rate": 4.5865753596996335e-06,
|
| 2297 |
+
"loss": 1.1148,
|
| 2298 |
+
"step": 1630
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.7161629434954008,
|
| 2302 |
+
"grad_norm": 0.2639599226097337,
|
| 2303 |
+
"learning_rate": 4.522433944684197e-06,
|
| 2304 |
+
"loss": 1.1156,
|
| 2305 |
+
"step": 1635
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.718353044240035,
|
| 2309 |
+
"grad_norm": 0.2597365152157971,
|
| 2310 |
+
"learning_rate": 4.45861287941363e-06,
|
| 2311 |
+
"loss": 1.0854,
|
| 2312 |
+
"step": 1640
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.7205431449846693,
|
| 2316 |
+
"grad_norm": 0.26268213694795733,
|
| 2317 |
+
"learning_rate": 4.395115896396457e-06,
|
| 2318 |
+
"loss": 1.1258,
|
| 2319 |
+
"step": 1645
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.7227332457293035,
|
| 2323 |
+
"grad_norm": 0.26221093408579443,
|
| 2324 |
+
"learning_rate": 4.331946709187578e-06,
|
| 2325 |
+
"loss": 1.1109,
|
| 2326 |
+
"step": 1650
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.7249233464739377,
|
| 2330 |
+
"grad_norm": 0.2607246299946745,
|
| 2331 |
+
"learning_rate": 4.269109012171112e-06,
|
| 2332 |
+
"loss": 1.1104,
|
| 2333 |
+
"step": 1655
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.7271134472185721,
|
| 2337 |
+
"grad_norm": 0.2574716134373636,
|
| 2338 |
+
"learning_rate": 4.206606480344303e-06,
|
| 2339 |
+
"loss": 1.0932,
|
| 2340 |
+
"step": 1660
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.7293035479632063,
|
| 2344 |
+
"grad_norm": 0.265233772269521,
|
| 2345 |
+
"learning_rate": 4.14444276910263e-06,
|
| 2346 |
+
"loss": 1.1332,
|
| 2347 |
+
"step": 1665
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.7314936487078406,
|
| 2351 |
+
"grad_norm": 0.25655122243148587,
|
| 2352 |
+
"learning_rate": 4.08262151402598e-06,
|
| 2353 |
+
"loss": 1.0891,
|
| 2354 |
+
"step": 1670
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.7336837494524748,
|
| 2358 |
+
"grad_norm": 0.25501243107566296,
|
| 2359 |
+
"learning_rate": 4.021146330666065e-06,
|
| 2360 |
+
"loss": 1.0902,
|
| 2361 |
+
"step": 1675
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.735873850197109,
|
| 2365 |
+
"grad_norm": 0.2612800818169169,
|
| 2366 |
+
"learning_rate": 3.960020814334936e-06,
|
| 2367 |
+
"loss": 1.1195,
|
| 2368 |
+
"step": 1680
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.7380639509417434,
|
| 2372 |
+
"grad_norm": 0.2643874720892855,
|
| 2373 |
+
"learning_rate": 3.899248539894756e-06,
|
| 2374 |
+
"loss": 1.1137,
|
| 2375 |
+
"step": 1685
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.7402540516863776,
|
| 2379 |
+
"grad_norm": 0.26838926718062733,
|
| 2380 |
+
"learning_rate": 3.838833061548682e-06,
|
| 2381 |
+
"loss": 1.1094,
|
| 2382 |
+
"step": 1690
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.7424441524310118,
|
| 2386 |
+
"grad_norm": 0.26970931976433926,
|
| 2387 |
+
"learning_rate": 3.7787779126330314e-06,
|
| 2388 |
+
"loss": 1.1297,
|
| 2389 |
+
"step": 1695
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.7446342531756461,
|
| 2393 |
+
"grad_norm": 0.26137631703110864,
|
| 2394 |
+
"learning_rate": 3.7190866054106358e-06,
|
| 2395 |
+
"loss": 1.1254,
|
| 2396 |
+
"step": 1700
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.7468243539202803,
|
| 2400 |
+
"grad_norm": 0.2614031117426139,
|
| 2401 |
+
"learning_rate": 3.659762630865411e-06,
|
| 2402 |
+
"loss": 1.1041,
|
| 2403 |
+
"step": 1705
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.7490144546649146,
|
| 2407 |
+
"grad_norm": 0.2580239989121174,
|
| 2408 |
+
"learning_rate": 3.600809458498219e-06,
|
| 2409 |
+
"loss": 1.1076,
|
| 2410 |
+
"step": 1710
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.7512045554095489,
|
| 2414 |
+
"grad_norm": 0.2672365887137131,
|
| 2415 |
+
