Add files using upload-large-folder tool
Browse files- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/special_tokens_map.json +51 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/tokenizer.json +0 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/tokenizer_config.json +55 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/trainer_state.json +595 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/special_tokens_map.json +51 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/tokenizer.json +0 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/tokenizer_config.json +55 -0
- uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/trainer_state.json +929 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/special_tokens_map.json +51 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/tokenizer.json +0 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/tokenizer_config.json +55 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/trainer_state.json +486 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/special_tokens_map.json +51 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/tokenizer.json +0 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/tokenizer_config.json +55 -0
- uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/trainer_state.json +820 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/special_tokens_map.json +51 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/tokenizer.json +0 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/tokenizer_config.json +55 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/trainer_state.json +370 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/special_tokens_map.json +51 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/tokenizer.json +0 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/tokenizer_config.json +55 -0
- uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/trainer_state.json +704 -0
- viclickbait2025-bamibert-s42-20260827T130938Z-ee19aa8e/trainer/checkpoint-150/special_tokens_map.json +51 -0
- viclickbait2025-bamibert-s42-20260827T130938Z-ee19aa8e/trainer/checkpoint-150/tokenizer_config.json +55 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/special_tokens_map.json +51 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/tokenizer.json +0 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/tokenizer_config.json +55 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/trainer_state.json +1590 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/special_tokens_map.json +51 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/tokenizer.json +0 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/tokenizer_config.json +55 -0
- vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/trainer_state.json +2253 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/special_tokens_map.json +51 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/tokenizer.json +0 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/tokenizer_config.json +55 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/trainer_state.json +1148 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/special_tokens_map.json +51 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/tokenizer.json +0 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/tokenizer_config.json +55 -0
- vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/trainer_state.json +1811 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/special_tokens_map.json +51 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/tokenizer.json +0 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/tokenizer_config.json +55 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/trainer_state.json +927 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/special_tokens_map.json +51 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/tokenizer.json +0 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/tokenizer_config.json +55 -0
- vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/trainer_state.json +1590 -0
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790/trainer_state.json
ADDED
|
@@ -0,0 +1,595 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1790,
|
| 3 |
+
"best_metric": 0.8535013300753359,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1790,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 22.502758026123047,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.0952,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 51.70573806762695,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 3.8986,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 31.190452575683594,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.6027,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 19.23320960998535,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4006,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 20.846994400024414,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 3.2288,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 39.98775100708008,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 3.1301,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 43.641353607177734,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 2.934,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 44.367496490478516,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 2.8103,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 51.65887451171875,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 2.0396,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 95.81674194335938,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.8671,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 38.34164047241211,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.7479,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 74.85626220703125,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.7033,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 45.92241668701172,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.6668,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 73.18209838867188,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.3044,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9121920404295641,
|
| 113 |
+
"eval_loss": 0.31590455770492554,
|
| 114 |
+
"eval_macro_f1": 0.6220951388509554,
|
| 115 |
+
"eval_runtime": 1.0649,
|
| 116 |
+
"eval_samples_per_second": 1486.534,
|
| 117 |
+
"eval_steps_per_second": 92.967,
|
| 118 |
+
"eval_weighted_f1": 0.890631832899203,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 81.14189910888672,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 1.1692,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 78.62606048583984,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 1.1165,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 72.6214828491211,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 1.2139,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 59.12213897705078,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 0.9952,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 24.40809440612793,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 1.2224,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 31.012418746948242,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 1.1047,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 58.590389251708984,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 1.2207,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 9.626352310180664,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 1.1276,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 11.062261581420898,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 1.0308,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 40.633460998535156,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.969,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 20.63018798828125,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 1.1539,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 31.195363998413086,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.9453,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 16.894248962402344,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.992,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 23.10975456237793,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 1.0665,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9330385344283006,
|
| 222 |
+
"eval_loss": 0.2521005868911743,
|
| 223 |
+
"eval_macro_f1": 0.7758143677827062,
|
| 224 |
+
"eval_runtime": 0.9105,
|
| 225 |
+
"eval_samples_per_second": 1738.662,
|
| 226 |
+
"eval_steps_per_second": 108.735,
|
| 227 |
+
"eval_weighted_f1": 0.9259481340778538,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 6.699418544769287,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.9025,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 8.981850624084473,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 1.0244,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 42.387176513671875,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.8874,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 33.133113861083984,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.9635,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 54.822845458984375,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.9101,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 34.682373046875,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.8393,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 16.338340759277344,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.6795,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 61.311893463134766,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.8481,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 42.22651672363281,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.8366,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 18.176904678344727,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.8754,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 29.46816062927246,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.8296,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 38.06634521484375,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.8433,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 15.29800796508789,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.7685,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 18.728879928588867,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.8457,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9380922299431459,
|
| 331 |
+
"eval_loss": 0.20598039031028748,
|
| 332 |
+
"eval_macro_f1": 0.8069122638572427,
|
| 333 |
+
"eval_runtime": 0.9741,
|
| 334 |
+
"eval_samples_per_second": 1625.139,
|
| 335 |
+
"eval_steps_per_second": 101.635,
|
| 336 |
+
"eval_weighted_f1": 0.9335173770294718,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 6.725589275360107,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.77,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 41.07616424560547,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.5628,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 15.49211597442627,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.666,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 48.635215759277344,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.8233,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 6.209376335144043,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.7932,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 2.8675124645233154,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.7105,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 6.533738613128662,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.613,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 51.26144027709961,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.6896,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 2.1346921920776367,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.5871,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 183.9330291748047,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.9003,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 16.037046432495117,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.5704,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 21.56624984741211,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.9196,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 44.50519943237305,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.716,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 18.84470558166504,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.8974,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 21.209745407104492,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.5805,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9450410612760581,
|
| 447 |
+
"eval_loss": 0.23549330234527588,
|
| 448 |
+
"eval_macro_f1": 0.8295937088063542,
|
| 449 |
+
"eval_runtime": 0.9283,
|
| 450 |
+
"eval_samples_per_second": 1705.209,
|
| 451 |
+
"eval_steps_per_second": 106.643,
|
| 452 |
+
"eval_weighted_f1": 0.9408894802580585,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 6.162760257720947,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.5816,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 1.8675271272659302,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.5444,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 10.477187156677246,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.676,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 95.41238403320312,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6629,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 14.249944686889648,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.582,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 30.288408279418945,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.5872,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 22.749391555786133,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.579,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 16.99974822998047,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.6561,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 47.46952438354492,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.5091,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 62.55239486694336,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.6304,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 36.24051284790039,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.5261,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 57.92010498046875,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.6513,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 1.1452276706695557,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.5149,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 7.032649993896484,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.6501,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9444093493367025,
|
| 556 |
+
"eval_loss": 0.2580535113811493,
|
| 557 |
+
"eval_macro_f1": 0.8535013300753359,
|
| 558 |
+
"eval_runtime": 1.2762,
|
| 559 |
+
"eval_samples_per_second": 1240.374,
|
| 560 |
+
"eval_steps_per_second": 77.572,
|
| 561 |
+
"eval_weighted_f1": 0.942839371797969,
|
| 562 |
+
"step": 1790
|
| 563 |
+
}
|
| 564 |
+
],
|
| 565 |
+
"logging_steps": 25,
|
| 566 |
+
"max_steps": 7160,
|
| 567 |
+
"num_input_tokens_seen": 0,
|
| 568 |
+
"num_train_epochs": 20,
|
| 569 |
+
"save_steps": 500,
|
| 570 |
+
"stateful_callbacks": {
|
| 571 |
+
"EarlyStoppingCallback": {
|
| 572 |
+
"args": {
|
| 573 |
+
"early_stopping_patience": 3,
|
| 574 |
+
"early_stopping_threshold": 0.0
|
| 575 |
+
},
|
| 576 |
+
"attributes": {
|
| 577 |
+
"early_stopping_patience_counter": 0
|
| 578 |
+
}
|
| 579 |
+
},
|
| 580 |
+
"TrainerControl": {
|
| 581 |
+
"args": {
|
| 582 |
+
"should_epoch_stop": false,
|
| 583 |
+
"should_evaluate": false,
|
| 584 |
+
"should_log": false,
|
| 585 |
+
"should_save": true,
|
| 586 |
+
"should_training_stop": false
|
| 587 |
+
},
|
| 588 |
+
"attributes": {}
|
| 589 |
+
}
|
| 590 |
+
},
|
| 591 |
+
"total_flos": 0.0,
|
| 592 |
+
"train_batch_size": 8,
|
| 593 |
+
"trial_name": null,
|
| 594 |
+
"trial_params": null
|
| 595 |
+
}
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-2864/trainer_state.json
ADDED
|
@@ -0,0 +1,929 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1790,
|
| 3 |
+
"best_metric": 0.8535013300753359,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-bamibert-s42-20260827T122234Z-f97f2036/trainer/checkpoint-1790",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2864,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 22.502758026123047,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.0952,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 51.70573806762695,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 3.8986,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 31.190452575683594,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.6027,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 19.23320960998535,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4006,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 20.846994400024414,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 3.2288,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 39.98775100708008,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 3.1301,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 43.641353607177734,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 2.934,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 44.367496490478516,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 2.8103,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 51.65887451171875,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 2.0396,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 95.81674194335938,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.8671,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 38.34164047241211,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.7479,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 74.85626220703125,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.7033,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 45.92241668701172,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.6668,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 73.18209838867188,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.3044,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9121920404295641,
|
| 113 |
+
"eval_loss": 0.31590455770492554,
|
| 114 |
+
"eval_macro_f1": 0.6220951388509554,
|
| 115 |
+
"eval_runtime": 1.0649,
|
| 116 |
+
"eval_samples_per_second": 1486.534,
|
| 117 |
+
"eval_steps_per_second": 92.967,
|
| 118 |
+
"eval_weighted_f1": 0.890631832899203,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 81.14189910888672,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 1.1692,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 78.62606048583984,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 1.1165,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 72.6214828491211,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 1.2139,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 59.12213897705078,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 0.9952,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 24.40809440612793,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 1.2224,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 31.012418746948242,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 1.1047,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 58.590389251708984,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 1.2207,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 9.626352310180664,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 1.1276,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 11.062261581420898,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 1.0308,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 40.633460998535156,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.969,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 20.63018798828125,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 1.1539,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 31.195363998413086,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.9453,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 16.894248962402344,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.992,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 23.10975456237793,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 1.0665,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9330385344283006,
|
| 222 |
+
"eval_loss": 0.2521005868911743,
|
| 223 |
+
"eval_macro_f1": 0.7758143677827062,
|
| 224 |
+
"eval_runtime": 0.9105,
|
| 225 |
+
"eval_samples_per_second": 1738.662,
|
| 226 |
+
"eval_steps_per_second": 108.735,
|
| 227 |
+
"eval_weighted_f1": 0.9259481340778538,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 6.699418544769287,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.9025,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 8.981850624084473,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 1.0244,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 42.387176513671875,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.8874,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 33.133113861083984,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.9635,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 54.822845458984375,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.9101,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 34.682373046875,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.8393,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 16.338340759277344,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.6795,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 61.311893463134766,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.8481,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 42.22651672363281,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.8366,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 18.176904678344727,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.8754,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 29.46816062927246,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.8296,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 38.06634521484375,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.8433,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 15.29800796508789,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.7685,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 18.728879928588867,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.8457,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9380922299431459,
|
| 331 |
+
"eval_loss": 0.20598039031028748,
|
| 332 |
+
"eval_macro_f1": 0.8069122638572427,
|
| 333 |
+
"eval_runtime": 0.9741,
|
| 334 |
+
"eval_samples_per_second": 1625.139,
|
| 335 |
+
"eval_steps_per_second": 101.635,
|
| 336 |
+
"eval_weighted_f1": 0.9335173770294718,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 6.725589275360107,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.77,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 41.07616424560547,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.5628,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 15.49211597442627,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.666,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 48.635215759277344,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.8233,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 6.209376335144043,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.7932,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 2.8675124645233154,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.7105,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 6.533738613128662,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.613,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 51.26144027709961,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.6896,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 2.1346921920776367,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.5871,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 183.9330291748047,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.9003,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 16.037046432495117,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.5704,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 21.56624984741211,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.9196,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 44.50519943237305,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.716,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 18.84470558166504,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.8974,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 21.209745407104492,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.5805,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9450410612760581,
|
| 447 |
+
"eval_loss": 0.23549330234527588,
|
| 448 |
+
"eval_macro_f1": 0.8295937088063542,
|
| 449 |
+
"eval_runtime": 0.9283,
|
| 450 |
+
"eval_samples_per_second": 1705.209,
|
| 451 |
+
"eval_steps_per_second": 106.643,
|
| 452 |
+
"eval_weighted_f1": 0.9408894802580585,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 6.162760257720947,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.5816,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 1.8675271272659302,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.5444,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 10.477187156677246,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.676,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 95.41238403320312,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6629,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 14.249944686889648,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.582,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 30.288408279418945,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.5872,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 22.749391555786133,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.579,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 16.99974822998047,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.6561,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 47.46952438354492,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.5091,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 62.55239486694336,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.6304,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 36.24051284790039,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.5261,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 57.92010498046875,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.6513,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 1.1452276706695557,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.5149,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 7.032649993896484,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.6501,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9444093493367025,
|
| 556 |
+
"eval_loss": 0.2580535113811493,
|
| 557 |
+
"eval_macro_f1": 0.8535013300753359,
|
| 558 |
+
"eval_runtime": 1.2762,
|
| 559 |
+
"eval_samples_per_second": 1240.374,
|
| 560 |
+
"eval_steps_per_second": 77.572,
|
| 561 |
+
"eval_weighted_f1": 0.942839371797969,
|
| 562 |
+
"step": 1790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.027991602519244,
|
| 566 |
+
"grad_norm": 7.7813310623168945,
|
| 567 |
+
"learning_rate": 1.6638733705772813e-05,
|
| 568 |
+
"loss": 0.8115,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.097970608817355,
|
| 573 |
+
"grad_norm": 0.16625691950321198,
|
| 574 |
+
"learning_rate": 1.6561142147734328e-05,
|
| 575 |
+
"loss": 0.5451,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.167949615115465,
|
| 580 |
+
"grad_norm": 23.16469955444336,
|
| 581 |
+
"learning_rate": 1.6483550589695843e-05,
|
| 582 |
+
"loss": 0.6489,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.237928621413576,
|
| 587 |
+
"grad_norm": 21.33266258239746,
|
| 588 |
+
"learning_rate": 1.6405959031657358e-05,
|
| 589 |
+
"loss": 0.6715,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.307907627711686,
|
| 594 |
+
"grad_norm": 1.7475327253341675,
|
| 595 |
+
"learning_rate": 1.632836747361887e-05,
|
| 596 |
+
"loss": 0.5272,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.377886634009797,
|
| 601 |
+
"grad_norm": 76.74285888671875,
|
| 602 |
+
"learning_rate": 1.6250775915580385e-05,
|
| 603 |
+
"loss": 0.5337,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.447865640307907,
|
| 608 |
+
"grad_norm": 4.064603328704834,
|
| 609 |
+
"learning_rate": 1.61731843575419e-05,
|
| 610 |
+
"loss": 0.3579,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.517844646606018,
|
| 615 |
+
"grad_norm": 5.200445652008057,
|
| 616 |
+
"learning_rate": 1.6095592799503415e-05,
|
| 617 |
+
"loss": 0.5669,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.587823652904129,
|
| 622 |
+
"grad_norm": 52.517234802246094,
|
| 623 |
+
"learning_rate": 1.601800124146493e-05,
|
| 624 |
+
"loss": 0.6752,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.65780265920224,
|
| 629 |
+
"grad_norm": 7.218233108520508,
|
| 630 |
+
"learning_rate": 1.5940409683426445e-05,
|
| 631 |
+
"loss": 0.5177,
|
| 632 |
+
"step": 2025
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.72778166550035,
|
| 636 |
+
"grad_norm": 21.226499557495117,
|
| 637 |
+
"learning_rate": 1.586281812538796e-05,
|
| 638 |
+
"loss": 0.4427,
|
| 639 |
+
"step": 2050
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.797760671798461,
|
| 643 |
+
"grad_norm": 20.382299423217773,
|
| 644 |
+
"learning_rate": 1.5785226567349475e-05,
|
| 645 |
+
"loss": 0.466,
|
| 646 |
+
"step": 2075
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.867739678096571,
|
| 650 |
+
"grad_norm": 5.556461334228516,
|
| 651 |
+
"learning_rate": 1.570763500931099e-05,
|
| 652 |
+
"loss": 0.4155,
|
| 653 |
+
"step": 2100
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.937718684394682,
|
| 657 |
+
"grad_norm": 20.228059768676758,
|
| 658 |
+
"learning_rate": 1.5630043451272502e-05,
|
| 659 |
+
"loss": 0.3849,
|
| 660 |
+
"step": 2125
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.0,
|
| 664 |
+
"eval_accuracy": 0.9418825015792799,
|
| 665 |
+
"eval_loss": 0.3186940550804138,
|
| 666 |
+
"eval_macro_f1": 0.8222524600169724,
|
| 667 |
+
"eval_runtime": 0.902,
|
| 668 |
+
"eval_samples_per_second": 1754.907,
|
| 669 |
+
"eval_steps_per_second": 109.751,
|
| 670 |
+
"eval_weighted_f1": 0.9370755357048892,
|
| 671 |
+
"step": 2148
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 6.005598320503849,
|
| 675 |
+
"grad_norm": 131.26980590820312,
|
| 676 |
+
"learning_rate": 1.5552451893234017e-05,
|
| 677 |
+
"loss": 0.4086,
|
| 678 |
+
"step": 2150
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 6.07557732680196,
|
| 682 |
+
"grad_norm": 102.93787384033203,
|
| 683 |
+
"learning_rate": 1.5474860335195532e-05,
|
| 684 |
+
"loss": 0.4317,
|
| 685 |
+
"step": 2175
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 6.14555633310007,
|
| 689 |
+
"grad_norm": 111.69163513183594,
|
| 690 |
+
"learning_rate": 1.5397268777157047e-05,
|
| 691 |
+
"loss": 0.3718,
|
| 692 |
+
"step": 2200
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.215535339398181,
|
| 696 |
+
"grad_norm": 76.61121368408203,
|
| 697 |
+
"learning_rate": 1.531967721911856e-05,
|
| 698 |
+
"loss": 0.5557,
|
| 699 |
+
"step": 2225
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 6.285514345696291,
|
| 703 |
+
"grad_norm": 3.659562110900879,
|
| 704 |
+
"learning_rate": 1.5242085661080076e-05,
|
| 705 |
+
"loss": 0.3799,
|
| 706 |
+
"step": 2250
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 6.355493351994402,
|
| 710 |
+
"grad_norm": 121.1417465209961,
|
| 711 |
+
"learning_rate": 1.5164494103041589e-05,
|
| 712 |
+
"loss": 0.2739,
|
| 713 |
+
"step": 2275
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 6.425472358292512,
|
| 717 |
+
"grad_norm": 66.07269287109375,
|
| 718 |
+
"learning_rate": 1.5086902545003106e-05,
|
| 719 |
+
"loss": 0.4065,
|
| 720 |
+
"step": 2300
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.495451364590623,
|
| 724 |
+
"grad_norm": 12.768668174743652,
|
| 725 |
+
"learning_rate": 1.5009310986964621e-05,
|
| 726 |
+
"loss": 0.4301,
|
| 727 |
+
"step": 2325
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 6.565430370888733,
|
| 731 |
+
"grad_norm": 3.3007452487945557,
|
| 732 |
+
"learning_rate": 1.4931719428926134e-05,
|
| 733 |
+
"loss": 0.4107,
|
| 734 |
+
"step": 2350
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 6.635409377186844,
|
| 738 |
+
"grad_norm": 0.19890382885932922,
|
| 739 |
+
"learning_rate": 1.485412787088765e-05,
|
| 740 |
+
"loss": 0.5636,
|
| 741 |
+
"step": 2375
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 6.705388383484954,
|
| 745 |
+
"grad_norm": 25.36736488342285,
|
| 746 |
+
"learning_rate": 1.4776536312849163e-05,
|
| 747 |
+
"loss": 0.2464,
|
| 748 |
+
"step": 2400
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 6.7753673897830655,
|
| 752 |
+
"grad_norm": 41.930152893066406,
|
| 753 |
+
"learning_rate": 1.4698944754810678e-05,
|
| 754 |
+
"loss": 0.4215,
|
| 755 |
+
"step": 2425
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 6.845346396081176,
|
| 759 |
+
"grad_norm": 1.9313616752624512,
|
| 760 |
+
"learning_rate": 1.4621353196772193e-05,
|
| 761 |
+
"loss": 0.336,
|
| 762 |
+
"step": 2450
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 6.9153254023792865,
|
| 766 |
+
"grad_norm": 0.8780106902122498,
|
| 767 |
+
"learning_rate": 1.4543761638733706e-05,
|
| 768 |
+
"loss": 0.2482,
|
| 769 |
+
"step": 2475
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 6.985304408677397,
|
| 773 |
+
"grad_norm": 0.2460872083902359,
|
| 774 |
+
"learning_rate": 1.4466170080695221e-05,
|
| 775 |
+
"loss": 0.6625,
|
| 776 |
+
"step": 2500
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.0,
|
| 780 |
+
"eval_accuracy": 0.9368288060644346,
|
| 781 |
+
"eval_loss": 0.34005007147789,
|
| 782 |
+
"eval_macro_f1": 0.8285966833655621,
|
| 783 |
+
"eval_runtime": 0.9993,
|
| 784 |
+
"eval_samples_per_second": 1584.152,
|
| 785 |
+
"eval_steps_per_second": 99.072,
|
| 786 |
+
"eval_weighted_f1": 0.9358783718251367,
|
| 787 |
+
"step": 2506
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 7.053184044786564,
|
| 791 |
+
"grad_norm": 0.14612752199172974,
|
| 792 |
+
"learning_rate": 1.4388578522656735e-05,
|
| 793 |
+
"loss": 0.3807,
|
| 794 |
+
"step": 2525
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 7.123163051084674,
|
| 798 |
+
"grad_norm": 0.1615884006023407,
|
| 799 |
+
"learning_rate": 1.431098696461825e-05,
|
| 800 |
+
"loss": 0.3867,
|
| 801 |
+
"step": 2550
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 7.193142057382786,
|
| 805 |
+
"grad_norm": 0.22993607819080353,
|
| 806 |
+
"learning_rate": 1.4233395406579767e-05,
|
| 807 |
+
"loss": 0.3853,
|
| 808 |
+
"step": 2575
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 7.263121063680896,
|
| 812 |
+
"grad_norm": 0.2886574864387512,
|
| 813 |
+
"learning_rate": 1.415580384854128e-05,
|
| 814 |
+
"loss": 0.206,
|
| 815 |
+
"step": 2600
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 7.3331000699790065,
|
| 819 |
+
"grad_norm": 0.11890740692615509,
|
| 820 |
+
"learning_rate": 1.4078212290502795e-05,
|
| 821 |
+
"loss": 0.2814,
|
| 822 |
+
"step": 2625
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 7.403079076277117,
|
| 826 |
+
"grad_norm": 2.4131717681884766,
|
| 827 |
+
"learning_rate": 1.400062073246431e-05,
|
| 828 |
+
"loss": 0.4293,
|
| 829 |
+
"step": 2650
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 7.4730580825752275,
|
| 833 |
+
"grad_norm": 32.73126220703125,
|
| 834 |
+
"learning_rate": 1.3923029174425823e-05,
|
| 835 |
+
"loss": 0.4089,
|
| 836 |
+
"step": 2675
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.543037088873338,
|
| 840 |
+
"grad_norm": 17.05631446838379,
|
| 841 |
+
"learning_rate": 1.3845437616387339e-05,
|
| 842 |
+
"loss": 0.4617,
|
| 843 |
+
"step": 2700
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.6130160951714485,
|
| 847 |
+
"grad_norm": 3.1084017753601074,
|
| 848 |
+
"learning_rate": 1.3767846058348852e-05,
|
| 849 |
+
"loss": 0.3543,
|
| 850 |
+
"step": 2725
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.682995101469559,
|
| 854 |
+
"grad_norm": 390.8155212402344,
|
| 855 |
+
"learning_rate": 1.3690254500310367e-05,
|
| 856 |
+
"loss": 0.4262,
|
| 857 |
+
"step": 2750
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 7.7529741077676695,
|
| 861 |
+
"grad_norm": 55.06918716430664,
|
| 862 |
+
"learning_rate": 1.3612662942271882e-05,
|
| 863 |
+
"loss": 0.2812,
|
| 864 |
+
"step": 2775
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.82295311406578,
|
| 868 |
+
"grad_norm": 13.030061721801758,
|
| 869 |
+
"learning_rate": 1.3535071384233395e-05,
|
| 870 |
+
"loss": 0.3767,
|
| 871 |
+
"step": 2800
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 7.8929321203638905,
|
| 875 |
+
"grad_norm": 31.446796417236328,
|
| 876 |
+
"learning_rate": 1.345747982619491e-05,
|
| 877 |
+
"loss": 0.2919,
|
| 878 |
+
"step": 2825
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 7.962911126662002,
|
| 882 |
+
"grad_norm": 0.3947957456111908,
|
| 883 |
+
"learning_rate": 1.3379888268156424e-05,
|
| 884 |
+
"loss": 0.3493,
|
| 885 |
+
"step": 2850
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 8.0,
|
| 889 |
+
"eval_accuracy": 0.9444093493367025,
|
| 890 |
+
"eval_loss": 0.31551989912986755,
|
| 891 |
+
"eval_macro_f1": 0.8371578041304867,
|
| 892 |
+
"eval_runtime": 0.9,
|
| 893 |
+
"eval_samples_per_second": 1758.942,
|
| 894 |
+
"eval_steps_per_second": 110.003,
|
| 895 |
+
"eval_weighted_f1": 0.9415404931299006,
|
| 896 |
+
"step": 2864
|
| 897 |
+
}
|
| 898 |
+
],
|
| 899 |
+
"logging_steps": 25,
|
| 900 |
+
"max_steps": 7160,
|
| 901 |
+
"num_input_tokens_seen": 0,
|
| 902 |
+
"num_train_epochs": 20,
|
| 903 |
+
"save_steps": 500,
|
| 904 |
+
"stateful_callbacks": {
|
| 905 |
+
"EarlyStoppingCallback": {
|
| 906 |
+
"args": {
|
| 907 |
+
"early_stopping_patience": 3,
|
| 908 |
+
"early_stopping_threshold": 0.0
|
| 909 |
+
},
|
| 910 |
+
"attributes": {
|
| 911 |
+
"early_stopping_patience_counter": 3
|
| 912 |
+
}
|
| 913 |
+
},
|
| 914 |
+
"TrainerControl": {
|
| 915 |
+
"args": {
|
| 916 |
+
"should_epoch_stop": false,
|
| 917 |
+
"should_evaluate": false,
|
| 918 |
+
"should_log": false,
|
| 919 |
+
"should_save": true,
|
| 920 |
+
"should_training_stop": true
|
| 921 |
+
},
|
| 922 |
+
"attributes": {}
|
| 923 |
+
}
|
| 924 |
+
},
|
| 925 |
+
"total_flos": 0.0,
|
| 926 |
+
"train_batch_size": 8,
|
| 927 |
+
"trial_name": null,
|
| 928 |
+
"trial_params": null
|
| 929 |
+
}
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432/trainer_state.json
ADDED
|
@@ -0,0 +1,486 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1432,
|
| 3 |
+
"best_metric": 0.8443713998163975,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432",
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1432,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 36.41475296020508,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.5921,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 36.4085578918457,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 4.3285,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 69.96395874023438,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.8186,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 180.90719604492188,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4549,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 29.332469940185547,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 3.3126,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 24.94131088256836,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 3.1176,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 26.75153160095215,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 2.6855,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 55.58729934692383,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 2.3268,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 27.88650131225586,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.67,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 57.269588470458984,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.5031,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 42.847984313964844,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.527,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 89.5159683227539,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.3595,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 36.81391525268555,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.2292,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 48.64099884033203,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.4035,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9109286165508528,
|
| 113 |
+
"eval_loss": 0.3039235770702362,
|
| 114 |
+
"eval_macro_f1": 0.6303599039769252,
|
| 115 |
+
"eval_runtime": 0.9278,
|
| 116 |
+
"eval_samples_per_second": 1706.275,
|
| 117 |
+
"eval_steps_per_second": 106.71,
|
| 118 |
+
"eval_weighted_f1": 0.890330086656749,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 45.324806213378906,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 1.1444,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 32.80961608886719,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 1.2518,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 29.984127044677734,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 0.9491,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 37.07761001586914,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 1.3461,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 29.19913673400879,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 0.9921,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 44.5896110534668,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.9285,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 27.994483947753906,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 0.9643,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 29.241727828979492,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.9931,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 52.123844146728516,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 1.0355,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 33.14244079589844,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.9322,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 6.311270236968994,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.9959,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 253.40066528320312,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 1.1593,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 84.322509765625,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.9566,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 22.185016632080078,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.8731,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.934301958307012,
|
| 222 |
+
"eval_loss": 0.23205561935901642,
|
| 223 |
+
"eval_macro_f1": 0.8139507372814744,
|
| 224 |
+
"eval_runtime": 0.9581,
|
| 225 |
+
"eval_samples_per_second": 1652.238,
|
| 226 |
+
"eval_steps_per_second": 103.33,
|
| 227 |
+
"eval_weighted_f1": 0.9297697779953376,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 31.749889373779297,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.9223,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 23.921693801879883,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.7971,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 21.33156394958496,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.9021,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 7.6090168952941895,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.7894,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 34.931556701660156,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.8078,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 15.035621643066406,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.8663,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 8.920546531677246,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.7591,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 38.21807098388672,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.8552,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 26.552431106567383,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.9426,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 37.04743576049805,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.9663,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 50.23114013671875,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.758,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 65.55677032470703,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.8147,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 35.47499465942383,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.9093,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 41.435394287109375,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.8266,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9418825015792799,
|
| 331 |
+
"eval_loss": 0.23250921070575714,
|
| 332 |
+
"eval_macro_f1": 0.8247827620621498,
|
| 333 |
+
"eval_runtime": 0.9802,
|
| 334 |
+
"eval_samples_per_second": 1614.976,
|
| 335 |
+
"eval_steps_per_second": 101.0,
|
| 336 |
+
"eval_weighted_f1": 0.9374505526690534,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 15.043100357055664,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.6459,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 19.535070419311523,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.7997,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 3.4376556873321533,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.5746,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 69.86431121826172,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.6904,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 31.638586044311523,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.7226,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 96.95050048828125,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.6862,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 21.67270278930664,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.7354,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 25.247961044311523,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.5942,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 17.575376510620117,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.7589,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 23.245197296142578,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.5925,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 33.69935989379883,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.8814,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 0.5149115920066833,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.6155,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 12.667987823486328,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.6186,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 21.41044807434082,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.7065,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 18.179180145263672,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.6432,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9399873657612129,
|
| 447 |
+
"eval_loss": 0.19400812685489655,
|
| 448 |
+
"eval_macro_f1": 0.8443713998163975,
|
| 449 |
+
"eval_runtime": 0.9233,
|
| 450 |
+
"eval_samples_per_second": 1714.574,
|
| 451 |
+
"eval_steps_per_second": 107.229,
|
| 452 |
+
