Add files using upload-large-folder tool
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/trainer_state.json +1371 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/trainer_state.json +1768 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/trainer_state.json +707 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/trainer_state.json +303 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/trainer_state.json +440 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/trainer_state.json +837 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/trainer_state.json +464 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/trainer_state.json +602 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/trainer_state.json +464 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/trainer_state.json +319 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/trainer_state.json +418 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/tokenizer.json +0 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/tokenizer_config.json +55 -0
- ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/trainer_state.json +556 -0
- ood-uit_vsfc-bamibert-topic-domain-02-s42-20260827T155652Z-f97f2036/trainer/checkpoint-2160/special_tokens_map.json +51 -0
- ood-uit_vsfc-bamibert-topic-domain-02-s42-20260827T155652Z-f97f2036/trainer/checkpoint-2160/tokenizer.json +0 -0
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350/trainer_state.json
ADDED
|
@@ -0,0 +1,1371 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 4350,
|
| 3 |
+
"best_metric": 0.845728676575657,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4350,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05747126436781609,
|
| 14 |
+
"grad_norm": 51.11894989013672,
|
| 15 |
+
"learning_rate": 5.517241379310345e-07,
|
| 16 |
+
"loss": 4.1595,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11494252873563218,
|
| 21 |
+
"grad_norm": 24.648967742919922,
|
| 22 |
+
"learning_rate": 1.1264367816091955e-06,
|
| 23 |
+
"loss": 3.9064,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1724137931034483,
|
| 28 |
+
"grad_norm": 9.497029304504395,
|
| 29 |
+
"learning_rate": 1.7011494252873565e-06,
|
| 30 |
+
"loss": 3.5984,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.22988505747126436,
|
| 35 |
+
"grad_norm": 15.340094566345215,
|
| 36 |
+
"learning_rate": 2.2758620689655173e-06,
|
| 37 |
+
"loss": 3.5112,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.28735632183908044,
|
| 42 |
+
"grad_norm": 17.248056411743164,
|
| 43 |
+
"learning_rate": 2.850574712643678e-06,
|
| 44 |
+
"loss": 3.3937,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3448275862068966,
|
| 49 |
+
"grad_norm": 36.462921142578125,
|
| 50 |
+
"learning_rate": 3.425287356321839e-06,
|
| 51 |
+
"loss": 3.1141,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.40229885057471265,
|
| 56 |
+
"grad_norm": 77.80730438232422,
|
| 57 |
+
"learning_rate": 4.000000000000001e-06,
|
| 58 |
+
"loss": 2.927,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.45977011494252873,
|
| 63 |
+
"grad_norm": 94.4952163696289,
|
| 64 |
+
"learning_rate": 4.5747126436781615e-06,
|
| 65 |
+
"loss": 2.7842,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5172413793103449,
|
| 70 |
+
"grad_norm": 63.12592315673828,
|
| 71 |
+
"learning_rate": 5.149425287356322e-06,
|
| 72 |
+
"loss": 2.2325,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5747126436781609,
|
| 77 |
+
"grad_norm": 57.453792572021484,
|
| 78 |
+
"learning_rate": 5.724137931034483e-06,
|
| 79 |
+
"loss": 1.9136,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.632183908045977,
|
| 84 |
+
"grad_norm": 54.622013092041016,
|
| 85 |
+
"learning_rate": 6.298850574712644e-06,
|
| 86 |
+
"loss": 1.7917,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6896551724137931,
|
| 91 |
+
"grad_norm": 65.43468475341797,
|
| 92 |
+
"learning_rate": 6.873563218390805e-06,
|
| 93 |
+
"loss": 1.6804,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7471264367816092,
|
| 98 |
+
"grad_norm": 37.64836883544922,
|
| 99 |
+
"learning_rate": 7.4482758620689665e-06,
|
| 100 |
+
"loss": 1.4561,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8045977011494253,
|
| 105 |
+
"grad_norm": 18.263620376586914,
|
| 106 |
+
"learning_rate": 8.022988505747126e-06,
|
| 107 |
+
"loss": 1.4172,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8620689655172413,
|
| 112 |
+
"grad_norm": 25.493741989135742,
|
| 113 |
+
"learning_rate": 8.597701149425289e-06,
|
| 114 |
+
"loss": 1.3644,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9195402298850575,
|
| 119 |
+
"grad_norm": 20.606971740722656,
|
| 120 |
+
"learning_rate": 9.172413793103448e-06,
|
| 121 |
+
"loss": 1.2567,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9770114942528736,
|
| 126 |
+
"grad_norm": 39.935821533203125,
|
| 127 |
+
"learning_rate": 9.74712643678161e-06,
|
| 128 |
+
"loss": 1.1238,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9159663865546218,
|
| 134 |
+
"eval_loss": 0.2868947386741638,
|
| 135 |
+
"eval_macro_f1": 0.6242481028650465,
|
| 136 |
+
"eval_runtime": 0.9269,
|
| 137 |
+
"eval_samples_per_second": 1669.028,
|
| 138 |
+
"eval_steps_per_second": 104.651,
|
| 139 |
+
"eval_weighted_f1": 0.8950748674120009,
|
| 140 |
+
"step": 435
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.0344827586206897,
|
| 144 |
+
"grad_norm": 45.07005310058594,
|
| 145 |
+
"learning_rate": 1.032183908045977e-05,
|
| 146 |
+
"loss": 1.4034,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.0919540229885056,
|
| 151 |
+
"grad_norm": 29.058692932128906,
|
| 152 |
+
"learning_rate": 1.0896551724137932e-05,
|
| 153 |
+
"loss": 1.2279,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.1494252873563218,
|
| 158 |
+
"grad_norm": 18.84539794921875,
|
| 159 |
+
"learning_rate": 1.1471264367816093e-05,
|
| 160 |
+
"loss": 1.046,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.206896551724138,
|
| 165 |
+
"grad_norm": 124.46168518066406,
|
| 166 |
+
"learning_rate": 1.2045977011494254e-05,
|
| 167 |
+
"loss": 1.1246,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.264367816091954,
|
| 172 |
+
"grad_norm": 29.491249084472656,
|
| 173 |
+
"learning_rate": 1.2620689655172415e-05,
|
| 174 |
+
"loss": 0.9138,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.3218390804597702,
|
| 179 |
+
"grad_norm": 38.48240661621094,
|
| 180 |
+
"learning_rate": 1.3195402298850574e-05,
|
| 181 |
+
"loss": 1.2376,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3793103448275863,
|
| 186 |
+
"grad_norm": 34.79777908325195,
|
| 187 |
+
"learning_rate": 1.3770114942528736e-05,
|
| 188 |
+
"loss": 1.0845,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4367816091954024,
|
| 193 |
+
"grad_norm": 20.485126495361328,
|
| 194 |
+
"learning_rate": 1.4344827586206897e-05,
|
| 195 |
+
"loss": 1.021,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4942528735632183,
|
| 200 |
+
"grad_norm": 24.968517303466797,
|
| 201 |
+
"learning_rate": 1.4919540229885058e-05,
|
| 202 |
+
"loss": 1.0578,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.5517241379310345,
|
| 207 |
+
"grad_norm": 27.639083862304688,
|
| 208 |
+
"learning_rate": 1.549425287356322e-05,
|
| 209 |
+
"loss": 1.1631,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6091954022988506,
|
| 214 |
+
"grad_norm": 40.84177780151367,
|
| 215 |
+
"learning_rate": 1.6068965517241383e-05,
|
| 216 |
+
"loss": 1.0554,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6666666666666665,
|
| 221 |
+
"grad_norm": 30.12960433959961,
|
| 222 |
+
"learning_rate": 1.664367816091954e-05,
|
| 223 |
+
"loss": 0.8529,
|
| 224 |
+
"step": 725
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7241379310344827,
|
| 228 |
+
"grad_norm": 56.6786003112793,
|
| 229 |
+
"learning_rate": 1.72183908045977e-05,
|
| 230 |
+
"loss": 0.9112,
|
| 231 |
+
"step": 750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.7816091954022988,
|
| 235 |
+
"grad_norm": 17.37928009033203,
|
| 236 |
+
"learning_rate": 1.7793103448275862e-05,
|
| 237 |
+
"loss": 0.9023,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.839080459770115,
|
| 242 |
+
"grad_norm": 17.32957649230957,
|
| 243 |
+
"learning_rate": 1.8367816091954027e-05,
|
| 244 |
+
"loss": 1.1465,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.896551724137931,
|
| 249 |
+
"grad_norm": 7.771543979644775,
|
| 250 |
+
"learning_rate": 1.8942528735632187e-05,
|
| 251 |
+
"loss": 0.9536,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.9540229885057472,
|
| 256 |
+
"grad_norm": 28.800386428833008,
|
| 257 |
+
"learning_rate": 1.9517241379310345e-05,
|
| 258 |
+
"loss": 0.9299,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_accuracy": 0.9243697478991597,
|
| 264 |
+
"eval_loss": 0.2366422712802887,
|
| 265 |
+
"eval_macro_f1": 0.81008186276009,
|
| 266 |
+
"eval_runtime": 0.8894,
|
| 267 |
+
"eval_samples_per_second": 1739.302,
|
| 268 |
+
"eval_steps_per_second": 109.058,
|
| 269 |
+
"eval_weighted_f1": 0.9241370987404892,
|
| 270 |
+
"step": 870
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.0114942528735633,
|
| 274 |
+
"grad_norm": 62.22730255126953,
|
| 275 |
+
"learning_rate": 1.9989782886334612e-05,
|
| 276 |
+
"loss": 1.2413,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0689655172413794,
|
| 281 |
+
"grad_norm": 16.838878631591797,
|
| 282 |
+
"learning_rate": 1.9925925925925928e-05,
|
| 283 |
+
"loss": 0.939,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.1264367816091956,
|
| 288 |
+
"grad_norm": 26.60405158996582,
|
| 289 |
+
"learning_rate": 1.9862068965517244e-05,
|
| 290 |
+
"loss": 0.8371,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.1839080459770113,
|
| 295 |
+
"grad_norm": 58.780487060546875,
|
| 296 |
+
"learning_rate": 1.979821200510856e-05,
|
| 297 |
+
"loss": 0.6944,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.2413793103448274,
|
| 302 |
+
"grad_norm": 15.469724655151367,
|
| 303 |
+
"learning_rate": 1.9734355044699872e-05,
|
| 304 |
+
"loss": 0.9043,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.2988505747126435,
|
| 309 |
+
"grad_norm": 37.76985168457031,
|
| 310 |
+
"learning_rate": 1.9670498084291188e-05,
|
| 311 |
+
"loss": 0.9597,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.3563218390804597,
|
| 316 |
+
"grad_norm": 95.69851684570312,
