Model save
Browse files- README.md +69 -0
- added_tokens.json +5 -0
- all_results.json +9 -0
- config.json +28 -0
- generation_config.json +6 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- special_tokens_map.json +20 -0
- tokenizer.json +0 -0
- tokenizer_config.json +43 -0
- train_results.json +9 -0
- trainer_state.json +2731 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: Qwen/Qwen2-0.5B
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- sft
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- generator
|
| 10 |
+
model-index:
|
| 11 |
+
- name: qwen2-0.5b-sft
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
+
|
| 18 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/zhaoyang1/huggingface/runs/h6gigsup)
|
| 19 |
+
# qwen2-0.5b-sft
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [Qwen/Qwen2-0.5B](https://huggingface.co/Qwen/Qwen2-0.5B) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 1.5079
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 2e-05
|
| 43 |
+
- train_batch_size: 8
|
| 44 |
+
- eval_batch_size: 8
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 2
|
| 48 |
+
- gradient_accumulation_steps: 8
|
| 49 |
+
- total_train_batch_size: 128
|
| 50 |
+
- total_eval_batch_size: 16
|
| 51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 52 |
+
- lr_scheduler_type: cosine
|
| 53 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 54 |
+
- num_epochs: 1
|
| 55 |
+
- mixed_precision_training: Native AMP
|
| 56 |
+
|
| 57 |
+
### Training results
|
| 58 |
+
|
| 59 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 60 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 61 |
+
| 1.5055 | 0.9997 | 1911 | 1.5079 |
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
### Framework versions
|
| 65 |
+
|
| 66 |
+
- Transformers 4.42.0
|
| 67 |
+
- Pytorch 2.3.0+cu121
|
| 68 |
+
- Datasets 2.19.1
|
| 69 |
+
- Tokenizers 0.19.1
|
added_tokens.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<|endoftext|>": 151643,
|
| 3 |
+
"<|im_end|>": 151645,
|
| 4 |
+
"<|im_start|>": 151644
|
| 5 |
+
}
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.999738425320429,
|
| 3 |
+
"total_flos": 107512363745280.0,
|
| 4 |
+
"train_loss": 1.5265680355653908,
|
| 5 |
+
"train_runtime": 15553.3872,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 15.73,
|
| 8 |
+
"train_steps_per_second": 0.123
|
| 9 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Qwen/Qwen2-0.5B",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Qwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"max_position_embeddings": 131072,
|
| 14 |
+
"max_window_layers": 24,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 14,
|
| 17 |
+
"num_hidden_layers": 24,
|
| 18 |
+
"num_key_value_heads": 2,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 1000000.0,
|
| 21 |
+
"sliding_window": 131072,
|
| 22 |
+
"tie_word_embeddings": true,
|
| 23 |
+
"torch_dtype": "float16",
|
| 24 |
+
"transformers_version": "4.42.0",
|
| 25 |
+
"use_cache": false,
|
| 26 |
+
"use_sliding_window": false,
|
| 27 |
+
"vocab_size": 151936
|
| 28 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.42.0"
|
| 6 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b385b5be4285f65f53e7bbddf58fecb4441ec7b33cdee2abd225ff77e34f2f5a
|
| 3 |
+
size 988097536
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>"
|
| 5 |
+
],
|
| 6 |
+
"eos_token": {
|
| 7 |
+
"content": "<|endoftext|>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false
|
| 12 |
+
},
|
| 13 |
+
"pad_token": {
|
| 14 |
+
"content": "<|endoftext|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false
|
| 19 |
+
}
|
| 20 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"151643": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"151644": {
|
| 13 |
+
"content": "<|im_start|>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"151645": {
|
| 21 |
+
"content": "<|im_end|>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"additional_special_tokens": [
|
| 30 |
+
"<|im_start|>",
|
| 31 |
+
"<|im_end|>"
|
| 32 |
+
],
|
| 33 |
+
"bos_token": null,
|
| 34 |
+
"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 %}",
|
| 35 |
+
"clean_up_tokenization_spaces": false,
|
| 36 |
+
"eos_token": "<|endoftext|>",
|
| 37 |
+
"errors": "replace",
|
| 38 |
+
"model_max_length": 32768,
|
| 39 |
+
"pad_token": "<|endoftext|>",
|
| 40 |
+
"split_special_tokens": false,
|
| 41 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 42 |
+
"unk_token": null
|
| 43 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.999738425320429,
|
| 3 |
+
"total_flos": 107512363745280.0,
|
| 4 |
+
"train_loss": 1.5265680355653908,
|
| 5 |
+
"train_runtime": 15553.3872,
|
| 6 |
+
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 15.73,
|
| 8 |
+
"train_steps_per_second": 0.123
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2731 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.999738425320429,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1911,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.000523149359142035,
|
| 13 |
+
"grad_norm": 8.50498175378257,
|
| 14 |
+
"learning_rate": 1.0416666666666667e-07,
|
| 15 |
+
"loss": 1.8225,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0026157467957101752,
|
| 20 |
+
"grad_norm": 8.7743077926545,
|
| 21 |
+
"learning_rate": 5.208333333333334e-07,
|
| 22 |
+
"loss": 1.8216,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0052314935914203504,
|
| 27 |
+
"grad_norm": 6.22485757813216,
|
| 28 |
+
"learning_rate": 1.0416666666666667e-06,
|
| 29 |
+
"loss": 1.7899,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.007847240387130526,
|
| 34 |
+
"grad_norm": 3.597307337373435,
|
| 35 |
+
"learning_rate": 1.5625e-06,
|
| 36 |
+
"loss": 1.709,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.010462987182840701,
|
| 41 |
+
"grad_norm": 1.81745766320769,
|
| 42 |
+
"learning_rate": 2.0833333333333334e-06,
|
| 43 |
+
"loss": 1.6422,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.013078733978550876,
|
| 48 |
+
"grad_norm": 1.3746421829291462,
|
| 49 |
+
"learning_rate": 2.604166666666667e-06,
|
| 50 |
+
"loss": 1.6256,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.01569448077426105,
|
| 55 |
+
"grad_norm": 1.1330404891483954,
|
| 56 |
+
"learning_rate": 3.125e-06,
|
| 57 |
+
"loss": 1.5745,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.01831022756997123,
|
| 62 |
+
"grad_norm": 1.070492621388286,
|
| 63 |
+
"learning_rate": 3.6458333333333333e-06,
|
| 64 |
+
"loss": 1.5935,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.020925974365681402,
|
| 69 |
+
"grad_norm": 1.113838910446101,
|
| 70 |
+
"learning_rate": 4.166666666666667e-06,
|
| 71 |
+
"loss": 1.5885,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.02354172116139158,
|
| 76 |
+
"grad_norm": 1.1078938402286853,
|
| 77 |
+
"learning_rate": 4.6875000000000004e-06,
|
| 78 |
+
"loss": 1.5775,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.026157467957101752,
|
| 83 |
+
"grad_norm": 1.0740497525043167,
|
| 84 |
+
"learning_rate": 5.208333333333334e-06,
|
| 85 |
+
"loss": 1.5596,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.02877321475281193,
|
| 90 |
+
"grad_norm": 1.0471464348725732,
|
| 91 |
+
"learning_rate": 5.729166666666667e-06,
|
| 92 |
+
"loss": 1.5727,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0313889615485221,
|
| 97 |
+
"grad_norm": 1.028076314326798,
|
| 98 |
+
"learning_rate": 6.25e-06,
|
| 99 |
+
"loss": 1.5529,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.034004708344232276,
|
| 104 |
+
"grad_norm": 1.1180193049504716,
|
| 105 |
+
"learning_rate": 6.770833333333334e-06,
|
| 106 |
+
"loss": 1.5908,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.03662045513994246,
|
| 111 |
+
"grad_norm": 1.062786102742324,
|
| 112 |
+
"learning_rate": 7.291666666666667e-06,
|
| 113 |
+
"loss": 1.571,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.03923620193565263,
|
| 118 |
+
"grad_norm": 1.0775438785201366,
|
| 119 |
+
"learning_rate": 7.8125e-06,
|
| 120 |
+
"loss": 1.5673,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.041851948731362804,
|
| 125 |
+
"grad_norm": 1.0138280109903437,
|
| 126 |
+
"learning_rate": 8.333333333333334e-06,
|
| 127 |
+
"loss": 1.5847,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04446769552707298,
|
| 132 |
+
"grad_norm": 1.0676890556901109,
|
| 133 |
+
"learning_rate": 8.854166666666667e-06,
|
| 134 |
+
"loss": 1.5777,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04708344232278316,
|
| 139 |
+
"grad_norm": 1.049439971963735,
|
| 140 |
+
"learning_rate": 9.375000000000001e-06,
|
| 141 |
+
"loss": 1.5863,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.04969918911849333,
|
| 146 |
+
"grad_norm": 1.0508764062113662,
|
| 147 |
+
"learning_rate": 9.895833333333334e-06,
|
| 148 |
+
"loss": 1.5648,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.052314935914203504,
|
| 153 |
+
"grad_norm": 1.0774966933407557,
|
| 154 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 155 |
+
"loss": 1.567,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.05493068270991368,
|
| 160 |
+
"grad_norm": 1.0820816348442155,
|
| 161 |
+
"learning_rate": 1.0937500000000002e-05,
|
| 162 |
+
"loss": 1.5794,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.05754642950562386,
|
| 167 |
+
"grad_norm": 1.1054344703983228,
|
| 168 |
+
"learning_rate": 1.1458333333333333e-05,
|
| 169 |
+
"loss": 1.5974,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.06016217630133403,
|
| 174 |
+
"grad_norm": 1.0979610972425926,
|
| 175 |
+
"learning_rate": 1.1979166666666669e-05,
|
| 176 |
+
"loss": 1.524,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.0627779230970442,
|
| 181 |
+
"grad_norm": 1.1061119519108826,
|
| 182 |
+
"learning_rate": 1.25e-05,
|
| 183 |
+
"loss": 1.557,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.06539366989275439,
|
| 188 |
+
"grad_norm": 1.0775672295340581,
|
| 189 |
+
"learning_rate": 1.3020833333333334e-05,
|
| 190 |
+
"loss": 1.5633,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.06800941668846455,
|
| 195 |
+
"grad_norm": 1.1271622797436758,
|
| 196 |
+
"learning_rate": 1.3541666666666668e-05,
|
| 197 |
+
"loss": 1.5657,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.07062516348417473,
|
| 202 |
+
"grad_norm": 1.1375583257158841,
|
| 203 |
+
"learning_rate": 1.4062500000000001e-05,
|
| 204 |
+
"loss": 1.5656,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.07324091027988491,
|
| 209 |
+
"grad_norm": 1.1021606611565173,
|
| 210 |
+
"learning_rate": 1.4583333333333333e-05,
|
| 211 |
+
"loss": 1.5814,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.07585665707559508,
|
| 216 |
+
"grad_norm": 1.142293742855133,
|
| 217 |
+
"learning_rate": 1.5104166666666668e-05,
|
| 218 |
+
"loss": 1.5442,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07847240387130526,
|
| 223 |
+
"grad_norm": 1.1243167952517432,
|
| 224 |
+
"learning_rate": 1.5625e-05,
|
| 225 |
+
"loss": 1.548,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.08108815066701543,
|
| 230 |
+
"grad_norm": 1.1145531449905433,
|
| 231 |
+
"learning_rate": 1.6145833333333334e-05,
|
| 232 |
+
"loss": 1.544,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.08370389746272561,
|
| 237 |
+
"grad_norm": 1.0668270140196932,
|
| 238 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 239 |
+