"eval_weighted_f1": 0.9394707465667415,
|
| 453 |
+
"step": 1432
|
| 454 |
+
}
|
| 455 |
+
],
|
| 456 |
+
"logging_steps": 25,
|
| 457 |
+
"max_steps": 7160,
|
| 458 |
+
"num_input_tokens_seen": 0,
|
| 459 |
+
"num_train_epochs": 20,
|
| 460 |
+
"save_steps": 500,
|
| 461 |
+
"stateful_callbacks": {
|
| 462 |
+
"EarlyStoppingCallback": {
|
| 463 |
+
"args": {
|
| 464 |
+
"early_stopping_patience": 3,
|
| 465 |
+
"early_stopping_threshold": 0.0
|
| 466 |
+
},
|
| 467 |
+
"attributes": {
|
| 468 |
+
"early_stopping_patience_counter": 0
|
| 469 |
+
}
|
| 470 |
+
},
|
| 471 |
+
"TrainerControl": {
|
| 472 |
+
"args": {
|
| 473 |
+
"should_epoch_stop": false,
|
| 474 |
+
"should_evaluate": false,
|
| 475 |
+
"should_log": false,
|
| 476 |
+
"should_save": true,
|
| 477 |
+
"should_training_stop": false
|
| 478 |
+
},
|
| 479 |
+
"attributes": {}
|
| 480 |
+
}
|
| 481 |
+
},
|
| 482 |
+
"total_flos": 0.0,
|
| 483 |
+
"train_batch_size": 8,
|
| 484 |
+
"trial_name": null,
|
| 485 |
+
"trial_params": null
|
| 486 |
+
}
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-2506/trainer_state.json
ADDED
|
@@ -0,0 +1,820 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1432,
|
| 3 |
+
"best_metric": 0.8443713998163975,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-bamibert-s43-20260827T122749Z-85e5496f/trainer/checkpoint-1432",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2506,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 36.41475296020508,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.5921,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 36.4085578918457,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 4.3285,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 69.96395874023438,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.8186,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 180.90719604492188,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4549,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 29.332469940185547,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 3.3126,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 24.94131088256836,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 3.1176,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 26.75153160095215,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 2.6855,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 55.58729934692383,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 2.3268,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 27.88650131225586,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.67,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 57.269588470458984,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.5031,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 42.847984313964844,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.527,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 89.5159683227539,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.3595,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 36.81391525268555,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.2292,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 48.64099884033203,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.4035,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9109286165508528,
|
| 113 |
+
"eval_loss": 0.3039235770702362,
|
| 114 |
+
"eval_macro_f1": 0.6303599039769252,
|
| 115 |
+
"eval_runtime": 0.9278,
|
| 116 |
+
"eval_samples_per_second": 1706.275,
|
| 117 |
+
"eval_steps_per_second": 106.71,
|
| 118 |
+
"eval_weighted_f1": 0.890330086656749,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 45.324806213378906,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 1.1444,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 32.80961608886719,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 1.2518,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 29.984127044677734,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 0.9491,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 37.07761001586914,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 1.3461,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 29.19913673400879,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 0.9921,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 44.5896110534668,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.9285,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 27.994483947753906,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 0.9643,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 29.241727828979492,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.9931,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 52.123844146728516,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 1.0355,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 33.14244079589844,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.9322,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 6.311270236968994,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.9959,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 253.40066528320312,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 1.1593,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 84.322509765625,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.9566,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 22.185016632080078,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.8731,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.934301958307012,
|
| 222 |
+
"eval_loss": 0.23205561935901642,
|
| 223 |
+
"eval_macro_f1": 0.8139507372814744,
|
| 224 |
+
"eval_runtime": 0.9581,
|
| 225 |
+
"eval_samples_per_second": 1652.238,
|
| 226 |
+
"eval_steps_per_second": 103.33,
|
| 227 |
+
"eval_weighted_f1": 0.9297697779953376,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 31.749889373779297,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.9223,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 23.921693801879883,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.7971,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 21.33156394958496,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.9021,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 7.6090168952941895,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.7894,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 34.931556701660156,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.8078,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 15.035621643066406,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.8663,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 8.920546531677246,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.7591,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 38.21807098388672,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.8552,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 26.552431106567383,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.9426,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 37.04743576049805,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.9663,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 50.23114013671875,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.758,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 65.55677032470703,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.8147,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 35.47499465942383,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.9093,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 41.435394287109375,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.8266,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9418825015792799,
|
| 331 |
+
"eval_loss": 0.23250921070575714,
|
| 332 |
+
"eval_macro_f1": 0.8247827620621498,
|
| 333 |
+
"eval_runtime": 0.9802,
|
| 334 |
+
"eval_samples_per_second": 1614.976,
|
| 335 |
+
"eval_steps_per_second": 101.0,
|
| 336 |
+
"eval_weighted_f1": 0.9374505526690534,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 15.043100357055664,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.6459,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 19.535070419311523,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.7997,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 3.4376556873321533,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.5746,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 69.86431121826172,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.6904,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 31.638586044311523,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.7226,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 96.95050048828125,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.6862,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 21.67270278930664,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.7354,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 25.247961044311523,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.5942,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 17.575376510620117,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.7589,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 23.245197296142578,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.5925,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 33.69935989379883,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.8814,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 0.5149115920066833,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.6155,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 12.667987823486328,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.6186,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 21.41044807434082,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.7065,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 18.179180145263672,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.6432,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9399873657612129,
|
| 447 |
+
"eval_loss": 0.19400812685489655,
|
| 448 |
+
"eval_macro_f1": 0.8443713998163975,
|
| 449 |
+
"eval_runtime": 0.9233,
|
| 450 |
+
"eval_samples_per_second": 1714.574,
|
| 451 |
+
"eval_steps_per_second": 107.229,
|
| 452 |
+
"eval_weighted_f1": 0.9394707465667415,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 3.2257983684539795,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.6633,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 19.694631576538086,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.6201,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 39.88555145263672,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.5799,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 13.165704727172852,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6309,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 74.67707824707031,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.5803,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 40.696807861328125,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.6139,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 49.941558837890625,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.6145,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 28.694549560546875,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.6588,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 4.543245792388916,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.6169,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 24.31388282775879,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.4412,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 52.120059967041016,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.5364,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 25.07593536376953,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.5834,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 5.423189163208008,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.5368,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 63.812931060791016,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.4511,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9418825015792799,
|
| 556 |
+
"eval_loss": 0.29054126143455505,
|
| 557 |
+
"eval_macro_f1": 0.8207679652958646,
|
| 558 |
+
"eval_runtime": 0.9089,
|
| 559 |
+
"eval_samples_per_second": 1741.599,
|
| 560 |
+
"eval_steps_per_second": 108.919,
|
| 561 |
+
"eval_weighted_f1": 0.9371481835590542,
|
| 562 |
+
"step": 1790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.027991602519244,
|
| 566 |
+
"grad_norm": 144.5280303955078,
|
| 567 |
+
"learning_rate": 1.6638733705772813e-05,
|
| 568 |
+
"loss": 0.6536,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.097970608817355,
|
| 573 |
+
"grad_norm": 15.186884880065918,
|
| 574 |
+
"learning_rate": 1.6561142147734328e-05,
|
| 575 |
+
"loss": 0.5327,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.167949615115465,
|
| 580 |
+
"grad_norm": 35.66469192504883,
|
| 581 |
+
"learning_rate": 1.6483550589695843e-05,
|
| 582 |
+
"loss": 0.3871,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.237928621413576,
|
| 587 |
+
"grad_norm": 11.442131996154785,
|
| 588 |
+
"learning_rate": 1.6405959031657358e-05,
|
| 589 |
+
"loss": 0.6045,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.307907627711686,
|
| 594 |
+
"grad_norm": 39.65973663330078,
|
| 595 |
+
"learning_rate": 1.632836747361887e-05,
|
| 596 |
+
"loss": 0.4038,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.377886634009797,
|
| 601 |
+
"grad_norm": 0.3154837489128113,
|
| 602 |
+
"learning_rate": 1.6250775915580385e-05,
|
| 603 |
+
"loss": 0.4626,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.447865640307907,
|
| 608 |
+
"grad_norm": 8.338847160339355,
|
| 609 |
+
"learning_rate": 1.61731843575419e-05,
|
| 610 |
+
"loss": 0.5832,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.517844646606018,
|
| 615 |
+
"grad_norm": 5.839235782623291,
|
| 616 |
+
"learning_rate": 1.6095592799503415e-05,
|
| 617 |
+
"loss": 0.3529,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.587823652904129,
|
| 622 |
+
"grad_norm": 11.817380905151367,
|
| 623 |
+
"learning_rate": 1.601800124146493e-05,
|
| 624 |
+
"loss": 0.5476,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.65780265920224,
|
| 629 |
+
"grad_norm": 3.6478872299194336,
|
| 630 |
+
"learning_rate": 1.5940409683426445e-05,
|
| 631 |
+
"loss": 0.3831,
|
| 632 |
+
"step": 2025
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.72778166550035,
|
| 636 |
+
"grad_norm": 44.632144927978516,
|
| 637 |
+
"learning_rate": 1.586281812538796e-05,
|
| 638 |
+
"loss": 0.544,
|
| 639 |
+
"step": 2050
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.797760671798461,
|
| 643 |
+
"grad_norm": 0.32747405767440796,
|
| 644 |
+
"learning_rate": 1.5785226567349475e-05,
|
| 645 |
+
"loss": 0.5064,
|
| 646 |
+
"step": 2075
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.867739678096571,
|
| 650 |
+
"grad_norm": 100.1316909790039,
|
| 651 |
+
"learning_rate": 1.570763500931099e-05,
|
| 652 |
+
"loss": 0.3017,
|
| 653 |
+
"step": 2100
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.937718684394682,
|
| 657 |
+
"grad_norm": 8.518050193786621,
|
| 658 |
+
"learning_rate": 1.5630043451272502e-05,
|
| 659 |
+
"loss": 0.5024,
|
| 660 |
+
"step": 2125
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.0,
|
| 664 |
+
"eval_accuracy": 0.9412507896399241,
|
| 665 |
+
"eval_loss": 0.2573198974132538,
|
| 666 |
+
"eval_macro_f1": 0.8391642911814072,
|
| 667 |
+
"eval_runtime": 0.9467,
|
| 668 |
+
"eval_samples_per_second": 1672.178,
|
| 669 |
+
"eval_steps_per_second": 104.577,
|
| 670 |
+
"eval_weighted_f1": 0.9397002786541704,
|
| 671 |
+
"step": 2148
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 6.005598320503849,
|
| 675 |
+
"grad_norm": 3.239985466003418,
|
| 676 |
+
"learning_rate": 1.5552451893234017e-05,
|
| 677 |
+
"loss": 0.5656,
|
| 678 |
+
"step": 2150
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 6.07557732680196,
|
| 682 |
+
"grad_norm": 3.581731081008911,
|
| 683 |
+
"learning_rate": 1.5474860335195532e-05,
|
| 684 |
+
"loss": 0.2974,
|
| 685 |
+
"step": 2175
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 6.14555633310007,
|
| 689 |
+
"grad_norm": 0.25982731580734253,
|
| 690 |
+
"learning_rate": 1.5397268777157047e-05,
|
| 691 |
+
"loss": 0.3872,
|
| 692 |
+
"step": 2200
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.215535339398181,
|
| 696 |
+
"grad_norm": 0.4616323709487915,
|
| 697 |
+
"learning_rate": 1.531967721911856e-05,
|
| 698 |
+
"loss": 0.4202,
|
| 699 |
+
"step": 2225
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 6.285514345696291,
|
| 703 |
+
"grad_norm": 1.0121649503707886,
|
| 704 |
+
"learning_rate": 1.5242085661080076e-05,
|
| 705 |
+
"loss": 0.4435,
|
| 706 |
+
"step": 2250
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 6.355493351994402,
|
| 710 |
+
"grad_norm": 57.74125671386719,
|
| 711 |
+
"learning_rate": 1.5164494103041589e-05,
|
| 712 |
+
"loss": 0.4316,
|
| 713 |
+
"step": 2275
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 6.425472358292512,
|
| 717 |
+
"grad_norm": 34.03230285644531,
|
| 718 |
+
"learning_rate": 1.5086902545003106e-05,
|
| 719 |
+
"loss": 0.4277,
|
| 720 |
+
"step": 2300
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.495451364590623,
|
| 724 |
+
"grad_norm": 63.19587707519531,
|
| 725 |
+
"learning_rate": 1.5009310986964621e-05,
|
| 726 |
+
"loss": 0.4668,
|
| 727 |
+
"step": 2325
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 6.565430370888733,
|
| 731 |
+
"grad_norm": 0.2165483832359314,
|
| 732 |
+
"learning_rate": 1.4931719428926134e-05,
|
| 733 |
+
"loss": 0.4616,
|
| 734 |
+
"step": 2350
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 6.635409377186844,
|
| 738 |
+
"grad_norm": 86.36817932128906,
|
| 739 |
+
"learning_rate": 1.485412787088765e-05,
|
| 740 |
+
"loss": 0.4504,
|
| 741 |
+
"step": 2375
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 6.705388383484954,
|
| 745 |
+
"grad_norm": 1.3609081506729126,
|
| 746 |
+
"learning_rate": 1.4776536312849163e-05,
|
| 747 |
+
"loss": 0.2468,
|
| 748 |
+
"step": 2400
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 6.7753673897830655,
|
| 752 |
+
"grad_norm": 2.522094964981079,
|
| 753 |
+
"learning_rate": 1.4698944754810678e-05,
|
| 754 |
+
"loss": 0.5077,
|
| 755 |
+
"step": 2425
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 6.845346396081176,
|
| 759 |
+
"grad_norm": 1.7295256853103638,
|
| 760 |
+
"learning_rate": 1.4621353196772193e-05,
|
| 761 |
+
"loss": 0.452,
|
| 762 |
+
"step": 2450
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 6.9153254023792865,
|
| 766 |
+
"grad_norm": 0.42052462697029114,
|
| 767 |
+
"learning_rate": 1.4543761638733706e-05,
|
| 768 |
+
"loss": 0.3788,
|
| 769 |
+
"step": 2475
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 6.985304408677397,
|
| 773 |
+
"grad_norm": 16.363544464111328,
|
| 774 |
+
"learning_rate": 1.4466170080695221e-05,
|
| 775 |
+
"loss": 0.2761,
|
| 776 |
+
"step": 2500
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.0,
|
| 780 |
+
"eval_accuracy": 0.9393556538218573,
|
| 781 |
+
"eval_loss": 0.32707327604293823,
|
| 782 |
+
"eval_macro_f1": 0.8271961017586303,
|
| 783 |
+
"eval_runtime": 0.9247,
|
| 784 |
+
"eval_samples_per_second": 1711.902,
|
| 785 |
+
"eval_steps_per_second": 107.061,
|
| 786 |
+
"eval_weighted_f1": 0.9352029070367505,
|
| 787 |
+
"step": 2506
|
| 788 |
+
}
|
| 789 |
+
],
|
| 790 |
+
"logging_steps": 25,
|
| 791 |
+
"max_steps": 7160,
|
| 792 |
+
"num_input_tokens_seen": 0,
|
| 793 |
+
"num_train_epochs": 20,
|
| 794 |
+
"save_steps": 500,
|
| 795 |
+
"stateful_callbacks": {
|
| 796 |
+
"EarlyStoppingCallback": {
|
| 797 |
+
"args": {
|
| 798 |
+
"early_stopping_patience": 3,
|
| 799 |
+
"early_stopping_threshold": 0.0
|
| 800 |
+
},
|
| 801 |
+
"attributes": {
|
| 802 |
+
"early_stopping_patience_counter": 3
|
| 803 |
+
}
|
| 804 |
+
},
|
| 805 |
+
"TrainerControl": {
|
| 806 |
+
"args": {
|
| 807 |
+
"should_epoch_stop": false,
|
| 808 |
+
"should_evaluate": false,
|
| 809 |
+
"should_log": false,
|
| 810 |
+
"should_save": true,
|
| 811 |
+
"should_training_stop": true
|
| 812 |
+
},
|
| 813 |
+
"attributes": {}
|
| 814 |
+
}
|
| 815 |
+
},
|
| 816 |
+
"total_flos": 0.0,
|
| 817 |
+
"train_batch_size": 8,
|
| 818 |
+
"trial_name": null,
|
| 819 |
+
"trial_params": null
|
| 820 |
+
}
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074/trainer_state.json
ADDED
|
@@ -0,0 +1,370 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1074,
|
| 3 |
+
"best_metric": 0.8419379600118683,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1074,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 28.010095596313477,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.3041,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 62.65412139892578,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 4.0315,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 35.63129806518555,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.7538,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 29.297677993774414,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4809,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 27.251455307006836,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 3.2293,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 26.070375442504883,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 2.9205,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 35.197288513183594,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 2.3483,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 46.078243255615234,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 1.7262,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 53.353885650634766,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.6336,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 132.63648986816406,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.4125,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 70.5636215209961,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.5112,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 93.6589126586914,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.3256,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 11.28514575958252,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.2971,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 103.91099548339844,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.0652,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9115603284902085,
|
| 113 |
+
"eval_loss": 0.27931907773017883,
|
| 114 |
+
"eval_macro_f1": 0.621900442488988,
|
| 115 |
+
"eval_runtime": 0.9773,
|
| 116 |
+
"eval_samples_per_second": 1619.741,
|
| 117 |
+
"eval_steps_per_second": 101.298,
|
| 118 |
+
"eval_weighted_f1": 0.8902744145737365,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 63.85672378540039,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 1.1415,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 34.2215576171875,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 1.1235,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 18.667619705200195,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 1.1183,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 110.78736114501953,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 0.9818,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 56.78545379638672,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 1.1776,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 64.2852554321289,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.8344,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 51.086185455322266,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 1.0726,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 19.155624389648438,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.9696,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 28.97992515563965,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 1.1268,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 48.032814025878906,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.935,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 593.2311401367188,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.8929,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 16.216840744018555,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.9148,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 19.12264060974121,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 1.0019,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 7.697551250457764,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.984,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9425142135186355,
|
| 222 |
+
"eval_loss": 0.219716414809227,
|
| 223 |
+
"eval_macro_f1": 0.821307570627719,
|
| 224 |
+
"eval_runtime": 0.921,
|
| 225 |
+
"eval_samples_per_second": 1718.807,
|
| 226 |
+
"eval_steps_per_second": 107.493,
|
| 227 |
+
"eval_weighted_f1": 0.9373076630234927,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 24.53575325012207,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.7385,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 48.4820556640625,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.8924,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 27.167097091674805,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.7517,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 16.005901336669922,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.9427,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 35.83484649658203,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.8933,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 23.349987030029297,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.7981,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 7.260409832000732,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.8023,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 22.78744888305664,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.841,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 18.904150009155273,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.7074,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 33.38016128540039,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.8701,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 52.98533248901367,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.9667,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 7.9378342628479,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.6651,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 30.187931060791016,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.8053,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 12.127618789672852,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.7133,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9399873657612129,
|
| 331 |
+
"eval_loss": 0.18643338978290558,
|
| 332 |
+
"eval_macro_f1": 0.8419379600118683,
|
| 333 |
+
"eval_runtime": 0.9566,
|
| 334 |
+
"eval_samples_per_second": 1654.866,
|
| 335 |
+
"eval_steps_per_second": 103.494,
|
| 336 |
+
"eval_weighted_f1": 0.9396843899893166,
|
| 337 |
+
"step": 1074
|
| 338 |
+
}
|
| 339 |
+
],
|
| 340 |
+
"logging_steps": 25,
|
| 341 |
+
"max_steps": 7160,
|
| 342 |
+
"num_input_tokens_seen": 0,
|
| 343 |
+
"num_train_epochs": 20,
|
| 344 |
+
"save_steps": 500,
|
| 345 |
+
"stateful_callbacks": {
|
| 346 |
+
"EarlyStoppingCallback": {
|
| 347 |
+
"args": {
|
| 348 |
+
"early_stopping_patience": 3,
|
| 349 |
+
"early_stopping_threshold": 0.0
|
| 350 |
+
},
|
| 351 |
+
"attributes": {
|
| 352 |
+
"early_stopping_patience_counter": 0
|
| 353 |
+
}
|
| 354 |
+
},
|
| 355 |
+
"TrainerControl": {
|
| 356 |
+
"args": {
|
| 357 |
+
"should_epoch_stop": false,
|
| 358 |
+
"should_evaluate": false,
|
| 359 |
+
"should_log": false,
|
| 360 |
+
"should_save": true,
|
| 361 |
+
"should_training_stop": false
|
| 362 |
+
},
|
| 363 |
+
"attributes": {}
|
| 364 |
+
}
|
| 365 |
+
},
|
| 366 |
+
"total_flos": 0.0,
|
| 367 |
+
"train_batch_size": 8,
|
| 368 |
+
"trial_name": null,
|
| 369 |
+
"trial_params": null
|
| 370 |
+
}
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-2148/trainer_state.json
ADDED
|
@@ -0,0 +1,704 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1074,
|
| 3 |
+
"best_metric": 0.8419379600118683,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-bamibert-s44-20260827T123219Z-b6645c77/trainer/checkpoint-1074",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2148,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 28.010095596313477,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.3041,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 62.65412139892578,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 4.0315,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 35.63129806518555,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.7538,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 29.297677993774414,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4809,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 27.251455307006836,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 3.2293,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 26.070375442504883,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 2.9205,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 35.197288513183594,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 2.3483,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 46.078243255615234,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 1.7262,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 53.353885650634766,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.6336,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 132.63648986816406,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.4125,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 70.5636215209961,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.5112,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 93.6589126586914,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.3256,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 11.28514575958252,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.2971,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 103.91099548339844,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.0652,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9115603284902085,
|
| 113 |
+
"eval_loss": 0.27931907773017883,
|
| 114 |
+
"eval_macro_f1": 0.621900442488988,
|
| 115 |
+
"eval_runtime": 0.9773,
|
| 116 |
+
"eval_samples_per_second": 1619.741,
|
| 117 |
+
"eval_steps_per_second": 101.298,
|
| 118 |
+
"eval_weighted_f1": 0.8902744145737365,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 63.85672378540039,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 1.1415,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 34.2215576171875,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 1.1235,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 18.667619705200195,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 1.1183,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 110.78736114501953,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 0.9818,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 56.78545379638672,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 1.1776,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 64.2852554321289,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.8344,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 51.086185455322266,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 1.0726,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 19.155624389648438,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.9696,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 28.97992515563965,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 1.1268,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 48.032814025878906,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.935,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 593.2311401367188,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.8929,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 16.216840744018555,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.9148,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 19.12264060974121,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 1.0019,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 7.697551250457764,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.984,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9425142135186355,
|
| 222 |
+
"eval_loss": 0.219716414809227,
|
| 223 |
+
"eval_macro_f1": 0.821307570627719,
|
| 224 |
+
"eval_runtime": 0.921,
|
| 225 |
+
"eval_samples_per_second": 1718.807,
|
| 226 |
+
"eval_steps_per_second": 107.493,
|
| 227 |
+
"eval_weighted_f1": 0.9373076630234927,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 24.53575325012207,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.7385,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 48.4820556640625,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.8924,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 27.167097091674805,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.7517,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 16.005901336669922,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.9427,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 35.83484649658203,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.8933,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 23.349987030029297,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.7981,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 7.260409832000732,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.8023,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 22.78744888305664,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.841,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 18.904150009155273,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.7074,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 33.38016128540039,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.8701,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 52.98533248901367,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.9667,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 7.9378342628479,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.6651,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 30.187931060791016,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.8053,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 12.127618789672852,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.7133,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9399873657612129,
|
| 331 |
+
"eval_loss": 0.18643338978290558,
|
| 332 |
+
"eval_macro_f1": 0.8419379600118683,
|
| 333 |
+
"eval_runtime": 0.9566,
|
| 334 |
+
"eval_samples_per_second": 1654.866,
|
| 335 |
+
"eval_steps_per_second": 103.494,
|
| 336 |
+
"eval_weighted_f1": 0.9396843899893166,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 19.3886775970459,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.8312,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 216.0979461669922,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.653,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 44.12488555908203,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.8774,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 41.19123458862305,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.7561,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 26.397462844848633,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.6819,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 25.49283790588379,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.6344,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 54.3346061706543,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.5906,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 28.17466926574707,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.8332,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 32.21369171142578,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.8445,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 4.817105770111084,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.6693,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 99.31521606445312,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.4555,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 4.231173992156982,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.5291,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 19.307300567626953,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.6594,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 1.2029757499694824,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.5974,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 19.153480529785156,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.591,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.932406822488945,
|
| 447 |
+
"eval_loss": 0.3084065318107605,
|
| 448 |
+
"eval_macro_f1": 0.8113670883275669,
|
| 449 |
+
"eval_runtime": 0.961,
|
| 450 |
+
"eval_samples_per_second": 1647.258,
|
| 451 |
+
"eval_steps_per_second": 103.019,
|
| 452 |
+
"eval_weighted_f1": 0.9269445917422482,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 57.419776916503906,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.6461,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 13.175628662109375,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.52,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 25.618328094482422,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.4319,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 21.055824279785156,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6668,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 49.4687614440918,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.5568,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 40.28314208984375,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.4406,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 42.87968826293945,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.7064,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 3.423809289932251,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.5626,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 23.562986373901367,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.6567,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 4.882331371307373,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.4247,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 39.2086181640625,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.709,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 84.05718231201172,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.5721,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 54.311683654785156,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.3393,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 27.99472999572754,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.7251,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9425142135186355,
|
| 556 |
+
"eval_loss": 0.2639157474040985,
|
| 557 |
+
"eval_macro_f1": 0.8350762120779,
|
| 558 |
+
"eval_runtime": 0.9244,
|
| 559 |
+
"eval_samples_per_second": 1712.538,
|
| 560 |
+
"eval_steps_per_second": 107.101,
|
| 561 |
+
"eval_weighted_f1": 0.9393037124039346,
|
| 562 |
+
"step": 1790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.027991602519244,
|
| 566 |
+
"grad_norm": 39.31532669067383,
|
| 567 |
+
"learning_rate": 1.6638733705772813e-05,
|
| 568 |
+
"loss": 0.4478,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.097970608817355,
|
| 573 |
+
"grad_norm": 8.918566703796387,
|
| 574 |
+
"learning_rate": 1.6561142147734328e-05,
|
| 575 |
+
"loss": 0.5089,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.167949615115465,
|
| 580 |
+
"grad_norm": 16.801895141601562,
|
| 581 |
+
"learning_rate": 1.6483550589695843e-05,
|
| 582 |
+
"loss": 0.456,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.237928621413576,
|
| 587 |
+
"grad_norm": 122.81635284423828,
|
| 588 |
+
"learning_rate": 1.6405959031657358e-05,
|
| 589 |
+
"loss": 0.4442,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.307907627711686,
|
| 594 |
+
"grad_norm": 0.3051159381866455,
|
| 595 |
+
"learning_rate": 1.632836747361887e-05,
|
| 596 |
+
"loss": 0.3605,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.377886634009797,
|
| 601 |
+
"grad_norm": 8.220174789428711,
|
| 602 |
+
"learning_rate": 1.6250775915580385e-05,
|
| 603 |
+
"loss": 0.6141,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.447865640307907,
|
| 608 |
+
"grad_norm": 14.740073204040527,
|
| 609 |
+
"learning_rate": 1.61731843575419e-05,
|
| 610 |
+
"loss": 0.6517,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.517844646606018,
|
| 615 |
+
"grad_norm": 90.922607421875,
|
| 616 |
+
"learning_rate": 1.6095592799503415e-05,
|
| 617 |
+
"loss": 0.4487,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.587823652904129,
|
| 622 |
+
"grad_norm": 172.3102569580078,
|
| 623 |
+
"learning_rate": 1.601800124146493e-05,
|
| 624 |
+
"loss": 0.4272,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.65780265920224,
|
| 629 |
+
"grad_norm": 1.0384031534194946,
|
| 630 |
+
"learning_rate": 1.5940409683426445e-05,
|
| 631 |
+
"loss": 0.5137,
|
| 632 |
+
"step": 2025
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.72778166550035,
|
| 636 |
+
"grad_norm": 58.63907241821289,
|
| 637 |
+
"learning_rate": 1.586281812538796e-05,
|
| 638 |
+
"loss": 0.4891,
|
| 639 |
+
"step": 2050
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.797760671798461,
|
| 643 |
+
"grad_norm": 0.17095279693603516,
|
| 644 |
+
"learning_rate": 1.5785226567349475e-05,
|
| 645 |
+
"loss": 0.518,
|
| 646 |
+
"step": 2075
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.867739678096571,
|
| 650 |
+
"grad_norm": 29.13418197631836,
|
| 651 |
+
"learning_rate": 1.570763500931099e-05,
|
| 652 |
+
"loss": 0.4642,
|
| 653 |
+
"step": 2100
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.937718684394682,
|
| 657 |
+
"grad_norm": 50.277984619140625,
|
| 658 |
+
"learning_rate": 1.5630043451272502e-05,
|
| 659 |
+
"loss": 0.5436,
|
| 660 |
+
"step": 2125
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.0,
|
| 664 |
+
"eval_accuracy": 0.9380922299431459,
|
| 665 |
+
"eval_loss": 0.3266836702823639,
|
| 666 |
+
"eval_macro_f1": 0.8150961990139621,
|
| 667 |
+
"eval_runtime": 0.9663,
|
| 668 |
+
"eval_samples_per_second": 1638.183,
|
| 669 |
+
"eval_steps_per_second": 102.451,
|
| 670 |
+
"eval_weighted_f1": 0.9331800515869827,
|
| 671 |
+
"step": 2148
|
| 672 |
+
}
|
| 673 |
+
],
|
| 674 |
+
"logging_steps": 25,
|
| 675 |
+
"max_steps": 7160,
|
| 676 |
+
"num_input_tokens_seen": 0,
|
| 677 |
+
"num_train_epochs": 20,
|
| 678 |
+
"save_steps": 500,
|
| 679 |
+
"stateful_callbacks": {
|
| 680 |
+
"EarlyStoppingCallback": {
|
| 681 |
+
"args": {
|
| 682 |
+
"early_stopping_patience": 3,
|
| 683 |
+
"early_stopping_threshold": 0.0
|
| 684 |
+
},
|
| 685 |
+
"attributes": {
|
| 686 |
+
"early_stopping_patience_counter": 3
|
| 687 |
+
}
|
| 688 |
+
},
|
| 689 |
+
"TrainerControl": {
|
| 690 |
+
"args": {
|
| 691 |
+
"should_epoch_stop": false,
|
| 692 |
+
"should_evaluate": false,
|
| 693 |
+
"should_log": false,
|
| 694 |
+
"should_save": true,
|
| 695 |
+
"should_training_stop": true
|
| 696 |
+
},
|
| 697 |
+
"attributes": {}
|
| 698 |
+
}
|
| 699 |
+
},
|
| 700 |
+
"total_flos": 0.0,
|
| 701 |
+
"train_batch_size": 8,
|
| 702 |
+
"trial_name": null,
|
| 703 |
+
"trial_params": null
|
| 704 |
+
}
|
viclickbait2025-bamibert-s42-20260827T130938Z-ee19aa8e/trainer/checkpoint-150/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
viclickbait2025-bamibert-s42-20260827T130938Z-ee19aa8e/trainer/checkpoint-150/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264/trainer_state.json
ADDED
|
@@ -0,0 +1,1590 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 5264,
|
| 3 |
+
"best_metric": 0.6915236970070544,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5264,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0332667997338656,
|
| 14 |
+
"grad_norm": 37.213836669921875,
|
| 15 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 16 |
+
"loss": 4.1936,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0665335994677312,
|
| 21 |
+
"grad_norm": 52.82760238647461,
|
| 22 |
+
"learning_rate": 6.515957446808512e-07,
|
| 23 |
+
"loss": 3.937,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0998003992015968,
|
| 28 |
+
"grad_norm": 129.2099151611328,
|
| 29 |
+
"learning_rate": 9.840425531914895e-07,
|
| 30 |
+
"loss": 3.4919,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1330671989354624,
|
| 35 |
+
"grad_norm": 22.686023712158203,
|
| 36 |
+
"learning_rate": 1.3164893617021277e-06,
|
| 37 |
+
"loss": 3.0222,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16633399866932802,
|
| 42 |
+
"grad_norm": 57.567840576171875,
|
| 43 |
+
"learning_rate": 1.648936170212766e-06,
|
| 44 |
+
"loss": 2.5057,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1996007984031936,
|
| 49 |
+
"grad_norm": 99.45436096191406,
|
| 50 |
+
"learning_rate": 1.9813829787234043e-06,
|
| 51 |
+
"loss": 2.3339,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2328675981370592,
|
| 56 |
+
"grad_norm": 13.004047393798828,
|
| 57 |
+
"learning_rate": 2.3138297872340428e-06,
|
| 58 |
+
"loss": 2.1973,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2661343978709248,
|
| 63 |
+
"grad_norm": 29.448951721191406,
|
| 64 |
+
"learning_rate": 2.646276595744681e-06,
|
| 65 |
+
"loss": 2.2666,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.2994011976047904,
|
| 70 |
+
"grad_norm": 16.099824905395508,
|
| 71 |
+
"learning_rate": 2.978723404255319e-06,
|
| 72 |
+
"loss": 2.1449,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.33266799733865604,
|
| 77 |
+
"grad_norm": 28.373823165893555,
|
| 78 |
+
"learning_rate": 3.311170212765958e-06,
|
| 79 |
+
"loss": 2.19,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.36593479707252163,
|
| 84 |
+
"grad_norm": 18.843015670776367,
|
| 85 |
+
"learning_rate": 3.643617021276596e-06,
|
| 86 |
+
"loss": 2.1845,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3992015968063872,
|
| 91 |
+
"grad_norm": 15.624969482421875,
|
| 92 |
+
"learning_rate": 3.976063829787235e-06,
|
| 93 |
+
"loss": 2.2532,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4324683965402528,
|
| 98 |
+
"grad_norm": 30.467538833618164,
|
| 99 |
+
"learning_rate": 4.308510638297873e-06,
|
| 100 |
+
"loss": 1.9754,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4657351962741184,
|
| 105 |
+
"grad_norm": 23.515451431274414,
|
| 106 |
+
"learning_rate": 4.640957446808511e-06,
|
| 107 |
+
"loss": 1.8119,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.499001996007984,
|
| 112 |
+
"grad_norm": 17.149723052978516,
|
| 113 |
+
"learning_rate": 4.973404255319149e-06,
|
| 114 |
+
"loss": 2.0686,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5322687957418496,
|
| 119 |
+
"grad_norm": 24.267086029052734,
|
| 120 |
+
"learning_rate": 5.305851063829788e-06,
|
| 121 |
+
"loss": 1.7767,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5655355954757152,
|
| 126 |
+
"grad_norm": 33.58161163330078,
|
| 127 |
+
"learning_rate": 5.638297872340426e-06,
|
| 128 |
+
"loss": 1.8802,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5988023952095808,
|
| 133 |
+
"grad_norm": 17.654483795166016,
|
| 134 |
+
"learning_rate": 5.970744680851064e-06,
|
| 135 |
+
"loss": 2.0599,
|
| 136 |
+
"step": 450
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6320691949434465,
|
| 140 |
+
"grad_norm": 37.77925491333008,
|
| 141 |
+
"learning_rate": 6.303191489361703e-06,
|
| 142 |
+
"loss": 2.0091,
|
| 143 |
+
"step": 475
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6653359946773121,
|
| 147 |
+
"grad_norm": 8.559550285339355,
|
| 148 |
+
"learning_rate": 6.635638297872341e-06,
|
| 149 |
+
"loss": 1.8012,
|
| 150 |
+
"step": 500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6986027944111777,
|
| 154 |
+
"grad_norm": 16.377788543701172,
|
| 155 |
+
"learning_rate": 6.968085106382979e-06,
|