|
| 317 |
+
"learning_rate": 1.9606641123882504e-05,
|
| 318 |
+
"loss": 0.9717,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.413793103448276,
|
| 323 |
+
"grad_norm": 19.623750686645508,
|
| 324 |
+
"learning_rate": 1.954278416347382e-05,
|
| 325 |
+
"loss": 0.7437,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.471264367816092,
|
| 330 |
+
"grad_norm": 39.535621643066406,
|
| 331 |
+
"learning_rate": 1.9478927203065136e-05,
|
| 332 |
+
"loss": 0.9575,
|
| 333 |
+
"step": 1075
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.528735632183908,
|
| 337 |
+
"grad_norm": 24.630657196044922,
|
| 338 |
+
"learning_rate": 1.941507024265645e-05,
|
| 339 |
+
"loss": 0.8364,
|
| 340 |
+
"step": 1100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.586206896551724,
|
| 344 |
+
"grad_norm": 5.918127536773682,
|
| 345 |
+
"learning_rate": 1.9351213282247767e-05,
|
| 346 |
+
"loss": 0.7122,
|
| 347 |
+
"step": 1125
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.6436781609195403,
|
| 351 |
+
"grad_norm": 30.433509826660156,
|
| 352 |
+
"learning_rate": 1.928735632183908e-05,
|
| 353 |
+
"loss": 0.9989,
|
| 354 |
+
"step": 1150
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.7011494252873565,
|
| 358 |
+
"grad_norm": 21.5908145904541,
|
| 359 |
+
"learning_rate": 1.92234993614304e-05,
|
| 360 |
+
"loss": 0.83,
|
| 361 |
+
"step": 1175
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.7586206896551726,
|
| 365 |
+
"grad_norm": 48.29234313964844,
|
| 366 |
+
"learning_rate": 1.915964240102171e-05,
|
| 367 |
+
"loss": 0.8289,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.8160919540229887,
|
| 372 |
+
"grad_norm": 22.402761459350586,
|
| 373 |
+
"learning_rate": 1.9095785440613027e-05,
|
| 374 |
+
"loss": 0.9382,
|
| 375 |
+
"step": 1225
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.873563218390805,
|
| 379 |
+
"grad_norm": 56.446250915527344,
|
| 380 |
+
"learning_rate": 1.9031928480204343e-05,
|
| 381 |
+
"loss": 0.894,
|
| 382 |
+
"step": 1250
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.9310344827586206,
|
| 386 |
+
"grad_norm": 30.15511131286621,
|
| 387 |
+
"learning_rate": 1.896807151979566e-05,
|
| 388 |
+
"loss": 0.8905,
|
| 389 |
+
"step": 1275
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.9885057471264367,
|
| 393 |
+
"grad_norm": 7.9579668045043945,
|
| 394 |
+
"learning_rate": 1.8904214559386975e-05,
|
| 395 |
+
"loss": 0.7984,
|
| 396 |
+
"step": 1300
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.0,
|
| 400 |
+
"eval_accuracy": 0.940530058177117,
|
| 401 |
+
"eval_loss": 0.24241545796394348,
|
| 402 |
+
"eval_macro_f1": 0.8204949787607205,
|
| 403 |
+
"eval_runtime": 0.9471,
|
| 404 |
+
"eval_samples_per_second": 1633.341,
|
| 405 |
+
"eval_steps_per_second": 102.414,
|
| 406 |
+
"eval_weighted_f1": 0.9357503071218877,
|
| 407 |
+
"step": 1305
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.045977011494253,
|
| 411 |
+
"grad_norm": 47.48753356933594,
|
| 412 |
+
"learning_rate": 1.884035759897829e-05,
|
| 413 |
+
"loss": 0.7366,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.103448275862069,
|
| 418 |
+
"grad_norm": 162.04103088378906,
|
| 419 |
+
"learning_rate": 1.8776500638569606e-05,
|
| 420 |
+
"loss": 0.6133,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.160919540229885,
|
| 425 |
+
"grad_norm": 47.9827880859375,
|
| 426 |
+
"learning_rate": 1.871264367816092e-05,
|
| 427 |
+
"loss": 0.7373,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.218390804597701,
|
| 432 |
+
"grad_norm": 5.845567226409912,
|
| 433 |
+
"learning_rate": 1.8648786717752235e-05,
|
| 434 |
+
"loss": 0.6238,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.2758620689655173,
|
| 439 |
+
"grad_norm": 15.353856086730957,
|
| 440 |
+
"learning_rate": 1.8584929757343554e-05,
|
| 441 |
+
"loss": 0.805,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.3333333333333335,
|
| 446 |
+
"grad_norm": 27.47763442993164,
|
| 447 |
+
"learning_rate": 1.8521072796934866e-05,
|
| 448 |
+
"loss": 0.7198,
|
| 449 |
+
"step": 1450
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.3908045977011496,
|
| 453 |
+
"grad_norm": 2.1405858993530273,
|
| 454 |
+
"learning_rate": 1.8457215836526182e-05,
|
| 455 |
+
"loss": 0.6049,
|
| 456 |
+
"step": 1475
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.4482758620689653,
|
| 460 |
+
"grad_norm": 40.75155258178711,
|
| 461 |
+
"learning_rate": 1.8393358876117498e-05,
|
| 462 |
+
"loss": 0.8064,
|
| 463 |
+
"step": 1500
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.5057471264367814,
|
| 467 |
+
"grad_norm": 46.65169906616211,
|
| 468 |
+
"learning_rate": 1.8329501915708814e-05,
|
| 469 |
+
"loss": 0.9762,
|
| 470 |
+
"step": 1525
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.5632183908045976,
|
| 474 |
+
"grad_norm": 1.9898571968078613,
|
| 475 |
+
"learning_rate": 1.826564495530013e-05,
|
| 476 |
+
"loss": 0.7079,
|
| 477 |
+
"step": 1550
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.6206896551724137,
|
| 481 |
+
"grad_norm": 42.3228874206543,
|
| 482 |
+
"learning_rate": 1.8201787994891446e-05,
|
| 483 |
+
"loss": 0.9295,
|
| 484 |
+
"step": 1575
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.67816091954023,
|
| 488 |
+
"grad_norm": 10.075776100158691,
|
| 489 |
+
"learning_rate": 1.813793103448276e-05,
|
| 490 |
+
"loss": 0.8212,
|
| 491 |
+
"step": 1600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.735632183908046,
|
| 495 |
+
"grad_norm": 63.30634307861328,
|
| 496 |
+
"learning_rate": 1.8074074074074074e-05,
|
| 497 |
+
"loss": 0.811,
|
| 498 |
+
"step": 1625
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.793103448275862,
|
| 502 |
+
"grad_norm": 25.703176498413086,
|
| 503 |
+
"learning_rate": 1.801021711366539e-05,
|
| 504 |
+
"loss": 0.5506,
|
| 505 |
+
"step": 1650
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.8505747126436782,
|
| 509 |
+
"grad_norm": 33.45174026489258,
|
| 510 |
+
"learning_rate": 1.7946360153256706e-05,
|
| 511 |
+
"loss": 0.736,
|
| 512 |
+
"step": 1675
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.9080459770114944,
|
| 516 |
+
"grad_norm": 1.0053305625915527,
|
| 517 |
+
"learning_rate": 1.788250319284802e-05,
|
| 518 |
+
"loss": 0.5411,
|
| 519 |
+
"step": 1700
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.9655172413793105,
|
| 523 |
+
"grad_norm": 48.85895538330078,
|
| 524 |
+
"learning_rate": 1.7818646232439337e-05,
|
| 525 |
+
"loss": 0.8589,
|
| 526 |
+
"step": 1725
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.0,
|
| 530 |
+
"eval_accuracy": 0.940530058177117,
|
| 531 |
+
"eval_loss": 0.2217937558889389,
|
| 532 |
+
"eval_macro_f1": 0.81702070732933,
|
| 533 |
+
"eval_runtime": 0.9359,
|
| 534 |
+
"eval_samples_per_second": 1652.994,
|
| 535 |
+
"eval_steps_per_second": 103.646,
|
| 536 |
+
"eval_weighted_f1": 0.9358381270109987,
|
| 537 |
+
"step": 1740
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.022988505747127,
|
| 541 |
+
"grad_norm": 18.975229263305664,
|
| 542 |
+
"learning_rate": 1.7754789272030653e-05,
|
| 543 |
+
"loss": 0.7437,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.080459770114943,
|
| 548 |
+
"grad_norm": 114.38420104980469,
|
| 549 |
+
"learning_rate": 1.769093231162197e-05,
|
| 550 |
+
"loss": 0.4905,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 4.137931034482759,
|
| 555 |
+
"grad_norm": 59.061588287353516,
|
| 556 |
+
"learning_rate": 1.762707535121328e-05,
|
| 557 |
+
"loss": 0.6036,
|
| 558 |
+
"step": 1800
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 4.195402298850575,
|
| 562 |
+
"grad_norm": 54.219871520996094,
|
| 563 |
+
"learning_rate": 1.75632183908046e-05,
|
| 564 |
+
"loss": 0.4817,
|
| 565 |
+
"step": 1825
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 4.252873563218391,
|
| 569 |
+
"grad_norm": 117.5795669555664,
|
| 570 |
+
"learning_rate": 1.7499361430395913e-05,
|
| 571 |
+
"loss": 0.6028,
|
| 572 |
+
"step": 1850
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 4.310344827586207,
|
| 576 |
+
"grad_norm": 124.32904052734375,
|
| 577 |
+
"learning_rate": 1.743550446998723e-05,
|
| 578 |
+
"loss": 0.6587,
|
| 579 |
+
"step": 1875
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 4.3678160919540225,
|
| 583 |
+
"grad_norm": 74.39202880859375,
|
| 584 |
+
"learning_rate": 1.7371647509578545e-05,
|
| 585 |
+
"loss": 0.6221,
|
| 586 |
+
"step": 1900
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.425287356321839,
|
| 590 |
+
"grad_norm": 23.316871643066406,
|
| 591 |
+
"learning_rate": 1.730779054916986e-05,
|
| 592 |
+
"loss": 0.4439,
|
| 593 |
+
"step": 1925
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 4.482758620689655,
|
| 597 |
+
"grad_norm": 37.18128967285156,
|
| 598 |
+
"learning_rate": 1.7243933588761176e-05,
|
| 599 |
+
"loss": 0.7173,
|
| 600 |
+
"step": 1950
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 4.540229885057471,
|
| 604 |
+
"grad_norm": 0.2895273268222809,
|
| 605 |
+
"learning_rate": 1.7180076628352492e-05,
|
| 606 |
+
"loss": 0.4822,
|
| 607 |
+
"step": 1975
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 4.597701149425287,
|
| 611 |
+
"grad_norm": 66.04633331298828,
|
| 612 |
+
"learning_rate": 1.7116219667943808e-05,
|
| 613 |
+
"loss": 0.6299,
|
| 614 |
+
"step": 2000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.655172413793103,
|
| 618 |
+
"grad_norm": 64.7670669555664,
|
| 619 |
+
"learning_rate": 1.705236270753512e-05,
|
| 620 |
+
"loss": 0.7058,
|
| 621 |
+
"step": 2025
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 4.712643678160919,
|
| 625 |
+
"grad_norm": 54.36949920654297,
|
| 626 |
+
"learning_rate": 1.6988505747126436e-05,
|
| 627 |
+
"loss": 0.578,
|
| 628 |
+
"step": 2050
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 4.7701149425287355,
|
| 632 |
+
"grad_norm": 24.543277740478516,
|
| 633 |
+
"learning_rate": 1.6924648786717756e-05,
|
| 634 |
+
"loss": 0.7193,
|
| 635 |
+
"step": 2075
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 4.827586206896552,
|
| 639 |
+
"grad_norm": 43.04127502441406,
|
| 640 |
+
"learning_rate": 1.6860791826309068e-05,
|
| 641 |
+
"loss": 0.588,
|
| 642 |
+
"step": 2100
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 4.885057471264368,
|
| 646 |
+
"grad_norm": 36.25173568725586,
|
| 647 |
+
