initial commit
Browse files- .gitattributes +1 -0
- README.md +55 -1
- added_tokens.json +5 -0
- all_results.json +7 -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_config.json +44 -0
- train_results.json +7 -0
- trainer_log.jsonl +0 -0
- trainer_state.json +3964 -0
- training_args.bin +3 -0
- training_loss.png +0 -0
- vocab.json +0 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
model.safetensors filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -1,3 +1,57 @@
|
|
| 1 |
---
|
| 2 |
-
license:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
license: other
|
| 3 |
+
base_model: Qwen/Qwen1.5-1.8B
|
| 4 |
+
tags:
|
| 5 |
+
- llama-factory
|
| 6 |
+
- full
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
model-index:
|
| 9 |
+
- name: train_2024-04-06-14-50-00
|
| 10 |
+
results: []
|
| 11 |
---
|
| 12 |
+
|
| 13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
+
|
| 16 |
+
# train_2024-04-06-14-50-00
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [Qwen/Qwen1.5-1.8B](https://huggingface.co/Qwen/Qwen1.5-1.8B) on the CMB_answer_merge, the belle_0.5M_random_122320 and the HuatuoGPT2_sft_instruct_GPT4 datasets.
|
| 19 |
+
|
| 20 |
+
## Model description
|
| 21 |
+
|
| 22 |
+
More information needed
|
| 23 |
+
|
| 24 |
+
## Intended uses & limitations
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Training and evaluation data
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training procedure
|
| 33 |
+
|
| 34 |
+
### Training hyperparameters
|
| 35 |
+
|
| 36 |
+
The following hyperparameters were used during training:
|
| 37 |
+
- learning_rate: 5e-05
|
| 38 |
+
- train_batch_size: 8
|
| 39 |
+
- eval_batch_size: 8
|
| 40 |
+
- seed: 42
|
| 41 |
+
- gradient_accumulation_steps: 4
|
| 42 |
+
- total_train_batch_size: 32
|
| 43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
+
- lr_scheduler_type: cosine
|
| 45 |
+
- lr_scheduler_warmup_steps: 20
|
| 46 |
+
- num_epochs: 3.0
|
| 47 |
+
|
| 48 |
+
### Training results
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
|
| 52 |
+
### Framework versions
|
| 53 |
+
|
| 54 |
+
- Transformers 4.39.1
|
| 55 |
+
- Pytorch 2.2.0+cu121
|
| 56 |
+
- Datasets 2.18.0
|
| 57 |
+
- Tokenizers 0.15.2
|
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 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 1.199945468658317,
|
| 4 |
+
"train_runtime": 5291.6066,
|
| 5 |
+
"train_samples_per_second": 17.008,
|
| 6 |
+
"train_steps_per_second": 0.531
|
| 7 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Qwen/Qwen1.5-1.8B",
|
| 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": 2048,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 5504,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"max_window_layers": 21,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 16,
|
| 17 |
+
"num_hidden_layers": 24,
|
| 18 |
+
"num_key_value_heads": 16,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 1000000.0,
|
| 21 |
+
"sliding_window": 32768,
|
| 22 |
+
"tie_word_embeddings": false,
|
| 23 |
+
"torch_dtype": "bfloat16",
|
| 24 |
+
"transformers_version": "4.39.1",
|
| 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": "4.39.1"
|
| 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 sha256:2e21376adbd00b7f458d01dd48659c973825a42ba3d9ee721d2cc6c3e1cdbf55
|
| 3 |
+
size 3673690696
|
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_config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message + '\\n' }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ 'Human: ' + content + '\\nAssistant: ' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + '\\n' }}{% endif %}{% endfor %}",
|
| 35 |
+
"clean_up_tokenization_spaces": false,
|
| 36 |
+
"eos_token": "<|endoftext|>",
|
| 37 |
+
"errors": "replace",
|
| 38 |
+
"model_max_length": 32768,
|
| 39 |
+
"pad_token": "<|endoftext|>",
|
| 40 |
+
"padding_side": "right",
|
| 41 |
+
"split_special_tokens": false,
|
| 42 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 43 |
+
"unk_token": null
|
| 44 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 1.199945468658317,
|
| 4 |
+
"train_runtime": 5291.6066,
|
| 5 |
+
"train_samples_per_second": 17.008,
|
| 6 |
+
"train_steps_per_second": 0.531
|
| 7 |
+
}
|
trainer_log.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3964 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9984,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2811,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"grad_norm": 8.1875,
|
| 14 |
+
"learning_rate": 1.25e-05,
|
| 15 |
+
"loss": 1.5548,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01,
|
| 20 |
+
"grad_norm": 7.59375,
|
| 21 |
+
"learning_rate": 2.5e-05,
|
| 22 |
+
"loss": 1.5637,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02,
|
| 27 |
+
"grad_norm": 7.90625,
|
| 28 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 29 |
+
"loss": 1.5441,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02,
|
| 34 |
+
"grad_norm": 7.5,
|
| 35 |
+
"learning_rate": 5e-05,
|
| 36 |
+
"loss": 1.5212,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03,
|
| 41 |
+
"grad_norm": 6.90625,
|
| 42 |
+
"learning_rate": 4.999960406039296e-05,
|
| 43 |
+
"loss": 1.3853,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
+
"grad_norm": 6.59375,
|
| 49 |
+
"learning_rate": 4.9998416254113275e-05,
|
| 50 |
+
"loss": 1.3669,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04,
|
| 55 |
+
"grad_norm": 4.6875,
|
| 56 |
+
"learning_rate": 4.9996436618784923e-05,
|
| 57 |
+
"loss": 1.339,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04,
|
| 62 |
+
"grad_norm": 5.125,
|
| 63 |
+
"learning_rate": 4.9993665217113186e-05,
|
| 64 |
+
"loss": 1.4381,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05,
|
| 69 |
+
"grad_norm": 6.4375,
|
| 70 |
+
"learning_rate": 4.9990102136882665e-05,
|
| 71 |
+
"loss": 1.3781,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.05,
|
| 76 |
+
"grad_norm": 4.875,
|
| 77 |
+
"learning_rate": 4.998574749095454e-05,
|
| 78 |
+
"loss": 1.3353,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.06,
|
| 83 |
+
"grad_norm": 5.90625,
|
| 84 |
+
"learning_rate": 4.998060141726295e-05,
|
| 85 |
+
"loss": 1.326,
|
| 86 |
+
"step": 55
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06,
|
| 90 |
+
"grad_norm": 5.125,
|
| 91 |
+
"learning_rate": 4.997466407881065e-05,
|
| 92 |
+
"loss": 1.314,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.07,
|
| 97 |
+
"grad_norm": 5.21875,
|
| 98 |
+
"learning_rate": 4.996793566366384e-05,
|
| 99 |
+
"loss": 1.2821,
|
| 100 |
+
"step": 65
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.07,
|
| 104 |
+
"grad_norm": 5.46875,
|
| 105 |
+
"learning_rate": 4.99604163849462e-05,
|
| 106 |
+
"loss": 1.3627,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.08,
|
| 111 |
+
"grad_norm": 6.25,
|
| 112 |
+
"learning_rate": 4.995210648083214e-05,
|
| 113 |
+
"loss": 1.3492,
|
| 114 |
+
"step": 75
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.09,
|
| 118 |
+
"grad_norm": 4.59375,
|
| 119 |
+
"learning_rate": 4.994300621453928e-05,
|
| 120 |
+
"loss": 1.3351,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.09,
|
| 125 |
+
"grad_norm": 5.53125,
|
| 126 |
+
"learning_rate": 4.99331158743201e-05,
|
| 127 |
+
"loss": 1.3716,
|
| 128 |
+
"step": 85
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1,
|
| 132 |
+
"grad_norm": 4.78125,
|
| 133 |
+
"learning_rate": 4.992243577345278e-05,
|
| 134 |
+
"loss": 1.3064,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1,
|
| 139 |
+
"grad_norm": 6.0625,
|
| 140 |
+
"learning_rate": 4.9910966250231324e-05,
|
| 141 |
+
"loss": 1.3729,
|
| 142 |
+
"step": 95
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.11,
|
| 146 |
+
"grad_norm": 5.59375,
|
| 147 |
+
"learning_rate": 4.989870766795481e-05,
|
| 148 |
+
"loss": 1.3587,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.11,
|
| 153 |
+
"grad_norm": 6.09375,
|
| 154 |
+
"learning_rate": 4.988566041491589e-05,
|
| 155 |
+
"loss": 1.3231,
|
| 156 |
+
"step": 105
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.12,
|
| 160 |
+
"grad_norm": 6.09375,
|
| 161 |
+
"learning_rate": 4.987182490438852e-05,
|
| 162 |
+
"loss": 1.2753,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.12,
|
| 167 |
+
"grad_norm": 6.09375,
|
| 168 |
+
"learning_rate": 4.985720157461482e-05,
|
| 169 |
+
"loss": 1.2592,
|
| 170 |
+
"step": 115
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.13,
|
| 174 |
+
"grad_norm": 5.1875,
|
| 175 |
+
"learning_rate": 4.984179088879122e-05,
|
| 176 |
+
"loss": 1.3175,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.13,
|
| 181 |
+
"grad_norm": 6.59375,
|
| 182 |
+
"learning_rate": 4.9825593335053796e-05,
|
| 183 |
+
"loss": 1.4169,
|
| 184 |
+
"step": 125
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.14,
|
| 188 |
+
"grad_norm": 5.46875,
|
| 189 |
+
"learning_rate": 4.9808609426462794e-05,
|
| 190 |
+
"loss": 1.3497,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.14,
|
| 195 |
+
"grad_norm": 6.6875,
|
| 196 |
+
"learning_rate": 4.979083970098638e-05,
|
| 197 |
+
"loss": 1.4249,
|
| 198 |
+
"step": 135
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.15,
|
| 202 |
+
"grad_norm": 5.40625,
|
| 203 |
+
"learning_rate": 4.977228472148361e-05,
|
| 204 |
+
"loss": 1.3558,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.15,
|
| 209 |
+
"grad_norm": 6.53125,
|
| 210 |
+
"learning_rate": 4.9752945075686574e-05,
|
| 211 |
+
"loss": 1.371,
|
| 212 |
+
"step": 145
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.16,
|
| 216 |
+
"grad_norm": 5.90625,
|
| 217 |
+
"learning_rate": 4.973282137618182e-05,
|
| 218 |
+
"loss": 1.2899,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.17,
|
| 223 |
+
"grad_norm": 7.96875,
|
| 224 |
+
"learning_rate": 4.9711914260390924e-05,
|
| 225 |
+
"loss": 1.3947,
|
| 226 |
+
"step": 155
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.17,
|
| 230 |
+
"grad_norm": 5.625,
|
| 231 |
+
"learning_rate": 4.96902243905503e-05,
|
| 232 |
+
"loss": 1.2762,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.18,
|
| 237 |
+
"grad_norm": 9.125,
|
| 238 |
+
"learning_rate": 4.966775245369024e-05,
|
| 239 |
+
"loss": 1.392,
|
| 240 |
+
"step": 165
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.18,
|
| 244 |
+
"grad_norm": 6.1875,
|
| 245 |
+
"learning_rate": 4.964449916161312e-05,
|
| 246 |
+
"loss": 1.3061,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.19,
|
| 251 |
+
"grad_norm": 5.53125,
|
| 252 |
+
"learning_rate": 4.9620465250870886e-05,
|
| 253 |
+
"loss": 1.2861,
|
| 254 |
+
"step": 175
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.19,
|
| 258 |
+
"grad_norm": 5.71875,
|
| 259 |
+
"learning_rate": 4.959565148274172e-05,
|
| 260 |
+
"loss": 1.3093,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2,
|
| 265 |
+
"grad_norm": 6.5,
|
| 266 |
+
"learning_rate": 4.95700586432059e-05,
|
| 267 |
+
"loss": 1.2894,
|
| 268 |
+
"step": 185
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2,
|
| 272 |
+
"grad_norm": 6.3125,
|
| 273 |
+
"learning_rate": 4.954368754292095e-05,
|
| 274 |
+
"loss": 1.2801,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.21,
|
| 279 |
+
"grad_norm": 4.75,
|
| 280 |
+
"learning_rate": 4.951653901719589e-05,
|
| 281 |
+
"loss": 1.3643,
|
| 282 |
+
"step": 195
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.21,
|
| 286 |
+
"grad_norm": 5.125,
|
| 287 |
+
"learning_rate": 4.9488613925964886e-05,
|
| 288 |
+
"loss": 1.2965,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.22,
|
| 293 |
+
"grad_norm": 5.125,
|
| 294 |
+
"learning_rate": 4.945991315375987e-05,
|
| 295 |
+
"loss": 1.3055,
|
| 296 |
+
"step": 205
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.22,
|
| 300 |
+
"grad_norm": 4.53125,
|
| 301 |
+
"learning_rate": 4.943043760968267e-05,
|
| 302 |
+
"loss": 1.2909,
|
| 303 |
+
"step": 210
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.23,
|
| 307 |
+
"grad_norm": 6.8125,
|
| 308 |
+
"learning_rate": 4.940018822737609e-05,
|
| 309 |
+
"loss": 1.3486,
|
| 310 |
+
"step": 215
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.23,
|
| 314 |
+
"grad_norm": 6.625,
|
| 315 |
+
"learning_rate": 4.9369165964994434e-05,
|
| 316 |
+
"loss": 1.3847,
|
| 317 |
+
"step": 220
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.24,
|
| 321 |
+
"grad_norm": 5.125,
|
| 322 |
+
"learning_rate": 4.9337371805173074e-05,
|
| 323 |
+
"loss": 1.3065,
|
| 324 |
+
"step": 225
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.25,
|
| 328 |
+
"grad_norm": 5.75,
|
| 329 |
+
"learning_rate": 4.93048067549974e-05,
|
| 330 |
+
"loss": 1.309,
|
| 331 |
+
"step": 230
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.25,
|
| 335 |
+
"grad_norm": 5.21875,
|
| 336 |
+
"learning_rate": 4.9271471845970854e-05,
|
| 337 |
+
"loss": 1.2953,
|
| 338 |
+
"step": 235
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.26,
|
| 342 |
+
"grad_norm": 5.59375,
|
| 343 |
+
"learning_rate": 4.9237368133982286e-05,
|
| 344 |
+
"loss": 1.3291,
|
| 345 |
+
"step": 240
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.26,
|
| 349 |
+
"grad_norm": 6.25,
|
| 350 |
+
"learning_rate": 4.920249669927255e-05,
|
| 351 |
+
"loss": 1.2906,
|
| 352 |
+
"step": 245
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.27,
|
| 356 |
+
"grad_norm": 5.46875,
|
| 357 |
+
"learning_rate": 4.916685864640021e-05,
|
| 358 |
+
"loss": 1.3583,
|
| 359 |
+
"step": 250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.27,
|
| 363 |
+
"grad_norm": 5.5625,
|
| 364 |
+
"learning_rate": 4.913045510420658e-05,
|
| 365 |
+
"loss": 1.2911,
|
| 366 |
+
"step": 255
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.28,
|
| 370 |
+
"grad_norm": 4.46875,
|
| 371 |
+
"learning_rate": 4.9093287225780004e-05,
|
| 372 |
+
"loss": 1.2865,
|
| 373 |
+
"step": 260
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.28,
|
| 377 |
+
"grad_norm": 4.9375,
|
| 378 |
+
"learning_rate": 4.905535618841931e-05,
|
| 379 |
+
"loss": 1.3438,
|
| 380 |
+
"step": 265
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.29,
|
| 384 |
+
"grad_norm": 6.53125,
|
| 385 |
+
"learning_rate": 4.901666319359649e-05,
|
| 386 |
+
"loss": 1.3012,
|
| 387 |
+
"step": 270
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.29,
|
| 391 |
+
"grad_norm": 5.84375,
|
| 392 |
+
"learning_rate": 4.897720946691867e-05,
|
| 393 |
+
"loss": 1.313,
|
| 394 |
+
"step": 275
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3,
|
| 398 |
+
"grad_norm": 4.71875,
|
| 399 |
+
"learning_rate": 4.893699625808931e-05,
|
| 400 |
+
"loss": 1.3882,
|
| 401 |
+
"step": 280