| 156 |
+
"loss": 1.5796,
|
| 157 |
+
"step": 525
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7318695941450433,
|
| 161 |
+
"grad_norm": 39.23353576660156,
|
| 162 |
+
"learning_rate": 7.300531914893617e-06,
|
| 163 |
+
"loss": 1.7269,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7651363938789089,
|
| 168 |
+
"grad_norm": 33.32931900024414,
|
| 169 |
+
"learning_rate": 7.632978723404256e-06,
|
| 170 |
+
"loss": 1.6038,
|
| 171 |
+
"step": 575
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.7984031936127745,
|
| 175 |
+
"grad_norm": 36.23918914794922,
|
| 176 |
+
"learning_rate": 7.965425531914895e-06,
|
| 177 |
+
"loss": 1.9594,
|
| 178 |
+
"step": 600
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.83166999334664,
|
| 182 |
+
"grad_norm": 28.780132293701172,
|
| 183 |
+
"learning_rate": 8.297872340425532e-06,
|
| 184 |
+
"loss": 1.8884,
|
| 185 |
+
"step": 625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8649367930805056,
|
| 189 |
+
"grad_norm": 32.994388580322266,
|
| 190 |
+
"learning_rate": 8.630319148936171e-06,
|
| 191 |
+
"loss": 1.669,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.8982035928143712,
|
| 196 |
+
"grad_norm": 26.472225189208984,
|
| 197 |
+
"learning_rate": 8.96276595744681e-06,
|
| 198 |
+
"loss": 1.6077,
|
| 199 |
+
"step": 675
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9314703925482368,
|
| 203 |
+
"grad_norm": 29.906286239624023,
|
| 204 |
+
"learning_rate": 9.295212765957447e-06,
|
| 205 |
+
"loss": 1.6924,
|
| 206 |
+
"step": 700
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9647371922821024,
|
| 210 |
+
"grad_norm": 25.40384292602539,
|
| 211 |
+
"learning_rate": 9.627659574468086e-06,
|
| 212 |
+
"loss": 1.7419,
|
| 213 |
+
"step": 725
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.998003992015968,
|
| 217 |
+
"grad_norm": 23.713624954223633,
|
| 218 |
+
"learning_rate": 9.960106382978723e-06,
|
| 219 |
+
"loss": 1.5993,
|
| 220 |
+
"step": 750
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0,
|
| 224 |
+
"eval_accuracy": 0.8480538922155688,
|
| 225 |
+
"eval_loss": 0.42238420248031616,
|
| 226 |
+
"eval_macro_f1": 0.51138076970829,
|
| 227 |
+
"eval_runtime": 1.5008,
|
| 228 |
+
"eval_samples_per_second": 1780.386,
|
| 229 |
+
"eval_steps_per_second": 111.274,
|
| 230 |
+
"eval_weighted_f1": 0.821597408229364,
|
| 231 |
+
"step": 752
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0306054557551563,
|
| 235 |
+
"grad_norm": 32.69313430786133,
|
| 236 |
+
"learning_rate": 1.0292553191489364e-05,
|
| 237 |
+
"loss": 1.6252,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.0638722554890219,
|
| 242 |
+
"grad_norm": 24.044479370117188,
|
| 243 |
+
"learning_rate": 1.0625e-05,
|
| 244 |
+
"loss": 1.5655,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.0971390552228875,
|
| 249 |
+
"grad_norm": 22.281408309936523,
|
| 250 |
+
"learning_rate": 1.095744680851064e-05,
|
| 251 |
+
"loss": 1.6928,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.130405854956753,
|
| 256 |
+
"grad_norm": 41.09510803222656,
|
| 257 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 258 |
+
"loss": 1.5383,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.1636726546906186,
|
| 263 |
+
"grad_norm": 23.410619735717773,
|
| 264 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 265 |
+
"loss": 1.4538,
|
| 266 |
+
"step": 875
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.1969394544244845,
|
| 270 |
+
"grad_norm": 34.236785888671875,
|
| 271 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 272 |
+
"loss": 1.7808,
|
| 273 |
+
"step": 900
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.23020625415835,
|
| 277 |
+
"grad_norm": 33.10630798339844,
|
| 278 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 279 |
+
"loss": 1.5242,
|
| 280 |
+
"step": 925
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.2634730538922156,
|
| 284 |
+
"grad_norm": 26.993452072143555,
|
| 285 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 286 |
+
"loss": 1.4654,
|
| 287 |
+
"step": 950
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.2967398536260812,
|
| 291 |
+
"grad_norm": 25.680158615112305,
|
| 292 |
+
"learning_rate": 1.295212765957447e-05,
|
| 293 |
+
"loss": 1.3583,
|
| 294 |
+
"step": 975
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.3300066533599468,
|
| 298 |
+
"grad_norm": 20.42882537841797,
|
| 299 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 300 |
+
"loss": 1.5349,
|
| 301 |
+
"step": 1000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.3632734530938124,
|
| 305 |
+
"grad_norm": 20.90618133544922,
|
| 306 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 307 |
+
"loss": 1.5473,
|
| 308 |
+
"step": 1025
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.396540252827678,
|
| 312 |
+
"grad_norm": 27.44194793701172,
|
| 313 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 314 |
+
"loss": 1.4876,
|
| 315 |
+
"step": 1050
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.4298070525615436,
|
| 319 |
+
"grad_norm": 26.702394485473633,
|
| 320 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 321 |
+
"loss": 1.5546,
|
| 322 |
+
"step": 1075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.4630738522954092,
|
| 326 |
+
"grad_norm": 28.11456871032715,
|
| 327 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 328 |
+
"loss": 1.7266,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.4963406520292748,
|
| 333 |
+
"grad_norm": 29.24340057373047,
|
| 334 |
+
"learning_rate": 1.49468085106383e-05,
|
| 335 |
+
"loss": 1.395,
|
| 336 |
+
"step": 1125
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.5296074517631404,
|
| 340 |
+
"grad_norm": 24.214881896972656,
|
| 341 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 342 |
+
"loss": 1.4127,
|
| 343 |
+
"step": 1150
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.562874251497006,
|
| 347 |
+
"grad_norm": 33.323997497558594,
|
| 348 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 349 |
+
"loss": 1.6986,
|
| 350 |
+
"step": 1175
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5961410512308716,
|
| 354 |
+
"grad_norm": 38.097965240478516,
|
| 355 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 356 |
+
"loss": 1.6189,
|
| 357 |
+
"step": 1200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6294078509647372,
|
| 361 |
+
"grad_norm": 39.18313980102539,
|
| 362 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 363 |
+
"loss": 1.6214,
|
| 364 |
+
"step": 1225
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.6626746506986028,
|
| 368 |
+
"grad_norm": 25.2917537689209,
|
| 369 |
+
"learning_rate": 1.660904255319149e-05,
|
| 370 |
+
"loss": 1.6657,
|
| 371 |
+
"step": 1250
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.6959414504324684,
|
| 375 |
+
"grad_norm": 27.77417755126953,
|
| 376 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 377 |
+
"loss": 1.4622,
|
| 378 |
+
"step": 1275
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.729208250166334,
|
| 382 |
+
"grad_norm": 26.38246726989746,
|
| 383 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 384 |
+
"loss": 1.4535,
|
| 385 |
+
"step": 1300
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.7624750499001998,
|
| 389 |
+
"grad_norm": 24.15390968322754,
|
| 390 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 391 |
+
"loss": 1.6517,
|
| 392 |
+
"step": 1325
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.7957418496340654,
|
| 396 |
+
"grad_norm": 25.304895401000977,
|
| 397 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 398 |
+
"loss": 1.4146,
|
| 399 |
+
"step": 1350
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.829008649367931,
|
| 403 |
+
"grad_norm": 30.78265953063965,
|
| 404 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 405 |
+
"loss": 1.5098,
|
| 406 |
+
"step": 1375
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.8622754491017965,
|
| 410 |
+
"grad_norm": 21.03673553466797,
|
| 411 |
+
"learning_rate": 1.860372340425532e-05,
|
| 412 |
+
"loss": 1.5286,
|
| 413 |
+
"step": 1400
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.8955422488356621,
|
| 417 |
+
"grad_norm": 26.255247116088867,
|
| 418 |
+
"learning_rate": 1.893617021276596e-05,
|
| 419 |
+
"loss": 1.564,
|
| 420 |
+
"step": 1425
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.9288090485695277,
|
| 424 |
+
"grad_norm": 48.96797561645508,
|
| 425 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 426 |
+
"loss": 1.5111,
|
| 427 |
+
"step": 1450
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.9620758483033933,
|
| 431 |
+
"grad_norm": 21.748788833618164,
|
| 432 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 433 |
+
"loss": 1.4308,
|
| 434 |
+
"step": 1475
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.995342648037259,
|
| 438 |
+
"grad_norm": 24.715282440185547,
|
| 439 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 440 |
+
"loss": 1.3674,
|
| 441 |
+
"step": 1500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0,
|
| 445 |
+
"eval_accuracy": 0.8637724550898204,
|
| 446 |
+
"eval_loss": 0.3624139428138733,
|
| 447 |
+
"eval_macro_f1": 0.6436672764733071,
|
| 448 |
+
"eval_runtime": 1.5434,
|
| 449 |
+
"eval_samples_per_second": 1731.26,
|
| 450 |
+
"eval_steps_per_second": 108.204,
|
| 451 |
+
"eval_weighted_f1": 0.8555428081072758,
|
| 452 |
+
"step": 1504
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.027944111776447,
|
| 456 |
+
"grad_norm": 23.2908935546875,
|
| 457 |
+
"learning_rate": 1.9970449172576832e-05,
|
| 458 |
+
"loss": 1.4192,
|
| 459 |
+
"step": 1525
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0612109115103125,
|
| 463 |
+
"grad_norm": 26.697948455810547,
|
| 464 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 465 |
+
"loss": 1.4633,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.094477711244178,
|
| 470 |
+
"grad_norm": 19.0259952545166,
|
| 471 |
+
"learning_rate": 1.9896572104018914e-05,
|
| 472 |
+
"loss": 1.3156,
|
| 473 |
+
"step": 1575
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.1277445109780437,
|
| 477 |
+
"grad_norm": 26.398216247558594,
|
| 478 |
+
"learning_rate": 1.9859633569739953e-05,
|
| 479 |
+
"loss": 1.1985,
|
| 480 |
+
"step": 1600
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.1610113107119093,
|
| 484 |
+
"grad_norm": 23.088871002197266,
|
| 485 |
+
"learning_rate": 1.9822695035460996e-05,
|
| 486 |
+
"loss": 1.2202,
|
| 487 |
+
"step": 1625
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.194278110445775,
|
| 491 |
+
"grad_norm": 26.138647079467773,
|
| 492 |
+
"learning_rate": 1.9785756501182035e-05,
|
| 493 |
+
"loss": 1.2849,
|
| 494 |
+
"step": 1650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.2275449101796405,
|
| 498 |
+
"grad_norm": 18.806598663330078,
|
| 499 |
+
"learning_rate": 1.9748817966903074e-05,
|
| 500 |
+
"loss": 1.2639,
|
| 501 |
+
"step": 1675
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.260811709913506,
|
| 505 |
+
"grad_norm": 15.955117225646973,
|
| 506 |
+
"learning_rate": 1.9711879432624113e-05,
|
| 507 |
+
"loss": 1.3084,
|
| 508 |
+
"step": 1700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.294078509647372,
|
| 512 |
+
"grad_norm": 26.74173355102539,
|
| 513 |
+
"learning_rate": 1.9674940898345156e-05,
|
| 514 |
+
"loss": 1.0918,
|
| 515 |
+
"step": 1725
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.3273453093812373,
|
| 519 |
+
"grad_norm": 25.826892852783203,
|
| 520 |
+
"learning_rate": 1.9638002364066195e-05,
|
| 521 |
+
"loss": 1.3421,
|
| 522 |
+
"step": 1750
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.3606121091151033,
|
| 526 |
+
"grad_norm": 19.300220489501953,
|
| 527 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 528 |
+
"loss": 1.2709,
|
| 529 |
+
"step": 1775
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.393878908848969,
|
| 533 |
+
"grad_norm": 33.41100311279297,
|
| 534 |
+
"learning_rate": 1.9564125295508277e-05,
|
| 535 |
+
"loss": 1.2728,
|
| 536 |
+
"step": 1800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.4271457085828345,
|
| 540 |
+
"grad_norm": 33.1726188659668,
|
| 541 |
+
"learning_rate": 1.9527186761229316e-05,
|
| 542 |
+
"loss": 1.3343,
|
| 543 |
+
"step": 1825
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.4604125083167,
|
| 547 |
+
"grad_norm": 16.324378967285156,
|
| 548 |
+
"learning_rate": 1.9490248226950355e-05,
|
| 549 |
+
"loss": 1.5619,
|
| 550 |
+
"step": 1850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.4936793080505657,
|
| 554 |
+
"grad_norm": 11.470542907714844,
|
| 555 |
+
"learning_rate": 1.9453309692671394e-05,
|
| 556 |
+
"loss": 1.3248,
|
| 557 |
+
"step": 1875
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.5269461077844313,
|
| 561 |
+
"grad_norm": 47.82541275024414,
|
| 562 |
+
"learning_rate": 1.9416371158392437e-05,
|
| 563 |
+
"loss": 1.4005,
|
| 564 |
+
"step": 1900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.560212907518297,
|
| 568 |
+
"grad_norm": 13.039468765258789,
|
| 569 |
+
"learning_rate": 1.9379432624113476e-05,
|
| 570 |
+
"loss": 1.2775,
|
| 571 |
+
"step": 1925
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.5934797072521625,
|
| 575 |
+
"grad_norm": 32.88628387451172,
|
| 576 |
+
"learning_rate": 1.9342494089834515e-05,
|
| 577 |
+
"loss": 1.3989,
|
| 578 |
+
"step": 1950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.626746506986028,
|
| 582 |
+
"grad_norm": 32.169254302978516,
|
| 583 |
+
"learning_rate": 1.9305555555555558e-05,
|
| 584 |
+
"loss": 1.2549,
|
| 585 |
+
"step": 1975
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.6600133067198937,
|
| 589 |
+
"grad_norm": 25.954526901245117,
|
| 590 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 591 |
+
"loss": 1.2969,
|
| 592 |
+
"step": 2000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.6932801064537593,
|
| 596 |
+
"grad_norm": 20.622583389282227,
|
| 597 |
+
"learning_rate": 1.9231678486997636e-05,
|
| 598 |
+
"loss": 1.3457,
|
| 599 |
+
"step": 2025
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.726546906187625,
|
| 603 |
+
"grad_norm": 43.02334213256836,
|
| 604 |
+
"learning_rate": 1.919473995271868e-05,
|
| 605 |
+
"loss": 1.0394,
|
| 606 |
+
"step": 2050
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.7598137059214904,
|
| 610 |
+
"grad_norm": 42.223323822021484,
|
| 611 |
+
"learning_rate": 1.9157801418439718e-05,
|
| 612 |
+
"loss": 1.389,
|
| 613 |
+
"step": 2075
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.793080505655356,
|
| 617 |
+
"grad_norm": 35.94285583496094,
|
| 618 |
+
"learning_rate": 1.9120862884160757e-05,
|
| 619 |
+
"loss": 1.2184,
|
| 620 |
+
"step": 2100
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.8263473053892216,
|
| 624 |
+
"grad_norm": 27.786062240600586,
|
| 625 |
+
"learning_rate": 1.9083924349881797e-05,
|
| 626 |
+
"loss": 1.2821,
|
| 627 |
+
"step": 2125
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.8596141051230872,
|
| 631 |
+
"grad_norm": 22.151962280273438,
|
| 632 |
+
"learning_rate": 1.904698581560284e-05,
|
| 633 |
+
"loss": 1.2209,
|
| 634 |
+
"step": 2150
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.892880904856953,
|
| 638 |
+
"grad_norm": 26.30831527709961,
|
| 639 |
+
"learning_rate": 1.9010047281323878e-05,
|
| 640 |
+
"loss": 1.306,
|
| 641 |
+
"step": 2175
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.9261477045908184,
|
| 645 |
+
"grad_norm": 21.87983512878418,
|
| 646 |
+
"learning_rate": 1.897310874704492e-05,
|
| 647 |
+
"loss": 1.327,
|
| 648 |
+
"step": 2200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.959414504324684,
|
| 652 |
+
"grad_norm": 33.40931701660156,
|
| 653 |
+
"learning_rate": 1.893617021276596e-05,
|
| 654 |
+
"loss": 1.11,
|
| 655 |
+
"step": 2225
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.9926813040585496,
|
| 659 |
+
"grad_norm": 22.679718017578125,
|
| 660 |
+
"learning_rate": 1.8899231678487e-05,
|
| 661 |
+
"loss": 1.3512,
|
| 662 |
+
"step": 2250
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.0,
|
| 666 |
+
"eval_accuracy": 0.8648952095808383,
|
| 667 |
+
"eval_loss": 0.3499581217765808,
|
| 668 |
+
"eval_macro_f1": 0.6764172148942992,
|
| 669 |
+
"eval_runtime": 1.5942,
|
| 670 |
+
"eval_samples_per_second": 1676.096,
|
| 671 |
+
"eval_steps_per_second": 104.756,
|
| 672 |
+
"eval_weighted_f1": 0.8652442921890966,
|
| 673 |
+
"step": 2256
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.0252827677977376,
|
| 677 |
+
"grad_norm": 34.03767776489258,
|
| 678 |
+
"learning_rate": 1.886229314420804e-05,
|
| 679 |
+
"loss": 1.139,
|
| 680 |
+
"step": 2275
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.0585495675316037,
|
| 684 |
+
"grad_norm": 19.097692489624023,
|
| 685 |
+
"learning_rate": 1.8825354609929078e-05,
|
| 686 |
+
"loss": 0.9606,
|
| 687 |
+
"step": 2300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.0918163672654693,
|
| 691 |
+
"grad_norm": 26.26155662536621,
|
| 692 |
+
"learning_rate": 1.878841607565012e-05,
|
| 693 |
+
"loss": 0.9136,
|
| 694 |
+
"step": 2325
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.125083166999335,
|
| 698 |
+
"grad_norm": 29.003942489624023,
|
| 699 |
+
"learning_rate": 1.875147754137116e-05,
|
| 700 |
+
"loss": 1.0377,
|
| 701 |
+
"step": 2350
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.1583499667332005,
|
| 705 |
+
"grad_norm": 55.168212890625,
|
| 706 |
+
"learning_rate": 1.8714539007092202e-05,
|
| 707 |
+
"loss": 0.9944,
|
| 708 |
+
"step": 2375
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.191616766467066,
|
| 712 |
+
"grad_norm": 31.195945739746094,
|
| 713 |
+
"learning_rate": 1.867760047281324e-05,
|
| 714 |
+
"loss": 0.9626,
|
| 715 |
+
"step": 2400
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.2248835662009316,
|
| 719 |
+
"grad_norm": 30.541507720947266,
|
| 720 |
+
"learning_rate": 1.864066193853428e-05,
|
| 721 |
+
"loss": 0.9487,
|
| 722 |
+
"step": 2425
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.2581503659347972,
|
| 726 |
+
"grad_norm": 32.93328857421875,
|
| 727 |
+
"learning_rate": 1.860372340425532e-05,
|
| 728 |
+
"loss": 1.0851,
|
| 729 |
+
"step": 2450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.291417165668663,
|
| 733 |
+
"grad_norm": 27.888490676879883,
|
| 734 |
+
"learning_rate": 1.856678486997636e-05,
|
| 735 |
+
"loss": 1.1047,
|
| 736 |
+
"step": 2475
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.3246839654025284,
|
| 740 |
+
"grad_norm": 24.827001571655273,
|
| 741 |
+
"learning_rate": 1.85298463356974e-05,
|
| 742 |
+
"loss": 0.9216,
|
| 743 |
+
"step": 2500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.357950765136394,
|
| 747 |
+
"grad_norm": 38.58675765991211,
|
| 748 |
+
"learning_rate": 1.849290780141844e-05,
|
| 749 |
+
"loss": 0.9583,
|
| 750 |
+
"step": 2525
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.3912175648702596,
|
| 754 |
+
"grad_norm": 38.48832702636719,
|
| 755 |
+
"learning_rate": 1.8455969267139483e-05,
|
| 756 |
+
"loss": 1.0392,
|
| 757 |
+
"step": 2550
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.424484364604125,
|
| 761 |
+
"grad_norm": 36.60333251953125,
|
| 762 |
+
"learning_rate": 1.8419030732860522e-05,
|
| 763 |
+
"loss": 0.9494,
|
| 764 |
+
"step": 2575
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.457751164337991,
|
| 768 |
+
"grad_norm": 29.002120971679688,
|
| 769 |
+
"learning_rate": 1.838209219858156e-05,
|
| 770 |
+
"loss": 0.9332,
|
| 771 |
+
"step": 2600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.4910179640718564,
|
| 775 |
+
"grad_norm": 21.734298706054688,
|
| 776 |
+
"learning_rate": 1.83451536643026e-05,
|
| 777 |
+
"loss": 1.2701,
|
| 778 |
+
"step": 2625
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.524284763805722,
|
| 782 |
+
"grad_norm": 21.092363357543945,
|
| 783 |
+
"learning_rate": 1.8308215130023643e-05,
|
| 784 |
+
"loss": 1.1146,
|
| 785 |
+
"step": 2650
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.5575515635395876,
|
| 789 |
+
"grad_norm": 17.316835403442383,
|
| 790 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 791 |
+
"loss": 0.9659,
|
| 792 |
+
"step": 2675
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.590818363273453,
|
| 796 |
+
"grad_norm": 42.5993766784668,
|
| 797 |
+
"learning_rate": 1.823433806146572e-05,
|
| 798 |
+
"loss": 1.1817,
|
| 799 |
+
"step": 2700
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.6240851630073188,
|
| 803 |
+
"grad_norm": 47.7999382019043,
|
| 804 |
+
"learning_rate": 1.8197399527186764e-05,
|
| 805 |
+
"loss": 1.0463,
|
| 806 |
+
"step": 2725
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.6573519627411843,
|
| 810 |
+
"grad_norm": 23.832273483276367,
|
| 811 |
+
"learning_rate": 1.8160460992907803e-05,
|
| 812 |
+
"loss": 0.9641,
|
| 813 |
+
"step": 2750
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.69061876247505,
|
| 817 |
+
"grad_norm": 18.98507308959961,
|
| 818 |
+
"learning_rate": 1.8123522458628842e-05,
|
| 819 |
+
"loss": 1.0791,
|
| 820 |
+
"step": 2775
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.7238855622089155,
|
| 824 |
+
"grad_norm": 30.13364028930664,
|
| 825 |
+
"learning_rate": 1.808658392434988e-05,
|
| 826 |
+
"loss": 0.8868,
|
| 827 |
+
"step": 2800
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.757152361942781,
|
| 831 |
+
"grad_norm": 32.605506896972656,
|
| 832 |
+
"learning_rate": 1.8049645390070924e-05,
|
| 833 |
+
"loss": 1.0201,
|
| 834 |
+
"step": 2825
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.7904191616766467,
|
| 838 |
+
"grad_norm": 27.201099395751953,
|
| 839 |
+
"learning_rate": 1.8012706855791963e-05,
|
| 840 |
+
"loss": 0.8997,
|
| 841 |
+
"step": 2850
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.8236859614105123,
|
| 845 |
+
"grad_norm": 49.87832260131836,
|
| 846 |
+
"learning_rate": 1.7975768321513003e-05,
|
| 847 |
+
"loss": 1.1678,
|
| 848 |
+
"step": 2875
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.856952761144378,
|
| 852 |
+
"grad_norm": 34.64746856689453,
|
| 853 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 854 |
+
"loss": 0.9894,
|
| 855 |
+
"step": 2900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.8902195608782435,
|
| 859 |
+
"grad_norm": 31.959882736206055,
|
| 860 |
+
"learning_rate": 1.7901891252955084e-05,
|
| 861 |
+
"loss": 0.9329,
|
| 862 |
+
"step": 2925
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.923486360612109,
|
| 866 |
+
"grad_norm": 14.112442016601562,
|
| 867 |
+
"learning_rate": 1.7864952718676124e-05,
|
| 868 |
+
"loss": 1.0693,
|
| 869 |
+
"step": 2950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.9567531603459747,
|
| 873 |
+
"grad_norm": 21.995716094970703,
|
| 874 |
+
"learning_rate": 1.7828014184397163e-05,
|
| 875 |
+
"loss": 1.0372,
|
| 876 |
+
"step": 2975
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.9900199600798403,
|
| 880 |
+
"grad_norm": 28.292930603027344,
|
| 881 |
+
"learning_rate": 1.7791075650118205e-05,
|
| 882 |
+
"loss": 1.0076,
|
| 883 |
+
"step": 3000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 4.0,
|
| 887 |
+
"eval_accuracy": 0.8682634730538922,
|
| 888 |
+
"eval_loss": 0.3861147463321686,
|
| 889 |
+
"eval_macro_f1": 0.6498392190640252,
|
| 890 |
+
"eval_runtime": 1.507,
|
| 891 |
+
"eval_samples_per_second": 1773.082,
|
| 892 |
+
"eval_steps_per_second": 110.818,
|
| 893 |
+
"eval_weighted_f1": 0.8641321592695009,
|
| 894 |
+
"step": 3008
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.022621423819029,
|
| 898 |
+
"grad_norm": 34.148014068603516,
|
| 899 |
+
"learning_rate": 1.7754137115839245e-05,
|
| 900 |
+
"loss": 0.8104,
|
| 901 |
+
"step": 3025
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 4.055888223552894,
|
| 905 |
+
"grad_norm": 23.081899642944336,
|
| 906 |
+
"learning_rate": 1.7717198581560284e-05,
|
| 907 |
+
"loss": 0.7902,
|
| 908 |
+
"step": 3050
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 4.08915502328676,
|
| 912 |
+
"grad_norm": 27.10505485534668,
|
| 913 |
+
"learning_rate": 1.7680260047281326e-05,
|
| 914 |
+
"loss": 0.8715,
|
| 915 |
+
"step": 3075
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.122421823020625,
|
| 919 |
+
"grad_norm": 53.181549072265625,
|
| 920 |
+
"learning_rate": 1.7643321513002365e-05,
|
| 921 |
+
"loss": 0.777,
|
| 922 |
+
"step": 3100
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 4.155688622754491,
|
| 926 |
+
"grad_norm": 32.25143814086914,
|
| 927 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 928 |
+
"loss": 0.7971,
|
| 929 |
+
"step": 3125
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 4.188955422488356,
|
| 933 |
+
"grad_norm": 38.85936737060547,
|
| 934 |
+
"learning_rate": 1.7569444444444447e-05,
|
| 935 |
+
"loss": 0.8528,
|
| 936 |
+
"step": 3150
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 4.222222222222222,
|
| 940 |
+
"grad_norm": 12.436830520629883,
|
| 941 |
+
"learning_rate": 1.7532505910165486e-05,
|
| 942 |
+
"loss": 0.6806,
|
| 943 |
+
"step": 3175
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 4.2554890219560875,
|
| 947 |
+
"grad_norm": 35.394046783447266,
|
| 948 |
+
"learning_rate": 1.7495567375886526e-05,
|
| 949 |
+
"loss": 0.7414,
|
| 950 |
+
"step": 3200
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 4.2887558216899535,
|
| 954 |
+
"grad_norm": 44.16730499267578,
|
| 955 |
+
"learning_rate": 1.7458628841607565e-05,
|
| 956 |
+
"loss": 0.8936,
|
| 957 |
+
"step": 3225
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 4.322022621423819,
|
| 961 |
+
"grad_norm": 25.659095764160156,
|
| 962 |
+
"learning_rate": 1.7421690307328607e-05,
|
| 963 |
+
"loss": 0.8213,
|
| 964 |
+
"step": 3250
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 4.355289421157685,
|
| 968 |
+
"grad_norm": 77.10603332519531,
|
| 969 |
+
"learning_rate": 1.7384751773049647e-05,
|
| 970 |
+
"loss": 0.7927,
|
| 971 |
+
"step": 3275
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 4.38855622089155,
|
| 975 |
+
"grad_norm": 30.05143928527832,
|
| 976 |
+
"learning_rate": 1.734781323877069e-05,
|
| 977 |
+
"loss": 0.7544,
|
| 978 |
+
"step": 3300
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.421823020625416,
|
| 982 |
+
"grad_norm": 46.79085922241211,
|
| 983 |
+
"learning_rate": 1.731087470449173e-05,
|
| 984 |
+
"loss": 0.9205,
|
| 985 |
+
"step": 3325
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 4.455089820359281,
|
| 989 |
+
"grad_norm": 40.14250946044922,
|
| 990 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 991 |
+
"loss": 0.8438,
|
| 992 |
+
"step": 3350
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.488356620093147,
|
| 996 |
+
"grad_norm": 28.33310317993164,
|
| 997 |
+
"learning_rate": 1.7236997635933807e-05,
|
| 998 |
+
"loss": 0.7885,
|
| 999 |
+
"step": 3375
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.521623419827012,
|
| 1003 |
+
"grad_norm": 32.68979263305664,
|
| 1004 |
+
"learning_rate": 1.7200059101654846e-05,
|
| 1005 |
+
"loss": 1.0292,
|
| 1006 |
+
"step": 3400
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.554890219560878,
|
| 1010 |
+
"grad_norm": 26.344491958618164,
|
| 1011 |
+
"learning_rate": 1.716312056737589e-05,
|
| 1012 |
+
"loss": 0.803,
|
| 1013 |
+
"step": 3425
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.588157019294744,
|
| 1017 |
+
"grad_norm": 14.852431297302246,
|
| 1018 |
+
"learning_rate": 1.7126182033096928e-05,
|
| 1019 |
+
"loss": 0.7835,
|
| 1020 |
+
"step": 3450
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.621423819028609,
|
| 1024 |
+
"grad_norm": 41.625064849853516,
|
| 1025 |
+
"learning_rate": 1.708924349881797e-05,
|
| 1026 |
+
"loss": 0.9212,
|
| 1027 |
+
"step": 3475
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 4.654690618762475,
|
| 1031 |
+
"grad_norm": 27.802034378051758,
|
| 1032 |
+
"learning_rate": 1.705230496453901e-05,
|
| 1033 |
+
"loss": 0.839,
|
| 1034 |
+
"step": 3500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.687957418496341,
|
| 1038 |
+
"grad_norm": 68.59847259521484,
|
| 1039 |
+
"learning_rate": 1.701536643026005e-05,
|
| 1040 |
+
"loss": 0.7907,
|
| 1041 |
+
"step": 3525
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.721224218230207,
|
| 1045 |
+
"grad_norm": 61.745582580566406,
|
| 1046 |
+
"learning_rate": 1.6978427895981088e-05,
|
| 1047 |
+
"loss": 0.7641,
|
| 1048 |
+
"step": 3550
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.754491017964072,
|
| 1052 |
+
"grad_norm": 39.64849853515625,
|
| 1053 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 1054 |
+
"loss": 0.6292,
|
| 1055 |
+
"step": 3575
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.787757817697938,
|
| 1059 |
+
"grad_norm": 38.32709503173828,
|
| 1060 |
+
"learning_rate": 1.690455082742317e-05,
|
| 1061 |
+
"loss": 0.9191,
|
| 1062 |
+
"step": 3600
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.821024617431803,
|
| 1066 |
+
"grad_norm": 21.010385513305664,
|
| 1067 |
+
"learning_rate": 1.686761229314421e-05,
|
| 1068 |
+
"loss": 0.8335,
|
| 1069 |
+
"step": 3625
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.854291417165669,
|
| 1073 |
+
"grad_norm": 24.61604118347168,
|
| 1074 |
+
"learning_rate": 1.683067375886525e-05,
|
| 1075 |
+
"loss": 0.9304,
|
| 1076 |
+
"step": 3650
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.887558216899534,
|
| 1080 |
+
"grad_norm": 70.51045227050781,
|
| 1081 |
+
"learning_rate": 1.679373522458629e-05,
|
| 1082 |
+
"loss": 0.7923,
|
| 1083 |
+
"step": 3675
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.9208250166334,
|
| 1087 |
+
"grad_norm": 38.28218078613281,
|
| 1088 |
+
"learning_rate": 1.675679669030733e-05,
|
| 1089 |
+
"loss": 0.7386,
|
| 1090 |
+
"step": 3700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.954091816367265,
|
| 1094 |
+
"grad_norm": 20.940914154052734,
|
| 1095 |
+
"learning_rate": 1.671985815602837e-05,
|
| 1096 |
+
"loss": 0.6568,
|
| 1097 |
+
"step": 3725
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.987358616101131,
|
| 1101 |
+
"grad_norm": 31.858694076538086,
|
| 1102 |
+
"learning_rate": 1.6682919621749408e-05,
|
| 1103 |
+
"loss": 0.9504,
|
| 1104 |
+
"step": 3750
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.0,
|
| 1108 |
+
"eval_accuracy": 0.8656437125748503,
|
| 1109 |
+
"eval_loss": 0.3926589787006378,
|
| 1110 |
+
"eval_macro_f1": 0.6748486242267919,
|
| 1111 |
+
"eval_runtime": 1.5,
|
| 1112 |
+
"eval_samples_per_second": 1781.337,
|
| 1113 |
+
"eval_steps_per_second": 111.334,
|
| 1114 |
+
"eval_weighted_f1": 0.8669454821698057,
|
| 1115 |
+
"step": 3760
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.0199600798403194,
|
| 1119 |
+
"grad_norm": 37.97001266479492,
|
| 1120 |
+
"learning_rate": 1.664598108747045e-05,
|
| 1121 |
+
"loss": 0.7584,
|
| 1122 |
+
"step": 3775
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.053226879574185,
|
| 1126 |
+
"grad_norm": 31.30186653137207,
|
| 1127 |
+
"learning_rate": 1.660904255319149e-05,
|
| 1128 |
+
"loss": 0.6099,
|
| 1129 |
+
"step": 3800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.086493679308051,
|
| 1133 |
+
"grad_norm": 27.994447708129883,
|
| 1134 |
+
"learning_rate": 1.6572104018912532e-05,
|
| 1135 |
+
"loss": 0.4493,
|
| 1136 |
+
"step": 3825
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.119760479041916,
|
| 1140 |
+
"grad_norm": 17.40340232849121,
|
| 1141 |
+
"learning_rate": 1.653516548463357e-05,
|
| 1142 |
+
"loss": 0.655,
|
| 1143 |
+
"step": 3850
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.153027278775782,
|
| 1147 |
+
"grad_norm": 27.897592544555664,
|
| 1148 |
+
"learning_rate": 1.649822695035461e-05,
|
| 1149 |
+
"loss": 0.7556,
|
| 1150 |
+
"step": 3875
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.186294078509647,
|
| 1154 |
+
"grad_norm": 45.2603759765625,
|
| 1155 |
+
"learning_rate": 1.646128841607565e-05,
|
| 1156 |
+
"loss": 0.6634,
|
| 1157 |
+
"step": 3900
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.219560878243513,
|
| 1161 |
+
"grad_norm": 40.82050323486328,
|
| 1162 |
+
"learning_rate": 1.642434988179669e-05,
|
| 1163 |
+
"loss": 0.6573,
|
| 1164 |
+
"step": 3925
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.252827677977379,
|
| 1168 |
+
"grad_norm": 34.0108757019043,
|
| 1169 |
+
"learning_rate": 1.6387411347517732e-05,
|
| 1170 |
+
"loss": 0.7301,
|
| 1171 |
+
"step": 3950
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.286094477711244,
|
| 1175 |
+
"grad_norm": 52.25631332397461,
|
| 1176 |
+
"learning_rate": 1.635047281323877e-05,
|
| 1177 |
+
"loss": 0.5258,
|
| 1178 |
+
"step": 3975
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.319361277445109,
|
| 1182 |
+
"grad_norm": 33.63074493408203,
|
| 1183 |
+
"learning_rate": 1.6313534278959813e-05,
|
| 1184 |
+
"loss": 0.5966,
|
| 1185 |
+
"step": 4000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.352628077178975,
|
| 1189 |
+
"grad_norm": 18.274282455444336,
|
| 1190 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 1191 |
+
"loss": 0.5739,
|
| 1192 |
+
"step": 4025
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.385894876912841,
|
| 1196 |
+
"grad_norm": 20.909400939941406,
|
| 1197 |
+
"learning_rate": 1.6239657210401892e-05,
|
| 1198 |
+
"loss": 0.4991,
|
| 1199 |
+
"step": 4050
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.419161676646707,
|
| 1203 |
+
"grad_norm": 10.55371379852295,
|
| 1204 |
+
"learning_rate": 1.620271867612293e-05,
|
| 1205 |
+
"loss": 0.7464,
|
| 1206 |
+
"step": 4075
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.452428476380573,
|
| 1210 |
+
"grad_norm": 49.42185592651367,
|
| 1211 |
+
"learning_rate": 1.6165780141843974e-05,
|
| 1212 |
+
"loss": 0.6877,
|
| 1213 |
+
"step": 4100
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.485695276114438,
|
| 1217 |
+
"grad_norm": 37.38511276245117,
|
| 1218 |
+
"learning_rate": 1.6128841607565013e-05,
|
| 1219 |
+
"loss": 0.6816,
|
| 1220 |
+
"step": 4125
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.518962075848304,
|
| 1224 |
+
"grad_norm": 6.26555871963501,
|
| 1225 |
+
"learning_rate": 1.6091903073286052e-05,
|
| 1226 |
+
"loss": 0.5693,
|
| 1227 |
+
"step": 4150
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.552228875582169,
|
| 1231 |
+
"grad_norm": 23.01319122314453,
|
| 1232 |
+
"learning_rate": 1.6054964539007095e-05,
|
| 1233 |
+
"loss": 0.792,
|
| 1234 |
+
"step": 4175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.585495675316035,
|
| 1238 |
+
"grad_norm": 48.40448760986328,
|
| 1239 |
+
"learning_rate": 1.6018026004728134e-05,
|
| 1240 |
+
"loss": 0.6381,
|
| 1241 |
+
"step": 4200
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.6187624750499,
|
| 1245 |
+
"grad_norm": 36.653079986572266,
|
| 1246 |
+
"learning_rate": 1.5981087470449176e-05,
|
| 1247 |
+
"loss": 0.7671,
|
| 1248 |
+
"step": 4225
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.652029274783766,
|
| 1252 |
+
"grad_norm": 27.41105842590332,
|
| 1253 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 1254 |
+
"loss": 0.6985,
|
| 1255 |
+
"step": 4250
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.685296074517631,
|
| 1259 |
+
"grad_norm": 66.96533966064453,
|
| 1260 |
+
"learning_rate": 1.5907210401891255e-05,
|
| 1261 |
+
"loss": 0.7069,
|
| 1262 |
+
"step": 4275
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.718562874251497,
|
| 1266 |
+
"grad_norm": 53.56682586669922,
|
| 1267 |
+
"learning_rate": 1.5870271867612294e-05,
|
| 1268 |
+
"loss": 0.6558,
|
| 1269 |
+
"step": 4300
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.7518296739853625,
|
| 1273 |
+
"grad_norm": 31.27998924255371,
|
| 1274 |
+
"learning_rate": 1.5833333333333333e-05,
|
| 1275 |
+
"loss": 0.6231,
|
| 1276 |
+
"step": 4325
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.7850964737192285,
|
| 1280 |
+
"grad_norm": 33.13193893432617,
|
| 1281 |
+
"learning_rate": 1.5796394799054376e-05,
|
| 1282 |
+
"loss": 0.7823,
|
| 1283 |
+
"step": 4350
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.818363273453094,
|
| 1287 |
+
"grad_norm": 25.645647048950195,
|
| 1288 |
+
"learning_rate": 1.5759456264775415e-05,
|
| 1289 |
+
"loss": 0.5854,
|
| 1290 |
+
"step": 4375
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.85163007318696,
|
| 1294 |
+
"grad_norm": 76.89608764648438,
|
| 1295 |
+
"learning_rate": 1.5722517730496457e-05,
|
| 1296 |
+
"loss": 0.8529,
|
| 1297 |
+
"step": 4400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.884896872920825,
|
| 1301 |
+
"grad_norm": 40.57425308227539,
|
| 1302 |
+
"learning_rate": 1.5685579196217497e-05,
|
| 1303 |
+
"loss": 0.6377,
|
| 1304 |
+
"step": 4425
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.918163672654691,
|
| 1308 |
+
"grad_norm": 7.283412456512451,
|
| 1309 |
+
"learning_rate": 1.5648640661938536e-05,
|
| 1310 |
+
"loss": 0.5438,
|
| 1311 |
+
"step": 4450
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.951430472388556,
|
| 1315 |
+
"grad_norm": 66.2386474609375,
|
| 1316 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 1317 |
+
"loss": 0.7596,
|
| 1318 |
+
"step": 4475
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.984697272122422,
|
| 1322 |
+
"grad_norm": 11.451156616210938,
|
| 1323 |
+
"learning_rate": 1.5574763593380614e-05,
|
| 1324 |
+
"loss": 0.7874,
|
| 1325 |
+
"step": 4500
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.0,
|
| 1329 |
+
"eval_accuracy": 0.8735029940119761,
|
| 1330 |
+
"eval_loss": 0.4388781189918518,
|
| 1331 |
+
"eval_macro_f1": 0.6827700263390525,
|
| 1332 |
+
"eval_runtime": 1.5303,
|
| 1333 |
+
"eval_samples_per_second": 1746.115,
|
| 1334 |
+
"eval_steps_per_second": 109.132,
|
| 1335 |
+
"eval_weighted_f1": 0.8711094177585016,
|
| 1336 |
+
"step": 4512
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 6.01729873586161,
|
| 1340 |
+
"grad_norm": 67.75423431396484,
|
| 1341 |
+
"learning_rate": 1.5537825059101657e-05,
|
| 1342 |
+
"loss": 0.5698,
|
| 1343 |
+
"step": 4525
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 6.050565535595475,
|
| 1347 |
+
"grad_norm": 7.571048736572266,
|
| 1348 |
+
"learning_rate": 1.5500886524822696e-05,
|
| 1349 |
+
"loss": 0.3304,
|
| 1350 |
+
"step": 4550
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 6.083832335329341,
|
| 1354 |
+
"grad_norm": 55.4453239440918,
|
| 1355 |
+
"learning_rate": 1.546394799054374e-05,
|
| 1356 |
+
"loss": 0.6612,
|
| 1357 |
+
"step": 4575
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 6.117099135063207,
|
| 1361 |
+
"grad_norm": 79.63700103759766,
|
| 1362 |
+
"learning_rate": 1.5427009456264778e-05,
|
| 1363 |
+
"loss": 0.3504,
|
| 1364 |
+
"step": 4600
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 6.1503659347970725,
|
| 1368 |
+
"grad_norm": 49.47695541381836,
|
| 1369 |
+
"learning_rate": 1.5390070921985817e-05,
|
| 1370 |
+
"loss": 0.5078,
|
| 1371 |
+
"step": 4625
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 6.1836327345309385,
|
| 1375 |
+
"grad_norm": 49.128597259521484,
|
| 1376 |
+
"learning_rate": 1.5353132387706856e-05,
|
| 1377 |
+
"loss": 0.5192,
|
| 1378 |
+
"step": 4650
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 6.216899534264804,
|
| 1382 |
+
"grad_norm": 19.36261749267578,
|
| 1383 |
+
"learning_rate": 1.5316193853427895e-05,
|
| 1384 |
+
"loss": 0.6108,
|
| 1385 |
+
"step": 4675
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 6.25016633399867,
|
| 1389 |
+
"grad_norm": 50.89854431152344,
|
| 1390 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 1391 |
+
"loss": 0.6622,
|
| 1392 |
+
"step": 4700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 6.283433133732535,
|
| 1396 |
+
"grad_norm": 57.39405059814453,
|
| 1397 |
+
"learning_rate": 1.5242316784869977e-05,
|
| 1398 |
+
"loss": 0.5008,
|
| 1399 |
+
"step": 4725
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 6.316699933466401,
|
| 1403 |
+
"grad_norm": 83.19384002685547,
|
| 1404 |
+
"learning_rate": 1.5205378250591018e-05,
|
| 1405 |
+
"loss": 0.6187,
|
| 1406 |
+
"step": 4750
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 6.349966733200266,
|
| 1410 |
+
"grad_norm": 30.342533111572266,
|
| 1411 |
+
"learning_rate": 1.5168439716312057e-05,
|
| 1412 |
+
"loss": 0.5589,
|
| 1413 |
+
"step": 4775
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 6.383233532934132,
|
| 1417 |
+
"grad_norm": 9.4254150390625,
|
| 1418 |
+
"learning_rate": 1.51315011820331e-05,
|
| 1419 |
+
"loss": 0.4618,
|
| 1420 |
+
"step": 4800
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 6.416500332667997,
|
| 1424 |
+
"grad_norm": 9.256645202636719,
|
| 1425 |
+
"learning_rate": 1.5094562647754139e-05,
|
| 1426 |
+
"loss": 0.5257,
|
| 1427 |
+
"step": 4825
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 6.449767132401863,
|
| 1431 |
+
"grad_norm": 10.17599105834961,
|
| 1432 |
+
"learning_rate": 1.5057624113475178e-05,
|
| 1433 |
+
"loss": 0.4557,
|
| 1434 |
+
"step": 4850
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 6.483033932135728,
|
| 1438 |
+
"grad_norm": 40.882362365722656,
|
| 1439 |
+
"learning_rate": 1.5020685579196219e-05,
|
| 1440 |
+
"loss": 0.7595,
|
| 1441 |
+
"step": 4875
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 6.5163007318695945,
|
| 1445 |
+
"grad_norm": 73.62734985351562,
|
| 1446 |
+
"learning_rate": 1.4983747044917258e-05,
|
| 1447 |
+
"loss": 0.7191,
|
| 1448 |
+
"step": 4900
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 6.54956753160346,
|
| 1452 |
+
"grad_norm": 45.64955520629883,
|
| 1453 |
+
"learning_rate": 1.49468085106383e-05,
|
| 1454 |
+
"loss": 0.5214,
|
| 1455 |
+
"step": 4925
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 6.582834331337326,
|
| 1459 |
+
"grad_norm": 85.1973648071289,
|
| 1460 |
+
"learning_rate": 1.490986997635934e-05,
|
| 1461 |
+
"loss": 0.6561,
|
| 1462 |
+
"step": 4950
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 6.616101131071191,
|
| 1466 |
+
"grad_norm": 55.2785530090332,
|
| 1467 |
+
"learning_rate": 1.487293144208038e-05,
|
| 1468 |
+
"loss": 0.4211,
|
| 1469 |
+
"step": 4975
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 6.649367930805057,
|
| 1473 |
+
"grad_norm": 75.65742492675781,
|
| 1474 |
+
"learning_rate": 1.483599290780142e-05,
|
| 1475 |
+
"loss": 0.5717,
|
| 1476 |
+
"step": 5000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 6.682634730538922,
|
| 1480 |
+
"grad_norm": 108.11344909667969,
|
| 1481 |
+
"learning_rate": 1.4799054373522459e-05,
|
| 1482 |
+
"loss": 0.4643,
|
| 1483 |
+
"step": 5025
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 6.715901530272788,
|
| 1487 |
+
"grad_norm": 37.75810241699219,
|
| 1488 |
+
"learning_rate": 1.47621158392435e-05,
|
| 1489 |
+
"loss": 0.5376,
|
| 1490 |
+
"step": 5050
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.749168330006653,
|
| 1494 |
+
"grad_norm": 55.700927734375,
|
| 1495 |
+
"learning_rate": 1.472517730496454e-05,
|
| 1496 |
+
"loss": 0.6029,
|
| 1497 |
+
"step": 5075
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 6.782435129740519,
|
| 1501 |
+
"grad_norm": 35.9907341003418,
|
| 1502 |
+
"learning_rate": 1.4688238770685582e-05,
|
| 1503 |
+
"loss": 0.5956,
|
| 1504 |
+
"step": 5100
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 6.815701929474384,
|
| 1508 |
+
"grad_norm": 47.296199798583984,
|
| 1509 |
+
"learning_rate": 1.4651300236406621e-05,
|
| 1510 |
+
"loss": 0.494,
|
| 1511 |
+
"step": 5125
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 6.84896872920825,
|
| 1515 |
+
"grad_norm": 65.1382827758789,
|
| 1516 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 1517 |
+
"loss": 0.6979,
|
| 1518 |
+
"step": 5150
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 6.8822355289421155,
|
| 1522 |
+
"grad_norm": 80.72993469238281,
|
| 1523 |
+
"learning_rate": 1.4577423167848701e-05,
|
| 1524 |
+
"loss": 0.5972,
|
| 1525 |
+
"step": 5175
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 6.915502328675982,
|
| 1529 |
+
"grad_norm": 103.25460052490234,
|
| 1530 |
+
"learning_rate": 1.454048463356974e-05,
|
| 1531 |
+
"loss": 0.5676,
|
| 1532 |
+
"step": 5200
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 6.948769128409847,
|
| 1536 |
+
"grad_norm": 34.944541931152344,
|
| 1537 |
+
"learning_rate": 1.4503546099290781e-05,
|
| 1538 |
+
"loss": 0.6839,
|
| 1539 |
+
"step": 5225
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 6.982035928143713,
|
| 1543 |
+
"grad_norm": 28.08905601501465,
|
| 1544 |
+
"learning_rate": 1.446660756501182e-05,
|
| 1545 |
+
"loss": 0.6631,
|
| 1546 |
+
"step": 5250
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 7.0,
|
| 1550 |
+
"eval_accuracy": 0.8764970059880239,
|
| 1551 |
+
"eval_loss": 0.4612451493740082,
|
| 1552 |
+
"eval_macro_f1": 0.6915236970070544,
|
| 1553 |
+
"eval_runtime": 1.5249,
|
| 1554 |
+
"eval_samples_per_second": 1752.252,
|
| 1555 |
+
"eval_steps_per_second": 109.516,
|
| 1556 |
+
"eval_weighted_f1": 0.8740941266993443,
|
| 1557 |
+
"step": 5264
|
| 1558 |
+
}
|
| 1559 |
+
],
|
| 1560 |
+
"logging_steps": 25,
|
| 1561 |
+
"max_steps": 15040,
|
| 1562 |
+
"num_input_tokens_seen": 0,
|
| 1563 |
+
"num_train_epochs": 20,
|
| 1564 |
+
"save_steps": 500,
|
| 1565 |
+
"stateful_callbacks": {
|
| 1566 |
+
"EarlyStoppingCallback": {
|
| 1567 |
+
"args": {
|
| 1568 |
+
"early_stopping_patience": 3,
|
| 1569 |
+
"early_stopping_threshold": 0.0
|
| 1570 |
+
},
|
| 1571 |
+
"attributes": {
|
| 1572 |
+
"early_stopping_patience_counter": 0
|
| 1573 |
+
}
|
| 1574 |
+
},
|
| 1575 |
+
"TrainerControl": {
|
| 1576 |
+
"args": {
|
| 1577 |
+
"should_epoch_stop": false,
|
| 1578 |
+
"should_evaluate": false,
|
| 1579 |
+
"should_log": false,
|
| 1580 |
+
"should_save": true,
|
| 1581 |
+
"should_training_stop": false
|
| 1582 |
+
},
|
| 1583 |
+
"attributes": {}
|
| 1584 |
+
}
|
| 1585 |
+
},
|
| 1586 |
+
"total_flos": 0.0,
|
| 1587 |
+
"train_batch_size": 8,
|
| 1588 |
+
"trial_name": null,
|
| 1589 |
+
"trial_params": null
|
| 1590 |
+
}
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-7520/trainer_state.json
ADDED
|
@@ -0,0 +1,2253 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 5264,
|
| 3 |
+
"best_metric": 0.6915236970070544,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vihsd-bamibert-s42-20260827T123619Z-f3636918/trainer/checkpoint-5264",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 7520,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0332667997338656,
|
| 14 |
+
"grad_norm": 37.213836669921875,
|
| 15 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 16 |
+
"loss": 4.1936,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0665335994677312,
|
| 21 |
+
"grad_norm": 52.82760238647461,
|
| 22 |
+
"learning_rate": 6.515957446808512e-07,
|
| 23 |
+
"loss": 3.937,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0998003992015968,
|
| 28 |
+
"grad_norm": 129.2099151611328,
|
| 29 |
+
"learning_rate": 9.840425531914895e-07,
|
| 30 |
+
"loss": 3.4919,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1330671989354624,
|
| 35 |
+
"grad_norm": 22.686023712158203,
|
| 36 |
+
"learning_rate": 1.3164893617021277e-06,
|
| 37 |
+
"loss": 3.0222,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16633399866932802,
|
| 42 |
+
"grad_norm": 57.567840576171875,
|
| 43 |
+
"learning_rate": 1.648936170212766e-06,
|
| 44 |
+
"loss": 2.5057,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1996007984031936,
|
| 49 |
+
"grad_norm": 99.45436096191406,
|
| 50 |
+
"learning_rate": 1.9813829787234043e-06,
|
| 51 |
+
"loss": 2.3339,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2328675981370592,
|
| 56 |
+
"grad_norm": 13.004047393798828,
|
| 57 |
+
"learning_rate": 2.3138297872340428e-06,
|
| 58 |
+
"loss": 2.1973,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2661343978709248,
|
| 63 |
+
"grad_norm": 29.448951721191406,
|
| 64 |
+
"learning_rate": 2.646276595744681e-06,
|
| 65 |
+
"loss": 2.2666,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.2994011976047904,
|
| 70 |
+
"grad_norm": 16.099824905395508,
|
| 71 |
+
"learning_rate": 2.978723404255319e-06,
|
| 72 |
+
"loss": 2.1449,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.33266799733865604,
|
| 77 |
+
"grad_norm": 28.373823165893555,
|
| 78 |
+
"learning_rate": 3.311170212765958e-06,
|
| 79 |
+
"loss": 2.19,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.36593479707252163,
|
| 84 |
+
"grad_norm": 18.843015670776367,
|
| 85 |
+
"learning_rate": 3.643617021276596e-06,
|
| 86 |
+
"loss": 2.1845,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3992015968063872,
|