"learning_rate": 1.6796934865900384e-05,
|
| 648 |
+
"loss": 0.5403,
|
| 649 |
+
"step": 2125
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 4.942528735632184,
|
| 653 |
+
"grad_norm": 37.643680572509766,
|
| 654 |
+
"learning_rate": 1.67330779054917e-05,
|
| 655 |
+
"loss": 0.5459,
|
| 656 |
+
"step": 2150
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.0,
|
| 660 |
+
"grad_norm": 80.67252349853516,
|
| 661 |
+
"learning_rate": 1.6669220945083016e-05,
|
| 662 |
+
"loss": 0.6906,
|
| 663 |
+
"step": 2175
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.0,
|
| 667 |
+
"eval_accuracy": 0.9444085326438267,
|
| 668 |
+
"eval_loss": 0.22898317873477936,
|
| 669 |
+
"eval_macro_f1": 0.8300968480747862,
|
| 670 |
+
"eval_runtime": 0.9158,
|
| 671 |
+
"eval_samples_per_second": 1689.295,
|
| 672 |
+
"eval_steps_per_second": 105.922,
|
| 673 |
+
"eval_weighted_f1": 0.9402869031491028,
|
| 674 |
+
"step": 2175
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 5.057471264367816,
|
| 678 |
+
"grad_norm": 9.40329360961914,
|
| 679 |
+
"learning_rate": 1.6605363984674328e-05,
|
| 680 |
+
"loss": 0.4648,
|
| 681 |
+
"step": 2200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 5.114942528735632,
|
| 685 |
+
"grad_norm": 47.423065185546875,
|
| 686 |
+
"learning_rate": 1.6541507024265647e-05,
|
| 687 |
+
"loss": 0.4028,
|
| 688 |
+
"step": 2225
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 5.172413793103448,
|
| 692 |
+
"grad_norm": 6.621703624725342,
|
| 693 |
+
"learning_rate": 1.6477650063856963e-05,
|
| 694 |
+
"loss": 0.5131,
|
| 695 |
+
"step": 2250
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 5.2298850574712645,
|
| 699 |
+
"grad_norm": 11.388619422912598,
|
| 700 |
+
"learning_rate": 1.6413793103448276e-05,
|
| 701 |
+
"loss": 0.4726,
|
| 702 |
+
"step": 2275
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.287356321839081,
|
| 706 |
+
"grad_norm": 38.415828704833984,
|
| 707 |
+
"learning_rate": 1.634993614303959e-05,
|
| 708 |
+
"loss": 0.4309,
|
| 709 |
+
"step": 2300
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 5.344827586206897,
|
| 713 |
+
"grad_norm": 6.844166278839111,
|
| 714 |
+
"learning_rate": 1.6286079182630907e-05,
|
| 715 |
+
"loss": 0.4982,
|
| 716 |
+
"step": 2325
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 5.402298850574713,
|
| 720 |
+
"grad_norm": 39.43610382080078,
|
| 721 |
+
"learning_rate": 1.6222222222222223e-05,
|
| 722 |
+
"loss": 0.5954,
|
| 723 |
+
"step": 2350
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 5.459770114942529,
|
| 727 |
+
"grad_norm": 0.5552355647087097,
|
| 728 |
+
"learning_rate": 1.615836526181354e-05,
|
| 729 |
+
"loss": 0.535,
|
| 730 |
+
"step": 2375
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 5.517241379310345,
|
| 734 |
+
"grad_norm": 21.38559341430664,
|
| 735 |
+
"learning_rate": 1.6094508301404855e-05,
|
| 736 |
+
"loss": 0.5216,
|
| 737 |
+
"step": 2400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 5.574712643678161,
|
| 741 |
+
"grad_norm": 5.072551250457764,
|
| 742 |
+
"learning_rate": 1.603065134099617e-05,
|
| 743 |
+
"loss": 0.5137,
|
| 744 |
+
"step": 2425
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.6321839080459775,
|
| 748 |
+
"grad_norm": 25.5581111907959,
|
| 749 |
+
"learning_rate": 1.5966794380587483e-05,
|
| 750 |
+
"loss": 0.6129,
|
| 751 |
+
"step": 2450
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 5.689655172413794,
|
| 755 |
+
"grad_norm": 69.6071548461914,
|
| 756 |
+
"learning_rate": 1.5902937420178802e-05,
|
| 757 |
+
"loss": 0.5287,
|
| 758 |
+
"step": 2475
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.747126436781609,
|
| 762 |
+
"grad_norm": 15.225367546081543,
|
| 763 |
+
"learning_rate": 1.5839080459770115e-05,
|
| 764 |
+
"loss": 0.4644,
|
| 765 |
+
"step": 2500
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 5.804597701149425,
|
| 769 |
+
"grad_norm": 38.04400634765625,
|
| 770 |
+
"learning_rate": 1.577522349936143e-05,
|
| 771 |
+
"loss": 0.7929,
|
| 772 |
+
"step": 2525
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 5.862068965517241,
|
| 776 |
+
"grad_norm": 48.28047561645508,
|
| 777 |
+
"learning_rate": 1.5711366538952746e-05,
|
| 778 |
+
"loss": 0.4603,
|
| 779 |
+
"step": 2550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 5.919540229885057,
|
| 783 |
+
"grad_norm": 0.3365640640258789,
|
| 784 |
+
"learning_rate": 1.5647509578544062e-05,
|
| 785 |
+
"loss": 0.5686,
|
| 786 |
+
"step": 2575
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 5.977011494252873,
|
| 790 |
+
"grad_norm": 125.91567993164062,
|
| 791 |
+
"learning_rate": 1.5583652618135378e-05,
|
| 792 |
+
"loss": 0.5187,
|
| 793 |
+
"step": 2600
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 6.0,
|
| 797 |
+
"eval_accuracy": 0.9437621202327084,
|
| 798 |
+
"eval_loss": 0.2934490442276001,
|
| 799 |
+
"eval_macro_f1": 0.8206915027747851,
|
| 800 |
+
"eval_runtime": 0.9319,
|
| 801 |
+
"eval_samples_per_second": 1660.106,
|
| 802 |
+
"eval_steps_per_second": 104.092,
|
| 803 |
+
"eval_weighted_f1": 0.9394797972100883,
|
| 804 |
+
"step": 2610
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 6.0344827586206895,
|
| 808 |
+
"grad_norm": 0.7695260047912598,
|
| 809 |
+
"learning_rate": 1.5519795657726694e-05,
|
| 810 |
+
"loss": 0.5836,
|
| 811 |
+
"step": 2625
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 6.091954022988506,
|
| 815 |
+
"grad_norm": 180.90708923339844,
|
| 816 |
+
"learning_rate": 1.545593869731801e-05,
|
| 817 |
+
"loss": 0.2913,
|
| 818 |
+
"step": 2650
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 6.149425287356322,
|
| 822 |
+
"grad_norm": 0.26224949955940247,
|
| 823 |
+
"learning_rate": 1.5392081736909322e-05,
|
| 824 |
+
"loss": 0.2249,
|
| 825 |
+
"step": 2675
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 6.206896551724138,
|
| 829 |
+
"grad_norm": 6.97501802444458,
|
| 830 |
+
"learning_rate": 1.5328224776500638e-05,
|
| 831 |
+
"loss": 0.7364,
|
| 832 |
+
"step": 2700
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 6.264367816091954,
|
| 836 |
+
"grad_norm": 0.2047957479953766,
|
| 837 |
+
"learning_rate": 1.5264367816091957e-05,
|
| 838 |
+
"loss": 0.4493,
|
| 839 |
+
"step": 2725
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 6.32183908045977,
|
| 843 |
+
"grad_norm": 28.957258224487305,
|
| 844 |
+
"learning_rate": 1.520051085568327e-05,
|
| 845 |
+
"loss": 0.3989,
|
| 846 |
+
"step": 2750
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 6.379310344827586,
|
| 850 |
+
"grad_norm": 5.395802974700928,
|
| 851 |
+
"learning_rate": 1.5136653895274587e-05,
|
| 852 |
+
"loss": 0.6056,
|
| 853 |
+
"step": 2775
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 6.436781609195402,
|
| 857 |
+
"grad_norm": 38.76792907714844,
|
| 858 |
+
"learning_rate": 1.5072796934865901e-05,
|
| 859 |
+
"loss": 0.2687,
|
| 860 |
+
"step": 2800
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 6.494252873563219,
|
| 864 |
+
"grad_norm": 5.454347133636475,
|
| 865 |
+
"learning_rate": 1.5008939974457217e-05,
|
| 866 |
+
"loss": 0.4046,
|
| 867 |
+
"step": 2825
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 6.551724137931035,
|
| 871 |
+
"grad_norm": 15.538047790527344,
|
| 872 |
+
"learning_rate": 1.4945083014048531e-05,
|
| 873 |
+
"loss": 0.4473,
|
| 874 |
+
"step": 2850
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 6.609195402298851,
|
| 878 |
+
"grad_norm": 0.30546435713768005,
|
| 879 |
+
"learning_rate": 1.4881226053639847e-05,
|
| 880 |
+
"loss": 0.3989,
|
| 881 |
+
"step": 2875
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 6.666666666666667,
|
| 885 |
+
"grad_norm": 29.2171573638916,
|
| 886 |
+
"learning_rate": 1.4817369093231165e-05,
|
| 887 |
+
"loss": 0.6199,
|
| 888 |
+
"step": 2900
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 6.724137931034483,
|
| 892 |
+
"grad_norm": 44.1796875,
|
| 893 |
+
"learning_rate": 1.4753512132822479e-05,
|
| 894 |
+
"loss": 0.2984,
|
| 895 |
+
"step": 2925
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 6.781609195402299,
|
| 899 |
+
"grad_norm": 50.609954833984375,
|
| 900 |
+
"learning_rate": 1.4689655172413795e-05,
|
| 901 |
+
"loss": 0.4964,
|
| 902 |
+
"step": 2950
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 6.8390804597701145,
|
| 906 |
+
"grad_norm": 1.0807713270187378,
|
| 907 |
+
"learning_rate": 1.4625798212005109e-05,
|
| 908 |
+
"loss": 0.5935,
|
| 909 |
+
"step": 2975
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 6.896551724137931,
|
| 913 |
+
"grad_norm": 5.837484836578369,
|
| 914 |
+
"learning_rate": 1.4561941251596425e-05,
|
| 915 |
+
"loss": 0.4504,
|
| 916 |
+
"step": 3000
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 6.954022988505747,
|
| 920 |
+
"grad_norm": 0.16583488881587982,
|
| 921 |
+
"learning_rate": 1.4498084291187742e-05,
|
| 922 |
+
"loss": 0.3312,
|
| 923 |
+
"step": 3025
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 7.0,
|
| 927 |
+
"eval_accuracy": 0.9411764705882353,
|
| 928 |
+
"eval_loss": 0.29707837104797363,
|
| 929 |
+
"eval_macro_f1": 0.8407276014218,
|
| 930 |
+
"eval_runtime": 0.9421,
|
| 931 |
+
"eval_samples_per_second": 1642.067,
|
| 932 |
+
"eval_steps_per_second": 102.961,
|
| 933 |
+
"eval_weighted_f1": 0.940398224164607,
|
| 934 |
+
"step": 3045
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 7.011494252873563,
|
| 938 |
+
"grad_norm": 1.0123372077941895,
|
| 939 |
+
"learning_rate": 1.4434227330779056e-05,
|
| 940 |
+
"loss": 0.5288,
|
| 941 |
+
"step": 3050
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 7.068965517241379,
|
| 945 |
+
"grad_norm": 0.2409793734550476,
|
| 946 |
+
"learning_rate": 1.4370370370370372e-05,
|
| 947 |
+
"loss": 0.3362,
|
| 948 |
+
"step": 3075
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 7.126436781609195,
|
| 952 |
+
"grad_norm": 33.50300216674805,
|
| 953 |
+
"learning_rate": 1.4306513409961686e-05,
|
| 954 |
+
"loss": 0.1966,
|
| 955 |
+
"step": 3100
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 7.183908045977011,
|
| 959 |
+
"grad_norm": 0.12312937527894974,
|
| 960 |
+
"learning_rate": 1.4242656449553002e-05,
|
| 961 |
+
"loss": 0.2149,
|
| 962 |
+
"step": 3125
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 7.241379310344827,
|
| 966 |
+
"grad_norm": 50.06519317626953,