"loss": 1.5385,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.08631964425843579,
|
| 244 |
+
"grad_norm": 1.1382110886452066,
|
| 245 |
+
"learning_rate": 1.71875e-05,
|
| 246 |
+
"loss": 1.5531,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.08893539105414595,
|
| 251 |
+
"grad_norm": 1.1117117441964515,
|
| 252 |
+
"learning_rate": 1.7708333333333335e-05,
|
| 253 |
+
"loss": 1.5526,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.09155113784985613,
|
| 258 |
+
"grad_norm": 1.0914528565872428,
|
| 259 |
+
"learning_rate": 1.8229166666666668e-05,
|
| 260 |
+
"loss": 1.5474,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.09416688464556631,
|
| 265 |
+
"grad_norm": 1.1858948011413253,
|
| 266 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 267 |
+
"loss": 1.551,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.09678263144127648,
|
| 272 |
+
"grad_norm": 1.1851194053435052,
|
| 273 |
+
"learning_rate": 1.9270833333333335e-05,
|
| 274 |
+
"loss": 1.5886,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.09939837823698666,
|
| 279 |
+
"grad_norm": 1.1262050443285683,
|
| 280 |
+
"learning_rate": 1.979166666666667e-05,
|
| 281 |
+
"loss": 1.5856,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.10201412503269683,
|
| 286 |
+
"grad_norm": 1.1136720509809253,
|
| 287 |
+
"learning_rate": 1.9999849699848662e-05,
|
| 288 |
+
"loss": 1.5626,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.10462987182840701,
|
| 293 |
+
"grad_norm": 1.072905424218808,
|
| 294 |
+
"learning_rate": 1.9998931215285354e-05,
|
| 295 |
+
"loss": 1.5521,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.10724561862411719,
|
| 300 |
+
"grad_norm": 1.1325042806937766,
|
| 301 |
+
"learning_rate": 1.9997177822843044e-05,
|
| 302 |
+
"loss": 1.5271,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.10986136541982736,
|
| 307 |
+
"grad_norm": 1.1034993768216363,
|
| 308 |
+
"learning_rate": 1.99945896689295e-05,
|
| 309 |
+
"loss": 1.5652,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.11247711221553754,
|
| 314 |
+
"grad_norm": 1.1367075404724407,
|
| 315 |
+
"learning_rate": 1.999116696965483e-05,
|
| 316 |
+
"loss": 1.5821,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.11509285901124772,
|
| 321 |
+
"grad_norm": 1.1281743008741738,
|
| 322 |
+
"learning_rate": 1.9986910010813433e-05,
|
| 323 |
+
"loss": 1.5624,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.11770860580695788,
|
| 328 |
+
"grad_norm": 1.1285921047508072,
|
| 329 |
+
"learning_rate": 1.998181914786015e-05,
|
| 330 |
+
"loss": 1.5608,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.12032435260266806,
|
| 335 |
+
"grad_norm": 1.0898629450432604,
|
| 336 |
+
"learning_rate": 1.997589480588057e-05,
|
| 337 |
+
"loss": 1.5758,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.12294009939837823,
|
| 342 |
+
"grad_norm": 1.1016685122122822,
|
| 343 |
+
"learning_rate": 1.9969137479555553e-05,
|
| 344 |
+
"loss": 1.5711,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.1255558461940884,
|
| 349 |
+
"grad_norm": 1.106453358196598,
|
| 350 |
+
"learning_rate": 1.996154773311989e-05,
|
| 351 |
+
"loss": 1.5631,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.12817159298979858,
|
| 356 |
+
"grad_norm": 1.1289498853101012,
|
| 357 |
+
"learning_rate": 1.995312620031523e-05,
|
| 358 |
+
"loss": 1.5219,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.13078733978550877,
|
| 363 |
+
"grad_norm": 1.1117110487338868,
|
| 364 |
+
"learning_rate": 1.9943873584337137e-05,
|
| 365 |
+
"loss": 1.5467,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.13340308658121894,
|
| 370 |
+
"grad_norm": 1.0796977913516912,
|
| 371 |
+
"learning_rate": 1.993379065777639e-05,
|
| 372 |
+
"loss": 1.5436,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.1360188333769291,
|
| 377 |
+
"grad_norm": 1.0981973791880635,
|
| 378 |
+
"learning_rate": 1.9922878262554445e-05,
|
| 379 |
+
"loss": 1.57,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.1386345801726393,
|
| 384 |
+
"grad_norm": 1.1583459064794652,
|
| 385 |
+
"learning_rate": 1.9911137309853164e-05,
|
| 386 |
+
"loss": 1.5545,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.14125032696834947,
|
| 391 |
+
"grad_norm": 1.0834355257389907,
|
| 392 |
+
"learning_rate": 1.9898568780038725e-05,
|
| 393 |
+
"loss": 1.5428,
|
| 394 |
+
"step": 270
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.14386607376405963,
|
| 398 |
+
"grad_norm": 1.1278027967007365,
|
| 399 |
+
"learning_rate": 1.9885173722579748e-05,
|
| 400 |
+
"loss": 1.5819,
|
| 401 |
+
"step": 275
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.14648182055976983,
|
| 405 |
+
"grad_norm": 1.0480209335705715,
|
| 406 |
+
"learning_rate": 1.9870953255959673e-05,
|
| 407 |
+
"loss": 1.5638,
|
| 408 |
+
"step": 280
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.14909756735548,
|
| 412 |
+
"grad_norm": 1.0857229449423735,
|
| 413 |
+
"learning_rate": 1.9855908567583374e-05,
|
| 414 |
+
"loss": 1.5513,
|
| 415 |
+
"step": 285
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.15171331415119016,
|
| 419 |
+
"grad_norm": 1.1067390682161593,
|
| 420 |
+
"learning_rate": 1.9840040913677997e-05,
|
| 421 |
+
"loss": 1.5337,
|
| 422 |
+
"step": 290
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.15432906094690035,
|
| 426 |
+
"grad_norm": 1.1037028207471695,
|
| 427 |
+
"learning_rate": 1.9823351619188084e-05,
|
| 428 |
+
"loss": 1.5897,
|
| 429 |
+
"step": 295
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.15694480774261052,
|
| 433 |
+
"grad_norm": 1.1201595290266912,
|
| 434 |
+
"learning_rate": 1.9805842077664913e-05,
|
| 435 |
+
"loss": 1.5425,
|
| 436 |
+
"step": 300
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.1595605545383207,
|
| 440 |
+
"grad_norm": 1.1636927375313533,
|
| 441 |
+
"learning_rate": 1.9787513751150165e-05,
|
| 442 |
+
"loss": 1.5542,
|
| 443 |
+
"step": 305
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.16217630133403085,
|
| 447 |
+
"grad_norm": 1.1700913216722832,
|
| 448 |
+
"learning_rate": 1.9768368170053836e-05,
|
| 449 |
+
"loss": 1.5563,
|
| 450 |
+
"step": 310
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.16479204812974105,
|
| 454 |
+
"grad_norm": 1.0665788517723835,
|
| 455 |
+
"learning_rate": 1.974840693302643e-05,
|
| 456 |
+
"loss": 1.5142,
|
| 457 |
+
"step": 315
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.16740779492545121,
|
| 461 |
+
"grad_norm": 1.159393707359845,
|
| 462 |
+
"learning_rate": 1.9727631706825497e-05,
|
| 463 |
+
"loss": 1.5603,
|
| 464 |
+
"step": 320
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.17002354172116138,
|
| 468 |
+
"grad_norm": 1.0550940360571934,
|
| 469 |
+
"learning_rate": 1.9706044226176446e-05,
|
| 470 |
+
"loss": 1.5442,
|
| 471 |
+
"step": 325
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.17263928851687158,
|
| 475 |
+
"grad_norm": 1.0709974323889713,
|
| 476 |
+
"learning_rate": 1.96836462936277e-05,
|
| 477 |
+
"loss": 1.5792,
|
| 478 |
+
"step": 330
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.17525503531258174,
|
| 482 |
+
"grad_norm": 1.088358661804479,
|
| 483 |
+
"learning_rate": 1.966043977940018e-05,
|
| 484 |
+
"loss": 1.5488,
|
| 485 |
+
"step": 335
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.1778707821082919,
|
| 489 |
+
"grad_norm": 1.020959330259616,
|
| 490 |
+
"learning_rate": 1.963642662123115e-05,
|
| 491 |
+
"loss": 1.5588,
|
| 492 |
+
"step": 340
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.1804865289040021,
|
| 496 |
+
"grad_norm": 1.098941584046506,
|
| 497 |
+
"learning_rate": 1.9611608824212395e-05,
|
| 498 |
+
"loss": 1.561,
|
| 499 |
+
"step": 345
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.18310227569971227,
|
| 503 |
+
"grad_norm": 1.1624909317107024,
|
| 504 |
+
"learning_rate": 1.9585988460622838e-05,
|
| 505 |
+
"loss": 1.5388,
|
| 506 |
+
"step": 350
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.18571802249542244,
|
| 510 |
+
"grad_norm": 1.0642789055045656,
|
| 511 |
+
"learning_rate": 1.9559567669755455e-05,
|
| 512 |
+
"loss": 1.5635,
|
| 513 |
+
"step": 355
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.18833376929113263,
|
| 517 |
+
"grad_norm": 1.08594425395068,
|
| 518 |
+
"learning_rate": 1.953234865773869e-05,
|
| 519 |
+
"loss": 1.5684,
|
| 520 |
+
"step": 360
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.1909495160868428,
|
| 524 |
+
"grad_norm": 1.1163677451407563,
|
| 525 |
+
"learning_rate": 1.9504333697352203e-05,
|
| 526 |
+
"loss": 1.5756,
|
| 527 |
+
"step": 365
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.19356526288255296,
|
| 531 |
+
"grad_norm": 1.0995787299596045,
|
| 532 |
+
"learning_rate": 1.9475525127837128e-05,
|
| 533 |
+
"loss": 1.5339,
|
| 534 |
+
"step": 370
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.19618100967826316,
|
| 538 |
+
"grad_norm": 1.0651385748289264,
|
| 539 |
+
"learning_rate": 1.9445925354700733e-05,
|
| 540 |
+
"loss": 1.5477,
|
| 541 |
+
"step": 375
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.19879675647397332,
|
| 545 |
+
"grad_norm": 1.1694817177791392,
|
| 546 |
+
"learning_rate": 1.9415536849515556e-05,
|
| 547 |
+
"loss": 1.5565,
|
| 548 |
+
"step": 380
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.2014125032696835,
|
| 552 |
+
"grad_norm": 1.1178248815532814,
|
| 553 |
+
"learning_rate": 1.938436214971303e-05,
|
| 554 |
+
"loss": 1.5474,
|
| 555 |
+
"step": 385
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.20402825006539366,
|
| 559 |
+
"grad_norm": 1.1467361846843396,
|
| 560 |
+
"learning_rate": 1.9352403858371618e-05,
|
| 561 |
+
"loss": 1.5521,
|
| 562 |
+
"step": 390
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.20664399686110385,
|
| 566 |
+
"grad_norm": 1.0833788576502061,
|
| 567 |
+
"learning_rate": 1.9319664643999454e-05,
|
| 568 |
+
"loss": 1.5607,
|
| 569 |
+
"step": 395
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.20925974365681402,
|
| 573 |
+
"grad_norm": 1.0463780529730196,
|
| 574 |
+
"learning_rate": 1.928614724031151e-05,
|
| 575 |
+
"loss": 1.5569,
|
| 576 |
+
"step": 400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.21187549045252418,
|
| 580 |
+
"grad_norm": 1.0984173620728173,
|
| 581 |
+
"learning_rate": 1.9251854446001358e-05,
|
| 582 |
+
"loss": 1.5697,
|
| 583 |
+
"step": 405
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.21449123724823438,
|
| 587 |
+
"grad_norm": 1.019288658733137,
|
| 588 |
+
"learning_rate": 1.921678912450745e-05,
|
| 589 |
+
"loss": 1.5472,
|
| 590 |
+
"step": 410
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.21710698404394455,
|
| 594 |
+
"grad_norm": 1.1331134806077368,
|
| 595 |
+
"learning_rate": 1.9180954203774032e-05,
|
| 596 |
+
"loss": 1.5734,
|
| 597 |
+
"step": 415
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2197227308396547,
|
| 601 |
+
"grad_norm": 1.062098050174779,
|
| 602 |
+
"learning_rate": 1.9144352676006683e-05,
|
| 603 |
+
"loss": 1.5266,
|
| 604 |
+
"step": 420
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.2223384776353649,
|
| 608 |
+
"grad_norm": 1.0398249601115788,
|
| 609 |
+
"learning_rate": 1.9106987597422434e-05,
|
| 610 |
+
"loss": 1.5495,
|
| 611 |
+
"step": 425
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.22495422443107507,
|
| 615 |
+
"grad_norm": 1.045791238747256,
