Upload 7 files
Browse files- generation_config.json +6 -0
- merges.txt +0 -0
- special_tokens_map.json +6 -0
- tokenizer.json +0 -0
- tokenizer_config.json +21 -0
- trainer_state.json +2161 -0
- vocab.json +0 -0
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.55.2"
|
| 6 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<|endoftext|>",
|
| 3 |
+
"eos_token": "<|endoftext|>",
|
| 4 |
+
"pad_token": "<|endoftext|>",
|
| 5 |
+
"unk_token": "<|endoftext|>"
|
| 6 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"50256": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
}
|
| 12 |
+
},
|
| 13 |
+
"bos_token": "<|endoftext|>",
|
| 14 |
+
"clean_up_tokenization_spaces": false,
|
| 15 |
+
"eos_token": "<|endoftext|>",
|
| 16 |
+
"extra_special_tokens": {},
|
| 17 |
+
"model_max_length": 1024,
|
| 18 |
+
"pad_token": "<|endoftext|>",
|
| 19 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 20 |
+
"unk_token": "<|endoftext|>"
|
| 21 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2161 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 2481,
|
| 7 |
+
"global_step": 14888,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.006717270101430779,
|
| 14 |
+
"grad_norm": 4.896524906158447,
|
| 15 |
+
"learning_rate": 5.480984340044743e-06,
|
| 16 |
+
"loss": 5.1811,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.013434540202861557,
|
| 21 |
+
"grad_norm": 2.687243700027466,
|
| 22 |
+
"learning_rate": 1.1073825503355706e-05,
|
| 23 |
+
"loss": 4.4733,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.020151810304292337,
|
| 28 |
+
"grad_norm": 1.7079490423202515,
|
| 29 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 30 |
+
"loss": 3.8,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.026869080405723115,
|
| 35 |
+
"grad_norm": 1.7120026350021362,
|
| 36 |
+
"learning_rate": 2.225950782997763e-05,
|
| 37 |
+
"loss": 3.3653,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03358635050715389,
|
| 42 |
+
"grad_norm": 1.4521100521087646,
|
| 43 |
+
"learning_rate": 2.785234899328859e-05,
|
| 44 |
+
"loss": 3.1308,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.040303620608584674,
|
| 49 |
+
"grad_norm": 1.4319677352905273,
|
| 50 |
+
"learning_rate": 3.3445190156599555e-05,
|
| 51 |
+
"loss": 2.9757,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.04702089071001545,
|
| 56 |
+
"grad_norm": 1.053918719291687,
|
| 57 |
+
"learning_rate": 3.903803131991052e-05,
|
| 58 |
+
"loss": 2.8591,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.05373816081144623,
|
| 63 |
+
"grad_norm": 0.9901153445243835,
|
| 64 |
+
"learning_rate": 4.463087248322148e-05,
|
| 65 |
+
"loss": 2.7635,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.06045543091287701,
|
| 70 |
+
"grad_norm": 1.0325074195861816,
|
| 71 |
+
"learning_rate": 4.999999763367056e-05,
|
| 72 |
+
"loss": 2.6867,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.06717270101430778,
|
| 77 |
+
"grad_norm": 0.7236190438270569,
|
| 78 |
+
"learning_rate": 4.999840037833225e-05,
|
| 79 |
+
"loss": 2.6204,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.07388997111573857,
|
| 84 |
+
"grad_norm": 0.7032467126846313,
|
| 85 |
+
"learning_rate": 4.9993845429571105e-05,
|
| 86 |
+
"loss": 2.5692,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.08060724121716935,
|
| 91 |
+
"grad_norm": 0.5874780416488647,
|
| 92 |
+
"learning_rate": 4.9986333326307276e-05,
|
| 93 |
+
"loss": 2.5271,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.08732451131860013,
|
| 98 |
+
"grad_norm": 0.5380228161811829,
|
| 99 |
+
"learning_rate": 4.997586495733758e-05,
|
| 100 |
+
"loss": 2.498,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.0940417814200309,
|
| 105 |
+
"grad_norm": 0.5316600203514099,
|
| 106 |
+
"learning_rate": 4.996244156123031e-05,
|
| 107 |
+
"loss": 2.4672,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.10075905152146168,
|
| 112 |
+
"grad_norm": 0.4045695662498474,
|
| 113 |
+
"learning_rate": 4.994606472617869e-05,
|
| 114 |
+
"loss": 2.4403,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.10747632162289246,
|
| 119 |
+
"grad_norm": 0.4827907979488373,
|
| 120 |
+
"learning_rate": 4.9926736389813e-05,
|
| 121 |
+
"loss": 2.4176,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.11419359172432324,
|
| 126 |
+
"grad_norm": 0.4095781743526459,
|
| 127 |
+
"learning_rate": 4.99044588389713e-05,
|
| 128 |
+
"loss": 2.4055,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.12091086182575402,
|
| 133 |
+
"grad_norm": 0.3878771662712097,
|
| 134 |
+
"learning_rate": 4.9879234709428855e-05,
|
| 135 |
+
"loss": 2.3713,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.12762813192718478,
|
| 140 |
+
"grad_norm": 0.3336116671562195,
|
| 141 |
+
"learning_rate": 4.9851066985586316e-05,
|
| 142 |
+
"loss": 2.3656,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.13434540202861556,
|
| 147 |
+
"grad_norm": 0.3157025873661041,
|
| 148 |
+
"learning_rate": 4.981995900011657e-05,
|
| 149 |
+
"loss": 2.3446,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.14106267213004636,
|
| 154 |
+
"grad_norm": 0.305578351020813,
|
| 155 |
+
"learning_rate": 4.978591443357048e-05,
|
| 156 |
+
"loss": 2.3291,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.14777994223147714,
|
| 161 |
+
"grad_norm": 0.34396952390670776,
|
| 162 |
+
"learning_rate": 4.9748937313941414e-05,
|
| 163 |
+
"loss": 2.3196,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.15449721233290792,
|
| 168 |
+
"grad_norm": 0.32559117674827576,
|
| 169 |
+
"learning_rate": 4.970903201618863e-05,
|
| 170 |
+
"loss": 2.3039,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1612144824343387,
|
| 175 |
+
"grad_norm": 0.28197962045669556,
|
| 176 |
+
"learning_rate": 4.966620326171969e-05,
|
| 177 |
+
"loss": 2.2978,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.16793175253576947,
|
| 182 |
+
"grad_norm": 0.3120637536048889,
|
| 183 |
+
"learning_rate": 4.962045611783186e-05,
|
| 184 |
+
"loss": 2.2814,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.17464902263720025,
|
| 189 |
+
"grad_norm": 0.31267744302749634,
|
| 190 |
+
"learning_rate": 4.9571795997112506e-05,
|
| 191 |
+
"loss": 2.2717,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.18136629273863103,
|
| 196 |
+
"grad_norm": 0.3214496374130249,
|
| 197 |
+
"learning_rate": 4.9520228656798784e-05,
|
| 198 |
+
"loss": 2.2573,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1880835628400618,
|
| 203 |
+
"grad_norm": 0.36649203300476074,
|
| 204 |
+
"learning_rate": 4.946576019809639e-05,
|
| 205 |
+
"loss": 2.246,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.19480083294149259,
|
| 210 |
+
"grad_norm": 0.31962043046951294,
|
| 211 |
+
"learning_rate": 4.940839706545777e-05,
|
| 212 |
+
"loss": 2.2388,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.20151810304292336,
|
| 217 |
+
"grad_norm": 0.395917683839798,
|
| 218 |
+
"learning_rate": 4.9348146045819585e-05,
|
| 219 |
+
"loss": 2.2331,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.20823537314435414,
|
| 224 |
+
"grad_norm": 0.3968123495578766,
|
| 225 |
+
"learning_rate": 4.928501426779974e-05,
|
| 226 |
+
"loss": 2.2212,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.21495264324578492,
|
| 231 |
+
"grad_norm": 0.3795957863330841,
|
| 232 |
+
"learning_rate": 4.921900920085394e-05,
|
| 233 |
+
"loss": 2.2153,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2216699133472157,
|
| 238 |
+
"grad_norm": 0.39966660737991333,
|
| 239 |
+
"learning_rate": 4.915013865439197e-05,
|
| 240 |
+
"loss": 2.2015,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.22838718344864647,
|
| 245 |
+
"grad_norm": 0.3845181465148926,
|
| 246 |
+
"learning_rate": 4.907841077685372e-05,
|
| 247 |
+
"loss": 2.204,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.23510445355007725,
|
| 252 |
+
"grad_norm": 0.48438650369644165,
|
| 253 |
+
"learning_rate": 4.900383405474503e-05,
|
| 254 |
+
"loss": 2.1953,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.24182172365150803,
|
| 259 |
+
"grad_norm": 0.4857676923274994,
|
| 260 |
+
"learning_rate": 4.892641731163372e-05,
|
| 261 |
+
"loss": 2.1815,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.2485389937529388,
|
| 266 |
+
"grad_norm": 0.6407597661018372,
|
| 267 |
+
"learning_rate": 4.8846169707105525e-05,
|
| 268 |
+
"loss": 2.1755,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.25525626385436956,
|
| 273 |
+
"grad_norm": 0.5669896006584167,
|
| 274 |
+
"learning_rate": 4.8763100735680445e-05,
|
| 275 |
+
"loss": 2.1698,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.26197353395580036,
|
| 280 |