"learning_rate": 3.542230536123925e-06,
|
| 2416 |
+
"loss": 1.0988,
|
| 2417 |
+
"step": 1715
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.7533946561541831,
|
| 2421 |
+
"grad_norm": 0.26577844660787864,
|
| 2422 |
+
"learning_rate": 3.484029289669797e-06,
|
| 2423 |
+
"loss": 1.098,
|
| 2424 |
+
"step": 1720
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.7555847568988173,
|
| 2428 |
+
"grad_norm": 0.2563307410321518,
|
| 2429 |
+
"learning_rate": 3.4262091229750973e-06,
|
| 2430 |
+
"loss": 1.1051,
|
| 2431 |
+
"step": 1725
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.7577748576434516,
|
| 2435 |
+
"grad_norm": 0.25998557620881213,
|
| 2436 |
+
"learning_rate": 3.3687734175920505e-06,
|
| 2437 |
+
"loss": 1.1121,
|
| 2438 |
+
"step": 1730
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.7599649583880859,
|
| 2442 |
+
"grad_norm": 0.2641859803660572,
|
| 2443 |
+
"learning_rate": 3.311725532588049e-06,
|
| 2444 |
+
"loss": 1.1223,
|
| 2445 |
+
"step": 1735
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.7621550591327201,
|
| 2449 |
+
"grad_norm": 0.26214013514430917,
|
| 2450 |
+
"learning_rate": 3.2550688043492273e-06,
|
| 2451 |
+
"loss": 1.11,
|
| 2452 |
+
"step": 1740
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.7643451598773544,
|
| 2456 |
+
"grad_norm": 0.2613558739863409,
|
| 2457 |
+
"learning_rate": 3.1988065463853204e-06,
|
| 2458 |
+
"loss": 1.0902,
|
| 2459 |
+
"step": 1745
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.7665352606219886,
|
| 2463 |
+
"grad_norm": 0.26742133303345345,
|
| 2464 |
+
"learning_rate": 3.1429420491358696e-06,
|
| 2465 |
+
"loss": 1.1262,
|
| 2466 |
+
"step": 1750
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.7687253613666228,
|
| 2470 |
+
"grad_norm": 0.2599204528063472,
|
| 2471 |
+
"learning_rate": 3.0874785797778096e-06,
|
| 2472 |
+
"loss": 1.148,
|
| 2473 |
+
"step": 1755
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.7709154621112572,
|
| 2477 |
+
"grad_norm": 0.2566302858053561,
|
| 2478 |
+
"learning_rate": 3.0324193820343685e-06,
|
| 2479 |
+
"loss": 1.1109,
|
| 2480 |
+
"step": 1760
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.7731055628558914,
|
| 2484 |
+
"grad_norm": 0.260139567695262,
|
| 2485 |
+
"learning_rate": 2.977767675985377e-06,
|
| 2486 |
+
"loss": 1.1344,
|
| 2487 |
+
"step": 1765
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.7752956636005256,
|
| 2491 |
+
"grad_norm": 0.2683465626714785,
|
| 2492 |
+
"learning_rate": 2.9235266578789268e-06,
|
| 2493 |
+
"loss": 1.1246,
|
| 2494 |
+
"step": 1770
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.7774857643451599,
|
| 2498 |
+
"grad_norm": 0.2594022825090287,
|
| 2499 |
+
"learning_rate": 2.8696994999444614e-06,
|
| 2500 |
+
"loss": 1.1074,
|
| 2501 |
+
"step": 1775
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.7796758650897941,
|
| 2505 |
+
"grad_norm": 0.26202834015198223,
|
| 2506 |
+
"learning_rate": 2.816289350207249e-06,
|
| 2507 |
+
"loss": 1.0943,
|
| 2508 |
+
"step": 1780
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.7818659658344284,
|
| 2512 |
+
"grad_norm": 0.2581965218546036,
|
| 2513 |
+
"learning_rate": 2.76329933230425e-06,
|
| 2514 |
+
"loss": 1.1088,
|
| 2515 |
+
"step": 1785
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.7840560665790627,
|
| 2519 |
+
"grad_norm": 0.262190008267224,
|
| 2520 |
+
"learning_rate": 2.7107325453014687e-06,
|
| 2521 |
+
"loss": 1.1098,
|
| 2522 |
+
"step": 1790
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.7862461673236969,
|
| 2526 |
+
"grad_norm": 0.2534248970210751,
|
| 2527 |
+
"learning_rate": 2.658592063512684e-06,
|
| 2528 |
+
"loss": 1.0893,
|
| 2529 |
+
"step": 1795
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.7884362680683311,
|
| 2533 |
+
"grad_norm": 0.2633626852571572,
|
| 2534 |
+