|
| 616 |
+
"learning_rate": 1.9068862087994596e-05,
|
| 617 |
+
"loss": 1.5646,
|
| 618 |
+
"step": 430
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.22756997122678524,
|
| 622 |
+
"grad_norm": 1.1332298558605771,
|
| 623 |
+
"learning_rate": 1.902997933119223e-05,
|
| 624 |
+
"loss": 1.5694,
|
| 625 |
+
"step": 435
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.23018571802249543,
|
| 629 |
+
"grad_norm": 1.0924934766248438,
|
| 630 |
+
"learning_rate": 1.8990342573714343e-05,
|
| 631 |
+
"loss": 1.5532,
|
| 632 |
+
"step": 440
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.2328014648182056,
|
| 636 |
+
"grad_norm": 1.050247774744565,
|
| 637 |
+
"learning_rate": 1.894995512521878e-05,
|
| 638 |
+
"loss": 1.5333,
|
| 639 |
+
"step": 445
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.23541721161391577,
|
| 643 |
+
"grad_norm": 1.0412618209618199,
|
| 644 |
+
"learning_rate": 1.8908820358045864e-05,
|
| 645 |
+
"loss": 1.5394,
|
| 646 |
+
"step": 450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.23803295840962596,
|
| 650 |
+
"grad_norm": 1.0745621857166745,
|
| 651 |
+
"learning_rate": 1.8866941706936818e-05,
|
| 652 |
+
"loss": 1.5541,
|
| 653 |
+
"step": 455
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.24064870520533613,
|
| 657 |
+
"grad_norm": 1.0159400612281335,
|
| 658 |
+
"learning_rate": 1.8824322668746954e-05,
|
| 659 |
+
"loss": 1.5244,
|
| 660 |
+
"step": 460
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.2432644520010463,
|
| 664 |
+
"grad_norm": 1.0282447285683234,
|
| 665 |
+
"learning_rate": 1.87809668021537e-05,
|
| 666 |
+
"loss": 1.5536,
|
| 667 |
+
"step": 465
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.24588019879675646,
|
| 671 |
+
"grad_norm": 1.0626399443770944,
|
| 672 |
+
"learning_rate": 1.873687772735944e-05,
|
| 673 |
+
"loss": 1.5368,
|
| 674 |
+
"step": 470
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.24849594559246665,
|
| 678 |
+
"grad_norm": 1.0509935242940092,
|
| 679 |
+
"learning_rate": 1.8692059125789232e-05,
|
| 680 |
+
"loss": 1.5332,
|
| 681 |
+
"step": 475
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.2511116923881768,
|
| 685 |
+
"grad_norm": 1.034717315829219,
|
| 686 |
+
"learning_rate": 1.8646514739783404e-05,
|
| 687 |
+
"loss": 1.5322,
|
| 688 |
+
"step": 480
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.253727439183887,
|
| 692 |
+
"grad_norm": 1.0777506677213722,
|
| 693 |
+
"learning_rate": 1.8600248372285085e-05,
|
| 694 |
+
"loss": 1.5293,
|
| 695 |
+
"step": 485
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.25634318597959715,
|
| 699 |
+
"grad_norm": 1.0806499250110846,
|
| 700 |
+
"learning_rate": 1.855326388652265e-05,
|
| 701 |
+
"loss": 1.5656,
|
| 702 |
+
"step": 490
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.25895893277530735,
|
| 706 |
+
"grad_norm": 1.0832360995200667,
|
| 707 |
+
"learning_rate": 1.850556520568713e-05,
|
| 708 |
+
"loss": 1.5278,
|
| 709 |
+
"step": 495
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.26157467957101754,
|
| 713 |
+
"grad_norm": 1.0575116001095382,
|
| 714 |
+
"learning_rate": 1.8457156312604654e-05,
|
| 715 |
+
"loss": 1.5403,
|
| 716 |
+
"step": 500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.2641904263667277,
|
| 720 |
+
"grad_norm": 1.019853383800297,
|
| 721 |
+
"learning_rate": 1.8408041249403863e-05,
|
| 722 |
+
"loss": 1.5427,
|
| 723 |
+
"step": 505
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.2668061731624379,
|
| 727 |
+
"grad_norm": 1.0628900028457942,
|
| 728 |
+
"learning_rate": 1.83582241171784e-05,
|
| 729 |
+
"loss": 1.5565,
|
| 730 |
+
"step": 510
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.26942191995814807,
|
| 734 |
+
"grad_norm": 1.0552839025677658,
|
| 735 |
+
"learning_rate": 1.830770907564447e-05,
|
| 736 |
+
"loss": 1.5573,
|
| 737 |
+
"step": 515
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.2720376667538582,
|
| 741 |
+
"grad_norm": 1.0188309963945468,
|
| 742 |
+
"learning_rate": 1.82565003427935e-05,
|
| 743 |
+
"loss": 1.5755,
|
| 744 |
+
"step": 520
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.2746534135495684,
|
| 748 |
+
"grad_norm": 1.0707482369174548,
|
| 749 |
+
"learning_rate": 1.8204602194539948e-05,
|
| 750 |
+
"loss": 1.5393,
|
| 751 |
+
"step": 525
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.2772691603452786,
|
| 755 |
+
"grad_norm": 1.0764393506648489,
|
| 756 |
+
"learning_rate": 1.8152018964364264e-05,
|
| 757 |
+
"loss": 1.5429,
|
| 758 |
+
"step": 530
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.27988490714098874,
|
| 762 |
+
"grad_norm": 1.0322843607035168,
|
| 763 |
+
"learning_rate": 1.809875504295103e-05,
|
| 764 |
+
"loss": 1.5147,
|
| 765 |
+
"step": 535
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.28250065393669893,
|
| 769 |
+
"grad_norm": 1.0590987228220765,
|
| 770 |
+
"learning_rate": 1.8044814877822353e-05,
|
| 771 |
+
"loss": 1.541,
|
| 772 |
+
"step": 540
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.2851164007324091,
|
| 776 |
+
"grad_norm": 1.0108769666699464,
|
| 777 |
+
"learning_rate": 1.79902029729665e-05,
|
| 778 |
+
"loss": 1.5481,
|
| 779 |
+
"step": 545
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.28773214752811926,
|
| 783 |
+
"grad_norm": 1.0793448502427236,
|
| 784 |
+
"learning_rate": 1.7934923888461816e-05,
|
| 785 |
+
"loss": 1.5596,
|
| 786 |
+
"step": 550
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.29034789432382946,
|
| 790 |
+
"grad_norm": 1.0339193335393733,
|
| 791 |
+
"learning_rate": 1.7878982240095948e-05,
|
| 792 |
+
"loss": 1.538,
|
| 793 |
+
"step": 555
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.29296364111953965,
|
| 797 |
+
"grad_norm": 1.0640817522557218,
|
| 798 |
+
"learning_rate": 1.782238269898043e-05,
|
| 799 |
+
"loss": 1.5515,
|
| 800 |
+
"step": 560
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2955793879152498,
|
| 804 |
+
"grad_norm": 1.0601138738375762,
|
| 805 |
+
"learning_rate": 1.7765129991160658e-05,
|
| 806 |
+
"loss": 1.5381,
|
| 807 |
+
"step": 565
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.29819513471096,
|
| 811 |
+
"grad_norm": 1.1210975652752035,
|
| 812 |
+
"learning_rate": 1.770722889722126e-05,
|
| 813 |
+
"loss": 1.5459,
|
| 814 |
+
"step": 570
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.3008108815066702,
|
| 818 |
+
"grad_norm": 1.04987498111599,
|
| 819 |
+
"learning_rate": 1.7648684251886923e-05,
|
| 820 |
+
"loss": 1.5271,
|
| 821 |
+
"step": 575
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.3034266283023803,
|
| 825 |
+
"grad_norm": 1.039388385836237,
|
| 826 |
+
"learning_rate": 1.758950094361868e-05,
|
| 827 |
+
"loss": 1.543,
|
| 828 |
+
"step": 580
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.3060423750980905,
|
| 832 |
+
"grad_norm": 1.080332419007695,
|
| 833 |
+
"learning_rate": 1.7529683914205732e-05,
|
| 834 |
+
"loss": 1.5498,
|
| 835 |
+
"step": 585
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.3086581218938007,
|
| 839 |
+
"grad_norm": 0.9922109335594432,
|
| 840 |
+
"learning_rate": 1.7469238158352833e-05,
|
| 841 |
+
"loss": 1.5419,
|
| 842 |
+
"step": 590
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.31127386868951085,
|
| 846 |
+
"grad_norm": 1.0591785912300846,
|
| 847 |
+
"learning_rate": 1.7408168723263195e-05,
|
| 848 |
+
"loss": 1.5597,
|
| 849 |
+
"step": 595
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.31388961548522104,
|
| 853 |
+
"grad_norm": 1.0228587979463524,
|
| 854 |
+
"learning_rate": 1.7346480708217077e-05,
|
| 855 |
+
"loss": 1.5365,
|
| 856 |
+
"step": 600
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.3165053622809312,
|
| 860 |
+
"grad_norm": 1.0130538528239692,
|
| 861 |
+
"learning_rate": 1.7284179264145983e-05,
|
| 862 |
+
"loss": 1.5252,
|
| 863 |
+
"step": 605
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.3191211090766414,
|
| 867 |
+
"grad_norm": 1.0847194376622917,
|
| 868 |
+
"learning_rate": 1.722126959320256e-05,
|
| 869 |
+
"loss": 1.5494,
|
| 870 |
+
"step": 610
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.32173685587235157,
|
| 874 |
+
"grad_norm": 1.1067302172616988,
|
| 875 |
+
"learning_rate": 1.715775694832623e-05,
|
| 876 |
+
"loss": 1.5971,
|
| 877 |
+
"step": 615
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.3243526026680617,
|
| 881 |
+
"grad_norm": 1.0880698671052864,
|
| 882 |
+
"learning_rate": 1.7093646632804563e-05,
|
| 883 |
+
"loss": 1.5512,
|
| 884 |
+
"step": 620
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.3269683494637719,
|
| 888 |
+
"grad_norm": 1.0200608253173582,
|
| 889 |
+
"learning_rate": 1.7028943999830457e-05,
|
| 890 |
+
"loss": 1.5146,
|
| 891 |
+
"step": 625
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.3295840962594821,
|
| 895 |
+
"grad_norm": 1.0481539736627015,
|
| 896 |
+
"learning_rate": 1.696365445205514e-05,
|
| 897 |
+
"loss": 1.5696,
|
| 898 |
+
"step": 630
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.33219984305519223,
|
| 902 |
+
"grad_norm": 1.1055045846810556,
|
| 903 |
+
"learning_rate": 1.689778344113707e-05,
|
| 904 |
+
"loss": 1.5122,
|
| 905 |
+
"step": 635
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.33481558985090243,
|
| 909 |
+
"grad_norm": 1.0353380382443884,
|
| 910 |
+
"learning_rate": 1.6831336467286692e-05,
|
| 911 |
+
"loss": 1.5316,
|
| 912 |
+
"step": 640
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.3374313366466126,
|
| 916 |
+
"grad_norm": 1.0188714551351619,
|
| 917 |
+
"learning_rate": 1.676431907880721e-05,
|
| 918 |
+
"loss": 1.5423,
|
| 919 |
+
"step": 645
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.34004708344232276,
|
| 923 |
+
"grad_norm": 1.131600008079283,
|
| 924 |
+
"learning_rate": 1.6696736871631273e-05,
|
| 925 |
+
"loss": 1.5354,
|
| 926 |
+
"step": 650
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.34266283023803296,
|
| 930 |
+
"grad_norm": 1.0515357663095917,
|
| 931 |
+
"learning_rate": 1.662859548885373e-05,
|
| 932 |
+
"loss": 1.5377,
|
| 933 |
+
"step": 655
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.34527857703374315,
|
| 937 |
+
"grad_norm": 1.0572294102012505,
|
| 938 |
+
"learning_rate": 1.6559900620260435e-05,
|
| 939 |
+
"loss": 1.5461,
|
| 940 |
+
"step": 660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.3478943238294533,
|
| 944 |
+
"grad_norm": 1.0152438536619777,
|
| 945 |
+
"learning_rate": 1.649065800185315e-05,
|
| 946 |
+
"loss": 1.5325,
|
| 947 |
+
"step": 665
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.3505100706251635,
|
| 951 |
+
"grad_norm": 1.0418201717103732,
|
| 952 |
+
"learning_rate": 1.642087341537057e-05,
|
| 953 |
+
"loss": 1.5448,
|
| 954 |
+
"step": 670
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.3531258174208737,
|
| 958 |
+
"grad_norm": 1.0480376036050463,
|
| 959 |
+
"learning_rate": 1.6350552687805604e-05,
|
| 960 |
+
"loss": 1.5099,
|
| 961 |
+
"step": 675
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.3557415642165838,
|
| 965 |
+
"grad_norm": 1.060841080063725,
|
| 966 |
+
"learning_rate": 1.6279701690918756e-05,
|
| 967 |
+
"loss": 1.563,
|
| 968 |
+
"step": 680
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.358357311012294,
|
| 972 |
+
"grad_norm": 1.0031599478310154,
|
| 973 |
+
"learning_rate": 1.6222643073599842e-05,
|
| 974 |
+
"loss": 1.531,