| 91 |
+
"grad_norm": 15.624969482421875,
|
| 92 |
+
"learning_rate": 3.976063829787235e-06,
|
| 93 |
+
"loss": 2.2532,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4324683965402528,
|
| 98 |
+
"grad_norm": 30.467538833618164,
|
| 99 |
+
"learning_rate": 4.308510638297873e-06,
|
| 100 |
+
"loss": 1.9754,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4657351962741184,
|
| 105 |
+
"grad_norm": 23.515451431274414,
|
| 106 |
+
"learning_rate": 4.640957446808511e-06,
|
| 107 |
+
"loss": 1.8119,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.499001996007984,
|
| 112 |
+
"grad_norm": 17.149723052978516,
|
| 113 |
+
"learning_rate": 4.973404255319149e-06,
|
| 114 |
+
"loss": 2.0686,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5322687957418496,
|
| 119 |
+
"grad_norm": 24.267086029052734,
|
| 120 |
+
"learning_rate": 5.305851063829788e-06,
|
| 121 |
+
"loss": 1.7767,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5655355954757152,
|
| 126 |
+
"grad_norm": 33.58161163330078,
|
| 127 |
+
"learning_rate": 5.638297872340426e-06,
|
| 128 |
+
"loss": 1.8802,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5988023952095808,
|
| 133 |
+
"grad_norm": 17.654483795166016,
|
| 134 |
+
"learning_rate": 5.970744680851064e-06,
|
| 135 |
+
"loss": 2.0599,
|
| 136 |
+
"step": 450
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6320691949434465,
|
| 140 |
+
"grad_norm": 37.77925491333008,
|
| 141 |
+
"learning_rate": 6.303191489361703e-06,
|
| 142 |
+
"loss": 2.0091,
|
| 143 |
+
"step": 475
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6653359946773121,
|
| 147 |
+
"grad_norm": 8.559550285339355,
|
| 148 |
+
"learning_rate": 6.635638297872341e-06,
|
| 149 |
+
"loss": 1.8012,
|
| 150 |
+
"step": 500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6986027944111777,
|
| 154 |
+
"grad_norm": 16.377788543701172,
|
| 155 |
+
"learning_rate": 6.968085106382979e-06,
|
| 156 |
+
"loss": 1.5796,
|
| 157 |
+
"step": 525
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7318695941450433,
|
| 161 |
+
"grad_norm": 39.23353576660156,
|
| 162 |
+
"learning_rate": 7.300531914893617e-06,
|
| 163 |
+
"loss": 1.7269,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7651363938789089,
|
| 168 |
+
"grad_norm": 33.32931900024414,
|
| 169 |
+
"learning_rate": 7.632978723404256e-06,
|
| 170 |
+
"loss": 1.6038,
|
| 171 |
+
"step": 575
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.7984031936127745,
|
| 175 |
+
"grad_norm": 36.23918914794922,
|
| 176 |
+
"learning_rate": 7.965425531914895e-06,
|
| 177 |
+
"loss": 1.9594,
|
| 178 |
+
"step": 600
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.83166999334664,
|
| 182 |
+
"grad_norm": 28.780132293701172,
|
| 183 |
+
"learning_rate": 8.297872340425532e-06,
|
| 184 |
+
"loss": 1.8884,
|
| 185 |
+
"step": 625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8649367930805056,
|
| 189 |
+
"grad_norm": 32.994388580322266,
|
| 190 |
+
"learning_rate": 8.630319148936171e-06,
|
| 191 |
+
"loss": 1.669,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.8982035928143712,
|
| 196 |
+
"grad_norm": 26.472225189208984,
|
| 197 |
+
"learning_rate": 8.96276595744681e-06,
|
| 198 |
+
"loss": 1.6077,
|
| 199 |
+
"step": 675
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9314703925482368,
|
| 203 |
+
"grad_norm": 29.906286239624023,
|
| 204 |
+
"learning_rate": 9.295212765957447e-06,
|
| 205 |
+
"loss": 1.6924,
|
| 206 |
+
"step": 700
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9647371922821024,
|
| 210 |
+
"grad_norm": 25.40384292602539,
|
| 211 |
+
"learning_rate": 9.627659574468086e-06,
|
| 212 |
+
"loss": 1.7419,
|
| 213 |
+
"step": 725
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.998003992015968,
|
| 217 |
+
"grad_norm": 23.713624954223633,
|
| 218 |
+
"learning_rate": 9.960106382978723e-06,
|
| 219 |
+
"loss": 1.5993,
|
| 220 |
+
"step": 750
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0,
|
| 224 |
+
"eval_accuracy": 0.8480538922155688,
|
| 225 |
+
"eval_loss": 0.42238420248031616,
|
| 226 |
+
"eval_macro_f1": 0.51138076970829,
|
| 227 |
+
"eval_runtime": 1.5008,
|
| 228 |
+
"eval_samples_per_second": 1780.386,
|
| 229 |
+
"eval_steps_per_second": 111.274,
|
| 230 |
+
"eval_weighted_f1": 0.821597408229364,
|
| 231 |
+
"step": 752
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0306054557551563,
|
| 235 |
+
"grad_norm": 32.69313430786133,
|
| 236 |
+
"learning_rate": 1.0292553191489364e-05,
|
| 237 |
+
"loss": 1.6252,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.0638722554890219,
|
| 242 |
+
"grad_norm": 24.044479370117188,
|
| 243 |
+
"learning_rate": 1.0625e-05,
|
| 244 |
+
"loss": 1.5655,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.0971390552228875,
|
| 249 |
+
"grad_norm": 22.281408309936523,
|
| 250 |
+
"learning_rate": 1.095744680851064e-05,
|
| 251 |
+
"loss": 1.6928,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.130405854956753,
|
| 256 |
+
"grad_norm": 41.09510803222656,
|
| 257 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 258 |
+
"loss": 1.5383,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.1636726546906186,
|
| 263 |
+
"grad_norm": 23.410619735717773,
|
| 264 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 265 |
+
"loss": 1.4538,
|
| 266 |
+
"step": 875
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.1969394544244845,
|
| 270 |
+
"grad_norm": 34.236785888671875,
|
| 271 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 272 |
+
"loss": 1.7808,
|
| 273 |
+
"step": 900
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.23020625415835,
|
| 277 |
+
"grad_norm": 33.10630798339844,
|
| 278 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 279 |
+
"loss": 1.5242,
|
| 280 |
+
"step": 925
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.2634730538922156,
|
| 284 |
+
"grad_norm": 26.993452072143555,
|
| 285 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 286 |
+
"loss": 1.4654,
|
| 287 |
+
"step": 950
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.2967398536260812,
|
| 291 |
+
"grad_norm": 25.680158615112305,
|
| 292 |
+
"learning_rate": 1.295212765957447e-05,
|
| 293 |
+
"loss": 1.3583,
|
| 294 |
+
"step": 975
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.3300066533599468,
|
| 298 |
+
"grad_norm": 20.42882537841797,
|
| 299 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 300 |
+
"loss": 1.5349,
|
| 301 |
+
"step": 1000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.3632734530938124,
|
| 305 |
+
"grad_norm": 20.90618133544922,
|
| 306 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 307 |
+
"loss": 1.5473,
|
| 308 |
+
"step": 1025
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.396540252827678,
|
| 312 |
+
"grad_norm": 27.44194793701172,
|
| 313 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 314 |
+
"loss": 1.4876,
|
| 315 |
+
"step": 1050
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.4298070525615436,
|
| 319 |
+
"grad_norm": 26.702394485473633,
|
| 320 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 321 |
+
"loss": 1.5546,
|
| 322 |
+
"step": 1075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.4630738522954092,
|
| 326 |
+
"grad_norm": 28.11456871032715,
|
| 327 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 328 |
+
"loss": 1.7266,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.4963406520292748,
|
| 333 |
+
"grad_norm": 29.24340057373047,
|
| 334 |
+
"learning_rate": 1.49468085106383e-05,
|
| 335 |
+
"loss": 1.395,
|
| 336 |
+
"step": 1125
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.5296074517631404,
|
| 340 |
+
"grad_norm": 24.214881896972656,
|
| 341 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 342 |
+
"loss": 1.4127,
|
| 343 |
+
"step": 1150
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.562874251497006,
|
| 347 |
+
"grad_norm": 33.323997497558594,
|
| 348 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 349 |
+
"loss": 1.6986,
|
| 350 |
+
"step": 1175
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5961410512308716,
|
| 354 |
+
"grad_norm": 38.097965240478516,
|
| 355 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 356 |
+
"loss": 1.6189,
|
| 357 |
+
"step": 1200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6294078509647372,
|
| 361 |
+
"grad_norm": 39.18313980102539,
|
| 362 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 363 |
+
"loss": 1.6214,
|
| 364 |
+
"step": 1225
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.6626746506986028,
|
| 368 |
+
"grad_norm": 25.2917537689209,
|
| 369 |
+
"learning_rate": 1.660904255319149e-05,
|
| 370 |
+
"loss": 1.6657,
|
| 371 |
+
"step": 1250
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.6959414504324684,
|
| 375 |
+
"grad_norm": 27.77417755126953,
|
| 376 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 377 |
+
"loss": 1.4622,
|
| 378 |
+
"step": 1275
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.729208250166334,
|
| 382 |
+
"grad_norm": 26.38246726989746,
|
| 383 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 384 |
+
"loss": 1.4535,
|
| 385 |
+
"step": 1300
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.7624750499001998,
|
| 389 |
+
"grad_norm": 24.15390968322754,
|
| 390 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 391 |
+
"loss": 1.6517,
|
| 392 |
+
"step": 1325
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.7957418496340654,
|
| 396 |
+
"grad_norm": 25.304895401000977,
|
| 397 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 398 |
+
"loss": 1.4146,
|
| 399 |
+
"step": 1350
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.829008649367931,
|
| 403 |
+
"grad_norm": 30.78265953063965,
|
| 404 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 405 |
+
"loss": 1.5098,
|
| 406 |
+
"step": 1375
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.8622754491017965,
|
| 410 |
+
"grad_norm": 21.03673553466797,
|
| 411 |
+
"learning_rate": 1.860372340425532e-05,
|
| 412 |
+
"loss": 1.5286,
|
| 413 |
+
"step": 1400
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.8955422488356621,
|
| 417 |
+
"grad_norm": 26.255247116088867,
|
| 418 |
+
"learning_rate": 1.893617021276596e-05,
|
| 419 |
+
"loss": 1.564,
|
| 420 |
+
"step": 1425
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.9288090485695277,
|
| 424 |
+
"grad_norm": 48.96797561645508,
|
| 425 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 426 |
+
"loss": 1.5111,
|
| 427 |
+
"step": 1450
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.9620758483033933,
|
| 431 |
+
"grad_norm": 21.748788833618164,
|
| 432 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 433 |
+
"loss": 1.4308,
|
| 434 |
+
"step": 1475
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.995342648037259,
|
| 438 |
+
"grad_norm": 24.715282440185547,
|
| 439 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 440 |
+
"loss": 1.3674,
|
| 441 |
+
"step": 1500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0,
|
| 445 |
+
"eval_accuracy": 0.8637724550898204,
|
| 446 |
+
"eval_loss": 0.3624139428138733,
|
| 447 |
+
"eval_macro_f1": 0.6436672764733071,
|
| 448 |
+
"eval_runtime": 1.5434,
|
| 449 |
+
"eval_samples_per_second": 1731.26,
|
| 450 |
+
"eval_steps_per_second": 108.204,
|
| 451 |
+
"eval_weighted_f1": 0.8555428081072758,
|
| 452 |
+
"step": 1504
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.027944111776447,
|
| 456 |
+
"grad_norm": 23.2908935546875,
|
| 457 |
+
"learning_rate": 1.9970449172576832e-05,
|
| 458 |
+
"loss": 1.4192,
|
| 459 |
+
"step": 1525
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0612109115103125,
|
| 463 |
+
"grad_norm": 26.697948455810547,
|
| 464 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 465 |
+
"loss": 1.4633,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.094477711244178,
|
| 470 |
+
"grad_norm": 19.0259952545166,
|
| 471 |
+
"learning_rate": 1.9896572104018914e-05,
|
| 472 |
+
"loss": 1.3156,
|
| 473 |
+
"step": 1575
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.1277445109780437,
|
| 477 |
+
"grad_norm": 26.398216247558594,
|
| 478 |
+
"learning_rate": 1.9859633569739953e-05,
|
| 479 |
+
"loss": 1.1985,
|
| 480 |
+
"step": 1600
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.1610113107119093,
|
| 484 |
+
"grad_norm": 23.088871002197266,
|
| 485 |
+
"learning_rate": 1.9822695035460996e-05,
|
| 486 |
+
"loss": 1.2202,
|
| 487 |
+
"step": 1625
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.194278110445775,
|
| 491 |
+
"grad_norm": 26.138647079467773,
|
| 492 |
+
"learning_rate": 1.9785756501182035e-05,
|
| 493 |
+
"loss": 1.2849,
|
| 494 |
+
"step": 1650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.2275449101796405,
|
| 498 |
+
"grad_norm": 18.806598663330078,
|
| 499 |
+
"learning_rate": 1.9748817966903074e-05,
|
| 500 |
+
"loss": 1.2639,
|
| 501 |
+
"step": 1675
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.260811709913506,
|
| 505 |
+
"grad_norm": 15.955117225646973,
|
| 506 |
+
"learning_rate": 1.9711879432624113e-05,
|
| 507 |
+
"loss": 1.3084,
|
| 508 |
+
"step": 1700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.294078509647372,
|
| 512 |
+
"grad_norm": 26.74173355102539,
|
| 513 |
+
"learning_rate": 1.9674940898345156e-05,
|
| 514 |
+
"loss": 1.0918,
|
| 515 |
+
"step": 1725
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.3273453093812373,
|
| 519 |
+
"grad_norm": 25.826892852783203,
|
| 520 |
+
"learning_rate": 1.9638002364066195e-05,
|
| 521 |
+
"loss": 1.3421,
|
| 522 |
+
"step": 1750
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.3606121091151033,
|
| 526 |
+
"grad_norm": 19.300220489501953,
|
| 527 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 528 |
+
"loss": 1.2709,
|
| 529 |
+
"step": 1775
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.393878908848969,
|
| 533 |
+
"grad_norm": 33.41100311279297,
|
| 534 |
+
"learning_rate": 1.9564125295508277e-05,
|
| 535 |
+
"loss": 1.2728,
|
| 536 |
+
"step": 1800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.4271457085828345,
|
| 540 |
+
"grad_norm": 33.1726188659668,
|
| 541 |
+
"learning_rate": 1.9527186761229316e-05,
|
| 542 |
+
"loss": 1.3343,
|
| 543 |
+
"step": 1825
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.4604125083167,
|
| 547 |
+
"grad_norm": 16.324378967285156,
|
| 548 |
+
"learning_rate": 1.9490248226950355e-05,
|
| 549 |
+
"loss": 1.5619,
|
| 550 |
+
"step": 1850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.4936793080505657,
|
| 554 |
+
"grad_norm": 11.470542907714844,
|
| 555 |
+
"learning_rate": 1.9453309692671394e-05,
|
| 556 |
+
"loss": 1.3248,
|
| 557 |
+
"step": 1875
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.5269461077844313,
|
| 561 |
+
"grad_norm": 47.82541275024414,
|
| 562 |
+
"learning_rate": 1.9416371158392437e-05,
|
| 563 |
+
"loss": 1.4005,
|
| 564 |
+
"step": 1900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.560212907518297,
|
| 568 |
+
"grad_norm": 13.039468765258789,
|
| 569 |
+
"learning_rate": 1.9379432624113476e-05,
|
| 570 |
+
"loss": 1.2775,
|
| 571 |
+
"step": 1925
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.5934797072521625,
|
| 575 |
+
"grad_norm": 32.88628387451172,
|
| 576 |
+
"learning_rate": 1.9342494089834515e-05,
|
| 577 |
+
"loss": 1.3989,
|
| 578 |
+
"step": 1950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.626746506986028,
|
| 582 |
+
"grad_norm": 32.169254302978516,
|
| 583 |
+
"learning_rate": 1.9305555555555558e-05,
|
| 584 |
+
"loss": 1.2549,
|
| 585 |
+
"step": 1975
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.6600133067198937,
|
| 589 |
+
"grad_norm": 25.954526901245117,
|
| 590 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 591 |
+
"loss": 1.2969,
|
| 592 |
+
"step": 2000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.6932801064537593,
|
| 596 |
+
"grad_norm": 20.622583389282227,
|
| 597 |
+
"learning_rate": 1.9231678486997636e-05,
|
| 598 |
+
"loss": 1.3457,
|
| 599 |
+
"step": 2025
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.726546906187625,
|
| 603 |
+
"grad_norm": 43.02334213256836,
|
| 604 |
+
"learning_rate": 1.919473995271868e-05,
|
| 605 |
+
"loss": 1.0394,
|
| 606 |
+
"step": 2050
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.7598137059214904,
|
| 610 |
+
"grad_norm": 42.223323822021484,
|
| 611 |
+
"learning_rate": 1.9157801418439718e-05,
|
| 612 |
+
"loss": 1.389,
|
| 613 |
+
"step": 2075
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.793080505655356,
|
| 617 |
+
"grad_norm": 35.94285583496094,
|
| 618 |
+
"learning_rate": 1.9120862884160757e-05,
|
| 619 |
+
"loss": 1.2184,
|
| 620 |
+
"step": 2100
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.8263473053892216,
|
| 624 |
+
"grad_norm": 27.786062240600586,
|
| 625 |
+
"learning_rate": 1.9083924349881797e-05,
|
| 626 |
+
"loss": 1.2821,
|
| 627 |
+
"step": 2125
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.8596141051230872,
|
| 631 |
+
"grad_norm": 22.151962280273438,
|
| 632 |
+
"learning_rate": 1.904698581560284e-05,
|
| 633 |
+
"loss": 1.2209,
|
| 634 |
+
"step": 2150
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.892880904856953,
|
| 638 |
+
"grad_norm": 26.30831527709961,
|
| 639 |
+
"learning_rate": 1.9010047281323878e-05,
|
| 640 |
+
"loss": 1.306,
|
| 641 |
+
"step": 2175
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.9261477045908184,
|
| 645 |
+
"grad_norm": 21.87983512878418,
|
| 646 |
+
"learning_rate": 1.897310874704492e-05,
|
| 647 |
+
"loss": 1.327,
|
| 648 |
+
"step": 2200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.959414504324684,
|
| 652 |
+
"grad_norm": 33.40931701660156,
|
| 653 |
+
"learning_rate": 1.893617021276596e-05,
|
| 654 |
+
"loss": 1.11,
|
| 655 |
+
"step": 2225
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.9926813040585496,
|
| 659 |
+
"grad_norm": 22.679718017578125,
|
| 660 |
+
"learning_rate": 1.8899231678487e-05,
|
| 661 |
+
"loss": 1.3512,
|
| 662 |
+
"step": 2250
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.0,
|
| 666 |
+
"eval_accuracy": 0.8648952095808383,
|
| 667 |
+
"eval_loss": 0.3499581217765808,
|
| 668 |
+
"eval_macro_f1": 0.6764172148942992,
|
| 669 |
+
"eval_runtime": 1.5942,
|
| 670 |
+
"eval_samples_per_second": 1676.096,
|
| 671 |
+
"eval_steps_per_second": 104.756,
|
| 672 |
+
"eval_weighted_f1": 0.8652442921890966,
|
| 673 |
+
"step": 2256
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.0252827677977376,
|
| 677 |
+
"grad_norm": 34.03767776489258,
|
| 678 |
+
"learning_rate": 1.886229314420804e-05,
|
| 679 |
+
"loss": 1.139,
|
| 680 |
+
"step": 2275
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.0585495675316037,
|
| 684 |
+
"grad_norm": 19.097692489624023,
|
| 685 |
+
"learning_rate": 1.8825354609929078e-05,
|
| 686 |
+
"loss": 0.9606,
|
| 687 |
+
"step": 2300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.0918163672654693,
|
| 691 |
+
"grad_norm": 26.26155662536621,
|
| 692 |
+
"learning_rate": 1.878841607565012e-05,
|
| 693 |
+
"loss": 0.9136,
|
| 694 |
+
"step": 2325
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.125083166999335,
|
| 698 |
+
"grad_norm": 29.003942489624023,
|
| 699 |
+
"learning_rate": 1.875147754137116e-05,
|
| 700 |
+
"loss": 1.0377,
|
| 701 |
+
"step": 2350
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.1583499667332005,
|
| 705 |
+
"grad_norm": 55.168212890625,
|
| 706 |
+
"learning_rate": 1.8714539007092202e-05,
|
| 707 |
+
"loss": 0.9944,
|
| 708 |
+
"step": 2375
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.191616766467066,
|
| 712 |
+
"grad_norm": 31.195945739746094,
|
| 713 |
+
"learning_rate": 1.867760047281324e-05,
|
| 714 |
+
"loss": 0.9626,
|
| 715 |
+
"step": 2400
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.2248835662009316,
|
| 719 |
+
"grad_norm": 30.541507720947266,
|
| 720 |
+
"learning_rate": 1.864066193853428e-05,
|
| 721 |
+
"loss": 0.9487,
|
| 722 |
+
"step": 2425
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.2581503659347972,
|
| 726 |
+
"grad_norm": 32.93328857421875,
|
| 727 |
+
"learning_rate": 1.860372340425532e-05,
|
| 728 |
+
"loss": 1.0851,
|
| 729 |
+
"step": 2450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.291417165668663,
|
| 733 |
+
"grad_norm": 27.888490676879883,
|
| 734 |
+
"learning_rate": 1.856678486997636e-05,
|
| 735 |
+
"loss": 1.1047,
|
| 736 |
+
"step": 2475
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.3246839654025284,
|
| 740 |
+
"grad_norm": 24.827001571655273,
|
| 741 |
+
"learning_rate": 1.85298463356974e-05,
|
| 742 |
+
"loss": 0.9216,
|
| 743 |
+
"step": 2500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.357950765136394,
|
| 747 |
+
"grad_norm": 38.58675765991211,
|
| 748 |
+
"learning_rate": 1.849290780141844e-05,
|
| 749 |
+
"loss": 0.9583,
|
| 750 |
+
"step": 2525
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.3912175648702596,
|
| 754 |
+
"grad_norm": 38.48832702636719,
|
| 755 |
+
"learning_rate": 1.8455969267139483e-05,
|
| 756 |
+
"loss": 1.0392,
|
| 757 |
+
"step": 2550
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.424484364604125,
|
| 761 |
+
"grad_norm": 36.60333251953125,
|
| 762 |
+
"learning_rate": 1.8419030732860522e-05,
|
| 763 |
+
"loss": 0.9494,
|
| 764 |
+
"step": 2575
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.457751164337991,
|
| 768 |
+
"grad_norm": 29.002120971679688,
|
| 769 |
+
"learning_rate": 1.838209219858156e-05,
|
| 770 |
+
"loss": 0.9332,
|
| 771 |
+
"step": 2600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.4910179640718564,
|
| 775 |
+
"grad_norm": 21.734298706054688,
|
| 776 |
+
"learning_rate": 1.83451536643026e-05,
|
| 777 |
+
"loss": 1.2701,
|
| 778 |
+
"step": 2625
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.524284763805722,
|
| 782 |
+
"grad_norm": 21.092363357543945,
|
| 783 |
+
"learning_rate": 1.8308215130023643e-05,
|
| 784 |
+
"loss": 1.1146,
|
| 785 |
+
"step": 2650
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.5575515635395876,
|
| 789 |
+
"grad_norm": 17.316835403442383,
|
| 790 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 791 |
+
"loss": 0.9659,
|
| 792 |
+
"step": 2675
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.590818363273453,
|
| 796 |
+
"grad_norm": 42.5993766784668,
|
| 797 |
+
"learning_rate": 1.823433806146572e-05,
|
| 798 |
+
"loss": 1.1817,
|
| 799 |
+
"step": 2700
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.6240851630073188,
|
| 803 |
+
"grad_norm": 47.7999382019043,
|
| 804 |
+
"learning_rate": 1.8197399527186764e-05,
|
| 805 |
+
"loss": 1.0463,
|
| 806 |
+
"step": 2725
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.6573519627411843,
|
| 810 |
+
"grad_norm": 23.832273483276367,
|
| 811 |
+
"learning_rate": 1.8160460992907803e-05,
|
| 812 |
+
"loss": 0.9641,
|
| 813 |
+
"step": 2750
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.69061876247505,
|
| 817 |
+
"grad_norm": 18.98507308959961,
|
| 818 |
+
"learning_rate": 1.8123522458628842e-05,
|
| 819 |
+
"loss": 1.0791,
|
| 820 |
+
"step": 2775
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.7238855622089155,
|
| 824 |
+
"grad_norm": 30.13364028930664,
|
| 825 |
+
"learning_rate": 1.808658392434988e-05,
|
| 826 |
+
"loss": 0.8868,
|
| 827 |
+
"step": 2800
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.757152361942781,
|
| 831 |
+
"grad_norm": 32.605506896972656,
|
| 832 |
+
"learning_rate": 1.8049645390070924e-05,
|
| 833 |
+
"loss": 1.0201,
|
| 834 |
+
"step": 2825
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.7904191616766467,
|
| 838 |
+
"grad_norm": 27.201099395751953,
|
| 839 |
+
"learning_rate": 1.8012706855791963e-05,
|
| 840 |
+
"loss": 0.8997,
|
| 841 |
+
"step": 2850
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.8236859614105123,
|
| 845 |
+
"grad_norm": 49.87832260131836,
|
| 846 |
+
"learning_rate": 1.7975768321513003e-05,
|
| 847 |
+
"loss": 1.1678,
|
| 848 |
+
"step": 2875
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.856952761144378,
|
| 852 |
+
"grad_norm": 34.64746856689453,
|
| 853 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 854 |
+
"loss": 0.9894,
|
| 855 |
+
"step": 2900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.8902195608782435,
|
| 859 |
+
"grad_norm": 31.959882736206055,
|
| 860 |
+
"learning_rate": 1.7901891252955084e-05,
|
| 861 |
+
"loss": 0.9329,
|
| 862 |
+
"step": 2925
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.923486360612109,
|
| 866 |
+
"grad_norm": 14.112442016601562,
|
| 867 |
+
"learning_rate": 1.7864952718676124e-05,
|
| 868 |
+
"loss": 1.0693,
|
| 869 |
+
"step": 2950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.9567531603459747,
|
| 873 |
+
"grad_norm": 21.995716094970703,
|
| 874 |
+
"learning_rate": 1.7828014184397163e-05,
|
| 875 |
+
"loss": 1.0372,
|
| 876 |
+
"step": 2975
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.9900199600798403,
|
| 880 |
+
"grad_norm": 28.292930603027344,
|
| 881 |
+
"learning_rate": 1.7791075650118205e-05,
|
| 882 |
+
"loss": 1.0076,
|
| 883 |
+
"step": 3000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 4.0,
|
| 887 |
+
"eval_accuracy": 0.8682634730538922,
|
| 888 |
+
"eval_loss": 0.3861147463321686,
|
| 889 |
+
"eval_macro_f1": 0.6498392190640252,
|
| 890 |
+
"eval_runtime": 1.507,
|
| 891 |
+
"eval_samples_per_second": 1773.082,
|
| 892 |
+
"eval_steps_per_second": 110.818,
|
| 893 |
+
"eval_weighted_f1": 0.8641321592695009,
|
| 894 |
+
"step": 3008
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.022621423819029,
|
| 898 |
+
"grad_norm": 34.148014068603516,
|
| 899 |
+
"learning_rate": 1.7754137115839245e-05,
|
| 900 |
+
"loss": 0.8104,
|
| 901 |
+
"step": 3025
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 4.055888223552894,
|
| 905 |
+
"grad_norm": 23.081899642944336,
|
| 906 |
+
"learning_rate": 1.7717198581560284e-05,
|
| 907 |
+
"loss": 0.7902,
|
| 908 |
+
"step": 3050
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 4.08915502328676,
|
| 912 |
+
"grad_norm": 27.10505485534668,
|
| 913 |
+
"learning_rate": 1.7680260047281326e-05,
|
| 914 |
+
"loss": 0.8715,
|
| 915 |
+
"step": 3075
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.122421823020625,
|
| 919 |
+
"grad_norm": 53.181549072265625,
|
| 920 |
+
"learning_rate": 1.7643321513002365e-05,
|
| 921 |
+
"loss": 0.777,
|
| 922 |
+
"step": 3100
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 4.155688622754491,
|
| 926 |
+
"grad_norm": 32.25143814086914,
|
| 927 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 928 |
+
"loss": 0.7971,
|
| 929 |
+
"step": 3125
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 4.188955422488356,
|
| 933 |
+
"grad_norm": 38.85936737060547,
|
| 934 |
+
"learning_rate": 1.7569444444444447e-05,
|
| 935 |
+
"loss": 0.8528,
|
| 936 |
+
"step": 3150
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 4.222222222222222,
|
| 940 |
+
"grad_norm": 12.436830520629883,
|
| 941 |
+
"learning_rate": 1.7532505910165486e-05,
|
| 942 |
+
"loss": 0.6806,
|
| 943 |
+
"step": 3175
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 4.2554890219560875,
|
| 947 |
+
"grad_norm": 35.394046783447266,
|
| 948 |
+
"learning_rate": 1.7495567375886526e-05,
|
| 949 |
+
"loss": 0.7414,
|
| 950 |
+
"step": 3200
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 4.2887558216899535,
|
| 954 |
+
"grad_norm": 44.16730499267578,
|
| 955 |
+
"learning_rate": 1.7458628841607565e-05,
|
| 956 |
+
"loss": 0.8936,
|
| 957 |
+
"step": 3225
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 4.322022621423819,
|
| 961 |
+
"grad_norm": 25.659095764160156,
|
| 962 |
+
"learning_rate": 1.7421690307328607e-05,
|
| 963 |
+
"loss": 0.8213,
|
| 964 |
+
"step": 3250
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 4.355289421157685,
|
| 968 |
+
"grad_norm": 77.10603332519531,
|
| 969 |
+
"learning_rate": 1.7384751773049647e-05,
|
| 970 |
+
"loss": 0.7927,
|
| 971 |
+
"step": 3275
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 4.38855622089155,
|
| 975 |
+
"grad_norm": 30.05143928527832,
|
| 976 |
+
"learning_rate": 1.734781323877069e-05,
|
| 977 |
+
"loss": 0.7544,
|
| 978 |
+
"step": 3300
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.421823020625416,
|
| 982 |
+
"grad_norm": 46.79085922241211,
|
| 983 |
+
"learning_rate": 1.731087470449173e-05,
|
| 984 |
+
"loss": 0.9205,
|
| 985 |
+
"step": 3325
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 4.455089820359281,
|
| 989 |
+
"grad_norm": 40.14250946044922,
|
| 990 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 991 |
+
"loss": 0.8438,
|
| 992 |
+
"step": 3350
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.488356620093147,
|
| 996 |
+
"grad_norm": 28.33310317993164,
|
| 997 |
+
"learning_rate": 1.7236997635933807e-05,
|
| 998 |
+
"loss": 0.7885,
|
| 999 |
+
"step": 3375
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.521623419827012,
|
| 1003 |
+
"grad_norm": 32.68979263305664,
|
| 1004 |
+
"learning_rate": 1.7200059101654846e-05,
|
| 1005 |
+
"loss": 1.0292,
|
| 1006 |
+
"step": 3400
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.554890219560878,
|
| 1010 |
+
"grad_norm": 26.344491958618164,
|
| 1011 |
+
"learning_rate": 1.716312056737589e-05,
|
| 1012 |
+
"loss": 0.803,
|
| 1013 |
+
"step": 3425
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.588157019294744,
|
| 1017 |
+
"grad_norm": 14.852431297302246,
|
| 1018 |
+
"learning_rate": 1.7126182033096928e-05,
|
| 1019 |
+
"loss": 0.7835,
|
| 1020 |
+
"step": 3450
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.621423819028609,
|
| 1024 |
+
"grad_norm": 41.625064849853516,
|
| 1025 |
+
"learning_rate": 1.708924349881797e-05,
|
| 1026 |
+
"loss": 0.9212,
|
| 1027 |
+
"step": 3475
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 4.654690618762475,
|
| 1031 |
+
"grad_norm": 27.802034378051758,
|
| 1032 |
+
"learning_rate": 1.705230496453901e-05,
|
| 1033 |
+
"loss": 0.839,
|
| 1034 |
+
"step": 3500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.687957418496341,
|
| 1038 |
+
"grad_norm": 68.59847259521484,
|
| 1039 |
+
"learning_rate": 1.701536643026005e-05,
|
| 1040 |
+
"loss": 0.7907,
|
| 1041 |
+
"step": 3525
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.721224218230207,
|
| 1045 |
+
"grad_norm": 61.745582580566406,
|
| 1046 |
+
"learning_rate": 1.6978427895981088e-05,
|
| 1047 |
+
"loss": 0.7641,
|
| 1048 |
+
"step": 3550
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.754491017964072,
|
| 1052 |
+
"grad_norm": 39.64849853515625,
|
| 1053 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 1054 |
+
"loss": 0.6292,
|
| 1055 |
+
"step": 3575
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.787757817697938,
|
| 1059 |
+
"grad_norm": 38.32709503173828,
|
| 1060 |
+
"learning_rate": 1.690455082742317e-05,
|
| 1061 |
+
"loss": 0.9191,
|
| 1062 |
+
"step": 3600
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.821024617431803,
|
| 1066 |
+
"grad_norm": 21.010385513305664,
|
| 1067 |
+
"learning_rate": 1.686761229314421e-05,
|
| 1068 |
+
"loss": 0.8335,
|
| 1069 |
+
"step": 3625
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.854291417165669,
|
| 1073 |
+
"grad_norm": 24.61604118347168,
|
| 1074 |
+
"learning_rate": 1.683067375886525e-05,
|
| 1075 |
+
"loss": 0.9304,
|
| 1076 |
+
"step": 3650
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.887558216899534,
|
| 1080 |
+
"grad_norm": 70.51045227050781,
|
| 1081 |
+
"learning_rate": 1.679373522458629e-05,
|
| 1082 |
+
"loss": 0.7923,
|
| 1083 |
+
"step": 3675
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.9208250166334,
|
| 1087 |
+
"grad_norm": 38.28218078613281,
|
| 1088 |
+
"learning_rate": 1.675679669030733e-05,
|
| 1089 |
+
"loss": 0.7386,
|
| 1090 |
+
"step": 3700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.954091816367265,
|
| 1094 |
+
"grad_norm": 20.940914154052734,
|
| 1095 |
+
"learning_rate": 1.671985815602837e-05,
|
| 1096 |
+
"loss": 0.6568,
|
| 1097 |
+
"step": 3725
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.987358616101131,
|
| 1101 |
+
"grad_norm": 31.858694076538086,
|
| 1102 |
+
"learning_rate": 1.6682919621749408e-05,
|
| 1103 |
+
"loss": 0.9504,
|
| 1104 |
+
"step": 3750
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.0,
|
| 1108 |
+
"eval_accuracy": 0.8656437125748503,
|
| 1109 |
+
"eval_loss": 0.3926589787006378,
|
| 1110 |
+
"eval_macro_f1": 0.6748486242267919,
|
| 1111 |
+
"eval_runtime": 1.5,
|
| 1112 |
+
"eval_samples_per_second": 1781.337,
|
| 1113 |
+
"eval_steps_per_second": 111.334,
|
| 1114 |
+
"eval_weighted_f1": 0.8669454821698057,
|
| 1115 |
+
"step": 3760
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.0199600798403194,
|
| 1119 |
+
"grad_norm": 37.97001266479492,
|
| 1120 |
+
"learning_rate": 1.664598108747045e-05,
|
| 1121 |
+
"loss": 0.7584,
|
| 1122 |
+
"step": 3775
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.053226879574185,
|
| 1126 |
+
"grad_norm": 31.30186653137207,
|
| 1127 |
+
"learning_rate": 1.660904255319149e-05,
|
| 1128 |
+
"loss": 0.6099,
|
| 1129 |
+
"step": 3800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.086493679308051,
|
| 1133 |
+
"grad_norm": 27.994447708129883,
|
| 1134 |
+
"learning_rate": 1.6572104018912532e-05,
|
| 1135 |
+
"loss": 0.4493,
|
| 1136 |
+
"step": 3825
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.119760479041916,
|
| 1140 |
+
"grad_norm": 17.40340232849121,
|
| 1141 |
+
"learning_rate": 1.653516548463357e-05,
|
| 1142 |
+
"loss": 0.655,
|
| 1143 |
+
"step": 3850
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.153027278775782,
|
| 1147 |
+
"grad_norm": 27.897592544555664,
|
| 1148 |
+
"learning_rate": 1.649822695035461e-05,
|
| 1149 |
+
"loss": 0.7556,
|
| 1150 |
+
"step": 3875
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.186294078509647,
|
| 1154 |
+
"grad_norm": 45.2603759765625,
|
| 1155 |
+
"learning_rate": 1.646128841607565e-05,
|
| 1156 |
+
"loss": 0.6634,
|
| 1157 |
+
"step": 3900
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.219560878243513,
|
| 1161 |
+
"grad_norm": 40.82050323486328,
|
| 1162 |
+
"learning_rate": 1.642434988179669e-05,
|
| 1163 |
+
"loss": 0.6573,
|
| 1164 |
+
"step": 3925
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.252827677977379,
|
| 1168 |
+
"grad_norm": 34.0108757019043,
|
| 1169 |
+
"learning_rate": 1.6387411347517732e-05,
|
| 1170 |
+
"loss": 0.7301,
|
| 1171 |
+
"step": 3950
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.286094477711244,
|
| 1175 |
+
"grad_norm": 52.25631332397461,
|
| 1176 |
+
"learning_rate": 1.635047281323877e-05,
|
| 1177 |
+
"loss": 0.5258,
|
| 1178 |
+
"step": 3975
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.319361277445109,
|
| 1182 |
+
"grad_norm": 33.63074493408203,
|
| 1183 |
+
"learning_rate": 1.6313534278959813e-05,
|
| 1184 |
+
"loss": 0.5966,
|
| 1185 |
+
"step": 4000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.352628077178975,
|
| 1189 |
+
"grad_norm": 18.274282455444336,
|
| 1190 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 1191 |
+
"loss": 0.5739,
|
| 1192 |
+
"step": 4025
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.385894876912841,
|
| 1196 |
+
"grad_norm": 20.909400939941406,
|
| 1197 |
+
"learning_rate": 1.6239657210401892e-05,
|
| 1198 |
+
"loss": 0.4991,
|
| 1199 |
+
"step": 4050
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.419161676646707,
|
| 1203 |
+
"grad_norm": 10.55371379852295,
|
| 1204 |
+
"learning_rate": 1.620271867612293e-05,
|
| 1205 |
+
"loss": 0.7464,
|
| 1206 |
+
"step": 4075
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.452428476380573,
|
| 1210 |
+
"grad_norm": 49.42185592651367,
|
| 1211 |
+
"learning_rate": 1.6165780141843974e-05,
|
| 1212 |
+
"loss": 0.6877,
|
| 1213 |
+
"step": 4100
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.485695276114438,
|
| 1217 |
+
"grad_norm": 37.38511276245117,
|
| 1218 |
+
"learning_rate": 1.6128841607565013e-05,
|
| 1219 |
+
"loss": 0.6816,
|
| 1220 |
+
"step": 4125
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.518962075848304,
|
| 1224 |
+
"grad_norm": 6.26555871963501,
|
| 1225 |
+
"learning_rate": 1.6091903073286052e-05,
|
| 1226 |
+
"loss": 0.5693,
|
| 1227 |
+
"step": 4150
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.552228875582169,
|
| 1231 |
+
"grad_norm": 23.01319122314453,
|
| 1232 |
+
"learning_rate": 1.6054964539007095e-05,
|
| 1233 |
+
"loss": 0.792,
|
| 1234 |
+
"step": 4175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.585495675316035,
|
| 1238 |
+
"grad_norm": 48.40448760986328,
|
| 1239 |
+
"learning_rate": 1.6018026004728134e-05,
|
| 1240 |
+
"loss": 0.6381,
|
| 1241 |
+
"step": 4200
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.6187624750499,
|
| 1245 |
+
"grad_norm": 36.653079986572266,
|
| 1246 |
+
"learning_rate": 1.5981087470449176e-05,
|
| 1247 |
+
"loss": 0.7671,
|
| 1248 |
+
"step": 4225
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.652029274783766,
|
| 1252 |
+
"grad_norm": 27.41105842590332,
|
| 1253 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 1254 |
+
"loss": 0.6985,
|
| 1255 |
+
"step": 4250
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.685296074517631,
|
| 1259 |
+
"grad_norm": 66.96533966064453,
|
| 1260 |
+
"learning_rate": 1.5907210401891255e-05,
|
| 1261 |
+
"loss": 0.7069,
|
| 1262 |
+
"step": 4275
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.718562874251497,
|
| 1266 |
+
"grad_norm": 53.56682586669922,
|
| 1267 |
+
"learning_rate": 1.5870271867612294e-05,
|
| 1268 |
+
"loss": 0.6558,
|
| 1269 |
+
"step": 4300
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.7518296739853625,
|
| 1273 |
+
"grad_norm": 31.27998924255371,
|
| 1274 |
+
"learning_rate": 1.5833333333333333e-05,
|
| 1275 |
+
"loss": 0.6231,
|
| 1276 |
+
"step": 4325
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.7850964737192285,
|
| 1280 |
+
"grad_norm": 33.13193893432617,
|
| 1281 |
+
"learning_rate": 1.5796394799054376e-05,
|
| 1282 |
+
"loss": 0.7823,
|
| 1283 |
+
"step": 4350
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.818363273453094,
|
| 1287 |
+
"grad_norm": 25.645647048950195,
|
| 1288 |
+
"learning_rate": 1.5759456264775415e-05,
|
| 1289 |
+
"loss": 0.5854,
|
| 1290 |
+
"step": 4375
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.85163007318696,
|
| 1294 |
+
"grad_norm": 76.89608764648438,
|
| 1295 |
+
"learning_rate": 1.5722517730496457e-05,
|
| 1296 |
+
"loss": 0.8529,
|
| 1297 |
+
"step": 4400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.884896872920825,
|
| 1301 |
+
"grad_norm": 40.57425308227539,
|
| 1302 |
+
"learning_rate": 1.5685579196217497e-05,
|
| 1303 |
+
"loss": 0.6377,
|
| 1304 |
+
"step": 4425
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.918163672654691,
|
| 1308 |
+
"grad_norm": 7.283412456512451,
|
| 1309 |
+
"learning_rate": 1.5648640661938536e-05,
|
| 1310 |
+
"loss": 0.5438,
|
| 1311 |
+
"step": 4450
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.951430472388556,
|
| 1315 |
+
"grad_norm": 66.2386474609375,
|
| 1316 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 1317 |
+
"loss": 0.7596,
|
| 1318 |
+
"step": 4475
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.984697272122422,
|
| 1322 |
+
"grad_norm": 11.451156616210938,
|
| 1323 |
+
"learning_rate": 1.5574763593380614e-05,
|
| 1324 |
+
"loss": 0.7874,
|
| 1325 |
+
"step": 4500
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.0,
|
| 1329 |
+
"eval_accuracy": 0.8735029940119761,
|
| 1330 |
+
"eval_loss": 0.4388781189918518,
|
| 1331 |
+
"eval_macro_f1": 0.6827700263390525,
|
| 1332 |
+
"eval_runtime": 1.5303,
|
| 1333 |
+
"eval_samples_per_second": 1746.115,
|
| 1334 |
+
"eval_steps_per_second": 109.132,
|
| 1335 |
+
"eval_weighted_f1": 0.8711094177585016,
|
| 1336 |
+
"step": 4512
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 6.01729873586161,
|
| 1340 |
+
"grad_norm": 67.75423431396484,
|
| 1341 |
+
"learning_rate": 1.5537825059101657e-05,
|
| 1342 |
+
"loss": 0.5698,
|
| 1343 |
+
"step": 4525
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 6.050565535595475,
|
| 1347 |
+
"grad_norm": 7.571048736572266,
|
| 1348 |
+
"learning_rate": 1.5500886524822696e-05,
|
| 1349 |
+
"loss": 0.3304,
|
| 1350 |
+
"step": 4550
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 6.083832335329341,
|
| 1354 |
+
"grad_norm": 55.4453239440918,
|
| 1355 |
+
"learning_rate": 1.546394799054374e-05,
|
| 1356 |
+
"loss": 0.6612,
|
| 1357 |
+
"step": 4575
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 6.117099135063207,
|
| 1361 |
+
"grad_norm": 79.63700103759766,
|
| 1362 |
+
"learning_rate": 1.5427009456264778e-05,
|
| 1363 |
+
"loss": 0.3504,
|
| 1364 |
+
"step": 4600
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 6.1503659347970725,
|
| 1368 |
+
"grad_norm": 49.47695541381836,
|
| 1369 |
+
"learning_rate": 1.5390070921985817e-05,
|
| 1370 |
+
"loss": 0.5078,
|
| 1371 |
+
"step": 4625
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 6.1836327345309385,
|
| 1375 |
+
"grad_norm": 49.128597259521484,
|
| 1376 |
+
"learning_rate": 1.5353132387706856e-05,
|
| 1377 |
+
"loss": 0.5192,
|
| 1378 |
+
"step": 4650
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 6.216899534264804,
|
| 1382 |
+
"grad_norm": 19.36261749267578,
|
| 1383 |
+
"learning_rate": 1.5316193853427895e-05,
|
| 1384 |
+
"loss": 0.6108,
|
| 1385 |
+
"step": 4675
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 6.25016633399867,
|
| 1389 |
+
"grad_norm": 50.89854431152344,
|
| 1390 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 1391 |
+
"loss": 0.6622,
|
| 1392 |
+
"step": 4700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 6.283433133732535,
|
| 1396 |
+
"grad_norm": 57.39405059814453,
|
| 1397 |
+
"learning_rate": 1.5242316784869977e-05,
|
| 1398 |
+
"loss": 0.5008,
|
| 1399 |
+
"step": 4725
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 6.316699933466401,
|
| 1403 |
+
"grad_norm": 83.19384002685547,
|
| 1404 |
+
"learning_rate": 1.5205378250591018e-05,
|
| 1405 |
+
"loss": 0.6187,
|
| 1406 |
+
"step": 4750
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 6.349966733200266,
|
| 1410 |
+
"grad_norm": 30.342533111572266,
|
| 1411 |
+
"learning_rate": 1.5168439716312057e-05,
|
| 1412 |
+
"loss": 0.5589,
|
| 1413 |
+
"step": 4775
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 6.383233532934132,
|
| 1417 |
+
"grad_norm": 9.4254150390625,
|
| 1418 |
+
"learning_rate": 1.51315011820331e-05,
|
| 1419 |
+
"loss": 0.4618,
|
| 1420 |
+
"step": 4800
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 6.416500332667997,
|
| 1424 |
+
"grad_norm": 9.256645202636719,
|
| 1425 |
+
"learning_rate": 1.5094562647754139e-05,
|
| 1426 |
+
"loss": 0.5257,
|
| 1427 |
+
"step": 4825
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 6.449767132401863,
|
| 1431 |
+
"grad_norm": 10.17599105834961,
|
| 1432 |
+
"learning_rate": 1.5057624113475178e-05,
|
| 1433 |
+
"loss": 0.4557,
|
| 1434 |
+
"step": 4850
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 6.483033932135728,
|
| 1438 |
+
"grad_norm": 40.882362365722656,
|
| 1439 |
+
"learning_rate": 1.5020685579196219e-05,
|
| 1440 |
+
"loss": 0.7595,
|
| 1441 |
+
"step": 4875
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 6.5163007318695945,
|
| 1445 |
+
"grad_norm": 73.62734985351562,
|
| 1446 |
+
"learning_rate": 1.4983747044917258e-05,
|
| 1447 |
+
"loss": 0.7191,
|
| 1448 |
+
"step": 4900
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 6.54956753160346,
|
| 1452 |
+
"grad_norm": 45.64955520629883,
|
| 1453 |
+
"learning_rate": 1.49468085106383e-05,
|
| 1454 |
+
"loss": 0.5214,
|
| 1455 |
+
"step": 4925
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 6.582834331337326,
|
| 1459 |
+
"grad_norm": 85.1973648071289,
|
| 1460 |
+
"learning_rate": 1.490986997635934e-05,
|
| 1461 |
+
"loss": 0.6561,
|
| 1462 |
+
"step": 4950
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 6.616101131071191,
|
| 1466 |
+
"grad_norm": 55.2785530090332,
|
| 1467 |
+
"learning_rate": 1.487293144208038e-05,
|
| 1468 |
+
"loss": 0.4211,
|
| 1469 |
+
"step": 4975
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 6.649367930805057,
|
| 1473 |
+
"grad_norm": 75.65742492675781,
|
| 1474 |
+
"learning_rate": 1.483599290780142e-05,
|
| 1475 |
+
"loss": 0.5717,
|
| 1476 |
+
"step": 5000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 6.682634730538922,
|
| 1480 |
+
"grad_norm": 108.11344909667969,
|
| 1481 |
+
"learning_rate": 1.4799054373522459e-05,
|
| 1482 |
+
"loss": 0.4643,
|
| 1483 |
+
"step": 5025
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 6.715901530272788,
|
| 1487 |
+
"grad_norm": 37.75810241699219,
|
| 1488 |
+
"learning_rate": 1.47621158392435e-05,
|
| 1489 |
+
"loss": 0.5376,
|
| 1490 |
+
"step": 5050
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.749168330006653,
|
| 1494 |
+
"grad_norm": 55.700927734375,
|
| 1495 |
+
"learning_rate": 1.472517730496454e-05,
|
| 1496 |
+
"loss": 0.6029,
|
| 1497 |
+
"step": 5075
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 6.782435129740519,
|
| 1501 |
+
"grad_norm": 35.9907341003418,
|
| 1502 |
+
"learning_rate": 1.4688238770685582e-05,
|
| 1503 |
+
"loss": 0.5956,
|
| 1504 |
+
"step": 5100
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 6.815701929474384,
|
| 1508 |
+
"grad_norm": 47.296199798583984,
|
| 1509 |
+
"learning_rate": 1.4651300236406621e-05,
|
| 1510 |
+
"loss": 0.494,
|
| 1511 |
+
"step": 5125
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 6.84896872920825,
|
| 1515 |
+
"grad_norm": 65.1382827758789,
|
| 1516 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 1517 |
+
"loss": 0.6979,
|
| 1518 |
+
"step": 5150
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 6.8822355289421155,
|
| 1522 |
+
"grad_norm": 80.72993469238281,
|
| 1523 |
+
"learning_rate": 1.4577423167848701e-05,
|
| 1524 |
+
"loss": 0.5972,
|
| 1525 |
+
"step": 5175
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 6.915502328675982,
|
| 1529 |
+
"grad_norm": 103.25460052490234,
|
| 1530 |
+
"learning_rate": 1.454048463356974e-05,
|
| 1531 |
+
"loss": 0.5676,
|
| 1532 |
+
"step": 5200
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 6.948769128409847,
|
| 1536 |
+
"grad_norm": 34.944541931152344,
|
| 1537 |
+
"learning_rate": 1.4503546099290781e-05,
|
| 1538 |
+
"loss": 0.6839,
|
| 1539 |
+
"step": 5225
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 6.982035928143713,
|
| 1543 |
+
"grad_norm": 28.08905601501465,
|
| 1544 |
+
"learning_rate": 1.446660756501182e-05,
|
| 1545 |
+
"loss": 0.6631,
|
| 1546 |
+
"step": 5250
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 7.0,
|
| 1550 |
+
"eval_accuracy": 0.8764970059880239,
|
| 1551 |
+
"eval_loss": 0.4612451493740082,
|
| 1552 |
+
"eval_macro_f1": 0.6915236970070544,
|
| 1553 |
+
"eval_runtime": 1.5249,
|
| 1554 |
+
"eval_samples_per_second": 1752.252,
|
| 1555 |
+
"eval_steps_per_second": 109.516,
|
| 1556 |
+
"eval_weighted_f1": 0.8740941266993443,
|
| 1557 |
+
"step": 5264
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 7.014637391882901,
|
| 1561 |
+
"grad_norm": 56.40662384033203,
|
| 1562 |
+
"learning_rate": 1.4429669030732863e-05,
|
| 1563 |
+
"loss": 0.4373,
|
| 1564 |
+
"step": 5275
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 7.047904191616767,
|
| 1568 |
+
"grad_norm": 14.445525169372559,
|
| 1569 |
+
"learning_rate": 1.4392730496453902e-05,
|
| 1570 |
+
"loss": 0.5497,
|
| 1571 |
+
"step": 5300
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 7.081170991350632,
|
| 1575 |
+
"grad_norm": 20.266786575317383,
|
| 1576 |
+
"learning_rate": 1.4355791962174943e-05,
|
| 1577 |
+
"loss": 0.5601,
|
| 1578 |
+
"step": 5325
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 7.114437791084498,
|
| 1582 |
+
"grad_norm": 33.82612991333008,
|
| 1583 |
+
"learning_rate": 1.4318853427895982e-05,
|
| 1584 |
+
"loss": 0.4247,
|
| 1585 |
+
"step": 5350
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 7.147704590818363,
|
| 1589 |
+
"grad_norm": 1.31230628490448,
|
| 1590 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 1591 |
+
"loss": 0.464,
|
| 1592 |
+
"step": 5375
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 7.180971390552229,
|
| 1596 |
+
"grad_norm": 95.05500030517578,
|
| 1597 |
+
"learning_rate": 1.4244976359338062e-05,
|
| 1598 |
+
"loss": 0.5411,
|
| 1599 |
+
"step": 5400
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 7.214238190286094,
|
| 1603 |
+
"grad_norm": 69.93002319335938,
|
| 1604 |
+
"learning_rate": 1.4208037825059103e-05,
|
| 1605 |
+
"loss": 0.4424,
|
| 1606 |
+
"step": 5425
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 7.24750499001996,