"learning_rate": 2.6068809363196633e-06,
|
| 2535 |
+
"loss": 1.091,
|
| 2536 |
+
"step": 1800
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.7906263688129654,
|
| 2540 |
+
"grad_norm": 0.25921581241790814,
|
| 2541 |
+
"learning_rate": 2.5556021879938074e-06,
|
| 2542 |
+
"loss": 1.093,
|
| 2543 |
+
"step": 1805
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.7928164695575997,
|
| 2547 |
+
"grad_norm": 0.2621990361536461,
|
| 2548 |
+
"learning_rate": 2.5047588175192995e-06,
|
| 2549 |
+
"loss": 1.1254,
|
| 2550 |
+
"step": 1810
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.7950065703022339,
|
| 2554 |
+
"grad_norm": 0.2535710412898093,
|
| 2555 |
+
"learning_rate": 2.454353798417698e-06,
|
| 2556 |
+
"loss": 1.1072,
|
| 2557 |
+
"step": 1815
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.7971966710468682,
|
| 2561 |
+
"grad_norm": 0.25964995240654326,
|
| 2562 |
+
"learning_rate": 2.4043900785740436e-06,
|
| 2563 |
+
"loss": 1.1023,
|
| 2564 |
+
"step": 1820
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.7993867717915024,
|
| 2568 |
+
"grad_norm": 0.2539381357636022,
|
| 2569 |
+
"learning_rate": 2.354870580064439e-06,
|
| 2570 |
+
"loss": 1.1133,
|
| 2571 |
+
"step": 1825
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.8015768725361366,
|
| 2575 |
+
"grad_norm": 0.2575538966366128,
|
| 2576 |
+
"learning_rate": 2.305798198985173e-06,
|
| 2577 |
+
"loss": 1.0879,
|
| 2578 |
+
"step": 1830
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.803766973280771,
|
| 2582 |
+
"grad_norm": 0.25244614295211687,
|
| 2583 |
+
"learning_rate": 2.257175805283338e-06,
|
| 2584 |
+
"loss": 1.1137,
|
| 2585 |
+
"step": 1835
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.8059570740254052,
|
| 2589 |
+
"grad_norm": 0.2668071511501433,
|
| 2590 |
+
"learning_rate": 2.209006242588977e-06,
|
| 2591 |
+
"loss": 1.1187,
|
| 2592 |
+
"step": 1840
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.8081471747700394,
|
| 2596 |
+
"grad_norm": 0.26833063688258807,
|
| 2597 |
+
"learning_rate": 2.1612923280487883e-06,
|
| 2598 |
+
"loss": 1.1049,
|
| 2599 |
+
"step": 1845
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.8103372755146737,
|
| 2603 |
+
"grad_norm": 0.25994289950129074,
|
| 2604 |
+
"learning_rate": 2.114036852161361e-06,
|
| 2605 |
+
"loss": 1.0926,
|
| 2606 |
+
"step": 1850
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.8125273762593079,
|
| 2610 |
+
"grad_norm": 0.26330093138785376,
|
| 2611 |
+
"learning_rate": 2.0672425786139794e-06,
|
| 2612 |
+
"loss": 1.0885,
|
| 2613 |
+
"step": 1855
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.8147174770039421,
|
| 2617 |
+
"grad_norm": 0.2627529440350032,
|
| 2618 |
+
"learning_rate": 2.020912244120983e-06,
|
| 2619 |
+
"loss": 1.0902,
|
| 2620 |
+
"step": 1860
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.8169075777485765,
|
| 2624 |
+
"grad_norm": 0.25955955375887674,
|
| 2625 |
+
"learning_rate": 1.9750485582637245e-06,
|
| 2626 |
+
"loss": 1.127,
|
| 2627 |
+
"step": 1865
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.8190976784932107,
|
| 2631 |
+
"grad_norm": 0.2615144960394142,
|
| 2632 |
+
"learning_rate": 1.9296542033321008e-06,
|
| 2633 |
+
"loss": 1.1074,
|
| 2634 |
+
"step": 1870
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.8212877792378449,
|
| 2638 |
+
"grad_norm": 0.2650186451464187,
|
| 2639 |
+
"learning_rate": 1.8847318341676657e-06,
|
| 2640 |
+
"loss": 1.1234,
|
| 2641 |
+
"step": 1875
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.8234778799824792,
|
| 2645 |
+
"grad_norm": 0.2641320073264704,
|
| 2646 |
+
"learning_rate": 1.840284078008393e-06,
|
| 2647 |
+
"loss": 1.0957,
|
| 2648 |
+
"step": 1880
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.8256679807271134,
|
| 2652 |
+
"grad_norm": 0.2581562950846561,
|
| 2653 |
+