|
| 975 |
+
"step": 685
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.3609730578080042,
|
| 979 |
+
"grad_norm": 1.0162970991108897,
|
| 980 |
+
"learning_rate": 1.6150852529788094e-05,
|
| 981 |
+
"loss": 1.5296,
|
| 982 |
+
"step": 690
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.36358880460371434,
|
| 986 |
+
"grad_norm": 1.0311841798916455,
|
| 987 |
+
"learning_rate": 1.6078548391560194e-05,
|
| 988 |
+
"loss": 1.523,
|
| 989 |
+
"step": 695
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.36620455139942454,
|
| 993 |
+
"grad_norm": 1.0425521172262444,
|
| 994 |
+
"learning_rate": 1.600573669629087e-05,
|
| 995 |
+
"loss": 1.5499,
|
| 996 |
+
"step": 700
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.36882029819513473,
|
| 1000 |
+
"grad_norm": 1.0077179207530171,
|
| 1001 |
+
"learning_rate": 1.593242352373573e-05,
|
| 1002 |
+
"loss": 1.521,
|
| 1003 |
+
"step": 705
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.37143604499084487,
|
| 1007 |
+
"grad_norm": 1.03371683112891,
|
| 1008 |
+
"learning_rate": 1.5858614995523582e-05,
|
| 1009 |
+
"loss": 1.5275,
|
| 1010 |
+
"step": 710
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.37405179178655507,
|
| 1014 |
+
"grad_norm": 1.027416326932866,
|
| 1015 |
+
"learning_rate": 1.5784317274645294e-05,
|
| 1016 |
+
"loss": 1.5334,
|
| 1017 |
+
"step": 715
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.37666753858226526,
|
| 1021 |
+
"grad_norm": 1.0347765017314998,
|
| 1022 |
+
"learning_rate": 1.5709536564939166e-05,
|
| 1023 |
+
"loss": 1.5258,
|
| 1024 |
+
"step": 720
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.3792832853779754,
|
| 1028 |
+
"grad_norm": 1.0914401604068136,
|
| 1029 |
+
"learning_rate": 1.5634279110572954e-05,
|
| 1030 |
+
"loss": 1.536,
|
| 1031 |
+
"step": 725
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.3818990321736856,
|
| 1035 |
+
"grad_norm": 1.0755937284063355,
|
| 1036 |
+
"learning_rate": 1.555855119552243e-05,
|
| 1037 |
+
"loss": 1.5056,
|
| 1038 |
+
"step": 730
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.3845147789693958,
|
| 1042 |
+
"grad_norm": 0.9920255452158145,
|
| 1043 |
+
"learning_rate": 1.548235914304671e-05,
|
| 1044 |
+
"loss": 1.4981,
|
| 1045 |
+
"step": 735
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.3871305257651059,
|
| 1049 |
+
"grad_norm": 1.0113153705885736,
|
| 1050 |
+
"learning_rate": 1.5405709315160253e-05,
|
| 1051 |
+
"loss": 1.5584,
|
| 1052 |
+
"step": 740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.3897462725608161,
|
| 1056 |
+
"grad_norm": 1.0413164080826058,
|
| 1057 |
+
"learning_rate": 1.5328608112101625e-05,
|
| 1058 |
+
"loss": 1.4892,
|
| 1059 |
+
"step": 745
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.3923620193565263,
|
| 1063 |
+
"grad_norm": 1.1064616624538501,
|
| 1064 |
+
"learning_rate": 1.5251061971799095e-05,
|
| 1065 |
+
"loss": 1.5259,
|
| 1066 |
+
"step": 750
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.39497776615223645,
|
| 1070 |
+
"grad_norm": 1.0539591793351195,
|
| 1071 |
+
"learning_rate": 1.5173077369333063e-05,
|
| 1072 |
+
"loss": 1.5237,
|
| 1073 |
+
"step": 755
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.39759351294794665,
|
| 1077 |
+
"grad_norm": 1.0288951094405483,
|
| 1078 |
+
"learning_rate": 1.509466081639539e-05,
|
| 1079 |
+
"loss": 1.5265,
|
| 1080 |
+
"step": 760
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.4002092597436568,
|
| 1084 |
+
"grad_norm": 0.9910808136275417,
|
| 1085 |
+
"learning_rate": 1.5015818860745683e-05,
|
| 1086 |
+
"loss": 1.4973,
|
| 1087 |
+
"step": 765
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.402825006539367,
|
| 1091 |
+
"grad_norm": 1.0563108868352806,
|
| 1092 |
+
"learning_rate": 1.4936558085664544e-05,
|
| 1093 |
+
"loss": 1.563,
|
| 1094 |
+
"step": 770
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.4054407533350772,
|
| 1098 |
+
"grad_norm": 1.0263473811764434,
|
| 1099 |
+
"learning_rate": 1.485688510940389e-05,
|
| 1100 |
+
"loss": 1.5151,
|
| 1101 |
+
"step": 775
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.4080565001307873,
|
| 1105 |
+
"grad_norm": 1.0154697112893702,
|
| 1106 |
+
"learning_rate": 1.4776806584634309e-05,
|
| 1107 |
+
"loss": 1.4925,
|
| 1108 |
+
"step": 780
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.4106722469264975,
|
| 1112 |
+
"grad_norm": 1.0553574640994388,
|
| 1113 |
+
"learning_rate": 1.4696329197889571e-05,
|
| 1114 |
+
"loss": 1.5146,
|
| 1115 |
+
"step": 785
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.4132879937222077,
|
| 1119 |
+
"grad_norm": 1.0211598871411691,
|
| 1120 |
+
"learning_rate": 1.461545966900831e-05,
|
| 1121 |
+
"loss": 1.5213,
|
| 1122 |
+
"step": 790
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.41590374051791784,
|
| 1126 |
+
"grad_norm": 1.0436672745470057,
|
| 1127 |
+
"learning_rate": 1.4534204750572918e-05,
|
| 1128 |
+
"loss": 1.5194,
|
| 1129 |
+
"step": 795
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.41851948731362804,
|
| 1133 |
+
"grad_norm": 1.0683987020624277,
|
| 1134 |
+
"learning_rate": 1.4452571227345698e-05,
|
| 1135 |
+
"loss": 1.5334,
|
| 1136 |
+
"step": 800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.42113523410933823,
|
| 1140 |
+
"grad_norm": 0.9987604732281646,
|
| 1141 |
+
"learning_rate": 1.437056591570235e-05,
|
| 1142 |
+
"loss": 1.523,
|
| 1143 |
+
"step": 805
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.42375098090504837,
|
| 1147 |
+
"grad_norm": 1.0218357841163621,
|
| 1148 |
+
"learning_rate": 1.4288195663062796e-05,
|
| 1149 |
+
"loss": 1.5217,
|
| 1150 |
+
"step": 810
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.42636672770075856,
|
| 1154 |
+
"grad_norm": 1.007902685397865,
|
| 1155 |
+
"learning_rate": 1.4205467347319425e-05,
|
| 1156 |
+
"loss": 1.4787,
|
| 1157 |
+
"step": 815
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.42898247449646876,
|
| 1161 |
+
"grad_norm": 1.0609807938633398,
|
| 1162 |
+
"learning_rate": 1.4122387876262794e-05,
|
| 1163 |
+
"loss": 1.5192,
|
| 1164 |
+
"step": 820
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.4315982212921789,
|
| 1168 |
+
"grad_norm": 1.0174184097028445,
|
| 1169 |
+
"learning_rate": 1.4038964187004833e-05,
|
| 1170 |
+
"loss": 1.5179,
|
| 1171 |
+
"step": 825
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4342139680878891,
|
| 1175 |
+
"grad_norm": 1.0228108210322504,
|
| 1176 |
+
"learning_rate": 1.395520324539958e-05,
|
| 1177 |
+
"loss": 1.5075,
|
| 1178 |
+
"step": 830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.4368297148835993,
|
| 1182 |
+
"grad_norm": 1.0785325718776333,
|
| 1183 |
+
"learning_rate": 1.387111204546156e-05,
|
| 1184 |
+
"loss": 1.5267,
|
| 1185 |
+
"step": 835
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.4394454616793094,
|
| 1189 |
+
"grad_norm": 0.9974032173436308,
|
| 1190 |
+
"learning_rate": 1.3786697608781768e-05,
|
| 1191 |
+
"loss": 1.5113,
|
| 1192 |
+
"step": 840
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.4420612084750196,
|
| 1196 |
+
"grad_norm": 1.0170077489667377,
|
| 1197 |
+
"learning_rate": 1.370196698394137e-05,
|
| 1198 |
+
"loss": 1.5644,
|
| 1199 |
+
"step": 845
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.4446769552707298,
|
| 1203 |
+
"grad_norm": 1.021249733750996,
|
| 1204 |
+
"learning_rate": 1.3616927245923157e-05,
|
| 1205 |
+
"loss": 1.5317,
|
| 1206 |
+
"step": 850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.44729270206643995,
|
| 1210 |
+
"grad_norm": 1.0490327472678453,
|
| 1211 |
+
"learning_rate": 1.3531585495520788e-05,
|
| 1212 |
+
"loss": 1.5326,
|
| 1213 |
+
"step": 855
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.44990844886215015,
|
| 1217 |
+
"grad_norm": 1.0605562385856955,
|
| 1218 |
+
"learning_rate": 1.3445948858745852e-05,
|
| 1219 |
+
"loss": 1.5336,
|
| 1220 |
+
"step": 860
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.45252419565786034,
|
| 1224 |
+
"grad_norm": 1.0200966540374112,
|
| 1225 |
+
"learning_rate": 1.3360024486232885e-05,
|
| 1226 |
+
"loss": 1.5144,
|
| 1227 |
+
"step": 865
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.4551399424535705,
|
| 1231 |
+
"grad_norm": 1.0527796533842588,
|
| 1232 |
+
"learning_rate": 1.3273819552642257e-05,
|
| 1233 |
+
"loss": 1.5396,
|
| 1234 |
+
"step": 870
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.4577556892492807,
|
| 1238 |
+
"grad_norm": 1.0693600047084866,
|
| 1239 |
+
"learning_rate": 1.3187341256061122e-05,
|
| 1240 |
+
"loss": 1.5505,
|
| 1241 |
+
"step": 875
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.46037143604499087,
|
| 1245 |
+
"grad_norm": 0.9890513679652697,
|
| 1246 |
+
"learning_rate": 1.310059681740235e-05,
|
| 1247 |
+
"loss": 1.5348,
|
| 1248 |
+
"step": 880
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.462987182840701,
|
| 1252 |
+
"grad_norm": 1.0099545630478501,
|
| 1253 |
+
"learning_rate": 1.3013593479801609e-05,
|
| 1254 |
+
"loss": 1.5298,
|
| 1255 |
+
"step": 885
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.4656029296364112,
|
| 1259 |
+
"grad_norm": 0.9723378721471712,
|
| 1260 |
+
"learning_rate": 1.2926338508012553e-05,
|
| 1261 |
+
"loss": 1.4968,
|
| 1262 |
+
"step": 890
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.4682186764321214,
|
| 1266 |
+
"grad_norm": 0.9889530864908892,
|
| 1267 |
+
"learning_rate": 1.2838839187800218e-05,
|
| 1268 |
+
"loss": 1.503,
|
| 1269 |
+
"step": 895
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.47083442322783153,
|
| 1273 |
+
"grad_norm": 0.9891864321287015,
|
| 1274 |
+
"learning_rate": 1.2751102825332654e-05,
|
| 1275 |
+
"loss": 1.5514,
|
| 1276 |
+
"step": 900
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.4734501700235417,
|
| 1280 |
+
"grad_norm": 1.0109535736344295,
|
| 1281 |
+
"learning_rate": 1.2663136746570883e-05,
|
| 1282 |
+
"loss": 1.5002,
|
| 1283 |
+
"step": 905
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.4760659168192519,
|
| 1287 |
+
"grad_norm": 1.0169628002748028,
|
| 1288 |
+
"learning_rate": 1.2574948296657167e-05,
|
| 1289 |
+
"loss": 1.4947,
|
| 1290 |
+
"step": 910
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.47868166361496206,
|
| 1294 |
+
"grad_norm": 1.0410822345832762,
|
| 1295 |
+
"learning_rate": 1.2486544839301698e-05,
|
| 1296 |
+
"loss": 1.5135,
|
| 1297 |
+
"step": 915
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.48129741041067226,
|
| 1301 |
+
"grad_norm": 1.0383523950765439,
|
| 1302 |
+
"learning_rate": 1.2397933756167717e-05,
|
| 1303 |
+
"loss": 1.5004,
|
| 1304 |
+
"step": 920
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.48391315720638245,
|
| 1308 |
+
"grad_norm": 1.026964587730628,
|
| 1309 |
+
"learning_rate": 1.2309122446255174e-05,
|
| 1310 |
+
"loss": 1.5356,
|
| 1311 |
+
"step": 925
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.4865289040020926,
|
| 1315 |
+
"grad_norm": 1.0591059166395116,
|
| 1316 |
+
"learning_rate": 1.2220118325282879e-05,
|
| 1317 |
+
"loss": 1.4936,
|
| 1318 |
+
"step": 930
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.4891446507978028,
|
| 1322 |
+
"grad_norm": 1.0331296995307677,
|
| 1323 |
+
"learning_rate": 1.2130928825069328e-05,
|
| 1324 |
+
"loss": 1.5266,
|
| 1325 |
+
"step": 935
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.4917603975935129,