|
| 967 |
+
"learning_rate": 1.4178799489144316e-05,
|
| 968 |
+
"loss": 0.5556,
|
| 969 |
+
"step": 3150
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 7.2988505747126435,
|
| 973 |
+
"grad_norm": 3.9388461112976074,
|
| 974 |
+
"learning_rate": 1.4114942528735634e-05,
|
| 975 |
+
"loss": 0.4871,
|
| 976 |
+
"step": 3175
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 7.35632183908046,
|
| 980 |
+
"grad_norm": 49.046329498291016,
|
| 981 |
+
"learning_rate": 1.405108556832695e-05,
|
| 982 |
+
"loss": 0.4448,
|
| 983 |
+
"step": 3200
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 7.413793103448276,
|
| 987 |
+
"grad_norm": 81.77159118652344,
|
| 988 |
+
"learning_rate": 1.3987228607918264e-05,
|
| 989 |
+
"loss": 0.4296,
|
| 990 |
+
"step": 3225
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 7.471264367816092,
|
| 994 |
+
"grad_norm": 19.393810272216797,
|
| 995 |
+
"learning_rate": 1.392337164750958e-05,
|
| 996 |
+
"loss": 0.4822,
|
| 997 |
+
"step": 3250
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 7.528735632183908,
|
| 1001 |
+
"grad_norm": 25.036033630371094,
|
| 1002 |
+
"learning_rate": 1.3859514687100894e-05,
|
| 1003 |
+
"loss": 0.4424,
|
| 1004 |
+
"step": 3275
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 7.586206896551724,
|
| 1008 |
+
"grad_norm": 37.77770233154297,
|
| 1009 |
+
"learning_rate": 1.3795657726692211e-05,
|
| 1010 |
+
"loss": 0.2608,
|
| 1011 |
+
"step": 3300
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 7.64367816091954,
|
| 1015 |
+
"grad_norm": 0.17635223269462585,
|
| 1016 |
+
"learning_rate": 1.3731800766283526e-05,
|
| 1017 |
+
"loss": 0.318,
|
| 1018 |
+
"step": 3325
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 7.7011494252873565,
|
| 1022 |
+
"grad_norm": 22.987146377563477,
|
| 1023 |
+
"learning_rate": 1.3667943805874841e-05,
|
| 1024 |
+
"loss": 0.4785,
|
| 1025 |
+
"step": 3350
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 7.758620689655173,
|
| 1029 |
+
"grad_norm": 0.2942962944507599,
|
| 1030 |
+
"learning_rate": 1.3604086845466157e-05,
|
| 1031 |
+
"loss": 0.3751,
|
| 1032 |
+
"step": 3375
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 7.816091954022989,
|
| 1036 |
+
"grad_norm": 70.56814575195312,
|
| 1037 |
+
"learning_rate": 1.3540229885057471e-05,
|
| 1038 |
+
"loss": 0.5053,
|
| 1039 |
+
"step": 3400
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 7.873563218390805,
|
| 1043 |
+
"grad_norm": 1.6916550397872925,
|
| 1044 |
+
"learning_rate": 1.3476372924648789e-05,
|
| 1045 |
+
"loss": 0.5785,
|
| 1046 |
+
"step": 3425
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 7.931034482758621,
|
| 1050 |
+
"grad_norm": 0.2030503749847412,
|
| 1051 |
+
"learning_rate": 1.3412515964240103e-05,
|
| 1052 |
+
"loss": 0.2199,
|
| 1053 |
+
"step": 3450
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 7.988505747126437,
|
| 1057 |
+
"grad_norm": 15.214978218078613,
|
| 1058 |
+
"learning_rate": 1.3348659003831419e-05,
|
| 1059 |
+
"loss": 0.5497,
|
| 1060 |
+
"step": 3475
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 8.0,
|
| 1064 |
+
"eval_accuracy": 0.9411764705882353,
|
| 1065 |
+
"eval_loss": 0.3194313049316406,
|
| 1066 |
+
"eval_macro_f1": 0.8359358872778418,
|
| 1067 |
+
"eval_runtime": 0.9088,
|
| 1068 |
+
"eval_samples_per_second": 1702.175,
|
| 1069 |
+
"eval_steps_per_second": 106.73,
|
| 1070 |
+
"eval_weighted_f1": 0.9403439511746985,
|
| 1071 |
+
"step": 3480
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 8.045977011494253,
|
| 1075 |
+
"grad_norm": 57.996795654296875,
|
| 1076 |
+
"learning_rate": 1.3284802043422733e-05,
|
| 1077 |
+
"loss": 0.488,
|
| 1078 |
+
"step": 3500
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 8.10344827586207,
|
| 1082 |
+
"grad_norm": 0.18005721271038055,
|
| 1083 |
+
"learning_rate": 1.3220945083014049e-05,
|
| 1084 |
+
"loss": 0.175,
|
| 1085 |
+
"step": 3525
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 8.160919540229886,
|
| 1089 |
+
"grad_norm": 0.10376998782157898,
|
| 1090 |
+
"learning_rate": 1.3157088122605367e-05,
|
| 1091 |
+
"loss": 0.1893,
|
| 1092 |
+
"step": 3550
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 8.218390804597702,
|
| 1096 |
+
"grad_norm": 0.15356546640396118,
|
| 1097 |
+
"learning_rate": 1.309323116219668e-05,
|
| 1098 |
+
"loss": 0.3906,
|
| 1099 |
+
"step": 3575
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 8.275862068965518,
|
| 1103 |
+
"grad_norm": 57.41118621826172,
|
| 1104 |
+
"learning_rate": 1.3029374201787996e-05,
|
| 1105 |
+
"loss": 0.5135,
|
| 1106 |
+
"step": 3600
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 8.333333333333334,
|
| 1110 |
+
"grad_norm": 0.10811824351549149,
|
| 1111 |
+
"learning_rate": 1.296551724137931e-05,
|
| 1112 |
+
"loss": 0.3121,
|
| 1113 |
+
"step": 3625
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 8.39080459770115,
|
| 1117 |
+
"grad_norm": 0.9773123860359192,
|
| 1118 |
+
"learning_rate": 1.2901660280970626e-05,
|
| 1119 |
+
"loss": 0.2917,
|
| 1120 |
+
"step": 3650
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 8.448275862068966,
|
| 1124 |
+
"grad_norm": 39.32749938964844,
|
| 1125 |
+
"learning_rate": 1.283780332056194e-05,
|
| 1126 |
+
"loss": 0.3271,
|
| 1127 |
+
"step": 3675
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 8.505747126436782,
|
| 1131 |
+
"grad_norm": 1.6585102081298828,
|
| 1132 |
+
"learning_rate": 1.2773946360153258e-05,
|
| 1133 |
+
"loss": 0.3852,
|
| 1134 |
+
"step": 3700
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 8.563218390804598,
|
| 1138 |
+
"grad_norm": 0.1653980165719986,
|
| 1139 |
+
"learning_rate": 1.2710089399744574e-05,
|
| 1140 |
+
"loss": 0.1754,
|
| 1141 |
+
"step": 3725
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 8.620689655172415,
|
| 1145 |
+
"grad_norm": 1.1534298658370972,
|
| 1146 |
+
"learning_rate": 1.2646232439335888e-05,
|
| 1147 |
+
"loss": 0.3556,
|
| 1148 |
+
"step": 3750
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 8.678160919540229,
|
| 1152 |
+
"grad_norm": 0.9390562176704407,
|
| 1153 |
+
"learning_rate": 1.2582375478927204e-05,
|
| 1154 |
+
"loss": 0.4791,
|
| 1155 |
+
"step": 3775
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 8.735632183908045,
|
| 1159 |
+
"grad_norm": 0.2481527179479599,
|
| 1160 |
+
"learning_rate": 1.2518518518518518e-05,
|
| 1161 |
+
"loss": 0.2192,
|
| 1162 |
+
"step": 3800
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 8.793103448275861,
|
| 1166 |
+
"grad_norm": 68.067138671875,
|
| 1167 |
+
"learning_rate": 1.2454661558109836e-05,
|
| 1168 |
+
"loss": 0.2412,
|
| 1169 |
+
"step": 3825
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 8.850574712643677,
|
| 1173 |
+
"grad_norm": 0.8862000703811646,
|
| 1174 |
+
"learning_rate": 1.2390804597701151e-05,
|
| 1175 |
+
"loss": 0.2749,
|
| 1176 |
+
"step": 3850
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 8.908045977011493,
|
| 1180 |
+
"grad_norm": 0.29746726155281067,
|
| 1181 |
+
"learning_rate": 1.2326947637292466e-05,
|
| 1182 |
+
"loss": 0.3163,
|
| 1183 |
+
"step": 3875
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 8.96551724137931,
|
| 1187 |
+
"grad_norm": 0.19681236147880554,
|
| 1188 |
+
"learning_rate": 1.2263090676883781e-05,
|
| 1189 |
+
"loss": 0.4417,
|
| 1190 |
+
"step": 3900
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 9.0,
|
| 1194 |
+
"eval_accuracy": 0.9411764705882353,
|
| 1195 |
+
"eval_loss": 0.3171326518058777,
|
| 1196 |
+
"eval_macro_f1": 0.8409499282861729,
|
| 1197 |
+
"eval_runtime": 0.9183,
|
| 1198 |
+
"eval_samples_per_second": 1684.554,
|
| 1199 |
+
"eval_steps_per_second": 105.625,
|
| 1200 |
+
"eval_weighted_f1": 0.9389039510277123,
|
| 1201 |
+
"step": 3915
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 9.022988505747126,
|
| 1205 |
+
"grad_norm": 37.357398986816406,
|
| 1206 |
+
"learning_rate": 1.2199233716475096e-05,
|
| 1207 |
+
"loss": 0.5702,
|
| 1208 |
+
"step": 3925
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 9.080459770114942,
|
| 1212 |
+
"grad_norm": 125.86402893066406,
|
| 1213 |
+
"learning_rate": 1.2135376756066413e-05,
|
| 1214 |
+
"loss": 0.2255,
|
| 1215 |
+
"step": 3950
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 9.137931034482758,
|
| 1219 |
+
"grad_norm": 161.5046844482422,
|
| 1220 |
+
"learning_rate": 1.2071519795657727e-05,
|
| 1221 |
+
"loss": 0.3536,
|
| 1222 |
+
"step": 3975
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 9.195402298850574,
|
| 1226 |
+
"grad_norm": 0.09798524528741837,
|
| 1227 |
+
"learning_rate": 1.2007662835249043e-05,
|
| 1228 |
+
"loss": 0.223,
|
| 1229 |
+
"step": 4000
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 9.25287356321839,
|
| 1233 |
+
"grad_norm": 5.217459678649902,
|
| 1234 |
+
"learning_rate": 1.1943805874840359e-05,
|
| 1235 |
+
"loss": 0.189,
|
| 1236 |
+
"step": 4025
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 9.310344827586206,
|
| 1240 |
+
"grad_norm": 84.28585815429688,
|
| 1241 |
+
"learning_rate": 1.1879948914431673e-05,
|
| 1242 |
+
"loss": 0.1744,
|
| 1243 |
+
"step": 4050
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 9.367816091954023,
|
| 1247 |
+
"grad_norm": 0.07747934758663177,
|
| 1248 |
+
"learning_rate": 1.181609195402299e-05,
|
| 1249 |
+
"loss": 0.2831,
|
| 1250 |
+
"step": 4075
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 9.425287356321839,
|
| 1254 |
+
"grad_norm": 125.41938781738281,
|
| 1255 |
+
"learning_rate": 1.1752234993614305e-05,
|
| 1256 |
+
"loss": 0.2265,
|
| 1257 |
+
"step": 4100
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 9.482758620689655,
|
| 1261 |
+
"grad_norm": 36.66695785522461,
|
| 1262 |
+
"learning_rate": 1.168837803320562e-05,
|
| 1263 |
+
"loss": 0.3566,
|
| 1264 |
+
"step": 4125
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 9.540229885057471,
|
| 1268 |
+
"grad_norm": 0.1046290472149849,
|
| 1269 |
+
"learning_rate": 1.1624521072796935e-05,
|
| 1270 |
+
"loss": 0.2652,
|
| 1271 |
+
"step": 4150
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 9.597701149425287,
|
| 1275 |
+
"grad_norm": 66.43321990966797,
|
| 1276 |
+
"learning_rate": 1.156066411238825e-05,
|
| 1277 |
+
"loss": 0.3305,
|
| 1278 |
+
"step": 4175
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 9.655172413793103,
|