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.3,
|
| 405 |
+
"grad_norm": 6.40625,
|
| 406 |
+
"learning_rate": 4.889602484086857e-05,
|
| 407 |
+
"loss": 1.2409,
|
| 408 |
+
"step": 285
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.31,
|
| 412 |
+
"grad_norm": 5.46875,
|
| 413 |
+
"learning_rate": 4.8854296513032996e-05,
|
| 414 |
+
"loss": 1.3014,
|
| 415 |
+
"step": 290
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.31,
|
| 419 |
+
"grad_norm": 5.125,
|
| 420 |
+
"learning_rate": 4.8811812596334395e-05,
|
| 421 |
+
"loss": 1.3349,
|
| 422 |
+
"step": 295
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.32,
|
| 426 |
+
"grad_norm": 6.34375,
|
| 427 |
+
"learning_rate": 4.8768574436458e-05,
|
| 428 |
+
"loss": 1.2216,
|
| 429 |
+
"step": 300
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.33,
|
| 433 |
+
"grad_norm": 5.5,
|
| 434 |
+
"learning_rate": 4.8724583402979826e-05,
|
| 435 |
+
"loss": 1.27,
|
| 436 |
+
"step": 305
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.33,
|
| 440 |
+
"grad_norm": 5.40625,
|
| 441 |
+
"learning_rate": 4.867984088932326e-05,
|
| 442 |
+
"loss": 1.2596,
|
| 443 |
+
"step": 310
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.34,
|
| 447 |
+
"grad_norm": 6.21875,
|
| 448 |
+
"learning_rate": 4.8634348312714956e-05,
|
| 449 |
+
"loss": 1.3084,
|
| 450 |
+
"step": 315
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.34,
|
| 454 |
+
"grad_norm": 5.625,
|
| 455 |
+
"learning_rate": 4.8588107114139955e-05,
|
| 456 |
+
"loss": 1.3364,
|
| 457 |
+
"step": 320
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.35,
|
| 461 |
+
"grad_norm": 5.40625,
|
| 462 |
+
"learning_rate": 4.8541118758296025e-05,
|
| 463 |
+
"loss": 1.1912,
|
| 464 |
+
"step": 325
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.35,
|
| 468 |
+
"grad_norm": 5.03125,
|
| 469 |
+
"learning_rate": 4.849338473354725e-05,
|
| 470 |
+
"loss": 1.3021,
|
| 471 |
+
"step": 330
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.36,
|
| 475 |
+
"grad_norm": 6.0,
|
| 476 |
+
"learning_rate": 4.8444906551876914e-05,
|
| 477 |
+
"loss": 1.2595,
|
| 478 |
+
"step": 335
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.36,
|
| 482 |
+
"grad_norm": 5.25,
|
| 483 |
+
"learning_rate": 4.839568574883958e-05,
|
| 484 |
+
"loss": 1.2704,
|
| 485 |
+
"step": 340
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.37,
|
| 489 |
+
"grad_norm": 5.25,
|
| 490 |
+
"learning_rate": 4.83457238835125e-05,
|
| 491 |
+
"loss": 1.4043,
|
| 492 |
+
"step": 345
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.37,
|
| 496 |
+
"grad_norm": 6.5625,
|
| 497 |
+
"learning_rate": 4.8295022538446164e-05,
|
| 498 |
+
"loss": 1.3419,
|
| 499 |
+
"step": 350
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.38,
|
| 503 |
+
"grad_norm": 5.75,
|
| 504 |
+
"learning_rate": 4.824358331961423e-05,
|
| 505 |
+
"loss": 1.2798,
|
| 506 |
+
"step": 355
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.38,
|
| 510 |
+
"grad_norm": 5.40625,
|
| 511 |
+
"learning_rate": 4.819140785636262e-05,
|
| 512 |
+
"loss": 1.3005,
|
| 513 |
+
"step": 360
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.39,
|
| 517 |
+
"grad_norm": 4.9375,
|
| 518 |
+
"learning_rate": 4.813849780135795e-05,
|
| 519 |
+
"loss": 1.3454,
|
| 520 |
+
"step": 365
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.39,
|
| 524 |
+
"grad_norm": 5.40625,
|
| 525 |
+
"learning_rate": 4.8084854830535095e-05,
|
| 526 |
+
"loss": 1.3375,
|
| 527 |
+
"step": 370
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.4,
|
| 531 |
+
"grad_norm": 6.09375,
|
| 532 |
+
"learning_rate": 4.803048064304423e-05,
|
| 533 |
+
"loss": 1.3316,
|
| 534 |
+
"step": 375
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.41,
|
| 538 |
+
"grad_norm": 4.875,
|
| 539 |
+
"learning_rate": 4.797537696119689e-05,
|
| 540 |
+
"loss": 1.2503,
|
| 541 |
+
"step": 380
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.41,
|
| 545 |
+
"grad_norm": 6.375,
|
| 546 |
+
"learning_rate": 4.7919545530411494e-05,
|
| 547 |
+
"loss": 1.2164,
|
| 548 |
+
"step": 385
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.42,
|
| 552 |
+
"grad_norm": 7.28125,
|
| 553 |
+
"learning_rate": 4.786298811915801e-05,
|
| 554 |
+
"loss": 1.2928,
|
| 555 |
+
"step": 390
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.42,
|
| 559 |
+
"grad_norm": 5.8125,
|
| 560 |
+
"learning_rate": 4.7805706518902e-05,
|
| 561 |
+
"loss": 1.2657,
|
| 562 |
+
"step": 395
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.43,
|
| 566 |
+
"grad_norm": 5.375,
|
| 567 |
+
"learning_rate": 4.774770254404779e-05,
|
| 568 |
+
"loss": 1.2557,
|
| 569 |
+
"step": 400
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.43,
|
| 573 |
+
"grad_norm": 7.90625,
|
| 574 |
+
"learning_rate": 4.768897803188105e-05,
|
| 575 |
+
"loss": 1.3147,
|
| 576 |
+
"step": 405
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.44,
|
| 580 |
+
"grad_norm": 5.8125,
|
| 581 |
+
"learning_rate": 4.7629534842510626e-05,
|
| 582 |
+
"loss": 1.289,
|
| 583 |
+
"step": 410
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.44,
|
| 587 |
+
"grad_norm": 5.96875,
|
| 588 |
+
"learning_rate": 4.7569374858809546e-05,
|
| 589 |
+
"loss": 1.2597,
|
| 590 |
+
"step": 415
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.45,
|
| 594 |
+
"grad_norm": 6.53125,
|
| 595 |
+
"learning_rate": 4.750849998635544e-05,
|
| 596 |
+
"loss": 1.247,
|
| 597 |
+
"step": 420
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.45,
|
| 601 |
+
"grad_norm": 8.0625,
|
| 602 |
+
"learning_rate": 4.744691215337015e-05,
|
| 603 |
+
"loss": 1.3946,
|
| 604 |
+
"step": 425
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.46,
|
| 608 |
+
"grad_norm": 8.8125,
|
| 609 |
+
"learning_rate": 4.738461331065867e-05,
|
| 610 |
+
"loss": 1.3241,
|
| 611 |
+
"step": 430
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.46,
|
| 615 |
+
"grad_norm": 5.9375,
|
| 616 |
+
"learning_rate": 4.732160543154735e-05,
|
| 617 |
+
"loss": 1.2764,
|
| 618 |
+
"step": 435
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.47,
|
| 622 |
+
"grad_norm": 10.125,
|
| 623 |
+
"learning_rate": 4.725789051182136e-05,
|
| 624 |
+
"loss": 1.2941,
|
| 625 |
+
"step": 440
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.47,
|
| 629 |
+
"grad_norm": 6.03125,
|
| 630 |
+
"learning_rate": 4.719347056966155e-05,
|
| 631 |
+
"loss": 1.3391,
|
| 632 |
+
"step": 445
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.48,
|
| 636 |
+
"grad_norm": 7.40625,
|
| 637 |
+
"learning_rate": 4.7128347645580434e-05,
|
| 638 |
+
"loss": 1.3739,
|
| 639 |
+
"step": 450
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.49,
|
| 643 |
+
"grad_norm": 6.34375,
|
| 644 |
+
"learning_rate": 4.7062523802357606e-05,
|
| 645 |
+
"loss": 1.2787,
|
| 646 |
+
"step": 455
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.49,
|
| 650 |
+
"grad_norm": 6.4375,
|
| 651 |
+
"learning_rate": 4.699600112497441e-05,
|
| 652 |
+
"loss": 1.2488,
|
| 653 |
+
"step": 460
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5,
|
| 657 |
+
"grad_norm": 5.3125,
|
| 658 |
+
"learning_rate": 4.692878172054784e-05,
|
| 659 |
+
"loss": 1.217,
|
| 660 |
+
"step": 465
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.5,
|
| 664 |
+
"grad_norm": 5.65625,
|
| 665 |
+
"learning_rate": 4.686086771826389e-05,
|
| 666 |
+
"loss": 1.2612,
|
| 667 |
+
"step": 470
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.51,
|
| 671 |
+
"grad_norm": 5.59375,
|
| 672 |
+
"learning_rate": 4.679226126931001e-05,
|
| 673 |
+
"loss": 1.2482,
|
| 674 |
+
"step": 475
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.51,
|
| 678 |
+
"grad_norm": 4.625,
|
| 679 |
+
"learning_rate": 4.672296454680705e-05,
|
| 680 |
+
"loss": 1.2179,
|
| 681 |
+
"step": 480
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.52,
|
| 685 |
+
"grad_norm": 7.28125,
|
| 686 |
+
"learning_rate": 4.665297974574037e-05,
|
| 687 |
+
"loss": 1.2427,
|
| 688 |
+
"step": 485
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.52,
|
| 692 |
+
"grad_norm": 5.34375,
|
| 693 |
+
"learning_rate": 4.6582309082890327e-05,
|
| 694 |
+
"loss": 1.3199,
|
| 695 |
+
"step": 490
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.53,
|
| 699 |
+
"grad_norm": 5.46875,
|
| 700 |
+
"learning_rate": 4.6510954796762094e-05,
|
| 701 |
+
"loss": 1.2884,
|
| 702 |
+
"step": 495
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.53,
|
| 706 |
+
"grad_norm": 5.8125,
|
| 707 |
+
"learning_rate": 4.643891914751472e-05,
|
| 708 |
+
"loss": 1.3176,
|
| 709 |
+
"step": 500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.54,
|
| 713 |
+
"grad_norm": 5.40625,
|
| 714 |
+
"learning_rate": 4.6366204416889526e-05,
|
| 715 |
+
"loss": 1.361,
|
| 716 |
+
"step": 505
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.54,
|
| 720 |
+
"grad_norm": 6.6875,
|
| 721 |
+
"learning_rate": 4.629281290813785e-05,
|
| 722 |
+
"loss": 1.3099,
|
| 723 |
+
"step": 510
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.55,
|
| 727 |
+
"grad_norm": 4.96875,
|
| 728 |
+
"learning_rate": 4.6218746945948124e-05,
|
| 729 |
+
"loss": 1.2626,
|
| 730 |
+
"step": 515
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.55,
|
| 734 |
+
"grad_norm": 5.71875,
|
| 735 |
+
"learning_rate": 4.614400887637218e-05,
|
| 736 |
+
"loss": 1.2902,
|
| 737 |
+
"step": 520
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.56,
|
| 741 |
+
"grad_norm": 4.6875,
|
| 742 |
+
"learning_rate": 4.6068601066750964e-05,
|
| 743 |
+
"loss": 1.242,
|
| 744 |
+
"step": 525
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.57,
|
| 748 |
+
"grad_norm": 5.53125,
|
| 749 |
+
"learning_rate": 4.599252590563956e-05,
|
| 750 |
+
"loss": 1.2424,
|
| 751 |
+
"step": 530
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.57,
|
| 755 |
+
"grad_norm": 4.9375,
|
| 756 |
+
"learning_rate": 4.5915785802731515e-05,
|
| 757 |
+
"loss": 1.263,
|
| 758 |
+
"step": 535
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.58,
|
| 762 |
+
"grad_norm": 5.625,
|
| 763 |
+
"learning_rate": 4.583838318878253e-05,
|
| 764 |
+
"loss": 1.237,
|
| 765 |
+
"step": 540
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.58,
|
| 769 |
+
"grad_norm": 5.8125,
|
| 770 |
+
"learning_rate": 4.576032051553345e-05,
|
| 771 |
+
"loss": 1.3625,
|
| 772 |
+
"step": 545
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.59,
|
| 776 |
+
"grad_norm": 6.34375,
|
| 777 |
+
"learning_rate": 4.5681600255632605e-05,
|
| 778 |
+
"loss": 1.2604,
|
| 779 |
+
"step": 550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.59,
|
| 783 |
+
"grad_norm": 4.875,
|
| 784 |
+
"learning_rate": 4.56022249025575e-05,
|
| 785 |
+
"loss": 1.4195,
|
| 786 |
+
"step": 555
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6,
|
| 790 |
+
"grad_norm": 6.03125,
|
| 791 |
+
"learning_rate": 4.5522196970535825e-05,
|
| 792 |
+
"loss": 1.2599,
|
| 793 |
+
"step": 560
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6,
|
| 797 |
+
"grad_norm": 5.21875,
|
| 798 |
+
"learning_rate": 4.5441518994465806e-05,
|
| 799 |
+
"loss": 1.3454,
|
| 800 |
+
"step": 565
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.61,
|
| 804 |
+
"grad_norm": 6.46875,
|
| 805 |
+
"learning_rate": 4.5360193529835935e-05,
|
| 806 |
+
"loss": 1.3825,
|
| 807 |
+
"step": 570
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.61,
|
| 811 |
+
"grad_norm": 5.8125,
|
| 812 |
+
"learning_rate": 4.527822315264403e-05,
|
| 813 |
+
"loss": 1.2646,
|
| 814 |
+
"step": 575
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.62,
|
| 818 |
+
"grad_norm": 5.78125,
|
| 819 |
+
"learning_rate": 4.519561045931559e-05,
|
| 820 |
+
"loss": 1.262,
|
| 821 |
+
"step": 580
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.62,
|
| 825 |
+
"grad_norm": 5.34375,
|
| 826 |
+
"learning_rate": 4.5112358066621605e-05,
|
| 827 |
+
"loss": 1.2725,
|
| 828 |
+
"step": 585
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.63,
|
| 832 |
+
"grad_norm": 8.1875,
|
| 833 |
+
"learning_rate": 4.502846861159565e-05,
|
| 834 |
+
"loss": 1.282,
|
| 835 |
+
"step": 590
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.63,
|
| 839 |
+
"grad_norm": 5.09375,
|
| 840 |
+
"learning_rate": 4.494394475145036e-05,
|
| 841 |
+
"loss": 1.292,
|
| 842 |
+
"step": 595
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.64,
|
| 846 |
+
"grad_norm": 5.84375,
|
| 847 |
+
"learning_rate": 4.485878916349324e-05,
|
| 848 |
+
"loss": 1.2546,
|
| 849 |
+
"step": 600
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.65,
|
| 853 |
+
"grad_norm": 6.0,
|
| 854 |
+
"learning_rate": 4.4773004545041894e-05,
|
| 855 |
+
"loss": 1.2673,
|
| 856 |
+
"step": 605
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.65,
|
| 860 |
+
"grad_norm": 5.09375,
|
| 861 |
+
"learning_rate": 4.468659361333857e-05,
|
| 862 |
+
"loss": 1.2391,
|
| 863 |
+
"step": 610
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.66,
|
| 867 |
+
"grad_norm": 8.4375,
|
| 868 |
+
"learning_rate": 4.4599559105464106e-05,
|
| 869 |
+
"loss": 1.3886,
|
| 870 |
+
"step": 615
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.66,
|
| 874 |
+
"grad_norm": 5.1875,
|
| 875 |
+
"learning_rate": 4.451190377825121e-05,
|
| 876 |
+
"loss": 1.1826,
|
| 877 |
+
"step": 620
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.67,
|
| 881 |
+
"grad_norm": 5.375,
|
| 882 |
+
"learning_rate": 4.442363040819714e-05,
|
| 883 |
+
"loss": 1.215,
|
| 884 |
+
"step": 625
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.67,
|
| 888 |
+
"grad_norm": 5.53125,
|
| 889 |
+
"learning_rate": 4.433474179137577e-05,
|
| 890 |
+
"loss": 1.18,
|
| 891 |
+
"step": 630
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.68,
|
| 895 |
+
"grad_norm": 5.875,
|
| 896 |
+
"learning_rate": 4.424524074334902e-05,
|
| 897 |
+
"loss": 1.2138,
|
| 898 |
+
"step": 635
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.68,
|
| 902 |
+
"grad_norm": 5.40625,
|
| 903 |
+
"learning_rate": 4.4155130099077676e-05,
|
| 904 |
+
"loss": 1.2433,
|
| 905 |
+
"step": 640
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.69,
|
| 909 |
+
"grad_norm": 5.0,
|
| 910 |
+
"learning_rate": 4.406441271283159e-05,
|
| 911 |
+
"loss": 1.2532,
|
| 912 |
+
"step": 645
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.69,
|
| 916 |
+
"grad_norm": 4.40625,
|
| 917 |
+
"learning_rate": 4.3973091458099266e-05,
|
| 918 |
+
"loss": 1.263,
|
| 919 |
+
"step": 650
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.7,
|
| 923 |
+
"grad_norm": 5.9375,
|
| 924 |
+
"learning_rate": 4.388116922749683e-05,
|
| 925 |
+
"loss": 1.2687,