+
"grad_norm": 0.5839398503303528,
|
| 281 |
+
"learning_rate": 4.867722022568936e-05,
|
| 282 |
+
"loss": 2.165,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.2686908040572311,
|
| 287 |
+
"grad_norm": 0.6011441946029663,
|
| 288 |
+
"learning_rate": 4.858853833811119e-05,
|
| 289 |
+
"loss": 2.1574,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2754080741586619,
|
| 294 |
+
"grad_norm": 0.562146008014679,
|
| 295 |
+
"learning_rate": 4.849706556537074e-05,
|
| 296 |
+
"loss": 2.146,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.2821253442600927,
|
| 301 |
+
"grad_norm": 0.5889083743095398,
|
| 302 |
+
"learning_rate": 4.840281273009719e-05,
|
| 303 |
+
"loss": 2.1465,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2888426143615235,
|
| 308 |
+
"grad_norm": 0.6204415559768677,
|
| 309 |
+
"learning_rate": 4.8305790983843744e-05,
|
| 310 |
+
"loss": 2.1457,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2955598844629543,
|
| 315 |
+
"grad_norm": 0.6884430050849915,
|
| 316 |
+
"learning_rate": 4.820601180576811e-05,
|
| 317 |
+
"loss": 2.1367,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.30227715456438503,
|
| 322 |
+
"grad_norm": 0.6785121560096741,
|
| 323 |
+
"learning_rate": 4.810348700127441e-05,
|
| 324 |
+
"loss": 2.1377,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.30899442466581584,
|
| 329 |
+
"grad_norm": 0.667234480381012,
|
| 330 |
+
"learning_rate": 4.7998228700616384e-05,
|
| 331 |
+
"loss": 2.1279,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3157116947672466,
|
| 336 |
+
"grad_norm": 0.5685933232307434,
|
| 337 |
+
"learning_rate": 4.789024935746223e-05,
|
| 338 |
+
"loss": 2.1165,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3224289648686774,
|
| 343 |
+
"grad_norm": 0.7456786036491394,
|
| 344 |
+
"learning_rate": 4.7779561747421106e-05,
|
| 345 |
+
"loss": 2.1148,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.32914623497010814,
|
| 350 |
+
"grad_norm": 0.7704935669898987,
|
| 351 |
+
"learning_rate": 4.766617896653162e-05,
|
| 352 |
+
"loss": 2.1035,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.33331094243299525,
|
| 357 |
+
"eval_loss": 2.031871795654297,
|
| 358 |
+
"eval_runtime": 1653.4784,
|
| 359 |
+
"eval_samples_per_second": 72.217,
|
| 360 |
+
"eval_steps_per_second": 9.028,
|
| 361 |
+
"step": 2481
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.33586350507153895,
|
| 365 |
+
"grad_norm": 0.6733186841011047,
|
| 366 |
+
"learning_rate": 4.755011442971233e-05,
|
| 367 |
+
"loss": 2.1033,
|
| 368 |
+
"step": 2500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.3425807751729697,
|
| 372 |
+
"grad_norm": 0.7341210246086121,
|
| 373 |
+
"learning_rate": 4.7431381869174574e-05,
|
| 374 |
+
"loss": 2.1,
|
| 375 |
+
"step": 2550
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.3492980452744005,
|
| 379 |
+
"grad_norm": 0.7517989277839661,
|
| 380 |
+
"learning_rate": 4.730999533279775e-05,
|
| 381 |
+
"loss": 2.0965,
|
| 382 |
+
"step": 2600
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.35601531537583125,
|
| 386 |
+
"grad_norm": 0.7430043816566467,
|
| 387 |
+
"learning_rate": 4.71859691824672e-05,
|
| 388 |
+
"loss": 2.0893,
|
| 389 |
+
"step": 2650
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.36273258547726206,
|
| 393 |
+
"grad_norm": 0.7112200856208801,
|
| 394 |
+
"learning_rate": 4.7059318092375016e-05,
|
| 395 |
+
"loss": 2.0856,
|
| 396 |
+
"step": 2700
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.3694498555786928,
|
| 400 |
+
"grad_norm": 0.8947717547416687,
|
| 401 |
+
"learning_rate": 4.693005704728384e-05,
|
| 402 |
+
"loss": 2.0832,
|
| 403 |
+
"step": 2750
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.3761671256801236,
|
| 407 |
+
"grad_norm": 0.8591889142990112,
|
| 408 |
+
"learning_rate": 4.679820134075395e-05,
|
| 409 |
+
"loss": 2.0761,
|
| 410 |
+
"step": 2800
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.38288439578155437,
|
| 414 |
+
"grad_norm": 0.933829665184021,
|
| 415 |
+
"learning_rate": 4.666376657333379e-05,
|
| 416 |
+
"loss": 2.0781,
|
| 417 |
+
"step": 2850
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.38960166588298517,
|
| 421 |
+
"grad_norm": 0.8115527629852295,
|
| 422 |
+
"learning_rate": 4.652676865071417e-05,
|
| 423 |
+
"loss": 2.0758,
|
| 424 |
+
"step": 2900
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.3963189359844159,
|
| 428 |
+
"grad_norm": 0.7666231989860535,
|
| 429 |
+
"learning_rate": 4.638722378184641e-05,
|
| 430 |
+
"loss": 2.0693,
|
| 431 |
+
"step": 2950
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.4030362060858467,
|
| 435 |
+
"grad_norm": 0.9425594806671143,
|
| 436 |
+
"learning_rate": 4.624514847702454e-05,
|
| 437 |
+
"loss": 2.0592,
|
| 438 |
+
"step": 3000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.4097534761872775,
|
| 442 |
+
"grad_norm": 0.8913134336471558,
|
| 443 |
+
"learning_rate": 4.610055954593192e-05,
|
| 444 |
+
"loss": 2.0595,
|
| 445 |
+
"step": 3050
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.4164707462887083,
|
| 449 |
+
"grad_norm": 0.9630874991416931,
|
| 450 |
+
"learning_rate": 4.595347409565237e-05,
|
| 451 |
+
"loss": 2.051,
|
| 452 |
+
"step": 3100
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.42318801639013903,
|
| 456 |
+
"grad_norm": 0.7468298673629761,
|
| 457 |
+
"learning_rate": 4.5803909528646125e-05,
|
| 458 |
+
"loss": 2.0541,
|
| 459 |
+
"step": 3150
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.42990528649156984,
|
| 463 |
+
"grad_norm": 0.9688680768013,
|
| 464 |
+
"learning_rate": 4.565188354069091e-05,
|
| 465 |
+
"loss": 2.0416,
|
| 466 |
+
"step": 3200
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.4366225565930006,
|
| 470 |
+
"grad_norm": 0.9443581700325012,
|
| 471 |
+
"learning_rate": 4.549741411878819e-05,
|
| 472 |
+
"loss": 2.0356,
|
| 473 |
+
"step": 3250
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4433398266944314,
|
| 477 |
+
"grad_norm": 0.8630342483520508,
|
| 478 |
+
"learning_rate": 4.534051953903511e-05,
|
| 479 |
+
"loss": 2.0455,
|
| 480 |
+
"step": 3300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.45005709679586214,
|
| 484 |
+
"grad_norm": 0.8253282904624939,
|
| 485 |
+
"learning_rate": 4.518121836446206e-05,
|
| 486 |
+
"loss": 2.0378,
|
| 487 |
+
"step": 3350
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.45677436689729295,
|
| 491 |
+
"grad_norm": 0.8722043633460999,
|
| 492 |
+
"learning_rate": 4.501952944283647e-05,
|
| 493 |
+
"loss": 2.0345,
|
| 494 |
+
"step": 3400
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.4634916369987237,
|
| 498 |
+
"grad_norm": 0.833014965057373,
|
| 499 |
+
"learning_rate": 4.4855471904432804e-05,
|
| 500 |
+
"loss": 2.0393,
|
| 501 |
+
"step": 3450
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4702089071001545,
|
| 505 |
+
"grad_norm": 0.8858036994934082,
|
| 506 |
+
"learning_rate": 4.468906515976912e-05,
|
| 507 |
+
"loss": 2.0342,
|
| 508 |
+
"step": 3500
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.47692617720158526,
|
| 512 |
+
"grad_norm": 0.7766168713569641,
|
| 513 |
+
"learning_rate": 4.452032889731056e-05,
|
| 514 |
+
"loss": 2.0255,
|
| 515 |
+
"step": 3550
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.48364344730301606,
|
| 519 |
+
"grad_norm": 1.0143591165542603,
|
| 520 |
+
"learning_rate": 4.434928308113986e-05,
|
| 521 |
+
"loss": 2.0183,
|
| 522 |
+
"step": 3600
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.4903607174044468,
|
| 526 |
+
"grad_norm": 0.8125320672988892,
|
| 527 |
+
"learning_rate": 4.417594794859533e-05,
|
| 528 |
+
"loss": 2.0202,
|
| 529 |
+
"step": 3650
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.4970779875058776,
|
| 533 |
+
"grad_norm": 0.7722158432006836,
|
| 534 |
+
"learning_rate": 4.4000344007876444e-05,
|
| 535 |
+
"loss": 2.0114,
|
| 536 |
+
"step": 3700
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.5037952576073084,
|
| 540 |
+
"grad_norm": 0.9473446607589722,
|
| 541 |
+
"learning_rate": 4.3822492035617404e-05,
|
| 542 |
+
"loss": 2.0125,
|
| 543 |
+
"step": 3750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.5105125277087391,
|
| 547 |
+
"grad_norm": 0.854096531867981,
|
| 548 |
+
"learning_rate": 4.3642413074428964e-05,
|
| 549 |
+
"loss": 2.0131,
|
| 550 |
+
"step": 3800
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.51722979781017,
|
| 554 |
+
"grad_norm": 0.9034811854362488,
|
| 555 |
+
"learning_rate": 4.346012843040877e-05,
|
| 556 |
+
"loss": 2.0056,
|
| 557 |
+
"step": 3850
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.5239470679116007,
|
| 561 |