| 1282 |
+
"grad_norm": 134.70994567871094,
|
| 1283 |
+
"learning_rate": 1.1496807151979568e-05,
|
| 1284 |
+
"loss": 0.196,
|
| 1285 |
+
"step": 4200
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 9.71264367816092,
|
| 1289 |
+
"grad_norm": 6.355610370635986,
|
| 1290 |
+
"learning_rate": 1.1432950191570882e-05,
|
| 1291 |
+
"loss": 0.3165,
|
| 1292 |
+
"step": 4225
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 9.770114942528735,
|
| 1296 |
+
"grad_norm": 14.816288948059082,
|
| 1297 |
+
"learning_rate": 1.1369093231162198e-05,
|
| 1298 |
+
"loss": 0.509,
|
| 1299 |
+
"step": 4250
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 9.827586206896552,
|
| 1303 |
+
"grad_norm": 0.14707352221012115,
|
| 1304 |
+
"learning_rate": 1.1305236270753512e-05,
|
| 1305 |
+
"loss": 0.1753,
|
| 1306 |
+
"step": 4275
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 9.885057471264368,
|
| 1310 |
+
"grad_norm": 0.12530837953090668,
|
| 1311 |
+
"learning_rate": 1.1241379310344828e-05,
|
| 1312 |
+
"loss": 0.2454,
|
| 1313 |
+
"step": 4300
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 9.942528735632184,
|
| 1317 |
+
"grad_norm": 133.88351440429688,
|
| 1318 |
+
"learning_rate": 1.1177522349936142e-05,
|
| 1319 |
+
"loss": 0.4458,
|
| 1320 |
+
"step": 4325
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 10.0,
|
| 1324 |
+
"grad_norm": 2.5613210201263428,
|
| 1325 |
+
"learning_rate": 1.111366538952746e-05,
|
| 1326 |
+
"loss": 0.1565,
|
| 1327 |
+
"step": 4350
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 10.0,
|
| 1331 |
+
"eval_accuracy": 0.9411764705882353,
|
| 1332 |
+
"eval_loss": 0.34651657938957214,
|
| 1333 |
+
"eval_macro_f1": 0.845728676575657,
|
| 1334 |
+
"eval_runtime": 0.9212,
|
| 1335 |
+
"eval_samples_per_second": 1679.294,
|
| 1336 |
+
"eval_steps_per_second": 105.295,
|
| 1337 |
+
"eval_weighted_f1": 0.9407380369896519,
|
| 1338 |
+
"step": 4350
|
| 1339 |
+
}
|
| 1340 |
+
],
|
| 1341 |
+
"logging_steps": 25,
|
| 1342 |
+
"max_steps": 8700,
|
| 1343 |
+
"num_input_tokens_seen": 0,
|
| 1344 |
+
"num_train_epochs": 20,
|
| 1345 |
+
"save_steps": 500,
|
| 1346 |
+
"stateful_callbacks": {
|
| 1347 |
+
"EarlyStoppingCallback": {
|
| 1348 |
+
"args": {
|
| 1349 |
+
"early_stopping_patience": 3,
|
| 1350 |
+
"early_stopping_threshold": 0.0
|
| 1351 |
+
},
|
| 1352 |
+
"attributes": {
|
| 1353 |
+
"early_stopping_patience_counter": 0
|
| 1354 |
+
}
|
| 1355 |
+
},
|
| 1356 |
+
"TrainerControl": {
|
| 1357 |
+
"args": {
|
| 1358 |
+
"should_epoch_stop": false,
|
| 1359 |
+
"should_evaluate": false,
|
| 1360 |
+
"should_log": false,
|
| 1361 |
+
"should_save": true,
|
| 1362 |
+
"should_training_stop": false
|
| 1363 |
+
},
|
| 1364 |
+
"attributes": {}
|
| 1365 |
+
}
|
| 1366 |
+
},
|
| 1367 |
+
"total_flos": 0.0,
|
| 1368 |
+
"train_batch_size": 8,
|
| 1369 |
+
"trial_name": null,
|
| 1370 |
+
"trial_params": null
|
| 1371 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-5655/trainer_state.json
ADDED
|
@@ -0,0 +1,1768 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 4350,
|
| 3 |
+
"best_metric": 0.845728676575657,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-00-s42-20260827T154343Z-f97f2036/trainer/checkpoint-4350",
|
| 5 |
+
"epoch": 13.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5655,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05747126436781609,
|
| 14 |
+
"grad_norm": 51.11894989013672,
|
| 15 |
+
"learning_rate": 5.517241379310345e-07,
|
| 16 |
+
"loss": 4.1595,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11494252873563218,
|
| 21 |
+
"grad_norm": 24.648967742919922,
|
| 22 |
+
"learning_rate": 1.1264367816091955e-06,
|
| 23 |
+
"loss": 3.9064,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1724137931034483,
|
| 28 |
+
"grad_norm": 9.497029304504395,
|
| 29 |
+
"learning_rate": 1.7011494252873565e-06,
|
| 30 |
+
"loss": 3.5984,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.22988505747126436,
|
| 35 |
+
"grad_norm": 15.340094566345215,
|
| 36 |
+
"learning_rate": 2.2758620689655173e-06,
|
| 37 |
+
"loss": 3.5112,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.28735632183908044,
|
| 42 |
+
"grad_norm": 17.248056411743164,
|
| 43 |
+
"learning_rate": 2.850574712643678e-06,
|
| 44 |
+
"loss": 3.3937,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3448275862068966,
|
| 49 |
+
"grad_norm": 36.462921142578125,
|
| 50 |
+
"learning_rate": 3.425287356321839e-06,
|
| 51 |
+
"loss": 3.1141,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.40229885057471265,
|
| 56 |
+
"grad_norm": 77.80730438232422,
|
| 57 |
+
"learning_rate": 4.000000000000001e-06,
|
| 58 |
+
"loss": 2.927,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.45977011494252873,
|
| 63 |
+
"grad_norm": 94.4952163696289,
|
| 64 |
+
"learning_rate": 4.5747126436781615e-06,
|
| 65 |
+
"loss": 2.7842,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5172413793103449,
|
| 70 |
+
"grad_norm": 63.12592315673828,
|
| 71 |
+
"learning_rate": 5.149425287356322e-06,
|
| 72 |
+
"loss": 2.2325,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5747126436781609,
|
| 77 |
+
"grad_norm": 57.453792572021484,
|
| 78 |
+
"learning_rate": 5.724137931034483e-06,
|
| 79 |
+
"loss": 1.9136,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.632183908045977,
|
| 84 |
+
"grad_norm": 54.622013092041016,
|
| 85 |
+
"learning_rate": 6.298850574712644e-06,
|
| 86 |
+
"loss": 1.7917,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6896551724137931,
|
| 91 |
+
"grad_norm": 65.43468475341797,
|
| 92 |
+
"learning_rate": 6.873563218390805e-06,
|
| 93 |
+
"loss": 1.6804,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7471264367816092,
|
| 98 |
+
"grad_norm": 37.64836883544922,
|
| 99 |
+
"learning_rate": 7.4482758620689665e-06,
|
| 100 |
+
"loss": 1.4561,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8045977011494253,
|
| 105 |
+
"grad_norm": 18.263620376586914,
|
| 106 |
+
"learning_rate": 8.022988505747126e-06,
|
| 107 |
+
"loss": 1.4172,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8620689655172413,
|
| 112 |
+
"grad_norm": 25.493741989135742,
|
| 113 |
+
"learning_rate": 8.597701149425289e-06,
|
| 114 |
+
"loss": 1.3644,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9195402298850575,
|
| 119 |
+
"grad_norm": 20.606971740722656,
|
| 120 |
+
"learning_rate": 9.172413793103448e-06,
|
| 121 |
+
"loss": 1.2567,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9770114942528736,
|
| 126 |
+
"grad_norm": 39.935821533203125,
|
| 127 |
+
"learning_rate": 9.74712643678161e-06,
|
| 128 |
+
"loss": 1.1238,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9159663865546218,
|
| 134 |
+
"eval_loss": 0.2868947386741638,
|
| 135 |
+
"eval_macro_f1": 0.6242481028650465,
|
| 136 |
+
"eval_runtime": 0.9269,
|
| 137 |
+
"eval_samples_per_second": 1669.028,
|
| 138 |
+
"eval_steps_per_second": 104.651,
|
| 139 |
+
"eval_weighted_f1": 0.8950748674120009,
|
| 140 |
+
"step": 435
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.0344827586206897,
|
| 144 |
+
"grad_norm": 45.07005310058594,
|
| 145 |
+
"learning_rate": 1.032183908045977e-05,
|
| 146 |
+
"loss": 1.4034,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.0919540229885056,
|
| 151 |
+
"grad_norm": 29.058692932128906,
|
| 152 |
+
"learning_rate": 1.0896551724137932e-05,
|
| 153 |
+
"loss": 1.2279,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.1494252873563218,
|
| 158 |
+
"grad_norm": 18.84539794921875,
|
| 159 |
+
"learning_rate": 1.1471264367816093e-05,
|
| 160 |
+
"loss": 1.046,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.206896551724138,
|
| 165 |
+
"grad_norm": 124.46168518066406,
|
| 166 |
+
"learning_rate": 1.2045977011494254e-05,
|
| 167 |
+
"loss": 1.1246,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.264367816091954,
|
| 172 |
+
"grad_norm": 29.491249084472656,
|
| 173 |
+
"learning_rate": 1.2620689655172415e-05,
|
| 174 |
+
"loss": 0.9138,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.3218390804597702,
|
| 179 |
+
"grad_norm": 38.48240661621094,
|
| 180 |
+
"learning_rate": 1.3195402298850574e-05,
|
| 181 |
+
"loss": 1.2376,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3793103448275863,
|
| 186 |
+
"grad_norm": 34.79777908325195,
|
| 187 |
+
"learning_rate": 1.3770114942528736e-05,
|
| 188 |
+
"loss": 1.0845,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4367816091954024,
|
| 193 |
+
"grad_norm": 20.485126495361328,
|
| 194 |
+
"learning_rate": 1.4344827586206897e-05,
|
| 195 |
+
"loss": 1.021,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4942528735632183,
|
| 200 |
+
"grad_norm": 24.968517303466797,
|
| 201 |
+
"learning_rate": 1.4919540229885058e-05,
|
| 202 |
+
"loss": 1.0578,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.5517241379310345,
|
| 207 |
+
"grad_norm": 27.639083862304688,
|
| 208 |
+
"learning_rate": 1.549425287356322e-05,
|
| 209 |
+
"loss": 1.1631,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6091954022988506,
|
| 214 |
+
"grad_norm": 40.84177780151367,
|
| 215 |
+
"learning_rate": 1.6068965517241383e-05,
|
| 216 |
+
"loss": 1.0554,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6666666666666665,
|
| 221 |
+
"grad_norm": 30.12960433959961,
|
| 222 |
+
"learning_rate": 1.664367816091954e-05,
|
| 223 |
+
"loss": 0.8529,
|
| 224 |
+
"step": 725
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7241379310344827,
|
| 228 |
+
"grad_norm": 56.6786003112793,
|
| 229 |
+
"learning_rate": 1.72183908045977e-05,
|
| 230 |
+
"loss": 0.9112,
|
| 231 |
+
"step": 750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.7816091954022988,
|
| 235 |
+
"grad_norm": 17.37928009033203,
|
| 236 |
+
"learning_rate": 1.7793103448275862e-05,
|
| 237 |
+
"loss": 0.9023,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.839080459770115,
|
| 242 |
+
"grad_norm": 17.32957649230957,
|
| 243 |
+
"learning_rate": 1.8367816091954027e-05,
|
| 244 |
+
"loss": 1.1465,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.896551724137931,
|
| 249 |
+
"grad_norm": 7.771543979644775,
|
| 250 |
+
"learning_rate": 1.8942528735632187e-05,
|
| 251 |
+
"loss": 0.9536,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.9540229885057472,
|
| 256 |
+
"grad_norm": 28.800386428833008,
|
| 257 |
+
"learning_rate": 1.9517241379310345e-05,
|
| 258 |
+