|
| 926 |
+
"step": 655
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.7,
|
| 930 |
+
"grad_norm": 6.46875,
|
| 931 |
+
"learning_rate": 4.3788648932676436e-05,
|
| 932 |
+
"loss": 1.2414,
|
| 933 |
+
"step": 660
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.71,
|
| 937 |
+
"grad_norm": 5.5,
|
| 938 |
+
"learning_rate": 4.369553350423401e-05,
|
| 939 |
+
"loss": 1.2109,
|
| 940 |
+
"step": 665
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.71,
|
| 944 |
+
"grad_norm": 5.34375,
|
| 945 |
+
"learning_rate": 4.3601825891616453e-05,
|
| 946 |
+
"loss": 1.3227,
|
| 947 |
+
"step": 670
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.72,
|
| 951 |
+
"grad_norm": 5.0625,
|
| 952 |
+
"learning_rate": 4.35075290630282e-05,
|
| 953 |
+
"loss": 1.2769,
|
| 954 |
+
"step": 675
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.73,
|
| 958 |
+
"grad_norm": 5.59375,
|
| 959 |
+
"learning_rate": 4.3412646005337166e-05,
|
| 960 |
+
"loss": 1.2736,
|
| 961 |
+
"step": 680
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.73,
|
| 965 |
+
"grad_norm": 5.875,
|
| 966 |
+
"learning_rate": 4.331717972398022e-05,
|
| 967 |
+
"loss": 1.1944,
|
| 968 |
+
"step": 685
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.74,
|
| 972 |
+
"grad_norm": 5.875,
|
| 973 |
+
"learning_rate": 4.32211332428679e-05,
|
| 974 |
+
"loss": 1.2209,
|
| 975 |
+
"step": 690
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.74,
|
| 979 |
+
"grad_norm": 8.0625,
|
| 980 |
+
"learning_rate": 4.31245096042887e-05,
|
| 981 |
+
"loss": 1.2117,
|
| 982 |
+
"step": 695
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.75,
|
| 986 |
+
"grad_norm": 6.03125,
|
| 987 |
+
"learning_rate": 4.302731186881265e-05,
|
| 988 |
+
"loss": 1.2603,
|
| 989 |
+
"step": 700
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.75,
|
| 993 |
+
"grad_norm": 5.5625,
|
| 994 |
+
"learning_rate": 4.29295431151944e-05,
|
| 995 |
+
"loss": 1.3085,
|
| 996 |
+
"step": 705
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.76,
|
| 1000 |
+
"grad_norm": 5.5,
|
| 1001 |
+
"learning_rate": 4.283120644027571e-05,
|
| 1002 |
+
"loss": 1.1815,
|
| 1003 |
+
"step": 710
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.76,
|
| 1007 |
+
"grad_norm": 5.1875,
|
| 1008 |
+
"learning_rate": 4.2732304958887336e-05,
|
| 1009 |
+
"loss": 1.2594,
|
| 1010 |
+
"step": 715
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.77,
|
| 1014 |
+
"grad_norm": 5.65625,
|
| 1015 |
+
"learning_rate": 4.2632841803750365e-05,
|
| 1016 |
+
"loss": 1.2502,
|
| 1017 |
+
"step": 720
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.77,
|
| 1021 |
+
"grad_norm": 7.09375,
|
| 1022 |
+
"learning_rate": 4.253282012537701e-05,
|
| 1023 |
+
"loss": 1.3242,
|
| 1024 |
+
"step": 725
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.78,
|
| 1028 |
+
"grad_norm": 6.65625,
|
| 1029 |
+
"learning_rate": 4.2432243091970784e-05,
|
| 1030 |
+
"loss": 1.286,
|
| 1031 |
+
"step": 730
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.78,
|
| 1035 |
+
"grad_norm": 5.625,
|
| 1036 |
+
"learning_rate": 4.233111388932618e-05,
|
| 1037 |
+
"loss": 1.2395,
|
| 1038 |
+
"step": 735
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.79,
|
| 1042 |
+
"grad_norm": 5.1875,
|
| 1043 |
+
"learning_rate": 4.222943572072773e-05,
|
| 1044 |
+
"loss": 1.2041,
|
| 1045 |
+
"step": 740
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.79,
|
| 1049 |
+
"grad_norm": 4.375,
|
| 1050 |
+
"learning_rate": 4.2127211806848575e-05,
|
| 1051 |
+
"loss": 1.2588,
|
| 1052 |
+
"step": 745
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.8,
|
| 1056 |
+
"grad_norm": 4.8125,
|
| 1057 |
+
"learning_rate": 4.202444538564842e-05,
|
| 1058 |
+
"loss": 1.2544,
|
| 1059 |
+
"step": 750
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.81,
|
| 1063 |
+
"grad_norm": 5.3125,
|
| 1064 |
+
"learning_rate": 4.192113971227097e-05,
|
| 1065 |
+
"loss": 1.2406,
|
| 1066 |
+
"step": 755
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.81,
|
| 1070 |
+
"grad_norm": 5.6875,
|
| 1071 |
+
"learning_rate": 4.181729805894083e-05,
|
| 1072 |
+
"loss": 1.3559,
|
| 1073 |
+
"step": 760
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.82,
|
| 1077 |
+
"grad_norm": 6.0625,
|
| 1078 |
+
"learning_rate": 4.171292371485991e-05,
|
| 1079 |
+
"loss": 1.2854,
|
| 1080 |
+
"step": 765
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.82,
|
| 1084 |
+
"grad_norm": 5.78125,
|
| 1085 |
+
"learning_rate": 4.160801998610312e-05,
|
| 1086 |
+
"loss": 1.2643,
|
| 1087 |
+
"step": 770
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.83,
|
| 1091 |
+
"grad_norm": 4.625,
|
| 1092 |
+
"learning_rate": 4.150259019551377e-05,
|
| 1093 |
+
"loss": 1.1756,
|
| 1094 |
+
"step": 775
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.83,
|
| 1098 |
+
"grad_norm": 5.78125,
|
| 1099 |
+
"learning_rate": 4.1396637682598235e-05,
|
| 1100 |
+
"loss": 1.2207,
|
| 1101 |
+
"step": 780
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.84,
|
| 1105 |
+
"grad_norm": 7.96875,
|
| 1106 |
+
"learning_rate": 4.1290165803420235e-05,
|
| 1107 |
+
"loss": 1.2095,
|
| 1108 |
+
"step": 785
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.84,
|
| 1112 |
+
"grad_norm": 5.9375,
|
| 1113 |
+
"learning_rate": 4.118317793049448e-05,
|
| 1114 |
+
"loss": 1.2021,
|
| 1115 |
+
"step": 790
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.85,
|
| 1119 |
+
"grad_norm": 5.25,
|
| 1120 |
+
"learning_rate": 4.107567745267987e-05,
|
| 1121 |
+
"loss": 1.2172,
|
| 1122 |
+
"step": 795
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.85,
|
| 1126 |
+
"grad_norm": 7.28125,
|
| 1127 |
+
"learning_rate": 4.096766777507218e-05,
|
| 1128 |
+
"loss": 1.1723,
|
| 1129 |
+
"step": 800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.86,
|
| 1133 |
+
"grad_norm": 5.3125,
|
| 1134 |
+
"learning_rate": 4.085915231889615e-05,
|
| 1135 |
+
"loss": 1.2254,
|
| 1136 |
+
"step": 805
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.86,
|
| 1140 |
+
"grad_norm": 6.6875,
|
| 1141 |
+
"learning_rate": 4.075013452139714e-05,
|
| 1142 |
+
"loss": 1.2538,
|
| 1143 |
+
"step": 810
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.87,
|
| 1147 |
+
"grad_norm": 5.09375,
|
| 1148 |
+
"learning_rate": 4.064061783573227e-05,
|
| 1149 |
+
"loss": 1.1936,
|
| 1150 |
+
"step": 815
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.87,
|
| 1154 |
+
"grad_norm": 5.1875,
|
| 1155 |
+
"learning_rate": 4.053060573086101e-05,
|
| 1156 |
+
"loss": 1.19,
|
| 1157 |
+
"step": 820
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.88,
|
| 1161 |
+
"grad_norm": 6.125,
|
| 1162 |
+
"learning_rate": 4.042010169143533e-05,
|
| 1163 |
+
"loss": 1.2572,
|
| 1164 |
+
"step": 825
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.89,
|
| 1168 |
+
"grad_norm": 6.6875,
|
| 1169 |
+
"learning_rate": 4.030910921768931e-05,
|
| 1170 |
+
"loss": 1.2529,
|
| 1171 |
+
"step": 830
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.89,
|
| 1175 |
+
"grad_norm": 7.34375,
|
| 1176 |
+
"learning_rate": 4.0197631825328264e-05,
|
| 1177 |
+
"loss": 1.2959,
|
| 1178 |
+
"step": 835
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.9,
|
| 1182 |
+
"grad_norm": 5.84375,
|
| 1183 |
+
"learning_rate": 4.008567304541737e-05,
|
| 1184 |
+
"loss": 1.3033,
|
| 1185 |
+
"step": 840
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.9,
|
| 1189 |
+
"grad_norm": 5.4375,
|
| 1190 |
+
"learning_rate": 3.9973236424269876e-05,
|
| 1191 |
+
"loss": 1.3391,
|
| 1192 |
+
"step": 845
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.91,
|
| 1196 |
+
"grad_norm": 4.9375,
|
| 1197 |
+
"learning_rate": 3.9860325523334695e-05,
|
| 1198 |
+
"loss": 1.2128,
|
| 1199 |
+
"step": 850
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.91,
|
| 1203 |
+
"grad_norm": 5.15625,
|
| 1204 |
+
"learning_rate": 3.9746943919083665e-05,
|
| 1205 |
+
"loss": 1.2255,
|
| 1206 |
+
"step": 855
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.92,
|
| 1210 |
+
"grad_norm": 7.46875,
|
| 1211 |
+
"learning_rate": 3.963309520289819e-05,
|
| 1212 |
+
"loss": 1.2787,
|
| 1213 |
+
"step": 860
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.92,
|
| 1217 |
+
"grad_norm": 7.625,
|
| 1218 |
+
"learning_rate": 3.951878298095556e-05,
|
| 1219 |
+
"loss": 1.4147,
|
| 1220 |
+
"step": 865
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.93,
|
| 1224 |
+
"grad_norm": 5.65625,
|
| 1225 |
+
"learning_rate": 3.940401087411467e-05,
|
| 1226 |
+
"loss": 1.2919,
|
| 1227 |
+
"step": 870
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.93,
|
| 1231 |
+
"grad_norm": 5.96875,
|
| 1232 |
+
"learning_rate": 3.928878251780135e-05,
|
| 1233 |
+
"loss": 1.291,
|
| 1234 |
+
"step": 875
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.94,
|
| 1238 |
+
"grad_norm": 5.78125,
|
| 1239 |
+
"learning_rate": 3.91731015618932e-05,
|
| 1240 |
+
"loss": 1.2669,
|
| 1241 |
+
"step": 880
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.94,
|
| 1245 |
+
"grad_norm": 6.28125,
|
| 1246 |
+
"learning_rate": 3.905697167060402e-05,
|
| 1247 |
+
"loss": 1.2935,
|
| 1248 |
+
"step": 885
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.95,
|
| 1252 |
+
"grad_norm": 5.03125,
|
| 1253 |
+
"learning_rate": 3.894039652236767e-05,
|
| 1254 |
+
"loss": 1.3249,
|
| 1255 |
+
"step": 890
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.95,
|
| 1259 |
+
"grad_norm": 6.59375,
|
| 1260 |
+
"learning_rate": 3.882337980972164e-05,
|
| 1261 |
+
"loss": 1.2558,
|
| 1262 |
+
"step": 895
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.96,
|
| 1266 |
+
"grad_norm": 4.84375,
|
| 1267 |
+
"learning_rate": 3.870592523919002e-05,
|
| 1268 |
+
"loss": 1.1875,
|
| 1269 |
+
"step": 900
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.97,
|
| 1273 |
+
"grad_norm": 6.375,
|
| 1274 |
+
"learning_rate": 3.8588036531166124e-05,
|
| 1275 |
+
"loss": 1.2176,
|
| 1276 |
+
"step": 905
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.97,
|
| 1280 |
+
"grad_norm": 7.21875,
|
| 1281 |
+
"learning_rate": 3.846971741979465e-05,
|
| 1282 |
+
"loss": 1.2816,
|
| 1283 |
+
"step": 910
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.98,
|
| 1287 |
+
"grad_norm": 7.34375,
|
| 1288 |
+
"learning_rate": 3.835097165285341e-05,
|
| 1289 |
+
"loss": 1.2478,
|
| 1290 |
+
"step": 915
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.98,
|
| 1294 |
+
"grad_norm": 5.75,
|
| 1295 |
+
"learning_rate": 3.823180299163458e-05,
|
| 1296 |
+
"loss": 1.2156,
|
| 1297 |
+
"step": 920
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.99,
|
| 1301 |
+
"grad_norm": 5.9375,
|
| 1302 |
+
"learning_rate": 3.811221521082558e-05,
|
| 1303 |
+
"loss": 1.3427,
|
| 1304 |
+
"step": 925
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.99,
|
| 1308 |
+
"grad_norm": 5.3125,
|
| 1309 |
+
"learning_rate": 3.7992212098389536e-05,
|
| 1310 |
+
"loss": 1.2286,
|
| 1311 |
+
"step": 930
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.0,
|
| 1315 |
+
"grad_norm": 5.4375,
|
| 1316 |
+
"learning_rate": 3.7871797455445256e-05,
|
| 1317 |
+
"loss": 1.2868,
|
| 1318 |
+
"step": 935
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.0,
|
| 1322 |
+
"grad_norm": 5.0625,
|
| 1323 |
+
"learning_rate": 3.775097509614687e-05,
|
| 1324 |
+
"loss": 1.2123,
|
| 1325 |
+
"step": 940
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.01,
|
| 1329 |
+
"grad_norm": 5.5625,
|
| 1330 |
+
"learning_rate": 3.762974884756296e-05,
|
| 1331 |
+
"loss": 1.114,
|
| 1332 |
+
"step": 945
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.01,
|
| 1336 |
+
"grad_norm": 4.5,
|
| 1337 |
+
"learning_rate": 3.7508122549555384e-05,
|
| 1338 |
+
"loss": 1.1734,
|
| 1339 |
+
"step": 950
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.02,
|
| 1343 |
+
"grad_norm": 5.21875,
|
| 1344 |
+
"learning_rate": 3.738610005465764e-05,
|
| 1345 |
+
"loss": 1.1267,
|
| 1346 |
+
"step": 955
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.02,
|
| 1350 |
+
"grad_norm": 4.9375,
|
| 1351 |
+
"learning_rate": 3.726368522795282e-05,
|
| 1352 |
+
"loss": 1.2459,
|
| 1353 |
+
"step": 960
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.03,
|
| 1357 |
+
"grad_norm": 5.28125,
|
| 1358 |
+
"learning_rate": 3.714088194695119e-05,
|
| 1359 |
+
"loss": 1.1431,
|
| 1360 |
+
"step": 965
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.03,
|
| 1364 |
+
"grad_norm": 9.1875,
|
| 1365 |
+
"learning_rate": 3.7017694101467384e-05,
|
| 1366 |
+
"loss": 1.2636,
|
| 1367 |
+
"step": 970
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.04,
|
| 1371 |
+
"grad_norm": 5.96875,
|
| 1372 |
+
"learning_rate": 3.6894125593497167e-05,
|
| 1373 |
+
"loss": 1.1851,
|
| 1374 |
+
"step": 975
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.05,
|
| 1378 |
+
"grad_norm": 5.75,
|
| 1379 |
+
"learning_rate": 3.677018033709386e-05,
|
| 1380 |
+
"loss": 1.134,
|
| 1381 |
+
"step": 980
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.05,
|
| 1385 |
+
"grad_norm": 5.1875,
|
| 1386 |
+
"learning_rate": 3.6645862258244344e-05,
|
| 1387 |
+
"loss": 1.1106,
|
| 1388 |
+
"step": 985
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.06,
|
| 1392 |
+
"grad_norm": 5.71875,
|
| 1393 |
+
"learning_rate": 3.6521175294744735e-05,
|
| 1394 |
+
"loss": 1.1718,
|
| 1395 |
+
"step": 990
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.06,
|
| 1399 |
+
"grad_norm": 4.8125,
|
| 1400 |
+
"learning_rate": 3.639612339607561e-05,
|
| 1401 |
+
"loss": 1.1642,
|
| 1402 |
+
"step": 995
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.07,
|
| 1406 |
+
"grad_norm": 5.6875,
|
| 1407 |
+
"learning_rate": 3.627071052327695e-05,
|
| 1408 |
+
"loss": 1.1399,
|
| 1409 |
+
"step": 1000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.07,
|
| 1413 |
+
"grad_norm": 6.4375,
|
| 1414 |
+
"learning_rate": 3.614494064882263e-05,
|
| 1415 |
+
"loss": 1.1741,
|
| 1416 |
+
"step": 1005
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.08,
|
| 1420 |
+
"grad_norm": 5.03125,
|
| 1421 |
+
"learning_rate": 3.6018817756494634e-05,
|
| 1422 |
+
"loss": 1.1551,
|
| 1423 |
+
"step": 1010
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.08,
|
| 1427 |
+
"grad_norm": 5.5,
|
| 1428 |
+
"learning_rate": 3.5892345841256825e-05,
|
| 1429 |
+
"loss": 1.1106,
|
| 1430 |
+
"step": 1015
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.09,
|
| 1434 |
+
"grad_norm": 7.75,
|
| 1435 |
+
"learning_rate": 3.576552890912844e-05,
|
| 1436 |
+
"loss": 1.1462,
|
| 1437 |
+
"step": 1020
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.09,
|
| 1441 |