"learning_rate": 1.7963135343349914e-06,
|
| 2654 |
+
"loss": 1.1094,
|
| 2655 |
+
"step": 1885
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.8278580814717477,
|
| 2659 |
+
"grad_norm": 0.25731784586468565,
|
| 2660 |
+
"learning_rate": 1.7528227747189152e-06,
|
| 2661 |
+
"loss": 1.1086,
|
| 2662 |
+
"step": 1890
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.830048182216382,
|
| 2666 |
+
"grad_norm": 0.2618515550927689,
|
| 2667 |
+
"learning_rate": 1.7098143426719293e-06,
|
| 2668 |
+
"loss": 1.0984,
|
| 2669 |
+
"step": 1895
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.8322382829610162,
|
| 2673 |
+
"grad_norm": 0.25314788080692846,
|
| 2674 |
+
"learning_rate": 1.667290753497386e-06,
|
| 2675 |
+
"loss": 1.0965,
|
| 2676 |
+
"step": 1900
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.8344283837056504,
|
| 2680 |
+
"grad_norm": 0.25504674250436493,
|
| 2681 |
+
"learning_rate": 1.6252544941430982e-06,
|
| 2682 |
+
"loss": 1.0957,
|
| 2683 |
+
"step": 1905
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.8366184844502847,
|
| 2687 |
+
"grad_norm": 0.2559793997504819,
|
| 2688 |
+
"learning_rate": 1.5837080230559077e-06,
|
| 2689 |
+
"loss": 1.0947,
|
| 2690 |
+
"step": 1910
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.838808585194919,
|
| 2694 |
+
"grad_norm": 0.26055333090211424,
|
| 2695 |
+
"learning_rate": 1.5426537700378985e-06,
|
| 2696 |
+
"loss": 1.0961,
|
| 2697 |
+
"step": 1915
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.8409986859395532,
|
| 2701 |
+
"grad_norm": 0.2576679394012241,
|
| 2702 |
+
"learning_rate": 1.5020941361042885e-06,
|
| 2703 |
+
"loss": 1.0934,
|
| 2704 |
+
"step": 1920
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.8431887866841875,
|
| 2708 |
+
"grad_norm": 0.25828962040009074,
|
| 2709 |
+
"learning_rate": 1.4620314933430269e-06,
|
| 2710 |
+
"loss": 1.1121,
|
| 2711 |
+
"step": 1925
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.8453788874288217,
|
| 2715 |
+
"grad_norm": 0.25513013439947374,
|
| 2716 |
+
"learning_rate": 1.4224681847760346e-06,
|
| 2717 |
+
"loss": 1.1145,
|
| 2718 |
+
"step": 1930
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.8475689881734559,
|
| 2722 |
+
"grad_norm": 0.26643194015943783,
|
| 2723 |
+
"learning_rate": 1.383406524222206e-06,
|
| 2724 |
+
"loss": 1.1002,
|
| 2725 |
+
"step": 1935
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.8497590889180903,
|
| 2729 |
+
"grad_norm": 0.25674411294345,
|
| 2730 |
+
"learning_rate": 1.3448487961620637e-06,
|
| 2731 |
+
"loss": 1.1355,
|
| 2732 |
+
"step": 1940
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.8519491896627245,
|
| 2736 |
+
"grad_norm": 0.2605245623523021,
|
| 2737 |
+
"learning_rate": 1.3067972556041753e-06,
|
| 2738 |
+
"loss": 1.1207,
|
| 2739 |
+
"step": 1945
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.8541392904073587,
|
| 2743 |
+
"grad_norm": 0.2616395396378427,
|
| 2744 |
+
"learning_rate": 1.2692541279532466e-06,
|
| 2745 |
+
"loss": 1.1129,
|
| 2746 |
+
"step": 1950
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.856329391151993,
|
| 2750 |
+
"grad_norm": 0.2558570340754774,
|
| 2751 |
+
"learning_rate": 1.2322216088799955e-06,
|
| 2752 |
+
"loss": 1.1156,
|
| 2753 |
+
"step": 1955
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.8585194918966272,
|
| 2757 |
+
"grad_norm": 0.25398464154559003,
|
| 2758 |
+
"learning_rate": 1.1957018641927187e-06,
|
| 2759 |
+
"loss": 1.1137,
|
| 2760 |
+
"step": 1960
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.8607095926412615,
|
| 2764 |
+
"grad_norm": 0.2614386394260854,
|
| 2765 |
+
"learning_rate": 1.1596970297106458e-06,
|
| 2766 |
+
"loss": 1.1254,
|
| 2767 |
+
"step": 1965
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.8628996933858958,
|
| 2771 |
+
"grad_norm": 0.2620955323370706,
|
| 2772 |
+