|
| 1329 |
+
"grad_norm": 1.0355969595335166,
|
| 1330 |
+
"learning_rate": 1.2041561392912118e-05,
|
| 1331 |
+
"loss": 1.513,
|
| 1332 |
+
"step": 940
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.4943761443892231,
|
| 1336 |
+
"grad_norm": 1.0156424092358933,
|
| 1337 |
+
"learning_rate": 1.1952023490966119e-05,
|
| 1338 |
+
"loss": 1.5141,
|
| 1339 |
+
"step": 945
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.4969918911849333,
|
| 1343 |
+
"grad_norm": 1.006170091022331,
|
| 1344 |
+
"learning_rate": 1.1862322595620372e-05,
|
| 1345 |
+
"loss": 1.5325,
|
| 1346 |
+
"step": 950
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.49960763798064345,
|
| 1350 |
+
"grad_norm": 1.0099928926730717,
|
| 1351 |
+
"learning_rate": 1.1772466196873828e-05,
|
| 1352 |
+
"loss": 1.5116,
|
| 1353 |
+
"step": 955
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.5022233847763536,
|
| 1357 |
+
"grad_norm": 1.0671464757515852,
|
| 1358 |
+
"learning_rate": 1.1682461797709921e-05,
|
| 1359 |
+
"loss": 1.5431,
|
| 1360 |
+
"step": 960
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.5048391315720638,
|
| 1364 |
+
"grad_norm": 1.0134823207599724,
|
| 1365 |
+
"learning_rate": 1.1592316913470083e-05,
|
| 1366 |
+
"loss": 1.5319,
|
| 1367 |
+
"step": 965
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.507454878367774,
|
| 1371 |
+
"grad_norm": 1.0027055074680191,
|
| 1372 |
+
"learning_rate": 1.1502039071226205e-05,
|
| 1373 |
+
"loss": 1.5061,
|
| 1374 |
+
"step": 970
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.5100706251634842,
|
| 1378 |
+
"grad_norm": 1.0184622709635633,
|
| 1379 |
+
"learning_rate": 1.1411635809152133e-05,
|
| 1380 |
+
"loss": 1.5342,
|
| 1381 |
+
"step": 975
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.5126863719591943,
|
| 1385 |
+
"grad_norm": 1.0686982981270088,
|
| 1386 |
+
"learning_rate": 1.1321114675894235e-05,
|
| 1387 |
+
"loss": 1.5446,
|
| 1388 |
+
"step": 980
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.5153021187549045,
|
| 1392 |
+
"grad_norm": 1.0011246745665834,
|
| 1393 |
+
"learning_rate": 1.1230483229941092e-05,
|
| 1394 |
+
"loss": 1.52,
|
| 1395 |
+
"step": 985
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.5179178655506147,
|
| 1399 |
+
"grad_norm": 1.029566422986875,
|
| 1400 |
+
"learning_rate": 1.1139749038992349e-05,
|
| 1401 |
+
"loss": 1.5272,
|
| 1402 |
+
"step": 990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.5205336123463249,
|
| 1406 |
+
"grad_norm": 1.0105863963908448,
|
| 1407 |
+
"learning_rate": 1.104891967932685e-05,
|
| 1408 |
+
"loss": 1.5324,
|
| 1409 |
+
"step": 995
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.5231493591420351,
|
| 1413 |
+
"grad_norm": 1.0551963943316993,
|
| 1414 |
+
"learning_rate": 1.0958002735169981e-05,
|
| 1415 |
+
"loss": 1.5059,
|
| 1416 |
+
"step": 1000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.5257651059377453,
|
| 1420 |
+
"grad_norm": 1.0039149878815243,
|
| 1421 |
+
"learning_rate": 1.086700579806042e-05,
|
| 1422 |
+
"loss": 1.4738,
|
| 1423 |
+
"step": 1005
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.5283808527334554,
|
| 1427 |
+
"grad_norm": 1.0151599862223728,
|
| 1428 |
+
"learning_rate": 1.0775936466216226e-05,
|
| 1429 |
+
"loss": 1.5345,
|
| 1430 |
+
"step": 1010
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.5309965995291656,
|
| 1434 |
+
"grad_norm": 0.9987481092934746,
|
| 1435 |
+
"learning_rate": 1.06848023439004e-05,
|
| 1436 |
+
"loss": 1.5349,
|
| 1437 |
+
"step": 1015
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.5336123463248758,
|
| 1441 |
+
"grad_norm": 0.9831986769359685,
|
| 1442 |
+
"learning_rate": 1.0593611040785927e-05,
|
| 1443 |
+
"loss": 1.551,
|
| 1444 |
+
"step": 1020
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.536228093120586,
|
| 1448 |
+
"grad_norm": 1.010583361512088,
|
| 1449 |
+
"learning_rate": 1.0502370171320382e-05,
|
| 1450 |
+
"loss": 1.5249,
|
| 1451 |
+
"step": 1025
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.5388438399162961,
|
| 1455 |
+
"grad_norm": 0.9983357698642957,
|
| 1456 |
+
"learning_rate": 1.04110873540901e-05,
|
| 1457 |
+
"loss": 1.5232,
|
| 1458 |
+
"step": 1030
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.5414595867120062,
|
| 1462 |
+
"grad_norm": 1.0843309523395082,
|
| 1463 |
+
"learning_rate": 1.0319770211184052e-05,
|
| 1464 |
+
"loss": 1.4754,
|
| 1465 |
+
"step": 1035
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.5440753335077164,
|
| 1469 |
+
"grad_norm": 1.0058326702680296,
|
| 1470 |
+
"learning_rate": 1.0228426367557396e-05,
|
| 1471 |
+
"loss": 1.5133,
|
| 1472 |
+
"step": 1040
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.5466910803034266,
|
| 1476 |
+
"grad_norm": 1.008493783415069,
|
| 1477 |
+
"learning_rate": 1.0137063450394786e-05,
|
| 1478 |
+
"loss": 1.522,
|
| 1479 |
+
"step": 1045
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.5493068270991368,
|
| 1483 |
+
"grad_norm": 1.0615454308769967,
|
| 1484 |
+
"learning_rate": 1.0045689088473509e-05,
|
| 1485 |
+
"loss": 1.5037,
|
| 1486 |
+
"step": 1050
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.551922573894847,
|
| 1490 |
+
"grad_norm": 0.9944300679828315,
|
| 1491 |
+
"learning_rate": 9.954310911526495e-06,
|
| 1492 |
+
"loss": 1.4847,
|
| 1493 |
+
"step": 1055
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.5545383206905572,
|
| 1497 |
+
"grad_norm": 0.9927554074330694,
|
| 1498 |
+
"learning_rate": 9.862936549605219e-06,
|
| 1499 |
+
"loss": 1.5106,
|
| 1500 |
+
"step": 1060
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.5571540674862673,
|
| 1504 |
+
"grad_norm": 0.9865550178283998,
|
| 1505 |
+
"learning_rate": 9.771573632442606e-06,
|
| 1506 |
+
"loss": 1.517,
|
| 1507 |
+
"step": 1065
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.5597698142819775,
|
| 1511 |
+
"grad_norm": 0.9966069424236838,
|
| 1512 |
+
"learning_rate": 9.68022978881595e-06,
|
| 1513 |
+
"loss": 1.5025,
|
| 1514 |
+
"step": 1070
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.5623855610776877,
|
| 1518 |
+
"grad_norm": 1.0095109151004296,
|
| 1519 |
+
"learning_rate": 9.588912645909905e-06,
|
| 1520 |
+
"loss": 1.5433,
|
| 1521 |
+
"step": 1075
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.5650013078733979,
|
| 1525 |
+
"grad_norm": 0.9952021046731035,
|
| 1526 |
+
"learning_rate": 9.497629828679625e-06,
|
| 1527 |
+
"loss": 1.5154,
|
| 1528 |
+
"step": 1080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.5676170546691081,
|
| 1532 |
+
"grad_norm": 1.0027294360396322,
|
| 1533 |
+
"learning_rate": 9.406388959214073e-06,
|
| 1534 |
+
"loss": 1.5111,
|
| 1535 |
+
"step": 1085
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.5702328014648183,
|
| 1539 |
+
"grad_norm": 0.9835810468995724,
|
| 1540 |
+
"learning_rate": 9.315197656099602e-06,
|
| 1541 |
+
"loss": 1.4865,
|
| 1542 |
+
"step": 1090
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.5728485482605283,
|
| 1546 |
+
"grad_norm": 0.9705328284796092,
|
| 1547 |
+
"learning_rate": 9.224063533783777e-06,
|
| 1548 |
+
"loss": 1.5083,
|
| 1549 |
+
"step": 1095
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.5754642950562385,
|
| 1553 |
+
"grad_norm": 1.0269985622108104,
|
| 1554 |
+
"learning_rate": 9.132994201939584e-06,
|
| 1555 |
+
"loss": 1.5497,
|
| 1556 |
+
"step": 1100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.5780800418519487,
|
| 1560 |
+
"grad_norm": 0.9943115221329752,
|
| 1561 |
+
"learning_rate": 9.04199726483002e-06,
|
| 1562 |
+
"loss": 1.5316,
|
| 1563 |
+
"step": 1105
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.5806957886476589,
|
| 1567 |
+
"grad_norm": 0.9851051801018973,
|
| 1568 |
+
"learning_rate": 8.951080320673154e-06,
|
| 1569 |
+
"loss": 1.5107,
|
| 1570 |
+
"step": 1110
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.5833115354433691,
|
| 1574 |
+
"grad_norm": 1.0084233617257117,
|
| 1575 |
+
"learning_rate": 8.860250961007653e-06,
|
| 1576 |
+
"loss": 1.4999,
|
| 1577 |
+
"step": 1115
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.5859272822390793,
|
| 1581 |
+
"grad_norm": 0.9792135269064121,
|
| 1582 |
+
"learning_rate": 8.769516770058915e-06,
|
| 1583 |
+
"loss": 1.4986,
|
| 1584 |
+
"step": 1120
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.5885430290347894,
|
| 1588 |
+
"grad_norm": 0.9814466857829048,
|
| 1589 |
+
"learning_rate": 8.678885324105766e-06,
|
| 1590 |
+
"loss": 1.4858,
|
| 1591 |
+
"step": 1125
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.5911587758304996,
|
| 1595 |
+
"grad_norm": 1.070197880908098,
|
| 1596 |
+
"learning_rate": 8.58836419084787e-06,
|
| 1597 |
+
"loss": 1.4895,
|
| 1598 |
+
"step": 1130
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.5937745226262098,
|
| 1602 |
+
"grad_norm": 1.0540676200666286,
|
| 1603 |
+
"learning_rate": 8.497960928773798e-06,
|
| 1604 |
+
"loss": 1.4991,
|
| 1605 |
+
"step": 1135
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.59639026942192,
|
| 1609 |
+
"grad_norm": 0.9756635550227876,
|
| 1610 |
+
"learning_rate": 8.40768308652992e-06,
|
| 1611 |
+
"loss": 1.5065,
|
| 1612 |
+
"step": 1140
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.5990060162176302,
|
| 1616 |
+
"grad_norm": 0.984044173838774,
|
| 1617 |
+
"learning_rate": 8.317538202290079e-06,
|
| 1618 |
+
"loss": 1.5074,
|
| 1619 |
+
"step": 1145
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.6016217630133404,
|
| 1623 |
+
"grad_norm": 0.9852521488294786,
|
| 1624 |
+
"learning_rate": 8.227533803126174e-06,
|
| 1625 |
+
"loss": 1.5018,
|
| 1626 |
+
"step": 1150
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.6042375098090504,
|
| 1630 |
+
"grad_norm": 0.9954972499796161,
|
| 1631 |
+
"learning_rate": 8.137677404379631e-06,
|
| 1632 |
+
"loss": 1.5167,
|
| 1633 |
+
"step": 1155
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.6068532566047606,
|
| 1637 |
+
"grad_norm": 0.9651848596891013,
|
| 1638 |
+
"learning_rate": 8.047976509033888e-06,
|
| 1639 |
+
"loss": 1.5076,
|
| 1640 |
+
"step": 1160
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.6094690034004708,
|
| 1644 |
+
"grad_norm": 0.9524703700980293,
|
| 1645 |
+
"learning_rate": 7.958438607087884e-06,
|
| 1646 |
+
"loss": 1.5182,
|
| 1647 |
+
"step": 1165
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.612084750196181,
|
| 1651 |
+
"grad_norm": 0.9794580890478966,
|
| 1652 |
+
"learning_rate": 7.869071174930675e-06,
|
| 1653 |
+
"loss": 1.512,
|
| 1654 |
+
"step": 1170
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.6147004969918912,
|
| 1658 |
+
"grad_norm": 1.0029058817463463,
|
| 1659 |
+
"learning_rate": 7.779881674717123e-06,
|
| 1660 |
+
"loss": 1.5269,
|
| 1661 |
+
"step": 1175
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.6173162437876014,
|
| 1665 |
+
"grad_norm": 1.0063375562285168,
|
| 1666 |
+
"learning_rate": 7.690877553744831e-06,
|
| 1667 |
+
"loss": 1.5111,
|
| 1668 |
+
"step": 1180
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.6199319905833115,
|
| 1672 |
+
"grad_norm": 1.0091077551574594,
|
| 1673 |
+
"learning_rate": 7.6020662438322845e-06,
|
| 1674 |
+
"loss": 1.5169,
|
| 1675 |
+
"step": 1185
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.6225477373790217,
|
| 1679 |
+
"grad_norm": 0.9943024589083075,
|
| 1680 |
+
"learning_rate": 7.513455160698307e-06,