+
"grad_norm": 4.90625,
|
| 1442 |
+
"learning_rate": 3.5638370977057186e-05,
|
| 1443 |
+
"loss": 1.1489,
|
| 1444 |
+
"step": 1025
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.1,
|
| 1448 |
+
"grad_norm": 5.125,
|
| 1449 |
+
"learning_rate": 3.5510876072791986e-05,
|
| 1450 |
+
"loss": 1.1427,
|
| 1451 |
+
"step": 1030
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.1,
|
| 1455 |
+
"grad_norm": 5.125,
|
| 1456 |
+
"learning_rate": 3.538304823475543e-05,
|
| 1457 |
+
"loss": 1.1172,
|
| 1458 |
+
"step": 1035
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.11,
|
| 1462 |
+
"grad_norm": 6.28125,
|
| 1463 |
+
"learning_rate": 3.5254891511915836e-05,
|
| 1464 |
+
"loss": 1.2579,
|
| 1465 |
+
"step": 1040
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.11,
|
| 1469 |
+
"grad_norm": 4.84375,
|
| 1470 |
+
"learning_rate": 3.5126409963659004e-05,
|
| 1471 |
+
"loss": 1.1606,
|
| 1472 |
+
"step": 1045
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.12,
|
| 1476 |
+
"grad_norm": 4.6875,
|
| 1477 |
+
"learning_rate": 3.4997607659659614e-05,
|
| 1478 |
+
"loss": 1.2115,
|
| 1479 |
+
"step": 1050
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.13,
|
| 1483 |
+
"grad_norm": 5.15625,
|
| 1484 |
+
"learning_rate": 3.4868488679752384e-05,
|
| 1485 |
+
"loss": 1.1618,
|
| 1486 |
+
"step": 1055
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.13,
|
| 1490 |
+
"grad_norm": 4.875,
|
| 1491 |
+
"learning_rate": 3.4739057113802756e-05,
|
| 1492 |
+
"loss": 1.1754,
|
| 1493 |
+
"step": 1060
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.14,
|
| 1497 |
+
"grad_norm": 6.125,
|
| 1498 |
+
"learning_rate": 3.46093170615774e-05,
|
| 1499 |
+
"loss": 1.2046,
|
| 1500 |
+
"step": 1065
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.14,
|
| 1504 |
+
"grad_norm": 6.03125,
|
| 1505 |
+
"learning_rate": 3.4479272632614326e-05,
|
| 1506 |
+
"loss": 1.2066,
|
| 1507 |
+
"step": 1070
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.15,
|
| 1511 |
+
"grad_norm": 6.90625,
|
| 1512 |
+
"learning_rate": 3.4348927946092766e-05,
|
| 1513 |
+
"loss": 1.1538,
|
| 1514 |
+
"step": 1075
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.15,
|
| 1518 |
+
"grad_norm": 5.5625,
|
| 1519 |
+
"learning_rate": 3.421828713070261e-05,
|
| 1520 |
+
"loss": 1.1924,
|
| 1521 |
+
"step": 1080
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.16,
|
| 1525 |
+
"grad_norm": 5.78125,
|
| 1526 |
+
"learning_rate": 3.4087354324513734e-05,
|
| 1527 |
+
"loss": 1.1463,
|
| 1528 |
+
"step": 1085
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.16,
|
| 1532 |
+
"grad_norm": 8.75,
|
| 1533 |
+
"learning_rate": 3.3956133674844826e-05,
|
| 1534 |
+
"loss": 1.2498,
|
| 1535 |
+
"step": 1090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.17,
|
| 1539 |
+
"grad_norm": 5.90625,
|
| 1540 |
+
"learning_rate": 3.38246293381321e-05,
|
| 1541 |
+
"loss": 1.1543,
|
| 1542 |
+
"step": 1095
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.17,
|
| 1546 |
+
"grad_norm": 5.3125,
|
| 1547 |
+
"learning_rate": 3.3692845479797564e-05,
|
| 1548 |
+
"loss": 1.1401,
|
| 1549 |
+
"step": 1100
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.18,
|
| 1553 |
+
"grad_norm": 5.53125,
|
| 1554 |
+
"learning_rate": 3.356078627411716e-05,
|
| 1555 |
+
"loss": 1.1909,
|
| 1556 |
+
"step": 1105
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.18,
|
| 1560 |
+
"grad_norm": 5.9375,
|
| 1561 |
+
"learning_rate": 3.342845590408849e-05,
|
| 1562 |
+
"loss": 1.1075,
|
| 1563 |
+
"step": 1110
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.19,
|
| 1567 |
+
"grad_norm": 5.15625,
|
| 1568 |
+
"learning_rate": 3.329585856129833e-05,
|
| 1569 |
+
"loss": 1.2045,
|
| 1570 |
+
"step": 1115
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.19,
|
| 1574 |
+
"grad_norm": 5.21875,
|
| 1575 |
+
"learning_rate": 3.316299844578985e-05,
|
| 1576 |
+
"loss": 1.1752,
|
| 1577 |
+
"step": 1120
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.2,
|
| 1581 |
+
"grad_norm": 4.5625,
|
| 1582 |
+
"learning_rate": 3.302987976592963e-05,
|
| 1583 |
+
"loss": 1.1077,
|
| 1584 |
+
"step": 1125
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.21,
|
| 1588 |
+
"grad_norm": 5.15625,
|
| 1589 |
+
"learning_rate": 3.289650673827427e-05,
|
| 1590 |
+
"loss": 1.1641,
|
| 1591 |
+
"step": 1130
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.21,
|
| 1595 |
+
"grad_norm": 4.96875,
|
| 1596 |
+
"learning_rate": 3.276288358743691e-05,
|
| 1597 |
+
"loss": 1.1683,
|
| 1598 |
+
"step": 1135
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.22,
|
| 1602 |
+
"grad_norm": 6.6875,
|
| 1603 |
+
"learning_rate": 3.262901454595338e-05,
|
| 1604 |
+
"loss": 1.1507,
|
| 1605 |
+
"step": 1140
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.22,
|
| 1609 |
+
"grad_norm": 6.1875,
|
| 1610 |
+
"learning_rate": 3.2494903854148134e-05,
|
| 1611 |
+
"loss": 1.1706,
|
| 1612 |
+
"step": 1145
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.23,
|
| 1616 |
+
"grad_norm": 5.5,
|
| 1617 |
+
"learning_rate": 3.236055575999994e-05,
|
| 1618 |
+
"loss": 1.1484,
|
| 1619 |
+
"step": 1150
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.23,
|
| 1623 |
+
"grad_norm": 5.03125,
|
| 1624 |
+
"learning_rate": 3.222597451900733e-05,
|
| 1625 |
+
"loss": 1.2247,
|
| 1626 |
+
"step": 1155
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.24,
|
| 1630 |
+
"grad_norm": 5.84375,
|
| 1631 |
+
"learning_rate": 3.2091164394053785e-05,
|
| 1632 |
+
"loss": 1.1492,
|
| 1633 |
+
"step": 1160
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.24,
|
| 1637 |
+
"grad_norm": 5.375,
|
| 1638 |
+
"learning_rate": 3.1956129655272755e-05,
|
| 1639 |
+
"loss": 1.1822,
|
| 1640 |
+
"step": 1165
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.25,
|
| 1644 |
+
"grad_norm": 6.59375,
|
| 1645 |
+
"learning_rate": 3.182087457991233e-05,
|
| 1646 |
+
"loss": 1.2857,
|
| 1647 |
+
"step": 1170
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.25,
|
| 1651 |
+
"grad_norm": 7.75,
|
| 1652 |
+
"learning_rate": 3.168540345219985e-05,
|
| 1653 |
+
"loss": 1.1828,
|
| 1654 |
+
"step": 1175
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.26,
|
| 1658 |
+
"grad_norm": 7.75,
|
| 1659 |
+
"learning_rate": 3.15497205632061e-05,
|
| 1660 |
+
"loss": 1.2212,
|
| 1661 |
+
"step": 1180
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.26,
|
| 1665 |
+
"grad_norm": 6.03125,
|
| 1666 |
+
"learning_rate": 3.141383021070947e-05,
|
| 1667 |
+
"loss": 1.1151,
|
| 1668 |
+
"step": 1185
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.27,
|
| 1672 |
+
"grad_norm": 4.71875,
|
| 1673 |
+
"learning_rate": 3.127773669905978e-05,
|
| 1674 |
+
"loss": 1.226,
|
| 1675 |
+
"step": 1190
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.27,
|
| 1679 |
+
"grad_norm": 6.09375,
|
| 1680 |
+
"learning_rate": 3.114144433904194e-05,
|
| 1681 |
+
"loss": 1.2121,
|
| 1682 |
+
"step": 1195
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.28,
|
| 1686 |
+
"grad_norm": 5.8125,
|
| 1687 |
+
"learning_rate": 3.100495744773944e-05,
|
| 1688 |
+
"loss": 1.1956,
|
| 1689 |
+
"step": 1200
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.29,
|
| 1693 |
+
"grad_norm": 6.90625,
|
| 1694 |
+
"learning_rate": 3.086828034839756e-05,
|
| 1695 |
+
"loss": 1.1705,
|
| 1696 |
+
"step": 1205
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.29,
|
| 1700 |
+
"grad_norm": 5.8125,
|
| 1701 |
+
"learning_rate": 3.073141737028648e-05,
|
| 1702 |
+
"loss": 1.1738,
|
| 1703 |
+
"step": 1210
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.3,
|
| 1707 |
+
"grad_norm": 6.28125,
|
| 1708 |
+
"learning_rate": 3.059437284856409e-05,
|
| 1709 |
+
"loss": 1.2063,
|
| 1710 |
+
"step": 1215
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.3,
|
| 1714 |
+
"grad_norm": 5.4375,
|
| 1715 |
+
"learning_rate": 3.04571511241387e-05,
|
| 1716 |
+
"loss": 1.1492,
|
| 1717 |
+
"step": 1220
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.31,
|
| 1721 |
+
"grad_norm": 5.78125,
|
| 1722 |
+
"learning_rate": 3.031975654353158e-05,
|
| 1723 |
+
"loss": 1.2052,
|
| 1724 |
+
"step": 1225
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.31,
|
| 1728 |
+
"grad_norm": 5.0,
|
| 1729 |
+
"learning_rate": 3.018219345873923e-05,
|
| 1730 |
+
"loss": 1.1688,
|
| 1731 |
+
"step": 1230
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.32,
|
| 1735 |
+
"grad_norm": 6.21875,
|
| 1736 |
+
"learning_rate": 3.004446622709554e-05,
|
| 1737 |
+
"loss": 1.0857,
|
| 1738 |
+
"step": 1235
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.32,
|
| 1742 |
+
"grad_norm": 6.40625,
|
| 1743 |
+
"learning_rate": 2.99065792111338e-05,
|
| 1744 |
+
"loss": 1.2369,
|
| 1745 |
+
"step": 1240
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.33,
|
| 1749 |
+
"grad_norm": 6.46875,
|
| 1750 |
+
"learning_rate": 2.976853677844847e-05,
|
| 1751 |
+
"loss": 1.1783,
|
| 1752 |
+
"step": 1245
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.33,
|
| 1756 |
+
"grad_norm": 5.65625,
|
| 1757 |
+
"learning_rate": 2.9630343301556883e-05,
|
| 1758 |
+
"loss": 1.1596,
|
| 1759 |
+
"step": 1250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.34,
|
| 1763 |
+
"grad_norm": 5.90625,
|
| 1764 |
+
"learning_rate": 2.9492003157760713e-05,
|
| 1765 |
+
"loss": 1.2145,
|
| 1766 |
+
"step": 1255
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.34,
|
| 1770 |
+
"grad_norm": 5.625,
|
| 1771 |
+
"learning_rate": 2.9353520729007326e-05,
|
| 1772 |
+
"loss": 1.157,
|
| 1773 |
+
"step": 1260
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.35,
|
| 1777 |
+
"grad_norm": 5.28125,
|
| 1778 |
+
"learning_rate": 2.921490040175101e-05,
|
| 1779 |
+
"loss": 1.2069,
|
| 1780 |
+
"step": 1265
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.35,
|
| 1784 |
+
"grad_norm": 5.90625,
|
| 1785 |
+
"learning_rate": 2.9076146566813994e-05,
|
| 1786 |
+
"loss": 1.1791,
|
| 1787 |
+
"step": 1270
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.36,
|
| 1791 |
+
"grad_norm": 5.21875,
|
| 1792 |
+
"learning_rate": 2.893726361924738e-05,
|
| 1793 |
+
"loss": 1.1243,
|
| 1794 |
+
"step": 1275
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.37,
|
| 1798 |
+
"grad_norm": 6.53125,
|
| 1799 |
+
"learning_rate": 2.8798255958191946e-05,
|
| 1800 |
+
"loss": 1.191,
|
| 1801 |
+
"step": 1280
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.37,
|
| 1805 |
+
"grad_norm": 5.625,
|
| 1806 |
+
"learning_rate": 2.8659127986738782e-05,
|
| 1807 |
+
"loss": 1.1712,
|
| 1808 |
+
"step": 1285
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.38,
|
| 1812 |
+
"grad_norm": 4.75,
|
| 1813 |
+
"learning_rate": 2.851988411178985e-05,
|
| 1814 |
+
"loss": 1.1619,
|
| 1815 |
+
"step": 1290
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.38,
|
| 1819 |
+
"grad_norm": 5.71875,
|
| 1820 |
+
"learning_rate": 2.8380528743918348e-05,
|
| 1821 |
+
"loss": 1.1699,
|
| 1822 |
+
"step": 1295
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.39,
|
| 1826 |
+
"grad_norm": 6.375,
|
| 1827 |
+
"learning_rate": 2.824106629722906e-05,
|
| 1828 |
+
"loss": 1.2141,
|
| 1829 |
+
"step": 1300
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.39,
|
| 1833 |
+
"grad_norm": 6.1875,
|
| 1834 |
+
"learning_rate": 2.810150118921847e-05,
|
| 1835 |
+
"loss": 1.1444,
|
| 1836 |
+
"step": 1305
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.4,
|
| 1840 |
+
"grad_norm": 5.875,
|
| 1841 |
+
"learning_rate": 2.7961837840634913e-05,
|
| 1842 |
+
"loss": 1.1494,
|
| 1843 |
+
"step": 1310
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.4,
|
| 1847 |
+
"grad_norm": 6.3125,
|
| 1848 |
+
"learning_rate": 2.7822080675338492e-05,
|
| 1849 |
+
"loss": 1.1783,
|
| 1850 |
+
"step": 1315
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.41,
|
| 1854 |
+
"grad_norm": 5.25,
|
| 1855 |
+
"learning_rate": 2.768223412016099e-05,
|
| 1856 |
+
"loss": 1.1855,
|
| 1857 |
+
"step": 1320
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.41,
|
| 1861 |
+
"grad_norm": 5.96875,
|
| 1862 |
+
"learning_rate": 2.7542302604765598e-05,
|
| 1863 |
+
"loss": 1.1509,
|
| 1864 |
+
"step": 1325
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.42,
|
| 1868 |
+
"grad_norm": 5.03125,
|
| 1869 |
+
"learning_rate": 2.740229056150666e-05,
|
| 1870 |
+
"loss": 1.1274,
|
| 1871 |
+
"step": 1330
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.42,
|
| 1875 |
+
"grad_norm": 4.53125,
|
| 1876 |
+
"learning_rate": 2.7262202425289253e-05,
|
| 1877 |
+
"loss": 1.1581,
|
| 1878 |
+
"step": 1335
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.43,
|
| 1882 |
+
"grad_norm": 5.28125,
|
| 1883 |
+
"learning_rate": 2.7122042633428697e-05,
|
| 1884 |
+
"loss": 1.0967,
|
| 1885 |
+
"step": 1340
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.43,
|
| 1889 |
+
"grad_norm": 5.71875,
|
| 1890 |
+
"learning_rate": 2.698181562551002e-05,
|
| 1891 |
+
"loss": 1.181,
|
| 1892 |
+
"step": 1345
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.44,
|
| 1896 |
+
"grad_norm": 5.625,
|
| 1897 |
+
"learning_rate": 2.684152584324735e-05,
|
| 1898 |
+
"loss": 1.151,
|
| 1899 |
+
"step": 1350
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.45,
|
| 1903 |
+
"grad_norm": 7.09375,
|
| 1904 |
+
"learning_rate": 2.6701177730343186e-05,
|
| 1905 |
+
"loss": 1.1978,
|
| 1906 |
+
"step": 1355
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.45,
|
| 1910 |
+
"grad_norm": 5.53125,
|
| 1911 |
+
"learning_rate": 2.6560775732347654e-05,
|
| 1912 |
+
"loss": 1.1315,
|
| 1913 |
+
"step": 1360
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.46,
|
| 1917 |
+
"grad_norm": 5.25,
|
| 1918 |
+
"learning_rate": 2.6420324296517708e-05,
|
| 1919 |
+
"loss": 1.1845,
|
| 1920 |
+
"step": 1365
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.46,
|
| 1924 |
+
"grad_norm": 5.53125,
|
| 1925 |
+
"learning_rate": 2.627982787167625e-05,
|
| 1926 |
+
"loss": 1.1386,
|
| 1927 |
+
"step": 1370
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.47,
|
| 1931 |
+
"grad_norm": 4.96875,
|
| 1932 |
+
"learning_rate": 2.613929090807123e-05,
|
| 1933 |
+
"loss": 1.159,
|
| 1934 |
+
"step": 1375
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.47,
|
| 1938 |
+
"grad_norm": 5.3125,
|
| 1939 |
+
"learning_rate": 2.599871785723465e-05,
|
| 1940 |
+
"loss": 1.1653,
|
| 1941 |
+
"step": 1380
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.48,
|
| 1945 |
+
"grad_norm": 5.53125,
|
| 1946 |
+
"learning_rate": 2.5858113171841604e-05,
|
| 1947 |
+
"loss": 1.2584,
|
| 1948 |
+
"step": 1385
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.48,