+
"grad_norm": 0.8732491135597229,
|
| 562 |
+
"learning_rate": 4.327565967062048e-05,
|
| 563 |
+
"loss": 2.0074,
|
| 564 |
+
"step": 3900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.5306643380130315,
|
| 568 |
+
"grad_norm": 1.0670524835586548,
|
| 569 |
+
"learning_rate": 4.3089028620542094e-05,
|
| 570 |
+
"loss": 2.0069,
|
| 571 |
+
"step": 3950
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.5373816081144622,
|
| 575 |
+
"grad_norm": 0.8823853135108948,
|
| 576 |
+
"learning_rate": 4.2900257361483666e-05,
|
| 577 |
+
"loss": 2.0071,
|
| 578 |
+
"step": 4000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.5440988782158931,
|
| 582 |
+
"grad_norm": 0.9907485842704773,
|
| 583 |
+
"learning_rate": 4.2709368227974724e-05,
|
| 584 |
+
"loss": 2.0083,
|
| 585 |
+
"step": 4050
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.5508161483173238,
|
| 589 |
+
"grad_norm": 1.0381219387054443,
|
| 590 |
+
"learning_rate": 4.251638380512174e-05,
|
| 591 |
+
"loss": 2.0028,
|
| 592 |
+
"step": 4100
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.5575334184187546,
|
| 596 |
+
"grad_norm": 0.8755475878715515,
|
| 597 |
+
"learning_rate": 4.232132692593602e-05,
|
| 598 |
+
"loss": 1.9974,
|
| 599 |
+
"step": 4150
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.5642506885201855,
|
| 603 |
+
"grad_norm": 1.062873125076294,
|
| 604 |
+
"learning_rate": 4.212422066863218e-05,
|
| 605 |
+
"loss": 1.9859,
|
| 606 |
+
"step": 4200
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.5709679586216162,
|
| 610 |
+
"grad_norm": 0.9898818731307983,
|
| 611 |
+
"learning_rate": 4.19250883538976e-05,
|
| 612 |
+
"loss": 1.9922,
|
| 613 |
+
"step": 4250
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.577685228723047,
|
| 617 |
+
"grad_norm": 0.9990688562393188,
|
| 618 |
+
"learning_rate": 4.172395354213331e-05,
|
| 619 |
+
"loss": 1.9935,
|
| 620 |
+
"step": 4300
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.5844024988244777,
|
| 624 |
+
"grad_norm": 0.8213431239128113,
|
| 625 |
+
"learning_rate": 4.152084003066636e-05,
|
| 626 |
+
"loss": 1.9867,
|
| 627 |
+
"step": 4350
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.5911197689259086,
|
| 631 |
+
"grad_norm": 0.913465142250061,
|
| 632 |
+
"learning_rate": 4.1315771850934295e-05,
|
| 633 |
+
"loss": 1.9809,
|
| 634 |
+
"step": 4400
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.5978370390273393,
|
| 638 |
+
"grad_norm": 0.84449303150177,
|
| 639 |
+
"learning_rate": 4.110877326564179e-05,
|
| 640 |
+
"loss": 1.9707,
|
| 641 |
+
"step": 4450
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.6045543091287701,
|
| 645 |
+
"grad_norm": 0.9786863327026367,
|
| 646 |
+
"learning_rate": 4.08998687658901e-05,
|
| 647 |
+
"loss": 1.9753,
|
| 648 |
+
"step": 4500
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.6112715792302008,
|
| 652 |
+
"grad_norm": 1.060813069343567,
|
| 653 |
+
"learning_rate": 4.06890830682793e-05,
|
| 654 |
+
"loss": 1.9804,
|
| 655 |
+
"step": 4550
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.6179888493316317,
|
| 659 |
+
"grad_norm": 0.9997825026512146,
|
| 660 |
+
"learning_rate": 4.047644111198398e-05,
|
| 661 |
+
"loss": 1.974,
|
| 662 |
+
"step": 4600
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.6247061194330624,
|
| 666 |
+
"grad_norm": 0.8858147859573364,
|
| 667 |
+
"learning_rate": 4.026196805580253e-05,
|
| 668 |
+
"loss": 1.9705,
|
| 669 |
+
"step": 4650
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.6314233895344932,
|
| 673 |
+
"grad_norm": 1.018781065940857,
|
| 674 |
+
"learning_rate": 4.004568927518054e-05,
|
| 675 |
+
"loss": 1.9666,
|
| 676 |
+
"step": 4700
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.6381406596359239,
|
| 680 |
+
"grad_norm": 0.9801031947135925,
|
| 681 |
+
"learning_rate": 3.982763035920836e-05,
|
| 682 |
+
"loss": 1.9634,
|
| 683 |
+
"step": 4750
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.6448579297373548,
|
| 687 |
+
"grad_norm": 0.9946137070655823,
|
| 688 |
+
"learning_rate": 3.960781710759365e-05,
|
| 689 |
+
"loss": 1.9661,
|
| 690 |
+
"step": 4800
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.6515751998387855,
|
| 694 |
+
"grad_norm": 0.9870684146881104,
|
| 695 |
+
"learning_rate": 3.9386275527608845e-05,
|
| 696 |
+
"loss": 1.9616,
|
| 697 |
+
"step": 4850
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.6582924699402163,
|
| 701 |
+
"grad_norm": 0.9816562533378601,
|
| 702 |
+
"learning_rate": 3.916303183101405e-05,
|
| 703 |
+
"loss": 1.969,
|
| 704 |
+
"step": 4900
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.665009740041647,
|
| 708 |
+
"grad_norm": 0.9809496402740479,
|
| 709 |
+
"learning_rate": 3.8938112430955834e-05,
|
| 710 |
+
"loss": 1.964,
|
| 711 |
+
"step": 4950
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.6666218848659905,
|
| 715 |
+
"eval_loss": 1.8941781520843506,
|
| 716 |
+
"eval_runtime": 1652.0834,
|
| 717 |
+
"eval_samples_per_second": 72.278,
|
| 718 |
+
"eval_steps_per_second": 9.035,
|
| 719 |
+
"step": 4962
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.6717270101430779,
|
| 723 |
+
"grad_norm": 0.8745808005332947,
|
| 724 |
+
"learning_rate": 3.871154393884212e-05,
|
| 725 |
+
"loss": 1.9555,
|
| 726 |
+
"step": 5000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.6784442802445086,
|
| 730 |
+
"grad_norm": 0.97087562084198,
|
| 731 |
+
"learning_rate": 3.848335316119369e-05,
|
| 732 |
+
"loss": 1.9614,
|
| 733 |
+
"step": 5050
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.6851615503459394,
|
| 737 |
+
"grad_norm": 1.2281529903411865,
|
| 738 |
+
"learning_rate": 3.825356709647252e-05,
|
| 739 |
+
"loss": 1.9554,
|
| 740 |
+
"step": 5100
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.6918788204473701,
|
| 744 |
+
"grad_norm": 0.9577364921569824,
|
| 745 |
+
"learning_rate": 3.802221293188748e-05,
|
| 746 |
+
"loss": 1.9569,
|
| 747 |
+
"step": 5150
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.698596090548801,
|
| 751 |
+
"grad_norm": 1.0820568799972534,
|
| 752 |
+
"learning_rate": 3.7789318040177636e-05,
|
| 753 |
+
"loss": 1.9489,
|
| 754 |
+
"step": 5200
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.7053133606502318,
|
| 758 |
+
"grad_norm": 1.0203666687011719,
|
| 759 |
+
"learning_rate": 3.7554909976373685e-05,
|
| 760 |
+
"loss": 1.9484,
|
| 761 |
+
"step": 5250
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.7120306307516625,
|
| 765 |
+
"grad_norm": 0.9432654976844788,
|
| 766 |
+
"learning_rate": 3.731901647453772e-05,
|
| 767 |
+
"loss": 1.9529,
|
| 768 |
+
"step": 5300
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.7187479008530933,
|
| 772 |
+
"grad_norm": 1.0531628131866455,
|
| 773 |
+
"learning_rate": 3.708166544448189e-05,
|
| 774 |
+
"loss": 1.9436,
|
| 775 |
+
"step": 5350
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.7254651709545241,
|
| 779 |
+
"grad_norm": 0.9235082864761353,
|
| 780 |
+
"learning_rate": 3.6842884968466276e-05,
|
| 781 |
+
"loss": 1.9427,
|
| 782 |
+
"step": 5400
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.7321824410559549,
|
| 786 |
+
"grad_norm": 1.1586941480636597,
|
| 787 |
+
"learning_rate": 3.6602703297876276e-05,
|
| 788 |
+
"loss": 1.9427,
|
| 789 |
+
"step": 5450
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.7388997111573856,
|
| 793 |
+
"grad_norm": 1.0403027534484863,
|
| 794 |
+
"learning_rate": 3.636114884988004e-05,
|
| 795 |
+
"loss": 1.9445,
|
| 796 |
+
"step": 5500
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.7456169812588164,
|
| 800 |
+
"grad_norm": 0.9487152099609375,
|
| 801 |
+
"learning_rate": 3.611825020406631e-05,
|
| 802 |
+
"loss": 1.9377,
|
| 803 |
+
"step": 5550
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.7523342513602472,
|
| 807 |
+
"grad_norm": 0.9370021224021912,
|
| 808 |
+
"learning_rate": 3.5874036099063025e-05,
|
| 809 |
+
"loss": 1.939,
|
| 810 |
+
"step": 5600
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.759051521461678,
|
| 814 |
+
"grad_norm": 1.0592392683029175,
|
| 815 |
+
"learning_rate": 3.562853542913706e-05,
|
| 816 |
+
"loss": 1.9331,
|
| 817 |
+
"step": 5650
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.7657687915631087,
|
| 821 |
+
"grad_norm": 0.9699749946594238,
|
| 822 |
+
"learning_rate": 3.538177724077562e-05,
|
| 823 |
+
"loss": 1.9332,
|
| 824 |
+
"step": 5700
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.7724860616645395,
|
| 828 |
+
"grad_norm": 1.1636828184127808,
|
| 829 |
+
"learning_rate": 3.5133790729249585e-05,
|
| 830 |
+
"loss": 1.9382,
|
| 831 |
+
"step": 5750