|
| 1681 |
+
"loss": 1.5213,
|
| 1682 |
+
"step": 1190
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.6251634841747319,
|
| 1686 |
+
"grad_norm": 0.9829867902427267,
|
| 1687 |
+
"learning_rate": 7.425051703342835e-06,
|
| 1688 |
+
"loss": 1.5159,
|
| 1689 |
+
"step": 1195
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.6277792309704421,
|
| 1693 |
+
"grad_norm": 0.970879702585121,
|
| 1694 |
+
"learning_rate": 7.336863253429121e-06,
|
| 1695 |
+
"loss": 1.492,
|
| 1696 |
+
"step": 1200
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.6303949777661523,
|
| 1700 |
+
"grad_norm": 1.015533293758364,
|
| 1701 |
+
"learning_rate": 7.248897174667346e-06,
|
| 1702 |
+
"loss": 1.5142,
|
| 1703 |
+
"step": 1205
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.6330107245618624,
|
| 1707 |
+
"grad_norm": 0.9983522324691199,
|
| 1708 |
+
"learning_rate": 7.161160812199785e-06,
|
| 1709 |
+
"loss": 1.5142,
|
| 1710 |
+
"step": 1210
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.6356264713575726,
|
| 1714 |
+
"grad_norm": 0.9981646652538129,
|
| 1715 |
+
"learning_rate": 7.0736614919874506e-06,
|
| 1716 |
+
"loss": 1.5147,
|
| 1717 |
+
"step": 1215
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.6382422181532827,
|
| 1721 |
+
"grad_norm": 1.0030807490520133,
|
| 1722 |
+
"learning_rate": 6.986406520198394e-06,
|
| 1723 |
+
"loss": 1.5281,
|
| 1724 |
+
"step": 1220
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.6408579649489929,
|
| 1728 |
+
"grad_norm": 0.9865616775503276,
|
| 1729 |
+
"learning_rate": 6.899403182597652e-06,
|
| 1730 |
+
"loss": 1.4889,
|
| 1731 |
+
"step": 1225
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.6434737117447031,
|
| 1735 |
+
"grad_norm": 0.9546372070962268,
|
| 1736 |
+
"learning_rate": 6.812658743938883e-06,
|
| 1737 |
+
"loss": 1.5075,
|
| 1738 |
+
"step": 1230
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.6460894585404133,
|
| 1742 |
+
"grad_norm": 0.9926030009168805,
|
| 1743 |
+
"learning_rate": 6.726180447357744e-06,
|
| 1744 |
+
"loss": 1.4862,
|
| 1745 |
+
"step": 1235
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.6487052053361234,
|
| 1749 |
+
"grad_norm": 1.0209154524235156,
|
| 1750 |
+
"learning_rate": 6.639975513767118e-06,
|
| 1751 |
+
"loss": 1.4983,
|
| 1752 |
+
"step": 1240
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.6513209521318336,
|
| 1756 |
+
"grad_norm": 0.9946339981848588,
|
| 1757 |
+
"learning_rate": 6.554051141254147e-06,
|
| 1758 |
+
"loss": 1.5164,
|
| 1759 |
+
"step": 1245
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.6539366989275438,
|
| 1763 |
+
"grad_norm": 0.9988896509581313,
|
| 1764 |
+
"learning_rate": 6.468414504479215e-06,
|
| 1765 |
+
"loss": 1.4842,
|
| 1766 |
+
"step": 1250
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.656552445723254,
|
| 1770 |
+
"grad_norm": 0.9887613888084512,
|
| 1771 |
+
"learning_rate": 6.3830727540768445e-06,
|
| 1772 |
+
"loss": 1.5049,
|
| 1773 |
+
"step": 1255
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.6591681925189642,
|
| 1777 |
+
"grad_norm": 1.0572212511536254,
|
| 1778 |
+
"learning_rate": 6.298033016058636e-06,
|
| 1779 |
+
"loss": 1.5263,
|
| 1780 |
+
"step": 1260
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.6617839393146744,
|
| 1784 |
+
"grad_norm": 0.9880149436409088,
|
| 1785 |
+
"learning_rate": 6.213302391218235e-06,
|
| 1786 |
+
"loss": 1.5245,
|
| 1787 |
+
"step": 1265
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.6643996861103845,
|
| 1791 |
+
"grad_norm": 0.9643400144599568,
|
| 1792 |
+
"learning_rate": 6.12888795453844e-06,
|
| 1793 |
+
"loss": 1.5109,
|
| 1794 |
+
"step": 1270
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.6670154329060947,
|
| 1798 |
+
"grad_norm": 1.000715533098279,
|
| 1799 |
+
"learning_rate": 6.0447967546004235e-06,
|
| 1800 |
+
"loss": 1.4894,
|
| 1801 |
+
"step": 1275
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.6696311797018049,
|
| 1805 |
+
"grad_norm": 0.9800856395483365,
|
| 1806 |
+
"learning_rate": 5.961035812995173e-06,
|
| 1807 |
+
"loss": 1.5,
|
| 1808 |
+
"step": 1280
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.672246926497515,
|
| 1812 |
+
"grad_norm": 1.0385638801148824,
|
| 1813 |
+
"learning_rate": 5.877612123737206e-06,
|
| 1814 |
+
"loss": 1.4934,
|
| 1815 |
+
"step": 1285
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.6748626732932252,
|
| 1819 |
+
"grad_norm": 0.987221117964989,
|
| 1820 |
+
"learning_rate": 5.79453265268058e-06,
|
| 1821 |
+
"loss": 1.528,
|
| 1822 |
+
"step": 1290
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.6774784200889354,
|
| 1826 |
+
"grad_norm": 0.94969028723842,
|
| 1827 |
+
"learning_rate": 5.71180433693721e-06,
|
| 1828 |
+
"loss": 1.5009,
|
| 1829 |
+
"step": 1295
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.6800941668846455,
|
| 1833 |
+
"grad_norm": 1.012692697283537,
|
| 1834 |
+
"learning_rate": 5.629434084297654e-06,
|
| 1835 |
+
"loss": 1.5085,
|
| 1836 |
+
"step": 1300
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.6827099136803557,
|
| 1840 |
+
"grad_norm": 0.9816607063868247,
|
| 1841 |
+
"learning_rate": 5.547428772654303e-06,
|
| 1842 |
+
"loss": 1.5086,
|
| 1843 |
+
"step": 1305
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.6853256604760659,
|
| 1847 |
+
"grad_norm": 1.0183355117538198,
|
| 1848 |
+
"learning_rate": 5.465795249427084e-06,
|
| 1849 |
+
"loss": 1.4997,
|
| 1850 |
+
"step": 1310
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.6879414072717761,
|
| 1854 |
+
"grad_norm": 0.970016627451196,
|
| 1855 |
+
"learning_rate": 5.384540330991692e-06,
|
| 1856 |
+
"loss": 1.4744,
|
| 1857 |
+
"step": 1315
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.6905571540674863,
|
| 1861 |
+
"grad_norm": 1.0178111067524296,
|
| 1862 |
+
"learning_rate": 5.303670802110435e-06,
|
| 1863 |
+
"loss": 1.509,
|
| 1864 |
+
"step": 1320
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.6931729008631965,
|
| 1868 |
+
"grad_norm": 0.9944664971116642,
|
| 1869 |
+
"learning_rate": 5.223193415365693e-06,
|
| 1870 |
+
"loss": 1.4929,
|
| 1871 |
+
"step": 1325
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.6957886476589066,
|
| 1875 |
+
"grad_norm": 0.9946825583101631,
|
| 1876 |
+
"learning_rate": 5.14311489059611e-06,
|
| 1877 |
+
"loss": 1.5002,
|
| 1878 |
+
"step": 1330
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.6984043944546168,
|
| 1882 |
+
"grad_norm": 0.9989762838121093,
|
| 1883 |
+
"learning_rate": 5.063441914335456e-06,
|
| 1884 |
+
"loss": 1.489,
|
| 1885 |
+
"step": 1335
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.701020141250327,
|
| 1889 |
+
"grad_norm": 0.9637542818577731,
|
| 1890 |
+
"learning_rate": 4.984181139254321e-06,
|
| 1891 |
+
"loss": 1.4972,
|
| 1892 |
+
"step": 1340
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.7036358880460372,
|
| 1896 |
+
"grad_norm": 0.9746542807450145,
|
| 1897 |
+
"learning_rate": 4.905339183604614e-06,
|
| 1898 |
+
"loss": 1.5135,
|
| 1899 |
+
"step": 1345
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.7062516348417474,
|
| 1903 |
+
"grad_norm": 0.9715476115436449,
|
| 1904 |
+
"learning_rate": 4.8269226306669415e-06,
|
| 1905 |
+
"loss": 1.4902,
|
| 1906 |
+
"step": 1350
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.7088673816374574,
|
| 1910 |
+
"grad_norm": 0.9807894006152054,
|
| 1911 |
+
"learning_rate": 4.748938028200909e-06,
|
| 1912 |
+
"loss": 1.5012,
|
| 1913 |
+
"step": 1355
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.7114831284331676,
|
| 1917 |
+
"grad_norm": 0.9908086485025123,
|
| 1918 |
+
"learning_rate": 4.67139188789838e-06,
|
| 1919 |
+
"loss": 1.506,
|
| 1920 |
+
"step": 1360
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.7140988752288778,
|
| 1924 |
+
"grad_norm": 0.9689040750872466,
|
| 1925 |
+
"learning_rate": 4.594290684839748e-06,
|
| 1926 |
+
"loss": 1.5207,
|
| 1927 |
+
"step": 1365
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.716714622024588,
|
| 1931 |
+
"grad_norm": 0.9963825669983652,
|
| 1932 |
+
"learning_rate": 4.51764085695329e-06,
|
| 1933 |
+
"loss": 1.4973,
|
| 1934 |
+
"step": 1370
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.7193303688202982,
|
| 1938 |
+
"grad_norm": 0.9629335061932984,
|
| 1939 |
+
"learning_rate": 4.441448804477572e-06,
|
| 1940 |
+
"loss": 1.5051,
|
| 1941 |
+
"step": 1375
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.7219461156160084,
|
| 1945 |
+
"grad_norm": 0.9660913246143501,
|
| 1946 |
+
"learning_rate": 4.36572088942705e-06,
|
| 1947 |
+
"loss": 1.5044,
|
| 1948 |
+
"step": 1380
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.7245618624117185,
|
| 1952 |
+
"grad_norm": 0.9743404732244383,
|
| 1953 |
+
"learning_rate": 4.290463435060832e-06,
|
| 1954 |
+
"loss": 1.5131,
|
| 1955 |
+
"step": 1385
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.7271776092074287,
|
| 1959 |
+
"grad_norm": 0.9760030839620752,
|
| 1960 |
+
"learning_rate": 4.2156827253547095e-06,
|
| 1961 |
+
"loss": 1.5067,
|
| 1962 |
+
"step": 1390
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.7297933560031389,
|
| 1966 |
+
"grad_norm": 0.9583252278034834,
|
| 1967 |
+
"learning_rate": 4.141385004476418e-06,
|
| 1968 |
+
"loss": 1.507,
|
| 1969 |
+
"step": 1395
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.7324091027988491,
|
| 1973 |
+
"grad_norm": 0.9948309507033296,
|
| 1974 |
+
"learning_rate": 4.06757647626427e-06,
|
| 1975 |
+
"loss": 1.5183,
|
| 1976 |
+
"step": 1400
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.7350248495945593,
|
| 1980 |
+
"grad_norm": 0.987442634040057,
|
| 1981 |
+
"learning_rate": 3.9942633037091305e-06,
|
| 1982 |
+
"loss": 1.509,
|
| 1983 |
+
"step": 1405
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.7376405963902695,
|
| 1987 |
+
"grad_norm": 0.9702680673369065,
|
| 1988 |
+
"learning_rate": 3.92145160843981e-06,
|
| 1989 |
+
"loss": 1.4877,
|
| 1990 |
+
"step": 1410
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.7402563431859795,
|
| 1994 |
+
"grad_norm": 0.9763959077401622,
|
| 1995 |
+
"learning_rate": 3.8491474702119104e-06,
|
| 1996 |
+
"loss": 1.5025,
|
| 1997 |
+
"step": 1415
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.7428720899816897,
|
| 2001 |
+
"grad_norm": 0.9601538393778235,
|
| 2002 |
+
"learning_rate": 3.7773569264001653e-06,
|
| 2003 |
+
"loss": 1.4817,
|
| 2004 |
+
"step": 1420
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.7454878367773999,
|
| 2008 |
+
"grad_norm": 0.9481115013329992,
|
| 2009 |
+
"learning_rate": 3.70608597149432e-06,
|
| 2010 |
+
"loss": 1.5248,
|
| 2011 |
+
"step": 1425
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.7481035835731101,
|
| 2015 |
+
"grad_norm": 0.9776589346839822,
|
| 2016 |
+
"learning_rate": 3.6353405565986057e-06,
|
| 2017 |
+
"loss": 1.4946,
|
| 2018 |
+
"step": 1430
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.7507193303688203,
|
| 2022 |
+
"grad_norm": 0.9366343276314563,
|
| 2023 |
+
"learning_rate": 3.565126588934803e-06,
|
| 2024 |
+
"loss": 1.494,
|
| 2025 |
+
"step": 1435
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.7533350771645305,
|
| 2029 |
+
"grad_norm": 0.9533093093428183,
|
| 2030 |
+
"learning_rate": 3.4954499313490076e-06,
|
| 2031 |
+
"loss": 1.4997,
|
| 2032 |
+
"step": 1440