|
| 1952 |
+
"grad_norm": 5.75,
|
| 1953 |
+
"learning_rate": 2.5717481305569186e-05,
|
| 1954 |
+
"loss": 1.2103,
|
| 1955 |
+
"step": 1390
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.49,
|
| 1959 |
+
"grad_norm": 5.875,
|
| 1960 |
+
"learning_rate": 2.5576826712955472e-05,
|
| 1961 |
+
"loss": 1.2361,
|
| 1962 |
+
"step": 1395
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.49,
|
| 1966 |
+
"grad_norm": 5.15625,
|
| 1967 |
+
"learning_rate": 2.5436153849258398e-05,
|
| 1968 |
+
"loss": 1.1491,
|
| 1969 |
+
"step": 1400
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.5,
|
| 1973 |
+
"grad_norm": 5.5625,
|
| 1974 |
+
"learning_rate": 2.5295467170314624e-05,
|
| 1975 |
+
"loss": 1.1003,
|
| 1976 |
+
"step": 1405
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.5,
|
| 1980 |
+
"grad_norm": 5.6875,
|
| 1981 |
+
"learning_rate": 2.515477113239843e-05,
|
| 1982 |
+
"loss": 1.2162,
|
| 1983 |
+
"step": 1410
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.51,
|
| 1987 |
+
"grad_norm": 6.1875,
|
| 1988 |
+
"learning_rate": 2.5014070192080526e-05,
|
| 1989 |
+
"loss": 1.1498,
|
| 1990 |
+
"step": 1415
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.51,
|
| 1994 |
+
"grad_norm": 6.28125,
|
| 1995 |
+
"learning_rate": 2.4873368806086918e-05,
|
| 1996 |
+
"loss": 1.2132,
|
| 1997 |
+
"step": 1420
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.52,
|
| 2001 |
+
"grad_norm": 9.4375,
|
| 2002 |
+
"learning_rate": 2.4732671431157716e-05,
|
| 2003 |
+
"loss": 1.1903,
|
| 2004 |
+
"step": 1425
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.53,
|
| 2008 |
+
"grad_norm": 7.625,
|
| 2009 |
+
"learning_rate": 2.4591982523906e-05,
|
| 2010 |
+
"loss": 1.1472,
|
| 2011 |
+
"step": 1430
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.53,
|
| 2015 |
+
"grad_norm": 5.25,
|
| 2016 |
+
"learning_rate": 2.4451306540676615e-05,
|
| 2017 |
+
"loss": 1.1085,
|
| 2018 |
+
"step": 1435
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.54,
|
| 2022 |
+
"grad_norm": 4.84375,
|
| 2023 |
+
"learning_rate": 2.431064793740505e-05,
|
| 2024 |
+
"loss": 1.0636,
|
| 2025 |
+
"step": 1440
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.54,
|
| 2029 |
+
"grad_norm": 6.1875,
|
| 2030 |
+
"learning_rate": 2.4170011169476254e-05,
|
| 2031 |
+
"loss": 1.1439,
|
| 2032 |
+
"step": 1445
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.55,
|
| 2036 |
+
"grad_norm": 4.96875,
|
| 2037 |
+
"learning_rate": 2.4029400691583574e-05,
|
| 2038 |
+
"loss": 1.2272,
|
| 2039 |
+
"step": 1450
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.55,
|
| 2043 |
+
"grad_norm": 5.71875,
|
| 2044 |
+
"learning_rate": 2.3888820957587605e-05,
|
| 2045 |
+
"loss": 1.158,
|
| 2046 |
+
"step": 1455
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.56,
|
| 2050 |
+
"grad_norm": 5.90625,
|
| 2051 |
+
"learning_rate": 2.3748276420375095e-05,
|
| 2052 |
+
"loss": 1.1893,
|
| 2053 |
+
"step": 1460
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.56,
|
| 2057 |
+
"grad_norm": 5.6875,
|
| 2058 |
+
"learning_rate": 2.3607771531717982e-05,
|
| 2059 |
+
"loss": 1.1827,
|
| 2060 |
+
"step": 1465
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.57,
|
| 2064 |
+
"grad_norm": 5.09375,
|
| 2065 |
+
"learning_rate": 2.3467310742132266e-05,
|
| 2066 |
+
"loss": 1.2413,
|
| 2067 |
+
"step": 1470
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.57,
|
| 2071 |
+
"grad_norm": 5.125,
|
| 2072 |
+
"learning_rate": 2.332689850073715e-05,
|
| 2073 |
+
"loss": 1.1588,
|
| 2074 |
+
"step": 1475
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 1.58,
|
| 2078 |
+
"grad_norm": 7.46875,
|
| 2079 |
+
"learning_rate": 2.3186539255114058e-05,
|
| 2080 |
+
"loss": 1.2112,
|
| 2081 |
+
"step": 1480
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.58,
|
| 2085 |
+
"grad_norm": 5.53125,
|
| 2086 |
+
"learning_rate": 2.304623745116573e-05,
|
| 2087 |
+
"loss": 1.1712,
|
| 2088 |
+
"step": 1485
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.59,
|
| 2092 |
+
"grad_norm": 5.625,
|
| 2093 |
+
"learning_rate": 2.2905997532975484e-05,
|
| 2094 |
+
"loss": 1.1619,
|
| 2095 |
+
"step": 1490
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.59,
|
| 2099 |
+
"grad_norm": 5.15625,
|
| 2100 |
+
"learning_rate": 2.2765823942666343e-05,
|
| 2101 |
+
"loss": 1.1737,
|
| 2102 |
+
"step": 1495
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.6,
|
| 2106 |
+
"grad_norm": 6.03125,
|
| 2107 |
+
"learning_rate": 2.2625721120260416e-05,
|
| 2108 |
+
"loss": 1.1432,
|
| 2109 |
+
"step": 1500
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.61,
|
| 2113 |
+
"grad_norm": 6.0,
|
| 2114 |
+
"learning_rate": 2.2485693503538232e-05,
|
| 2115 |
+
"loss": 1.1612,
|
| 2116 |
+
"step": 1505
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.61,
|
| 2120 |
+
"grad_norm": 4.78125,
|
| 2121 |
+
"learning_rate": 2.2345745527898138e-05,
|
| 2122 |
+
"loss": 1.1659,
|
| 2123 |
+
"step": 1510
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.62,
|
| 2127 |
+
"grad_norm": 5.78125,
|
| 2128 |
+
"learning_rate": 2.2205881626215868e-05,
|
| 2129 |
+
"loss": 1.1377,
|
| 2130 |
+
"step": 1515
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.62,
|
| 2134 |
+
"grad_norm": 5.28125,
|
| 2135 |
+
"learning_rate": 2.2066106228704058e-05,
|
| 2136 |
+
"loss": 1.1139,
|
| 2137 |
+
"step": 1520
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.63,
|
| 2141 |
+
"grad_norm": 6.03125,
|
| 2142 |
+
"learning_rate": 2.1926423762772017e-05,
|
| 2143 |
+
"loss": 1.1575,
|
| 2144 |
+
"step": 1525
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.63,
|
| 2148 |
+
"grad_norm": 6.1875,
|
| 2149 |
+
"learning_rate": 2.1786838652885373e-05,
|
| 2150 |
+
"loss": 1.1132,
|
| 2151 |
+
"step": 1530
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.64,
|
| 2155 |
+
"grad_norm": 4.84375,
|
| 2156 |
+
"learning_rate": 2.1647355320426017e-05,
|
| 2157 |
+
"loss": 1.1038,
|
| 2158 |
+
"step": 1535
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.64,
|
| 2162 |
+
"grad_norm": 5.625,
|
| 2163 |
+
"learning_rate": 2.1507978183552034e-05,
|
| 2164 |
+
"loss": 1.064,
|
| 2165 |
+
"step": 1540
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.65,
|
| 2169 |
+
"grad_norm": 5.1875,
|
| 2170 |
+
"learning_rate": 2.1368711657057705e-05,
|
| 2171 |
+
"loss": 1.1183,
|
| 2172 |
+
"step": 1545
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.65,
|
| 2176 |
+
"grad_norm": 4.71875,
|
| 2177 |
+
"learning_rate": 2.122956015223375e-05,
|
| 2178 |
+
"loss": 1.1619,
|
| 2179 |
+
"step": 1550
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.66,
|
| 2183 |
+
"grad_norm": 5.0,
|
| 2184 |
+
"learning_rate": 2.109052807672752e-05,
|
| 2185 |
+
"loss": 1.2064,
|
| 2186 |
+
"step": 1555
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.66,
|
| 2190 |
+
"grad_norm": 5.9375,
|
| 2191 |
+
"learning_rate": 2.0951619834403462e-05,
|
| 2192 |
+
"loss": 1.1686,
|
| 2193 |
+
"step": 1560
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.67,
|
| 2197 |
+
"grad_norm": 5.375,
|
| 2198 |
+
"learning_rate": 2.081283982520356e-05,
|
| 2199 |
+
"loss": 1.1372,
|
| 2200 |
+
"step": 1565
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.67,
|
| 2204 |
+
"grad_norm": 5.375,
|
| 2205 |
+
"learning_rate": 2.0674192445007997e-05,
|
| 2206 |
+
"loss": 1.1645,
|
| 2207 |
+
"step": 1570
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.68,
|
| 2211 |
+
"grad_norm": 5.75,
|
| 2212 |
+
"learning_rate": 2.0535682085495912e-05,
|
| 2213 |
+
"loss": 1.1098,
|
| 2214 |
+
"step": 1575
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.69,
|
| 2218 |
+
"grad_norm": 6.40625,
|
| 2219 |
+
"learning_rate": 2.039731313400629e-05,
|
| 2220 |
+
"loss": 1.1762,
|
| 2221 |
+
"step": 1580
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.69,
|
| 2225 |
+
"grad_norm": 6.1875,
|
| 2226 |
+
"learning_rate": 2.025908997339899e-05,
|
| 2227 |
+
"loss": 1.1623,
|
| 2228 |
+
"step": 1585
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.7,
|
| 2232 |
+
"grad_norm": 4.65625,
|
| 2233 |
+
"learning_rate": 2.0121016981915934e-05,
|
| 2234 |
+
"loss": 1.169,
|
| 2235 |
+
"step": 1590
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.7,
|
| 2239 |
+
"grad_norm": 5.6875,
|
| 2240 |
+
"learning_rate": 1.9983098533042384e-05,
|
| 2241 |
+
"loss": 1.1352,
|
| 2242 |
+
"step": 1595
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.71,
|
| 2246 |
+
"grad_norm": 5.125,
|
| 2247 |
+
"learning_rate": 1.9845338995368483e-05,
|
| 2248 |
+
"loss": 1.1707,
|
| 2249 |
+
"step": 1600
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.71,
|
| 2253 |
+
"grad_norm": 5.78125,
|
| 2254 |
+
"learning_rate": 1.9707742732450775e-05,
|
| 2255 |
+
"loss": 1.1958,
|
| 2256 |
+
"step": 1605
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.72,
|
| 2260 |
+
"grad_norm": 5.28125,
|
| 2261 |
+
"learning_rate": 1.9570314102674108e-05,
|
| 2262 |
+
"loss": 1.1699,
|
| 2263 |
+
"step": 1610
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.72,
|
| 2267 |
+
"grad_norm": 5.8125,
|
| 2268 |
+
"learning_rate": 1.9433057459113485e-05,
|
| 2269 |
+
"loss": 1.1558,
|
| 2270 |
+
"step": 1615
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.73,
|
| 2274 |
+
"grad_norm": 5.78125,
|
| 2275 |
+
"learning_rate": 1.9295977149396225e-05,
|
| 2276 |
+
"loss": 1.1885,
|
| 2277 |
+
"step": 1620
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.73,
|
| 2281 |
+
"grad_norm": 5.875,
|
| 2282 |
+
"learning_rate": 1.9159077515564257e-05,
|
| 2283 |
+
"loss": 1.1141,
|
| 2284 |
+
"step": 1625
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.74,
|
| 2288 |
+
"grad_norm": 4.375,
|
| 2289 |
+
"learning_rate": 1.902236289393654e-05,
|
| 2290 |
+
"loss": 1.1866,
|
| 2291 |
+
"step": 1630
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 1.74,
|
| 2295 |
+
"grad_norm": 4.90625,
|
| 2296 |
+
"learning_rate": 1.888583761497178e-05,
|
| 2297 |
+
"loss": 1.1655,
|
| 2298 |
+
"step": 1635
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.75,
|
| 2302 |
+
"grad_norm": 5.25,
|
| 2303 |
+
"learning_rate": 1.874950600313118e-05,
|
| 2304 |
+
"loss": 1.167,
|
| 2305 |
+
"step": 1640
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 1.75,
|
| 2309 |
+
"grad_norm": 6.625,
|
| 2310 |
+
"learning_rate": 1.861337237674154e-05,
|
| 2311 |
+
"loss": 1.1927,
|
| 2312 |
+
"step": 1645
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.76,
|
| 2316 |
+
"grad_norm": 5.375,
|
| 2317 |
+
"learning_rate": 1.8477441047858425e-05,
|
| 2318 |
+
"loss": 1.1902,
|
| 2319 |
+
"step": 1650
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.77,
|
| 2323 |
+
"grad_norm": 5.65625,
|
| 2324 |
+
"learning_rate": 1.8341716322129575e-05,
|
| 2325 |
+
"loss": 1.1921,
|
| 2326 |
+
"step": 1655
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 1.77,
|
| 2330 |
+
"grad_norm": 5.25,
|
| 2331 |
+
"learning_rate": 1.820620249865858e-05,
|
| 2332 |
+
"loss": 1.1864,
|
| 2333 |
+
"step": 1660
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 1.78,
|
| 2337 |
+
"grad_norm": 6.875,
|
| 2338 |
+
"learning_rate": 1.8070903869868615e-05,
|
| 2339 |
+
"loss": 1.1866,
|
| 2340 |
+
"step": 1665
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 1.78,
|
| 2344 |
+
"grad_norm": 5.65625,
|
| 2345 |
+
"learning_rate": 1.7935824721366574e-05,
|
| 2346 |
+
"loss": 1.1595,
|
| 2347 |
+
"step": 1670
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 1.79,
|
| 2351 |
+
"grad_norm": 7.1875,
|
| 2352 |
+
"learning_rate": 1.7800969331807237e-05,
|
| 2353 |
+
"loss": 1.1422,
|
| 2354 |
+
"step": 1675
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.79,
|
| 2358 |
+
"grad_norm": 5.8125,
|
| 2359 |
+
"learning_rate": 1.766634197275781e-05,
|
| 2360 |
+
"loss": 1.1428,
|
| 2361 |
+
"step": 1680
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 1.8,
|
| 2365 |
+
"grad_norm": 5.9375,
|
| 2366 |
+
"learning_rate": 1.7531946908562596e-05,
|
| 2367 |
+
"loss": 1.1056,
|
| 2368 |
+
"step": 1685
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 1.8,
|
| 2372 |
+
"grad_norm": 7.1875,
|
| 2373 |
+
"learning_rate": 1.7397788396207883e-05,
|
| 2374 |
+
"loss": 1.2066,
|
| 2375 |
+
"step": 1690
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 1.81,
|
| 2379 |
+
"grad_norm": 5.1875,
|
| 2380 |
+
"learning_rate": 1.7263870685187188e-05,
|
| 2381 |
+
"loss": 1.1753,
|
| 2382 |
+
"step": 1695
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 1.81,
|
| 2386 |
+
"grad_norm": 6.15625,
|
| 2387 |
+
"learning_rate": 1.7130198017366556e-05,
|
| 2388 |
+
"loss": 1.1243,
|
| 2389 |
+
"step": 1700
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 1.82,
|
| 2393 |
+
"grad_norm": 5.09375,
|
| 2394 |
+
"learning_rate": 1.699677462685029e-05,
|
| 2395 |
+
"loss": 1.1959,
|
| 2396 |
+
"step": 1705
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.82,
|
| 2400 |
+
"grad_norm": 5.25,
|
| 2401 |
+
"learning_rate": 1.686360473984678e-05,
|
| 2402 |
+
"loss": 1.1313,
|
| 2403 |
+
"step": 1710
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.83,
|
| 2407 |
+
"grad_norm": 5.21875,
|
| 2408 |
+
"learning_rate": 1.6730692574534622e-05,
|
| 2409 |
+
"loss": 1.1578,
|
| 2410 |
+
"step": 1715
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 1.83,
|
| 2414 |
+
"grad_norm": 6.5625,
|
| 2415 |
+
"learning_rate": 1.659804234092908e-05,
|
| 2416 |
+
"loss": 1.1415,
|
| 2417 |
+
"step": 1720
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 1.84,
|
| 2421 |
+
"grad_norm": 6.1875,
|
| 2422 |
+
"learning_rate": 1.646565824074865e-05,
|
| 2423 |
+
"loss": 1.1565,
|
| 2424 |
+
"step": 1725
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 1.85,
|
| 2428 |
+
"grad_norm": 5.3125,
|
| 2429 |
+
"learning_rate": 1.6333544467282028e-05,
|
| 2430 |
+
"loss": 1.1677,
|
| 2431 |
+
"step": 1730
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 1.85,
|
| 2435 |
+
"grad_norm": 9.375,
|
| 2436 |
+
"learning_rate": 1.6201705205255244e-05,
|
| 2437 |
+
"loss": 1.1842,
|
| 2438 |
+
"step": 1735
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 1.86,
|
| 2442 |
+
"grad_norm": 7.8125,
|
| 2443 |
+
"learning_rate": 1.6070144630699165e-05,
|
| 2444 |
+
"loss": 1.2085,
|
| 2445 |
+
"step": 1740
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 1.86,
|
| 2449 |
+
"grad_norm": 6.0625,
|
| 2450 |
+
"learning_rate": 1.5938866910817152e-05,
|
| 2451 |
+
"loss": 1.156,
|
| 2452 |
+
"step": 1745
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 1.87,
|
| 2456 |
+
"grad_norm": 5.28125,
|
| 2457 |
+
"learning_rate": 1.5807876203853118e-05,
|
| 2458 |
+
"loss": 1.1278,