|
| 1610 |
+
"grad_norm": 1.278869390487671,
|
| 1611 |
+
"learning_rate": 1.4171099290780144e-05,
|
| 1612 |
+
"loss": 0.4006,
|
| 1613 |
+
"step": 5450
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 7.2807717897538256,
|
| 1617 |
+
"grad_norm": 4.12618350982666,
|
| 1618 |
+
"learning_rate": 1.4134160756501183e-05,
|
| 1619 |
+
"loss": 0.5958,
|
| 1620 |
+
"step": 5475
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 7.314038589487692,
|
| 1624 |
+
"grad_norm": 135.7259521484375,
|
| 1625 |
+
"learning_rate": 1.4097222222222224e-05,
|
| 1626 |
+
"loss": 0.6047,
|
| 1627 |
+
"step": 5500
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 7.347305389221557,
|
| 1631 |
+
"grad_norm": 62.319541931152344,
|
| 1632 |
+
"learning_rate": 1.4060283687943263e-05,
|
| 1633 |
+
"loss": 0.4441,
|
| 1634 |
+
"step": 5525
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 7.380572188955423,
|
| 1638 |
+
"grad_norm": 55.307979583740234,
|
| 1639 |
+
"learning_rate": 1.4023345153664302e-05,
|
| 1640 |
+
"loss": 0.5032,
|
| 1641 |
+
"step": 5550
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 7.413838988689288,
|
| 1645 |
+
"grad_norm": 94.30623626708984,
|
| 1646 |
+
"learning_rate": 1.3986406619385345e-05,
|
| 1647 |
+
"loss": 0.4143,
|
| 1648 |
+
"step": 5575
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 7.447105788423154,
|
| 1652 |
+
"grad_norm": 29.69417953491211,
|
| 1653 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 1654 |
+
"loss": 0.4291,
|
| 1655 |
+
"step": 5600
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 7.480372588157019,
|
| 1659 |
+
"grad_norm": 1.3788127899169922,
|
| 1660 |
+
"learning_rate": 1.3912529550827425e-05,
|
| 1661 |
+
"loss": 0.3212,
|
| 1662 |
+
"step": 5625
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 7.513639387890885,
|
| 1666 |
+
"grad_norm": 48.75605773925781,
|
| 1667 |
+
"learning_rate": 1.3875591016548464e-05,
|
| 1668 |
+
"loss": 0.5268,
|
| 1669 |
+
"step": 5650
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 7.54690618762475,
|
| 1673 |
+
"grad_norm": 103.29325103759766,
|
| 1674 |
+
"learning_rate": 1.3838652482269505e-05,
|
| 1675 |
+
"loss": 0.5023,
|
| 1676 |
+
"step": 5675
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 7.580172987358616,
|
| 1680 |
+
"grad_norm": 4.039899826049805,
|
| 1681 |
+
"learning_rate": 1.3801713947990544e-05,
|
| 1682 |
+
"loss": 0.4368,
|
| 1683 |
+
"step": 5700
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 7.6134397870924815,
|
| 1687 |
+
"grad_norm": 84.64030456542969,
|
| 1688 |
+
"learning_rate": 1.3764775413711583e-05,
|
| 1689 |
+
"loss": 0.4767,
|
| 1690 |
+
"step": 5725
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 7.6467065868263475,
|
| 1694 |
+
"grad_norm": 58.82286834716797,
|
| 1695 |
+
"learning_rate": 1.3727836879432626e-05,
|
| 1696 |
+
"loss": 0.4143,
|
| 1697 |
+
"step": 5750
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 7.679973386560213,
|
| 1701 |
+
"grad_norm": 39.76670837402344,
|
| 1702 |
+
"learning_rate": 1.3690898345153665e-05,
|
| 1703 |
+
"loss": 0.4539,
|
| 1704 |
+
"step": 5775
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 7.713240186294079,
|
| 1708 |
+
"grad_norm": 79.04641723632812,
|
| 1709 |
+
"learning_rate": 1.3653959810874706e-05,
|
| 1710 |
+
"loss": 0.4439,
|
| 1711 |
+
"step": 5800
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 7.746506986027944,
|
| 1715 |
+
"grad_norm": 40.229087829589844,
|
| 1716 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 1717 |
+
"loss": 0.5551,
|
| 1718 |
+
"step": 5825
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 7.77977378576181,
|
| 1722 |
+
"grad_norm": 90.43767547607422,
|
| 1723 |
+
"learning_rate": 1.3580082742316786e-05,
|
| 1724 |
+
"loss": 0.3941,
|
| 1725 |
+
"step": 5850
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 7.813040585495675,
|
| 1729 |
+
"grad_norm": 24.041860580444336,
|
| 1730 |
+
"learning_rate": 1.3543144208037825e-05,
|
| 1731 |
+
"loss": 0.4246,
|
| 1732 |
+
"step": 5875
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 7.846307385229541,
|
| 1736 |
+
"grad_norm": 26.003265380859375,
|
| 1737 |
+
"learning_rate": 1.3506205673758865e-05,
|
| 1738 |
+
"loss": 0.4047,
|
| 1739 |
+
"step": 5900
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 7.879574184963406,
|
| 1743 |
+
"grad_norm": 36.54729080200195,
|
| 1744 |
+
"learning_rate": 1.3469267139479907e-05,
|
| 1745 |
+
"loss": 0.3484,
|
| 1746 |
+
"step": 5925
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 7.912840984697272,
|
| 1750 |
+
"grad_norm": 128.67298889160156,
|
| 1751 |
+
"learning_rate": 1.3432328605200946e-05,
|
| 1752 |
+
"loss": 0.6038,
|
| 1753 |
+
"step": 5950
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 7.946107784431137,
|
| 1757 |
+
"grad_norm": 21.814119338989258,
|
| 1758 |
+
"learning_rate": 1.3395390070921987e-05,
|
| 1759 |
+
"loss": 0.4007,
|
| 1760 |
+
"step": 5975
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 7.9793745841650034,
|
| 1764 |
+
"grad_norm": 51.24502182006836,
|
| 1765 |
+
"learning_rate": 1.3358451536643026e-05,
|
| 1766 |
+
"loss": 0.3045,
|
| 1767 |
+
"step": 6000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 8.0,
|
| 1771 |
+
"eval_accuracy": 0.8723802395209581,
|
| 1772 |
+
"eval_loss": 0.6036049127578735,
|
| 1773 |
+
"eval_macro_f1": 0.682344327797904,
|
| 1774 |
+
"eval_runtime": 1.5342,
|
| 1775 |
+
"eval_samples_per_second": 1741.674,
|
| 1776 |
+
"eval_steps_per_second": 108.855,
|
| 1777 |
+
"eval_weighted_f1": 0.8692917887779061,
|
| 1778 |
+
"step": 6016
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 8.011976047904191,
|
| 1782 |
+
"grad_norm": 10.092239379882812,
|
| 1783 |
+
"learning_rate": 1.3321513002364069e-05,
|
| 1784 |
+
"loss": 0.5365,
|
| 1785 |
+
"step": 6025
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 8.045242847638058,
|
| 1789 |
+
"grad_norm": 50.38981628417969,
|
| 1790 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 1791 |
+
"loss": 0.3193,
|
| 1792 |
+
"step": 6050
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 8.078509647371924,
|
| 1796 |
+
"grad_norm": 35.566551208496094,
|
| 1797 |
+
"learning_rate": 1.3247635933806147e-05,
|
| 1798 |
+
"loss": 0.3672,
|
| 1799 |
+
"step": 6075
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 8.111776447105788,
|
| 1803 |
+
"grad_norm": 2.3392438888549805,
|
| 1804 |
+
"learning_rate": 1.3210697399527188e-05,
|
| 1805 |
+
"loss": 0.4582,
|
| 1806 |
+
"step": 6100
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 8.145043246839654,
|
| 1810 |
+
"grad_norm": 20.907745361328125,
|
| 1811 |
+
"learning_rate": 1.3173758865248227e-05,
|
| 1812 |
+
"loss": 0.4062,
|
| 1813 |
+
"step": 6125
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 8.17831004657352,
|
| 1817 |
+
"grad_norm": 33.94376754760742,
|
| 1818 |
+
"learning_rate": 1.3136820330969268e-05,
|
| 1819 |
+
"loss": 0.4065,
|
| 1820 |
+
"step": 6150
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 8.211576846307386,
|
| 1824 |
+
"grad_norm": 68.93928527832031,
|
| 1825 |
+
"learning_rate": 1.3099881796690307e-05,
|
| 1826 |
+
"loss": 0.3423,
|
| 1827 |
+
"step": 6175
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 8.24484364604125,
|
| 1831 |
+
"grad_norm": 28.12694549560547,
|
| 1832 |
+
"learning_rate": 1.306294326241135e-05,
|
| 1833 |
+
"loss": 0.4134,
|
| 1834 |
+
"step": 6200
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 8.278110445775116,
|
| 1838 |
+
"grad_norm": 15.582269668579102,
|
| 1839 |
+
"learning_rate": 1.302600472813239e-05,
|
| 1840 |
+
"loss": 0.2289,
|
| 1841 |
+
"step": 6225
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 8.311377245508982,
|
| 1845 |
+
"grad_norm": 66.87283325195312,
|
| 1846 |
+
"learning_rate": 1.2989066193853428e-05,
|
| 1847 |
+
"loss": 0.2798,
|
| 1848 |
+
"step": 6250
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 8.344644045242848,
|
| 1852 |
+
"grad_norm": 47.40612030029297,
|
| 1853 |
+
"learning_rate": 1.295212765957447e-05,
|
| 1854 |
+
"loss": 0.4283,
|
| 1855 |
+
"step": 6275
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 8.377910844976713,
|
| 1859 |
+
"grad_norm": 83.83490753173828,
|
| 1860 |
+
"learning_rate": 1.2915189125295508e-05,
|
| 1861 |
+
"loss": 0.4423,
|
| 1862 |
+
"step": 6300
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 8.411177644710579,
|
| 1866 |
+
"grad_norm": 12.353107452392578,
|
| 1867 |
+
"learning_rate": 1.287825059101655e-05,
|
| 1868 |
+
"loss": 0.3033,
|
| 1869 |
+
"step": 6325
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 8.444444444444445,
|
| 1873 |
+
"grad_norm": 108.77391815185547,
|
| 1874 |
+
"learning_rate": 1.2841312056737589e-05,
|
| 1875 |
+
"loss": 0.3304,
|
| 1876 |
+
"step": 6350
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 8.47771124417831,
|
| 1880 |
+
"grad_norm": 82.73484802246094,
|
| 1881 |
+
"learning_rate": 1.2804373522458631e-05,
|
| 1882 |
+
"loss": 0.3412,
|
| 1883 |
+
"step": 6375
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 8.510978043912175,
|
| 1887 |
+
"grad_norm": 1.6888914108276367,
|
| 1888 |
+
"learning_rate": 1.276743498817967e-05,
|
| 1889 |
+
"loss": 0.6313,
|
| 1890 |
+
"step": 6400
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 8.544244843646041,
|
| 1894 |
+
"grad_norm": 37.53441619873047,
|
| 1895 |
+
"learning_rate": 1.273049645390071e-05,
|
| 1896 |
+
"loss": 0.5788,
|
| 1897 |
+
"step": 6425
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 8.577511643379907,
|
| 1901 |
+
"grad_norm": 26.874143600463867,
|
| 1902 |
+
"learning_rate": 1.269355791962175e-05,
|
| 1903 |
+
"loss": 0.5573,
|
| 1904 |
+
"step": 6450
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 8.610778443113773,
|
| 1908 |
+
"grad_norm": 22.276689529418945,
|
| 1909 |
+
"learning_rate": 1.265661938534279e-05,
|
| 1910 |
+
"loss": 0.5416,
|
| 1911 |
+
"step": 6475
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 8.644045242847637,
|
| 1915 |
+
"grad_norm": 59.7324333190918,
|
| 1916 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 1917 |
+
"loss": 0.451,
|
| 1918 |
+
"step": 6500
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 8.677312042581503,
|
| 1922 |
+
"grad_norm": 29.890644073486328,
|
| 1923 |
+
"learning_rate": 1.2582742316784871e-05,
|
| 1924 |
+
"loss": 0.3457,
|
| 1925 |
+
"step": 6525
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 8.71057884231537,
|
| 1929 |
+
"grad_norm": 50.45320510864258,
|
| 1930 |
+
"learning_rate": 1.2545803782505912e-05,
|
| 1931 |
+
"loss": 0.4717,
|
| 1932 |
+
"step": 6550
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 8.743845642049235,
|
| 1936 |
+
"grad_norm": 88.98976135253906,
|
| 1937 |
+
"learning_rate": 1.2508865248226951e-05,
|
| 1938 |
+
"loss": 0.4307,
|
| 1939 |
+
"step": 6575
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 8.7771124417831,
|
| 1943 |
+
"grad_norm": 51.34151840209961,
|
| 1944 |
+
"learning_rate": 1.247192671394799e-05,
|
| 1945 |
+
"loss": 0.4547,
|
| 1946 |
+
"step": 6600
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 8.810379241516966,
|
| 1950 |
+
"grad_norm": 6.570895195007324,
|
| 1951 |
+
"learning_rate": 1.2434988179669031e-05,
|
| 1952 |
+
"loss": 0.5577,
|
| 1953 |
+
"step": 6625
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 8.843646041250832,
|
| 1957 |
+
"grad_norm": 54.829036712646484,
|
| 1958 |
+
"learning_rate": 1.239804964539007e-05,
|
| 1959 |
+
"loss": 0.4702,
|
| 1960 |
+
"step": 6650
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 8.876912840984698,
|
| 1964 |
+
"grad_norm": 35.923702239990234,
|
| 1965 |
+
"learning_rate": 1.2361111111111113e-05,
|
| 1966 |
+
"loss": 0.3955,
|
| 1967 |
+
"step": 6675
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 8.910179640718562,
|
| 1971 |
+
"grad_norm": 66.6717758178711,
|
| 1972 |
+
"learning_rate": 1.2324172576832152e-05,
|
| 1973 |
+
"loss": 0.3124,
|
| 1974 |
+
"step": 6700
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 8.943446440452428,
|
| 1978 |
+
"grad_norm": 124.40660095214844,
|
| 1979 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 1980 |
+
"loss": 0.441,
|
| 1981 |
+
"step": 6725
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 8.976713240186294,
|
| 1985 |
+
"grad_norm": 0.379179447889328,
|
| 1986 |
+
"learning_rate": 1.2250295508274232e-05,
|
| 1987 |
+
"loss": 0.4675,
|
| 1988 |
+
"step": 6750
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 9.0,
|
| 1992 |
+
"eval_accuracy": 0.8626497005988024,
|
| 1993 |
+
"eval_loss": 0.6842963099479675,
|
| 1994 |
+
"eval_macro_f1": 0.6828302184092205,
|
| 1995 |
+
"eval_runtime": 1.6072,
|
| 1996 |
+
"eval_samples_per_second": 1662.52,
|
| 1997 |
+
"eval_steps_per_second": 103.907,
|
| 1998 |
+
"eval_weighted_f1": 0.8664097413648523,
|
| 1999 |
+
"step": 6768
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 9.009314703925483,
|
| 2003 |
+
"grad_norm": 0.3325503468513489,
|
| 2004 |
+
"learning_rate": 1.2213356973995272e-05,
|
| 2005 |
+
"loss": 0.4008,
|
| 2006 |
+
"step": 6775
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 9.042581503659347,
|
| 2010 |
+
"grad_norm": 1.5962432622909546,
|
| 2011 |
+
"learning_rate": 1.2176418439716313e-05,
|
| 2012 |
+
"loss": 0.2985,
|
| 2013 |
+
"step": 6800
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 9.075848303393213,
|
| 2017 |
+
"grad_norm": 48.94321823120117,
|
| 2018 |
+
"learning_rate": 1.2139479905437352e-05,
|
| 2019 |
+
"loss": 0.3142,
|
| 2020 |
+
"step": 6825
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 9.10911510312708,
|
| 2024 |
+
"grad_norm": 111.9867935180664,
|
| 2025 |
+
"learning_rate": 1.2102541371158394e-05,
|
| 2026 |
+
"loss": 0.3504,
|
| 2027 |
+
"step": 6850
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 9.142381902860945,
|
| 2031 |
+
"grad_norm": 0.2738988995552063,
|
| 2032 |
+
"learning_rate": 1.2065602836879434e-05,
|
| 2033 |
+
"loss": 0.3851,
|
| 2034 |
+
"step": 6875
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 9.17564870259481,
|
| 2038 |
+
"grad_norm": 190.9883270263672,
|
| 2039 |
+
"learning_rate": 1.2028664302600474e-05,
|
| 2040 |
+
"loss": 0.4273,
|
| 2041 |
+
"step": 6900
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 9.208915502328676,
|
| 2045 |
+
"grad_norm": 12.919103622436523,
|
| 2046 |
+
"learning_rate": 1.1991725768321514e-05,
|
| 2047 |
+
"loss": 0.2955,
|
| 2048 |
+
"step": 6925
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 9.242182302062542,
|
| 2052 |
+
"grad_norm": 100.3220443725586,
|
| 2053 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 2054 |
+
"loss": 0.5116,
|
| 2055 |
+
"step": 6950
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 9.275449101796408,
|
| 2059 |
+
"grad_norm": 53.51063919067383,
|
| 2060 |
+
"learning_rate": 1.1917848699763595e-05,
|
| 2061 |
+
"loss": 0.3791,
|
| 2062 |
+
"step": 6975
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 9.308715901530272,
|
| 2066 |
+
"grad_norm": 69.34867858886719,
|
| 2067 |
+
"learning_rate": 1.1880910165484635e-05,
|
| 2068 |
+
"loss": 0.3056,
|
| 2069 |
+
"step": 7000
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 9.341982701264138,
|
| 2073 |
+
"grad_norm": 34.36561584472656,
|
| 2074 |
+
"learning_rate": 1.1843971631205675e-05,
|
| 2075 |
+
"loss": 0.3061,
|
| 2076 |
+
"step": 7025
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 9.375249500998004,
|
| 2080 |
+
"grad_norm": 43.30571365356445,
|
| 2081 |
+
"learning_rate": 1.1807033096926715e-05,
|
| 2082 |
+
"loss": 0.3637,
|
| 2083 |
+
"step": 7050
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 9.40851630073187,
|
| 2087 |
+
"grad_norm": 11.80591869354248,
|
| 2088 |
+
"learning_rate": 1.1770094562647755e-05,
|
| 2089 |
+
"loss": 0.4694,
|
| 2090 |
+
"step": 7075
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 9.441783100465734,
|
| 2094 |
+
"grad_norm": 0.2933409810066223,
|
| 2095 |
+
"learning_rate": 1.1733156028368795e-05,
|
| 2096 |
+
"loss": 0.2603,
|
| 2097 |
+
"step": 7100
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 9.4750499001996,
|
| 2101 |
+
"grad_norm": 4.330387592315674,
|
| 2102 |
+
"learning_rate": 1.1696217494089837e-05,
|
| 2103 |
+
"loss": 0.4118,
|
| 2104 |
+
"step": 7125
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 9.508316699933467,
|
| 2108 |
+
"grad_norm": 35.95737838745117,
|
| 2109 |
+
"learning_rate": 1.1659278959810876e-05,
|
| 2110 |
+
"loss": 0.4323,
|
| 2111 |
+
"step": 7150
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 9.541583499667333,
|
| 2115 |
+
"grad_norm": 5.509078025817871,
|
| 2116 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 2117 |
+
"loss": 0.3734,
|
| 2118 |
+
"step": 7175
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 9.574850299401197,
|
| 2122 |
+
"grad_norm": 36.61819839477539,
|
| 2123 |
+
"learning_rate": 1.1585401891252956e-05,
|
| 2124 |
+
"loss": 0.4831,
|
| 2125 |
+
"step": 7200
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 9.608117099135063,
|
| 2129 |
+
"grad_norm": 41.197086334228516,
|
| 2130 |
+
"learning_rate": 1.1548463356973996e-05,
|
| 2131 |
+
"loss": 0.4023,
|
| 2132 |
+
"step": 7225
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 9.641383898868929,
|
| 2136 |
+
"grad_norm": 91.30474853515625,
|
| 2137 |
+
"learning_rate": 1.1511524822695037e-05,
|
| 2138 |
+
"loss": 0.3154,
|
| 2139 |
+
"step": 7250
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 9.674650698602795,
|
| 2143 |
+
"grad_norm": 64.01599884033203,
|
| 2144 |
+
"learning_rate": 1.1474586288416076e-05,
|
| 2145 |
+
"loss": 0.3316,
|
| 2146 |
+
"step": 7275
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 9.70791749833666,
|
| 2150 |
+
"grad_norm": 13.01900863647461,
|
| 2151 |
+
"learning_rate": 1.1437647754137118e-05,
|
| 2152 |
+
"loss": 0.2857,
|
| 2153 |
+
"step": 7300
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 9.741184298070525,
|
| 2157 |
+
"grad_norm": 55.60533905029297,
|
| 2158 |
+
"learning_rate": 1.1400709219858158e-05,
|
| 2159 |
+
"loss": 0.5138,
|
| 2160 |
+
"step": 7325
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 9.774451097804391,
|
| 2164 |
+
"grad_norm": 50.6915397644043,
|
| 2165 |
+
"learning_rate": 1.1363770685579197e-05,
|
| 2166 |
+
"loss": 0.2967,
|
| 2167 |
+
"step": 7350
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 9.807717897538257,
|
| 2171 |
+
"grad_norm": 2.413247585296631,
|
| 2172 |
+
"learning_rate": 1.1326832151300238e-05,
|
| 2173 |
+
"loss": 0.2323,
|
| 2174 |
+
"step": 7375
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 9.840984697272123,
|
| 2178 |
+
"grad_norm": 65.9951400756836,
|
| 2179 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 2180 |
+
"loss": 0.4549,
|
| 2181 |
+
"step": 7400
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 9.874251497005988,
|
| 2185 |
+
"grad_norm": 17.142255783081055,
|
| 2186 |
+
"learning_rate": 1.1252955082742318e-05,
|
| 2187 |
+
"loss": 0.3073,
|
| 2188 |
+
"step": 7425
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 9.907518296739854,
|
| 2192 |
+
"grad_norm": 158.14410400390625,
|
| 2193 |
+
"learning_rate": 1.1216016548463357e-05,
|
| 2194 |
+
"loss": 0.5155,
|
| 2195 |
+
"step": 7450
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 9.94078509647372,
|
| 2199 |
+
"grad_norm": 51.309715270996094,
|
| 2200 |
+
"learning_rate": 1.11790780141844e-05,
|
| 2201 |
+
"loss": 0.2647,
|
| 2202 |
+
"step": 7475
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 9.974051896207584,
|
| 2206 |
+
"grad_norm": 136.2278594970703,
|
| 2207 |
+
"learning_rate": 1.1142139479905439e-05,
|
| 2208 |
+
"loss": 0.4401,
|
| 2209 |
+
"step": 7500
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 10.0,
|
| 2213 |
+
"eval_accuracy": 0.8648952095808383,
|
| 2214 |
+
"eval_loss": 0.755864143371582,
|
| 2215 |
+
"eval_macro_f1": 0.6629953580597392,
|
| 2216 |
+
"eval_runtime": 1.5361,
|
| 2217 |
+
"eval_samples_per_second": 1739.452,
|
| 2218 |
+
"eval_steps_per_second": 108.716,
|
| 2219 |
+
"eval_weighted_f1": 0.8632223290721509,
|
| 2220 |
+
"step": 7520
|
| 2221 |
+
}
|
| 2222 |
+
],
|
| 2223 |
+
"logging_steps": 25,
|
| 2224 |
+
"max_steps": 15040,
|
| 2225 |
+
"num_input_tokens_seen": 0,
|
| 2226 |
+
"num_train_epochs": 20,
|
| 2227 |
+
"save_steps": 500,
|
| 2228 |
+
"stateful_callbacks": {
|
| 2229 |
+
"EarlyStoppingCallback": {
|
| 2230 |
+
"args": {
|
| 2231 |
+
"early_stopping_patience": 3,
|
| 2232 |
+
"early_stopping_threshold": 0.0
|
| 2233 |
+
},
|
| 2234 |
+
"attributes": {
|
| 2235 |
+
"early_stopping_patience_counter": 3
|
| 2236 |
+
}
|
| 2237 |
+
},
|
| 2238 |
+
"TrainerControl": {
|
| 2239 |
+
"args": {
|
| 2240 |
+
"should_epoch_stop": false,
|
| 2241 |
+
"should_evaluate": false,
|
| 2242 |
+
"should_log": false,
|
| 2243 |
+
"should_save": true,
|
| 2244 |
+
"should_training_stop": true
|
| 2245 |
+
},
|
| 2246 |
+
"attributes": {}
|
| 2247 |
+
}
|
| 2248 |
+
},
|
| 2249 |
+
"total_flos": 0.0,
|
| 2250 |
+
"train_batch_size": 8,
|
| 2251 |
+
"trial_name": null,
|
| 2252 |
+
"trial_params": null
|
| 2253 |
+
}
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760/trainer_state.json
ADDED
|
@@ -0,0 +1,1148 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3760,
|
| 3 |
+
"best_metric": 0.6784753176089526,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3760,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0332667997338656,
|
| 14 |
+
"grad_norm": 44.579559326171875,
|
| 15 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 16 |
+
"loss": 4.5196,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0665335994677312,
|
| 21 |
+
"grad_norm": 70.95801544189453,
|
| 22 |
+
"learning_rate": 6.515957446808512e-07,
|
| 23 |
+
"loss": 4.1274,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0998003992015968,
|
| 28 |
+
"grad_norm": 51.73350524902344,
|
| 29 |
+
"learning_rate": 9.840425531914895e-07,
|
| 30 |
+
"loss": 3.37,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1330671989354624,
|
| 35 |
+
"grad_norm": 46.476234436035156,
|
| 36 |
+
"learning_rate": 1.3164893617021277e-06,
|
| 37 |
+
"loss": 2.5589,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16633399866932802,
|
| 42 |
+
"grad_norm": 22.159822463989258,
|
| 43 |
+
"learning_rate": 1.648936170212766e-06,
|
| 44 |
+
"loss": 2.0841,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1996007984031936,
|
| 49 |
+
"grad_norm": 13.729100227355957,
|
| 50 |
+
"learning_rate": 1.9813829787234043e-06,
|
| 51 |
+
"loss": 2.3818,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2328675981370592,
|
| 56 |
+
"grad_norm": 31.293073654174805,
|
| 57 |
+
"learning_rate": 2.3138297872340428e-06,
|
| 58 |
+
"loss": 2.1685,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2661343978709248,
|
| 63 |
+
"grad_norm": 9.639074325561523,
|
| 64 |
+
"learning_rate": 2.646276595744681e-06,
|
| 65 |
+
"loss": 2.0098,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.2994011976047904,
|
| 70 |
+
"grad_norm": 29.225324630737305,
|
| 71 |
+
"learning_rate": 2.978723404255319e-06,
|
| 72 |
+
"loss": 2.1219,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.33266799733865604,
|
| 77 |
+
"grad_norm": 12.81136417388916,
|
| 78 |
+
"learning_rate": 3.311170212765958e-06,
|
| 79 |
+
"loss": 2.1715,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.36593479707252163,
|
| 84 |
+
"grad_norm": 15.273274421691895,
|
| 85 |
+
"learning_rate": 3.643617021276596e-06,
|
| 86 |
+
"loss": 2.195,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3992015968063872,
|
| 91 |
+
"grad_norm": 17.43927001953125,
|
| 92 |
+
"learning_rate": 3.976063829787235e-06,
|
| 93 |
+
"loss": 2.0343,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4324683965402528,
|
| 98 |
+
"grad_norm": 19.527477264404297,
|
| 99 |
+
"learning_rate": 4.308510638297873e-06,
|
| 100 |
+
"loss": 1.9757,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4657351962741184,
|
| 105 |
+
"grad_norm": 30.783679962158203,
|
| 106 |
+
"learning_rate": 4.640957446808511e-06,
|
| 107 |
+
"loss": 2.1451,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.499001996007984,
|
| 112 |
+
"grad_norm": 27.373443603515625,
|
| 113 |
+
"learning_rate": 4.973404255319149e-06,
|
| 114 |
+
"loss": 1.9268,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5322687957418496,
|
| 119 |
+
"grad_norm": 36.12935256958008,
|
| 120 |
+
"learning_rate": 5.305851063829788e-06,
|
| 121 |
+
"loss": 1.8605,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5655355954757152,
|
| 126 |
+
"grad_norm": 39.421878814697266,
|
| 127 |
+
"learning_rate": 5.638297872340426e-06,
|
| 128 |
+
"loss": 2.1257,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5988023952095808,
|
| 133 |
+
"grad_norm": 25.742048263549805,
|
| 134 |
+
"learning_rate": 5.970744680851064e-06,
|
| 135 |
+
"loss": 1.975,
|
| 136 |
+
"step": 450
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6320691949434465,
|
| 140 |
+
"grad_norm": 18.563602447509766,
|
| 141 |
+
"learning_rate": 6.303191489361703e-06,
|
| 142 |
+
"loss": 1.9787,
|
| 143 |
+
"step": 475
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6653359946773121,
|
| 147 |
+
"grad_norm": 27.189918518066406,
|
| 148 |
+
"learning_rate": 6.635638297872341e-06,
|
| 149 |
+
"loss": 1.9924,
|
| 150 |
+
"step": 500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6986027944111777,
|
| 154 |
+
"grad_norm": 38.091251373291016,
|
| 155 |
+
"learning_rate": 6.968085106382979e-06,
|
| 156 |
+
"loss": 1.8647,
|
| 157 |
+
"step": 525
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7318695941450433,
|
| 161 |
+
"grad_norm": 22.76192855834961,
|
| 162 |
+
"learning_rate": 7.300531914893617e-06,
|
| 163 |
+
"loss": 1.7466,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7651363938789089,
|
| 168 |
+
"grad_norm": 27.64113426208496,
|
| 169 |
+
"learning_rate": 7.632978723404256e-06,
|
| 170 |
+
"loss": 1.9297,
|
| 171 |
+
"step": 575
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.7984031936127745,
|
| 175 |
+
"grad_norm": 27.07349395751953,
|
| 176 |
+
"learning_rate": 7.965425531914895e-06,
|
| 177 |
+
"loss": 1.6819,
|
| 178 |
+
"step": 600
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.83166999334664,
|
| 182 |
+
"grad_norm": 36.454010009765625,
|
| 183 |
+
"learning_rate": 8.297872340425532e-06,
|
| 184 |
+
"loss": 1.7133,
|
| 185 |
+
"step": 625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8649367930805056,
|
| 189 |
+
"grad_norm": 23.96141815185547,
|
| 190 |
+
"learning_rate": 8.630319148936171e-06,
|
| 191 |
+
"loss": 1.8537,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.8982035928143712,
|
| 196 |
+
"grad_norm": 14.69851016998291,
|
| 197 |
+
"learning_rate": 8.96276595744681e-06,
|
| 198 |
+
"loss": 1.8554,
|
| 199 |
+
"step": 675
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9314703925482368,
|
| 203 |
+
"grad_norm": 27.234848022460938,
|
| 204 |
+
"learning_rate": 9.295212765957447e-06,
|
| 205 |
+
"loss": 1.7566,
|
| 206 |
+
"step": 700
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9647371922821024,
|
| 210 |
+
"grad_norm": 27.243558883666992,
|
| 211 |
+
"learning_rate": 9.627659574468086e-06,
|
| 212 |
+
"loss": 1.5349,
|
| 213 |
+
"step": 725
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.998003992015968,
|
| 217 |
+
"grad_norm": 44.65242385864258,
|
| 218 |
+
"learning_rate": 9.960106382978723e-06,
|
| 219 |
+
"loss": 1.6613,
|
| 220 |
+
"step": 750
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0,
|
| 224 |
+
"eval_accuracy": 0.8592814371257484,
|
| 225 |
+
"eval_loss": 0.401942640542984,
|
| 226 |
+
"eval_macro_f1": 0.5934326975706287,
|
| 227 |
+
"eval_runtime": 1.5543,
|
| 228 |
+
"eval_samples_per_second": 1719.061,
|
| 229 |
+
"eval_steps_per_second": 107.441,
|
| 230 |
+
"eval_weighted_f1": 0.8410192971389543,
|
| 231 |
+
"step": 752
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0306054557551563,
|
| 235 |
+
"grad_norm": 27.092920303344727,
|
| 236 |
+
"learning_rate": 1.0292553191489364e-05,
|
| 237 |
+
"loss": 1.621,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.0638722554890219,
|
| 242 |
+
"grad_norm": 28.554630279541016,
|
| 243 |
+
"learning_rate": 1.0625e-05,
|
| 244 |
+
"loss": 1.7632,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.0971390552228875,
|
| 249 |
+
"grad_norm": 18.251523971557617,
|
| 250 |
+
"learning_rate": 1.095744680851064e-05,
|
| 251 |
+
"loss": 1.6875,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.130405854956753,
|
| 256 |
+
"grad_norm": 25.526334762573242,
|
| 257 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 258 |
+
"loss": 1.5827,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.1636726546906186,
|
| 263 |
+
"grad_norm": 34.136436462402344,
|
| 264 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 265 |
+
"loss": 1.5267,
|
| 266 |
+
"step": 875
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.1969394544244845,
|
| 270 |
+
"grad_norm": 38.70103454589844,
|
| 271 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 272 |
+
"loss": 1.5552,
|
| 273 |
+
"step": 900
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.23020625415835,
|
| 277 |
+
"grad_norm": 30.68389320373535,
|
| 278 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 279 |
+
"loss": 1.5197,
|
| 280 |
+
"step": 925
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.2634730538922156,
|
| 284 |
+
"grad_norm": 28.493099212646484,
|
| 285 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 286 |
+
"loss": 1.6468,
|
| 287 |
+
"step": 950
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.2967398536260812,
|
| 291 |
+
"grad_norm": 25.333066940307617,
|
| 292 |
+
"learning_rate": 1.295212765957447e-05,
|
| 293 |
+
"loss": 1.3778,
|
| 294 |
+
"step": 975
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.3300066533599468,
|
| 298 |
+
"grad_norm": 21.102970123291016,
|
| 299 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 300 |
+
"loss": 1.5565,
|
| 301 |
+
"step": 1000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.3632734530938124,
|
| 305 |
+
"grad_norm": 26.531511306762695,
|
| 306 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 307 |
+
"loss": 1.4961,
|
| 308 |
+
"step": 1025
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.396540252827678,
|
| 312 |
+
"grad_norm": 23.80035400390625,
|
| 313 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 314 |
+
"loss": 1.6439,
|
| 315 |
+
"step": 1050
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.4298070525615436,
|
| 319 |
+
"grad_norm": 29.40401840209961,
|
| 320 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 321 |
+
"loss": 1.6489,
|
| 322 |
+
"step": 1075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.4630738522954092,
|
| 326 |
+
"grad_norm": 23.498512268066406,
|
| 327 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 328 |
+
"loss": 1.7361,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.4963406520292748,
|
| 333 |
+
"grad_norm": 25.611040115356445,
|
| 334 |
+
"learning_rate": 1.49468085106383e-05,
|
| 335 |
+
"loss": 1.5585,
|
| 336 |
+
"step": 1125
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.5296074517631404,
|
| 340 |
+
"grad_norm": 22.148107528686523,
|
| 341 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 342 |
+
"loss": 1.7262,
|
| 343 |
+
"step": 1150
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.562874251497006,
|
| 347 |
+
"grad_norm": 18.81505012512207,
|
| 348 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 349 |
+
"loss": 1.5674,
|
| 350 |
+
"step": 1175
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5961410512308716,
|
| 354 |
+
"grad_norm": 41.842350006103516,
|
| 355 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 356 |
+
"loss": 1.6489,
|
| 357 |
+
"step": 1200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6294078509647372,
|
| 361 |
+
"grad_norm": 18.060359954833984,
|
| 362 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 363 |
+
"loss": 1.4792,
|
| 364 |
+
"step": 1225
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.6626746506986028,
|
| 368 |
+
"grad_norm": 26.489307403564453,
|
| 369 |
+
"learning_rate": 1.660904255319149e-05,
|
| 370 |
+
"loss": 1.6544,
|
| 371 |
+
"step": 1250
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.6959414504324684,
|
| 375 |
+
"grad_norm": 22.181209564208984,
|
| 376 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 377 |
+
"loss": 1.4514,
|
| 378 |
+
"step": 1275
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.729208250166334,
|
| 382 |
+
"grad_norm": 14.94635009765625,
|
| 383 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 384 |
+
"loss": 1.3136,
|
| 385 |
+
"step": 1300
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.7624750499001998,
|
| 389 |
+
"grad_norm": 14.453506469726562,
|
| 390 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 391 |
+
"loss": 1.6663,
|
| 392 |
+
"step": 1325
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.7957418496340654,
|
| 396 |
+
"grad_norm": 16.815460205078125,
|
| 397 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 398 |
+
"loss": 1.4437,
|
| 399 |
+
"step": 1350
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.829008649367931,
|
| 403 |
+
"grad_norm": 20.052541732788086,
|
| 404 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 405 |
+
"loss": 1.3574,
|
| 406 |
+
"step": 1375
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.8622754491017965,
|
| 410 |
+
"grad_norm": 17.54789161682129,
|
| 411 |
+
"learning_rate": 1.860372340425532e-05,
|
| 412 |
+
"loss": 1.4221,
|
| 413 |
+
"step": 1400
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.8955422488356621,
|
| 417 |
+
"grad_norm": 25.357349395751953,
|
| 418 |
+
"learning_rate": 1.893617021276596e-05,
|
| 419 |
+
"loss": 1.5318,
|
| 420 |
+
"step": 1425
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.9288090485695277,
|
| 424 |
+
"grad_norm": 18.421524047851562,
|
| 425 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 426 |
+
"loss": 1.5834,
|
| 427 |
+
"step": 1450
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.9620758483033933,
|
| 431 |
+
"grad_norm": 34.23824691772461,
|
| 432 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 433 |
+
"loss": 1.4096,
|
| 434 |
+
"step": 1475
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.995342648037259,
|
| 438 |
+
"grad_norm": 21.318138122558594,
|
| 439 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 440 |
+
"loss": 1.5665,
|
| 441 |
+
"step": 1500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0,
|
| 445 |
+
"eval_accuracy": 0.8581586826347305,
|
| 446 |
+
"eval_loss": 0.37113580107688904,
|
| 447 |
+
"eval_macro_f1": 0.6351347089894025,
|
| 448 |
+
"eval_runtime": 1.527,
|
| 449 |
+
"eval_samples_per_second": 1749.794,
|
| 450 |
+
"eval_steps_per_second": 109.362,
|
| 451 |
+
"eval_weighted_f1": 0.8540414267172031,
|
| 452 |
+
"step": 1504
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.027944111776447,
|
| 456 |
+
"grad_norm": 30.614301681518555,
|
| 457 |
+
"learning_rate": 1.9970449172576832e-05,
|
| 458 |
+
"loss": 1.3339,
|
| 459 |
+
"step": 1525
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0612109115103125,
|
| 463 |
+
"grad_norm": 21.835119247436523,
|
| 464 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 465 |
+
"loss": 1.2229,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.094477711244178,
|
| 470 |
+
"grad_norm": 36.414981842041016,
|
| 471 |
+
"learning_rate": 1.9896572104018914e-05,
|
| 472 |
+
"loss": 1.3,
|
| 473 |
+
"step": 1575
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.1277445109780437,
|
| 477 |
+
"grad_norm": 19.14915657043457,
|
| 478 |
+
"learning_rate": 1.9859633569739953e-05,
|
| 479 |
+
"loss": 1.4414,
|
| 480 |
+
"step": 1600
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.1610113107119093,
|
| 484 |
+
"grad_norm": 25.200891494750977,
|
| 485 |
+
"learning_rate": 1.9822695035460996e-05,
|
| 486 |
+
"loss": 1.4353,
|
| 487 |
+
"step": 1625
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.194278110445775,
|
| 491 |
+
"grad_norm": 23.982807159423828,
|
| 492 |
+
"learning_rate": 1.9785756501182035e-05,
|
| 493 |
+
"loss": 1.159,
|
| 494 |
+
"step": 1650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.2275449101796405,
|
| 498 |
+
"grad_norm": 23.8933162689209,
|
| 499 |
+
"learning_rate": 1.9748817966903074e-05,
|
| 500 |
+
"loss": 1.2949,
|
| 501 |
+
"step": 1675
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.260811709913506,
|
| 505 |
+
"grad_norm": 24.900924682617188,
|
| 506 |
+
"learning_rate": 1.9711879432624113e-05,
|
| 507 |
+
"loss": 1.4499,
|
| 508 |
+
"step": 1700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.294078509647372,
|
| 512 |
+
"grad_norm": 26.51698112487793,
|
| 513 |
+
"learning_rate": 1.9674940898345156e-05,
|
| 514 |
+
"loss": 1.4441,
|
| 515 |
+
"step": 1725
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.3273453093812373,
|
| 519 |
+
"grad_norm": 30.009052276611328,
|
| 520 |
+
"learning_rate": 1.9638002364066195e-05,
|
| 521 |
+
"loss": 1.1753,
|
| 522 |
+
"step": 1750
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.3606121091151033,
|
| 526 |
+
"grad_norm": 27.629823684692383,
|
| 527 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 528 |
+
"loss": 1.2432,
|
| 529 |
+
"step": 1775
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.393878908848969,
|
| 533 |
+
"grad_norm": 14.804618835449219,
|
| 534 |
+
"learning_rate": 1.9564125295508277e-05,
|
| 535 |
+
"loss": 1.303,
|
| 536 |
+
"step": 1800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.4271457085828345,
|
| 540 |
+
"grad_norm": 18.46533966064453,
|
| 541 |
+
"learning_rate": 1.9527186761229316e-05,
|
| 542 |
+
"loss": 1.3121,
|
| 543 |
+
"step": 1825
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.4604125083167,
|
| 547 |
+
"grad_norm": 42.512123107910156,
|
| 548 |
+
"learning_rate": 1.9490248226950355e-05,
|
| 549 |
+
"loss": 1.2911,
|
| 550 |
+
"step": 1850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.4936793080505657,
|
| 554 |
+
"grad_norm": 34.890411376953125,
|
| 555 |
+
"learning_rate": 1.9453309692671394e-05,
|
| 556 |
+
"loss": 1.4884,
|
| 557 |
+
"step": 1875
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.5269461077844313,
|
| 561 |
+
"grad_norm": 26.789302825927734,
|
| 562 |
+
"learning_rate": 1.9416371158392437e-05,
|
| 563 |
+
"loss": 1.4215,
|
| 564 |
+
"step": 1900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.560212907518297,
|
| 568 |
+
"grad_norm": 22.254226684570312,
|
| 569 |
+
"learning_rate": 1.9379432624113476e-05,
|
| 570 |
+
"loss": 1.2067,
|
| 571 |
+
"step": 1925
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.5934797072521625,
|
| 575 |
+
"grad_norm": 13.718234062194824,
|
| 576 |
+
"learning_rate": 1.9342494089834515e-05,
|
| 577 |
+
"loss": 1.3911,
|
| 578 |
+
"step": 1950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.626746506986028,
|
| 582 |
+
"grad_norm": 35.57254409790039,
|
| 583 |
+
"learning_rate": 1.9305555555555558e-05,
|
| 584 |
+
"loss": 1.2174,
|
| 585 |
+
"step": 1975
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.6600133067198937,
|
| 589 |
+
"grad_norm": 18.326045989990234,
|
| 590 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 591 |
+
"loss": 1.1263,
|
| 592 |
+
"step": 2000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.6932801064537593,
|
| 596 |
+
"grad_norm": 36.05733108520508,
|
| 597 |
+
"learning_rate": 1.9231678486997636e-05,
|
| 598 |
+
"loss": 1.3456,
|
| 599 |
+
"step": 2025
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.726546906187625,
|
| 603 |
+
"grad_norm": 19.417469024658203,
|
| 604 |
+
"learning_rate": 1.919473995271868e-05,
|
| 605 |
+
"loss": 1.0996,
|
| 606 |
+
"step": 2050
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.7598137059214904,
|
| 610 |
+
"grad_norm": 35.408267974853516,
|
| 611 |
+
"learning_rate": 1.9157801418439718e-05,
|
| 612 |
+
"loss": 1.4205,
|
| 613 |
+
"step": 2075
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.793080505655356,
|
| 617 |
+
"grad_norm": 22.593019485473633,
|
| 618 |
+
"learning_rate": 1.9120862884160757e-05,
|
| 619 |
+
"loss": 1.1586,
|
| 620 |
+
"step": 2100
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.8263473053892216,
|
| 624 |
+
"grad_norm": 35.65427780151367,
|
| 625 |
+
"learning_rate": 1.9083924349881797e-05,
|
| 626 |
+
"loss": 1.3719,
|
| 627 |
+
"step": 2125
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.8596141051230872,
|
| 631 |
+
"grad_norm": 27.764564514160156,
|
| 632 |
+
"learning_rate": 1.904698581560284e-05,
|
| 633 |
+
"loss": 1.2774,
|
| 634 |
+
"step": 2150
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.892880904856953,
|
| 638 |
+
"grad_norm": 54.58634948730469,
|
| 639 |
+
"learning_rate": 1.9010047281323878e-05,
|
| 640 |
+
"loss": 1.1423,
|
| 641 |
+
"step": 2175
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.9261477045908184,
|
| 645 |
+
"grad_norm": 13.277375221252441,
|
| 646 |
+
"learning_rate": 1.897310874704492e-05,
|
| 647 |
+
"loss": 1.3565,
|
| 648 |
+
"step": 2200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.959414504324684,
|
| 652 |
+
"grad_norm": 20.953813552856445,
|
| 653 |
+
"learning_rate": 1.893617021276596e-05,
|
| 654 |
+
"loss": 1.2279,
|
| 655 |
+
"step": 2225
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.9926813040585496,
|
| 659 |
+
"grad_norm": 19.979778289794922,
|
| 660 |
+
"learning_rate": 1.8899231678487e-05,
|
| 661 |
+
"loss": 1.2266,
|
| 662 |
+
"step": 2250
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.0,
|
| 666 |
+
"eval_accuracy": 0.8712574850299402,
|
| 667 |
+
"eval_loss": 0.3697344958782196,
|
| 668 |
+
"eval_macro_f1": 0.6181104325673971,
|
| 669 |
+
"eval_runtime": 1.5151,
|
| 670 |
+
"eval_samples_per_second": 1763.605,
|
| 671 |
+
"eval_steps_per_second": 110.225,
|
| 672 |
+
"eval_weighted_f1": 0.8580288580313121,
|
| 673 |
+
"step": 2256
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.0252827677977376,
|
| 677 |
+
"grad_norm": 27.134233474731445,
|
| 678 |
+
"learning_rate": 1.886229314420804e-05,
|
| 679 |
+
"loss": 1.1056,
|
| 680 |
+
"step": 2275
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.0585495675316037,
|
| 684 |
+
"grad_norm": 22.94137191772461,
|
| 685 |
+
"learning_rate": 1.8825354609929078e-05,
|
| 686 |
+
"loss": 1.1479,
|
| 687 |
+
"step": 2300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.0918163672654693,
|
| 691 |
+
"grad_norm": 16.680999755859375,
|
| 692 |
+
"learning_rate": 1.878841607565012e-05,
|
| 693 |
+
"loss": 1.0197,
|
| 694 |
+
"step": 2325
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.125083166999335,
|
| 698 |
+
"grad_norm": 28.6317138671875,
|
| 699 |
+
"learning_rate": 1.875147754137116e-05,
|
| 700 |
+
"loss": 1.0265,
|
| 701 |
+
"step": 2350
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.1583499667332005,
|
| 705 |
+
"grad_norm": 24.19533348083496,
|
| 706 |
+
"learning_rate": 1.8714539007092202e-05,
|
| 707 |
+
"loss": 1.0724,
|
| 708 |
+
"step": 2375
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.191616766467066,
|
| 712 |
+
"grad_norm": 20.041053771972656,
|
| 713 |
+
"learning_rate": 1.867760047281324e-05,
|
| 714 |
+
"loss": 1.0137,
|
| 715 |
+
"step": 2400
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.2248835662009316,
|
| 719 |
+
"grad_norm": 49.500205993652344,
|
| 720 |
+
"learning_rate": 1.864066193853428e-05,
|
| 721 |
+
"loss": 1.0362,
|
| 722 |
+
"step": 2425
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.2581503659347972,
|
| 726 |
+
"grad_norm": 33.01273727416992,
|
| 727 |
+
"learning_rate": 1.860372340425532e-05,
|
| 728 |
+
"loss": 1.0825,
|
| 729 |
+
"step": 2450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.291417165668663,
|
| 733 |
+
"grad_norm": 29.107446670532227,
|
| 734 |
+
"learning_rate": 1.856678486997636e-05,
|
| 735 |
+
"loss": 1.1413,
|
| 736 |
+
"step": 2475
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.3246839654025284,
|
| 740 |
+
"grad_norm": 26.056676864624023,
|
| 741 |
+
"learning_rate": 1.85298463356974e-05,
|
| 742 |
+
"loss": 1.0489,
|
| 743 |
+
"step": 2500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.357950765136394,
|
| 747 |
+
"grad_norm": 20.73958396911621,
|
| 748 |
+
"learning_rate": 1.849290780141844e-05,
|
| 749 |
+
"loss": 1.0098,
|
| 750 |
+
"step": 2525
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.3912175648702596,
|
| 754 |
+
"grad_norm": 25.818767547607422,
|
| 755 |
+
"learning_rate": 1.8455969267139483e-05,
|
| 756 |
+
"loss": 0.9775,
|
| 757 |
+
"step": 2550
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.424484364604125,
|
| 761 |
+
"grad_norm": 34.23649978637695,
|
| 762 |
+
"learning_rate": 1.8419030732860522e-05,
|
| 763 |
+
"loss": 1.179,
|
| 764 |
+
"step": 2575
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.457751164337991,
|
| 768 |
+
"grad_norm": 30.995899200439453,
|
| 769 |
+
"learning_rate": 1.838209219858156e-05,
|
| 770 |
+
"loss": 0.9971,
|
| 771 |
+
"step": 2600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.4910179640718564,
|
| 775 |
+
"grad_norm": 16.175426483154297,
|
| 776 |
+
"learning_rate": 1.83451536643026e-05,
|
| 777 |
+
"loss": 0.9482,
|
| 778 |
+
"step": 2625
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.524284763805722,
|
| 782 |
+
"grad_norm": 20.30826187133789,
|
| 783 |
+
"learning_rate": 1.8308215130023643e-05,
|
| 784 |
+
"loss": 1.2234,
|
| 785 |
+
"step": 2650
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.5575515635395876,
|
| 789 |
+
"grad_norm": 46.85417175292969,
|
| 790 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 791 |
+
"loss": 1.094,
|
| 792 |
+
"step": 2675
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.590818363273453,
|
| 796 |
+
"grad_norm": 25.728748321533203,
|
| 797 |
+
"learning_rate": 1.823433806146572e-05,
|
| 798 |
+
"loss": 1.0115,
|
| 799 |
+
"step": 2700
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.6240851630073188,
|
| 803 |
+
"grad_norm": 30.966445922851562,
|
| 804 |
+
"learning_rate": 1.8197399527186764e-05,
|
| 805 |
+
"loss": 1.1783,
|
| 806 |
+
"step": 2725
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.6573519627411843,
|
| 810 |
+
"grad_norm": 33.87657928466797,
|
| 811 |
+
"learning_rate": 1.8160460992907803e-05,
|
| 812 |
+
"loss": 1.2113,
|
| 813 |
+
"step": 2750
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.69061876247505,
|
| 817 |
+
"grad_norm": 27.6673641204834,
|
| 818 |
+
"learning_rate": 1.8123522458628842e-05,
|
| 819 |
+
"loss": 1.1515,
|
| 820 |
+
"step": 2775
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.7238855622089155,
|
| 824 |
+
"grad_norm": 24.126436233520508,
|
| 825 |
+
"learning_rate": 1.808658392434988e-05,
|
| 826 |
+
"loss": 0.9524,
|
| 827 |
+
"step": 2800
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.757152361942781,
|
| 831 |
+
"grad_norm": 48.614654541015625,
|
| 832 |
+
"learning_rate": 1.8049645390070924e-05,
|
| 833 |
+
"loss": 0.8609,
|
| 834 |
+
"step": 2825
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.7904191616766467,
|
| 838 |
+
"grad_norm": 28.923856735229492,
|
| 839 |
+
"learning_rate": 1.8012706855791963e-05,
|
| 840 |
+
"loss": 1.1387,
|
| 841 |
+
"step": 2850
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.8236859614105123,
|
| 845 |
+
"grad_norm": 41.61885070800781,
|
| 846 |
+
"learning_rate": 1.7975768321513003e-05,
|
| 847 |
+
"loss": 1.0666,
|
| 848 |
+
"step": 2875
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.856952761144378,
|
| 852 |
+
"grad_norm": 16.2569522857666,
|
| 853 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 854 |
+
"loss": 1.1988,
|
| 855 |
+
"step": 2900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.8902195608782435,
|
| 859 |
+
"grad_norm": 15.537834167480469,
|
| 860 |
+
"learning_rate": 1.7901891252955084e-05,
|
| 861 |
+
"loss": 1.0413,
|
| 862 |
+
"step": 2925
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.923486360612109,
|
| 866 |
+
"grad_norm": 30.901418685913086,
|
| 867 |
+
"learning_rate": 1.7864952718676124e-05,
|
| 868 |
+
"loss": 0.9529,
|
| 869 |
+
"step": 2950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.9567531603459747,
|
| 873 |
+
"grad_norm": 33.222023010253906,
|
| 874 |
+
"learning_rate": 1.7828014184397163e-05,
|
| 875 |
+
"loss": 0.8885,
|
| 876 |
+
"step": 2975
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.9900199600798403,
|
| 880 |
+
"grad_norm": 34.731224060058594,
|
| 881 |
+
"learning_rate": 1.7791075650118205e-05,
|
| 882 |
+
"loss": 1.1978,
|
| 883 |
+
"step": 3000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 4.0,
|
| 887 |