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.7792033317659703,
|
| 835 |
+
"grad_norm": 0.9427869915962219,
|
| 836 |
+
"learning_rate": 3.488460523515927e-05,
|
| 837 |
+
"loss": 1.9295,
|
| 838 |
+
"step": 5800
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.7859206018674011,
|
| 842 |
+
"grad_norm": 1.0565423965454102,
|
| 843 |
+
"learning_rate": 3.4634250240963e-05,
|
| 844 |
+
"loss": 1.9335,
|
| 845 |
+
"step": 5850
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.7926378719688318,
|
| 849 |
+
"grad_norm": 1.0577119588851929,
|
| 850 |
+
"learning_rate": 3.4382755367488845e-05,
|
| 851 |
+
"loss": 1.9296,
|
| 852 |
+
"step": 5900
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.7993551420702626,
|
| 856 |
+
"grad_norm": 1.0106233358383179,
|
| 857 |
+
"learning_rate": 3.413015037043003e-05,
|
| 858 |
+
"loss": 1.9314,
|
| 859 |
+
"step": 5950
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.8060724121716935,
|
| 863 |
+
"grad_norm": 1.0226566791534424,
|
| 864 |
+
"learning_rate": 3.387646513682442e-05,
|
| 865 |
+
"loss": 1.9265,
|
| 866 |
+
"step": 6000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.8127896822731242,
|
| 870 |
+
"grad_norm": 0.9026387333869934,
|
| 871 |
+
"learning_rate": 3.362172968151838e-05,
|
| 872 |
+
"loss": 1.9248,
|
| 873 |
+
"step": 6050
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.819506952374555,
|
| 877 |
+
"grad_norm": 0.881735622882843,
|
| 878 |
+
"learning_rate": 3.33659741436156e-05,
|
| 879 |
+
"loss": 1.9239,
|
| 880 |
+
"step": 6100
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.8262242224759858,
|
| 884 |
+
"grad_norm": 1.0480873584747314,
|
| 885 |
+
"learning_rate": 3.3109228782911125e-05,
|
| 886 |
+
"loss": 1.9172,
|
| 887 |
+
"step": 6150
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.8329414925774166,
|
| 891 |
+
"grad_norm": 0.9652755260467529,
|
| 892 |
+
"learning_rate": 3.2851523976311214e-05,
|
| 893 |
+
"loss": 1.9212,
|
| 894 |
+
"step": 6200
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.8396587626788473,
|
| 898 |
+
"grad_norm": 0.9710230231285095,
|
| 899 |
+
"learning_rate": 3.2592890214239254e-05,
|
| 900 |
+
"loss": 1.9129,
|
| 901 |
+
"step": 6250
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.8463760327802781,
|
| 905 |
+
"grad_norm": 1.0008606910705566,
|
| 906 |
+
"learning_rate": 3.2333358097028284e-05,
|
| 907 |
+
"loss": 1.9194,
|
| 908 |
+
"step": 6300
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.8530933028817089,
|
| 912 |
+
"grad_norm": 0.9385756850242615,
|
| 913 |
+
"learning_rate": 3.207295833130049e-05,
|
| 914 |
+
"loss": 1.9222,
|
| 915 |
+
"step": 6350
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.8598105729831397,
|
| 919 |
+
"grad_norm": 1.0019196271896362,
|
| 920 |
+
"learning_rate": 3.18117217263342e-05,
|
| 921 |
+
"loss": 1.9175,
|
| 922 |
+
"step": 6400
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.8665278430845704,
|
| 926 |
+
"grad_norm": 0.9140141010284424,
|
| 927 |
+
"learning_rate": 3.154967919041859e-05,
|
| 928 |
+
"loss": 1.9108,
|
| 929 |
+
"step": 6450
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.8732451131860012,
|
| 933 |
+
"grad_norm": 0.8793635368347168,
|
| 934 |
+
"learning_rate": 3.128686172719684e-05,
|
| 935 |
+
"loss": 1.908,
|
| 936 |
+
"step": 6500
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.879962383287432,
|
| 940 |
+
"grad_norm": 1.009559988975525,
|
| 941 |
+
"learning_rate": 3.102330043199787e-05,
|
| 942 |
+
"loss": 1.9126,
|
| 943 |
+
"step": 6550
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.8866796533888628,
|
| 947 |
+
"grad_norm": 0.9987242817878723,
|
| 948 |
+
"learning_rate": 3.07590264881573e-05,
|
| 949 |
+
"loss": 1.915,
|
| 950 |
+
"step": 6600
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8933969234902935,
|
| 954 |
+
"grad_norm": 1.0023444890975952,
|
| 955 |
+
"learning_rate": 3.049407116332802e-05,
|
| 956 |
+
"loss": 1.9125,
|
| 957 |
+
"step": 6650
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.9001141935917243,
|
| 961 |
+
"grad_norm": 0.9578864574432373,
|
| 962 |
+
"learning_rate": 3.022846580578071e-05,
|
| 963 |
+
"loss": 1.9116,
|
| 964 |
+
"step": 6700
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.9068314636931551,
|
| 968 |
+
"grad_norm": 0.9548916220664978,
|
| 969 |
+
"learning_rate": 2.9962241840694872e-05,
|
| 970 |
+
"loss": 1.9059,
|
| 971 |
+
"step": 6750
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.9135487337945859,
|
| 975 |
+
"grad_norm": 1.1118357181549072,
|
| 976 |
+
"learning_rate": 2.9695430766440736e-05,
|
| 977 |
+
"loss": 1.9027,
|
| 978 |
+
"step": 6800
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.9202660038960166,
|
| 982 |
+
"grad_norm": 1.0644875764846802,
|
| 983 |
+
"learning_rate": 2.942806415085255e-05,
|
| 984 |
+
"loss": 1.9093,
|
| 985 |
+
"step": 6850
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.9269832739974474,
|
| 989 |
+
"grad_norm": 0.9042929410934448,
|
| 990 |
+
"learning_rate": 2.9160173627493603e-05,
|
| 991 |
+
"loss": 1.9007,
|
| 992 |
+
"step": 6900
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.9337005440988783,
|
| 996 |
+
"grad_norm": 0.9487243890762329,
|
| 997 |
+
"learning_rate": 2.889179089191349e-05,
|
| 998 |
+
"loss": 1.9017,
|
| 999 |
+
"step": 6950
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.940417814200309,
|
| 1003 |
+
"grad_norm": 1.0207464694976807,
|
| 1004 |
+
"learning_rate": 2.862294769789804e-05,
|
| 1005 |
+
"loss": 1.9048,
|
| 1006 |
+
"step": 7000
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.9471350843017398,
|
| 1010 |
+
"grad_norm": 0.8947831392288208,
|
| 1011 |
+
"learning_rate": 2.8353675853712365e-05,
|
| 1012 |
+
"loss": 1.8949,
|
| 1013 |
+
"step": 7050
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.9538523544031705,
|
| 1017 |
+
"grad_norm": 1.0309513807296753,
|
| 1018 |
+
"learning_rate": 2.8084007218337467e-05,
|
| 1019 |
+
"loss": 1.8972,
|
| 1020 |
+
"step": 7100
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.9605696245046014,
|
| 1024 |
+
"grad_norm": 0.9352649450302124,
|
| 1025 |
+
"learning_rate": 2.7813973697700813e-05,
|
| 1026 |
+
"loss": 1.9065,
|
| 1027 |
+
"step": 7150
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.9672868946060321,
|
| 1031 |
+
"grad_norm": 0.9510485529899597,
|
| 1032 |
+
"learning_rate": 2.754360724090137e-05,
|
| 1033 |
+
"loss": 1.8943,
|
| 1034 |
+
"step": 7200
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.9740041647074629,
|
| 1038 |
+
"grad_norm": 0.9971142411231995,
|
| 1039 |
+
"learning_rate": 2.7272939836429563e-05,
|
| 1040 |
+
"loss": 1.8983,
|
| 1041 |
+
"step": 7250
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.9807214348088936,
|
| 1045 |
+
"grad_norm": 0.9219502806663513,
|
| 1046 |
+
"learning_rate": 2.700200350838253e-05,
|
| 1047 |
+
"loss": 1.9019,
|
| 1048 |
+
"step": 7300
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.9874387049103245,
|
| 1052 |
+
"grad_norm": 1.0496368408203125,
|
| 1053 |
+
"learning_rate": 2.6730830312675182e-05,
|
| 1054 |
+
"loss": 1.8948,
|
| 1055 |
+
"step": 7350
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.9941559750117552,
|
| 1059 |
+
"grad_norm": 0.9791837930679321,
|
| 1060 |
+
"learning_rate": 2.6459452333247497e-05,
|
| 1061 |
+
"loss": 1.8907,
|
| 1062 |
+
"step": 7400
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.9999328272989857,
|
| 1066 |
+
"eval_loss": 1.831755518913269,
|
| 1067 |
+
"eval_runtime": 1652.2475,
|
| 1068 |
+
"eval_samples_per_second": 72.271,
|
| 1069 |
+
"eval_steps_per_second": 9.034,
|
| 1070 |
+
"step": 7443
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.0008060724121717,
|
| 1074 |
+
"grad_norm": 0.8881465196609497,
|
| 1075 |
+
"learning_rate": 2.618790167826851e-05,
|
| 1076 |
+
"loss": 1.8913,
|
| 1077 |
+
"step": 7450
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.0075233425136025,
|
| 1081 |
+
"grad_norm": 0.9141615033149719,
|
| 1082 |
+
"learning_rate": 2.5916210476337416e-05,
|
| 1083 |
+
"loss": 1.8922,
|
| 1084 |
+
"step": 7500
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 1.0142406126150332,
|
| 1088 |
+
"grad_norm": 0.9290676116943359,
|
| 1089 |
+
"learning_rate": 2.5644410872682262e-05,
|
| 1090 |
+
"loss": 1.8921,
|
| 1091 |
+
"step": 7550
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.020957882716464,
|
| 1095 |
+
"grad_norm": 0.9989870190620422,
|
| 1096 |
+
"learning_rate": 2.5372535025356674e-05,
|
| 1097 |
+
"loss": 1.8913,
|
| 1098 |
+
"step": 7600
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.027675152817895,
|
| 1102 |
+
"grad_norm": 0.8889585137367249,