|
| 2459 |
+
"step": 1750
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 1.87,
|
| 2463 |
+
"grad_norm": 5.59375,
|
| 2464 |
+
"learning_rate": 1.5677176658959782e-05,
|
| 2465 |
+
"loss": 1.1558,
|
| 2466 |
+
"step": 1755
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 1.88,
|
| 2470 |
+
"grad_norm": 5.75,
|
| 2471 |
+
"learning_rate": 1.554677241606726e-05,
|
| 2472 |
+
"loss": 1.1674,
|
| 2473 |
+
"step": 1760
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 1.88,
|
| 2477 |
+
"grad_norm": 5.0625,
|
| 2478 |
+
"learning_rate": 1.5416667605751933e-05,
|
| 2479 |
+
"loss": 1.1816,
|
| 2480 |
+
"step": 1765
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 1.89,
|
| 2484 |
+
"grad_norm": 5.65625,
|
| 2485 |
+
"learning_rate": 1.5286866349105593e-05,
|
| 2486 |
+
"loss": 1.1665,
|
| 2487 |
+
"step": 1770
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 1.89,
|
| 2491 |
+
"grad_norm": 5.96875,
|
| 2492 |
+
"learning_rate": 1.5157372757604932e-05,
|
| 2493 |
+
"loss": 1.1067,
|
| 2494 |
+
"step": 1775
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 1.9,
|
| 2498 |
+
"grad_norm": 6.75,
|
| 2499 |
+
"learning_rate": 1.502819093298128e-05,
|
| 2500 |
+
"loss": 1.2465,
|
| 2501 |
+
"step": 1780
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 1.9,
|
| 2505 |
+
"grad_norm": 5.65625,
|
| 2506 |
+
"learning_rate": 1.4899324967090705e-05,
|
| 2507 |
+
"loss": 1.1388,
|
| 2508 |
+
"step": 1785
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 1.91,
|
| 2512 |
+
"grad_norm": 6.125,
|
| 2513 |
+
"learning_rate": 1.4770778941784401e-05,
|
| 2514 |
+
"loss": 1.1855,
|
| 2515 |
+
"step": 1790
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 1.91,
|
| 2519 |
+
"grad_norm": 5.03125,
|
| 2520 |
+
"learning_rate": 1.4642556928779403e-05,
|
| 2521 |
+
"loss": 1.1639,
|
| 2522 |
+
"step": 1795
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 1.92,
|
| 2526 |
+
"grad_norm": 5.625,
|
| 2527 |
+
"learning_rate": 1.4514662989529562e-05,
|
| 2528 |
+
"loss": 1.1454,
|
| 2529 |
+
"step": 1800
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 1.93,
|
| 2533 |
+
"grad_norm": 5.46875,
|
| 2534 |
+
"learning_rate": 1.4387101175096985e-05,
|
| 2535 |
+
"loss": 1.1658,
|
| 2536 |
+
"step": 1805
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 1.93,
|
| 2540 |
+
"grad_norm": 5.3125,
|
| 2541 |
+
"learning_rate": 1.4259875526023622e-05,
|
| 2542 |
+
"loss": 1.1015,
|
| 2543 |
+
"step": 1810
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 1.94,
|
| 2547 |
+
"grad_norm": 4.84375,
|
| 2548 |
+
"learning_rate": 1.4132990072203378e-05,
|
| 2549 |
+
"loss": 1.1869,
|
| 2550 |
+
"step": 1815
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 1.94,
|
| 2554 |
+
"grad_norm": 5.90625,
|
| 2555 |
+
"learning_rate": 1.400644883275437e-05,
|
| 2556 |
+
"loss": 1.1581,
|
| 2557 |
+
"step": 1820
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 1.95,
|
| 2561 |
+
"grad_norm": 5.4375,
|
| 2562 |
+
"learning_rate": 1.3880255815891696e-05,
|
| 2563 |
+
"loss": 1.2049,
|
| 2564 |
+
"step": 1825
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 1.95,
|
| 2568 |
+
"grad_norm": 5.3125,
|
| 2569 |
+
"learning_rate": 1.3754415018800448e-05,
|
| 2570 |
+
"loss": 1.1779,
|
| 2571 |
+
"step": 1830
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 1.96,
|
| 2575 |
+
"grad_norm": 5.03125,
|
| 2576 |
+
"learning_rate": 1.3628930427509068e-05,
|
| 2577 |
+
"loss": 1.1909,
|
| 2578 |
+
"step": 1835
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 1.96,
|
| 2582 |
+
"grad_norm": 4.625,
|
| 2583 |
+
"learning_rate": 1.3503806016763152e-05,
|
| 2584 |
+
"loss": 1.1462,
|
| 2585 |
+
"step": 1840
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 1.97,
|
| 2589 |
+
"grad_norm": 6.03125,
|
| 2590 |
+
"learning_rate": 1.3379045749899485e-05,
|
| 2591 |
+
"loss": 1.1609,
|
| 2592 |
+
"step": 1845
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 1.97,
|
| 2596 |
+
"grad_norm": 6.0625,
|
| 2597 |
+
"learning_rate": 1.325465357872056e-05,
|
| 2598 |
+
"loss": 1.1463,
|
| 2599 |
+
"step": 1850
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 1.98,
|
| 2603 |
+
"grad_norm": 5.6875,
|
| 2604 |
+
"learning_rate": 1.3130633443369377e-05,
|
| 2605 |
+
"loss": 1.1447,
|
| 2606 |
+
"step": 1855
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 1.98,
|
| 2610 |
+
"grad_norm": 5.78125,
|
| 2611 |
+
"learning_rate": 1.3006989272204611e-05,
|
| 2612 |
+
"loss": 1.1105,
|
| 2613 |
+
"step": 1860
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 1.99,
|
| 2617 |
+
"grad_norm": 8.3125,
|
| 2618 |
+
"learning_rate": 1.2883724981676243e-05,
|
| 2619 |
+
"loss": 1.2189,
|
| 2620 |
+
"step": 1865
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 1.99,
|
| 2624 |
+
"grad_norm": 5.25,
|
| 2625 |
+
"learning_rate": 1.2760844476201429e-05,
|
| 2626 |
+
"loss": 1.1292,
|
| 2627 |
+
"step": 1870
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 2.0,
|
| 2631 |
+
"grad_norm": 5.21875,
|
| 2632 |
+
"learning_rate": 1.2638351648040918e-05,
|
| 2633 |
+
"loss": 1.1375,
|
| 2634 |
+
"step": 1875
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 2.01,
|
| 2638 |
+
"grad_norm": 4.65625,
|
| 2639 |
+
"learning_rate": 1.2516250377175673e-05,
|
| 2640 |
+
"loss": 1.1576,
|
| 2641 |
+
"step": 1880
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 2.01,
|
| 2645 |
+
"grad_norm": 4.84375,
|
| 2646 |
+
"learning_rate": 1.2394544531184036e-05,
|
| 2647 |
+
"loss": 1.168,
|
| 2648 |
+
"step": 1885
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 2.02,
|
| 2652 |
+
"grad_norm": 6.3125,
|
| 2653 |
+
"learning_rate": 1.2273237965119202e-05,
|
| 2654 |
+
"loss": 1.1362,
|
| 2655 |
+
"step": 1890
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 2.02,
|
| 2659 |
+
"grad_norm": 6.0,
|
| 2660 |
+
"learning_rate": 1.2152334521387082e-05,
|
| 2661 |
+
"loss": 1.147,
|
| 2662 |
+
"step": 1895
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 2.03,
|
| 2666 |
+
"grad_norm": 7.28125,
|
| 2667 |
+
"learning_rate": 1.2031838029624657e-05,
|
| 2668 |
+
"loss": 1.0798,
|
| 2669 |
+
"step": 1900
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 2.03,
|
| 2673 |
+
"grad_norm": 4.375,
|
| 2674 |
+
"learning_rate": 1.1911752306578593e-05,
|
| 2675 |
+
"loss": 1.0446,
|
| 2676 |
+
"step": 1905
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 2.04,
|
| 2680 |
+
"grad_norm": 5.21875,
|
| 2681 |
+
"learning_rate": 1.1792081155984422e-05,
|
| 2682 |
+
"loss": 1.0947,
|
| 2683 |
+
"step": 1910
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 2.04,
|
| 2687 |
+
"grad_norm": 5.25,
|
| 2688 |
+
"learning_rate": 1.1672828368446018e-05,
|
| 2689 |
+
"loss": 1.1792,
|
| 2690 |
+
"step": 1915
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 2.05,
|
| 2694 |
+
"grad_norm": 5.96875,
|
| 2695 |
+
"learning_rate": 1.1553997721315509e-05,
|
| 2696 |
+
"loss": 1.1014,
|
| 2697 |
+
"step": 1920
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 2.05,
|
| 2701 |
+
"grad_norm": 5.53125,
|
| 2702 |
+
"learning_rate": 1.1435592978573695e-05,
|
| 2703 |
+
"loss": 1.1151,
|
| 2704 |
+
"step": 1925
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 2.06,
|
| 2708 |
+
"grad_norm": 5.71875,
|
| 2709 |
+
"learning_rate": 1.1317617890710738e-05,
|
| 2710 |
+
"loss": 1.1495,
|
| 2711 |
+
"step": 1930
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.06,
|
| 2715 |
+
"grad_norm": 5.125,
|
| 2716 |
+
"learning_rate": 1.1200076194607448e-05,
|
| 2717 |
+
"loss": 1.203,
|
| 2718 |
+
"step": 1935
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 2.07,
|
| 2722 |
+
"grad_norm": 6.84375,
|
| 2723 |
+
"learning_rate": 1.1082971613416865e-05,
|
| 2724 |
+
"loss": 1.1496,
|
| 2725 |
+
"step": 1940
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 2.07,
|
| 2729 |
+
"grad_norm": 5.5625,
|
| 2730 |
+
"learning_rate": 1.0966307856446322e-05,
|
| 2731 |
+
"loss": 1.098,
|
| 2732 |
+
"step": 1945
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 2.08,
|
| 2736 |
+
"grad_norm": 5.3125,
|
| 2737 |
+
"learning_rate": 1.0850088619040005e-05,
|
| 2738 |
+
"loss": 1.1397,
|
| 2739 |
+
"step": 1950
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 2.09,
|
| 2743 |
+
"grad_norm": 5.0,
|
| 2744 |
+
"learning_rate": 1.0734317582461833e-05,
|
| 2745 |
+
"loss": 1.2423,
|
| 2746 |
+
"step": 1955
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 2.09,
|
| 2750 |
+
"grad_norm": 4.875,
|
| 2751 |
+
"learning_rate": 1.061899841377892e-05,
|
| 2752 |
+
"loss": 1.1154,
|
| 2753 |
+
"step": 1960
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 2.1,
|
| 2757 |
+
"grad_norm": 4.78125,
|
| 2758 |
+
"learning_rate": 1.0504134765745351e-05,
|
| 2759 |
+
"loss": 1.2164,
|
| 2760 |
+
"step": 1965
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 2.1,
|
| 2764 |
+
"grad_norm": 5.0,
|
| 2765 |
+
"learning_rate": 1.0389730276686557e-05,
|
| 2766 |
+
"loss": 1.0513,
|
| 2767 |
+
"step": 1970
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 2.11,
|
| 2771 |
+
"grad_norm": 4.9375,
|
| 2772 |
+
"learning_rate": 1.0275788570384016e-05,
|
| 2773 |
+
"loss": 1.0955,
|
| 2774 |
+
"step": 1975
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 2.11,
|
| 2778 |
+
"grad_norm": 5.59375,
|
| 2779 |
+
"learning_rate": 1.0162313255960465e-05,
|
| 2780 |
+
"loss": 1.075,
|
| 2781 |
+
"step": 1980
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 2.12,
|
| 2785 |
+
"grad_norm": 5.03125,
|
| 2786 |
+
"learning_rate": 1.0049307927765634e-05,
|
| 2787 |
+
"loss": 1.1331,
|
| 2788 |
+
"step": 1985
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 2.12,
|
| 2792 |
+
"grad_norm": 6.0,
|
| 2793 |
+
"learning_rate": 9.936776165262324e-06,
|
| 2794 |
+
"loss": 1.1564,
|
| 2795 |
+
"step": 1990
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 2.13,
|
| 2799 |
+
"grad_norm": 4.75,
|
| 2800 |
+
"learning_rate": 9.824721532913091e-06,
|
| 2801 |
+
"loss": 1.0781,
|
| 2802 |
+
"step": 1995
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 2.13,
|
| 2806 |
+
"grad_norm": 5.46875,
|
| 2807 |
+
"learning_rate": 9.71314758006731e-06,
|
| 2808 |
+
"loss": 1.1172,
|
| 2809 |
+
"step": 2000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 2.14,
|
| 2813 |
+
"grad_norm": 6.09375,
|
| 2814 |
+
"learning_rate": 9.602057840848732e-06,
|
| 2815 |
+
"loss": 1.1993,
|
| 2816 |
+
"step": 2005
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.14,
|
| 2820 |
+
"grad_norm": 4.75,
|
| 2821 |
+
"learning_rate": 9.491455834043583e-06,
|
| 2822 |
+
"loss": 1.0923,
|
| 2823 |
+
"step": 2010
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 2.15,
|
| 2827 |
+
"grad_norm": 5.75,
|
| 2828 |
+
"learning_rate": 9.38134506298906e-06,
|
| 2829 |
+
"loss": 1.1308,
|
| 2830 |
+
"step": 2015
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 2.15,
|
| 2834 |
+
"grad_norm": 4.84375,
|
| 2835 |
+
"learning_rate": 9.271729015462411e-06,
|
| 2836 |
+
"loss": 1.1036,
|
| 2837 |
+
"step": 2020
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 2.16,
|
| 2841 |
+
"grad_norm": 4.65625,
|
| 2842 |
+
"learning_rate": 9.162611163570403e-06,
|
| 2843 |
+
"loss": 1.1123,
|
| 2844 |
+
"step": 2025
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 2.17,
|
| 2848 |
+
"grad_norm": 6.21875,
|
| 2849 |
+
"learning_rate": 9.053994963639406e-06,
|
| 2850 |
+
"loss": 1.1922,
|
| 2851 |
+
"step": 2030
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 2.17,
|
| 2855 |
+
"grad_norm": 6.3125,
|
| 2856 |
+
"learning_rate": 8.945883856105841e-06,
|
| 2857 |
+
"loss": 1.2439,
|
| 2858 |
+
"step": 2035
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 2.18,
|
| 2862 |
+
"grad_norm": 5.9375,
|
| 2863 |
+
"learning_rate": 8.838281265407284e-06,
|
| 2864 |
+
"loss": 1.1455,
|
| 2865 |
+
"step": 2040
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 2.18,
|
| 2869 |
+
"grad_norm": 5.09375,
|
| 2870 |
+
"learning_rate": 8.731190599873912e-06,
|
| 2871 |
+
"loss": 1.1547,
|
| 2872 |
+
"step": 2045
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 2.19,
|
| 2876 |
+
"grad_norm": 5.40625,
|
| 2877 |
+
"learning_rate": 8.624615251620627e-06,
|
| 2878 |
+
"loss": 1.1167,
|
| 2879 |
+
"step": 2050
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 2.19,
|
| 2883 |
+
"grad_norm": 6.53125,
|
| 2884 |
+
"learning_rate": 8.518558596439532e-06,
|
| 2885 |
+
"loss": 1.1272,
|
| 2886 |
+
"step": 2055
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 2.2,
|
| 2890 |
+
"grad_norm": 6.4375,
|
| 2891 |
+
"learning_rate": 8.413023993693075e-06,
|
| 2892 |
+
"loss": 1.1031,
|
| 2893 |
+
"step": 2060
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 2.2,
|
| 2897 |
+
"grad_norm": 5.71875,
|
| 2898 |
+
"learning_rate": 8.308014786207571e-06,
|
| 2899 |
+
"loss": 1.1688,
|
| 2900 |
+
"step": 2065
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 2.21,
|
| 2904 |
+
"grad_norm": 4.84375,
|
| 2905 |
+
"learning_rate": 8.203534300167382e-06,
|
| 2906 |
+
"loss": 1.1888,
|
| 2907 |
+
"step": 2070
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 2.21,
|
| 2911 |
+
"grad_norm": 5.1875,
|
| 2912 |
+
"learning_rate": 8.0995858450095e-06,
|
| 2913 |
+
"loss": 1.1494,
|
| 2914 |
+
"step": 2075
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 2.22,
|
| 2918 |
+
"grad_norm": 4.8125,
|
| 2919 |
+
"learning_rate": 7.99617271331877e-06,
|
| 2920 |
+
"loss": 1.1709,
|
| 2921 |
+
"step": 2080
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 2.22,
|
| 2925 |
+
"grad_norm": 5.46875,
|
| 2926 |
+
"learning_rate": 7.893298180723582e-06,
|
| 2927 |
+
"loss": 1.1323,
|
| 2928 |
+
"step": 2085
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 2.23,
|
| 2932 |
+
"grad_norm": 5.96875,
|
| 2933 |
+
"learning_rate": 7.790965505792078e-06,
|
| 2934 |
+
"loss": 1.1689,
|
| 2935 |
+
"step": 2090
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 2.23,
|
| 2939 |
+
"grad_norm": 6.15625,
|
| 2940 |
+
"learning_rate": 7.689177929929004e-06,
|
| 2941 |
+
"loss": 1.1852,
|
| 2942 |
+
"step": 2095
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 2.24,
|
| 2946 |
+
"grad_norm": 5.34375,
|
| 2947 |
+
"learning_rate": 7.5879386772729625e-06,
|
| 2948 |
+
"loss": 1.1371,
|
| 2949 |
+
"step": 2100
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 2.25,
|
| 2953 |
+
"grad_norm": 4.5625,
|
| 2954 |
+
"learning_rate": 7.4872509545943715e-06,
|
| 2955 |
+
"loss": 1.0184,
|
| 2956 |
+
"step": 2105
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 2.25,
|
| 2960 |
+
"grad_norm": 4.78125,
|
| 2961 |
+
"learning_rate": 7.387117951193792e-06,
|
| 2962 |
+
"loss": 1.1552,
|
| 2963 |
+