+
"eval_accuracy": 0.8660179640718563,
|
| 888 |
+
"eval_loss": 0.35781410336494446,
|
| 889 |
+
"eval_macro_f1": 0.663992514269203,
|
| 890 |
+
"eval_runtime": 1.5449,
|
| 891 |
+
"eval_samples_per_second": 1729.565,
|
| 892 |
+
"eval_steps_per_second": 108.098,
|
| 893 |
+
"eval_weighted_f1": 0.865035833325726,
|
| 894 |
+
"step": 3008
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.022621423819029,
|
| 898 |
+
"grad_norm": 19.01364517211914,
|
| 899 |
+
"learning_rate": 1.7754137115839245e-05,
|
| 900 |
+
"loss": 0.8489,
|
| 901 |
+
"step": 3025
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 4.055888223552894,
|
| 905 |
+
"grad_norm": 26.636028289794922,
|
| 906 |
+
"learning_rate": 1.7717198581560284e-05,
|
| 907 |
+
"loss": 0.8252,
|
| 908 |
+
"step": 3050
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 4.08915502328676,
|
| 912 |
+
"grad_norm": 33.308753967285156,
|
| 913 |
+
"learning_rate": 1.7680260047281326e-05,
|
| 914 |
+
"loss": 0.7289,
|
| 915 |
+
"step": 3075
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.122421823020625,
|
| 919 |
+
"grad_norm": 11.095755577087402,
|
| 920 |
+
"learning_rate": 1.7643321513002365e-05,
|
| 921 |
+
"loss": 0.8243,
|
| 922 |
+
"step": 3100
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 4.155688622754491,
|
| 926 |
+
"grad_norm": 21.71223258972168,
|
| 927 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 928 |
+
"loss": 1.0173,
|
| 929 |
+
"step": 3125
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 4.188955422488356,
|
| 933 |
+
"grad_norm": 26.51653480529785,
|
| 934 |
+
"learning_rate": 1.7569444444444447e-05,
|
| 935 |
+
"loss": 0.868,
|
| 936 |
+
"step": 3150
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 4.222222222222222,
|
| 940 |
+
"grad_norm": 22.54006004333496,
|
| 941 |
+
"learning_rate": 1.7532505910165486e-05,
|
| 942 |
+
"loss": 0.7639,
|
| 943 |
+
"step": 3175
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 4.2554890219560875,
|
| 947 |
+
"grad_norm": 64.81362915039062,
|
| 948 |
+
"learning_rate": 1.7495567375886526e-05,
|
| 949 |
+
"loss": 0.9346,
|
| 950 |
+
"step": 3200
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 4.2887558216899535,
|
| 954 |
+
"grad_norm": 47.427085876464844,
|
| 955 |
+
"learning_rate": 1.7458628841607565e-05,
|
| 956 |
+
"loss": 0.8158,
|
| 957 |
+
"step": 3225
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 4.322022621423819,
|
| 961 |
+
"grad_norm": 15.681526184082031,
|
| 962 |
+
"learning_rate": 1.7421690307328607e-05,
|
| 963 |
+
"loss": 0.7279,
|
| 964 |
+
"step": 3250
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 4.355289421157685,
|
| 968 |
+
"grad_norm": 29.830162048339844,
|
| 969 |
+
"learning_rate": 1.7384751773049647e-05,
|
| 970 |
+
"loss": 0.6306,
|
| 971 |
+
"step": 3275
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 4.38855622089155,
|
| 975 |
+
"grad_norm": 22.46377182006836,
|
| 976 |
+
"learning_rate": 1.734781323877069e-05,
|
| 977 |
+
"loss": 0.7833,
|
| 978 |
+
"step": 3300
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.421823020625416,
|
| 982 |
+
"grad_norm": 43.262149810791016,
|
| 983 |
+
"learning_rate": 1.731087470449173e-05,
|
| 984 |
+
"loss": 0.8915,
|
| 985 |
+
"step": 3325
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 4.455089820359281,
|
| 989 |
+
"grad_norm": 38.07517623901367,
|
| 990 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 991 |
+
"loss": 0.8497,
|
| 992 |
+
"step": 3350
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.488356620093147,
|
| 996 |
+
"grad_norm": 54.30280685424805,
|
| 997 |
+
"learning_rate": 1.7236997635933807e-05,
|
| 998 |
+
"loss": 0.7794,
|
| 999 |
+
"step": 3375
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.521623419827012,
|
| 1003 |
+
"grad_norm": 72.16100311279297,
|
| 1004 |
+
"learning_rate": 1.7200059101654846e-05,
|
| 1005 |
+
"loss": 0.8375,
|
| 1006 |
+
"step": 3400
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.554890219560878,
|
| 1010 |
+
"grad_norm": 40.60910415649414,
|
| 1011 |
+
"learning_rate": 1.716312056737589e-05,
|
| 1012 |
+
"loss": 0.889,
|
| 1013 |
+
"step": 3425
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.588157019294744,
|
| 1017 |
+
"grad_norm": 52.0064582824707,
|
| 1018 |
+
"learning_rate": 1.7126182033096928e-05,
|
| 1019 |
+
"loss": 0.8153,
|
| 1020 |
+
"step": 3450
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.621423819028609,
|
| 1024 |
+
"grad_norm": 38.7625846862793,
|
| 1025 |
+
"learning_rate": 1.708924349881797e-05,
|
| 1026 |
+
"loss": 0.9168,
|
| 1027 |
+
"step": 3475
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 4.654690618762475,
|
| 1031 |
+
"grad_norm": 31.68454360961914,
|
| 1032 |
+
"learning_rate": 1.705230496453901e-05,
|
| 1033 |
+
"loss": 0.7938,
|
| 1034 |
+
"step": 3500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.687957418496341,
|
| 1038 |
+
"grad_norm": 28.17580223083496,
|
| 1039 |
+
"learning_rate": 1.701536643026005e-05,
|
| 1040 |
+
"loss": 0.9412,
|
| 1041 |
+
"step": 3525
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.721224218230207,
|
| 1045 |
+
"grad_norm": 22.04960060119629,
|
| 1046 |
+
"learning_rate": 1.6978427895981088e-05,
|
| 1047 |
+
"loss": 0.8231,
|
| 1048 |
+
"step": 3550
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.754491017964072,
|
| 1052 |
+
"grad_norm": 45.03123474121094,
|
| 1053 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 1054 |
+
"loss": 0.8961,
|
| 1055 |
+
"step": 3575
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.787757817697938,
|
| 1059 |
+
"grad_norm": 17.203054428100586,
|
| 1060 |
+
"learning_rate": 1.690455082742317e-05,
|
| 1061 |
+
"loss": 0.8676,
|
| 1062 |
+
"step": 3600
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.821024617431803,
|
| 1066 |
+
"grad_norm": 25.068689346313477,
|
| 1067 |
+
"learning_rate": 1.686761229314421e-05,
|
| 1068 |
+
"loss": 0.7191,
|
| 1069 |
+
"step": 3625
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.854291417165669,
|
| 1073 |
+
"grad_norm": 18.855440139770508,
|
| 1074 |
+
"learning_rate": 1.683067375886525e-05,
|
| 1075 |
+
"loss": 0.9022,
|
| 1076 |
+
"step": 3650
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.887558216899534,
|
| 1080 |
+
"grad_norm": 39.43722915649414,
|
| 1081 |
+
"learning_rate": 1.679373522458629e-05,
|
| 1082 |
+
"loss": 0.9583,
|
| 1083 |
+
"step": 3675
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.9208250166334,
|
| 1087 |
+
"grad_norm": 49.7006950378418,
|
| 1088 |
+
"learning_rate": 1.675679669030733e-05,
|
| 1089 |
+
"loss": 0.781,
|
| 1090 |
+
"step": 3700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.954091816367265,
|
| 1094 |
+
"grad_norm": 54.37468338012695,
|
| 1095 |
+
"learning_rate": 1.671985815602837e-05,
|
| 1096 |
+
"loss": 0.8906,
|
| 1097 |
+
"step": 3725
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.987358616101131,
|
| 1101 |
+
"grad_norm": 52.063228607177734,
|
| 1102 |
+
"learning_rate": 1.6682919621749408e-05,
|
| 1103 |
+
"loss": 0.8871,
|
| 1104 |
+
"step": 3750
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.0,
|
| 1108 |
+
"eval_accuracy": 0.8660179640718563,
|
| 1109 |
+
"eval_loss": 0.46085795760154724,
|
| 1110 |
+
"eval_macro_f1": 0.6784753176089526,
|
| 1111 |
+
"eval_runtime": 1.5487,
|
| 1112 |
+
"eval_samples_per_second": 1725.332,
|
| 1113 |
+
"eval_steps_per_second": 107.833,
|
| 1114 |
+
"eval_weighted_f1": 0.8662687400734955,
|
| 1115 |
+
"step": 3760
|
| 1116 |
+
}
|
| 1117 |
+
],
|
| 1118 |
+
"logging_steps": 25,
|
| 1119 |
+
"max_steps": 15040,
|
| 1120 |
+
"num_input_tokens_seen": 0,
|
| 1121 |
+
"num_train_epochs": 20,
|
| 1122 |
+
"save_steps": 500,
|
| 1123 |
+
"stateful_callbacks": {
|
| 1124 |
+
"EarlyStoppingCallback": {
|
| 1125 |
+
"args": {
|
| 1126 |
+
"early_stopping_patience": 3,
|
| 1127 |
+
"early_stopping_threshold": 0.0
|
| 1128 |
+
},
|
| 1129 |
+
"attributes": {
|
| 1130 |
+
"early_stopping_patience_counter": 0
|
| 1131 |
+
}
|
| 1132 |
+
},
|
| 1133 |
+
"TrainerControl": {
|
| 1134 |
+
"args": {
|
| 1135 |
+
"should_epoch_stop": false,
|
| 1136 |
+
"should_evaluate": false,
|
| 1137 |
+
"should_log": false,
|
| 1138 |
+
"should_save": true,
|
| 1139 |
+
"should_training_stop": false
|
| 1140 |
+
},
|
| 1141 |
+
"attributes": {}
|
| 1142 |
+
}
|
| 1143 |
+
},
|
| 1144 |
+
"total_flos": 0.0,
|
| 1145 |
+
"train_batch_size": 8,
|
| 1146 |
+
"trial_name": null,
|
| 1147 |
+
"trial_params": null
|
| 1148 |
+
}
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-6016/trainer_state.json
ADDED
|
@@ -0,0 +1,1811 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3760,
|
| 3 |
+
"best_metric": 0.6784753176089526,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vihsd-bamibert-s43-20260827T124937Z-57f9616a/trainer/checkpoint-3760",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 6016,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0332667997338656,
|
| 14 |
+
"grad_norm": 44.579559326171875,
|
| 15 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 16 |
+
"loss": 4.5196,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0665335994677312,
|
| 21 |
+
"grad_norm": 70.95801544189453,
|
| 22 |
+
"learning_rate": 6.515957446808512e-07,
|
| 23 |
+
"loss": 4.1274,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0998003992015968,
|
| 28 |
+
"grad_norm": 51.73350524902344,
|
| 29 |
+
"learning_rate": 9.840425531914895e-07,
|
| 30 |
+
"loss": 3.37,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1330671989354624,
|
| 35 |
+
"grad_norm": 46.476234436035156,
|
| 36 |
+
"learning_rate": 1.3164893617021277e-06,
|
| 37 |
+
"loss": 2.5589,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16633399866932802,
|
| 42 |
+
"grad_norm": 22.159822463989258,
|
| 43 |
+
"learning_rate": 1.648936170212766e-06,
|
| 44 |
+
"loss": 2.0841,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1996007984031936,
|
| 49 |
+
"grad_norm": 13.729100227355957,
|
| 50 |
+
"learning_rate": 1.9813829787234043e-06,
|
| 51 |
+
"loss": 2.3818,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2328675981370592,
|
| 56 |
+
"grad_norm": 31.293073654174805,
|
| 57 |
+
"learning_rate": 2.3138297872340428e-06,
|
| 58 |
+
"loss": 2.1685,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2661343978709248,
|
| 63 |
+
"grad_norm": 9.639074325561523,
|
| 64 |
+
"learning_rate": 2.646276595744681e-06,
|
| 65 |
+
"loss": 2.0098,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.2994011976047904,
|
| 70 |
+
"grad_norm": 29.225324630737305,
|
| 71 |
+
"learning_rate": 2.978723404255319e-06,
|
| 72 |
+
"loss": 2.1219,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.33266799733865604,
|
| 77 |
+
"grad_norm": 12.81136417388916,
|
| 78 |
+
"learning_rate": 3.311170212765958e-06,
|
| 79 |
+
"loss": 2.1715,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.36593479707252163,
|
| 84 |
+
"grad_norm": 15.273274421691895,
|
| 85 |
+
"learning_rate": 3.643617021276596e-06,
|
| 86 |
+
"loss": 2.195,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3992015968063872,
|
| 91 |
+
"grad_norm": 17.43927001953125,
|
| 92 |
+
"learning_rate": 3.976063829787235e-06,
|
| 93 |
+
"loss": 2.0343,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4324683965402528,
|
| 98 |
+
"grad_norm": 19.527477264404297,
|
| 99 |
+
"learning_rate": 4.308510638297873e-06,
|
| 100 |
+
"loss": 1.9757,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4657351962741184,
|
| 105 |
+
"grad_norm": 30.783679962158203,
|
| 106 |
+
"learning_rate": 4.640957446808511e-06,
|
| 107 |
+
"loss": 2.1451,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.499001996007984,
|
| 112 |
+
"grad_norm": 27.373443603515625,
|
| 113 |
+
"learning_rate": 4.973404255319149e-06,
|
| 114 |
+
"loss": 1.9268,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5322687957418496,
|
| 119 |
+
"grad_norm": 36.12935256958008,
|
| 120 |
+
"learning_rate": 5.305851063829788e-06,
|
| 121 |
+
"loss": 1.8605,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5655355954757152,
|
| 126 |
+
"grad_norm": 39.421878814697266,
|
| 127 |
+
"learning_rate": 5.638297872340426e-06,
|
| 128 |
+
"loss": 2.1257,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5988023952095808,
|
| 133 |
+
"grad_norm": 25.742048263549805,
|
| 134 |
+
"learning_rate": 5.970744680851064e-06,
|
| 135 |
+
"loss": 1.975,
|
| 136 |
+
"step": 450
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6320691949434465,
|
| 140 |
+
"grad_norm": 18.563602447509766,
|
| 141 |
+
"learning_rate": 6.303191489361703e-06,
|
| 142 |
+
"loss": 1.9787,
|
| 143 |
+
"step": 475
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6653359946773121,
|
| 147 |
+
"grad_norm": 27.189918518066406,
|
| 148 |
+
"learning_rate": 6.635638297872341e-06,
|
| 149 |
+
"loss": 1.9924,
|
| 150 |
+
"step": 500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6986027944111777,
|
| 154 |
+
"grad_norm": 38.091251373291016,
|
| 155 |
+
"learning_rate": 6.968085106382979e-06,
|
| 156 |
+
"loss": 1.8647,
|
| 157 |
+
"step": 525
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7318695941450433,
|
| 161 |
+
"grad_norm": 22.76192855834961,
|
| 162 |
+
"learning_rate": 7.300531914893617e-06,
|
| 163 |
+
"loss": 1.7466,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7651363938789089,
|
| 168 |
+
"grad_norm": 27.64113426208496,
|
| 169 |
+
"learning_rate": 7.632978723404256e-06,
|
| 170 |
+
"loss": 1.9297,
|
| 171 |
+
"step": 575
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.7984031936127745,
|
| 175 |
+
"grad_norm": 27.07349395751953,
|
| 176 |
+
"learning_rate": 7.965425531914895e-06,
|
| 177 |
+
"loss": 1.6819,
|
| 178 |
+
"step": 600
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.83166999334664,
|
| 182 |
+
"grad_norm": 36.454010009765625,
|
| 183 |
+
"learning_rate": 8.297872340425532e-06,
|
| 184 |
+
"loss": 1.7133,
|
| 185 |
+
"step": 625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8649367930805056,
|
| 189 |
+
"grad_norm": 23.96141815185547,
|
| 190 |
+
"learning_rate": 8.630319148936171e-06,
|
| 191 |
+
"loss": 1.8537,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.8982035928143712,
|
| 196 |
+
"grad_norm": 14.69851016998291,
|
| 197 |
+
"learning_rate": 8.96276595744681e-06,
|
| 198 |
+
"loss": 1.8554,
|
| 199 |
+
"step": 675
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9314703925482368,
|
| 203 |
+
"grad_norm": 27.234848022460938,
|
| 204 |
+
"learning_rate": 9.295212765957447e-06,
|
| 205 |
+
"loss": 1.7566,
|
| 206 |
+
"step": 700
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9647371922821024,
|
| 210 |
+
"grad_norm": 27.243558883666992,
|
| 211 |
+
"learning_rate": 9.627659574468086e-06,
|
| 212 |
+
"loss": 1.5349,
|
| 213 |
+
"step": 725
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.998003992015968,
|
| 217 |
+
"grad_norm": 44.65242385864258,
|
| 218 |
+
"learning_rate": 9.960106382978723e-06,
|
| 219 |
+
"loss": 1.6613,
|
| 220 |
+
"step": 750
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0,
|
| 224 |
+
"eval_accuracy": 0.8592814371257484,
|
| 225 |
+
"eval_loss": 0.401942640542984,
|
| 226 |
+
"eval_macro_f1": 0.5934326975706287,
|
| 227 |
+
"eval_runtime": 1.5543,
|
| 228 |
+
"eval_samples_per_second": 1719.061,
|
| 229 |
+
"eval_steps_per_second": 107.441,
|
| 230 |
+
"eval_weighted_f1": 0.8410192971389543,
|
| 231 |
+
"step": 752
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0306054557551563,
|
| 235 |
+
"grad_norm": 27.092920303344727,
|
| 236 |
+
"learning_rate": 1.0292553191489364e-05,
|
| 237 |
+
"loss": 1.621,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.0638722554890219,
|
| 242 |
+
"grad_norm": 28.554630279541016,
|
| 243 |
+
"learning_rate": 1.0625e-05,
|
| 244 |
+
"loss": 1.7632,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.0971390552228875,
|
| 249 |
+
"grad_norm": 18.251523971557617,
|
| 250 |
+
"learning_rate": 1.095744680851064e-05,
|
| 251 |
+
"loss": 1.6875,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.130405854956753,
|
| 256 |
+
"grad_norm": 25.526334762573242,
|
| 257 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 258 |
+
"loss": 1.5827,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.1636726546906186,
|
| 263 |
+
"grad_norm": 34.136436462402344,
|
| 264 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 265 |
+
"loss": 1.5267,
|
| 266 |
+
"step": 875
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.1969394544244845,
|
| 270 |
+
"grad_norm": 38.70103454589844,
|
| 271 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 272 |
+
"loss": 1.5552,
|
| 273 |
+
"step": 900
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.23020625415835,
|
| 277 |
+
"grad_norm": 30.68389320373535,
|
| 278 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 279 |
+
"loss": 1.5197,
|
| 280 |
+
"step": 925
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.2634730538922156,
|
| 284 |
+
"grad_norm": 28.493099212646484,
|
| 285 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 286 |
+
"loss": 1.6468,
|
| 287 |
+
"step": 950
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.2967398536260812,
|
| 291 |
+
"grad_norm": 25.333066940307617,
|
| 292 |
+
"learning_rate": 1.295212765957447e-05,
|
| 293 |
+
"loss": 1.3778,
|
| 294 |
+
"step": 975
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.3300066533599468,
|
| 298 |
+
"grad_norm": 21.102970123291016,
|
| 299 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 300 |
+
"loss": 1.5565,
|
| 301 |
+
"step": 1000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.3632734530938124,
|
| 305 |
+
"grad_norm": 26.531511306762695,
|
| 306 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 307 |
+
"loss": 1.4961,
|
| 308 |
+
"step": 1025
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.396540252827678,
|
| 312 |
+
"grad_norm": 23.80035400390625,
|
| 313 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 314 |
+
"loss": 1.6439,
|
| 315 |
+
"step": 1050
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.4298070525615436,
|
| 319 |
+
"grad_norm": 29.40401840209961,
|
| 320 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 321 |
+
"loss": 1.6489,
|
| 322 |
+
"step": 1075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.4630738522954092,
|
| 326 |
+
"grad_norm": 23.498512268066406,
|
| 327 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 328 |
+
"loss": 1.7361,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.4963406520292748,
|
| 333 |
+
"grad_norm": 25.611040115356445,
|
| 334 |
+
"learning_rate": 1.49468085106383e-05,
|
| 335 |
+
"loss": 1.5585,
|
| 336 |
+
"step": 1125
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.5296074517631404,
|
| 340 |
+
"grad_norm": 22.148107528686523,
|
| 341 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 342 |
+
"loss": 1.7262,
|
| 343 |
+
"step": 1150
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.562874251497006,
|
| 347 |
+
"grad_norm": 18.81505012512207,
|
| 348 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 349 |
+
"loss": 1.5674,
|
| 350 |
+
"step": 1175
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5961410512308716,
|
| 354 |
+
"grad_norm": 41.842350006103516,
|
| 355 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 356 |
+
"loss": 1.6489,
|
| 357 |
+
"step": 1200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6294078509647372,
|
| 361 |
+
"grad_norm": 18.060359954833984,
|
| 362 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 363 |
+
"loss": 1.4792,
|
| 364 |
+
"step": 1225
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.6626746506986028,
|
| 368 |
+
"grad_norm": 26.489307403564453,
|
| 369 |
+
"learning_rate": 1.660904255319149e-05,
|
| 370 |
+
"loss": 1.6544,
|
| 371 |
+
"step": 1250
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.6959414504324684,
|
| 375 |
+
"grad_norm": 22.181209564208984,
|
| 376 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 377 |
+
"loss": 1.4514,
|
| 378 |
+
"step": 1275
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.729208250166334,
|
| 382 |
+
"grad_norm": 14.94635009765625,
|
| 383 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 384 |
+
"loss": 1.3136,
|
| 385 |
+
"step": 1300
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.7624750499001998,
|
| 389 |
+
"grad_norm": 14.453506469726562,
|
| 390 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 391 |
+
"loss": 1.6663,
|
| 392 |
+
"step": 1325
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.7957418496340654,
|
| 396 |
+
"grad_norm": 16.815460205078125,
|
| 397 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 398 |
+
"loss": 1.4437,
|
| 399 |
+
"step": 1350
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.829008649367931,
|
| 403 |
+
"grad_norm": 20.052541732788086,
|
| 404 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 405 |
+
"loss": 1.3574,
|
| 406 |
+
"step": 1375
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.8622754491017965,
|
| 410 |
+
"grad_norm": 17.54789161682129,
|
| 411 |
+
"learning_rate": 1.860372340425532e-05,
|
| 412 |
+
"loss": 1.4221,
|
| 413 |
+
"step": 1400
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.8955422488356621,
|
| 417 |
+
"grad_norm": 25.357349395751953,
|
| 418 |
+
"learning_rate": 1.893617021276596e-05,
|
| 419 |
+
"loss": 1.5318,
|
| 420 |
+
"step": 1425
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.9288090485695277,
|
| 424 |
+
"grad_norm": 18.421524047851562,
|
| 425 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 426 |
+
"loss": 1.5834,
|
| 427 |
+
"step": 1450
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.9620758483033933,
|
| 431 |
+
"grad_norm": 34.23824691772461,
|
| 432 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 433 |
+
"loss": 1.4096,
|
| 434 |
+
"step": 1475
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.995342648037259,
|
| 438 |
+
"grad_norm": 21.318138122558594,
|
| 439 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 440 |
+
"loss": 1.5665,
|
| 441 |
+
"step": 1500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0,
|
| 445 |
+
"eval_accuracy": 0.8581586826347305,
|
| 446 |
+
"eval_loss": 0.37113580107688904,
|
| 447 |
+
"eval_macro_f1": 0.6351347089894025,
|
| 448 |
+
"eval_runtime": 1.527,
|
| 449 |
+
"eval_samples_per_second": 1749.794,
|
| 450 |
+
"eval_steps_per_second": 109.362,
|
| 451 |
+
"eval_weighted_f1": 0.8540414267172031,
|
| 452 |
+
"step": 1504
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.027944111776447,
|
| 456 |
+
"grad_norm": 30.614301681518555,
|
| 457 |
+
"learning_rate": 1.9970449172576832e-05,
|
| 458 |
+
"loss": 1.3339,
|
| 459 |
+
"step": 1525
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0612109115103125,
|
| 463 |
+
"grad_norm": 21.835119247436523,
|
| 464 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 465 |
+
"loss": 1.2229,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.094477711244178,
|
| 470 |
+
"grad_norm": 36.414981842041016,
|
| 471 |
+
"learning_rate": 1.9896572104018914e-05,
|
| 472 |
+
"loss": 1.3,
|
| 473 |
+
"step": 1575
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.1277445109780437,
|
| 477 |
+
"grad_norm": 19.14915657043457,
|
| 478 |
+
"learning_rate": 1.9859633569739953e-05,
|
| 479 |
+
"loss": 1.4414,
|
| 480 |
+
"step": 1600
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.1610113107119093,
|
| 484 |
+
"grad_norm": 25.200891494750977,
|
| 485 |
+
"learning_rate": 1.9822695035460996e-05,
|
| 486 |
+
"loss": 1.4353,
|
| 487 |
+
"step": 1625
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.194278110445775,
|
| 491 |
+
"grad_norm": 23.982807159423828,
|
| 492 |
+
"learning_rate": 1.9785756501182035e-05,
|
| 493 |
+
"loss": 1.159,
|
| 494 |
+
"step": 1650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.2275449101796405,
|
| 498 |
+
"grad_norm": 23.8933162689209,
|
| 499 |
+
"learning_rate": 1.9748817966903074e-05,
|
| 500 |
+
"loss": 1.2949,
|
| 501 |
+
"step": 1675
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.260811709913506,
|
| 505 |
+
"grad_norm": 24.900924682617188,
|
| 506 |
+
"learning_rate": 1.9711879432624113e-05,
|
| 507 |
+
"loss": 1.4499,
|
| 508 |
+
"step": 1700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.294078509647372,
|
| 512 |
+
"grad_norm": 26.51698112487793,
|
| 513 |
+
"learning_rate": 1.9674940898345156e-05,
|
| 514 |
+
"loss": 1.4441,
|
| 515 |
+
"step": 1725
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.3273453093812373,
|
| 519 |
+
"grad_norm": 30.009052276611328,
|
| 520 |
+
"learning_rate": 1.9638002364066195e-05,
|
| 521 |
+
"loss": 1.1753,
|
| 522 |
+
"step": 1750
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.3606121091151033,
|
| 526 |
+
"grad_norm": 27.629823684692383,
|
| 527 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 528 |
+
"loss": 1.2432,
|
| 529 |
+
"step": 1775
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.393878908848969,
|
| 533 |
+
"grad_norm": 14.804618835449219,
|
| 534 |
+
"learning_rate": 1.9564125295508277e-05,
|
| 535 |
+
"loss": 1.303,
|
| 536 |
+
"step": 1800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.4271457085828345,
|
| 540 |
+
"grad_norm": 18.46533966064453,
|
| 541 |
+
"learning_rate": 1.9527186761229316e-05,
|
| 542 |
+
"loss": 1.3121,
|
| 543 |
+
"step": 1825
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.4604125083167,
|
| 547 |
+
"grad_norm": 42.512123107910156,
|
| 548 |
+
"learning_rate": 1.9490248226950355e-05,
|
| 549 |
+
"loss": 1.2911,
|
| 550 |
+
"step": 1850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.4936793080505657,
|
| 554 |
+
"grad_norm": 34.890411376953125,
|
| 555 |
+
"learning_rate": 1.9453309692671394e-05,
|
| 556 |
+
"loss": 1.4884,
|
| 557 |
+
"step": 1875
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.5269461077844313,
|
| 561 |
+
"grad_norm": 26.789302825927734,
|
| 562 |
+
"learning_rate": 1.9416371158392437e-05,
|
| 563 |
+
"loss": 1.4215,
|
| 564 |
+
"step": 1900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.560212907518297,
|
| 568 |
+
"grad_norm": 22.254226684570312,
|
| 569 |
+
"learning_rate": 1.9379432624113476e-05,
|
| 570 |
+
"loss": 1.2067,
|
| 571 |
+
"step": 1925
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.5934797072521625,
|
| 575 |
+
"grad_norm": 13.718234062194824,
|
| 576 |
+
"learning_rate": 1.9342494089834515e-05,
|
| 577 |
+
"loss": 1.3911,
|
| 578 |
+
"step": 1950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.626746506986028,
|
| 582 |
+
"grad_norm": 35.57254409790039,
|
| 583 |
+
"learning_rate": 1.9305555555555558e-05,
|
| 584 |
+
"loss": 1.2174,
|
| 585 |
+
"step": 1975
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.6600133067198937,
|
| 589 |
+
"grad_norm": 18.326045989990234,
|
| 590 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 591 |
+
"loss": 1.1263,
|
| 592 |
+
"step": 2000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.6932801064537593,
|
| 596 |
+
"grad_norm": 36.05733108520508,
|
| 597 |
+
"learning_rate": 1.9231678486997636e-05,
|
| 598 |
+
"loss": 1.3456,
|
| 599 |
+
"step": 2025
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.726546906187625,
|
| 603 |
+
"grad_norm": 19.417469024658203,
|
| 604 |
+
"learning_rate": 1.919473995271868e-05,
|
| 605 |
+
"loss": 1.0996,
|
| 606 |
+
"step": 2050
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.7598137059214904,
|
| 610 |
+
"grad_norm": 35.408267974853516,
|
| 611 |
+
"learning_rate": 1.9157801418439718e-05,
|
| 612 |
+
"loss": 1.4205,
|
| 613 |
+
"step": 2075
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.793080505655356,
|
| 617 |
+
"grad_norm": 22.593019485473633,
|
| 618 |
+
"learning_rate": 1.9120862884160757e-05,
|
| 619 |
+
"loss": 1.1586,
|
| 620 |
+
"step": 2100
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.8263473053892216,
|
| 624 |
+
"grad_norm": 35.65427780151367,
|
| 625 |
+
"learning_rate": 1.9083924349881797e-05,
|
| 626 |
+
"loss": 1.3719,
|
| 627 |
+
"step": 2125
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.8596141051230872,
|
| 631 |
+
"grad_norm": 27.764564514160156,
|
| 632 |
+
"learning_rate": 1.904698581560284e-05,
|
| 633 |
+
"loss": 1.2774,
|
| 634 |
+
"step": 2150
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.892880904856953,
|
| 638 |
+
"grad_norm": 54.58634948730469,
|
| 639 |
+
"learning_rate": 1.9010047281323878e-05,
|
| 640 |
+
"loss": 1.1423,
|
| 641 |
+
"step": 2175
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.9261477045908184,
|
| 645 |
+
"grad_norm": 13.277375221252441,
|
| 646 |
+
"learning_rate": 1.897310874704492e-05,
|
| 647 |
+
"loss": 1.3565,
|
| 648 |
+
"step": 2200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.959414504324684,
|
| 652 |
+
"grad_norm": 20.953813552856445,
|
| 653 |
+
"learning_rate": 1.893617021276596e-05,
|
| 654 |
+
"loss": 1.2279,
|
| 655 |
+
"step": 2225
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.9926813040585496,
|
| 659 |
+
"grad_norm": 19.979778289794922,
|
| 660 |
+
"learning_rate": 1.8899231678487e-05,
|
| 661 |
+
"loss": 1.2266,
|
| 662 |
+
"step": 2250
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.0,
|
| 666 |
+
"eval_accuracy": 0.8712574850299402,
|
| 667 |
+
"eval_loss": 0.3697344958782196,
|
| 668 |
+
"eval_macro_f1": 0.6181104325673971,
|
| 669 |
+
"eval_runtime": 1.5151,
|
| 670 |
+
"eval_samples_per_second": 1763.605,
|
| 671 |
+
"eval_steps_per_second": 110.225,
|
| 672 |
+
"eval_weighted_f1": 0.8580288580313121,
|
| 673 |
+
"step": 2256
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.0252827677977376,
|
| 677 |
+
"grad_norm": 27.134233474731445,
|
| 678 |
+
"learning_rate": 1.886229314420804e-05,
|
| 679 |
+
"loss": 1.1056,
|
| 680 |
+
"step": 2275
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.0585495675316037,
|
| 684 |
+
"grad_norm": 22.94137191772461,
|
| 685 |
+
"learning_rate": 1.8825354609929078e-05,
|
| 686 |
+
"loss": 1.1479,
|
| 687 |
+
"step": 2300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.0918163672654693,
|
| 691 |
+
"grad_norm": 16.680999755859375,
|
| 692 |
+
"learning_rate": 1.878841607565012e-05,
|
| 693 |
+
"loss": 1.0197,
|
| 694 |
+
"step": 2325
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.125083166999335,
|
| 698 |
+
"grad_norm": 28.6317138671875,
|
| 699 |
+
"learning_rate": 1.875147754137116e-05,
|
| 700 |
+
"loss": 1.0265,
|
| 701 |
+
"step": 2350
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.1583499667332005,
|
| 705 |
+
"grad_norm": 24.19533348083496,
|
| 706 |
+
"learning_rate": 1.8714539007092202e-05,
|
| 707 |
+
"loss": 1.0724,
|
| 708 |
+
"step": 2375
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.191616766467066,
|
| 712 |
+
"grad_norm": 20.041053771972656,
|
| 713 |
+
"learning_rate": 1.867760047281324e-05,
|
| 714 |
+
"loss": 1.0137,
|
| 715 |
+
"step": 2400
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.2248835662009316,
|
| 719 |
+
"grad_norm": 49.500205993652344,
|
| 720 |
+
"learning_rate": 1.864066193853428e-05,
|
| 721 |
+
"loss": 1.0362,
|
| 722 |
+
"step": 2425
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.2581503659347972,
|
| 726 |
+
"grad_norm": 33.01273727416992,
|
| 727 |
+
"learning_rate": 1.860372340425532e-05,
|
| 728 |
+
"loss": 1.0825,
|
| 729 |
+
"step": 2450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.291417165668663,
|
| 733 |
+
"grad_norm": 29.107446670532227,
|
| 734 |
+
"learning_rate": 1.856678486997636e-05,
|
| 735 |
+
"loss": 1.1413,
|
| 736 |
+
"step": 2475
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.3246839654025284,
|
| 740 |
+
"grad_norm": 26.056676864624023,
|
| 741 |
+
"learning_rate": 1.85298463356974e-05,
|
| 742 |
+
"loss": 1.0489,
|
| 743 |
+
"step": 2500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.357950765136394,
|
| 747 |
+
"grad_norm": 20.73958396911621,
|
| 748 |
+
"learning_rate": 1.849290780141844e-05,
|
| 749 |
+
"loss": 1.0098,
|
| 750 |
+
"step": 2525
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.3912175648702596,
|
| 754 |
+
"grad_norm": 25.818767547607422,
|
| 755 |
+
"learning_rate": 1.8455969267139483e-05,
|
| 756 |
+
"loss": 0.9775,
|
| 757 |
+
"step": 2550
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.424484364604125,
|
| 761 |
+
"grad_norm": 34.23649978637695,
|
| 762 |
+
"learning_rate": 1.8419030732860522e-05,
|
| 763 |
+
"loss": 1.179,
|
| 764 |
+
"step": 2575
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.457751164337991,
|
| 768 |
+
"grad_norm": 30.995899200439453,
|
| 769 |
+
"learning_rate": 1.838209219858156e-05,
|
| 770 |
+
"loss": 0.9971,
|
| 771 |
+
"step": 2600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.4910179640718564,
|
| 775 |
+
"grad_norm": 16.175426483154297,
|
| 776 |
+
"learning_rate": 1.83451536643026e-05,
|
| 777 |
+
"loss": 0.9482,
|
| 778 |
+
"step": 2625
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.524284763805722,
|
| 782 |
+
"grad_norm": 20.30826187133789,
|
| 783 |
+
"learning_rate": 1.8308215130023643e-05,
|
| 784 |
+
"loss": 1.2234,
|
| 785 |
+
"step": 2650
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.5575515635395876,
|
| 789 |
+
"grad_norm": 46.85417175292969,
|
| 790 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 791 |
+
"loss": 1.094,
|
| 792 |
+
"step": 2675
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.590818363273453,
|
| 796 |
+
"grad_norm": 25.728748321533203,
|
| 797 |
+
"learning_rate": 1.823433806146572e-05,
|
| 798 |
+
"loss": 1.0115,
|
| 799 |
+
"step": 2700
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.6240851630073188,
|
| 803 |
+
"grad_norm": 30.966445922851562,
|
| 804 |
+
"learning_rate": 1.8197399527186764e-05,
|
| 805 |
+
"loss": 1.1783,
|
| 806 |
+
"step": 2725
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.6573519627411843,
|
| 810 |
+
"grad_norm": 33.87657928466797,
|
| 811 |
+
"learning_rate": 1.8160460992907803e-05,
|
| 812 |
+
"loss": 1.2113,
|
| 813 |
+
"step": 2750
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.69061876247505,
|
| 817 |
+
"grad_norm": 27.6673641204834,
|
| 818 |
+
"learning_rate": 1.8123522458628842e-05,
|
| 819 |
+
"loss": 1.1515,
|
| 820 |
+
"step": 2775
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.7238855622089155,
|
| 824 |
+
"grad_norm": 24.126436233520508,
|
| 825 |
+
"learning_rate": 1.808658392434988e-05,
|
| 826 |
+
"loss": 0.9524,
|
| 827 |
+
"step": 2800
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.757152361942781,
|
| 831 |
+
"grad_norm": 48.614654541015625,
|
| 832 |
+
"learning_rate": 1.8049645390070924e-05,
|
| 833 |
+
"loss": 0.8609,
|
| 834 |
+
"step": 2825
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.7904191616766467,
|
| 838 |
+
"grad_norm": 28.923856735229492,
|
| 839 |
+
"learning_rate": 1.8012706855791963e-05,
|
| 840 |
+
"loss": 1.1387,
|
| 841 |
+
"step": 2850
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.8236859614105123,
|
| 845 |
+
"grad_norm": 41.61885070800781,
|
| 846 |
+
"learning_rate": 1.7975768321513003e-05,
|
| 847 |
+
"loss": 1.0666,
|
| 848 |
+
"step": 2875
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.856952761144378,
|
| 852 |
+
"grad_norm": 16.2569522857666,
|
| 853 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 854 |
+
"loss": 1.1988,
|
| 855 |
+
"step": 2900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.8902195608782435,
|
| 859 |
+
"grad_norm": 15.537834167480469,
|
| 860 |
+
"learning_rate": 1.7901891252955084e-05,
|
| 861 |
+
"loss": 1.0413,
|
| 862 |
+
"step": 2925
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.923486360612109,
|
| 866 |
+
"grad_norm": 30.901418685913086,
|
| 867 |
+
"learning_rate": 1.7864952718676124e-05,
|
| 868 |
+
"loss": 0.9529,
|
| 869 |
+
"step": 2950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.9567531603459747,
|
| 873 |
+
"grad_norm": 33.222023010253906,
|
| 874 |
+
"learning_rate": 1.7828014184397163e-05,
|
| 875 |
+
"loss": 0.8885,
|
| 876 |
+
"step": 2975
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.9900199600798403,
|
| 880 |
+
"grad_norm": 34.731224060058594,
|
| 881 |
+
"learning_rate": 1.7791075650118205e-05,
|
| 882 |
+
"loss": 1.1978,
|
| 883 |
+
"step": 3000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 4.0,
|
| 887 |
+
"eval_accuracy": 0.8660179640718563,
|
| 888 |
+
"eval_loss": 0.35781410336494446,
|
| 889 |
+
"eval_macro_f1": 0.663992514269203,
|
| 890 |
+
"eval_runtime": 1.5449,
|
| 891 |
+
"eval_samples_per_second": 1729.565,
|
| 892 |
+
"eval_steps_per_second": 108.098,
|
| 893 |
+
"eval_weighted_f1": 0.865035833325726,
|
| 894 |
+
"step": 3008
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.022621423819029,
|
| 898 |
+
"grad_norm": 19.01364517211914,
|
| 899 |
+
"learning_rate": 1.7754137115839245e-05,
|
| 900 |
+
"loss": 0.8489,
|
| 901 |
+
"step": 3025
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 4.055888223552894,
|
| 905 |
+
"grad_norm": 26.636028289794922,
|
| 906 |
+
"learning_rate": 1.7717198581560284e-05,
|
| 907 |
+
"loss": 0.8252,
|
| 908 |
+
"step": 3050
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 4.08915502328676,
|
| 912 |
+
"grad_norm": 33.308753967285156,
|
| 913 |
+
"learning_rate": 1.7680260047281326e-05,
|
| 914 |
+
"loss": 0.7289,
|
| 915 |
+
"step": 3075
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.122421823020625,
|
| 919 |
+
"grad_norm": 11.095755577087402,
|
| 920 |
+
"learning_rate": 1.7643321513002365e-05,
|
| 921 |
+
"loss": 0.8243,
|
| 922 |
+
"step": 3100
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 4.155688622754491,
|
| 926 |
+
"grad_norm": 21.71223258972168,
|
| 927 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 928 |
+
"loss": 1.0173,
|
| 929 |
+
"step": 3125
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 4.188955422488356,
|
| 933 |
+
"grad_norm": 26.51653480529785,
|
| 934 |
+
"learning_rate": 1.7569444444444447e-05,
|
| 935 |
+
"loss": 0.868,
|
| 936 |
+
"step": 3150
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 4.222222222222222,
|
| 940 |
+
"grad_norm": 22.54006004333496,
|
| 941 |
+
"learning_rate": 1.7532505910165486e-05,
|
| 942 |
+
"loss": 0.7639,
|
| 943 |
+
"step": 3175
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 4.2554890219560875,
|
| 947 |
+
"grad_norm": 64.81362915039062,
|
| 948 |
+
"learning_rate": 1.7495567375886526e-05,
|
| 949 |
+
"loss": 0.9346,
|
| 950 |
+
"step": 3200
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 4.2887558216899535,
|
| 954 |
+
"grad_norm": 47.427085876464844,
|
| 955 |
+
"learning_rate": 1.7458628841607565e-05,
|
| 956 |
+
"loss": 0.8158,
|
| 957 |
+
"step": 3225
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 4.322022621423819,
|
| 961 |
+
"grad_norm": 15.681526184082031,
|
| 962 |
+
"learning_rate": 1.7421690307328607e-05,
|
| 963 |
+
"loss": 0.7279,
|
| 964 |
+
"step": 3250
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 4.355289421157685,
|
| 968 |
+
"grad_norm": 29.830162048339844,
|
| 969 |
+
"learning_rate": 1.7384751773049647e-05,
|
| 970 |
+
"loss": 0.6306,
|
| 971 |
+
"step": 3275
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 4.38855622089155,
|
| 975 |
+
"grad_norm": 22.46377182006836,
|
| 976 |
+
"learning_rate": 1.734781323877069e-05,
|
| 977 |
+
"loss": 0.7833,
|
| 978 |
+
"step": 3300
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.421823020625416,
|
| 982 |
+
"grad_norm": 43.262149810791016,
|
| 983 |
+
"learning_rate": 1.731087470449173e-05,
|
| 984 |
+
"loss": 0.8915,
|
| 985 |
+
"step": 3325
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 4.455089820359281,
|
| 989 |
+
"grad_norm": 38.07517623901367,
|
| 990 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 991 |
+
"loss": 0.8497,
|
| 992 |
+
"step": 3350
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.488356620093147,
|
| 996 |
+
"grad_norm": 54.30280685424805,
|
| 997 |
+
"learning_rate": 1.7236997635933807e-05,
|
| 998 |
+
"loss": 0.7794,
|
| 999 |
+
"step": 3375
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.521623419827012,
|
| 1003 |
+
"grad_norm": 72.16100311279297,
|
| 1004 |
+
"learning_rate": 1.7200059101654846e-05,
|
| 1005 |
+
"loss": 0.8375,
|
| 1006 |
+
"step": 3400
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.554890219560878,
|
| 1010 |
+
"grad_norm": 40.60910415649414,
|
| 1011 |
+
"learning_rate": 1.716312056737589e-05,
|
| 1012 |
+
"loss": 0.889,
|
| 1013 |
+
"step": 3425
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.588157019294744,
|
| 1017 |
+
"grad_norm": 52.0064582824707,
|
| 1018 |
+
"learning_rate": 1.7126182033096928e-05,
|
| 1019 |
+
"loss": 0.8153,
|
| 1020 |
+
"step": 3450
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.621423819028609,
|
| 1024 |
+
"grad_norm": 38.7625846862793,
|
| 1025 |
+
"learning_rate": 1.708924349881797e-05,
|
| 1026 |
+
"loss": 0.9168,
|
| 1027 |
+
"step": 3475
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 4.654690618762475,
|
| 1031 |
+
"grad_norm": 31.68454360961914,
|
| 1032 |
+
"learning_rate": 1.705230496453901e-05,
|
| 1033 |
+
"loss": 0.7938,
|
| 1034 |
+
"step": 3500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.687957418496341,
|
| 1038 |
+
"grad_norm": 28.17580223083496,
|
| 1039 |
+
"learning_rate": 1.701536643026005e-05,
|
| 1040 |
+
"loss": 0.9412,
|
| 1041 |
+
"step": 3525
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.721224218230207,
|
| 1045 |
+
"grad_norm": 22.04960060119629,
|
| 1046 |
+
"learning_rate": 1.6978427895981088e-05,
|
| 1047 |
+
"loss": 0.8231,
|
| 1048 |
+
"step": 3550
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.754491017964072,
|
| 1052 |
+
"grad_norm": 45.03123474121094,
|
| 1053 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 1054 |
+
"loss": 0.8961,
|
| 1055 |
+
"step": 3575
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.787757817697938,
|
| 1059 |
+
"grad_norm": 17.203054428100586,
|
| 1060 |
+
"learning_rate": 1.690455082742317e-05,
|
| 1061 |
+
"loss": 0.8676,
|
| 1062 |
+
"step": 3600
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.821024617431803,
|
| 1066 |
+
"grad_norm": 25.068689346313477,
|
| 1067 |
+
"learning_rate": 1.686761229314421e-05,
|
| 1068 |
+
"loss": 0.7191,
|
| 1069 |
+
"step": 3625
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.854291417165669,
|
| 1073 |
+
"grad_norm": 18.855440139770508,
|
| 1074 |
+
"learning_rate": 1.683067375886525e-05,
|
| 1075 |
+
"loss": 0.9022,
|
| 1076 |
+
"step": 3650
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.887558216899534,
|
| 1080 |
+
"grad_norm": 39.43722915649414,
|
| 1081 |
+
"learning_rate": 1.679373522458629e-05,
|
| 1082 |
+
"loss": 0.9583,
|
| 1083 |
+
"step": 3675
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.9208250166334,
|
| 1087 |
+
"grad_norm": 49.7006950378418,
|
| 1088 |
+
"learning_rate": 1.675679669030733e-05,
|
| 1089 |
+
"loss": 0.781,
|
| 1090 |
+
"step": 3700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.954091816367265,
|
| 1094 |
+
"grad_norm": 54.37468338012695,
|
| 1095 |
+
"learning_rate": 1.671985815602837e-05,
|
| 1096 |
+
"loss": 0.8906,
|
| 1097 |
+
"step": 3725
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.987358616101131,
|
| 1101 |
+
"grad_norm": 52.063228607177734,
|
| 1102 |
+
"learning_rate": 1.6682919621749408e-05,
|
| 1103 |
+
"loss": 0.8871,
|
| 1104 |
+
"step": 3750
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.0,
|
| 1108 |
+
"eval_accuracy": 0.8660179640718563,
|
| 1109 |
+
"eval_loss": 0.46085795760154724,
|
| 1110 |
+
"eval_macro_f1": 0.6784753176089526,
|
| 1111 |
+
"eval_runtime": 1.5487,
|
| 1112 |
+
"eval_samples_per_second": 1725.332,
|
| 1113 |
+
"eval_steps_per_second": 107.833,
|
| 1114 |
+
"eval_weighted_f1": 0.8662687400734955,
|
| 1115 |
+
"step": 3760
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.0199600798403194,
|
| 1119 |
+
"grad_norm": 58.26340866088867,
|
| 1120 |
+
"learning_rate": 1.664598108747045e-05,
|
| 1121 |
+
"loss": 0.6121,
|
| 1122 |
+
"step": 3775
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.053226879574185,
|
| 1126 |
+
"grad_norm": 23.749174118041992,
|
| 1127 |
+
"learning_rate": 1.660904255319149e-05,
|
| 1128 |
+
"loss": 0.5646,
|
| 1129 |
+
"step": 3800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.086493679308051,
|
| 1133 |
+
"grad_norm": 53.65329360961914,
|
| 1134 |
+
"learning_rate": 1.6572104018912532e-05,
|
| 1135 |
+
"loss": 0.624,
|
| 1136 |
+
"step": 3825
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.119760479041916,
|
| 1140 |
+
"grad_norm": 36.52328109741211,
|
| 1141 |
+
"learning_rate": 1.653516548463357e-05,
|
| 1142 |
+
"loss": 0.6409,
|
| 1143 |
+
"step": 3850
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.153027278775782,
|
| 1147 |
+
"grad_norm": 47.592708587646484,
|
| 1148 |
+
"learning_rate": 1.649822695035461e-05,
|
| 1149 |
+
"loss": 0.6812,
|
| 1150 |
+
"step": 3875
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.186294078509647,
|
| 1154 |
+
"grad_norm": 36.178470611572266,
|
| 1155 |
+
"learning_rate": 1.646128841607565e-05,
|
| 1156 |
+
"loss": 0.657,
|
| 1157 |
+
"step": 3900
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.219560878243513,
|
| 1161 |
+
"grad_norm": 63.940887451171875,
|
| 1162 |
+
"learning_rate": 1.642434988179669e-05,
|
| 1163 |
+
"loss": 0.744,
|
| 1164 |
+
"step": 3925
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.252827677977379,
|
| 1168 |
+
"grad_norm": 41.73306655883789,
|
| 1169 |
+
"learning_rate": 1.6387411347517732e-05,
|
| 1170 |
+
"loss": 0.7452,
|
| 1171 |
+
"step": 3950
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.286094477711244,
|
| 1175 |
+
"grad_norm": 23.160594940185547,
|
| 1176 |
+
"learning_rate": 1.635047281323877e-05,
|
| 1177 |
+
"loss": 0.6592,
|
| 1178 |
+
"step": 3975
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.319361277445109,
|
| 1182 |
+
"grad_norm": 34.606117248535156,
|
| 1183 |
+
"learning_rate": 1.6313534278959813e-05,
|
| 1184 |
+
"loss": 0.7744,
|
| 1185 |
+
"step": 4000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.352628077178975,
|
| 1189 |
+
"grad_norm": 69.0250244140625,
|
| 1190 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 1191 |
+
"loss": 0.6746,
|
| 1192 |
+
"step": 4025
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.385894876912841,
|
| 1196 |
+
"grad_norm": 21.54619789123535,
|
| 1197 |
+
"learning_rate": 1.6239657210401892e-05,
|
| 1198 |
+
"loss": 0.5286,
|
| 1199 |
+
"step": 4050
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.419161676646707,
|
| 1203 |
+
"grad_norm": 43.221920013427734,
|
| 1204 |
+
"learning_rate": 1.620271867612293e-05,
|
| 1205 |
+
"loss": 0.7407,
|
| 1206 |
+
"step": 4075
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.452428476380573,
|
| 1210 |
+
"grad_norm": 40.18708038330078,
|
| 1211 |
+
"learning_rate": 1.6165780141843974e-05,
|
| 1212 |
+
"loss": 0.6513,
|
| 1213 |
+
"step": 4100
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.485695276114438,
|
| 1217 |
+
"grad_norm": 35.7791862487793,
|
| 1218 |
+
"learning_rate": 1.6128841607565013e-05,
|
| 1219 |
+
"loss": 0.8098,
|
| 1220 |
+
"step": 4125
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.518962075848304,
|
| 1224 |
+
"grad_norm": 46.182525634765625,
|
| 1225 |
+
"learning_rate": 1.6091903073286052e-05,
|
| 1226 |
+
"loss": 0.7406,
|
| 1227 |
+
"step": 4150
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.552228875582169,
|
| 1231 |
+
"grad_norm": 34.99147415161133,
|
| 1232 |
+
"learning_rate": 1.6054964539007095e-05,
|
| 1233 |
+
"loss": 0.5702,
|
| 1234 |
+
"step": 4175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.585495675316035,
|
| 1238 |
+
"grad_norm": 29.88373374938965,
|
| 1239 |
+
"learning_rate": 1.6018026004728134e-05,
|
| 1240 |
+
"loss": 0.7216,
|
| 1241 |
+
"step": 4200
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.6187624750499,
|
| 1245 |
+
"grad_norm": 24.29672622680664,
|
| 1246 |
+
"learning_rate": 1.5981087470449176e-05,
|
| 1247 |
+
"loss": 0.6033,
|
| 1248 |
+
"step": 4225
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.652029274783766,
|
| 1252 |
+
"grad_norm": 26.062042236328125,
|
| 1253 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 1254 |
+
"loss": 0.6893,
|
| 1255 |
+
"step": 4250
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.685296074517631,
|
| 1259 |
+
"grad_norm": 23.462562561035156,
|
| 1260 |
+
"learning_rate": 1.5907210401891255e-05,
|
| 1261 |
+
"loss": 0.5354,
|
| 1262 |
+
"step": 4275
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.718562874251497,
|
| 1266 |
+
"grad_norm": 55.017372131347656,
|
| 1267 |
+
"learning_rate": 1.5870271867612294e-05,
|
| 1268 |
+
"loss": 0.5262,
|
| 1269 |
+
"step": 4300