|
| 1103 |
+
"learning_rate": 2.5100615101435078e-05,
|
| 1104 |
+
"loss": 1.8875,
|
| 1105 |
+
"step": 7650
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.0343924229193255,
|
| 1109 |
+
"grad_norm": 0.906791627407074,
|
| 1110 |
+
"learning_rate": 2.4828683273206837e-05,
|
| 1111 |
+
"loss": 1.8811,
|
| 1112 |
+
"step": 7700
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.0411096930207564,
|
| 1116 |
+
"grad_norm": 0.933408260345459,
|
| 1117 |
+
"learning_rate": 2.4556771714369775e-05,
|
| 1118 |
+
"loss": 1.8829,
|
| 1119 |
+
"step": 7750
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.047826963122187,
|
| 1123 |
+
"grad_norm": 0.9643042683601379,
|
| 1124 |
+
"learning_rate": 2.4284912596223532e-05,
|
| 1125 |
+
"loss": 1.8757,
|
| 1126 |
+
"step": 7800
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.054544233223618,
|
| 1130 |
+
"grad_norm": 1.0680428743362427,
|
| 1131 |
+
"learning_rate": 2.4013138083863217e-05,
|
| 1132 |
+
"loss": 1.8896,
|
| 1133 |
+
"step": 7850
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.0612615033250488,
|
| 1137 |
+
"grad_norm": 0.9255633354187012,
|
| 1138 |
+
"learning_rate": 2.3741480332373772e-05,
|
| 1139 |
+
"loss": 1.885,
|
| 1140 |
+
"step": 7900
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 1.0679787734264794,
|
| 1144 |
+
"grad_norm": 0.9109947681427002,
|
| 1145 |
+
"learning_rate": 2.346997148302555e-05,
|
| 1146 |
+
"loss": 1.8732,
|
| 1147 |
+
"step": 7950
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 1.0746960435279103,
|
| 1151 |
+
"grad_norm": 0.9063991904258728,
|
| 1152 |
+
"learning_rate": 2.3198643659471493e-05,
|
| 1153 |
+
"loss": 1.8901,
|
| 1154 |
+
"step": 8000
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.0814133136293411,
|
| 1158 |
+
"grad_norm": 1.0128731727600098,
|
| 1159 |
+
"learning_rate": 2.2927528963946435e-05,
|
| 1160 |
+
"loss": 1.8882,
|
| 1161 |
+
"step": 8050
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.0881305837307718,
|
| 1165 |
+
"grad_norm": 0.8854324221611023,
|
| 1166 |
+
"learning_rate": 2.2656659473468877e-05,
|
| 1167 |
+
"loss": 1.8848,
|
| 1168 |
+
"step": 8100
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 1.0948478538322026,
|
| 1172 |
+
"grad_norm": 1.0122106075286865,
|
| 1173 |
+
"learning_rate": 2.238606723604583e-05,
|
| 1174 |
+
"loss": 1.8738,
|
| 1175 |
+
"step": 8150
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.1015651239336335,
|
| 1179 |
+
"grad_norm": 1.0020389556884766,
|
| 1180 |
+
"learning_rate": 2.2115784266881022e-05,
|
| 1181 |
+
"loss": 1.8758,
|
| 1182 |
+
"step": 8200
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.1082823940350641,
|
| 1186 |
+
"grad_norm": 1.013856291770935,
|
| 1187 |
+
"learning_rate": 2.1845842544587014e-05,
|
| 1188 |
+
"loss": 1.879,
|
| 1189 |
+
"step": 8250
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 1.114999664136495,
|
| 1193 |
+
"grad_norm": 1.0340920686721802,
|
| 1194 |
+
"learning_rate": 2.157627400740161e-05,
|
| 1195 |
+
"loss": 1.874,
|
| 1196 |
+
"step": 8300
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.1217169342379256,
|
| 1200 |
+
"grad_norm": 0.9874864220619202,
|
| 1201 |
+
"learning_rate": 2.1307110549409143e-05,
|
| 1202 |
+
"loss": 1.8789,
|
| 1203 |
+
"step": 8350
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.1284342043393565,
|
| 1207 |
+
"grad_norm": 0.9384549856185913,
|
| 1208 |
+
"learning_rate": 2.1038384016766856e-05,
|
| 1209 |
+
"loss": 1.8736,
|
| 1210 |
+
"step": 8400
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.1351514744407873,
|
| 1214 |
+
"grad_norm": 1.1180434226989746,
|
| 1215 |
+
"learning_rate": 2.0770126203937057e-05,
|
| 1216 |
+
"loss": 1.8709,
|
| 1217 |
+
"step": 8450
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.141868744542218,
|
| 1221 |
+
"grad_norm": 1.0236446857452393,
|
| 1222 |
+
"learning_rate": 2.0502368849925268e-05,
|
| 1223 |
+
"loss": 1.8701,
|
| 1224 |
+
"step": 8500
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 1.1485860146436488,
|
| 1228 |
+
"grad_norm": 0.8483227491378784,
|
| 1229 |
+
"learning_rate": 2.0235143634525144e-05,
|
| 1230 |
+
"loss": 1.8744,
|
| 1231 |
+
"step": 8550
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 1.1553032847450795,
|
| 1235 |
+
"grad_norm": 1.0728415250778198,
|
| 1236 |
+
"learning_rate": 1.9968482174570154e-05,
|
| 1237 |
+
"loss": 1.8707,
|
| 1238 |
+
"step": 8600
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.1620205548465103,
|
| 1242 |
+
"grad_norm": 0.9894828796386719,
|
| 1243 |
+
"learning_rate": 1.970241602019288e-05,
|
| 1244 |
+
"loss": 1.8722,
|
| 1245 |
+
"step": 8650
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.1687378249479412,
|
| 1249 |
+
"grad_norm": 0.9536831378936768,
|
| 1250 |
+
"learning_rate": 1.9436976651092144e-05,
|
| 1251 |
+
"loss": 1.875,
|
| 1252 |
+
"step": 8700
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 1.1754550950493718,
|
| 1256 |
+
"grad_norm": 0.8805471658706665,
|
| 1257 |
+
"learning_rate": 1.9172195472808457e-05,
|
| 1258 |
+
"loss": 1.8671,
|
| 1259 |
+
"step": 8750
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.1821723651508027,
|
| 1263 |
+
"grad_norm": 0.9444993734359741,
|
| 1264 |
+
"learning_rate": 1.890810381300831e-05,
|
| 1265 |
+
"loss": 1.8639,
|
| 1266 |
+
"step": 8800
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.1888896352522336,
|
| 1270 |
+
"grad_norm": 0.9312206506729126,
|
| 1271 |
+
"learning_rate": 1.8644732917777578e-05,
|
| 1272 |
+
"loss": 1.8704,
|
| 1273 |
+
"step": 8850
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 1.1956069053536642,
|
| 1277 |
+
"grad_norm": 0.8733024001121521,
|
| 1278 |
+
"learning_rate": 1.838211394792468e-05,
|
| 1279 |
+
"loss": 1.8661,
|
| 1280 |
+
"step": 8900
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.202324175455095,
|
| 1284 |
+
"grad_norm": 0.9139480590820312,
|
| 1285 |
+
"learning_rate": 1.812027797529372e-05,
|
| 1286 |
+
"loss": 1.8733,
|
| 1287 |
+
"step": 8950
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.209041445556526,
|
| 1291 |
+
"grad_norm": 0.8931354880332947,
|
| 1292 |
+
"learning_rate": 1.7859255979088268e-05,
|
| 1293 |
+
"loss": 1.8613,
|
| 1294 |
+
"step": 9000
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 1.2157587156579566,
|
| 1298 |
+
"grad_norm": 0.8285694122314453,
|
| 1299 |
+
"learning_rate": 1.7599078842206024e-05,
|
| 1300 |
+
"loss": 1.8596,
|
| 1301 |
+
"step": 9050
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.2224759857593874,
|
| 1305 |
+
"grad_norm": 0.9247407913208008,
|
| 1306 |
+
"learning_rate": 1.7339777347584896e-05,
|
| 1307 |
+
"loss": 1.8789,
|
| 1308 |
+
"step": 9100
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 1.229193255860818,
|
| 1312 |
+
"grad_norm": 1.0588265657424927,
|
| 1313 |
+
"learning_rate": 1.708138217456088e-05,
|
| 1314 |
+
"loss": 1.8627,
|
| 1315 |
+
"step": 9150
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.235910525962249,
|
| 1319 |
+
"grad_norm": 0.8769893646240234,
|
| 1320 |
+
"learning_rate": 1.6823923895238303e-05,
|
| 1321 |
+
"loss": 1.8663,
|
| 1322 |
+
"step": 9200
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 1.2426277960636798,
|
| 1326 |
+
"grad_norm": 0.9970062375068665,
|
| 1327 |
+
"learning_rate": 1.6567432970872587e-05,
|
| 1328 |
+
"loss": 1.8632,
|
| 1329 |
+
"step": 9250
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 1.2493450661651104,
|
| 1333 |
+
"grad_norm": 0.9387636780738831,
|
| 1334 |
+
"learning_rate": 1.6311939748266282e-05,
|
| 1335 |
+
"loss": 1.8658,
|
| 1336 |
+
"step": 9300
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 1.2560623362665413,
|
| 1340 |
+
"grad_norm": 0.9504996538162231,
|
| 1341 |
+
"learning_rate": 1.605747445617851e-05,
|
| 1342 |
+
"loss": 1.8655,
|
| 1343 |
+
"step": 9350
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.262779606367972,
|
| 1347 |
+
"grad_norm": 0.9407314658164978,
|
| 1348 |
+
"learning_rate": 1.5804067201748526e-05,
|
| 1349 |
+
"loss": 1.863,
|
| 1350 |
+
"step": 9400
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.2694968764694028,
|
| 1354 |
+
"grad_norm": 0.8854203224182129,
|
| 1355 |
+
"learning_rate": 1.55517479669335e-05,
|
| 1356 |
+
"loss": 1.8636,
|
| 1357 |
+
"step": 9450
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 1.2762141465708337,
|
| 1361 |
+
"grad_norm": 0.9684790372848511,
|
| 1362 |
+
"learning_rate": 1.530054660496125e-05,
|
| 1363 |
+
"loss": 1.8571,
|
| 1364 |
+
"step": 9500
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 1.2829314166722643,
|
| 1368 |
+
"grad_norm": 0.896827757358551,
|
| 1369 |
+