"step": 2110
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 2.26,
|
| 2967 |
+
"grad_norm": 6.125,
|
| 2968 |
+
"learning_rate": 7.287542838801003e-06,
|
| 2969 |
+
"loss": 1.1553,
|
| 2970 |
+
"step": 2115
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 2.26,
|
| 2974 |
+
"grad_norm": 5.28125,
|
| 2975 |
+
"learning_rate": 7.188528771474478e-06,
|
| 2976 |
+
"loss": 1.1062,
|
| 2977 |
+
"step": 2120
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 2.27,
|
| 2981 |
+
"grad_norm": 5.125,
|
| 2982 |
+
"learning_rate": 7.090078885501475e-06,
|
| 2983 |
+
"loss": 1.0785,
|
| 2984 |
+
"step": 2125
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 2.27,
|
| 2988 |
+
"grad_norm": 4.90625,
|
| 2989 |
+
"learning_rate": 6.992196299298739e-06,
|
| 2990 |
+
"loss": 1.0447,
|
| 2991 |
+
"step": 2130
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 2.28,
|
| 2995 |
+
"grad_norm": 6.375,
|
| 2996 |
+
"learning_rate": 6.894884113313679e-06,
|
| 2997 |
+
"loss": 1.1757,
|
| 2998 |
+
"step": 2135
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 2.28,
|
| 3002 |
+
"grad_norm": 5.6875,
|
| 3003 |
+
"learning_rate": 6.798145409926191e-06,
|
| 3004 |
+
"loss": 1.1345,
|
| 3005 |
+
"step": 2140
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 2.29,
|
| 3009 |
+
"grad_norm": 5.46875,
|
| 3010 |
+
"learning_rate": 6.701983253351021e-06,
|
| 3011 |
+
"loss": 1.1477,
|
| 3012 |
+
"step": 2145
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 2.29,
|
| 3016 |
+
"grad_norm": 5.78125,
|
| 3017 |
+
"learning_rate": 6.606400689540673e-06,
|
| 3018 |
+
"loss": 1.1445,
|
| 3019 |
+
"step": 2150
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 2.3,
|
| 3023 |
+
"grad_norm": 5.0625,
|
| 3024 |
+
"learning_rate": 6.511400746088986e-06,
|
| 3025 |
+
"loss": 1.1672,
|
| 3026 |
+
"step": 2155
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 2.3,
|
| 3030 |
+
"grad_norm": 5.75,
|
| 3031 |
+
"learning_rate": 6.416986432135161e-06,
|
| 3032 |
+
"loss": 1.0833,
|
| 3033 |
+
"step": 2160
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 2.31,
|
| 3037 |
+
"grad_norm": 5.03125,
|
| 3038 |
+
"learning_rate": 6.323160738268524e-06,
|
| 3039 |
+
"loss": 1.0998,
|
| 3040 |
+
"step": 2165
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 2.31,
|
| 3044 |
+
"grad_norm": 4.53125,
|
| 3045 |
+
"learning_rate": 6.229926636433733e-06,
|
| 3046 |
+
"loss": 1.0866,
|
| 3047 |
+
"step": 2170
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 2.32,
|
| 3051 |
+
"grad_norm": 5.96875,
|
| 3052 |
+
"learning_rate": 6.137287079836681e-06,
|
| 3053 |
+
"loss": 1.142,
|
| 3054 |
+
"step": 2175
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 2.33,
|
| 3058 |
+
"grad_norm": 5.15625,
|
| 3059 |
+
"learning_rate": 6.045245002850952e-06,
|
| 3060 |
+
"loss": 1.0761,
|
| 3061 |
+
"step": 2180
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 2.33,
|
| 3065 |
+
"grad_norm": 6.15625,
|
| 3066 |
+
"learning_rate": 5.95380332092483e-06,
|
| 3067 |
+
"loss": 1.1538,
|
| 3068 |
+
"step": 2185
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 2.34,
|
| 3072 |
+
"grad_norm": 4.9375,
|
| 3073 |
+
"learning_rate": 5.8629649304890235e-06,
|
| 3074 |
+
"loss": 1.1283,
|
| 3075 |
+
"step": 2190
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 2.34,
|
| 3079 |
+
"grad_norm": 6.3125,
|
| 3080 |
+
"learning_rate": 5.772732708864842e-06,
|
| 3081 |
+
"loss": 1.2129,
|
| 3082 |
+
"step": 2195
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 2.35,
|
| 3086 |
+
"grad_norm": 5.625,
|
| 3087 |
+
"learning_rate": 5.683109514173124e-06,
|
| 3088 |
+
"loss": 1.1755,
|
| 3089 |
+
"step": 2200
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 2.35,
|
| 3093 |
+
"grad_norm": 5.125,
|
| 3094 |
+
"learning_rate": 5.594098185243676e-06,
|
| 3095 |
+
"loss": 1.072,
|
| 3096 |
+
"step": 2205
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 2.36,
|
| 3100 |
+
"grad_norm": 5.5625,
|
| 3101 |
+
"learning_rate": 5.505701541525335e-06,
|
| 3102 |
+
"loss": 1.1341,
|
| 3103 |
+
"step": 2210
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 2.36,
|
| 3107 |
+
"grad_norm": 5.09375,
|
| 3108 |
+
"learning_rate": 5.417922382996699e-06,
|
| 3109 |
+
"loss": 1.1623,
|
| 3110 |
+
"step": 2215
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 2.37,
|
| 3114 |
+
"grad_norm": 5.875,
|
| 3115 |
+
"learning_rate": 5.330763490077404e-06,
|
| 3116 |
+
"loss": 1.09,
|
| 3117 |
+
"step": 2220
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 2.37,
|
| 3121 |
+
"grad_norm": 7.875,
|
| 3122 |
+
"learning_rate": 5.244227623540074e-06,
|
| 3123 |
+
"loss": 1.1628,
|
| 3124 |
+
"step": 2225
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 2.38,
|
| 3128 |
+
"grad_norm": 5.59375,
|
| 3129 |
+
"learning_rate": 5.1583175244228785e-06,
|
| 3130 |
+
"loss": 1.1654,
|
| 3131 |
+
"step": 2230
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 2.38,
|
| 3135 |
+
"grad_norm": 5.5,
|
| 3136 |
+
"learning_rate": 5.073035913942678e-06,
|
| 3137 |
+
"loss": 1.117,
|
| 3138 |
+
"step": 2235
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 2.39,
|
| 3142 |
+
"grad_norm": 6.28125,
|
| 3143 |
+
"learning_rate": 4.988385493408871e-06,
|
| 3144 |
+
"loss": 1.1308,
|
| 3145 |
+
"step": 2240
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 2.39,
|
| 3149 |
+
"grad_norm": 5.0,
|
| 3150 |
+
"learning_rate": 4.904368944137783e-06,
|
| 3151 |
+
"loss": 1.1502,
|
| 3152 |
+
"step": 2245
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 2.4,
|
| 3156 |
+
"grad_norm": 5.84375,
|
| 3157 |
+
"learning_rate": 4.82098892736779e-06,
|
| 3158 |
+
"loss": 1.137,
|
| 3159 |
+
"step": 2250
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 2.41,
|
| 3163 |
+
"grad_norm": 4.96875,
|
| 3164 |
+
"learning_rate": 4.738248084174959e-06,
|
| 3165 |
+
"loss": 1.1274,
|
| 3166 |
+
"step": 2255
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 2.41,
|
| 3170 |
+
"grad_norm": 6.125,
|
| 3171 |
+
"learning_rate": 4.656149035389457e-06,
|
| 3172 |
+
"loss": 1.0886,
|
| 3173 |
+
"step": 2260
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 2.42,
|
| 3177 |
+
"grad_norm": 4.90625,
|
| 3178 |
+
"learning_rate": 4.574694381512498e-06,
|
| 3179 |
+
"loss": 1.2045,
|
| 3180 |
+
"step": 2265
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 2.42,
|
| 3184 |
+
"grad_norm": 5.875,
|
| 3185 |
+
"learning_rate": 4.493886702633962e-06,
|
| 3186 |
+
"loss": 1.1118,
|
| 3187 |
+
"step": 2270
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 2.43,
|
| 3191 |
+
"grad_norm": 5.21875,
|
| 3192 |
+
"learning_rate": 4.41372855835071e-06,
|
| 3193 |
+
"loss": 1.1354,
|
| 3194 |
+
"step": 2275
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 2.43,
|
| 3198 |
+
"grad_norm": 5.5,
|
| 3199 |
+
"learning_rate": 4.3342224876854604e-06,
|
| 3200 |
+
"loss": 1.1385,
|
| 3201 |
+
"step": 2280
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 2.44,
|
| 3205 |
+
"grad_norm": 5.3125,
|
| 3206 |
+
"learning_rate": 4.255371009006415e-06,
|
| 3207 |
+
"loss": 1.2304,
|
| 3208 |
+
"step": 2285
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 2.44,
|
| 3212 |
+
"grad_norm": 5.8125,
|
| 3213 |
+
"learning_rate": 4.177176619947451e-06,
|
| 3214 |
+
"loss": 1.1728,
|
| 3215 |
+
"step": 2290
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 2.45,
|
| 3219 |
+
"grad_norm": 5.625,
|
| 3220 |
+
"learning_rate": 4.099641797329015e-06,
|
| 3221 |
+
"loss": 1.14,
|
| 3222 |
+
"step": 2295
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 2.45,
|
| 3226 |
+
"grad_norm": 6.15625,
|
| 3227 |
+
"learning_rate": 4.0227689970796956e-06,
|
| 3228 |
+
"loss": 1.1576,
|
| 3229 |
+
"step": 2300
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 2.46,
|
| 3233 |
+
"grad_norm": 5.90625,
|
| 3234 |
+
"learning_rate": 3.946560654158385e-06,
|
| 3235 |
+
"loss": 1.1663,
|
| 3236 |
+
"step": 2305
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 2.46,
|
| 3240 |
+
"grad_norm": 7.5,
|
| 3241 |
+
"learning_rate": 3.871019182477203e-06,
|
| 3242 |
+
"loss": 1.2139,
|
| 3243 |
+
"step": 2310
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 2.47,
|
| 3247 |
+
"grad_norm": 5.09375,
|
| 3248 |
+
"learning_rate": 3.796146974824988e-06,
|
| 3249 |
+
"loss": 1.142,
|
| 3250 |
+
"step": 2315
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 2.47,
|
| 3254 |
+
"grad_norm": 4.8125,
|
| 3255 |
+
"learning_rate": 3.7219464027915403e-06,
|
| 3256 |
+
"loss": 1.1319,
|
| 3257 |
+
"step": 2320
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 2.48,
|
| 3261 |
+
"grad_norm": 5.03125,
|
| 3262 |
+
"learning_rate": 3.6484198166924973e-06,
|
| 3263 |
+
"loss": 1.1338,
|
| 3264 |
+
"step": 2325
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 2.49,
|
| 3268 |
+
"grad_norm": 5.375,
|
| 3269 |
+
"learning_rate": 3.5755695454948558e-06,
|
| 3270 |
+
"loss": 1.1259,
|
| 3271 |
+
"step": 2330
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 2.49,
|
| 3275 |
+
"grad_norm": 6.40625,
|
| 3276 |
+
"learning_rate": 3.5033978967432437e-06,
|
| 3277 |
+
"loss": 1.226,
|
| 3278 |
+
"step": 2335
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 2.5,
|
| 3282 |
+
"grad_norm": 4.96875,
|
| 3283 |
+
"learning_rate": 3.431907156486791e-06,
|
| 3284 |
+
"loss": 1.1347,
|
| 3285 |
+
"step": 2340
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 2.5,
|
| 3289 |
+
"grad_norm": 5.25,
|
| 3290 |
+
"learning_rate": 3.361099589206751e-06,
|
| 3291 |
+
"loss": 1.133,
|
| 3292 |
+
"step": 2345
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 2.51,
|
| 3296 |
+
"grad_norm": 4.5,
|
| 3297 |
+
"learning_rate": 3.2909774377447606e-06,
|
| 3298 |
+
"loss": 1.0546,
|
| 3299 |
+
"step": 2350
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 2.51,
|
| 3303 |
+
"grad_norm": 6.125,
|
| 3304 |
+
"learning_rate": 3.221542923231774e-06,
|
| 3305 |
+
"loss": 1.1597,
|
| 3306 |
+
"step": 2355
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 2.52,
|
| 3310 |
+
"grad_norm": 5.75,
|
| 3311 |
+
"learning_rate": 3.1527982450177547e-06,
|
| 3312 |
+
"loss": 1.1044,
|
| 3313 |
+
"step": 2360
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 2.52,
|
| 3317 |
+
"grad_norm": 6.1875,
|
| 3318 |
+
"learning_rate": 3.084745580601964e-06,
|
| 3319 |
+
"loss": 1.1435,
|
| 3320 |
+
"step": 2365
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 2.53,
|
| 3324 |
+
"grad_norm": 5.5,
|
| 3325 |
+
"learning_rate": 3.0173870855640246e-06,
|
| 3326 |
+
"loss": 1.2131,
|
| 3327 |
+
"step": 2370
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 2.53,
|
| 3331 |
+
"grad_norm": 6.25,
|
| 3332 |
+
"learning_rate": 2.9507248934956155e-06,
|
| 3333 |
+
"loss": 1.1718,
|
| 3334 |
+
"step": 2375
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 2.54,
|
| 3338 |
+
"grad_norm": 4.40625,
|
| 3339 |
+
"learning_rate": 2.8847611159329072e-06,
|
| 3340 |
+
"loss": 1.1415,
|
| 3341 |
+
"step": 2380
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 2.54,
|
| 3345 |
+
"grad_norm": 5.5625,
|
| 3346 |
+
"learning_rate": 2.819497842289684e-06,
|
| 3347 |
+
"loss": 1.1762,
|
| 3348 |
+
"step": 2385
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 2.55,
|
| 3352 |
+
"grad_norm": 5.375,
|
| 3353 |
+
"learning_rate": 2.754937139791125e-06,
|
| 3354 |
+
"loss": 1.1139,
|
| 3355 |
+
"step": 2390
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 2.55,
|
| 3359 |
+
"grad_norm": 5.78125,
|
| 3360 |
+
"learning_rate": 2.691081053408376e-06,
|
| 3361 |
+
"loss": 1.1348,
|
| 3362 |
+
"step": 2395
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 2.56,
|
| 3366 |
+
"grad_norm": 5.59375,
|
| 3367 |
+
"learning_rate": 2.6279316057937263e-06,
|
| 3368 |
+
"loss": 1.2093,
|
| 3369 |
+
"step": 2400
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 2.57,
|
| 3373 |
+
"grad_norm": 6.9375,
|
| 3374 |
+
"learning_rate": 2.5654907972165853e-06,
|
| 3375 |
+
"loss": 1.1454,
|
| 3376 |
+
"step": 2405
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 2.57,
|
| 3380 |
+
"grad_norm": 5.46875,
|
| 3381 |
+
"learning_rate": 2.50376060550008e-06,
|
| 3382 |
+
"loss": 1.1167,
|
| 3383 |
+
"step": 2410
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 2.58,
|
| 3387 |
+
"grad_norm": 5.59375,
|
| 3388 |
+
"learning_rate": 2.442742985958446e-06,
|
| 3389 |
+
"loss": 1.1659,
|
| 3390 |
+
"step": 2415
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 2.58,
|
| 3394 |
+
"grad_norm": 5.125,
|
| 3395 |
+
"learning_rate": 2.3824398713350597e-06,
|
| 3396 |
+
"loss": 1.1642,
|
| 3397 |
+
"step": 2420
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 2.59,
|
| 3401 |
+
"grad_norm": 6.0625,
|
| 3402 |
+
"learning_rate": 2.3228531717412515e-06,
|
| 3403 |
+
"loss": 1.1461,
|
| 3404 |
+
"step": 2425
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 2.59,
|
| 3408 |
+
"grad_norm": 5.15625,
|
| 3409 |
+
"learning_rate": 2.2639847745957647e-06,
|
| 3410 |
+
"loss": 1.1391,
|
| 3411 |
+
"step": 2430
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 2.6,
|
| 3415 |
+
"grad_norm": 5.6875,
|
| 3416 |
+
"learning_rate": 2.2058365445650126e-06,
|
| 3417 |
+
"loss": 1.1331,
|
| 3418 |
+
"step": 2435
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 2.6,
|
| 3422 |
+
"grad_norm": 5.8125,
|
| 3423 |
+
"learning_rate": 2.1484103235039714e-06,
|
| 3424 |
+
"loss": 1.108,
|
| 3425 |
+
"step": 2440
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 2.61,
|
| 3429 |
+
"grad_norm": 5.65625,
|
| 3430 |
+
"learning_rate": 2.091707930397885e-06,
|
| 3431 |
+
"loss": 1.142,
|
| 3432 |
+
"step": 2445
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 2.61,
|
| 3436 |
+
"grad_norm": 6.1875,
|
| 3437 |
+
"learning_rate": 2.035731161304602e-06,
|
| 3438 |
+
"loss": 1.0998,
|
| 3439 |
+
"step": 2450
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 2.62,
|
| 3443 |
+
"grad_norm": 4.40625,
|
| 3444 |
+
"learning_rate": 1.9804817892977277e-06,
|
| 3445 |
+
"loss": 1.1734,
|
| 3446 |
+
"step": 2455
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 2.62,
|
| 3450 |
+
"grad_norm": 6.90625,
|
| 3451 |
+
"learning_rate": 1.925961564410428e-06,
|
| 3452 |
+
"loss": 1.2022,
|
| 3453 |
+
"step": 2460
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 2.63,
|
| 3457 |
+
"grad_norm": 5.46875,
|
| 3458 |
+
"learning_rate": 1.8721722135800174e-06,
|
| 3459 |
+
"loss": 1.1813,
|
| 3460 |
+
"step": 2465
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 2.63,
|
| 3464 |
+
"grad_norm": 5.46875,
|
| 3465 |
+
"learning_rate": 1.819115440593258e-06,
|
| 3466 |
+
"loss": 1.1642,
|
| 3467 |
+
"step": 2470
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 2.64,