"learning_rate": 1.1242092111390157e-06,
|
| 2773 |
+
"loss": 1.085,
|
| 2774 |
+
"step": 1970
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.86508979413053,
|
| 2778 |
+
"grad_norm": 0.2640807465997075,
|
| 2779 |
+
"learning_rate": 1.0892404839459269e-06,
|
| 2780 |
+
"loss": 1.1074,
|
| 2781 |
+
"step": 1975
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.8672798948751642,
|
| 2785 |
+
"grad_norm": 0.2580694157592811,
|
| 2786 |
+
"learning_rate": 1.054792893240969e-06,
|
| 2787 |
+
"loss": 1.1098,
|
| 2788 |
+
"step": 1980
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.8694699956197985,
|
| 2792 |
+
"grad_norm": 0.2651420557425447,
|
| 2793 |
+
"learning_rate": 1.0208684536555968e-06,
|
| 2794 |
+
"loss": 1.1275,
|
| 2795 |
+
"step": 1985
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.8716600963644328,
|
| 2799 |
+
"grad_norm": 0.262002558002591,
|
| 2800 |
+
"learning_rate": 9.874691492253242e-07,
|
| 2801 |
+
"loss": 1.1039,
|
| 2802 |
+
"step": 1990
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.873850197109067,
|
| 2806 |
+
"grad_norm": 0.25919863963901085,
|
| 2807 |
+
"learning_rate": 9.545969332736748e-07,
|
| 2808 |
+
"loss": 1.1195,
|
| 2809 |
+
"step": 1995
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.8760402978537013,
|
| 2813 |
+
"grad_norm": 0.2668241542423085,
|
| 2814 |
+
"learning_rate": 9.222537282979593e-07,
|
| 2815 |
+
"loss": 1.1004,
|
| 2816 |
+
"step": 2000
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 0.8782303985983355,
|
| 2820 |
+
"grad_norm": 0.2601150473608183,
|
| 2821 |
+
"learning_rate": 8.904414258568306e-07,
|
| 2822 |
+
"loss": 1.1027,
|
| 2823 |
+
"step": 2005
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 0.8804204993429697,
|
| 2827 |
+
"grad_norm": 0.2548383260510458,
|
| 2828 |
+
"learning_rate": 8.591618864596541e-07,
|
| 2829 |
+
"loss": 1.1051,
|
| 2830 |
+
"step": 2010
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 0.8826106000876041,
|
| 2834 |
+
"grad_norm": 0.2600128178676384,
|
| 2835 |
+
"learning_rate": 8.284169394577124e-07,
|
| 2836 |
+
"loss": 1.1236,
|
| 2837 |
+
"step": 2015
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 0.8848007008322383,
|
| 2841 |
+
"grad_norm": 0.26066874164079284,
|
| 2842 |
+
"learning_rate": 7.982083829372023e-07,
|
| 2843 |
+
"loss": 1.1242,
|
| 2844 |
+
"step": 2020
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.8869908015768725,
|
| 2848 |
+
"grad_norm": 0.25692211657826886,
|
| 2849 |
+
"learning_rate": 7.685379836140872e-07,
|
| 2850 |
+
"loss": 1.1406,
|
| 2851 |
+
"step": 2025
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 0.8891809023215068,
|
| 2855 |
+
"grad_norm": 0.26481424368233436,
|
| 2856 |
+
"learning_rate": 7.394074767307591e-07,
|
| 2857 |
+
"loss": 1.1313,
|
| 2858 |
+
"step": 2030
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 0.891371003066141,
|
| 2862 |
+
"grad_norm": 0.2618210555897095,
|
| 2863 |
+
"learning_rate": 7.10818565954573e-07,
|
| 2864 |
+
"loss": 1.1193,
|
| 2865 |
+
"step": 2035
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 0.8935611038107752,
|
| 2869 |
+
"grad_norm": 0.26537735037912497,
|
| 2870 |
+
"learning_rate": 6.827729232781943e-07,
|
| 2871 |
+
"loss": 1.1273,
|
| 2872 |
+
"step": 2040
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 0.8957512045554096,
|
| 2876 |
+
"grad_norm": 0.2565351465887861,
|
| 2877 |
+
"learning_rate": 6.552721889218194e-07,
|
| 2878 |
+
"loss": 1.1074,
|
| 2879 |
+
"step": 2045
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 0.8979413053000438,
|
| 2883 |
+
"grad_norm": 0.2641157557567522,
|
| 2884 |
+
"learning_rate": 6.283179712372489e-07,
|
| 2885 |
+
"loss": 1.1102,
|
| 2886 |
+
"step": 2050
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 0.900131406044678,
|
| 2890 |
+
"grad_norm": 0.258380894950019,
|