"learning_rate": 1.5050492836798091e-05,
|
| 1370 |
+
"loss": 1.8549,
|
| 1371 |
+
"step": 9550
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 1.2896486867736952,
|
| 1375 |
+
"grad_norm": 0.8551830649375916,
|
| 1376 |
+
"learning_rate": 1.4801616247632455e-05,
|
| 1377 |
+
"loss": 1.8664,
|
| 1378 |
+
"step": 9600
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 1.296365956875126,
|
| 1382 |
+
"grad_norm": 0.911526620388031,
|
| 1383 |
+
"learning_rate": 1.4553946283374475e-05,
|
| 1384 |
+
"loss": 1.8639,
|
| 1385 |
+
"step": 9650
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.3030832269765567,
|
| 1389 |
+
"grad_norm": 0.8315178155899048,
|
| 1390 |
+
"learning_rate": 1.4307512247172077e-05,
|
| 1391 |
+
"loss": 1.854,
|
| 1392 |
+
"step": 9700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 1.3098004970779875,
|
| 1396 |
+
"grad_norm": 0.8541605472564697,
|
| 1397 |
+
"learning_rate": 1.4062343295943998e-05,
|
| 1398 |
+
"loss": 1.8609,
|
| 1399 |
+
"step": 9750
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 1.3165177671794184,
|
| 1403 |
+
"grad_norm": 0.8474893569946289,
|
| 1404 |
+
"learning_rate": 1.381846843693002e-05,
|
| 1405 |
+
"loss": 1.8603,
|
| 1406 |
+
"step": 9800
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.323235037280849,
|
| 1410 |
+
"grad_norm": 0.8238804340362549,
|
| 1411 |
+
"learning_rate": 1.357591652425904e-05,
|
| 1412 |
+
"loss": 1.8509,
|
| 1413 |
+
"step": 9850
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.3299523073822799,
|
| 1417 |
+
"grad_norm": 0.8796683549880981,
|
| 1418 |
+
"learning_rate": 1.3334716255535146e-05,
|
| 1419 |
+
"loss": 1.8595,
|
| 1420 |
+
"step": 9900
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.3331765970309666,
|
| 1424 |
+
"eval_loss": 1.7995295524597168,
|
| 1425 |
+
"eval_runtime": 1651.9195,
|
| 1426 |
+
"eval_samples_per_second": 72.285,
|
| 1427 |
+
"eval_steps_per_second": 9.036,
|
| 1428 |
+
"step": 9924
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 1.3366695774837107,
|
| 1432 |
+
"grad_norm": 0.8274658918380737,
|
| 1433 |
+
"learning_rate": 1.309489616844225e-05,
|
| 1434 |
+
"loss": 1.8547,
|
| 1435 |
+
"step": 9950
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 1.3433868475851414,
|
| 1439 |
+
"grad_norm": 1.110520362854004,
|
| 1440 |
+
"learning_rate": 1.2856484637367655e-05,
|
| 1441 |
+
"loss": 1.8543,
|
| 1442 |
+
"step": 10000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 1.3501041176865722,
|
| 1446 |
+
"grad_norm": 0.991307258605957,
|
| 1447 |
+
"learning_rate": 1.2619509870044926e-05,
|
| 1448 |
+
"loss": 1.8542,
|
| 1449 |
+
"step": 10050
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 1.3568213877880029,
|
| 1453 |
+
"grad_norm": 1.0729873180389404,
|
| 1454 |
+
"learning_rate": 1.2383999904216485e-05,
|
| 1455 |
+
"loss": 1.855,
|
| 1456 |
+
"step": 10100
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 1.3635386578894337,
|
| 1460 |
+
"grad_norm": 0.9162331819534302,
|
| 1461 |
+
"learning_rate": 1.2149982604316311e-05,
|
| 1462 |
+
"loss": 1.862,
|
| 1463 |
+
"step": 10150
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 1.3702559279908644,
|
| 1467 |
+
"grad_norm": 1.0340936183929443,
|
| 1468 |
+
"learning_rate": 1.1917485658173145e-05,
|
| 1469 |
+
"loss": 1.845,
|
| 1470 |
+
"step": 10200
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 1.3769731980922952,
|
| 1474 |
+
"grad_norm": 0.9109688401222229,
|
| 1475 |
+
"learning_rate": 1.1686536573734625e-05,
|
| 1476 |
+
"loss": 1.854,
|
| 1477 |
+
"step": 10250
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 1.383690468193726,
|
| 1481 |
+
"grad_norm": 0.901867151260376,
|
| 1482 |
+
"learning_rate": 1.1457162675812647e-05,
|
| 1483 |
+
"loss": 1.8607,
|
| 1484 |
+
"step": 10300
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 1.3904077382951567,
|
| 1488 |
+
"grad_norm": 0.9229015111923218,
|
| 1489 |
+
"learning_rate": 1.1229391102850428e-05,
|
| 1490 |
+
"loss": 1.8508,
|
| 1491 |
+
"step": 10350
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1.3971250083965876,
|
| 1495 |
+
"grad_norm": 0.8118278384208679,
|
| 1496 |
+
"learning_rate": 1.1003248803711625e-05,
|
| 1497 |
+
"loss": 1.8478,
|
| 1498 |
+
"step": 10400
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 1.4038422784980185,
|
| 1502 |
+
"grad_norm": 0.8655158281326294,
|
| 1503 |
+
"learning_rate": 1.0778762534491849e-05,
|
| 1504 |
+
"loss": 1.8583,
|
| 1505 |
+
"step": 10450
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.410559548599449,
|
| 1509 |
+
"grad_norm": 0.8829045295715332,
|
| 1510 |
+
"learning_rate": 1.0555958855353029e-05,
|
| 1511 |
+
"loss": 1.8536,
|
| 1512 |
+
"step": 10500
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 1.41727681870088,
|
| 1516 |
+
"grad_norm": 0.8320277333259583,
|
| 1517 |
+
"learning_rate": 1.0334864127380931e-05,
|
| 1518 |
+
"loss": 1.8546,
|
| 1519 |
+
"step": 10550
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 1.4239940888023108,
|
| 1523 |
+
"grad_norm": 0.8472936749458313,
|
| 1524 |
+
"learning_rate": 1.0115504509466244e-05,
|
| 1525 |
+
"loss": 1.8528,
|
| 1526 |
+
"step": 10600
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 1.4307113589037415,
|
| 1530 |
+
"grad_norm": 0.9551612734794617,
|
| 1531 |
+
"learning_rate": 9.89790595520956e-06,
|
| 1532 |
+
"loss": 1.846,
|
| 1533 |
+
"step": 10650
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1.4374286290051723,
|
| 1537 |
+
"grad_norm": 0.8682100772857666,
|
| 1538 |
+
"learning_rate": 9.6820942098507e-06,
|
| 1539 |
+
"loss": 1.8541,
|
| 1540 |
+
"step": 10700
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.4441458991066032,
|
| 1544 |
+
"grad_norm": 0.7923955321311951,
|
| 1545 |
+
"learning_rate": 9.468094807222633e-06,
|
| 1546 |
+
"loss": 1.8502,
|
| 1547 |
+
"step": 10750
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1.4508631692080338,
|
| 1551 |
+
"grad_norm": 0.7568480968475342,
|
| 1552 |
+
"learning_rate": 9.255933066730449e-06,
|
| 1553 |
+
"loss": 1.8622,
|
| 1554 |
+
"step": 10800
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.4575804393094647,
|
| 1558 |
+
"grad_norm": 0.9057785868644714,
|
| 1559 |
+
"learning_rate": 9.045634090355667e-06,
|
| 1560 |
+
"loss": 1.852,
|
| 1561 |
+
"step": 10850
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 1.4642977094108955,
|
| 1565 |
+
"grad_norm": 0.8364023566246033,
|
| 1566 |
+
"learning_rate": 8.837222759686306e-06,
|
| 1567 |
+
"loss": 1.8495,
|
| 1568 |
+
"step": 10900
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 1.4710149795123262,
|
| 1572 |
+
"grad_norm": 0.9518344402313232,
|
| 1573 |
+
"learning_rate": 8.630723732972998e-06,
|
| 1574 |
+
"loss": 1.8545,
|
| 1575 |
+
"step": 10950
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 1.477732249613757,
|
| 1579 |
+
"grad_norm": 0.8617038130760193,
|
| 1580 |
+
"learning_rate": 8.426161442211552e-06,
|
| 1581 |
+
"loss": 1.8491,
|
| 1582 |
+
"step": 11000
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 1.4844495197151877,
|
| 1586 |
+
"grad_norm": 0.9056522846221924,
|
| 1587 |
+
"learning_rate": 8.22356009025225e-06,
|
| 1588 |
+
"loss": 1.8539,
|
| 1589 |
+
"step": 11050
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 1.4911667898166185,
|
| 1593 |
+
"grad_norm": 0.7397828102111816,
|
| 1594 |
+
"learning_rate": 8.022943647936315e-06,
|
| 1595 |
+
"loss": 1.8494,
|
| 1596 |
+
"step": 11100
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 1.4978840599180492,
|
| 1600 |
+
"grad_norm": 0.7917934656143188,
|
| 1601 |
+
"learning_rate": 7.82433585125977e-06,
|
| 1602 |
+
"loss": 1.8504,
|
| 1603 |
+
"step": 11150
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 1.50460133001948,
|
| 1607 |
+
"grad_norm": 0.8637117147445679,
|
| 1608 |
+
"learning_rate": 7.627760198565112e-06,
|
| 1609 |
+
"loss": 1.8502,
|
| 1610 |
+
"step": 11200
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 1.511318600120911,
|
| 1614 |
+
"grad_norm": 1.025888204574585,
|
| 1615 |
+
"learning_rate": 7.433239947761095e-06,
|
| 1616 |
+
"loss": 1.8503,
|
| 1617 |
+
"step": 11250
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.5180358702223415,
|
| 1621 |
+
"grad_norm": 0.7819210886955261,
|
| 1622 |
+
"learning_rate": 7.2407981135709735e-06,
|
| 1623 |
+
"loss": 1.8465,
|
| 1624 |
+
"step": 11300
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 1.5247531403237724,
|
| 1628 |
+
"grad_norm": 0.9122748374938965,
|
| 1629 |
+
"learning_rate": 7.050457464809495e-06,
|
| 1630 |
+
"loss": 1.8467,
|
| 1631 |
+
"step": 11350
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 1.5314704104252033,