"loss": 0.9299,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_accuracy": 0.9243697478991597,
|
| 264 |
+
"eval_loss": 0.2366422712802887,
|
| 265 |
+
"eval_macro_f1": 0.81008186276009,
|
| 266 |
+
"eval_runtime": 0.8894,
|
| 267 |
+
"eval_samples_per_second": 1739.302,
|
| 268 |
+
"eval_steps_per_second": 109.058,
|
| 269 |
+
"eval_weighted_f1": 0.9241370987404892,
|
| 270 |
+
"step": 870
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.0114942528735633,
|
| 274 |
+
"grad_norm": 62.22730255126953,
|
| 275 |
+
"learning_rate": 1.9989782886334612e-05,
|
| 276 |
+
"loss": 1.2413,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0689655172413794,
|
| 281 |
+
"grad_norm": 16.838878631591797,
|
| 282 |
+
"learning_rate": 1.9925925925925928e-05,
|
| 283 |
+
"loss": 0.939,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.1264367816091956,
|
| 288 |
+
"grad_norm": 26.60405158996582,
|
| 289 |
+
"learning_rate": 1.9862068965517244e-05,
|
| 290 |
+
"loss": 0.8371,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.1839080459770113,
|
| 295 |
+
"grad_norm": 58.780487060546875,
|
| 296 |
+
"learning_rate": 1.979821200510856e-05,
|
| 297 |
+
"loss": 0.6944,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.2413793103448274,
|
| 302 |
+
"grad_norm": 15.469724655151367,
|
| 303 |
+
"learning_rate": 1.9734355044699872e-05,
|
| 304 |
+
"loss": 0.9043,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.2988505747126435,
|
| 309 |
+
"grad_norm": 37.76985168457031,
|
| 310 |
+
"learning_rate": 1.9670498084291188e-05,
|
| 311 |
+
"loss": 0.9597,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.3563218390804597,
|
| 316 |
+
"grad_norm": 95.69851684570312,
|
| 317 |
+
"learning_rate": 1.9606641123882504e-05,
|
| 318 |
+
"loss": 0.9717,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.413793103448276,
|
| 323 |
+
"grad_norm": 19.623750686645508,
|
| 324 |
+
"learning_rate": 1.954278416347382e-05,
|
| 325 |
+
"loss": 0.7437,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.471264367816092,
|
| 330 |
+
"grad_norm": 39.535621643066406,
|
| 331 |
+
"learning_rate": 1.9478927203065136e-05,
|
| 332 |
+
"loss": 0.9575,
|
| 333 |
+
"step": 1075
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.528735632183908,
|
| 337 |
+
"grad_norm": 24.630657196044922,
|
| 338 |
+
"learning_rate": 1.941507024265645e-05,
|
| 339 |
+
"loss": 0.8364,
|
| 340 |
+
"step": 1100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.586206896551724,
|
| 344 |
+
"grad_norm": 5.918127536773682,
|
| 345 |
+
"learning_rate": 1.9351213282247767e-05,
|
| 346 |
+
"loss": 0.7122,
|
| 347 |
+
"step": 1125
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.6436781609195403,
|
| 351 |
+
"grad_norm": 30.433509826660156,
|
| 352 |
+
"learning_rate": 1.928735632183908e-05,
|
| 353 |
+
"loss": 0.9989,
|
| 354 |
+
"step": 1150
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.7011494252873565,
|
| 358 |
+
"grad_norm": 21.5908145904541,
|
| 359 |
+
"learning_rate": 1.92234993614304e-05,
|
| 360 |
+
"loss": 0.83,
|
| 361 |
+
"step": 1175
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.7586206896551726,
|
| 365 |
+
"grad_norm": 48.29234313964844,
|
| 366 |
+
"learning_rate": 1.915964240102171e-05,
|
| 367 |
+
"loss": 0.8289,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.8160919540229887,
|
| 372 |
+
"grad_norm": 22.402761459350586,
|
| 373 |
+
"learning_rate": 1.9095785440613027e-05,
|
| 374 |
+
"loss": 0.9382,
|
| 375 |
+
"step": 1225
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.873563218390805,
|
| 379 |
+
"grad_norm": 56.446250915527344,
|
| 380 |
+
"learning_rate": 1.9031928480204343e-05,
|
| 381 |
+
"loss": 0.894,
|
| 382 |
+
"step": 1250
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.9310344827586206,
|
| 386 |
+
"grad_norm": 30.15511131286621,
|
| 387 |
+
"learning_rate": 1.896807151979566e-05,
|
| 388 |
+
"loss": 0.8905,
|
| 389 |
+
"step": 1275
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.9885057471264367,
|
| 393 |
+
"grad_norm": 7.9579668045043945,
|
| 394 |
+
"learning_rate": 1.8904214559386975e-05,
|
| 395 |
+
"loss": 0.7984,
|
| 396 |
+
"step": 1300
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.0,
|
| 400 |
+
"eval_accuracy": 0.940530058177117,
|
| 401 |
+
"eval_loss": 0.24241545796394348,
|
| 402 |
+
"eval_macro_f1": 0.8204949787607205,
|
| 403 |
+
"eval_runtime": 0.9471,
|
| 404 |
+
"eval_samples_per_second": 1633.341,
|
| 405 |
+
"eval_steps_per_second": 102.414,
|
| 406 |
+
"eval_weighted_f1": 0.9357503071218877,
|
| 407 |
+
"step": 1305
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.045977011494253,
|
| 411 |
+
"grad_norm": 47.48753356933594,
|
| 412 |
+
"learning_rate": 1.884035759897829e-05,
|
| 413 |
+
"loss": 0.7366,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.103448275862069,
|
| 418 |
+
"grad_norm": 162.04103088378906,
|
| 419 |
+
"learning_rate": 1.8776500638569606e-05,
|
| 420 |
+
"loss": 0.6133,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.160919540229885,
|
| 425 |
+
"grad_norm": 47.9827880859375,
|
| 426 |
+
"learning_rate": 1.871264367816092e-05,
|
| 427 |
+
"loss": 0.7373,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.218390804597701,
|
| 432 |
+
"grad_norm": 5.845567226409912,
|
| 433 |
+
"learning_rate": 1.8648786717752235e-05,
|
| 434 |
+
"loss": 0.6238,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.2758620689655173,
|
| 439 |
+
"grad_norm": 15.353856086730957,
|
| 440 |
+
"learning_rate": 1.8584929757343554e-05,
|
| 441 |
+
"loss": 0.805,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.3333333333333335,
|
| 446 |
+
"grad_norm": 27.47763442993164,
|
| 447 |
+
"learning_rate": 1.8521072796934866e-05,
|
| 448 |
+
"loss": 0.7198,
|
| 449 |
+
"step": 1450
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.3908045977011496,
|
| 453 |
+
"grad_norm": 2.1405858993530273,
|
| 454 |
+
"learning_rate": 1.8457215836526182e-05,
|
| 455 |
+
"loss": 0.6049,
|
| 456 |
+
"step": 1475
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.4482758620689653,
|
| 460 |
+
"grad_norm": 40.75155258178711,
|
| 461 |
+
"learning_rate": 1.8393358876117498e-05,
|
| 462 |
+
"loss": 0.8064,
|
| 463 |
+
"step": 1500
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.5057471264367814,
|
| 467 |
+
"grad_norm": 46.65169906616211,
|
| 468 |
+
"learning_rate": 1.8329501915708814e-05,
|
| 469 |
+
"loss": 0.9762,
|
| 470 |
+
"step": 1525
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.5632183908045976,
|
| 474 |
+
"grad_norm": 1.9898571968078613,
|
| 475 |
+
"learning_rate": 1.826564495530013e-05,
|
| 476 |
+
"loss": 0.7079,
|
| 477 |
+
"step": 1550
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.6206896551724137,
|
| 481 |
+
"grad_norm": 42.3228874206543,
|
| 482 |
+
"learning_rate": 1.8201787994891446e-05,
|
| 483 |
+
"loss": 0.9295,
|
| 484 |
+
"step": 1575
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.67816091954023,
|
| 488 |
+
"grad_norm": 10.075776100158691,
|
| 489 |
+
"learning_rate": 1.813793103448276e-05,
|
| 490 |
+
"loss": 0.8212,
|
| 491 |
+
"step": 1600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.735632183908046,
|
| 495 |
+
"grad_norm": 63.30634307861328,
|
| 496 |
+
"learning_rate": 1.8074074074074074e-05,
|
| 497 |
+
"loss": 0.811,
|
| 498 |
+
"step": 1625
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.793103448275862,
|
| 502 |
+
"grad_norm": 25.703176498413086,
|
| 503 |
+
"learning_rate": 1.801021711366539e-05,
|
| 504 |
+
"loss": 0.5506,
|
| 505 |
+
"step": 1650
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.8505747126436782,
|
| 509 |
+
"grad_norm": 33.45174026489258,
|
| 510 |
+
"learning_rate": 1.7946360153256706e-05,
|
| 511 |
+
"loss": 0.736,
|
| 512 |
+
"step": 1675
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.9080459770114944,
|
| 516 |
+
"grad_norm": 1.0053305625915527,
|
| 517 |
+
"learning_rate": 1.788250319284802e-05,
|
| 518 |
+
"loss": 0.5411,
|
| 519 |
+
"step": 1700
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.9655172413793105,
|
| 523 |
+
"grad_norm": 48.85895538330078,
|
| 524 |
+
"learning_rate": 1.7818646232439337e-05,
|
| 525 |
+
"loss": 0.8589,
|
| 526 |
+
"step": 1725
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.0,
|
| 530 |
+
"eval_accuracy": 0.940530058177117,
|
| 531 |
+
"eval_loss": 0.2217937558889389,
|
| 532 |
+
"eval_macro_f1": 0.81702070732933,
|
| 533 |
+
"eval_runtime": 0.9359,
|
| 534 |
+
"eval_samples_per_second": 1652.994,
|
| 535 |
+
"eval_steps_per_second": 103.646,
|
| 536 |
+
"eval_weighted_f1": 0.9358381270109987,
|
| 537 |
+
"step": 1740
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.022988505747127,
|
| 541 |
+
"grad_norm": 18.975229263305664,
|
| 542 |
+
"learning_rate": 1.7754789272030653e-05,
|
| 543 |
+
"loss": 0.7437,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.080459770114943,
|
| 548 |
+
"grad_norm": 114.38420104980469,
|
| 549 |
+
"learning_rate": 1.769093231162197e-05,
|
| 550 |
+
"loss": 0.4905,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 4.137931034482759,
|
| 555 |
+
"grad_norm": 59.061588287353516,
|
| 556 |
+
"learning_rate": 1.762707535121328e-05,
|
| 557 |
+
"loss": 0.6036,
|
| 558 |
+
"step": 1800
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 4.195402298850575,
|
| 562 |
+
"grad_norm": 54.219871520996094,
|
| 563 |
+
"learning_rate": 1.75632183908046e-05,
|
| 564 |
+
"loss": 0.4817,
|
| 565 |
+
"step": 1825
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 4.252873563218391,
|
| 569 |
+
"grad_norm": 117.5795669555664,
|
| 570 |
+
"learning_rate": 1.7499361430395913e-05,
|
| 571 |
+
"loss": 0.6028,
|
| 572 |
+
"step": 1850
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 4.310344827586207,
|
| 576 |
+
"grad_norm": 124.32904052734375,
|
| 577 |
+
"learning_rate": 1.743550446998723e-05,
|
| 578 |
+
"loss": 0.6587,
|
| 579 |
+
"step": 1875
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 4.3678160919540225,
|
| 583 |
+
"grad_norm": 74.39202880859375,
|
| 584 |
+
"learning_rate": 1.7371647509578545e-05,
|
| 585 |
+
"loss": 0.6221,
|
| 586 |
+
"step": 1900
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.425287356321839,
|
| 590 |
+
"grad_norm": 23.316871643066406,
|
| 591 |
+
"learning_rate": 1.730779054916986e-05,
|
| 592 |
+
"loss": 0.4439,
|
| 593 |
+
"step": 1925