| 2891 |
+
"learning_rate": 6.019118466138285e-07,
|
| 2892 |
+
"loss": 1.0936,
|
| 2893 |
+
"step": 2055
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 0.9023215067893123,
|
| 2897 |
+
"grad_norm": 0.2621614154084412,
|
| 2898 |
+
"learning_rate": 5.760553593862506e-07,
|
| 2899 |
+
"loss": 1.0959,
|
| 2900 |
+
"step": 2060
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 0.9045116075339465,
|
| 2904 |
+
"grad_norm": 0.2560826189907596,
|
| 2905 |
+
"learning_rate": 5.507500217442341e-07,
|
| 2906 |
+
"loss": 1.1016,
|
| 2907 |
+
"step": 2065
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 0.9067017082785808,
|
| 2911 |
+
"grad_norm": 0.2590724071414488,
|
| 2912 |
+
"learning_rate": 5.259973136440921e-07,
|
| 2913 |
+
"loss": 1.1082,
|
| 2914 |
+
"step": 2070
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 0.9088918090232151,
|
| 2918 |
+
"grad_norm": 0.2573319556492272,
|
| 2919 |
+
"learning_rate": 5.017986827221733e-07,
|
| 2920 |
+
"loss": 1.1082,
|
| 2921 |
+
"step": 2075
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 0.9110819097678493,
|
| 2925 |
+
"grad_norm": 0.34516100541646555,
|
| 2926 |
+
"learning_rate": 4.781555442102015e-07,
|
| 2927 |
+
"loss": 1.1355,
|
| 2928 |
+
"step": 2080
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 0.9132720105124835,
|
| 2932 |
+
"grad_norm": 0.26180704480049366,
|
| 2933 |
+
"learning_rate": 4.5506928085250033e-07,
|
| 2934 |
+
"loss": 1.0871,
|
| 2935 |
+
"step": 2085
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 0.9154621112571178,
|
| 2939 |
+
"grad_norm": 0.25833323890857174,
|
| 2940 |
+
"learning_rate": 4.3254124282513634e-07,
|
| 2941 |
+
"loss": 1.1062,
|
| 2942 |
+
"step": 2090
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 0.9176522120017521,
|
| 2946 |
+
"grad_norm": 0.26187121720127504,
|
| 2947 |
+
"learning_rate": 4.1057274765694765e-07,
|
| 2948 |
+
"loss": 1.0939,
|
| 2949 |
+
"step": 2095
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 0.9198423127463863,
|
| 2953 |
+
"grad_norm": 0.2573692262543959,
|
| 2954 |
+
"learning_rate": 3.891650801524871e-07,
|
| 2955 |
+
"loss": 1.1074,
|
| 2956 |
+
"step": 2100
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 0.9220324134910206,
|
| 2960 |
+
"grad_norm": 0.2649955238574841,
|
| 2961 |
+
"learning_rate": 3.6831949231689203e-07,
|
| 2962 |
+
"loss": 1.1168,
|
| 2963 |
+
"step": 2105
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 0.9242225142356548,
|
| 2967 |
+
"grad_norm": 0.2553634372156294,
|
| 2968 |
+
"learning_rate": 3.4803720328264733e-07,
|
| 2969 |
+
"loss": 1.1168,
|
| 2970 |
+
"step": 2110
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 0.926412614980289,
|
| 2974 |
+
"grad_norm": 0.26728677917909055,
|
| 2975 |
+
"learning_rate": 3.283193992383049e-07,
|
| 2976 |
+
"loss": 1.1273,
|
| 2977 |
+
"step": 2115
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 0.9286027157249234,
|
| 2981 |
+
"grad_norm": 0.2607892654338355,
|
| 2982 |
+
"learning_rate": 3.091672333590923e-07,
|
| 2983 |
+
"loss": 1.0867,
|
| 2984 |
+
"step": 2120
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 0.9307928164695576,
|
| 2988 |
+
"grad_norm": 0.2527633048942072,
|
| 2989 |
+
"learning_rate": 2.905818257394799e-07,
|
| 2990 |
+
"loss": 1.1133,
|
| 2991 |
+
"step": 2125
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 0.9329829172141918,
|
| 2995 |
+
"grad_norm": 0.25568945394844267,
|
| 2996 |
+
"learning_rate": 2.725642633276737e-07,
|
| 2997 |
+
"loss": 1.1098,
|
| 2998 |
+
"step": 2130
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 0.9351730179588261,
|
| 3002 |
+
"grad_norm": 0.2597970875903515,
|
| 3003 |
+
"learning_rate": 2.5511559986204247e-07,
|
| 3004 |
+
"loss": 1.1184,
|
| 3005 |
+
"step": 2135
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 0.9373631187034603,
|
| 3009 |
+