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.7559508239602406,
|
| 2036 |
+
"grad_norm": 0.9973613426926783,
|
| 2037 |
+
"learning_rate": 3.426316401822073e-06,
|
| 2038 |
+
"loss": 1.4915,
|
| 2039 |
+
"step": 1445
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.7585665707559508,
|
| 2043 |
+
"grad_norm": 0.9578958384727457,
|
| 2044 |
+
"learning_rate": 3.3577317729838242e-06,
|
| 2045 |
+
"loss": 1.4901,
|
| 2046 |
+
"step": 1450
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.761182317551661,
|
| 2050 |
+
"grad_norm": 0.9780765605337923,
|
| 2051 |
+
"learning_rate": 3.2897017716310296e-06,
|
| 2052 |
+
"loss": 1.489,
|
| 2053 |
+
"step": 1455
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.7637980643473712,
|
| 2057 |
+
"grad_norm": 0.9637546669711574,
|
| 2058 |
+
"learning_rate": 3.2222320782492267e-06,
|
| 2059 |
+
"loss": 1.4762,
|
| 2060 |
+
"step": 1460
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.7664138111430814,
|
| 2064 |
+
"grad_norm": 0.9767551262257389,
|
| 2065 |
+
"learning_rate": 3.1553283265383938e-06,
|
| 2066 |
+
"loss": 1.4969,
|
| 2067 |
+
"step": 1465
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.7690295579387916,
|
| 2071 |
+
"grad_norm": 1.0140641021578174,
|
| 2072 |
+
"learning_rate": 3.0889961029425453e-06,
|
| 2073 |
+
"loss": 1.5161,
|
| 2074 |
+
"step": 1470
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.7716453047345017,
|
| 2078 |
+
"grad_norm": 0.9936236915799145,
|
| 2079 |
+
"learning_rate": 3.0232409461832647e-06,
|
| 2080 |
+
"loss": 1.4892,
|
| 2081 |
+
"step": 1475
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.7742610515302119,
|
| 2085 |
+
"grad_norm": 1.0085205030096636,
|
| 2086 |
+
"learning_rate": 2.958068346797217e-06,
|
| 2087 |
+
"loss": 1.49,
|
| 2088 |
+
"step": 1480
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.776876798325922,
|
| 2092 |
+
"grad_norm": 0.9735744368097925,
|
| 2093 |
+
"learning_rate": 2.893483746677692e-06,
|
| 2094 |
+
"loss": 1.5317,
|
| 2095 |
+
"step": 1485
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.7794925451216322,
|
| 2099 |
+
"grad_norm": 0.9585549703903414,
|
| 2100 |
+
"learning_rate": 2.829492538620222e-06,
|
| 2101 |
+
"loss": 1.4837,
|
| 2102 |
+
"step": 1490
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.7821082919173424,
|
| 2106 |
+
"grad_norm": 0.9511945475302451,
|
| 2107 |
+
"learning_rate": 2.766100065872259e-06,
|
| 2108 |
+
"loss": 1.5412,
|
| 2109 |
+
"step": 1495
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.7847240387130526,
|
| 2113 |
+
"grad_norm": 0.9821297615247657,
|
| 2114 |
+
"learning_rate": 2.7033116216870358e-06,
|
| 2115 |
+
"loss": 1.4978,
|
| 2116 |
+
"step": 1500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.7873397855087627,
|
| 2120 |
+
"grad_norm": 0.9870765810855415,
|
| 2121 |
+
"learning_rate": 2.6411324488815694e-06,
|
| 2122 |
+
"loss": 1.4994,
|
| 2123 |
+
"step": 1505
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.7899555323044729,
|
| 2127 |
+
"grad_norm": 0.988122505793901,
|
| 2128 |
+
"learning_rate": 2.5795677393989015e-06,
|
| 2129 |
+
"loss": 1.4847,
|
| 2130 |
+
"step": 1510
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.7925712791001831,
|
| 2134 |
+
"grad_norm": 0.9815521201411757,
|
| 2135 |
+
"learning_rate": 2.5186226338745532e-06,
|
| 2136 |
+
"loss": 1.5133,
|
| 2137 |
+
"step": 1515
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.7951870258958933,
|
| 2141 |
+
"grad_norm": 0.9788766473931083,
|
| 2142 |
+
"learning_rate": 2.4583022212072983e-06,
|
| 2143 |
+
"loss": 1.5082,
|
| 2144 |
+
"step": 1520
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.7978027726916035,
|
| 2148 |
+
"grad_norm": 0.9782316028601571,
|
| 2149 |
+
"learning_rate": 2.3986115381342347e-06,
|
| 2150 |
+
"loss": 1.4823,
|
| 2151 |
+
"step": 1525
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.8004185194873136,
|
| 2155 |
+
"grad_norm": 0.9601611522592316,
|
| 2156 |
+
"learning_rate": 2.339555568810221e-06,
|
| 2157 |
+
"loss": 1.4809,
|
| 2158 |
+
"step": 1530
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.8030342662830238,
|
| 2162 |
+
"grad_norm": 0.953372174372059,
|
| 2163 |
+
"learning_rate": 2.292771102778739e-06,
|
| 2164 |
+
"loss": 1.5263,
|
| 2165 |
+
"step": 1535
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.805650013078734,
|
| 2169 |
+
"grad_norm": 0.9833349846687767,
|
| 2170 |
+
"learning_rate": 2.2348700088393436e-06,
|
| 2171 |
+
"loss": 1.489,
|
| 2172 |
+
"step": 1540
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.8082657598744442,
|
| 2176 |
+
"grad_norm": 0.9821605598666732,
|
| 2177 |
+
"learning_rate": 2.1776173010195733e-06,
|
| 2178 |
+
"loss": 1.5206,
|
| 2179 |
+
"step": 1545
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.8108815066701544,
|
| 2183 |
+
"grad_norm": 1.0910299524698535,
|
| 2184 |
+
"learning_rate": 2.121017759904056e-06,
|
| 2185 |
+
"loss": 1.4858,
|
| 2186 |
+
"step": 1550
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.8134972534658645,
|
| 2190 |
+
"grad_norm": 0.98556632266862,
|
| 2191 |
+
"learning_rate": 2.0650761115381846e-06,
|
| 2192 |
+
"loss": 1.4898,
|
| 2193 |
+
"step": 1555
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.8161130002615746,
|
| 2197 |
+
"grad_norm": 0.9760806468628352,
|
| 2198 |
+
"learning_rate": 2.009797027033501e-06,
|
| 2199 |
+
"loss": 1.5021,
|
| 2200 |
+
"step": 1560
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.8187287470572848,
|
| 2204 |
+
"grad_norm": 0.9764277586594998,
|
| 2205 |
+
"learning_rate": 1.9551851221776486e-06,
|
| 2206 |
+
"loss": 1.4881,
|
| 2207 |
+
"step": 1565
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.821344493852995,
|
| 2211 |
+
"grad_norm": 1.0023116441587456,
|
| 2212 |
+
"learning_rate": 1.9012449570489744e-06,
|
| 2213 |
+
"loss": 1.4805,
|
| 2214 |
+
"step": 1570
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.8239602406487052,
|
| 2218 |
+
"grad_norm": 0.9501198689907971,
|
| 2219 |
+
"learning_rate": 1.8479810356357385e-06,
|
| 2220 |
+
"loss": 1.4928,
|
| 2221 |
+
"step": 1575
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.8265759874444154,
|
| 2225 |
+
"grad_norm": 0.9636163461126281,
|
| 2226 |
+
"learning_rate": 1.795397805460053e-06,
|
| 2227 |
+
"loss": 1.4976,
|
| 2228 |
+
"step": 1580
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.8291917342401256,
|
| 2232 |
+
"grad_norm": 0.9643978441145337,
|
| 2233 |
+
"learning_rate": 1.743499657206501e-06,
|
| 2234 |
+
"loss": 1.5107,
|
| 2235 |
+
"step": 1585
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.8318074810358357,
|
| 2239 |
+
"grad_norm": 0.9915986002147755,
|
| 2240 |
+
"learning_rate": 1.692290924355533e-06,
|
| 2241 |
+
"loss": 1.5085,
|
| 2242 |
+
"step": 1590
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.8344232278315459,
|
| 2246 |
+
"grad_norm": 0.9909029262132248,
|
| 2247 |
+
"learning_rate": 1.6417758828216013e-06,
|
| 2248 |
+
"loss": 1.5184,
|
| 2249 |
+
"step": 1595
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.8370389746272561,
|
| 2253 |
+
"grad_norm": 0.9747563567696917,
|
| 2254 |
+
"learning_rate": 1.5919587505961387e-06,
|
| 2255 |
+
"loss": 1.5566,
|
| 2256 |
+
"step": 1600
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.8396547214229663,
|
| 2260 |
+
"grad_norm": 0.9653148757059233,
|
| 2261 |
+
"learning_rate": 1.5428436873953478e-06,
|
| 2262 |
+
"loss": 1.4915,
|
| 2263 |
+
"step": 1605
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.8422704682186765,
|
| 2267 |
+
"grad_norm": 1.00376588771245,
|
| 2268 |
+
"learning_rate": 1.4944347943128722e-06,
|
| 2269 |
+
"loss": 1.5186,
|
| 2270 |
+
"step": 1610
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.8448862150143867,
|
| 2274 |
+
"grad_norm": 0.9791331327246297,
|
| 2275 |
+
"learning_rate": 1.446736113477355e-06,
|
| 2276 |
+
"loss": 1.4757,
|
| 2277 |
+
"step": 1615
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.8475019618100967,
|
| 2281 |
+
"grad_norm": 0.9755886397420878,
|
| 2282 |
+
"learning_rate": 1.3997516277149158e-06,
|
| 2283 |
+
"loss": 1.4765,
|
| 2284 |
+
"step": 1620
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.8501177086058069,
|
| 2288 |
+
"grad_norm": 0.991899631468335,
|
| 2289 |
+
"learning_rate": 1.353485260216596e-06,
|
| 2290 |
+
"loss": 1.4987,
|
| 2291 |
+
"step": 1625
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.8527334554015171,
|
| 2295 |
+
"grad_norm": 0.9880862198388226,
|
| 2296 |
+
"learning_rate": 1.3079408742107702e-06,
|
| 2297 |
+
"loss": 1.4803,
|
| 2298 |
+
"step": 1630
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.8553492021972273,
|
| 2302 |
+
"grad_norm": 0.9798766288497291,
|
| 2303 |
+
"learning_rate": 1.2631222726405601e-06,
|
| 2304 |
+
"loss": 1.5138,
|
| 2305 |
+
"step": 1635
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.8579649489929375,
|
| 2309 |
+
"grad_norm": 0.9549696237423514,
|
| 2310 |
+
"learning_rate": 1.2190331978463e-06,
|
| 2311 |
+
"loss": 1.4819,
|
| 2312 |
+
"step": 1640
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.8605806957886477,
|
| 2316 |
+
"grad_norm": 0.9696751794271421,
|
| 2317 |
+
"learning_rate": 1.1756773312530456e-06,
|
| 2318 |
+
"loss": 1.4947,
|
| 2319 |
+
"step": 1645
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.8631964425843578,
|
| 2323 |
+
"grad_norm": 0.9876616871522614,
|
| 2324 |
+
"learning_rate": 1.133058293063184e-06,
|
| 2325 |
+
"loss": 1.4821,
|
| 2326 |
+
"step": 1650
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.865812189380068,
|
| 2330 |
+
"grad_norm": 0.96373367990197,
|
| 2331 |
+
"learning_rate": 1.0911796419541364e-06,
|
| 2332 |
+
"loss": 1.5175,
|
| 2333 |
+
"step": 1655
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.8684279361757782,
|
| 2337 |
+
"grad_norm": 0.9403141810504254,
|
| 2338 |
+
"learning_rate": 1.0500448747812209e-06,
|
| 2339 |
+
"loss": 1.4997,
|
| 2340 |
+
"step": 1660
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.8710436829714884,
|
| 2344 |
+
"grad_norm": 0.9600461810654733,
|
| 2345 |
+
"learning_rate": 1.0096574262856561e-06,
|
| 2346 |
+
"loss": 1.4909,
|
| 2347 |
+
"step": 1665
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.8736594297671986,
|
| 2351 |
+
"grad_norm": 0.9684357751570363,
|
| 2352 |
+
"learning_rate": 9.700206688077707e-07,
|
| 2353 |
+
"loss": 1.4871,
|
| 2354 |
+
"step": 1670
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.8762751765629088,
|
| 2358 |
+
"grad_norm": 0.9527398435644013,
|
| 2359 |
+
"learning_rate": 9.311379120054076e-07,
|
| 2360 |
+
"loss": 1.4863,
|
| 2361 |
+
"step": 1675
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.8788909233586188,
|
| 2365 |
+
"grad_norm": 0.9754611983606123,
|
| 2366 |
+
"learning_rate": 8.930124025775677e-07,
|
| 2367 |
+
"loss": 1.4957,
|
| 2368 |
+
"step": 1680
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.881506670154329,
|
| 2372 |
+
"grad_norm": 0.9580136767362725,
|
| 2373 |
+
"learning_rate": 8.556473239933183e-07,
|
| 2374 |
+
"loss": 1.5122,
|
| 2375 |
+
"step": 1685
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.8841224169500392,
|
| 2379 |
+
"grad_norm": 0.9617716799464289,
|
| 2380 |
+
"learning_rate": 8.19045796225969e-07,
|
| 2381 |
+
"loss": 1.4871,
|
| 2382 |
+
"step": 1690
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.8867381637457494,
|
| 2386 |
+