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 4.482758620689655,
|
| 597 |
+
"grad_norm": 37.18128967285156,
|
| 598 |
+
"learning_rate": 1.7243933588761176e-05,
|
| 599 |
+
"loss": 0.7173,
|
| 600 |
+
"step": 1950
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 4.540229885057471,
|
| 604 |
+
"grad_norm": 0.2895273268222809,
|
| 605 |
+
"learning_rate": 1.7180076628352492e-05,
|
| 606 |
+
"loss": 0.4822,
|
| 607 |
+
"step": 1975
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 4.597701149425287,
|
| 611 |
+
"grad_norm": 66.04633331298828,
|
| 612 |
+
"learning_rate": 1.7116219667943808e-05,
|
| 613 |
+
"loss": 0.6299,
|
| 614 |
+
"step": 2000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.655172413793103,
|
| 618 |
+
"grad_norm": 64.7670669555664,
|
| 619 |
+
"learning_rate": 1.705236270753512e-05,
|
| 620 |
+
"loss": 0.7058,
|
| 621 |
+
"step": 2025
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 4.712643678160919,
|
| 625 |
+
"grad_norm": 54.36949920654297,
|
| 626 |
+
"learning_rate": 1.6988505747126436e-05,
|
| 627 |
+
"loss": 0.578,
|
| 628 |
+
"step": 2050
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 4.7701149425287355,
|
| 632 |
+
"grad_norm": 24.543277740478516,
|
| 633 |
+
"learning_rate": 1.6924648786717756e-05,
|
| 634 |
+
"loss": 0.7193,
|
| 635 |
+
"step": 2075
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 4.827586206896552,
|
| 639 |
+
"grad_norm": 43.04127502441406,
|
| 640 |
+
"learning_rate": 1.6860791826309068e-05,
|
| 641 |
+
"loss": 0.588,
|
| 642 |
+
"step": 2100
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 4.885057471264368,
|
| 646 |
+
"grad_norm": 36.25173568725586,
|
| 647 |
+
"learning_rate": 1.6796934865900384e-05,
|
| 648 |
+
"loss": 0.5403,
|
| 649 |
+
"step": 2125
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 4.942528735632184,
|
| 653 |
+
"grad_norm": 37.643680572509766,
|
| 654 |
+
"learning_rate": 1.67330779054917e-05,
|
| 655 |
+
"loss": 0.5459,
|
| 656 |
+
"step": 2150
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.0,
|
| 660 |
+
"grad_norm": 80.67252349853516,
|
| 661 |
+
"learning_rate": 1.6669220945083016e-05,
|
| 662 |
+
"loss": 0.6906,
|
| 663 |
+
"step": 2175
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.0,
|
| 667 |
+
"eval_accuracy": 0.9444085326438267,
|
| 668 |
+
"eval_loss": 0.22898317873477936,
|
| 669 |
+
"eval_macro_f1": 0.8300968480747862,
|
| 670 |
+
"eval_runtime": 0.9158,
|
| 671 |
+
"eval_samples_per_second": 1689.295,
|
| 672 |
+
"eval_steps_per_second": 105.922,
|
| 673 |
+
"eval_weighted_f1": 0.9402869031491028,
|
| 674 |
+
"step": 2175
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 5.057471264367816,
|
| 678 |
+
"grad_norm": 9.40329360961914,
|
| 679 |
+
"learning_rate": 1.6605363984674328e-05,
|
| 680 |
+
"loss": 0.4648,
|
| 681 |
+
"step": 2200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 5.114942528735632,
|
| 685 |
+
"grad_norm": 47.423065185546875,
|
| 686 |
+
"learning_rate": 1.6541507024265647e-05,
|
| 687 |
+
"loss": 0.4028,
|
| 688 |
+
"step": 2225
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 5.172413793103448,
|
| 692 |
+
"grad_norm": 6.621703624725342,
|
| 693 |
+
"learning_rate": 1.6477650063856963e-05,
|
| 694 |
+
"loss": 0.5131,
|
| 695 |
+
"step": 2250
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 5.2298850574712645,
|
| 699 |
+
"grad_norm": 11.388619422912598,
|
| 700 |
+
"learning_rate": 1.6413793103448276e-05,
|
| 701 |
+
"loss": 0.4726,
|
| 702 |
+
"step": 2275
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.287356321839081,
|
| 706 |
+
"grad_norm": 38.415828704833984,
|
| 707 |
+
"learning_rate": 1.634993614303959e-05,
|
| 708 |
+
"loss": 0.4309,
|
| 709 |
+
"step": 2300
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 5.344827586206897,
|
| 713 |
+
"grad_norm": 6.844166278839111,
|
| 714 |
+
"learning_rate": 1.6286079182630907e-05,
|
| 715 |
+
"loss": 0.4982,
|
| 716 |
+
"step": 2325
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 5.402298850574713,
|
| 720 |
+
"grad_norm": 39.43610382080078,
|
| 721 |
+
"learning_rate": 1.6222222222222223e-05,
|
| 722 |
+
"loss": 0.5954,
|
| 723 |
+
"step": 2350
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 5.459770114942529,
|
| 727 |
+
"grad_norm": 0.5552355647087097,
|
| 728 |
+
"learning_rate": 1.615836526181354e-05,
|
| 729 |
+
"loss": 0.535,
|
| 730 |
+
"step": 2375
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 5.517241379310345,
|
| 734 |
+
"grad_norm": 21.38559341430664,
|
| 735 |
+
"learning_rate": 1.6094508301404855e-05,
|
| 736 |
+
"loss": 0.5216,
|
| 737 |
+
"step": 2400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 5.574712643678161,
|
| 741 |
+
"grad_norm": 5.072551250457764,
|
| 742 |
+
"learning_rate": 1.603065134099617e-05,
|
| 743 |
+
"loss": 0.5137,
|
| 744 |
+
"step": 2425
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.6321839080459775,
|
| 748 |
+
"grad_norm": 25.5581111907959,
|
| 749 |
+
"learning_rate": 1.5966794380587483e-05,
|
| 750 |
+
"loss": 0.6129,
|
| 751 |
+
"step": 2450
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 5.689655172413794,
|
| 755 |
+
"grad_norm": 69.6071548461914,
|
| 756 |
+
"learning_rate": 1.5902937420178802e-05,
|
| 757 |
+
"loss": 0.5287,
|
| 758 |
+
"step": 2475
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.747126436781609,
|
| 762 |
+
"grad_norm": 15.225367546081543,
|
| 763 |
+
"learning_rate": 1.5839080459770115e-05,
|
| 764 |
+
"loss": 0.4644,
|
| 765 |
+
"step": 2500
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 5.804597701149425,
|
| 769 |
+
"grad_norm": 38.04400634765625,
|
| 770 |
+
"learning_rate": 1.577522349936143e-05,
|
| 771 |
+
"loss": 0.7929,
|
| 772 |
+
"step": 2525
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 5.862068965517241,
|
| 776 |
+
"grad_norm": 48.28047561645508,
|
| 777 |
+
"learning_rate": 1.5711366538952746e-05,
|
| 778 |
+
"loss": 0.4603,
|
| 779 |
+
"step": 2550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 5.919540229885057,
|
| 783 |
+
"grad_norm": 0.3365640640258789,
|
| 784 |
+
"learning_rate": 1.5647509578544062e-05,
|
| 785 |
+
"loss": 0.5686,
|
| 786 |
+
"step": 2575
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 5.977011494252873,
|
| 790 |
+
"grad_norm": 125.91567993164062,
|
| 791 |
+
"learning_rate": 1.5583652618135378e-05,
|
| 792 |
+
"loss": 0.5187,
|
| 793 |
+
"step": 2600
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 6.0,
|
| 797 |
+
"eval_accuracy": 0.9437621202327084,
|
| 798 |
+
"eval_loss": 0.2934490442276001,
|
| 799 |
+
"eval_macro_f1": 0.8206915027747851,
|
| 800 |
+
"eval_runtime": 0.9319,
|
| 801 |
+
"eval_samples_per_second": 1660.106,
|
| 802 |
+
"eval_steps_per_second": 104.092,
|
| 803 |
+
"eval_weighted_f1": 0.9394797972100883,
|
| 804 |
+
"step": 2610
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 6.0344827586206895,
|
| 808 |
+
"grad_norm": 0.7695260047912598,
|
| 809 |
+
"learning_rate": 1.5519795657726694e-05,
|
| 810 |
+
"loss": 0.5836,
|
| 811 |
+
"step": 2625
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 6.091954022988506,
|
| 815 |
+
"grad_norm": 180.90708923339844,
|
| 816 |
+
"learning_rate": 1.545593869731801e-05,
|
| 817 |
+
"loss": 0.2913,
|
| 818 |
+
"step": 2650
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 6.149425287356322,
|
| 822 |
+
"grad_norm": 0.26224949955940247,
|
| 823 |
+
"learning_rate": 1.5392081736909322e-05,
|
| 824 |
+
"loss": 0.2249,
|
| 825 |
+
"step": 2675
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 6.206896551724138,
|
| 829 |
+
"grad_norm": 6.97501802444458,
|
| 830 |
+
"learning_rate": 1.5328224776500638e-05,
|
| 831 |
+
"loss": 0.7364,
|
| 832 |
+
"step": 2700
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 6.264367816091954,
|
| 836 |
+
"grad_norm": 0.2047957479953766,
|
| 837 |
+
"learning_rate": 1.5264367816091957e-05,
|
| 838 |
+
"loss": 0.4493,
|
| 839 |
+
"step": 2725
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 6.32183908045977,
|
| 843 |
+
"grad_norm": 28.957258224487305,
|
| 844 |
+
"learning_rate": 1.520051085568327e-05,
|
| 845 |
+
"loss": 0.3989,
|
| 846 |
+
"step": 2750
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 6.379310344827586,
|
| 850 |
+
"grad_norm": 5.395802974700928,
|
| 851 |
+
"learning_rate": 1.5136653895274587e-05,
|
| 852 |
+
"loss": 0.6056,
|
| 853 |
+
"step": 2775
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 6.436781609195402,
|
| 857 |
+
"grad_norm": 38.76792907714844,
|
| 858 |
+
"learning_rate": 1.5072796934865901e-05,
|
| 859 |
+
"loss": 0.2687,
|
| 860 |
+
"step": 2800
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 6.494252873563219,
|
| 864 |
+
"grad_norm": 5.454347133636475,
|
| 865 |
+
"learning_rate": 1.5008939974457217e-05,
|
| 866 |
+
"loss": 0.4046,
|
| 867 |
+
"step": 2825
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 6.551724137931035,
|
| 871 |
+
"grad_norm": 15.538047790527344,
|
| 872 |
+
"learning_rate": 1.4945083014048531e-05,
|
| 873 |
+
"loss": 0.4473,
|
| 874 |
+
"step": 2850
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 6.609195402298851,
|
| 878 |
+
"grad_norm": 0.30546435713768005,
|
| 879 |
+
"learning_rate": 1.4881226053639847e-05,
|
| 880 |
+
"loss": 0.3989,
|
| 881 |
+
"step": 2875
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 6.666666666666667,
|
| 885 |
+
"grad_norm": 29.2171573638916,
|
| 886 |
+
"learning_rate": 1.4817369093231165e-05,
|
| 887 |
+
"loss": 0.6199,
|
| 888 |
+
"step": 2900
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 6.724137931034483,
|
| 892 |
+
"grad_norm": 44.1796875,
|
| 893 |
+
"learning_rate": 1.4753512132822479e-05,
|
| 894 |
+
"loss": 0.2984,
|
| 895 |
+
"step": 2925
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 6.781609195402299,
|
| 899 |
+
"grad_norm": 50.609954833984375,
|
| 900 |
+
"learning_rate": 1.4689655172413795e-05,
|
| 901 |
+
"loss": 0.4964,
|
| 902 |
+
"step": 2950
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 6.8390804597701145,
|
| 906 |
+
"grad_norm": 1.0807713270187378,
|
| 907 |
+
"learning_rate": 1.4625798212005109e-05,
|
| 908 |
+
"loss": 0.5935,
|
| 909 |
+
"step": 2975
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 6.896551724137931,
|
| 913 |
+
"grad_norm": 5.837484836578369,
|
| 914 |
+
"learning_rate": 1.4561941251596425e-05,
|
| 915 |
+
"loss": 0.4504,