"grad_norm": 0.26327747871941953,
|
| 3010 |
+
"learning_rate": 2.3823685580949273e-07,
|
| 3011 |
+
"loss": 1.0965,
|
| 3012 |
+
"step": 2140
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 0.9395532194480947,
|
| 3016 |
+
"grad_norm": 0.2607010813003693,
|
| 3017 |
+
"learning_rate": 2.219290183057865e-07,
|
| 3018 |
+
"loss": 1.1352,
|
| 3019 |
+
"step": 2145
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 0.9417433201927289,
|
| 3023 |
+
"grad_norm": 0.2530017456095944,
|
| 3024 |
+
"learning_rate": 2.0619304109781636e-07,
|
| 3025 |
+
"loss": 1.1012,
|
| 3026 |
+
"step": 2150
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 0.9439334209373631,
|
| 3030 |
+
"grad_norm": 0.2519870547260623,
|
| 3031 |
+
"learning_rate": 1.9102984448781337e-07,
|
| 3032 |
+
"loss": 1.0752,
|
| 3033 |
+
"step": 2155
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.9461235216819974,
|
| 3037 |
+
"grad_norm": 0.2530254165097107,
|
| 3038 |
+
"learning_rate": 1.764403152795402e-07,
|
| 3039 |
+
"loss": 1.0934,
|
| 3040 |
+
"step": 2160
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 0.9483136224266316,
|
| 3044 |
+
"grad_norm": 0.25397273342292553,
|
| 3045 |
+
"learning_rate": 1.6242530672641143e-07,
|
| 3046 |
+
"loss": 1.1219,
|
| 3047 |
+
"step": 2165
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 0.9505037231712659,
|
| 3051 |
+
"grad_norm": 0.26174974744061347,
|
| 3052 |
+
"learning_rate": 1.4898563848160464e-07,
|
| 3053 |
+
"loss": 1.1473,
|
| 3054 |
+
"step": 2170
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 0.9526938239159002,
|
| 3058 |
+
"grad_norm": 0.24955513279870836,
|
| 3059 |
+
"learning_rate": 1.3612209655011866e-07,
|
| 3060 |
+
"loss": 1.1195,
|
| 3061 |
+
"step": 2175
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 0.9548839246605344,
|
| 3065 |
+
"grad_norm": 0.2583751624724583,
|
| 3066 |
+
"learning_rate": 1.2383543324280044e-07,
|
| 3067 |
+
"loss": 1.1189,
|
| 3068 |
+
"step": 2180
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 0.9570740254051686,
|
| 3072 |
+
"grad_norm": 0.2588289009510903,
|
| 3073 |
+
"learning_rate": 1.1212636713235581e-07,
|
| 3074 |
+
"loss": 1.1416,
|
| 3075 |
+
"step": 2185
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 0.9592641261498029,
|
| 3079 |
+
"grad_norm": 0.2583550968959351,
|
| 3080 |
+
"learning_rate": 1.0099558301132072e-07,
|
| 3081 |
+
"loss": 1.1254,
|
| 3082 |
+
"step": 2190
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.9614542268944372,
|
| 3086 |
+
"grad_norm": 0.2527293991159154,
|
| 3087 |
+
"learning_rate": 9.044373185200906e-08,
|
| 3088 |
+
"loss": 1.1031,
|
| 3089 |
+
"step": 2195
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.9636443276390714,
|
| 3093 |
+
"grad_norm": 0.2521840148118415,
|
| 3094 |
+
"learning_rate": 8.047143076844399e-08,
|
| 3095 |
+
"loss": 1.0783,
|
| 3096 |
+
"step": 2200
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.9658344283837057,
|
| 3100 |
+
"grad_norm": 0.2606898444853039,
|
| 3101 |
+
"learning_rate": 7.10792629802659e-08,
|
| 3102 |
+
"loss": 1.1113,
|
| 3103 |
+
"step": 2205
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 0.9680245291283399,
|
| 3107 |
+
"grad_norm": 0.25532227510506406,
|
| 3108 |
+
"learning_rate": 6.226777777862513e-08,
|
| 3109 |
+
"loss": 1.1293,
|
| 3110 |
+
"step": 2210
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 0.9702146298729741,
|
| 3114 |
+
"grad_norm": 0.2565349165845633,
|
| 3115 |
+
"learning_rate": 5.403749049405438e-08,
|
| 3116 |
+
"loss": 1.1025,
|
| 3117 |
+
"step": 2215
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 0.9724047306176085,
|
| 3121 |
+
"grad_norm": 0.25458880932233896,
|
| 3122 |
+
"learning_rate": 4.638888246633166e-08,
|
| 3123 |
+
"loss": 1.1121,
|
| 3124 |
+
"step": 2220
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.9745948313622427,