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.7518296739853625,
|
| 1273 |
+
"grad_norm": 24.62998390197754,
|
| 1274 |
+
"learning_rate": 1.5833333333333333e-05,
|
| 1275 |
+
"loss": 0.7531,
|
| 1276 |
+
"step": 4325
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.7850964737192285,
|
| 1280 |
+
"grad_norm": 23.63026237487793,
|
| 1281 |
+
"learning_rate": 1.5796394799054376e-05,
|
| 1282 |
+
"loss": 0.7022,
|
| 1283 |
+
"step": 4350
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.818363273453094,
|
| 1287 |
+
"grad_norm": 41.58454895019531,
|
| 1288 |
+
"learning_rate": 1.5759456264775415e-05,
|
| 1289 |
+
"loss": 0.8444,
|
| 1290 |
+
"step": 4375
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.85163007318696,
|
| 1294 |
+
"grad_norm": 49.92122268676758,
|
| 1295 |
+
"learning_rate": 1.5722517730496457e-05,
|
| 1296 |
+
"loss": 0.7502,
|
| 1297 |
+
"step": 4400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.884896872920825,
|
| 1301 |
+
"grad_norm": 38.15485382080078,
|
| 1302 |
+
"learning_rate": 1.5685579196217497e-05,
|
| 1303 |
+
"loss": 0.5652,
|
| 1304 |
+
"step": 4425
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.918163672654691,
|
| 1308 |
+
"grad_norm": 41.067874908447266,
|
| 1309 |
+
"learning_rate": 1.5648640661938536e-05,
|
| 1310 |
+
"loss": 0.6824,
|
| 1311 |
+
"step": 4450
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.951430472388556,
|
| 1315 |
+
"grad_norm": 47.181846618652344,
|
| 1316 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 1317 |
+
"loss": 0.7414,
|
| 1318 |
+
"step": 4475
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.984697272122422,
|
| 1322 |
+
"grad_norm": 39.01291275024414,
|
| 1323 |
+
"learning_rate": 1.5574763593380614e-05,
|
| 1324 |
+
"loss": 0.6981,
|
| 1325 |
+
"step": 4500
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.0,
|
| 1329 |
+
"eval_accuracy": 0.8764970059880239,
|
| 1330 |
+
"eval_loss": 0.45932307839393616,
|
| 1331 |
+
"eval_macro_f1": 0.6636949067600075,
|
| 1332 |
+
"eval_runtime": 1.5179,
|
| 1333 |
+
"eval_samples_per_second": 1760.342,
|
| 1334 |
+
"eval_steps_per_second": 110.021,
|
| 1335 |
+
"eval_weighted_f1": 0.8683884169779571,
|
| 1336 |
+
"step": 4512
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 6.01729873586161,
|
| 1340 |
+
"grad_norm": 63.35742950439453,
|
| 1341 |
+
"learning_rate": 1.5537825059101657e-05,
|
| 1342 |
+
"loss": 0.64,
|
| 1343 |
+
"step": 4525
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 6.050565535595475,
|
| 1347 |
+
"grad_norm": 92.18560791015625,
|
| 1348 |
+
"learning_rate": 1.5500886524822696e-05,
|
| 1349 |
+
"loss": 0.6807,
|
| 1350 |
+
"step": 4550
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 6.083832335329341,
|
| 1354 |
+
"grad_norm": 34.119571685791016,
|
| 1355 |
+
"learning_rate": 1.546394799054374e-05,
|
| 1356 |
+
"loss": 0.5575,
|
| 1357 |
+
"step": 4575
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 6.117099135063207,
|
| 1361 |
+
"grad_norm": 4.943152904510498,
|
| 1362 |
+
"learning_rate": 1.5427009456264778e-05,
|
| 1363 |
+
"loss": 0.5165,
|
| 1364 |
+
"step": 4600
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 6.1503659347970725,
|
| 1368 |
+
"grad_norm": 85.80406951904297,
|
| 1369 |
+
"learning_rate": 1.5390070921985817e-05,
|
| 1370 |
+
"loss": 0.5439,
|
| 1371 |
+
"step": 4625
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 6.1836327345309385,
|
| 1375 |
+
"grad_norm": 95.17032623291016,
|
| 1376 |
+
"learning_rate": 1.5353132387706856e-05,
|
| 1377 |
+
"loss": 0.7291,
|
| 1378 |
+
"step": 4650
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 6.216899534264804,
|
| 1382 |
+
"grad_norm": 61.599830627441406,
|
| 1383 |
+
"learning_rate": 1.5316193853427895e-05,
|
| 1384 |
+
"loss": 0.6088,
|
| 1385 |
+
"step": 4675
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 6.25016633399867,
|
| 1389 |
+
"grad_norm": 20.04920196533203,
|
| 1390 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 1391 |
+
"loss": 0.4175,
|
| 1392 |
+
"step": 4700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 6.283433133732535,
|
| 1396 |
+
"grad_norm": 34.03755187988281,
|
| 1397 |
+
"learning_rate": 1.5242316784869977e-05,
|
| 1398 |
+
"loss": 0.689,
|
| 1399 |
+
"step": 4725
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 6.316699933466401,
|
| 1403 |
+
"grad_norm": 6.635862350463867,
|
| 1404 |
+
"learning_rate": 1.5205378250591018e-05,
|
| 1405 |
+
"loss": 0.5189,
|
| 1406 |
+
"step": 4750
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 6.349966733200266,
|
| 1410 |
+
"grad_norm": 53.87474822998047,
|
| 1411 |
+
"learning_rate": 1.5168439716312057e-05,
|
| 1412 |
+
"loss": 0.6227,
|
| 1413 |
+
"step": 4775
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 6.383233532934132,
|
| 1417 |
+
"grad_norm": 46.599002838134766,
|
| 1418 |
+
"learning_rate": 1.51315011820331e-05,
|
| 1419 |
+
"loss": 0.534,
|
| 1420 |
+
"step": 4800
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 6.416500332667997,
|
| 1424 |
+
"grad_norm": 33.50902557373047,
|
| 1425 |
+
"learning_rate": 1.5094562647754139e-05,
|
| 1426 |
+
"loss": 0.5636,
|
| 1427 |
+
"step": 4825
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 6.449767132401863,
|
| 1431 |
+
"grad_norm": 1.7829655408859253,
|
| 1432 |
+
"learning_rate": 1.5057624113475178e-05,
|
| 1433 |
+
"loss": 0.4675,
|
| 1434 |
+
"step": 4850
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 6.483033932135728,
|
| 1438 |
+
"grad_norm": 79.02617645263672,
|
| 1439 |
+
"learning_rate": 1.5020685579196219e-05,
|
| 1440 |
+
"loss": 0.6349,
|
| 1441 |
+
"step": 4875
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 6.5163007318695945,
|
| 1445 |
+
"grad_norm": 4.111961364746094,
|
| 1446 |
+
"learning_rate": 1.4983747044917258e-05,
|
| 1447 |
+
"loss": 0.5121,
|
| 1448 |
+
"step": 4900
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 6.54956753160346,
|
| 1452 |
+
"grad_norm": 36.20473098754883,
|
| 1453 |
+
"learning_rate": 1.49468085106383e-05,
|
| 1454 |
+
"loss": 0.4947,
|
| 1455 |
+
"step": 4925
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 6.582834331337326,
|
| 1459 |
+
"grad_norm": 71.45379638671875,
|
| 1460 |
+
"learning_rate": 1.490986997635934e-05,
|
| 1461 |
+
"loss": 0.6426,
|
| 1462 |
+
"step": 4950
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 6.616101131071191,
|
| 1466 |
+
"grad_norm": 19.379718780517578,
|
| 1467 |
+
"learning_rate": 1.487293144208038e-05,
|
| 1468 |
+
"loss": 0.5489,
|
| 1469 |
+
"step": 4975
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 6.649367930805057,
|
| 1473 |
+
"grad_norm": 62.92333221435547,
|
| 1474 |
+
"learning_rate": 1.483599290780142e-05,
|
| 1475 |
+
"loss": 0.5921,
|
| 1476 |
+
"step": 5000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 6.682634730538922,
|
| 1480 |
+
"grad_norm": 38.973960876464844,
|
| 1481 |
+
"learning_rate": 1.4799054373522459e-05,
|
| 1482 |
+
"loss": 0.7771,
|
| 1483 |
+
"step": 5025
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 6.715901530272788,
|
| 1487 |
+
"grad_norm": 34.819427490234375,
|
| 1488 |
+
"learning_rate": 1.47621158392435e-05,
|
| 1489 |
+
"loss": 0.5188,
|
| 1490 |
+
"step": 5050
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.749168330006653,
|
| 1494 |
+
"grad_norm": 14.319246292114258,
|
| 1495 |
+
"learning_rate": 1.472517730496454e-05,
|
| 1496 |
+
"loss": 0.5,
|
| 1497 |
+
"step": 5075
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 6.782435129740519,
|
| 1501 |
+
"grad_norm": 41.829811096191406,
|
| 1502 |
+
"learning_rate": 1.4688238770685582e-05,
|
| 1503 |
+
"loss": 0.4302,
|
| 1504 |
+
"step": 5100
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 6.815701929474384,
|
| 1508 |
+
"grad_norm": 62.93412399291992,
|
| 1509 |
+
"learning_rate": 1.4651300236406621e-05,
|
| 1510 |
+
"loss": 0.5171,
|
| 1511 |
+
"step": 5125
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 6.84896872920825,
|
| 1515 |
+
"grad_norm": 22.422821044921875,
|
| 1516 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 1517 |
+
"loss": 0.61,
|
| 1518 |
+
"step": 5150
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 6.8822355289421155,
|
| 1522 |
+
"grad_norm": 27.06107521057129,
|
| 1523 |
+
"learning_rate": 1.4577423167848701e-05,
|
| 1524 |
+
"loss": 0.5891,
|
| 1525 |
+
"step": 5175
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 6.915502328675982,
|
| 1529 |
+
"grad_norm": 36.85319900512695,
|
| 1530 |
+
"learning_rate": 1.454048463356974e-05,
|
| 1531 |
+
"loss": 0.584,
|
| 1532 |
+
"step": 5200
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 6.948769128409847,
|
| 1536 |
+
"grad_norm": 48.67154312133789,
|
| 1537 |
+
"learning_rate": 1.4503546099290781e-05,
|
| 1538 |
+
"loss": 0.479,
|
| 1539 |
+
"step": 5225
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 6.982035928143713,
|
| 1543 |
+
"grad_norm": 60.36023712158203,
|
| 1544 |
+
"learning_rate": 1.446660756501182e-05,
|
| 1545 |
+
"loss": 0.5479,
|
| 1546 |
+
"step": 5250
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 7.0,
|
| 1550 |
+
"eval_accuracy": 0.8708832335329342,
|
| 1551 |
+
"eval_loss": 0.49154114723205566,
|
| 1552 |
+
"eval_macro_f1": 0.652422018861267,
|
| 1553 |
+
"eval_runtime": 1.55,
|
| 1554 |
+
"eval_samples_per_second": 1723.886,
|
| 1555 |
+
"eval_steps_per_second": 107.743,
|
| 1556 |
+
"eval_weighted_f1": 0.8625886334625845,
|
| 1557 |
+
"step": 5264
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 7.014637391882901,
|
| 1561 |
+
"grad_norm": 41.701473236083984,
|
| 1562 |
+
"learning_rate": 1.4429669030732863e-05,
|
| 1563 |
+
"loss": 0.541,
|
| 1564 |
+
"step": 5275
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 7.047904191616767,
|
| 1568 |
+
"grad_norm": 22.2352237701416,
|
| 1569 |
+
"learning_rate": 1.4392730496453902e-05,
|
| 1570 |
+
"loss": 0.3331,
|
| 1571 |
+
"step": 5300
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 7.081170991350632,
|
| 1575 |
+
"grad_norm": 108.27775573730469,
|
| 1576 |
+
"learning_rate": 1.4355791962174943e-05,
|
| 1577 |
+
"loss": 0.4123,
|
| 1578 |
+
"step": 5325
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 7.114437791084498,
|
| 1582 |
+
"grad_norm": 82.6246337890625,
|
| 1583 |
+
"learning_rate": 1.4318853427895982e-05,
|
| 1584 |
+
"loss": 0.5651,
|
| 1585 |
+
"step": 5350
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 7.147704590818363,
|
| 1589 |
+
"grad_norm": 44.44942092895508,
|
| 1590 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 1591 |
+
"loss": 0.5589,
|
| 1592 |
+
"step": 5375
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 7.180971390552229,
|
| 1596 |
+
"grad_norm": 41.803279876708984,
|
| 1597 |
+
"learning_rate": 1.4244976359338062e-05,
|
| 1598 |
+
"loss": 0.559,
|
| 1599 |
+
"step": 5400
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 7.214238190286094,
|
| 1603 |
+
"grad_norm": 26.026790618896484,
|
| 1604 |
+
"learning_rate": 1.4208037825059103e-05,
|
| 1605 |
+
"loss": 0.4932,
|
| 1606 |
+
"step": 5425
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 7.24750499001996,
|
| 1610 |
+
"grad_norm": 13.97793960571289,
|
| 1611 |
+
"learning_rate": 1.4171099290780144e-05,
|
| 1612 |
+
"loss": 0.4862,
|
| 1613 |
+
"step": 5450
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 7.2807717897538256,
|
| 1617 |
+
"grad_norm": 0.895022451877594,
|
| 1618 |
+
"learning_rate": 1.4134160756501183e-05,
|
| 1619 |
+
"loss": 0.4033,
|
| 1620 |
+
"step": 5475
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 7.314038589487692,
|
| 1624 |
+
"grad_norm": 0.39349615573883057,
|
| 1625 |
+
"learning_rate": 1.4097222222222224e-05,
|
| 1626 |
+
"loss": 0.3902,
|
| 1627 |
+
"step": 5500
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 7.347305389221557,
|
| 1631 |
+
"grad_norm": 52.431339263916016,
|
| 1632 |
+
"learning_rate": 1.4060283687943263e-05,
|
| 1633 |
+
"loss": 0.5077,
|
| 1634 |
+
"step": 5525
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 7.380572188955423,
|
| 1638 |
+
"grad_norm": 42.235626220703125,
|
| 1639 |
+
"learning_rate": 1.4023345153664302e-05,
|
| 1640 |
+
"loss": 0.4848,
|
| 1641 |
+
"step": 5550
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 7.413838988689288,
|
| 1645 |
+
"grad_norm": 61.47570037841797,
|
| 1646 |
+
"learning_rate": 1.3986406619385345e-05,
|
| 1647 |
+
"loss": 0.4168,
|
| 1648 |
+
"step": 5575
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 7.447105788423154,
|
| 1652 |
+
"grad_norm": 14.914517402648926,
|
| 1653 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 1654 |
+
"loss": 0.4495,
|
| 1655 |
+
"step": 5600
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 7.480372588157019,
|
| 1659 |
+
"grad_norm": 44.91154479980469,
|
| 1660 |
+
"learning_rate": 1.3912529550827425e-05,
|
| 1661 |
+
"loss": 0.4127,
|
| 1662 |
+
"step": 5625
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 7.513639387890885,
|
| 1666 |
+
"grad_norm": 30.53471565246582,
|
| 1667 |
+
"learning_rate": 1.3875591016548464e-05,
|
| 1668 |
+
"loss": 0.5258,
|
| 1669 |
+
"step": 5650
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 7.54690618762475,
|
| 1673 |
+
"grad_norm": 101.61994934082031,
|
| 1674 |
+
"learning_rate": 1.3838652482269505e-05,
|
| 1675 |
+
"loss": 0.662,
|
| 1676 |
+
"step": 5675
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 7.580172987358616,
|
| 1680 |
+
"grad_norm": 52.47076416015625,
|
| 1681 |
+
"learning_rate": 1.3801713947990544e-05,
|
| 1682 |
+
"loss": 0.5044,
|
| 1683 |
+
"step": 5700
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 7.6134397870924815,
|
| 1687 |
+
"grad_norm": 1.6895707845687866,
|
| 1688 |
+
"learning_rate": 1.3764775413711583e-05,
|
| 1689 |
+
"loss": 0.663,
|
| 1690 |
+
"step": 5725
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 7.6467065868263475,
|
| 1694 |
+
"grad_norm": 87.18041229248047,
|
| 1695 |
+
"learning_rate": 1.3727836879432626e-05,
|
| 1696 |
+
"loss": 0.423,
|
| 1697 |
+
"step": 5750
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 7.679973386560213,
|
| 1701 |
+
"grad_norm": 66.78807067871094,
|
| 1702 |
+
"learning_rate": 1.3690898345153665e-05,
|
| 1703 |
+
"loss": 0.4324,
|
| 1704 |
+
"step": 5775
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 7.713240186294079,
|
| 1708 |
+
"grad_norm": 45.631587982177734,
|
| 1709 |
+
"learning_rate": 1.3653959810874706e-05,
|
| 1710 |
+
"loss": 0.4707,
|
| 1711 |
+
"step": 5800
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 7.746506986027944,
|
| 1715 |
+
"grad_norm": 48.78377914428711,
|
| 1716 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 1717 |
+
"loss": 0.564,
|
| 1718 |
+
"step": 5825
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 7.77977378576181,
|
| 1722 |
+
"grad_norm": 75.56221008300781,
|
| 1723 |
+
"learning_rate": 1.3580082742316786e-05,
|
| 1724 |
+
"loss": 0.674,
|
| 1725 |
+
"step": 5850
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 7.813040585495675,
|
| 1729 |
+
"grad_norm": 66.46585845947266,
|
| 1730 |
+
"learning_rate": 1.3543144208037825e-05,
|
| 1731 |
+
"loss": 0.4366,
|
| 1732 |
+
"step": 5875
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 7.846307385229541,
|
| 1736 |
+
"grad_norm": 52.85831832885742,
|
| 1737 |
+
"learning_rate": 1.3506205673758865e-05,
|
| 1738 |
+
"loss": 0.4353,
|
| 1739 |
+
"step": 5900
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 7.879574184963406,
|
| 1743 |
+
"grad_norm": 57.075138092041016,
|
| 1744 |
+
"learning_rate": 1.3469267139479907e-05,
|
| 1745 |
+
"loss": 0.5894,
|
| 1746 |
+
"step": 5925
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 7.912840984697272,
|
| 1750 |
+
"grad_norm": 76.97833251953125,
|
| 1751 |
+
"learning_rate": 1.3432328605200946e-05,
|
| 1752 |
+
"loss": 0.4212,
|
| 1753 |
+
"step": 5950
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 7.946107784431137,
|
| 1757 |
+
"grad_norm": 113.99784088134766,
|
| 1758 |
+
"learning_rate": 1.3395390070921987e-05,
|
| 1759 |
+
"loss": 0.5023,
|
| 1760 |
+
"step": 5975
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 7.9793745841650034,
|
| 1764 |
+
"grad_norm": 1.6755313873291016,
|
| 1765 |
+
"learning_rate": 1.3358451536643026e-05,
|
| 1766 |
+
"loss": 0.5633,
|
| 1767 |
+
"step": 6000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 8.0,
|
| 1771 |
+
"eval_accuracy": 0.8667664670658682,
|
| 1772 |
+
"eval_loss": 0.5947781205177307,
|
| 1773 |
+
"eval_macro_f1": 0.6740485232066526,
|
| 1774 |
+
"eval_runtime": 1.5609,
|
| 1775 |
+
"eval_samples_per_second": 1711.817,
|
| 1776 |
+
"eval_steps_per_second": 106.989,
|
| 1777 |
+
"eval_weighted_f1": 0.8676462616309832,
|
| 1778 |
+
"step": 6016
|
| 1779 |
+
}
|
| 1780 |
+
],
|
| 1781 |
+
"logging_steps": 25,
|
| 1782 |
+
"max_steps": 15040,
|
| 1783 |
+
"num_input_tokens_seen": 0,
|
| 1784 |
+
"num_train_epochs": 20,
|
| 1785 |
+
"save_steps": 500,
|
| 1786 |
+
"stateful_callbacks": {
|
| 1787 |
+
"EarlyStoppingCallback": {
|
| 1788 |
+
"args": {
|
| 1789 |
+
"early_stopping_patience": 3,
|
| 1790 |
+
"early_stopping_threshold": 0.0
|
| 1791 |
+
},
|
| 1792 |
+
"attributes": {
|
| 1793 |
+
"early_stopping_patience_counter": 3
|
| 1794 |
+
}
|
| 1795 |
+
},
|
| 1796 |
+
"TrainerControl": {
|
| 1797 |
+
"args": {
|
| 1798 |
+
"should_epoch_stop": false,
|
| 1799 |
+
"should_evaluate": false,
|
| 1800 |
+
"should_log": false,
|
| 1801 |
+
"should_save": true,
|
| 1802 |
+
"should_training_stop": true
|
| 1803 |
+
},
|
| 1804 |
+
"attributes": {}
|
| 1805 |
+
}
|
| 1806 |
+
},
|
| 1807 |
+
"total_flos": 0.0,
|
| 1808 |
+
"train_batch_size": 8,
|
| 1809 |
+
"trial_name": null,
|
| 1810 |
+
"trial_params": null
|
| 1811 |
+
}
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008/trainer_state.json
ADDED
|
@@ -0,0 +1,927 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3008,
|
| 3 |
+
"best_metric": 0.6974282738120429,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008",
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3008,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0332667997338656,
|
| 14 |
+
"grad_norm": 34.15004348754883,
|
| 15 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 16 |
+
"loss": 4.3193,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0665335994677312,
|
| 21 |
+
"grad_norm": 35.99433517456055,
|
| 22 |
+
"learning_rate": 6.515957446808512e-07,
|
| 23 |
+
"loss": 4.0656,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0998003992015968,
|
| 28 |
+
"grad_norm": 64.98133087158203,
|
| 29 |
+
"learning_rate": 9.840425531914895e-07,
|
| 30 |
+
"loss": 3.5941,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1330671989354624,
|
| 35 |
+
"grad_norm": 31.218713760375977,
|
| 36 |
+
"learning_rate": 1.3164893617021277e-06,
|
| 37 |
+
"loss": 2.9458,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16633399866932802,
|
| 42 |
+
"grad_norm": 27.85391616821289,
|
| 43 |
+
"learning_rate": 1.648936170212766e-06,
|
| 44 |
+
"loss": 2.3829,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1996007984031936,
|
| 49 |
+
"grad_norm": 31.22490692138672,
|
| 50 |
+
"learning_rate": 1.9813829787234043e-06,
|
| 51 |
+
"loss": 2.3445,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2328675981370592,
|
| 56 |
+
"grad_norm": 21.99728012084961,
|
| 57 |
+
"learning_rate": 2.3138297872340428e-06,
|
| 58 |
+
"loss": 2.0081,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2661343978709248,
|
| 63 |
+
"grad_norm": 21.359256744384766,
|
| 64 |
+
"learning_rate": 2.646276595744681e-06,
|
| 65 |
+
"loss": 2.1865,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.2994011976047904,
|
| 70 |
+
"grad_norm": 22.947093963623047,
|
| 71 |
+
"learning_rate": 2.978723404255319e-06,
|
| 72 |
+
"loss": 2.1144,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.33266799733865604,
|
| 77 |
+
"grad_norm": 15.964245796203613,
|
| 78 |
+
"learning_rate": 3.311170212765958e-06,
|
| 79 |
+
"loss": 2.2028,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.36593479707252163,
|
| 84 |
+
"grad_norm": 15.935502052307129,
|
| 85 |
+
"learning_rate": 3.643617021276596e-06,
|
| 86 |
+
"loss": 2.0956,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3992015968063872,
|
| 91 |
+
"grad_norm": 28.799392700195312,
|
| 92 |
+
"learning_rate": 3.976063829787235e-06,
|
| 93 |
+
"loss": 2.2256,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4324683965402528,
|
| 98 |
+
"grad_norm": 21.484956741333008,
|
| 99 |
+
"learning_rate": 4.308510638297873e-06,
|
| 100 |
+
"loss": 2.1551,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4657351962741184,
|
| 105 |
+
"grad_norm": 18.2623348236084,
|
| 106 |
+
"learning_rate": 4.640957446808511e-06,
|
| 107 |
+
"loss": 2.063,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.499001996007984,
|
| 112 |
+
"grad_norm": 28.219589233398438,
|
| 113 |
+
"learning_rate": 4.973404255319149e-06,
|
| 114 |
+
"loss": 1.9974,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5322687957418496,
|
| 119 |
+
"grad_norm": 51.115116119384766,
|
| 120 |
+
"learning_rate": 5.305851063829788e-06,
|
| 121 |
+
"loss": 2.0349,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5655355954757152,
|
| 126 |
+
"grad_norm": 34.70194625854492,
|
| 127 |
+
"learning_rate": 5.638297872340426e-06,
|
| 128 |
+
"loss": 1.9449,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5988023952095808,
|
| 133 |
+
"grad_norm": 27.660011291503906,
|
| 134 |
+
"learning_rate": 5.970744680851064e-06,
|
| 135 |
+
"loss": 1.9879,
|
| 136 |
+
"step": 450
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6320691949434465,
|
| 140 |
+
"grad_norm": 52.268550872802734,
|
| 141 |
+
"learning_rate": 6.303191489361703e-06,
|
| 142 |
+
"loss": 1.9235,
|
| 143 |
+
"step": 475
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6653359946773121,
|
| 147 |
+
"grad_norm": 22.55914878845215,
|
| 148 |
+
"learning_rate": 6.635638297872341e-06,
|
| 149 |
+
"loss": 1.9964,
|
| 150 |
+
"step": 500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6986027944111777,
|
| 154 |
+
"grad_norm": 29.36838722229004,
|
| 155 |
+
"learning_rate": 6.968085106382979e-06,
|
| 156 |
+
"loss": 1.7544,
|
| 157 |
+
"step": 525
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7318695941450433,
|
| 161 |
+
"grad_norm": 29.081684112548828,
|
| 162 |
+
"learning_rate": 7.300531914893617e-06,
|
| 163 |
+
"loss": 1.5827,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7651363938789089,
|
| 168 |
+
"grad_norm": 39.61648178100586,
|
| 169 |
+
"learning_rate": 7.632978723404256e-06,
|
| 170 |
+
"loss": 1.7948,
|
| 171 |
+
"step": 575
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.7984031936127745,
|
| 175 |
+
"grad_norm": 31.700763702392578,
|
| 176 |
+
"learning_rate": 7.965425531914895e-06,
|
| 177 |
+
"loss": 1.5821,
|
| 178 |
+
"step": 600
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.83166999334664,
|
| 182 |
+
"grad_norm": 20.465805053710938,
|
| 183 |
+
"learning_rate": 8.297872340425532e-06,
|
| 184 |
+
"loss": 1.5987,
|
| 185 |
+
"step": 625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8649367930805056,
|
| 189 |
+
"grad_norm": 31.21770668029785,
|
| 190 |
+
"learning_rate": 8.630319148936171e-06,
|
| 191 |
+
"loss": 1.6804,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.8982035928143712,
|
| 196 |
+
"grad_norm": 28.831789016723633,
|
| 197 |
+
"learning_rate": 8.96276595744681e-06,
|
| 198 |
+
"loss": 1.7875,
|
| 199 |
+
"step": 675
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9314703925482368,
|
| 203 |
+
"grad_norm": 24.137300491333008,
|
| 204 |
+
"learning_rate": 9.295212765957447e-06,
|
| 205 |
+
"loss": 1.883,
|
| 206 |
+
"step": 700
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9647371922821024,
|
| 210 |
+
"grad_norm": 35.83584976196289,
|
| 211 |
+
"learning_rate": 9.627659574468086e-06,
|
| 212 |
+
"loss": 1.729,
|
| 213 |
+
"step": 725
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.998003992015968,
|
| 217 |
+
"grad_norm": 26.05014419555664,
|
| 218 |
+
"learning_rate": 9.960106382978723e-06,
|
| 219 |
+
"loss": 1.6761,
|
| 220 |
+
"step": 750
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0,
|
| 224 |
+
"eval_accuracy": 0.842814371257485,
|
| 225 |
+
"eval_loss": 0.4147079288959503,
|
| 226 |
+
"eval_macro_f1": 0.5715291591376829,
|
| 227 |
+
"eval_runtime": 1.5487,
|
| 228 |
+
"eval_samples_per_second": 1725.328,
|
| 229 |
+
"eval_steps_per_second": 107.833,
|
| 230 |
+
"eval_weighted_f1": 0.8313170510138047,
|
| 231 |
+
"step": 752
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0306054557551563,
|
| 235 |
+
"grad_norm": 35.20064163208008,
|
| 236 |
+
"learning_rate": 1.0292553191489364e-05,
|
| 237 |
+
"loss": 1.6047,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.0638722554890219,
|
| 242 |
+
"grad_norm": 18.61759376525879,
|
| 243 |
+
"learning_rate": 1.0625e-05,
|
| 244 |
+
"loss": 1.4676,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.0971390552228875,
|
| 249 |
+
"grad_norm": 30.29521369934082,
|
| 250 |
+
"learning_rate": 1.095744680851064e-05,
|
| 251 |
+
"loss": 1.5534,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.130405854956753,
|
| 256 |
+
"grad_norm": 49.2490234375,
|
| 257 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 258 |
+
"loss": 1.6714,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.1636726546906186,
|
| 263 |
+
"grad_norm": 21.129003524780273,
|
| 264 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 265 |
+
"loss": 1.7271,
|
| 266 |
+
"step": 875
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.1969394544244845,
|
| 270 |
+
"grad_norm": 50.25038528442383,
|
| 271 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 272 |
+
"loss": 1.5317,
|
| 273 |
+
"step": 900
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.23020625415835,
|
| 277 |
+
"grad_norm": 37.23767852783203,
|
| 278 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 279 |
+
"loss": 1.6333,
|
| 280 |
+
"step": 925
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.2634730538922156,
|
| 284 |
+
"grad_norm": 17.01172637939453,
|
| 285 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 286 |
+
"loss": 1.5422,
|
| 287 |
+
"step": 950
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.2967398536260812,
|
| 291 |
+
"grad_norm": 21.008949279785156,
|
| 292 |
+
"learning_rate": 1.295212765957447e-05,
|
| 293 |
+
"loss": 1.5974,
|
| 294 |
+
"step": 975
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.3300066533599468,
|
| 298 |
+
"grad_norm": 28.253963470458984,
|
| 299 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 300 |
+
"loss": 1.4676,
|
| 301 |
+
"step": 1000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.3632734530938124,
|
| 305 |
+
"grad_norm": 27.492841720581055,
|
| 306 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 307 |
+
"loss": 1.5826,
|
| 308 |
+
"step": 1025
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.396540252827678,
|
| 312 |
+
"grad_norm": 20.68940544128418,
|
| 313 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 314 |
+
"loss": 1.5191,
|
| 315 |
+
"step": 1050
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.4298070525615436,
|
| 319 |
+
"grad_norm": 34.80696105957031,
|
| 320 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 321 |
+
"loss": 1.5666,
|
| 322 |
+
"step": 1075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.4630738522954092,
|
| 326 |
+
"grad_norm": 43.78744888305664,
|
| 327 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 328 |
+
"loss": 1.4861,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.4963406520292748,
|
| 333 |
+
"grad_norm": 29.718017578125,
|
| 334 |
+
"learning_rate": 1.49468085106383e-05,
|
| 335 |
+
"loss": 1.8153,
|
| 336 |
+
"step": 1125
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.5296074517631404,
|
| 340 |
+
"grad_norm": 25.93970489501953,
|
| 341 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 342 |
+
"loss": 1.6267,
|
| 343 |
+
"step": 1150
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.562874251497006,
|
| 347 |
+
"grad_norm": 21.3642578125,
|
| 348 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 349 |
+
"loss": 1.5005,
|
| 350 |
+
"step": 1175
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5961410512308716,
|
| 354 |
+
"grad_norm": 32.647422790527344,
|
| 355 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 356 |
+
"loss": 1.631,
|
| 357 |
+
"step": 1200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6294078509647372,
|
| 361 |
+
"grad_norm": 13.734174728393555,
|
| 362 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 363 |
+
"loss": 1.4731,
|
| 364 |
+
"step": 1225
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.6626746506986028,
|
| 368 |
+
"grad_norm": 28.250856399536133,
|
| 369 |
+
"learning_rate": 1.660904255319149e-05,
|
| 370 |
+
"loss": 1.5175,
|
| 371 |
+
"step": 1250
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.6959414504324684,
|
| 375 |
+
"grad_norm": 22.97903823852539,
|
| 376 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 377 |
+
"loss": 1.5888,
|
| 378 |
+
"step": 1275
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.729208250166334,
|
| 382 |
+
"grad_norm": 22.442594528198242,
|
| 383 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 384 |
+
"loss": 1.3336,
|
| 385 |
+
"step": 1300
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.7624750499001998,
|
| 389 |
+
"grad_norm": 34.93794250488281,
|
| 390 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 391 |
+
"loss": 1.5893,
|
| 392 |
+
"step": 1325
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.7957418496340654,
|
| 396 |
+
"grad_norm": 29.638469696044922,
|
| 397 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 398 |
+
"loss": 1.3456,
|
| 399 |
+
"step": 1350
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.829008649367931,
|
| 403 |
+
"grad_norm": 23.77605438232422,
|
| 404 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 405 |
+
"loss": 1.5958,
|
| 406 |
+
"step": 1375
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.8622754491017965,
|
| 410 |
+
"grad_norm": 34.77972412109375,
|
| 411 |
+
"learning_rate": 1.860372340425532e-05,
|
| 412 |
+
"loss": 1.5545,
|
| 413 |
+
"step": 1400
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.8955422488356621,
|
| 417 |
+
"grad_norm": 18.019433975219727,
|
| 418 |
+
"learning_rate": 1.893617021276596e-05,
|
| 419 |
+
"loss": 1.3322,
|
| 420 |
+
"step": 1425
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.9288090485695277,
|
| 424 |
+
"grad_norm": 25.30364227294922,
|
| 425 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 426 |
+
"loss": 1.625,
|
| 427 |
+
"step": 1450
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.9620758483033933,
|
| 431 |
+
"grad_norm": 18.51852035522461,
|
| 432 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 433 |
+
"loss": 1.4318,
|
| 434 |
+
"step": 1475
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.995342648037259,
|
| 438 |
+
"grad_norm": 26.217803955078125,
|
| 439 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 440 |
+
"loss": 1.4811,
|
| 441 |
+
"step": 1500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0,
|
| 445 |
+
"eval_accuracy": 0.8619011976047904,
|
| 446 |
+
"eval_loss": 0.3896377980709076,
|
| 447 |
+
"eval_macro_f1": 0.5782126976453429,
|
| 448 |
+
"eval_runtime": 1.5856,
|
| 449 |
+
"eval_samples_per_second": 1685.217,
|
| 450 |
+
"eval_steps_per_second": 105.326,
|
| 451 |
+
"eval_weighted_f1": 0.8409251664533528,
|
| 452 |
+
"step": 1504
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.027944111776447,
|
| 456 |
+
"grad_norm": 56.660160064697266,
|
| 457 |
+
"learning_rate": 1.9970449172576832e-05,
|
| 458 |
+
"loss": 1.4506,
|
| 459 |
+
"step": 1525
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0612109115103125,
|
| 463 |
+
"grad_norm": 37.306522369384766,
|
| 464 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 465 |
+
"loss": 1.4542,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.094477711244178,
|
| 470 |
+
"grad_norm": 22.323488235473633,
|
| 471 |
+
"learning_rate": 1.9896572104018914e-05,
|
| 472 |
+
"loss": 1.4015,
|
| 473 |
+
"step": 1575
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.1277445109780437,
|
| 477 |
+
"grad_norm": 39.91142654418945,
|
| 478 |
+
"learning_rate": 1.9859633569739953e-05,
|
| 479 |
+
"loss": 1.3846,
|
| 480 |
+
"step": 1600
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.1610113107119093,
|
| 484 |
+
"grad_norm": 26.0784969329834,
|
| 485 |
+
"learning_rate": 1.9822695035460996e-05,
|
| 486 |
+
"loss": 1.3649,
|
| 487 |
+
"step": 1625
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.194278110445775,
|
| 491 |
+
"grad_norm": 39.20205307006836,
|
| 492 |
+
"learning_rate": 1.9785756501182035e-05,
|
| 493 |
+
"loss": 1.1681,
|
| 494 |
+
"step": 1650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.2275449101796405,
|
| 498 |
+
"grad_norm": 24.542713165283203,
|
| 499 |
+
"learning_rate": 1.9748817966903074e-05,
|
| 500 |
+
"loss": 1.3143,
|
| 501 |
+
"step": 1675
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.260811709913506,
|
| 505 |
+
"grad_norm": 19.09051513671875,
|
| 506 |
+
"learning_rate": 1.9711879432624113e-05,
|
| 507 |
+
"loss": 1.3256,
|
| 508 |
+
"step": 1700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.294078509647372,
|
| 512 |
+
"grad_norm": 17.27608299255371,
|
| 513 |
+
"learning_rate": 1.9674940898345156e-05,
|
| 514 |
+
"loss": 1.2525,
|
| 515 |
+
"step": 1725
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.3273453093812373,
|
| 519 |
+
"grad_norm": 26.802358627319336,
|
| 520 |
+
"learning_rate": 1.9638002364066195e-05,
|
| 521 |
+
"loss": 1.2224,
|
| 522 |
+
"step": 1750
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.3606121091151033,
|
| 526 |
+
"grad_norm": 25.98259162902832,
|
| 527 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 528 |
+
"loss": 1.3073,
|
| 529 |
+
"step": 1775
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.393878908848969,
|
| 533 |
+
"grad_norm": 60.687034606933594,
|
| 534 |
+
"learning_rate": 1.9564125295508277e-05,
|
| 535 |
+
"loss": 1.3168,
|
| 536 |
+
"step": 1800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.4271457085828345,
|
| 540 |
+
"grad_norm": 13.699719429016113,
|
| 541 |
+
"learning_rate": 1.9527186761229316e-05,
|
| 542 |
+
"loss": 1.349,
|
| 543 |
+
"step": 1825
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.4604125083167,
|
| 547 |
+
"grad_norm": 22.38040542602539,
|
| 548 |
+
"learning_rate": 1.9490248226950355e-05,
|
| 549 |
+
"loss": 1.2184,
|
| 550 |
+
"step": 1850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.4936793080505657,
|
| 554 |
+
"grad_norm": 28.865779876708984,
|
| 555 |
+
"learning_rate": 1.9453309692671394e-05,
|
| 556 |
+
"loss": 1.3009,
|
| 557 |
+
"step": 1875
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.5269461077844313,
|
| 561 |
+
"grad_norm": 27.044086456298828,
|
| 562 |
+
"learning_rate": 1.9416371158392437e-05,
|
| 563 |
+
"loss": 1.4822,
|
| 564 |
+
"step": 1900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.560212907518297,
|
| 568 |
+
"grad_norm": 27.432329177856445,
|
| 569 |
+
"learning_rate": 1.9379432624113476e-05,
|
| 570 |
+
"loss": 1.3222,
|
| 571 |
+
"step": 1925
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.5934797072521625,
|
| 575 |
+
"grad_norm": 23.156661987304688,
|
| 576 |
+
"learning_rate": 1.9342494089834515e-05,
|
| 577 |
+
"loss": 1.3166,
|
| 578 |
+
"step": 1950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.626746506986028,
|
| 582 |
+
"grad_norm": 31.14120864868164,
|
| 583 |
+
"learning_rate": 1.9305555555555558e-05,
|
| 584 |
+
"loss": 1.3456,
|
| 585 |
+
"step": 1975
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.6600133067198937,
|
| 589 |
+
"grad_norm": 38.80833435058594,
|
| 590 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 591 |
+
"loss": 1.4191,
|
| 592 |
+
"step": 2000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.6932801064537593,
|
| 596 |
+
"grad_norm": 50.219032287597656,
|
| 597 |
+
"learning_rate": 1.9231678486997636e-05,
|
| 598 |
+
"loss": 1.3517,
|
| 599 |
+
"step": 2025
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.726546906187625,
|
| 603 |
+
"grad_norm": 33.395103454589844,
|
| 604 |
+
"learning_rate": 1.919473995271868e-05,
|
| 605 |
+
"loss": 1.1377,
|
| 606 |
+
"step": 2050
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.7598137059214904,
|
| 610 |
+
"grad_norm": 32.816558837890625,
|
| 611 |
+
"learning_rate": 1.9157801418439718e-05,
|
| 612 |
+
"loss": 1.0702,
|
| 613 |
+
"step": 2075
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.793080505655356,
|
| 617 |
+
"grad_norm": 17.633621215820312,
|
| 618 |
+
"learning_rate": 1.9120862884160757e-05,
|
| 619 |
+
"loss": 1.401,
|
| 620 |
+
"step": 2100
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.8263473053892216,
|
| 624 |
+
"grad_norm": 17.697147369384766,
|
| 625 |
+
"learning_rate": 1.9083924349881797e-05,
|
| 626 |
+
"loss": 1.3274,
|
| 627 |
+
"step": 2125
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.8596141051230872,
|
| 631 |
+
"grad_norm": 29.65160369873047,
|
| 632 |
+
"learning_rate": 1.904698581560284e-05,
|
| 633 |
+
"loss": 1.4492,
|
| 634 |
+
"step": 2150
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.892880904856953,
|
| 638 |
+
"grad_norm": 21.899288177490234,
|
| 639 |
+
"learning_rate": 1.9010047281323878e-05,
|
| 640 |
+
"loss": 1.2815,
|
| 641 |
+
"step": 2175
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.9261477045908184,
|
| 645 |
+
"grad_norm": 30.655946731567383,
|
| 646 |
+
"learning_rate": 1.897310874704492e-05,
|
| 647 |
+
"loss": 1.1365,
|
| 648 |
+
"step": 2200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.959414504324684,
|
| 652 |
+
"grad_norm": 23.340801239013672,
|
| 653 |
+
"learning_rate": 1.893617021276596e-05,
|
| 654 |
+
"loss": 0.8965,
|
| 655 |
+
"step": 2225
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.9926813040585496,
|
| 659 |
+
"grad_norm": 33.06778335571289,
|
| 660 |
+
"learning_rate": 1.8899231678487e-05,
|
| 661 |
+
"loss": 1.6078,
|
| 662 |
+
"step": 2250
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.0,
|
| 666 |
+
"eval_accuracy": 0.8720059880239521,
|
| 667 |
+
"eval_loss": 0.3429495692253113,
|
| 668 |
+
"eval_macro_f1": 0.6385853572483872,
|
| 669 |
+
"eval_runtime": 1.5534,
|
| 670 |
+
"eval_samples_per_second": 1720.058,
|
| 671 |
+
"eval_steps_per_second": 107.504,
|
| 672 |
+
"eval_weighted_f1": 0.8599077627946274,
|
| 673 |
+
"step": 2256
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.0252827677977376,
|
| 677 |
+
"grad_norm": 21.779008865356445,
|
| 678 |
+
"learning_rate": 1.886229314420804e-05,
|
| 679 |
+
"loss": 1.017,
|
| 680 |
+
"step": 2275
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.0585495675316037,
|
| 684 |
+
"grad_norm": 34.79258346557617,
|
| 685 |
+
"learning_rate": 1.8825354609929078e-05,
|
| 686 |
+
"loss": 1.07,
|
| 687 |
+
"step": 2300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.0918163672654693,
|
| 691 |
+
"grad_norm": 27.5338134765625,
|
| 692 |
+
"learning_rate": 1.878841607565012e-05,
|
| 693 |
+
"loss": 0.8897,
|
| 694 |
+
"step": 2325
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.125083166999335,
|
| 698 |
+
"grad_norm": 35.044342041015625,
|
| 699 |
+
"learning_rate": 1.875147754137116e-05,
|
| 700 |
+
"loss": 0.9177,
|
| 701 |
+
"step": 2350
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.1583499667332005,
|
| 705 |
+
"grad_norm": 22.801095962524414,
|
| 706 |
+
"learning_rate": 1.8714539007092202e-05,
|
| 707 |
+
"loss": 1.1781,
|
| 708 |
+
"step": 2375
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.191616766467066,
|
| 712 |
+
"grad_norm": 10.759162902832031,
|
| 713 |
+
"learning_rate": 1.867760047281324e-05,
|
| 714 |
+
"loss": 1.2204,
|
| 715 |
+
"step": 2400
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.2248835662009316,
|
| 719 |
+
"grad_norm": 38.22099685668945,
|
| 720 |
+
"learning_rate": 1.864066193853428e-05,
|
| 721 |
+
"loss": 0.9673,
|
| 722 |
+
"step": 2425
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.2581503659347972,
|
| 726 |
+
"grad_norm": 29.876724243164062,
|
| 727 |
+
"learning_rate": 1.860372340425532e-05,
|
| 728 |
+
"loss": 1.1241,
|
| 729 |
+
"step": 2450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.291417165668663,
|
| 733 |
+
"grad_norm": 47.733001708984375,
|
| 734 |
+
"learning_rate": 1.856678486997636e-05,
|
| 735 |
+
"loss": 1.0585,
|
| 736 |
+
"step": 2475
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.3246839654025284,
|
| 740 |
+
"grad_norm": 46.068023681640625,
|
| 741 |
+
"learning_rate": 1.85298463356974e-05,
|
| 742 |
+
"loss": 0.8631,
|
| 743 |
+
"step": 2500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.357950765136394,
|
| 747 |
+
"grad_norm": 29.83353042602539,
|
| 748 |
+
"learning_rate": 1.849290780141844e-05,
|
| 749 |
+
"loss": 0.8431,
|
| 750 |
+
"step": 2525
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.3912175648702596,
|
| 754 |
+
"grad_norm": 29.349613189697266,
|
| 755 |
+
"learning_rate": 1.8455969267139483e-05,
|
| 756 |
+
"loss": 0.9994,
|
| 757 |
+
"step": 2550
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.424484364604125,
|
| 761 |
+
"grad_norm": 15.252015113830566,
|
| 762 |
+
"learning_rate": 1.8419030732860522e-05,
|
| 763 |
+
"loss": 1.1265,
|
| 764 |
+
"step": 2575
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.457751164337991,
|
| 768 |
+
"grad_norm": 13.561429023742676,
|
| 769 |
+
"learning_rate": 1.838209219858156e-05,
|
| 770 |
+
"loss": 1.0097,
|
| 771 |
+
"step": 2600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.4910179640718564,
|
| 775 |
+
"grad_norm": 32.128536224365234,
|
| 776 |
+
"learning_rate": 1.83451536643026e-05,
|
| 777 |
+
"loss": 0.9359,
|
| 778 |
+
"step": 2625
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.524284763805722,
|
| 782 |
+
"grad_norm": 21.812347412109375,
|
| 783 |
+
"learning_rate": 1.8308215130023643e-05,
|
| 784 |
+
"loss": 1.0516,
|
| 785 |
+
"step": 2650
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.5575515635395876,
|
| 789 |
+
"grad_norm": 13.613457679748535,
|
| 790 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 791 |
+
"loss": 1.072,
|
| 792 |
+
"step": 2675
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.590818363273453,
|
| 796 |
+
"grad_norm": 50.27625274658203,
|
| 797 |
+
"learning_rate": 1.823433806146572e-05,
|
| 798 |
+
"loss": 1.2114,
|
| 799 |
+
"step": 2700
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.6240851630073188,
|
| 803 |
+
"grad_norm": 20.310781478881836,
|
| 804 |
+
"learning_rate": 1.8197399527186764e-05,
|
| 805 |
+
"loss": 1.1355,
|
| 806 |
+
"step": 2725
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.6573519627411843,
|
| 810 |
+
"grad_norm": 27.120084762573242,
|
| 811 |
+
"learning_rate": 1.8160460992907803e-05,
|
| 812 |
+
"loss": 0.9515,
|
| 813 |
+
"step": 2750
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.69061876247505,
|
| 817 |
+
"grad_norm": 31.807022094726562,
|
| 818 |
+
"learning_rate": 1.8123522458628842e-05,
|
| 819 |
+
"loss": 1.0553,
|
| 820 |
+
"step": 2775
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.7238855622089155,
|
| 824 |
+
"grad_norm": 15.850422859191895,
|
| 825 |
+
"learning_rate": 1.808658392434988e-05,
|
| 826 |
+
"loss": 1.0379,
|
| 827 |
+
"step": 2800
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.757152361942781,
|
| 831 |
+
"grad_norm": 19.942224502563477,
|
| 832 |
+
"learning_rate": 1.8049645390070924e-05,
|
| 833 |
+
"loss": 1.0981,
|
| 834 |
+
"step": 2825
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.7904191616766467,
|
| 838 |
+
"grad_norm": 24.405765533447266,
|
| 839 |
+
"learning_rate": 1.8012706855791963e-05,
|
| 840 |
+
"loss": 1.0724,
|
| 841 |
+
"step": 2850
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.8236859614105123,
|
| 845 |
+
"grad_norm": 23.55417251586914,
|
| 846 |
+
"learning_rate": 1.7975768321513003e-05,
|
| 847 |
+
"loss": 1.0048,
|
| 848 |
+
"step": 2875
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.856952761144378,
|
| 852 |
+
"grad_norm": 24.6857852935791,
|
| 853 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 854 |
+
"loss": 1.0738,
|
| 855 |
+
"step": 2900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.8902195608782435,
|
| 859 |
+
"grad_norm": 26.437158584594727,
|
| 860 |
+
"learning_rate": 1.7901891252955084e-05,
|
| 861 |
+
"loss": 1.0814,
|
| 862 |
+
"step": 2925
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.923486360612109,
|
| 866 |
+
"grad_norm": 30.49480438232422,
|
| 867 |
+
"learning_rate": 1.7864952718676124e-05,
|
| 868 |
+
"loss": 0.951,
|
| 869 |
+
"step": 2950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.9567531603459747,
|
| 873 |
+
"grad_norm": 35.565547943115234,
|
| 874 |
+
"learning_rate": 1.7828014184397163e-05,
|
| 875 |
+
"loss": 1.2132,
|
| 876 |
+
"step": 2975
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.9900199600798403,
|
| 880 |
+
"grad_norm": 29.539398193359375,
|
| 881 |
+
"learning_rate": 1.7791075650118205e-05,
|
| 882 |
+
"loss": 1.0817,
|
| 883 |
+
"step": 3000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 4.0,
|
| 887 |
+
"eval_accuracy": 0.873877245508982,
|
| 888 |
+
"eval_loss": 0.3812912106513977,
|
| 889 |
+
"eval_macro_f1": 0.6974282738120429,
|
| 890 |
+
"eval_runtime": 1.5114,
|
| 891 |
+
"eval_samples_per_second": 1767.947,
|
| 892 |
+
"eval_steps_per_second": 110.497,
|
| 893 |
+
"eval_weighted_f1": 0.8741648663273964,
|
| 894 |
+
"step": 3008
|
| 895 |
+
}
|
| 896 |
+
],
|
| 897 |
+
"logging_steps": 25,
|
| 898 |
+
"max_steps": 15040,
|
| 899 |
+
"num_input_tokens_seen": 0,
|
| 900 |
+
"num_train_epochs": 20,
|
| 901 |
+
"save_steps": 500,
|
| 902 |
+
"stateful_callbacks": {
|
| 903 |
+
"EarlyStoppingCallback": {
|
| 904 |
+
"args": {
|
| 905 |
+
"early_stopping_patience": 3,
|
| 906 |
+
"early_stopping_threshold": 0.0
|
| 907 |
+
},
|
| 908 |
+
"attributes": {
|
| 909 |
+
"early_stopping_patience_counter": 0
|
| 910 |
+
}
|
| 911 |
+
},
|
| 912 |
+
"TrainerControl": {
|
| 913 |
+
"args": {
|
| 914 |
+
"should_epoch_stop": false,
|
| 915 |
+
"should_evaluate": false,
|
| 916 |
+
"should_log": false,
|
| 917 |
+
"should_save": true,
|
| 918 |
+
"should_training_stop": false
|
| 919 |
+
},
|
| 920 |
+
"attributes": {}
|
| 921 |
+
}
|
| 922 |
+
},
|
| 923 |
+
"total_flos": 0.0,
|
| 924 |
+
"train_batch_size": 8,
|
| 925 |
+
"trial_name": null,
|
| 926 |
+
"trial_params": null
|
| 927 |
+
}
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"20480": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 2048,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "XLMRobertaTokenizerFast",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-5264/trainer_state.json
ADDED
|
@@ -0,0 +1,1590 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3008,
|
| 3 |
+
"best_metric": 0.6974282738120429,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/vihsd-bamibert-s44-20260827T130014Z-78fe7de2/trainer/checkpoint-3008",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5264,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0332667997338656,
|
| 14 |
+
"grad_norm": 34.15004348754883,
|
| 15 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 16 |
+
"loss": 4.3193,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0665335994677312,
|
| 21 |
+
"grad_norm": 35.99433517456055,
|
| 22 |
+
"learning_rate": 6.515957446808512e-07,
|
| 23 |
+
"loss": 4.0656,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0998003992015968,
|
| 28 |
+
"grad_norm": 64.98133087158203,
|
| 29 |
+
"learning_rate": 9.840425531914895e-07,
|
| 30 |
+
"loss": 3.5941,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1330671989354624,
|
| 35 |
+
"grad_norm": 31.218713760375977,
|
| 36 |
+
"learning_rate": 1.3164893617021277e-06,
|
| 37 |
+
"loss": 2.9458,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16633399866932802,
|
| 42 |
+
"grad_norm": 27.85391616821289,
|
| 43 |
+
"learning_rate": 1.648936170212766e-06,
|
| 44 |
+
"loss": 2.3829,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1996007984031936,
|
| 49 |
+
"grad_norm": 31.22490692138672,
|
| 50 |
+
"learning_rate": 1.9813829787234043e-06,
|
| 51 |
+
"loss": 2.3445,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2328675981370592,
|
| 56 |
+
"grad_norm": 21.99728012084961,
|
| 57 |
+
"learning_rate": 2.3138297872340428e-06,
|
| 58 |
+
"loss": 2.0081,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2661343978709248,
|
| 63 |
+
"grad_norm": 21.359256744384766,
|
| 64 |