|
| 1635 |
+
"grad_norm": 0.915520429611206,
|
| 1636 |
+
"learning_rate": 6.862240521689011e-06,
|
| 1637 |
+
"loss": 1.8435,
|
| 1638 |
+
"step": 11400
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 1.538187680526634,
|
| 1642 |
+
"grad_norm": 0.7819466590881348,
|
| 1643 |
+
"learning_rate": 6.676169553154993e-06,
|
| 1644 |
+
"loss": 1.8476,
|
| 1645 |
+
"step": 11450
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.5449049506280648,
|
| 1649 |
+
"grad_norm": 0.9129722714424133,
|
| 1650 |
+
"learning_rate": 6.492266574251249e-06,
|
| 1651 |
+
"loss": 1.8534,
|
| 1652 |
+
"step": 11500
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 1.5516222207294956,
|
| 1656 |
+
"grad_norm": 0.834613561630249,
|
| 1657 |
+
"learning_rate": 6.310553343515249e-06,
|
| 1658 |
+
"loss": 1.8522,
|
| 1659 |
+
"step": 11550
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.5583394908309263,
|
| 1663 |
+
"grad_norm": 0.8443191647529602,
|
| 1664 |
+
"learning_rate": 6.131051360403731e-06,
|
| 1665 |
+
"loss": 1.844,
|
| 1666 |
+
"step": 11600
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 1.5650567609323571,
|
| 1670 |
+
"grad_norm": 0.858828604221344,
|
| 1671 |
+
"learning_rate": 5.953781862748983e-06,
|
| 1672 |
+
"loss": 1.8446,
|
| 1673 |
+
"step": 11650
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 1.571774031033788,
|
| 1677 |
+
"grad_norm": 0.7843521237373352,
|
| 1678 |
+
"learning_rate": 5.778765824246099e-06,
|
| 1679 |
+
"loss": 1.8465,
|
| 1680 |
+
"step": 11700
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 1.5784913011352186,
|
| 1684 |
+
"grad_norm": 0.9034683704376221,
|
| 1685 |
+
"learning_rate": 5.6060239519714565e-06,
|
| 1686 |
+
"loss": 1.8485,
|
| 1687 |
+
"step": 11750
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 1.5852085712366493,
|
| 1691 |
+
"grad_norm": 0.8224745392799377,
|
| 1692 |
+
"learning_rate": 5.435576683932758e-06,
|
| 1693 |
+
"loss": 1.844,
|
| 1694 |
+
"step": 11800
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 1.5919258413380803,
|
| 1698 |
+
"grad_norm": 0.8487153649330139,
|
| 1699 |
+
"learning_rate": 5.267444186650908e-06,
|
| 1700 |
+
"loss": 1.8399,
|
| 1701 |
+
"step": 11850
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 1.598643111439511,
|
| 1705 |
+
"grad_norm": 0.9923579692840576,
|
| 1706 |
+
"learning_rate": 5.101646352773973e-06,
|
| 1707 |
+
"loss": 1.8438,
|
| 1708 |
+
"step": 11900
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 1.6053603815409416,
|
| 1712 |
+
"grad_norm": 1.0088372230529785,
|
| 1713 |
+
"learning_rate": 4.938202798723632e-06,
|
| 1714 |
+
"loss": 1.8533,
|
| 1715 |
+
"step": 11950
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 1.6120776516423725,
|
| 1719 |
+
"grad_norm": 0.9827088713645935,
|
| 1720 |
+
"learning_rate": 4.777132862374201e-06,
|
| 1721 |
+
"loss": 1.8482,
|
| 1722 |
+
"step": 12000
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 1.6187949217438034,
|
| 1726 |
+
"grad_norm": 0.7875204086303711,
|
| 1727 |
+
"learning_rate": 4.618455600764701e-06,
|
| 1728 |
+
"loss": 1.846,
|
| 1729 |
+
"step": 12050
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 1.625512191845234,
|
| 1733 |
+
"grad_norm": 0.8284263610839844,
|
| 1734 |
+
"learning_rate": 4.462189787844101e-06,
|
| 1735 |
+
"loss": 1.8462,
|
| 1736 |
+
"step": 12100
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 1.6322294619466649,
|
| 1740 |
+
"grad_norm": 0.9832016229629517,
|
| 1741 |
+
"learning_rate": 4.308353912250077e-06,
|
| 1742 |
+
"loss": 1.8467,
|
| 1743 |
+
"step": 12150
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 1.6389467320480957,
|
| 1747 |
+
"grad_norm": 0.8272432088851929,
|
| 1748 |
+
"learning_rate": 4.156966175121524e-06,
|
| 1749 |
+
"loss": 1.8507,
|
| 1750 |
+
"step": 12200
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 1.6456640021495264,
|
| 1754 |
+
"grad_norm": 0.8643631935119629,
|
| 1755 |
+
"learning_rate": 4.008044487945087e-06,
|
| 1756 |
+
"loss": 1.851,
|
| 1757 |
+
"step": 12250
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 1.6523812722509572,
|
| 1761 |
+
"grad_norm": 0.821950376033783,
|
| 1762 |
+
"learning_rate": 3.861606470435939e-06,
|
| 1763 |
+
"loss": 1.8422,
|
| 1764 |
+
"step": 12300
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 1.659098542352388,
|
| 1768 |
+
"grad_norm": 0.8796666860580444,
|
| 1769 |
+
"learning_rate": 3.717669448453126e-06,
|
| 1770 |
+
"loss": 1.8456,
|
| 1771 |
+
"step": 12350
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 1.6658158124538187,
|
| 1775 |
+
"grad_norm": 0.8530693650245667,
|
| 1776 |
+
"learning_rate": 3.5762504519496365e-06,
|
| 1777 |
+
"loss": 1.8408,
|
| 1778 |
+
"step": 12400
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 1.666487539463962,
|
| 1782 |
+
"eval_loss": 1.7879396677017212,
|
| 1783 |
+
"eval_runtime": 1652.1563,
|
| 1784 |
+
"eval_samples_per_second": 72.275,
|
| 1785 |
+
"eval_steps_per_second": 9.035,
|
| 1786 |
+
"step": 12405
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 1.6725330825552496,
|
| 1790 |
+
"grad_norm": 0.836791455745697,
|
| 1791 |
+
"learning_rate": 3.437366212957502e-06,
|
| 1792 |
+
"loss": 1.8465,
|
| 1793 |
+
"step": 12450
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 1.6792503526566804,
|
| 1797 |
+
"grad_norm": 0.7426266670227051,
|
| 1798 |
+
"learning_rate": 3.3010331636081387e-06,
|
| 1799 |
+
"loss": 1.8475,
|
| 1800 |
+
"step": 12500
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 1.685967622758111,
|
| 1804 |
+
"grad_norm": 0.8120975494384766,
|
| 1805 |
+
"learning_rate": 3.167267434188173e-06,
|
| 1806 |
+
"loss": 1.847,
|
| 1807 |
+
"step": 12550
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 1.692684892859542,
|
| 1811 |
+
"grad_norm": 0.7585613131523132,
|
| 1812 |
+
"learning_rate": 3.0360848512309887e-06,
|
| 1813 |
+
"loss": 1.8452,
|
| 1814 |
+
"step": 12600
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 1.6994021629609728,
|
| 1818 |
+
"grad_norm": 0.8330528736114502,
|
| 1819 |
+
"learning_rate": 2.907500935644203e-06,
|
| 1820 |
+
"loss": 1.8455,
|
| 1821 |
+
"step": 12650
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.7061194330624034,
|
| 1825 |
+
"grad_norm": 0.998671293258667,
|
| 1826 |
+
"learning_rate": 2.781530900873305e-06,
|
| 1827 |
+
"loss": 1.8499,
|
| 1828 |
+
"step": 12700
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 1.712836703163834,
|
| 1832 |
+
"grad_norm": 0.8430862426757812,
|
| 1833 |
+
"learning_rate": 2.6581896511016614e-06,
|
| 1834 |
+
"loss": 1.8425,
|
| 1835 |
+
"step": 12750
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 1.7195539732652652,
|
| 1839 |
+
"grad_norm": 0.9203202724456787,
|
| 1840 |
+
"learning_rate": 2.537491779487147e-06,
|
| 1841 |
+
"loss": 1.8374,
|
| 1842 |
+
"step": 12800
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 1.7262712433666958,
|
| 1846 |
+
"grad_norm": 1.0353131294250488,
|
| 1847 |
+
"learning_rate": 2.419451566435532e-06,
|
| 1848 |
+
"loss": 1.8439,
|
| 1849 |
+
"step": 12850
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 1.7329885134681264,
|
| 1853 |
+
"grad_norm": 0.8823174834251404,
|
| 1854 |
+
"learning_rate": 2.3040829779108985e-06,
|
| 1855 |
+
"loss": 1.8408,
|
| 1856 |
+
"step": 12900
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 1.7397057835695573,
|
| 1860 |
+
"grad_norm": 0.8599026799201965,
|
| 1861 |
+
"learning_rate": 2.19139966378325e-06,
|
| 1862 |
+
"loss": 1.8518,
|
| 1863 |
+
"step": 12950
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 1.7464230536709882,
|
| 1867 |
+
"grad_norm": 0.8370893597602844,
|
| 1868 |
+
"learning_rate": 2.081414956213526e-06,
|
| 1869 |
+
"loss": 1.8489,
|
| 1870 |
+
"step": 13000
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 1.7531403237724188,
|
| 1874 |
+
"grad_norm": 0.7799704670906067,
|
| 1875 |
+
"learning_rate": 1.9741418680762013e-06,
|
| 1876 |
+
"loss": 1.845,
|
| 1877 |
+
"step": 13050
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 1.7598575938738497,
|
| 1881 |
+
"grad_norm": 0.8190951943397522,
|
| 1882 |
+
"learning_rate": 1.8695930914196664e-06,
|
| 1883 |
+
"loss": 1.8458,
|
| 1884 |
+
"step": 13100
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 1.7665748639752805,
|
| 1888 |
+
"grad_norm": 0.8451455235481262,
|
| 1889 |
+
"learning_rate": 1.7677809959645548e-06,
|
| 1890 |
+
"loss": 1.8446,
|
| 1891 |
+
"step": 13150
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 1.7732921340767112,
|
| 1895 |
+
"grad_norm": 0.8454949855804443,
|
| 1896 |
+
"learning_rate": 1.6687176276402261e-06,
|
| 1897 |
+
"loss": 1.8373,