|
| 3128 |
+
"grad_norm": 0.25398250551108936,
|
| 3129 |
+
"learning_rate": 3.932240101633178e-08,
|
| 3130 |
+
"loss": 1.1023,
|
| 3131 |
+
"step": 2225
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 0.9767849321068769,
|
| 3135 |
+
"grad_norm": 0.25422167947403135,
|
| 3136 |
+
"learning_rate": 3.283845941985941e-08,
|
| 3137 |
+
"loss": 1.0758,
|
| 3138 |
+
"step": 2230
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 0.9789750328515112,
|
| 3142 |
+
"grad_norm": 0.2569295535003476,
|
| 3143 |
+
"learning_rate": 2.693743688348627e-08,
|
| 3144 |
+
"loss": 1.1168,
|
| 3145 |
+
"step": 2235
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.9811651335961454,
|
| 3149 |
+
"grad_norm": 0.2611140820134084,
|
| 3150 |
+
"learning_rate": 2.1619678522366573e-08,
|
| 3151 |
+
"loss": 1.098,
|
| 3152 |
+
"step": 2240
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 0.9833552343407796,
|
| 3156 |
+
"grad_norm": 0.2585042183570841,
|
| 3157 |
+
"learning_rate": 1.68854953400599e-08,
|
| 3158 |
+
"loss": 1.1182,
|
| 3159 |
+
"step": 2245
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.985545335085414,
|
| 3163 |
+
"grad_norm": 0.2699764722438751,
|
| 3164 |
+
"learning_rate": 1.2735164210337935e-08,
|
| 3165 |
+
"loss": 1.1629,
|
| 3166 |
+
"step": 2250
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.9877354358300482,
|
| 3170 |
+
"grad_norm": 0.2604976252616973,
|
| 3171 |
+
"learning_rate": 9.168927860994104e-09,
|
| 3172 |
+
"loss": 1.1049,
|
| 3173 |
+
"step": 2255
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 0.9899255365746824,
|
| 3177 |
+
"grad_norm": 0.2570709645581014,
|
| 3178 |
+
"learning_rate": 6.186994859646023e-09,
|
| 3179 |
+
"loss": 1.0938,
|
| 3180 |
+
"step": 2260
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.9921156373193167,
|
| 3184 |
+
"grad_norm": 0.2534472927063691,
|
| 3185 |
+
"learning_rate": 3.7895396015374955e-09,
|
| 3186 |
+
"loss": 1.1094,
|
| 3187 |
+
"step": 2265
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 0.994305738063951,
|
| 3191 |
+
"grad_norm": 0.2532795250404029,
|
| 3192 |
+
"learning_rate": 1.9767022993444353e-09,
|
| 3193 |
+
"loss": 1.1031,
|
| 3194 |
+
"step": 2270
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 0.9964958388085852,
|
| 3198 |
+
"grad_norm": 0.25604334514879273,
|
| 3199 |
+
"learning_rate": 7.485889749658803e-10,
|
| 3200 |
+
"loss": 1.1129,
|
| 3201 |
+
"step": 2275
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.9986859395532195,
|
| 3205 |
+
"grad_norm": 0.26226473069552947,
|
| 3206 |
+
"learning_rate": 1.0527145333227851e-10,
|
| 3207 |
+
"loss": 1.1066,
|
| 3208 |
+
"step": 2280
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 1.0,
|
| 3212 |
+
"step": 2283,
|
| 3213 |
+
"total_flos": 216025446481920.0,
|
| 3214 |
+
"train_loss": 1.2513469975087603,
|
| 3215 |
+
"train_runtime": 7749.9665,
|
| 3216 |
+
"train_samples_per_second": 18.846,
|
| 3217 |
+
"train_steps_per_second": 0.295
|
| 3218 |
+
}
|
| 3219 |
+
],
|
| 3220 |
+
"logging_steps": 5,
|
| 3221 |
+
"max_steps": 2283,
|
| 3222 |
+
"num_input_tokens_seen": 0,
|
| 3223 |
+
"num_train_epochs": 1,
|
| 3224 |
+
"save_steps": 500,
|
| 3225 |
+
"stateful_callbacks": {
|
| 3226 |
+
"TrainerControl": {
|
| 3227 |
+
"args": {
|
| 3228 |
+
"should_epoch_stop": false,
|
| 3229 |
+
"should_evaluate": false,
|
| 3230 |
+
"should_log": false,
|
| 3231 |
+
"should_save": false,
|
| 3232 |
+
"should_training_stop": false
|
| 3233 |
+
},
|
| 3234 |
+
"attributes": {}
|
| 3235 |
+
}
|
| 3236 |
+
},
|
| 3237 |
+
"total_flos": 216025446481920.0,
|
| 3238 |
+
"train_batch_size": 16,
|
| 3239 |
+
"trial_name": null,
|
| 3240 |
+
"trial_params": null
|
| 3241 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5bc5d8a1b59c97e707deff93ff7a253b0217f3d649a60f0f9d2a0c03c0f7f856
|
| 3 |
+
size 6904
|