|
| 3471 |
+
"grad_norm": 5.71875,
|
| 3472 |
+
"learning_rate": 1.766792926032365e-06,
|
| 3473 |
+
"loss": 1.1271,
|
| 3474 |
+
"step": 2475
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 2.65,
|
| 3478 |
+
"grad_norm": 7.1875,
|
| 3479 |
+
"learning_rate": 1.715206327221816e-06,
|
| 3480 |
+
"loss": 1.1829,
|
| 3481 |
+
"step": 2480
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 2.65,
|
| 3485 |
+
"grad_norm": 5.53125,
|
| 3486 |
+
"learning_rate": 1.6643572781758181e-06,
|
| 3487 |
+
"loss": 1.1764,
|
| 3488 |
+
"step": 2485
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 2.66,
|
| 3492 |
+
"grad_norm": 7.03125,
|
| 3493 |
+
"learning_rate": 1.614247389546575e-06,
|
| 3494 |
+
"loss": 1.0975,
|
| 3495 |
+
"step": 2490
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 2.66,
|
| 3499 |
+
"grad_norm": 5.09375,
|
| 3500 |
+
"learning_rate": 1.5648782485732566e-06,
|
| 3501 |
+
"loss": 1.1742,
|
| 3502 |
+
"step": 2495
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 2.67,
|
| 3506 |
+
"grad_norm": 6.46875,
|
| 3507 |
+
"learning_rate": 1.5162514190317195e-06,
|
| 3508 |
+
"loss": 1.1686,
|
| 3509 |
+
"step": 2500
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 2.67,
|
| 3513 |
+
"grad_norm": 6.0,
|
| 3514 |
+
"learning_rate": 1.4683684411849947e-06,
|
| 3515 |
+
"loss": 1.1135,
|
| 3516 |
+
"step": 2505
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 2.68,
|
| 3520 |
+
"grad_norm": 4.96875,
|
| 3521 |
+
"learning_rate": 1.4212308317344686e-06,
|
| 3522 |
+
"loss": 1.1648,
|
| 3523 |
+
"step": 2510
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 2.68,
|
| 3527 |
+
"grad_norm": 5.96875,
|
| 3528 |
+
"learning_rate": 1.3748400837718696e-06,
|
| 3529 |
+
"loss": 1.1331,
|
| 3530 |
+
"step": 2515
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 2.69,
|
| 3534 |
+
"grad_norm": 5.0,
|
| 3535 |
+
"learning_rate": 1.32919766673196e-06,
|
| 3536 |
+
"loss": 1.1018,
|
| 3537 |
+
"step": 2520
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 2.69,
|
| 3541 |
+
"grad_norm": 6.125,
|
| 3542 |
+
"learning_rate": 1.2843050263459888e-06,
|
| 3543 |
+
"loss": 1.1353,
|
| 3544 |
+
"step": 2525
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 2.7,
|
| 3548 |
+
"grad_norm": 5.8125,
|
| 3549 |
+
"learning_rate": 1.2401635845959159e-06,
|
| 3550 |
+
"loss": 1.1662,
|
| 3551 |
+
"step": 2530
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 2.7,
|
| 3555 |
+
"grad_norm": 6.4375,
|
| 3556 |
+
"learning_rate": 1.1967747396693401e-06,
|
| 3557 |
+
"loss": 1.2042,
|
| 3558 |
+
"step": 2535
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 2.71,
|
| 3562 |
+
"grad_norm": 5.59375,
|
| 3563 |
+
"learning_rate": 1.1541398659152464e-06,
|
| 3564 |
+
"loss": 1.132,
|
| 3565 |
+
"step": 2540
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 2.71,
|
| 3569 |
+
"grad_norm": 6.5625,
|
| 3570 |
+
"learning_rate": 1.1122603138004422e-06,
|
| 3571 |
+
"loss": 1.1555,
|
| 3572 |
+
"step": 2545
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 2.72,
|
| 3576 |
+
"grad_norm": 5.8125,
|
| 3577 |
+
"learning_rate": 1.0711374098667987e-06,
|
| 3578 |
+
"loss": 1.1753,
|
| 3579 |
+
"step": 2550
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 2.73,
|
| 3583 |
+
"grad_norm": 5.46875,
|
| 3584 |
+
"learning_rate": 1.0307724566892385e-06,
|
| 3585 |
+
"loss": 1.1657,
|
| 3586 |
+
"step": 2555
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 2.73,
|
| 3590 |
+
"grad_norm": 5.0,
|
| 3591 |
+
"learning_rate": 9.911667328344477e-07,
|
| 3592 |
+
"loss": 1.1786,
|
| 3593 |
+
"step": 2560
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 2.74,
|
| 3597 |
+
"grad_norm": 5.03125,
|
| 3598 |
+
"learning_rate": 9.523214928204127e-07,
|
| 3599 |
+
"loss": 1.1759,
|
| 3600 |
+
"step": 2565
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 2.74,
|
| 3604 |
+
"grad_norm": 6.25,
|
| 3605 |
+
"learning_rate": 9.142379670766532e-07,
|
| 3606 |
+
"loss": 1.19,
|
| 3607 |
+
"step": 2570
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 2.75,
|
| 3611 |
+
"grad_norm": 6.0,
|
| 3612 |
+
"learning_rate": 8.76917361905269e-07,
|
| 3613 |
+
"loss": 1.1569,
|
| 3614 |
+
"step": 2575
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 2.75,
|
| 3618 |
+
"grad_norm": 7.28125,
|
| 3619 |
+
"learning_rate": 8.403608594427237e-07,
|
| 3620 |
+
"loss": 1.1427,
|
| 3621 |
+
"step": 2580
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 2.76,
|
| 3625 |
+
"grad_norm": 5.1875,
|
| 3626 |
+
"learning_rate": 8.045696176223888e-07,
|
| 3627 |
+
"loss": 1.1103,
|
| 3628 |
+
"step": 2585
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 2.76,
|
| 3632 |
+
"grad_norm": 5.75,
|
| 3633 |
+
"learning_rate": 7.695447701378866e-07,
|
| 3634 |
+
"loss": 1.1479,
|
| 3635 |
+
"step": 2590
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 2.77,
|
| 3639 |
+
"grad_norm": 5.0,
|
| 3640 |
+
"learning_rate": 7.352874264071579e-07,
|
| 3641 |
+
"loss": 1.1034,
|
| 3642 |
+
"step": 2595
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 2.77,
|
| 3646 |
+
"grad_norm": 5.84375,
|
| 3647 |
+
"learning_rate": 7.017986715373515e-07,
|
| 3648 |
+
"loss": 1.0887,
|
| 3649 |
+
"step": 2600
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 2.78,
|
| 3653 |
+
"grad_norm": 6.21875,
|
| 3654 |
+
"learning_rate": 6.69079566290412e-07,
|
| 3655 |
+
"loss": 1.2236,
|
| 3656 |
+
"step": 2605
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 2.78,
|
| 3660 |
+
"grad_norm": 6.46875,
|
| 3661 |
+
"learning_rate": 6.371311470495162e-07,
|
| 3662 |
+
"loss": 1.1793,
|
| 3663 |
+
"step": 2610
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 2.79,
|
| 3667 |
+
"grad_norm": 5.15625,
|
| 3668 |
+
"learning_rate": 6.059544257862343e-07,
|
| 3669 |
+
"loss": 1.0842,
|
| 3670 |
+
"step": 2615
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 2.79,
|
| 3674 |
+
"grad_norm": 4.71875,
|
| 3675 |
+
"learning_rate": 5.75550390028462e-07,
|
| 3676 |
+
"loss": 1.077,
|
| 3677 |
+
"step": 2620
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 2.8,
|
| 3681 |
+
"grad_norm": 6.03125,
|
| 3682 |
+
"learning_rate": 5.459200028291617e-07,
|
| 3683 |
+
"loss": 1.1295,
|
| 3684 |
+
"step": 2625
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 2.81,
|
| 3688 |
+
"grad_norm": 5.90625,
|
| 3689 |
+
"learning_rate": 5.170642027358341e-07,
|
| 3690 |
+
"loss": 1.1019,
|
| 3691 |
+
"step": 2630
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 2.81,
|
| 3695 |
+
"grad_norm": 5.96875,
|
| 3696 |
+
"learning_rate": 4.889839037608202e-07,
|
| 3697 |
+
"loss": 1.1931,
|
| 3698 |
+
"step": 2635
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 2.82,
|
| 3702 |
+
"grad_norm": 5.75,
|
| 3703 |
+
"learning_rate": 4.616799953523182e-07,
|
| 3704 |
+
"loss": 1.1147,
|
| 3705 |
+
"step": 2640
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 2.82,
|
| 3709 |
+
"grad_norm": 5.8125,
|
| 3710 |
+
"learning_rate": 4.351533423662318e-07,
|
| 3711 |
+
"loss": 1.1505,
|
| 3712 |
+
"step": 2645
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 2.83,
|
| 3716 |
+
"grad_norm": 5.375,
|
| 3717 |
+
"learning_rate": 4.094047850387667e-07,
|
| 3718 |
+
"loss": 1.1301,
|
| 3719 |
+
"step": 2650
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 2.83,
|
| 3723 |
+
"grad_norm": 5.6875,
|
| 3724 |
+
"learning_rate": 3.8443513895981e-07,
|
| 3725 |
+
"loss": 1.1204,
|
| 3726 |
+
"step": 2655
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 2.84,
|
| 3730 |
+
"grad_norm": 5.6875,
|
| 3731 |
+
"learning_rate": 3.6024519504711806e-07,
|
| 3732 |
+
"loss": 1.0925,
|
| 3733 |
+
"step": 2660
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 2.84,
|
| 3737 |
+
"grad_norm": 5.84375,
|
| 3738 |
+
"learning_rate": 3.3683571952123906e-07,
|
| 3739 |
+
"loss": 1.1177,
|
| 3740 |
+
"step": 2665
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 2.85,
|
| 3744 |
+
"grad_norm": 6.09375,
|
| 3745 |
+
"learning_rate": 3.142074538812545e-07,
|
| 3746 |
+
"loss": 1.1465,
|
| 3747 |
+
"step": 2670
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 2.85,
|
| 3751 |
+
"grad_norm": 5.25,
|
| 3752 |
+
"learning_rate": 2.9236111488129545e-07,
|
| 3753 |
+
"loss": 1.1202,
|
| 3754 |
+
"step": 2675
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 2.86,
|
| 3758 |
+
"grad_norm": 6.84375,
|
| 3759 |
+
"learning_rate": 2.712973945078301e-07,
|
| 3760 |
+
"loss": 1.1668,
|
| 3761 |
+
"step": 2680
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 2.86,
|
| 3765 |
+
"grad_norm": 5.09375,
|
| 3766 |
+
"learning_rate": 2.510169599577533e-07,
|
| 3767 |
+
"loss": 1.1126,
|
| 3768 |
+
"step": 2685
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 2.87,
|
| 3772 |
+
"grad_norm": 6.28125,
|
| 3773 |
+
"learning_rate": 2.315204536172455e-07,
|
| 3774 |
+
"loss": 1.1607,
|
| 3775 |
+
"step": 2690
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 2.87,
|
| 3779 |
+
"grad_norm": 5.84375,
|
| 3780 |
+
"learning_rate": 2.128084930414359e-07,
|
| 3781 |
+
"loss": 1.1588,
|
| 3782 |
+
"step": 2695
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 2.88,
|
| 3786 |
+
"grad_norm": 6.90625,
|
| 3787 |
+
"learning_rate": 1.9488167093482667e-07,
|
| 3788 |
+
"loss": 1.1142,
|
| 3789 |
+
"step": 2700
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 2.89,
|
| 3793 |
+
"grad_norm": 5.90625,
|
| 3794 |
+
"learning_rate": 1.7774055513253006e-07,
|
| 3795 |
+
"loss": 1.1313,
|
| 3796 |
+
"step": 2705
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 2.89,
|
| 3800 |
+
"grad_norm": 5.46875,
|
| 3801 |
+
"learning_rate": 1.613856885822801e-07,
|
| 3802 |
+
"loss": 1.0984,
|
| 3803 |
+
"step": 2710
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 2.9,
|
| 3807 |
+
"grad_norm": 5.34375,
|
| 3808 |
+
"learning_rate": 1.4581758932722956e-07,
|
| 3809 |
+
"loss": 1.0768,
|
| 3810 |
+
"step": 2715
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 2.9,
|
| 3814 |
+
"grad_norm": 5.375,
|
| 3815 |
+
"learning_rate": 1.3103675048954934e-07,
|
| 3816 |
+
"loss": 1.1069,
|
| 3817 |
+
"step": 2720
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 2.91,
|
| 3821 |
+
"grad_norm": 5.25,
|
| 3822 |
+
"learning_rate": 1.1704364025479642e-07,
|
| 3823 |
+
"loss": 1.1131,
|
| 3824 |
+
"step": 2725
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 2.91,
|
| 3828 |
+
"grad_norm": 6.46875,
|
| 3829 |
+
"learning_rate": 1.0383870185710076e-07,
|
| 3830 |
+
"loss": 1.1476,
|
| 3831 |
+
"step": 2730
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 2.92,
|
| 3835 |
+
"grad_norm": 6.71875,
|
| 3836 |
+
"learning_rate": 9.142235356510987e-08,
|
| 3837 |
+
"loss": 1.1705,
|
| 3838 |
+
"step": 2735
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 2.92,
|
| 3842 |
+
"grad_norm": 5.0,
|
| 3843 |
+
"learning_rate": 7.979498866874935e-08,
|
| 3844 |
+
"loss": 1.069,
|
| 3845 |
+
"step": 2740
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 2.93,
|
| 3849 |
+
"grad_norm": 5.4375,
|
| 3850 |
+
"learning_rate": 6.895697546676349e-08,
|
| 3851 |
+
"loss": 1.1401,
|
| 3852 |
+
"step": 2745
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 2.93,
|
| 3856 |
+
"grad_norm": 6.65625,
|
| 3857 |
+
"learning_rate": 5.890865725504402e-08,
|
| 3858 |
+
"loss": 1.119,
|
| 3859 |
+
"step": 2750
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 2.94,
|
| 3863 |
+
"grad_norm": 6.09375,
|
| 3864 |
+
"learning_rate": 4.9650352315766535e-08,
|
| 3865 |
+
"loss": 1.1177,
|
| 3866 |
+
"step": 2755
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 2.94,
|
| 3870 |
+
"grad_norm": 5.6875,
|
| 3871 |
+
"learning_rate": 4.118235390730141e-08,
|
| 3872 |
+
"loss": 1.0892,
|
| 3873 |
+
"step": 2760
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 2.95,
|
| 3877 |
+
"grad_norm": 8.9375,
|
| 3878 |
+
"learning_rate": 3.350493025492396e-08,
|
| 3879 |
+
"loss": 1.1853,
|
| 3880 |
+
"step": 2765
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 2.95,
|
| 3884 |
+
"grad_norm": 5.21875,
|
| 3885 |
+
"learning_rate": 2.661832454232127e-08,
|
| 3886 |
+
"loss": 1.1004,
|
| 3887 |
+
"step": 2770
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 2.96,
|
| 3891 |
+
"grad_norm": 5.71875,
|
| 3892 |
+
"learning_rate": 2.0522754903892772e-08,
|
| 3893 |
+
"loss": 1.113,
|
| 3894 |
+
"step": 2775
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 2.97,
|
| 3898 |
+
"grad_norm": 5.5,
|
| 3899 |
+
"learning_rate": 1.5218414417833558e-08,
|
| 3900 |
+
"loss": 1.1525,
|
| 3901 |
+
"step": 2780
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 2.97,
|
| 3905 |
+
"grad_norm": 5.34375,
|
| 3906 |
+
"learning_rate": 1.0705471100022624e-08,
|
| 3907 |
+
"loss": 1.1145,
|
| 3908 |
+
"step": 2785
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 2.98,
|
| 3912 |
+
"grad_norm": 4.46875,
|
| 3913 |
+
"learning_rate": 6.9840678986993205e-09,
|
| 3914 |
+
"loss": 1.0893,
|
| 3915 |
+
"step": 2790
|
| 3916 |
+
},
|
| 3917 |
+
{
|
| 3918 |
+
"epoch": 2.98,
|
| 3919 |
+
"grad_norm": 6.3125,
|
| 3920 |
+
"learning_rate": 4.054322689936441e-09,
|
| 3921 |
+
"loss": 1.1528,
|
| 3922 |
+
"step": 2795
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 2.99,
|
| 3926 |
+
"grad_norm": 5.6875,
|
| 3927 |
+
"learning_rate": 1.9163282739126376e-09,
|
| 3928 |
+
"loss": 1.1335,
|
| 3929 |
+
"step": 2800
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 2.99,
|
| 3933 |
+
"grad_norm": 7.53125,
|
| 3934 |
+
"learning_rate": 5.701523719536805e-10,
|
| 3935 |
+
"loss": 1.1066,
|
| 3936 |
+
"step": 2805
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 3.0,
|
| 3940 |
+
"grad_norm": 5.875,
|
| 3941 |
+
"learning_rate": 1.5837624414705154e-11,
|
| 3942 |
+
"loss": 1.2276,
|
| 3943 |
+
"step": 2810
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 3.0,
|
| 3947 |
+
"step": 2811,
|
| 3948 |
+
"total_flos": 3.2918248209933926e+17,
|
| 3949 |
+
"train_loss": 1.199945468658317,
|
| 3950 |
+
"train_runtime": 5291.6066,
|
| 3951 |
+
"train_samples_per_second": 17.008,
|
| 3952 |
+
"train_steps_per_second": 0.531
|
| 3953 |
+
}
|
| 3954 |
+
],
|
| 3955 |
+
"logging_steps": 5,
|
| 3956 |
+
"max_steps": 2811,
|
| 3957 |
+
"num_input_tokens_seen": 0,
|
| 3958 |
+
"num_train_epochs": 3,
|
| 3959 |
+
"save_steps": 30000,
|
| 3960 |
+
"total_flos": 3.2918248209933926e+17,
|
| 3961 |
+
"train_batch_size": 8,
|
| 3962 |
+
"trial_name": null,
|
| 3963 |
+
"trial_params": null
|
| 3964 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38fedf6be0e07c0425685dc8679590ba54e808a58489299659d5bf0cb4a7f824
|
| 3 |
+
size 5112
|
training_loss.png
ADDED
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|