|
| 1898 |
+
"step": 13200
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 1.780009404178142,
|
| 1902 |
+
"grad_norm": 0.9424586296081543,
|
| 1903 |
+
"learning_rate": 1.572414707159553e-06,
|
| 1904 |
+
"loss": 1.8441,
|
| 1905 |
+
"step": 13250
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.7867266742795729,
|
| 1909 |
+
"grad_norm": 0.8610557913780212,
|
| 1910 |
+
"learning_rate": 1.4788836286321606e-06,
|
| 1911 |
+
"loss": 1.8398,
|
| 1912 |
+
"step": 13300
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 1.7934439443810035,
|
| 1916 |
+
"grad_norm": 0.9130797982215881,
|
| 1917 |
+
"learning_rate": 1.3881354582163525e-06,
|
| 1918 |
+
"loss": 1.8398,
|
| 1919 |
+
"step": 13350
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 1.8001612144824344,
|
| 1923 |
+
"grad_norm": 0.9039607048034668,
|
| 1924 |
+
"learning_rate": 1.3001809328097914e-06,
|
| 1925 |
+
"loss": 1.8472,
|
| 1926 |
+
"step": 13400
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 1.8068784845838652,
|
| 1930 |
+
"grad_norm": 0.8576995730400085,
|
| 1931 |
+
"learning_rate": 1.2150304587791873e-06,
|
| 1932 |
+
"loss": 1.8481,
|
| 1933 |
+
"step": 13450
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 1.8135957546852959,
|
| 1937 |
+
"grad_norm": 0.9080734252929688,
|
| 1938 |
+
"learning_rate": 1.1326941107290351e-06,
|
| 1939 |
+
"loss": 1.8402,
|
| 1940 |
+
"step": 13500
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 1.8203130247867265,
|
| 1944 |
+
"grad_norm": 0.8428635597229004,
|
| 1945 |
+
"learning_rate": 1.053181630309666e-06,
|
| 1946 |
+
"loss": 1.846,
|
| 1947 |
+
"step": 13550
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.8270302948881576,
|
| 1951 |
+
"grad_norm": 0.890152633190155,
|
| 1952 |
+
"learning_rate": 9.765024250646238e-07,
|
| 1953 |
+
"loss": 1.8529,
|
| 1954 |
+
"step": 13600
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 1.8337475649895882,
|
| 1958 |
+
"grad_norm": 0.803615391254425,
|
| 1959 |
+
"learning_rate": 9.026655673176454e-07,
|
| 1960 |
+
"loss": 1.8444,
|
| 1961 |
+
"step": 13650
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 1.8404648350910189,
|
| 1965 |
+
"grad_norm": 0.8416168093681335,
|
| 1966 |
+
"learning_rate": 8.316797930992465e-07,
|
| 1967 |
+
"loss": 1.8413,
|
| 1968 |
+
"step": 13700
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 1.8471821051924497,
|
| 1972 |
+
"grad_norm": 0.9183295369148254,
|
| 1973 |
+
"learning_rate": 7.635535011131178e-07,
|
| 1974 |
+
"loss": 1.8438,
|
| 1975 |
+
"step": 13750
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 1.8538993752938806,
|
| 1979 |
+
"grad_norm": 0.8079262375831604,
|
| 1980 |
+
"learning_rate": 6.982947517424315e-07,
|
| 1981 |
+
"loss": 1.8446,
|
| 1982 |
+
"step": 13800
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 1.8606166453953112,
|
| 1986 |
+
"grad_norm": 0.7838028073310852,
|
| 1987 |
+
"learning_rate": 6.35911266096173e-07,
|
| 1988 |
+
"loss": 1.845,
|
| 1989 |
+
"step": 13850
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 1.867333915496742,
|
| 1993 |
+
"grad_norm": 0.8391366004943848,
|
| 1994 |
+
"learning_rate": 5.764104250956165e-07,
|
| 1995 |
+
"loss": 1.8483,
|
| 1996 |
+
"step": 13900
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 1.874051185598173,
|
| 2000 |
+
"grad_norm": 0.9014426469802856,
|
| 2001 |
+
"learning_rate": 5.197992686010511e-07,
|
| 2002 |
+
"loss": 1.8438,
|
| 2003 |
+
"step": 13950
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 1.8807684556996036,
|
| 2007 |
+
"grad_norm": 0.9228802919387817,
|
| 2008 |
+
"learning_rate": 4.660844945788501e-07,
|
| 2009 |
+
"loss": 1.8474,
|
| 2010 |
+
"step": 14000
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 1.8874857258010345,
|
| 2014 |
+
"grad_norm": 0.7464848756790161,
|
| 2015 |
+
"learning_rate": 4.1527245830901563e-07,
|
| 2016 |
+
"loss": 1.8436,
|
| 2017 |
+
"step": 14050
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 1.8942029959024653,
|
| 2021 |
+
"grad_norm": 0.771722674369812,
|
| 2022 |
+
"learning_rate": 3.6736917163322505e-07,
|
| 2023 |
+
"loss": 1.8438,
|
| 2024 |
+
"step": 14100
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 1.900920266003896,
|
| 2028 |
+
"grad_norm": 0.7891733050346375,
|
| 2029 |
+
"learning_rate": 3.2238030224356897e-07,
|
| 2030 |
+
"loss": 1.8423,
|
| 2031 |
+
"step": 14150
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 1.9076375361053268,
|
| 2035 |
+
"grad_norm": 0.8450707197189331,
|
| 2036 |
+
"learning_rate": 2.803111730119545e-07,
|
| 2037 |
+
"loss": 1.843,
|
| 2038 |
+
"step": 14200
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 1.9143548062067577,
|
| 2042 |
+
"grad_norm": 0.9002699255943298,
|
| 2043 |
+
"learning_rate": 2.4116676136033135e-07,
|
| 2044 |
+
"loss": 1.8433,
|
| 2045 |
+
"step": 14250
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 1.9210720763081883,
|
| 2049 |
+
"grad_norm": 0.8603649735450745,
|
| 2050 |
+
"learning_rate": 2.049516986717931e-07,
|
| 2051 |
+
"loss": 1.8396,
|
| 2052 |
+
"step": 14300
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 1.927789346409619,
|
| 2056 |
+
"grad_norm": 0.8522045612335205,
|
| 2057 |
+
"learning_rate": 1.7167026974261313e-07,
|
| 2058 |
+
"loss": 1.8498,
|
| 2059 |
+
"step": 14350
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 1.93450661651105,
|
| 2063 |
+
"grad_norm": 0.8563119769096375,
|
| 2064 |
+
"learning_rate": 1.4132641227528054e-07,
|
| 2065 |
+
"loss": 1.841,
|
| 2066 |
+
"step": 14400
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 1.9412238866124807,
|
| 2070 |
+
"grad_norm": 0.8482009172439575,
|
| 2071 |
+
"learning_rate": 1.1392371641262001e-07,
|
| 2072 |
+
"loss": 1.8502,
|
| 2073 |
+
"step": 14450
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 1.9479411567139113,
|
| 2077 |
+
"grad_norm": 0.8035847544670105,
|
| 2078 |
+
"learning_rate": 8.946542431300942e-08,
|
| 2079 |
+
"loss": 1.8467,
|
| 2080 |
+
"step": 14500
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 1.9546584268153424,
|
| 2084 |
+
"grad_norm": 0.8062044382095337,
|
| 2085 |
+
"learning_rate": 6.795442976679501e-08,
|
| 2086 |
+
"loss": 1.8438,
|
| 2087 |
+
"step": 14550
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 1.961375696916773,
|
| 2091 |
+
"grad_norm": 0.7858136296272278,
|
| 2092 |
+
"learning_rate": 4.939327785390691e-08,
|
| 2093 |
+
"loss": 1.8431,
|
| 2094 |
+
"step": 14600
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 1.9680929670182037,
|
| 2098 |
+
"grad_norm": 0.7691812515258789,
|
| 2099 |
+
"learning_rate": 3.37841646427417e-08,
|
| 2100 |
+
"loss": 1.8442,
|
| 2101 |
+
"step": 14650
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 1.9748102371196345,
|
| 2105 |
+
"grad_norm": 0.8699848055839539,
|
| 2106 |
+
"learning_rate": 2.1128936930320254e-08,
|
| 2107 |
+
"loss": 1.8388,
|
| 2108 |
+
"step": 14700
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 1.9815275072210654,
|
| 2112 |
+
"grad_norm": 0.7812600135803223,
|
| 2113 |
+
"learning_rate": 1.142909202380138e-08,
|
| 2114 |
+
"loss": 1.8403,
|
| 2115 |
+
"step": 14750
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 1.988244777322496,
|
| 2119 |
+
"grad_norm": 0.856472373008728,
|
| 2120 |
+
"learning_rate": 4.6857775633152305e-09,
|
| 2121 |
+
"loss": 1.837,
|
| 2122 |
+
"step": 14800
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 1.994962047423927,
|
| 2126 |
+
"grad_norm": 0.7702723741531372,
|
| 2127 |
+
"learning_rate": 8.997913861857888e-10,
|
| 2128 |
+
"loss": 1.8449,
|
| 2129 |
+
"step": 14850
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 1.9997984818969572,
|
| 2133 |
+
"eval_loss": 1.786035418510437,
|
| 2134 |
+
"eval_runtime": 1652.2214,
|
| 2135 |
+
"eval_samples_per_second": 72.272,
|
| 2136 |
+
"eval_steps_per_second": 9.035,
|
| 2137 |
+
"step": 14886
|
| 2138 |
+
}
|
| 2139 |
+
],
|
| 2140 |
+
"logging_steps": 50,
|
| 2141 |
+
"max_steps": 14888,
|
| 2142 |
+
"num_input_tokens_seen": 0,
|
| 2143 |
+
"num_train_epochs": 2,
|
| 2144 |
+
"save_steps": 2481,
|
| 2145 |
+
"stateful_callbacks": {
|
| 2146 |
+
"TrainerControl": {
|
| 2147 |
+
"args": {
|
| 2148 |
+
"should_epoch_stop": false,
|
| 2149 |
+
"should_evaluate": false,
|
| 2150 |
+
"should_log": false,
|
| 2151 |
+
"should_save": true,
|
| 2152 |
+
"should_training_stop": true
|
| 2153 |
+
},
|
| 2154 |
+
"attributes": {}
|
| 2155 |
+
}
|
| 2156 |
+
},
|
| 2157 |
+
"total_flos": 3.539345213296214e+18,
|
| 2158 |
+
"train_batch_size": 2,
|
| 2159 |
+
"trial_name": null,
|
| 2160 |
+
"trial_params": null
|
| 2161 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|