+
"learning_rate": 2.646276595744681e-06,
|
| 65 |
+
"loss": 2.1865,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.2994011976047904,
|
| 70 |
+
"grad_norm": 22.947093963623047,
|
| 71 |
+
"learning_rate": 2.978723404255319e-06,
|
| 72 |
+
"loss": 2.1144,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.33266799733865604,
|
| 77 |
+
"grad_norm": 15.964245796203613,
|
| 78 |
+
"learning_rate": 3.311170212765958e-06,
|
| 79 |
+
"loss": 2.2028,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.36593479707252163,
|
| 84 |
+
"grad_norm": 15.935502052307129,
|
| 85 |
+
"learning_rate": 3.643617021276596e-06,
|
| 86 |
+
"loss": 2.0956,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.3992015968063872,
|
| 91 |
+
"grad_norm": 28.799392700195312,
|
| 92 |
+
"learning_rate": 3.976063829787235e-06,
|
| 93 |
+
"loss": 2.2256,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4324683965402528,
|
| 98 |
+
"grad_norm": 21.484956741333008,
|
| 99 |
+
"learning_rate": 4.308510638297873e-06,
|
| 100 |
+
"loss": 2.1551,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4657351962741184,
|
| 105 |
+
"grad_norm": 18.2623348236084,
|
| 106 |
+
"learning_rate": 4.640957446808511e-06,
|
| 107 |
+
"loss": 2.063,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.499001996007984,
|
| 112 |
+
"grad_norm": 28.219589233398438,
|
| 113 |
+
"learning_rate": 4.973404255319149e-06,
|
| 114 |
+
"loss": 1.9974,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5322687957418496,
|
| 119 |
+
"grad_norm": 51.115116119384766,
|
| 120 |
+
"learning_rate": 5.305851063829788e-06,
|
| 121 |
+
"loss": 2.0349,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5655355954757152,
|
| 126 |
+
"grad_norm": 34.70194625854492,
|
| 127 |
+
"learning_rate": 5.638297872340426e-06,
|
| 128 |
+
"loss": 1.9449,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5988023952095808,
|
| 133 |
+
"grad_norm": 27.660011291503906,
|
| 134 |
+
"learning_rate": 5.970744680851064e-06,
|
| 135 |
+
"loss": 1.9879,
|
| 136 |
+
"step": 450
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6320691949434465,
|
| 140 |
+
"grad_norm": 52.268550872802734,
|
| 141 |
+
"learning_rate": 6.303191489361703e-06,
|
| 142 |
+
"loss": 1.9235,
|
| 143 |
+
"step": 475
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6653359946773121,
|
| 147 |
+
"grad_norm": 22.55914878845215,
|
| 148 |
+
"learning_rate": 6.635638297872341e-06,
|
| 149 |
+
"loss": 1.9964,
|
| 150 |
+
"step": 500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.6986027944111777,
|
| 154 |
+
"grad_norm": 29.36838722229004,
|
| 155 |
+
"learning_rate": 6.968085106382979e-06,
|
| 156 |
+
"loss": 1.7544,
|
| 157 |
+
"step": 525
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7318695941450433,
|
| 161 |
+
"grad_norm": 29.081684112548828,
|
| 162 |
+
"learning_rate": 7.300531914893617e-06,
|
| 163 |
+
"loss": 1.5827,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7651363938789089,
|
| 168 |
+
"grad_norm": 39.61648178100586,
|
| 169 |
+
"learning_rate": 7.632978723404256e-06,
|
| 170 |
+
"loss": 1.7948,
|
| 171 |
+
"step": 575
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.7984031936127745,
|
| 175 |
+
"grad_norm": 31.700763702392578,
|
| 176 |
+
"learning_rate": 7.965425531914895e-06,
|
| 177 |
+
"loss": 1.5821,
|
| 178 |
+
"step": 600
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.83166999334664,
|
| 182 |
+
"grad_norm": 20.465805053710938,
|
| 183 |
+
"learning_rate": 8.297872340425532e-06,
|
| 184 |
+
"loss": 1.5987,
|
| 185 |
+
"step": 625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8649367930805056,
|
| 189 |
+
"grad_norm": 31.21770668029785,
|
| 190 |
+
"learning_rate": 8.630319148936171e-06,
|
| 191 |
+
"loss": 1.6804,
|
| 192 |
+
"step": 650
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.8982035928143712,
|
| 196 |
+
"grad_norm": 28.831789016723633,
|
| 197 |
+
"learning_rate": 8.96276595744681e-06,
|
| 198 |
+
"loss": 1.7875,
|
| 199 |
+
"step": 675
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9314703925482368,
|
| 203 |
+
"grad_norm": 24.137300491333008,
|
| 204 |
+
"learning_rate": 9.295212765957447e-06,
|
| 205 |
+
"loss": 1.883,
|
| 206 |
+
"step": 700
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9647371922821024,
|
| 210 |
+
"grad_norm": 35.83584976196289,
|
| 211 |
+
"learning_rate": 9.627659574468086e-06,
|
| 212 |
+
"loss": 1.729,
|
| 213 |
+
"step": 725
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.998003992015968,
|
| 217 |
+
"grad_norm": 26.05014419555664,
|
| 218 |
+
"learning_rate": 9.960106382978723e-06,
|
| 219 |
+
"loss": 1.6761,
|
| 220 |
+
"step": 750
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0,
|
| 224 |
+
"eval_accuracy": 0.842814371257485,
|
| 225 |
+
"eval_loss": 0.4147079288959503,
|
| 226 |
+
"eval_macro_f1": 0.5715291591376829,
|
| 227 |
+
"eval_runtime": 1.5487,
|
| 228 |
+
"eval_samples_per_second": 1725.328,
|
| 229 |
+
"eval_steps_per_second": 107.833,
|
| 230 |
+
"eval_weighted_f1": 0.8313170510138047,
|
| 231 |
+
"step": 752
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.0306054557551563,
|
| 235 |
+
"grad_norm": 35.20064163208008,
|
| 236 |
+
"learning_rate": 1.0292553191489364e-05,
|
| 237 |
+
"loss": 1.6047,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.0638722554890219,
|
| 242 |
+
"grad_norm": 18.61759376525879,
|
| 243 |
+
"learning_rate": 1.0625e-05,
|
| 244 |
+
"loss": 1.4676,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.0971390552228875,
|
| 249 |
+
"grad_norm": 30.29521369934082,
|
| 250 |
+
"learning_rate": 1.095744680851064e-05,
|
| 251 |
+
"loss": 1.5534,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.130405854956753,
|
| 256 |
+
"grad_norm": 49.2490234375,
|
| 257 |
+
"learning_rate": 1.1289893617021277e-05,
|
| 258 |
+
"loss": 1.6714,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.1636726546906186,
|
| 263 |
+
"grad_norm": 21.129003524780273,
|
| 264 |
+
"learning_rate": 1.1622340425531916e-05,
|
| 265 |
+
"loss": 1.7271,
|
| 266 |
+
"step": 875
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.1969394544244845,
|
| 270 |
+
"grad_norm": 50.25038528442383,
|
| 271 |
+
"learning_rate": 1.1954787234042554e-05,
|
| 272 |
+
"loss": 1.5317,
|
| 273 |
+
"step": 900
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.23020625415835,
|
| 277 |
+
"grad_norm": 37.23767852783203,
|
| 278 |
+
"learning_rate": 1.2287234042553193e-05,
|
| 279 |
+
"loss": 1.6333,
|
| 280 |
+
"step": 925
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.2634730538922156,
|
| 284 |
+
"grad_norm": 17.01172637939453,
|
| 285 |
+
"learning_rate": 1.2619680851063832e-05,
|
| 286 |
+
"loss": 1.5422,
|
| 287 |
+
"step": 950
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.2967398536260812,
|
| 291 |
+
"grad_norm": 21.008949279785156,
|
| 292 |
+
"learning_rate": 1.295212765957447e-05,
|
| 293 |
+
"loss": 1.5974,
|
| 294 |
+
"step": 975
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.3300066533599468,
|
| 298 |
+
"grad_norm": 28.253963470458984,
|
| 299 |
+
"learning_rate": 1.3284574468085108e-05,
|
| 300 |
+
"loss": 1.4676,
|
| 301 |
+
"step": 1000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.3632734530938124,
|
| 305 |
+
"grad_norm": 27.492841720581055,
|
| 306 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 307 |
+
"loss": 1.5826,
|
| 308 |
+
"step": 1025
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.396540252827678,
|
| 312 |
+
"grad_norm": 20.68940544128418,
|
| 313 |
+
"learning_rate": 1.3949468085106384e-05,
|
| 314 |
+
"loss": 1.5191,
|
| 315 |
+
"step": 1050
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.4298070525615436,
|
| 319 |
+
"grad_norm": 34.80696105957031,
|
| 320 |
+
"learning_rate": 1.4281914893617021e-05,
|
| 321 |
+
"loss": 1.5666,
|
| 322 |
+
"step": 1075
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.4630738522954092,
|
| 326 |
+
"grad_norm": 43.78744888305664,
|
| 327 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 328 |
+
"loss": 1.4861,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.4963406520292748,
|
| 333 |
+
"grad_norm": 29.718017578125,
|
| 334 |
+
"learning_rate": 1.49468085106383e-05,
|
| 335 |
+
"loss": 1.8153,
|
| 336 |
+
"step": 1125
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.5296074517631404,
|
| 340 |
+
"grad_norm": 25.93970489501953,
|
| 341 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 342 |
+
"loss": 1.6267,
|
| 343 |
+
"step": 1150
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.562874251497006,
|
| 347 |
+
"grad_norm": 21.3642578125,
|
| 348 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 349 |
+
"loss": 1.5005,
|
| 350 |
+
"step": 1175
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5961410512308716,
|
| 354 |
+
"grad_norm": 32.647422790527344,
|
| 355 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 356 |
+
"loss": 1.631,
|
| 357 |
+
"step": 1200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6294078509647372,
|
| 361 |
+
"grad_norm": 13.734174728393555,
|
| 362 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 363 |
+
"loss": 1.4731,
|
| 364 |
+
"step": 1225
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.6626746506986028,
|
| 368 |
+
"grad_norm": 28.250856399536133,
|
| 369 |
+
"learning_rate": 1.660904255319149e-05,
|
| 370 |
+
"loss": 1.5175,
|
| 371 |
+
"step": 1250
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.6959414504324684,
|
| 375 |
+
"grad_norm": 22.97903823852539,
|
| 376 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 377 |
+
"loss": 1.5888,
|
| 378 |
+
"step": 1275
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.729208250166334,
|
| 382 |
+
"grad_norm": 22.442594528198242,
|
| 383 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 384 |
+
"loss": 1.3336,
|
| 385 |
+
"step": 1300
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.7624750499001998,
|
| 389 |
+
"grad_norm": 34.93794250488281,
|
| 390 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 391 |
+
"loss": 1.5893,
|
| 392 |
+
"step": 1325
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.7957418496340654,
|
| 396 |
+
"grad_norm": 29.638469696044922,
|
| 397 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 398 |
+
"loss": 1.3456,
|
| 399 |
+
"step": 1350
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.829008649367931,
|
| 403 |
+
"grad_norm": 23.77605438232422,
|
| 404 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 405 |
+
"loss": 1.5958,
|
| 406 |
+
"step": 1375
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.8622754491017965,
|
| 410 |
+
"grad_norm": 34.77972412109375,
|
| 411 |
+
"learning_rate": 1.860372340425532e-05,
|
| 412 |
+
"loss": 1.5545,
|
| 413 |
+
"step": 1400
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.8955422488356621,
|
| 417 |
+
"grad_norm": 18.019433975219727,
|
| 418 |
+
"learning_rate": 1.893617021276596e-05,
|
| 419 |
+
"loss": 1.3322,
|
| 420 |
+
"step": 1425
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.9288090485695277,
|
| 424 |
+
"grad_norm": 25.30364227294922,
|
| 425 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 426 |
+
"loss": 1.625,
|
| 427 |
+
"step": 1450
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.9620758483033933,
|
| 431 |
+
"grad_norm": 18.51852035522461,
|
| 432 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 433 |
+
"loss": 1.4318,
|
| 434 |
+
"step": 1475
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.995342648037259,
|
| 438 |
+
"grad_norm": 26.217803955078125,
|
| 439 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 440 |
+
"loss": 1.4811,
|
| 441 |
+
"step": 1500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 2.0,
|
| 445 |
+
"eval_accuracy": 0.8619011976047904,
|
| 446 |
+
"eval_loss": 0.3896377980709076,
|
| 447 |
+
"eval_macro_f1": 0.5782126976453429,
|
| 448 |
+
"eval_runtime": 1.5856,
|
| 449 |
+
"eval_samples_per_second": 1685.217,
|
| 450 |
+
"eval_steps_per_second": 105.326,
|
| 451 |
+
"eval_weighted_f1": 0.8409251664533528,
|
| 452 |
+
"step": 1504
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.027944111776447,
|
| 456 |
+
"grad_norm": 56.660160064697266,
|
| 457 |
+
"learning_rate": 1.9970449172576832e-05,
|
| 458 |
+
"loss": 1.4506,
|
| 459 |
+
"step": 1525
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0612109115103125,
|
| 463 |
+
"grad_norm": 37.306522369384766,
|
| 464 |
+
"learning_rate": 1.9933510638297875e-05,
|
| 465 |
+
"loss": 1.4542,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.094477711244178,
|
| 470 |
+
"grad_norm": 22.323488235473633,
|
| 471 |
+
"learning_rate": 1.9896572104018914e-05,
|
| 472 |
+
"loss": 1.4015,
|
| 473 |
+
"step": 1575
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.1277445109780437,
|
| 477 |
+
"grad_norm": 39.91142654418945,
|
| 478 |
+
"learning_rate": 1.9859633569739953e-05,
|
| 479 |
+
"loss": 1.3846,
|
| 480 |
+
"step": 1600
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.1610113107119093,
|
| 484 |
+
"grad_norm": 26.0784969329834,
|
| 485 |
+
"learning_rate": 1.9822695035460996e-05,
|
| 486 |
+
"loss": 1.3649,
|
| 487 |
+
"step": 1625
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.194278110445775,
|
| 491 |
+
"grad_norm": 39.20205307006836,
|
| 492 |
+
"learning_rate": 1.9785756501182035e-05,
|
| 493 |
+
"loss": 1.1681,
|
| 494 |
+
"step": 1650
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.2275449101796405,
|
| 498 |
+
"grad_norm": 24.542713165283203,
|
| 499 |
+
"learning_rate": 1.9748817966903074e-05,
|
| 500 |
+
"loss": 1.3143,
|
| 501 |
+
"step": 1675
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.260811709913506,
|
| 505 |
+
"grad_norm": 19.09051513671875,
|
| 506 |
+
"learning_rate": 1.9711879432624113e-05,
|
| 507 |
+
"loss": 1.3256,
|
| 508 |
+
"step": 1700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.294078509647372,
|
| 512 |
+
"grad_norm": 17.27608299255371,
|
| 513 |
+
"learning_rate": 1.9674940898345156e-05,
|
| 514 |
+
"loss": 1.2525,
|
| 515 |
+
"step": 1725
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.3273453093812373,
|
| 519 |
+
"grad_norm": 26.802358627319336,
|
| 520 |
+
"learning_rate": 1.9638002364066195e-05,
|
| 521 |
+
"loss": 1.2224,
|
| 522 |
+
"step": 1750
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.3606121091151033,
|
| 526 |
+
"grad_norm": 25.98259162902832,
|
| 527 |
+
"learning_rate": 1.9601063829787234e-05,
|
| 528 |
+
"loss": 1.3073,
|
| 529 |
+
"step": 1775
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.393878908848969,
|
| 533 |
+
"grad_norm": 60.687034606933594,
|
| 534 |
+
"learning_rate": 1.9564125295508277e-05,
|
| 535 |
+
"loss": 1.3168,
|
| 536 |
+
"step": 1800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.4271457085828345,
|
| 540 |
+
"grad_norm": 13.699719429016113,
|
| 541 |
+
"learning_rate": 1.9527186761229316e-05,
|
| 542 |
+
"loss": 1.349,
|
| 543 |
+
"step": 1825
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.4604125083167,
|
| 547 |
+
"grad_norm": 22.38040542602539,
|
| 548 |
+
"learning_rate": 1.9490248226950355e-05,
|
| 549 |
+
"loss": 1.2184,
|
| 550 |
+
"step": 1850
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.4936793080505657,
|
| 554 |
+
"grad_norm": 28.865779876708984,
|
| 555 |
+
"learning_rate": 1.9453309692671394e-05,
|
| 556 |
+
"loss": 1.3009,
|
| 557 |
+
"step": 1875
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.5269461077844313,
|
| 561 |
+
"grad_norm": 27.044086456298828,
|
| 562 |
+
"learning_rate": 1.9416371158392437e-05,
|
| 563 |
+
"loss": 1.4822,
|
| 564 |
+
"step": 1900
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.560212907518297,
|
| 568 |
+
"grad_norm": 27.432329177856445,
|
| 569 |
+
"learning_rate": 1.9379432624113476e-05,
|
| 570 |
+
"loss": 1.3222,
|
| 571 |
+
"step": 1925
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.5934797072521625,
|
| 575 |
+
"grad_norm": 23.156661987304688,
|
| 576 |
+
"learning_rate": 1.9342494089834515e-05,
|
| 577 |
+
"loss": 1.3166,
|
| 578 |
+
"step": 1950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.626746506986028,
|
| 582 |
+
"grad_norm": 31.14120864868164,
|
| 583 |
+
"learning_rate": 1.9305555555555558e-05,
|
| 584 |
+
"loss": 1.3456,
|
| 585 |
+
"step": 1975
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.6600133067198937,
|
| 589 |
+
"grad_norm": 38.80833435058594,
|
| 590 |
+
"learning_rate": 1.9268617021276597e-05,
|
| 591 |
+
"loss": 1.4191,
|
| 592 |
+
"step": 2000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.6932801064537593,
|
| 596 |
+
"grad_norm": 50.219032287597656,
|
| 597 |
+
"learning_rate": 1.9231678486997636e-05,
|
| 598 |
+
"loss": 1.3517,
|
| 599 |
+
"step": 2025
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.726546906187625,
|
| 603 |
+
"grad_norm": 33.395103454589844,
|
| 604 |
+
"learning_rate": 1.919473995271868e-05,
|
| 605 |
+
"loss": 1.1377,
|
| 606 |
+
"step": 2050
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.7598137059214904,
|
| 610 |
+
"grad_norm": 32.816558837890625,
|
| 611 |
+
"learning_rate": 1.9157801418439718e-05,
|
| 612 |
+
"loss": 1.0702,
|
| 613 |
+
"step": 2075
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.793080505655356,
|
| 617 |
+
"grad_norm": 17.633621215820312,
|
| 618 |
+
"learning_rate": 1.9120862884160757e-05,
|
| 619 |
+
"loss": 1.401,
|
| 620 |
+
"step": 2100
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.8263473053892216,
|
| 624 |
+
"grad_norm": 17.697147369384766,
|
| 625 |
+
"learning_rate": 1.9083924349881797e-05,
|
| 626 |
+
"loss": 1.3274,
|
| 627 |
+
"step": 2125
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.8596141051230872,
|
| 631 |
+
"grad_norm": 29.65160369873047,
|
| 632 |
+
"learning_rate": 1.904698581560284e-05,
|
| 633 |
+
"loss": 1.4492,
|
| 634 |
+
"step": 2150
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.892880904856953,
|
| 638 |
+
"grad_norm": 21.899288177490234,
|
| 639 |
+
"learning_rate": 1.9010047281323878e-05,
|
| 640 |
+
"loss": 1.2815,
|
| 641 |
+
"step": 2175
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.9261477045908184,
|
| 645 |
+
"grad_norm": 30.655946731567383,
|
| 646 |
+
"learning_rate": 1.897310874704492e-05,
|
| 647 |
+
"loss": 1.1365,
|
| 648 |
+
"step": 2200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.959414504324684,
|
| 652 |
+
"grad_norm": 23.340801239013672,
|
| 653 |
+
"learning_rate": 1.893617021276596e-05,
|
| 654 |
+
"loss": 0.8965,
|
| 655 |
+
"step": 2225
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.9926813040585496,
|
| 659 |
+
"grad_norm": 33.06778335571289,
|
| 660 |
+
"learning_rate": 1.8899231678487e-05,
|
| 661 |
+
"loss": 1.6078,
|
| 662 |
+
"step": 2250
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.0,
|
| 666 |
+
"eval_accuracy": 0.8720059880239521,
|
| 667 |
+
"eval_loss": 0.3429495692253113,
|
| 668 |
+
"eval_macro_f1": 0.6385853572483872,
|
| 669 |
+
"eval_runtime": 1.5534,
|
| 670 |
+
"eval_samples_per_second": 1720.058,
|
| 671 |
+
"eval_steps_per_second": 107.504,
|
| 672 |
+
"eval_weighted_f1": 0.8599077627946274,
|
| 673 |
+
"step": 2256
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.0252827677977376,
|
| 677 |
+
"grad_norm": 21.779008865356445,
|
| 678 |
+
"learning_rate": 1.886229314420804e-05,
|
| 679 |
+
"loss": 1.017,
|
| 680 |
+
"step": 2275
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.0585495675316037,
|
| 684 |
+
"grad_norm": 34.79258346557617,
|
| 685 |
+
"learning_rate": 1.8825354609929078e-05,
|
| 686 |
+
"loss": 1.07,
|
| 687 |
+
"step": 2300
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.0918163672654693,
|
| 691 |
+
"grad_norm": 27.5338134765625,
|
| 692 |
+
"learning_rate": 1.878841607565012e-05,
|
| 693 |
+
"loss": 0.8897,
|
| 694 |
+
"step": 2325
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.125083166999335,
|
| 698 |
+
"grad_norm": 35.044342041015625,
|
| 699 |
+
"learning_rate": 1.875147754137116e-05,
|
| 700 |
+
"loss": 0.9177,
|
| 701 |
+
"step": 2350
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.1583499667332005,
|
| 705 |
+
"grad_norm": 22.801095962524414,
|
| 706 |
+
"learning_rate": 1.8714539007092202e-05,
|
| 707 |
+
"loss": 1.1781,
|
| 708 |
+
"step": 2375
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.191616766467066,
|
| 712 |
+
"grad_norm": 10.759162902832031,
|
| 713 |
+
"learning_rate": 1.867760047281324e-05,
|
| 714 |
+
"loss": 1.2204,
|
| 715 |
+
"step": 2400
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.2248835662009316,
|
| 719 |
+
"grad_norm": 38.22099685668945,
|
| 720 |
+
"learning_rate": 1.864066193853428e-05,
|
| 721 |
+
"loss": 0.9673,
|
| 722 |
+
"step": 2425
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.2581503659347972,
|
| 726 |
+
"grad_norm": 29.876724243164062,
|
| 727 |
+
"learning_rate": 1.860372340425532e-05,
|
| 728 |
+
"loss": 1.1241,
|
| 729 |
+
"step": 2450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.291417165668663,
|
| 733 |
+
"grad_norm": 47.733001708984375,
|
| 734 |
+
"learning_rate": 1.856678486997636e-05,
|
| 735 |
+
"loss": 1.0585,
|
| 736 |
+
"step": 2475
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.3246839654025284,
|
| 740 |
+
"grad_norm": 46.068023681640625,
|
| 741 |
+
"learning_rate": 1.85298463356974e-05,
|
| 742 |
+
"loss": 0.8631,
|
| 743 |
+
"step": 2500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.357950765136394,
|
| 747 |
+
"grad_norm": 29.83353042602539,
|
| 748 |
+
"learning_rate": 1.849290780141844e-05,
|
| 749 |
+
"loss": 0.8431,
|
| 750 |
+
"step": 2525
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.3912175648702596,
|
| 754 |
+
"grad_norm": 29.349613189697266,
|
| 755 |
+
"learning_rate": 1.8455969267139483e-05,
|
| 756 |
+
"loss": 0.9994,
|
| 757 |
+
"step": 2550
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.424484364604125,
|
| 761 |
+
"grad_norm": 15.252015113830566,
|
| 762 |
+
"learning_rate": 1.8419030732860522e-05,
|
| 763 |
+
"loss": 1.1265,
|
| 764 |
+
"step": 2575
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.457751164337991,
|
| 768 |
+
"grad_norm": 13.561429023742676,
|
| 769 |
+
"learning_rate": 1.838209219858156e-05,
|
| 770 |
+
"loss": 1.0097,
|
| 771 |
+
"step": 2600
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.4910179640718564,
|
| 775 |
+
"grad_norm": 32.128536224365234,
|
| 776 |
+
"learning_rate": 1.83451536643026e-05,
|
| 777 |
+
"loss": 0.9359,
|
| 778 |
+
"step": 2625
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.524284763805722,
|
| 782 |
+
"grad_norm": 21.812347412109375,
|
| 783 |
+
"learning_rate": 1.8308215130023643e-05,
|
| 784 |
+
"loss": 1.0516,
|
| 785 |
+
"step": 2650
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.5575515635395876,
|
| 789 |
+
"grad_norm": 13.613457679748535,
|
| 790 |
+
"learning_rate": 1.8271276595744682e-05,
|
| 791 |
+
"loss": 1.072,
|
| 792 |
+
"step": 2675
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.590818363273453,
|
| 796 |
+
"grad_norm": 50.27625274658203,
|
| 797 |
+
"learning_rate": 1.823433806146572e-05,
|
| 798 |
+
"loss": 1.2114,
|
| 799 |
+
"step": 2700
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.6240851630073188,
|
| 803 |
+
"grad_norm": 20.310781478881836,
|
| 804 |
+
"learning_rate": 1.8197399527186764e-05,
|
| 805 |
+
"loss": 1.1355,
|
| 806 |
+
"step": 2725
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.6573519627411843,
|
| 810 |
+
"grad_norm": 27.120084762573242,
|
| 811 |
+
"learning_rate": 1.8160460992907803e-05,
|
| 812 |
+
"loss": 0.9515,
|
| 813 |
+
"step": 2750
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.69061876247505,
|
| 817 |
+
"grad_norm": 31.807022094726562,
|
| 818 |
+
"learning_rate": 1.8123522458628842e-05,
|
| 819 |
+
"loss": 1.0553,
|
| 820 |
+
"step": 2775
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.7238855622089155,
|
| 824 |
+
"grad_norm": 15.850422859191895,
|
| 825 |
+
"learning_rate": 1.808658392434988e-05,
|
| 826 |
+
"loss": 1.0379,
|
| 827 |
+
"step": 2800
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.757152361942781,
|
| 831 |
+
"grad_norm": 19.942224502563477,
|
| 832 |
+
"learning_rate": 1.8049645390070924e-05,
|
| 833 |
+
"loss": 1.0981,
|
| 834 |
+
"step": 2825
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.7904191616766467,
|
| 838 |
+
"grad_norm": 24.405765533447266,
|
| 839 |
+
"learning_rate": 1.8012706855791963e-05,
|
| 840 |
+
"loss": 1.0724,
|
| 841 |
+
"step": 2850
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.8236859614105123,
|
| 845 |
+
"grad_norm": 23.55417251586914,
|
| 846 |
+
"learning_rate": 1.7975768321513003e-05,
|
| 847 |
+
"loss": 1.0048,
|
| 848 |
+
"step": 2875
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.856952761144378,
|
| 852 |
+
"grad_norm": 24.6857852935791,
|
| 853 |
+
"learning_rate": 1.7938829787234045e-05,
|
| 854 |
+
"loss": 1.0738,
|
| 855 |
+
"step": 2900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.8902195608782435,
|
| 859 |
+
"grad_norm": 26.437158584594727,
|
| 860 |
+
"learning_rate": 1.7901891252955084e-05,
|
| 861 |
+
"loss": 1.0814,
|
| 862 |
+
"step": 2925
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.923486360612109,
|
| 866 |
+
"grad_norm": 30.49480438232422,
|
| 867 |
+
"learning_rate": 1.7864952718676124e-05,
|
| 868 |
+
"loss": 0.951,
|
| 869 |
+
"step": 2950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.9567531603459747,
|
| 873 |
+
"grad_norm": 35.565547943115234,
|
| 874 |
+
"learning_rate": 1.7828014184397163e-05,
|
| 875 |
+
"loss": 1.2132,
|
| 876 |
+
"step": 2975
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.9900199600798403,
|
| 880 |
+
"grad_norm": 29.539398193359375,
|
| 881 |
+
"learning_rate": 1.7791075650118205e-05,
|
| 882 |
+
"loss": 1.0817,
|
| 883 |
+
"step": 3000
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 4.0,
|
| 887 |
+
"eval_accuracy": 0.873877245508982,
|
| 888 |
+
"eval_loss": 0.3812912106513977,
|
| 889 |
+
"eval_macro_f1": 0.6974282738120429,
|
| 890 |
+
"eval_runtime": 1.5114,
|
| 891 |
+
"eval_samples_per_second": 1767.947,
|
| 892 |
+
"eval_steps_per_second": 110.497,
|
| 893 |
+
"eval_weighted_f1": 0.8741648663273964,
|
| 894 |
+
"step": 3008
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.022621423819029,
|
| 898 |
+
"grad_norm": 33.87923812866211,
|
| 899 |
+
"learning_rate": 1.7754137115839245e-05,
|
| 900 |
+
"loss": 0.8396,
|
| 901 |
+
"step": 3025
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 4.055888223552894,
|
| 905 |
+
"grad_norm": 52.33735656738281,
|
| 906 |
+
"learning_rate": 1.7717198581560284e-05,
|
| 907 |
+
"loss": 0.7258,
|
| 908 |
+
"step": 3050
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 4.08915502328676,
|
| 912 |
+
"grad_norm": 15.482519149780273,
|
| 913 |
+
"learning_rate": 1.7680260047281326e-05,
|
| 914 |
+
"loss": 0.822,
|
| 915 |
+
"step": 3075
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.122421823020625,
|
| 919 |
+
"grad_norm": 33.11475372314453,
|
| 920 |
+
"learning_rate": 1.7643321513002365e-05,
|
| 921 |
+
"loss": 0.832,
|
| 922 |
+
"step": 3100
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 4.155688622754491,
|
| 926 |
+
"grad_norm": 15.799664497375488,
|
| 927 |
+
"learning_rate": 1.7606382978723408e-05,
|
| 928 |
+
"loss": 0.89,
|
| 929 |
+
"step": 3125
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 4.188955422488356,
|
| 933 |
+
"grad_norm": 58.16423034667969,
|
| 934 |
+
"learning_rate": 1.7569444444444447e-05,
|
| 935 |
+
"loss": 0.7332,
|
| 936 |
+
"step": 3150
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 4.222222222222222,
|
| 940 |
+
"grad_norm": 28.870731353759766,
|
| 941 |
+
"learning_rate": 1.7532505910165486e-05,
|
| 942 |
+
"loss": 0.8978,
|
| 943 |
+
"step": 3175
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 4.2554890219560875,
|
| 947 |
+
"grad_norm": 39.65562057495117,
|
| 948 |
+
"learning_rate": 1.7495567375886526e-05,
|
| 949 |
+
"loss": 1.0445,
|
| 950 |
+
"step": 3200
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 4.2887558216899535,
|
| 954 |
+
"grad_norm": 23.383996963500977,
|
| 955 |
+
"learning_rate": 1.7458628841607565e-05,
|
| 956 |
+
"loss": 0.6852,
|
| 957 |
+
"step": 3225
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 4.322022621423819,
|
| 961 |
+
"grad_norm": 39.20933532714844,
|
| 962 |
+
"learning_rate": 1.7421690307328607e-05,
|
| 963 |
+
"loss": 0.9351,
|
| 964 |
+
"step": 3250
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 4.355289421157685,
|
| 968 |
+
"grad_norm": 45.878814697265625,
|
| 969 |
+
"learning_rate": 1.7384751773049647e-05,
|
| 970 |
+
"loss": 0.8399,
|
| 971 |
+
"step": 3275
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 4.38855622089155,
|
| 975 |
+
"grad_norm": 40.72806167602539,
|
| 976 |
+
"learning_rate": 1.734781323877069e-05,
|
| 977 |
+
"loss": 0.7156,
|
| 978 |
+
"step": 3300
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.421823020625416,
|
| 982 |
+
"grad_norm": 71.23174285888672,
|
| 983 |
+
"learning_rate": 1.731087470449173e-05,
|
| 984 |
+
"loss": 0.7086,
|
| 985 |
+
"step": 3325
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 4.455089820359281,
|
| 989 |
+
"grad_norm": 25.345460891723633,
|
| 990 |
+
"learning_rate": 1.7273936170212768e-05,
|
| 991 |
+
"loss": 0.8167,
|
| 992 |
+
"step": 3350
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 4.488356620093147,
|
| 996 |
+
"grad_norm": 27.16160774230957,
|
| 997 |
+
"learning_rate": 1.7236997635933807e-05,
|
| 998 |
+
"loss": 1.03,
|
| 999 |
+
"step": 3375
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.521623419827012,
|
| 1003 |
+
"grad_norm": 38.71554183959961,
|
| 1004 |
+
"learning_rate": 1.7200059101654846e-05,
|
| 1005 |
+
"loss": 1.0428,
|
| 1006 |
+
"step": 3400
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.554890219560878,
|
| 1010 |
+
"grad_norm": 44.60442352294922,
|
| 1011 |
+
"learning_rate": 1.716312056737589e-05,
|
| 1012 |
+
"loss": 0.645,
|
| 1013 |
+
"step": 3425
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.588157019294744,
|
| 1017 |
+
"grad_norm": 35.653175354003906,
|
| 1018 |
+
"learning_rate": 1.7126182033096928e-05,
|
| 1019 |
+
"loss": 0.8255,
|
| 1020 |
+
"step": 3450
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.621423819028609,
|
| 1024 |
+
"grad_norm": 17.62716293334961,
|
| 1025 |
+
"learning_rate": 1.708924349881797e-05,
|
| 1026 |
+
"loss": 0.7571,
|
| 1027 |
+
"step": 3475
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 4.654690618762475,
|
| 1031 |
+
"grad_norm": 55.70232009887695,
|
| 1032 |
+
"learning_rate": 1.705230496453901e-05,
|
| 1033 |
+
"loss": 0.866,
|
| 1034 |
+
"step": 3500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 4.687957418496341,
|
| 1038 |
+
"grad_norm": 65.43431091308594,
|
| 1039 |
+
"learning_rate": 1.701536643026005e-05,
|
| 1040 |
+
"loss": 0.712,
|
| 1041 |
+
"step": 3525
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.721224218230207,
|
| 1045 |
+
"grad_norm": 45.67749786376953,
|
| 1046 |
+
"learning_rate": 1.6978427895981088e-05,
|
| 1047 |
+
"loss": 0.7377,
|
| 1048 |
+
"step": 3550
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.754491017964072,
|
| 1052 |
+
"grad_norm": 19.373743057250977,
|
| 1053 |
+
"learning_rate": 1.6941489361702127e-05,
|
| 1054 |
+
"loss": 0.9145,
|
| 1055 |
+
"step": 3575
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.787757817697938,
|
| 1059 |
+
"grad_norm": 42.380680084228516,
|
| 1060 |
+
"learning_rate": 1.690455082742317e-05,
|
| 1061 |
+
"loss": 0.8456,
|
| 1062 |
+
"step": 3600
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.821024617431803,
|
| 1066 |
+
"grad_norm": 19.278717041015625,
|
| 1067 |
+
"learning_rate": 1.686761229314421e-05,
|
| 1068 |
+
"loss": 0.8307,
|
| 1069 |
+
"step": 3625
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 4.854291417165669,
|
| 1073 |
+
"grad_norm": 27.214744567871094,
|
| 1074 |
+
"learning_rate": 1.683067375886525e-05,
|
| 1075 |
+
"loss": 0.8568,
|
| 1076 |
+
"step": 3650
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 4.887558216899534,
|
| 1080 |
+
"grad_norm": 24.036231994628906,
|
| 1081 |
+
"learning_rate": 1.679373522458629e-05,
|
| 1082 |
+
"loss": 0.8592,
|
| 1083 |
+
"step": 3675
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.9208250166334,
|
| 1087 |
+
"grad_norm": 37.26359558105469,
|
| 1088 |
+
"learning_rate": 1.675679669030733e-05,
|
| 1089 |
+
"loss": 0.8617,
|
| 1090 |
+
"step": 3700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.954091816367265,
|
| 1094 |
+
"grad_norm": 20.701065063476562,
|
| 1095 |
+
"learning_rate": 1.671985815602837e-05,
|
| 1096 |
+
"loss": 0.9248,
|
| 1097 |
+
"step": 3725
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.987358616101131,
|
| 1101 |
+
"grad_norm": 26.032941818237305,
|
| 1102 |
+
"learning_rate": 1.6682919621749408e-05,
|
| 1103 |
+
"loss": 0.7982,
|
| 1104 |
+
"step": 3750
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.0,
|
| 1108 |
+
"eval_accuracy": 0.874251497005988,
|
| 1109 |
+
"eval_loss": 0.3948594033718109,
|
| 1110 |
+
"eval_macro_f1": 0.6742414894316765,
|
| 1111 |
+
"eval_runtime": 1.562,
|
| 1112 |
+
"eval_samples_per_second": 1710.675,
|
| 1113 |
+
"eval_steps_per_second": 106.917,
|
| 1114 |
+
"eval_weighted_f1": 0.868610204929646,
|
| 1115 |
+
"step": 3760
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.0199600798403194,
|
| 1119 |
+
"grad_norm": 48.44776153564453,
|
| 1120 |
+
"learning_rate": 1.664598108747045e-05,
|
| 1121 |
+
"loss": 0.7793,
|
| 1122 |
+
"step": 3775
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.053226879574185,
|
| 1126 |
+
"grad_norm": 27.714279174804688,
|
| 1127 |
+
"learning_rate": 1.660904255319149e-05,
|
| 1128 |
+
"loss": 0.7261,
|
| 1129 |
+
"step": 3800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.086493679308051,
|
| 1133 |
+
"grad_norm": 60.48532485961914,
|
| 1134 |
+
"learning_rate": 1.6572104018912532e-05,
|
| 1135 |
+
"loss": 0.6221,
|
| 1136 |
+
"step": 3825
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.119760479041916,
|
| 1140 |
+
"grad_norm": 60.350746154785156,
|
| 1141 |
+
"learning_rate": 1.653516548463357e-05,
|
| 1142 |
+
"loss": 0.6676,
|
| 1143 |
+
"step": 3850
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.153027278775782,
|
| 1147 |
+
"grad_norm": 37.48871994018555,
|
| 1148 |
+
"learning_rate": 1.649822695035461e-05,
|
| 1149 |
+
"loss": 0.6285,
|
| 1150 |
+
"step": 3875
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.186294078509647,
|
| 1154 |
+
"grad_norm": 33.875877380371094,
|
| 1155 |
+
"learning_rate": 1.646128841607565e-05,
|
| 1156 |
+
"loss": 0.6937,
|
| 1157 |
+
"step": 3900
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.219560878243513,
|
| 1161 |
+
"grad_norm": 16.855091094970703,
|
| 1162 |
+
"learning_rate": 1.642434988179669e-05,
|
| 1163 |
+
"loss": 0.6802,
|
| 1164 |
+
"step": 3925
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.252827677977379,
|
| 1168 |
+
"grad_norm": 18.034183502197266,
|
| 1169 |
+
"learning_rate": 1.6387411347517732e-05,
|
| 1170 |
+
"loss": 0.5597,
|
| 1171 |
+
"step": 3950
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.286094477711244,
|
| 1175 |
+
"grad_norm": 23.325111389160156,
|
| 1176 |
+
"learning_rate": 1.635047281323877e-05,
|
| 1177 |
+
"loss": 0.8569,
|
| 1178 |
+
"step": 3975
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.319361277445109,
|
| 1182 |
+
"grad_norm": 19.467321395874023,
|
| 1183 |
+
"learning_rate": 1.6313534278959813e-05,
|
| 1184 |
+
"loss": 0.7593,
|
| 1185 |
+
"step": 4000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.352628077178975,
|
| 1189 |
+
"grad_norm": 35.11878204345703,
|
| 1190 |
+
"learning_rate": 1.6276595744680853e-05,
|
| 1191 |
+
"loss": 0.6568,
|
| 1192 |
+
"step": 4025
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.385894876912841,
|
| 1196 |
+
"grad_norm": 39.558387756347656,
|
| 1197 |
+
"learning_rate": 1.6239657210401892e-05,
|
| 1198 |
+
"loss": 0.6274,
|
| 1199 |
+
"step": 4050
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.419161676646707,
|
| 1203 |
+
"grad_norm": 21.97646713256836,
|
| 1204 |
+
"learning_rate": 1.620271867612293e-05,
|
| 1205 |
+
"loss": 0.6535,
|
| 1206 |
+
"step": 4075
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.452428476380573,
|
| 1210 |
+
"grad_norm": 19.23145866394043,
|
| 1211 |
+
"learning_rate": 1.6165780141843974e-05,
|
| 1212 |
+
"loss": 0.5834,
|
| 1213 |
+
"step": 4100
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.485695276114438,
|
| 1217 |
+
"grad_norm": 38.08762741088867,
|
| 1218 |
+
"learning_rate": 1.6128841607565013e-05,
|
| 1219 |
+
"loss": 0.5348,
|
| 1220 |
+
"step": 4125
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.518962075848304,
|
| 1224 |
+
"grad_norm": 25.441452026367188,
|
| 1225 |
+
"learning_rate": 1.6091903073286052e-05,
|
| 1226 |
+
"loss": 0.7722,
|
| 1227 |
+
"step": 4150
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.552228875582169,
|
| 1231 |
+
"grad_norm": 44.96171951293945,
|
| 1232 |
+
"learning_rate": 1.6054964539007095e-05,
|
| 1233 |
+
"loss": 0.6307,
|
| 1234 |
+
"step": 4175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.585495675316035,
|
| 1238 |
+
"grad_norm": 24.344436645507812,
|
| 1239 |
+
"learning_rate": 1.6018026004728134e-05,
|
| 1240 |
+
"loss": 0.5773,
|
| 1241 |
+
"step": 4200
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.6187624750499,
|
| 1245 |
+
"grad_norm": 39.54812240600586,
|
| 1246 |
+
"learning_rate": 1.5981087470449176e-05,
|
| 1247 |
+
"loss": 0.5922,
|
| 1248 |
+
"step": 4225
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.652029274783766,
|
| 1252 |
+
"grad_norm": 34.13394546508789,
|
| 1253 |
+
"learning_rate": 1.5944148936170216e-05,
|
| 1254 |
+
"loss": 0.9566,
|
| 1255 |
+
"step": 4250
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.685296074517631,
|
| 1259 |
+
"grad_norm": 89.45663452148438,
|
| 1260 |
+
"learning_rate": 1.5907210401891255e-05,
|
| 1261 |
+
"loss": 0.7392,
|
| 1262 |
+
"step": 4275
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.718562874251497,
|
| 1266 |
+
"grad_norm": 45.16569137573242,
|
| 1267 |
+
"learning_rate": 1.5870271867612294e-05,
|
| 1268 |
+
"loss": 0.5687,
|
| 1269 |
+
"step": 4300
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.7518296739853625,
|
| 1273 |
+
"grad_norm": 67.9870834350586,
|
| 1274 |
+
"learning_rate": 1.5833333333333333e-05,
|
| 1275 |
+
"loss": 0.6997,
|
| 1276 |
+
"step": 4325
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.7850964737192285,
|
| 1280 |
+
"grad_norm": 34.026763916015625,
|
| 1281 |
+
"learning_rate": 1.5796394799054376e-05,
|
| 1282 |
+
"loss": 0.5136,
|
| 1283 |
+
"step": 4350
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.818363273453094,
|
| 1287 |
+
"grad_norm": 37.53065872192383,
|
| 1288 |
+
"learning_rate": 1.5759456264775415e-05,
|
| 1289 |
+
"loss": 0.5431,
|
| 1290 |
+
"step": 4375
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.85163007318696,
|
| 1294 |
+
"grad_norm": 15.42143440246582,
|
| 1295 |
+
"learning_rate": 1.5722517730496457e-05,
|
| 1296 |
+
"loss": 0.9806,
|
| 1297 |
+
"step": 4400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.884896872920825,
|
| 1301 |
+
"grad_norm": 36.47323226928711,
|
| 1302 |
+
"learning_rate": 1.5685579196217497e-05,
|
| 1303 |
+
"loss": 0.6511,
|
| 1304 |
+
"step": 4425
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.918163672654691,
|
| 1308 |
+
"grad_norm": 25.335269927978516,
|
| 1309 |
+
"learning_rate": 1.5648640661938536e-05,
|
| 1310 |
+
"loss": 0.7454,
|
| 1311 |
+
"step": 4450
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.951430472388556,
|
| 1315 |
+
"grad_norm": 8.366436004638672,
|
| 1316 |
+
"learning_rate": 1.5611702127659575e-05,
|
| 1317 |
+
"loss": 0.5721,
|
| 1318 |
+
"step": 4475
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.984697272122422,
|
| 1322 |
+
"grad_norm": 58.66536331176758,
|
| 1323 |
+
"learning_rate": 1.5574763593380614e-05,
|
| 1324 |
+
"loss": 0.7008,
|
| 1325 |
+
"step": 4500
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.0,
|
| 1329 |
+
"eval_accuracy": 0.8787425149700598,
|
| 1330 |
+
"eval_loss": 0.5009787678718567,
|
| 1331 |
+
"eval_macro_f1": 0.686657575927525,
|
| 1332 |
+
"eval_runtime": 1.5198,
|
| 1333 |
+
"eval_samples_per_second": 1758.169,
|
| 1334 |
+
"eval_steps_per_second": 109.886,
|
| 1335 |
+
"eval_weighted_f1": 0.87233306244317,
|
| 1336 |
+
"step": 4512
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 6.01729873586161,
|
| 1340 |
+
"grad_norm": 14.830575942993164,
|
| 1341 |
+
"learning_rate": 1.5537825059101657e-05,
|
| 1342 |
+
"loss": 0.785,
|
| 1343 |
+
"step": 4525
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 6.050565535595475,
|
| 1347 |
+
"grad_norm": 2.5058681964874268,
|
| 1348 |
+
"learning_rate": 1.5500886524822696e-05,
|
| 1349 |
+
"loss": 0.4656,
|
| 1350 |
+
"step": 4550
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 6.083832335329341,
|
| 1354 |
+
"grad_norm": 8.88479232788086,
|
| 1355 |
+
"learning_rate": 1.546394799054374e-05,
|
| 1356 |
+
"loss": 0.5767,
|
| 1357 |
+
"step": 4575
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 6.117099135063207,
|
| 1361 |
+
"grad_norm": 31.7122859954834,
|
| 1362 |
+
"learning_rate": 1.5427009456264778e-05,
|
| 1363 |
+
"loss": 0.4134,
|
| 1364 |
+
"step": 4600
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 6.1503659347970725,
|
| 1368 |
+
"grad_norm": 12.102670669555664,
|
| 1369 |
+
"learning_rate": 1.5390070921985817e-05,
|
| 1370 |
+
"loss": 0.5668,
|
| 1371 |
+
"step": 4625
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 6.1836327345309385,
|
| 1375 |
+
"grad_norm": 8.906182289123535,
|
| 1376 |
+
"learning_rate": 1.5353132387706856e-05,
|
| 1377 |
+
"loss": 0.4697,
|
| 1378 |
+
"step": 4650
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 6.216899534264804,
|
| 1382 |
+
"grad_norm": 36.15642166137695,
|
| 1383 |
+
"learning_rate": 1.5316193853427895e-05,
|
| 1384 |
+
"loss": 0.5326,
|
| 1385 |
+
"step": 4675
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 6.25016633399867,
|
| 1389 |
+
"grad_norm": 75.987060546875,
|
| 1390 |
+
"learning_rate": 1.5279255319148938e-05,
|
| 1391 |
+
"loss": 0.4873,
|
| 1392 |
+
"step": 4700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 6.283433133732535,
|
| 1396 |
+
"grad_norm": 46.1754264831543,
|
| 1397 |
+
"learning_rate": 1.5242316784869977e-05,
|
| 1398 |
+
"loss": 0.54,
|
| 1399 |
+
"step": 4725
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 6.316699933466401,
|
| 1403 |
+
"grad_norm": 61.719181060791016,
|
| 1404 |
+
"learning_rate": 1.5205378250591018e-05,
|
| 1405 |
+
"loss": 0.5285,
|
| 1406 |
+
"step": 4750
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 6.349966733200266,
|
| 1410 |
+
"grad_norm": 54.186676025390625,
|
| 1411 |
+
"learning_rate": 1.5168439716312057e-05,
|
| 1412 |
+
"loss": 0.5224,
|
| 1413 |
+
"step": 4775
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 6.383233532934132,
|
| 1417 |
+
"grad_norm": 33.198463439941406,
|
| 1418 |
+
"learning_rate": 1.51315011820331e-05,
|
| 1419 |
+
"loss": 0.4514,
|
| 1420 |
+
"step": 4800
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 6.416500332667997,
|
| 1424 |
+
"grad_norm": 22.454734802246094,
|
| 1425 |
+
"learning_rate": 1.5094562647754139e-05,
|
| 1426 |
+
"loss": 0.5112,
|
| 1427 |
+
"step": 4825
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 6.449767132401863,
|
| 1431 |
+
"grad_norm": 43.916725158691406,
|
| 1432 |
+
"learning_rate": 1.5057624113475178e-05,
|
| 1433 |
+
"loss": 0.4969,
|
| 1434 |
+
"step": 4850
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 6.483033932135728,
|
| 1438 |
+
"grad_norm": 63.36553192138672,
|
| 1439 |
+
"learning_rate": 1.5020685579196219e-05,
|
| 1440 |
+
"loss": 0.4341,
|
| 1441 |
+
"step": 4875
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 6.5163007318695945,
|
| 1445 |
+
"grad_norm": 67.7337646484375,
|
| 1446 |
+
"learning_rate": 1.4983747044917258e-05,
|
| 1447 |
+
"loss": 0.5819,
|
| 1448 |
+
"step": 4900
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 6.54956753160346,
|
| 1452 |
+
"grad_norm": 27.86078453063965,
|
| 1453 |
+
"learning_rate": 1.49468085106383e-05,
|
| 1454 |
+
"loss": 0.6794,
|
| 1455 |
+
"step": 4925
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 6.582834331337326,
|
| 1459 |
+
"grad_norm": 94.15121459960938,
|
| 1460 |
+
"learning_rate": 1.490986997635934e-05,
|
| 1461 |
+
"loss": 0.5984,
|
| 1462 |
+
"step": 4950
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 6.616101131071191,
|
| 1466 |
+
"grad_norm": 66.33749389648438,
|
| 1467 |
+
"learning_rate": 1.487293144208038e-05,
|
| 1468 |
+
"loss": 0.6164,
|
| 1469 |
+
"step": 4975
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 6.649367930805057,
|
| 1473 |
+
"grad_norm": 44.44607162475586,
|
| 1474 |
+
"learning_rate": 1.483599290780142e-05,
|
| 1475 |
+
"loss": 0.5336,
|
| 1476 |
+
"step": 5000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 6.682634730538922,
|
| 1480 |
+
"grad_norm": 9.470718383789062,
|
| 1481 |
+
"learning_rate": 1.4799054373522459e-05,
|
| 1482 |
+
"loss": 0.4993,
|
| 1483 |
+
"step": 5025
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 6.715901530272788,
|
| 1487 |
+
"grad_norm": 26.956151962280273,
|
| 1488 |
+
"learning_rate": 1.47621158392435e-05,
|
| 1489 |
+
"loss": 0.5878,
|
| 1490 |
+
"step": 5050
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.749168330006653,
|
| 1494 |
+
"grad_norm": 44.748477935791016,
|
| 1495 |
+
"learning_rate": 1.472517730496454e-05,
|
| 1496 |
+
"loss": 0.7479,
|
| 1497 |
+
"step": 5075
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 6.782435129740519,
|
| 1501 |
+
"grad_norm": 26.0183162689209,
|
| 1502 |
+
"learning_rate": 1.4688238770685582e-05,
|
| 1503 |
+
"loss": 0.4997,
|
| 1504 |
+
"step": 5100
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 6.815701929474384,
|
| 1508 |
+
"grad_norm": 16.654342651367188,
|
| 1509 |
+
"learning_rate": 1.4651300236406621e-05,
|
| 1510 |
+
"loss": 0.4583,
|
| 1511 |
+
"step": 5125
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 6.84896872920825,
|
| 1515 |
+
"grad_norm": 60.56449508666992,
|
| 1516 |
+
"learning_rate": 1.4614361702127662e-05,
|
| 1517 |
+
"loss": 0.6139,
|
| 1518 |
+
"step": 5150
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 6.8822355289421155,
|
| 1522 |
+
"grad_norm": 57.17264175415039,
|
| 1523 |
+
"learning_rate": 1.4577423167848701e-05,
|
| 1524 |
+
"loss": 0.592,
|
| 1525 |
+
"step": 5175
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 6.915502328675982,
|
| 1529 |
+
"grad_norm": 54.279685974121094,
|
| 1530 |
+
"learning_rate": 1.454048463356974e-05,
|
| 1531 |
+
"loss": 0.5221,
|
| 1532 |
+
"step": 5200
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 6.948769128409847,
|
| 1536 |
+
"grad_norm": 16.250139236450195,
|
| 1537 |
+
"learning_rate": 1.4503546099290781e-05,
|
| 1538 |
+
"loss": 0.6532,
|
| 1539 |
+
"step": 5225
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 6.982035928143713,
|
| 1543 |
+
"grad_norm": 22.75954818725586,
|
| 1544 |
+
"learning_rate": 1.446660756501182e-05,
|
| 1545 |
+
"loss": 0.5646,
|
| 1546 |
+
"step": 5250
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 7.0,
|
| 1550 |
+
"eval_accuracy": 0.8619011976047904,
|
| 1551 |
+
"eval_loss": 0.4982662498950958,
|
| 1552 |
+
"eval_macro_f1": 0.6768491427935445,
|
| 1553 |
+
"eval_runtime": 1.6622,
|
| 1554 |
+
"eval_samples_per_second": 1607.536,
|
| 1555 |
+
"eval_steps_per_second": 100.471,
|
| 1556 |
+
"eval_weighted_f1": 0.8656540220162909,
|
| 1557 |
+
"step": 5264
|
| 1558 |
+
}
|
| 1559 |
+
],
|
| 1560 |
+
"logging_steps": 25,
|
| 1561 |
+
"max_steps": 15040,
|
| 1562 |
+
"num_input_tokens_seen": 0,
|
| 1563 |
+
"num_train_epochs": 20,
|
| 1564 |
+
"save_steps": 500,
|
| 1565 |
+
"stateful_callbacks": {
|
| 1566 |
+
"EarlyStoppingCallback": {
|
| 1567 |
+
"args": {
|
| 1568 |
+
"early_stopping_patience": 3,
|
| 1569 |
+
"early_stopping_threshold": 0.0
|
| 1570 |
+
},
|
| 1571 |
+
"attributes": {
|
| 1572 |
+
"early_stopping_patience_counter": 3
|
| 1573 |
+
}
|
| 1574 |
+
},
|
| 1575 |
+
"TrainerControl": {
|
| 1576 |
+
"args": {
|
| 1577 |
+
"should_epoch_stop": false,
|
| 1578 |
+
"should_evaluate": false,
|
| 1579 |
+
"should_log": false,
|
| 1580 |
+
"should_save": true,
|
| 1581 |
+
"should_training_stop": true
|
| 1582 |
+
},
|
| 1583 |
+
"attributes": {}
|
| 1584 |
+
}
|
| 1585 |
+
},
|
| 1586 |
+
"total_flos": 0.0,
|
| 1587 |
+
"train_batch_size": 8,
|
| 1588 |
+
"trial_name": null,
|
| 1589 |
+
"trial_params": null
|
| 1590 |
+
}
|