"grad_norm": 0.9589643389521932,
|
| 2387 |
+
"learning_rate": 7.832108754925549e-07,
|
| 2388 |
+
"loss": 1.4914,
|
| 2389 |
+
"step": 1695
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.8893539105414596,
|
| 2393 |
+
"grad_norm": 0.9894178941556601,
|
| 2394 |
+
"learning_rate": 7.481455539986437e-07,
|
| 2395 |
+
"loss": 1.504,
|
| 2396 |
+
"step": 1700
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.8919696573371697,
|
| 2400 |
+
"grad_norm": 0.9795542957714503,
|
| 2401 |
+
"learning_rate": 7.138527596884881e-07,
|
| 2402 |
+
"loss": 1.4883,
|
| 2403 |
+
"step": 1705
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.8945854041328799,
|
| 2407 |
+
"grad_norm": 0.985817609953172,
|
| 2408 |
+
"learning_rate": 6.803353560005488e-07,
|
| 2409 |
+
"loss": 1.5182,
|
| 2410 |
+
"step": 1710
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.8972011509285901,
|
| 2414 |
+
"grad_norm": 0.9716109322305025,
|
| 2415 |
+
"learning_rate": 6.475961416283838e-07,
|
| 2416 |
+
"loss": 1.5221,
|
| 2417 |
+
"step": 1715
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.8998168977243003,
|
| 2421 |
+
"grad_norm": 0.9780675454207901,
|
| 2422 |
+
"learning_rate": 6.156378502869742e-07,
|
| 2423 |
+
"loss": 1.5115,
|
| 2424 |
+
"step": 1720
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.9024326445200105,
|
| 2428 |
+
"grad_norm": 0.9467432563981524,
|
| 2429 |
+
"learning_rate": 5.844631504844478e-07,
|
| 2430 |
+
"loss": 1.4729,
|
| 2431 |
+
"step": 1725
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.9050483913157207,
|
| 2435 |
+
"grad_norm": 0.9548396375299275,
|
| 2436 |
+
"learning_rate": 5.540746452992684e-07,
|
| 2437 |
+
"loss": 1.4944,
|
| 2438 |
+
"step": 1730
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.9076641381114308,
|
| 2442 |
+
"grad_norm": 1.0117478938427757,
|
| 2443 |
+
"learning_rate": 5.244748721628735e-07,
|
| 2444 |
+
"loss": 1.507,
|
| 2445 |
+
"step": 1735
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.910279884907141,
|
| 2449 |
+
"grad_norm": 0.9458384045652506,
|
| 2450 |
+
"learning_rate": 4.956663026478015e-07,
|
| 2451 |
+
"loss": 1.4857,
|
| 2452 |
+
"step": 1740
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.9128956317028512,
|
| 2456 |
+
"grad_norm": 0.9464520527021981,
|
| 2457 |
+
"learning_rate": 4.676513422613138e-07,
|
| 2458 |
+
"loss": 1.4927,
|
| 2459 |
+
"step": 1745
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.9155113784985613,
|
| 2463 |
+
"grad_norm": 0.9656923991728011,
|
| 2464 |
+
"learning_rate": 4.404323302445457e-07,
|
| 2465 |
+
"loss": 1.501,
|
| 2466 |
+
"step": 1750
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.9181271252942715,
|
| 2470 |
+
"grad_norm": 0.9684809511897617,
|
| 2471 |
+
"learning_rate": 4.140115393771649e-07,
|
| 2472 |
+
"loss": 1.5071,
|
| 2473 |
+
"step": 1755
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.9207428720899817,
|
| 2477 |
+
"grad_norm": 0.954800117328748,
|
| 2478 |
+
"learning_rate": 3.883911757876058e-07,
|
| 2479 |
+
"loss": 1.5018,
|
| 2480 |
+
"step": 1760
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.9233586188856918,
|
| 2484 |
+
"grad_norm": 0.9431176921856136,
|
| 2485 |
+
"learning_rate": 3.6357337876885444e-07,
|
| 2486 |
+
"loss": 1.4717,
|
| 2487 |
+
"step": 1765
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.925974365681402,
|
| 2491 |
+
"grad_norm": 0.9500467270915927,
|
| 2492 |
+
"learning_rate": 3.395602205998205e-07,
|
| 2493 |
+
"loss": 1.5015,
|
| 2494 |
+
"step": 1770
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.9285901124771122,
|
| 2498 |
+
"grad_norm": 0.9619915809450086,
|
| 2499 |
+
"learning_rate": 3.16353706372301e-07,
|
| 2500 |
+
"loss": 1.4883,
|
| 2501 |
+
"step": 1775
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.9312058592728224,
|
| 2505 |
+
"grad_norm": 0.9711037002937901,
|
| 2506 |
+
"learning_rate": 2.9395577382355546e-07,
|
| 2507 |
+
"loss": 1.4865,
|
| 2508 |
+
"step": 1780
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.9338216060685326,
|
| 2512 |
+
"grad_norm": 0.9691613024591422,
|
| 2513 |
+
"learning_rate": 2.723682931745053e-07,
|
| 2514 |
+
"loss": 1.5153,
|
| 2515 |
+
"step": 1785
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.9364373528642428,
|
| 2519 |
+
"grad_norm": 0.9487204339032221,
|
| 2520 |
+
"learning_rate": 2.515930669735733e-07,
|
| 2521 |
+
"loss": 1.5183,
|
| 2522 |
+
"step": 1790
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.9390530996599529,
|
| 2526 |
+
"grad_norm": 0.962148596880566,
|
| 2527 |
+
"learning_rate": 2.3163182994616728e-07,
|
| 2528 |
+
"loss": 1.5133,
|
| 2529 |
+
"step": 1795
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.9416688464556631,
|
| 2533 |
+
"grad_norm": 0.9664661545604885,
|
| 2534 |
+
"learning_rate": 2.1248624884983582e-07,
|
| 2535 |
+
"loss": 1.4835,
|
| 2536 |
+
"step": 1800
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.9442845932513733,
|
| 2540 |
+
"grad_norm": 0.9543308181657987,
|
| 2541 |
+
"learning_rate": 1.941579223350898e-07,
|
| 2542 |
+
"loss": 1.4991,
|
| 2543 |
+
"step": 1805
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.9469003400470835,
|
| 2547 |
+
"grad_norm": 0.9890140540984144,
|
| 2548 |
+
"learning_rate": 1.7664838081192016e-07,
|
| 2549 |
+
"loss": 1.5116,
|
| 2550 |
+
"step": 1810
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.9495160868427936,
|
| 2554 |
+
"grad_norm": 0.9431547186579136,
|
| 2555 |
+
"learning_rate": 1.5995908632200352e-07,
|
| 2556 |
+
"loss": 1.487,
|
| 2557 |
+
"step": 1815
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.9521318336385038,
|
| 2561 |
+
"grad_norm": 0.9424565377771105,
|
| 2562 |
+
"learning_rate": 1.4409143241662871e-07,
|
| 2563 |
+
"loss": 1.506,
|
| 2564 |
+
"step": 1820
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.9547475804342139,
|
| 2568 |
+
"grad_norm": 0.9703580714859561,
|
| 2569 |
+
"learning_rate": 1.2904674404032758e-07,
|
| 2570 |
+
"loss": 1.4905,
|
| 2571 |
+
"step": 1825
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.9573633272299241,
|
| 2575 |
+
"grad_norm": 0.9914509011094129,
|
| 2576 |
+
"learning_rate": 1.1482627742025354e-07,
|
| 2577 |
+
"loss": 1.4905,
|
| 2578 |
+
"step": 1830
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.9599790740256343,
|
| 2582 |
+
"grad_norm": 0.9606889211690712,
|
| 2583 |
+
"learning_rate": 1.0143121996127547e-07,
|
| 2584 |
+
"loss": 1.5137,
|
| 2585 |
+
"step": 1835
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.9625948208213445,
|
| 2589 |
+
"grad_norm": 0.9819397585460637,
|
| 2590 |
+
"learning_rate": 8.886269014683591e-08,
|
| 2591 |
+
"loss": 1.5014,
|
| 2592 |
+
"step": 1840
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.9652105676170547,
|
| 2596 |
+
"grad_norm": 0.9506559349034596,
|
| 2597 |
+
"learning_rate": 7.712173744555796e-08,
|
| 2598 |
+
"loss": 1.5044,
|
| 2599 |
+
"step": 1845
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.9678263144127649,
|
| 2603 |
+
"grad_norm": 0.9654399449302187,
|
| 2604 |
+
"learning_rate": 6.62093422236132e-08,
|
| 2605 |
+
"loss": 1.473,
|
| 2606 |
+
"step": 1850
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.970442061208475,
|
| 2610 |
+
"grad_norm": 0.9744217661983238,
|
| 2611 |
+
"learning_rate": 5.61264156628627e-08,
|
| 2612 |
+
"loss": 1.4749,
|
| 2613 |
+
"step": 1855
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.9730578080041852,
|
| 2617 |
+
"grad_norm": 0.9550439712817645,
|
| 2618 |
+
"learning_rate": 4.687379968477235e-08,
|
| 2619 |
+
"loss": 1.5015,
|
| 2620 |
+
"step": 1860
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.9756735547998954,
|
| 2624 |
+
"grad_norm": 0.9572079676618851,
|
| 2625 |
+
"learning_rate": 3.845226688011239e-08,
|
| 2626 |
+
"loss": 1.4919,
|
| 2627 |
+
"step": 1865
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.9782893015956056,
|
| 2631 |
+
"grad_norm": 0.9649896511988153,
|
| 2632 |
+
"learning_rate": 3.086252044445126e-08,
|
| 2633 |
+
"loss": 1.499,
|
| 2634 |
+
"step": 1870
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.9809050483913158,
|
| 2638 |
+
"grad_norm": 0.9668539201382029,
|
| 2639 |
+
"learning_rate": 2.4105194119429242e-08,
|
| 2640 |
+
"loss": 1.5089,
|
| 2641 |
+
"step": 1875
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.9835207951870258,
|
| 2645 |
+
"grad_norm": 0.9469577423540613,
|
| 2646 |
+
"learning_rate": 1.8180852139851878e-08,
|
| 2647 |
+
"loss": 1.4917,
|
| 2648 |
+
"step": 1880
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.986136541982736,
|
| 2652 |
+
"grad_norm": 0.9720110181031226,
|
| 2653 |
+
"learning_rate": 1.3089989186568785e-08,
|
| 2654 |
+
"loss": 1.5159,
|
| 2655 |
+
"step": 1885
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.9887522887784462,
|
| 2659 |
+
"grad_norm": 0.9562880241516497,
|
| 2660 |
+
"learning_rate": 8.833030345172244e-09,
|
| 2661 |
+
"loss": 1.5145,
|
| 2662 |
+
"step": 1890
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.9913680355741564,
|
| 2666 |
+
"grad_norm": 0.9611852075134648,
|
| 2667 |
+
"learning_rate": 5.410331070498931e-09,
|
| 2668 |
+
"loss": 1.5114,
|
| 2669 |
+
"step": 1895
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.9939837823698666,
|
| 2673 |
+
"grad_norm": 0.958078042967506,
|
| 2674 |
+
"learning_rate": 2.822177156954764e-09,
|
| 2675 |
+
"loss": 1.5175,
|
| 2676 |
+
"step": 1900
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.9965995291655768,
|
| 2680 |
+
"grad_norm": 0.9612631923896605,
|
| 2681 |
+
"learning_rate": 1.0687847146462206e-09,
|
| 2682 |
+
"loss": 1.5166,
|
| 2683 |
+
"step": 1905
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.9992152759612869,
|
| 2687 |
+
"grad_norm": 0.9480114376607924,
|
| 2688 |
+
"learning_rate": 1.5030015133699062e-10,
|
| 2689 |
+
"loss": 1.5055,
|
| 2690 |
+
"step": 1910
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.999738425320429,
|
| 2694 |
+
"eval_loss": 1.5079056024551392,
|
| 2695 |
+
"eval_runtime": 428.141,
|
| 2696 |
+
"eval_samples_per_second": 63.239,
|
| 2697 |
+
"eval_steps_per_second": 3.954,
|
| 2698 |
+
"step": 1911
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.999738425320429,
|
| 2702 |
+
"step": 1911,
|
| 2703 |
+
"total_flos": 107512363745280.0,
|
| 2704 |
+
"train_loss": 1.5265680355653908,
|
| 2705 |
+
"train_runtime": 15553.3872,
|
| 2706 |
+
"train_samples_per_second": 15.73,
|
| 2707 |
+
"train_steps_per_second": 0.123
|
| 2708 |
+
}
|
| 2709 |
+
],
|
| 2710 |
+
"logging_steps": 5,
|
| 2711 |
+
"max_steps": 1911,
|
| 2712 |
+
"num_input_tokens_seen": 0,
|
| 2713 |
+
"num_train_epochs": 1,
|
| 2714 |
+
"save_steps": 500,
|
| 2715 |
+
"stateful_callbacks": {
|
| 2716 |
+
"TrainerControl": {
|
| 2717 |
+
"args": {
|
| 2718 |
+
"should_epoch_stop": false,
|
| 2719 |
+
"should_evaluate": false,
|
| 2720 |
+
"should_log": false,
|
| 2721 |
+
"should_save": false,
|
| 2722 |
+
"should_training_stop": false
|
| 2723 |
+
},
|
| 2724 |
+
"attributes": {}
|
| 2725 |
+
}
|
| 2726 |
+
},
|
| 2727 |
+
"total_flos": 107512363745280.0,
|
| 2728 |
+
"train_batch_size": 8,
|
| 2729 |
+
"trial_name": null,
|
| 2730 |
+
"trial_params": null
|
| 2731 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f492d9e15d1c86f0ab5946584cc5b2239088843e04a1809e6dcc8352888ca551
|
| 3 |
+
size 7032
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|