|
| 916 |
+
"step": 3000
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 6.954022988505747,
|
| 920 |
+
"grad_norm": 0.16583488881587982,
|
| 921 |
+
"learning_rate": 1.4498084291187742e-05,
|
| 922 |
+
"loss": 0.3312,
|
| 923 |
+
"step": 3025
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 7.0,
|
| 927 |
+
"eval_accuracy": 0.9411764705882353,
|
| 928 |
+
"eval_loss": 0.29707837104797363,
|
| 929 |
+
"eval_macro_f1": 0.8407276014218,
|
| 930 |
+
"eval_runtime": 0.9421,
|
| 931 |
+
"eval_samples_per_second": 1642.067,
|
| 932 |
+
"eval_steps_per_second": 102.961,
|
| 933 |
+
"eval_weighted_f1": 0.940398224164607,
|
| 934 |
+
"step": 3045
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 7.011494252873563,
|
| 938 |
+
"grad_norm": 1.0123372077941895,
|
| 939 |
+
"learning_rate": 1.4434227330779056e-05,
|
| 940 |
+
"loss": 0.5288,
|
| 941 |
+
"step": 3050
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 7.068965517241379,
|
| 945 |
+
"grad_norm": 0.2409793734550476,
|
| 946 |
+
"learning_rate": 1.4370370370370372e-05,
|
| 947 |
+
"loss": 0.3362,
|
| 948 |
+
"step": 3075
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 7.126436781609195,
|
| 952 |
+
"grad_norm": 33.50300216674805,
|
| 953 |
+
"learning_rate": 1.4306513409961686e-05,
|
| 954 |
+
"loss": 0.1966,
|
| 955 |
+
"step": 3100
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 7.183908045977011,
|
| 959 |
+
"grad_norm": 0.12312937527894974,
|
| 960 |
+
"learning_rate": 1.4242656449553002e-05,
|
| 961 |
+
"loss": 0.2149,
|
| 962 |
+
"step": 3125
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 7.241379310344827,
|
| 966 |
+
"grad_norm": 50.06519317626953,
|
| 967 |
+
"learning_rate": 1.4178799489144316e-05,
|
| 968 |
+
"loss": 0.5556,
|
| 969 |
+
"step": 3150
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 7.2988505747126435,
|
| 973 |
+
"grad_norm": 3.9388461112976074,
|
| 974 |
+
"learning_rate": 1.4114942528735634e-05,
|
| 975 |
+
"loss": 0.4871,
|
| 976 |
+
"step": 3175
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 7.35632183908046,
|
| 980 |
+
"grad_norm": 49.046329498291016,
|
| 981 |
+
"learning_rate": 1.405108556832695e-05,
|
| 982 |
+
"loss": 0.4448,
|
| 983 |
+
"step": 3200
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 7.413793103448276,
|
| 987 |
+
"grad_norm": 81.77159118652344,
|
| 988 |
+
"learning_rate": 1.3987228607918264e-05,
|
| 989 |
+
"loss": 0.4296,
|
| 990 |
+
"step": 3225
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 7.471264367816092,
|
| 994 |
+
"grad_norm": 19.393810272216797,
|
| 995 |
+
"learning_rate": 1.392337164750958e-05,
|
| 996 |
+
"loss": 0.4822,
|
| 997 |
+
"step": 3250
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 7.528735632183908,
|
| 1001 |
+
"grad_norm": 25.036033630371094,
|
| 1002 |
+
"learning_rate": 1.3859514687100894e-05,
|
| 1003 |
+
"loss": 0.4424,
|
| 1004 |
+
"step": 3275
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 7.586206896551724,
|
| 1008 |
+
"grad_norm": 37.77770233154297,
|
| 1009 |
+
"learning_rate": 1.3795657726692211e-05,
|
| 1010 |
+
"loss": 0.2608,
|
| 1011 |
+
"step": 3300
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 7.64367816091954,
|
| 1015 |
+
"grad_norm": 0.17635223269462585,
|
| 1016 |
+
"learning_rate": 1.3731800766283526e-05,
|
| 1017 |
+
"loss": 0.318,
|
| 1018 |
+
"step": 3325
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 7.7011494252873565,
|
| 1022 |
+
"grad_norm": 22.987146377563477,
|
| 1023 |
+
"learning_rate": 1.3667943805874841e-05,
|
| 1024 |
+
"loss": 0.4785,
|
| 1025 |
+
"step": 3350
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 7.758620689655173,
|
| 1029 |
+
"grad_norm": 0.2942962944507599,
|
| 1030 |
+
"learning_rate": 1.3604086845466157e-05,
|
| 1031 |
+
"loss": 0.3751,
|
| 1032 |
+
"step": 3375
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 7.816091954022989,
|
| 1036 |
+
"grad_norm": 70.56814575195312,
|
| 1037 |
+
"learning_rate": 1.3540229885057471e-05,
|
| 1038 |
+
"loss": 0.5053,
|
| 1039 |
+
"step": 3400
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 7.873563218390805,
|
| 1043 |
+
"grad_norm": 1.6916550397872925,
|
| 1044 |
+
"learning_rate": 1.3476372924648789e-05,
|
| 1045 |
+
"loss": 0.5785,
|
| 1046 |
+
"step": 3425
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 7.931034482758621,
|
| 1050 |
+
"grad_norm": 0.2030503749847412,
|
| 1051 |
+
"learning_rate": 1.3412515964240103e-05,
|
| 1052 |
+
"loss": 0.2199,
|
| 1053 |
+
"step": 3450
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 7.988505747126437,
|
| 1057 |
+
"grad_norm": 15.214978218078613,
|
| 1058 |
+
"learning_rate": 1.3348659003831419e-05,
|
| 1059 |
+
"loss": 0.5497,
|
| 1060 |
+
"step": 3475
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 8.0,
|
| 1064 |
+
"eval_accuracy": 0.9411764705882353,
|
| 1065 |
+
"eval_loss": 0.3194313049316406,
|
| 1066 |
+
"eval_macro_f1": 0.8359358872778418,
|
| 1067 |
+
"eval_runtime": 0.9088,
|
| 1068 |
+
"eval_samples_per_second": 1702.175,
|
| 1069 |
+
"eval_steps_per_second": 106.73,
|
| 1070 |
+
"eval_weighted_f1": 0.9403439511746985,
|
| 1071 |
+
"step": 3480
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 8.045977011494253,
|
| 1075 |
+
"grad_norm": 57.996795654296875,
|
| 1076 |
+
"learning_rate": 1.3284802043422733e-05,
|
| 1077 |
+
"loss": 0.488,
|
| 1078 |
+
"step": 3500
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 8.10344827586207,
|
| 1082 |
+
"grad_norm": 0.18005721271038055,
|
| 1083 |
+
"learning_rate": 1.3220945083014049e-05,
|
| 1084 |
+
"loss": 0.175,
|
| 1085 |
+
"step": 3525
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 8.160919540229886,
|
| 1089 |
+
"grad_norm": 0.10376998782157898,
|
| 1090 |
+
"learning_rate": 1.3157088122605367e-05,
|
| 1091 |
+
"loss": 0.1893,
|
| 1092 |
+
"step": 3550
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 8.218390804597702,
|
| 1096 |
+
"grad_norm": 0.15356546640396118,
|
| 1097 |
+
"learning_rate": 1.309323116219668e-05,
|
| 1098 |
+
"loss": 0.3906,
|
| 1099 |
+
"step": 3575
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 8.275862068965518,
|
| 1103 |
+
"grad_norm": 57.41118621826172,
|
| 1104 |
+
"learning_rate": 1.3029374201787996e-05,
|
| 1105 |
+
"loss": 0.5135,
|
| 1106 |
+
"step": 3600
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 8.333333333333334,
|
| 1110 |
+
"grad_norm": 0.10811824351549149,
|
| 1111 |
+
"learning_rate": 1.296551724137931e-05,
|
| 1112 |
+
"loss": 0.3121,
|
| 1113 |
+
"step": 3625
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 8.39080459770115,
|
| 1117 |
+
"grad_norm": 0.9773123860359192,
|
| 1118 |
+
"learning_rate": 1.2901660280970626e-05,
|
| 1119 |
+
"loss": 0.2917,
|
| 1120 |
+
"step": 3650
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 8.448275862068966,
|
| 1124 |
+
"grad_norm": 39.32749938964844,
|
| 1125 |
+
"learning_rate": 1.283780332056194e-05,
|
| 1126 |
+
"loss": 0.3271,
|
| 1127 |
+
"step": 3675
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 8.505747126436782,
|
| 1131 |
+
"grad_norm": 1.6585102081298828,
|
| 1132 |
+
"learning_rate": 1.2773946360153258e-05,
|
| 1133 |
+
"loss": 0.3852,
|
| 1134 |
+
"step": 3700
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 8.563218390804598,
|
| 1138 |
+
"grad_norm": 0.1653980165719986,
|
| 1139 |
+
"learning_rate": 1.2710089399744574e-05,
|
| 1140 |
+
"loss": 0.1754,
|
| 1141 |
+
"step": 3725
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 8.620689655172415,
|
| 1145 |
+
"grad_norm": 1.1534298658370972,
|
| 1146 |
+
"learning_rate": 1.2646232439335888e-05,
|
| 1147 |
+
"loss": 0.3556,
|
| 1148 |
+
"step": 3750
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 8.678160919540229,
|
| 1152 |
+
"grad_norm": 0.9390562176704407,
|
| 1153 |
+
"learning_rate": 1.2582375478927204e-05,
|
| 1154 |
+
"loss": 0.4791,
|
| 1155 |
+
"step": 3775
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 8.735632183908045,
|
| 1159 |
+
"grad_norm": 0.2481527179479599,
|
| 1160 |
+
"learning_rate": 1.2518518518518518e-05,
|
| 1161 |
+
"loss": 0.2192,
|
| 1162 |
+
"step": 3800
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 8.793103448275861,
|
| 1166 |
+
"grad_norm": 68.067138671875,
|
| 1167 |
+
"learning_rate": 1.2454661558109836e-05,
|
| 1168 |
+
"loss": 0.2412,
|
| 1169 |
+
"step": 3825
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 8.850574712643677,
|
| 1173 |
+
"grad_norm": 0.8862000703811646,
|
| 1174 |
+
"learning_rate": 1.2390804597701151e-05,
|
| 1175 |
+
"loss": 0.2749,
|
| 1176 |
+
"step": 3850
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 8.908045977011493,
|
| 1180 |
+
"grad_norm": 0.29746726155281067,
|
| 1181 |
+
"learning_rate": 1.2326947637292466e-05,
|
| 1182 |
+
"loss": 0.3163,
|
| 1183 |
+
"step": 3875
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 8.96551724137931,
|
| 1187 |
+
"grad_norm": 0.19681236147880554,
|
| 1188 |
+
"learning_rate": 1.2263090676883781e-05,
|
| 1189 |
+
"loss": 0.4417,
|
| 1190 |
+
"step": 3900
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 9.0,
|
| 1194 |
+
"eval_accuracy": 0.9411764705882353,
|
| 1195 |
+
"eval_loss": 0.3171326518058777,
|
| 1196 |
+
"eval_macro_f1": 0.8409499282861729,
|
| 1197 |
+
"eval_runtime": 0.9183,
|
| 1198 |
+
"eval_samples_per_second": 1684.554,
|
| 1199 |
+
"eval_steps_per_second": 105.625,
|
| 1200 |
+
"eval_weighted_f1": 0.9389039510277123,
|
| 1201 |
+
"step": 3915
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 9.022988505747126,
|
| 1205 |
+
"grad_norm": 37.357398986816406,
|
| 1206 |
+
"learning_rate": 1.2199233716475096e-05,
|
| 1207 |
+
"loss": 0.5702,
|
| 1208 |
+
"step": 3925
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 9.080459770114942,
|
| 1212 |
+
"grad_norm": 125.86402893066406,
|
| 1213 |
+
"learning_rate": 1.2135376756066413e-05,
|
| 1214 |
+
"loss": 0.2255,
|
| 1215 |
+
"step": 3950
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 9.137931034482758,
|
| 1219 |
+
"grad_norm": 161.5046844482422,
|
| 1220 |
+
"learning_rate": 1.2071519795657727e-05,
|
| 1221 |
+
"loss": 0.3536,
|
| 1222 |
+
"step": 3975
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 9.195402298850574,
|
| 1226 |
+
"grad_norm": 0.09798524528741837,
|
| 1227 |
+
"learning_rate": 1.2007662835249043e-05,
|
| 1228 |
+
"loss": 0.223,
|
| 1229 |
+
"step": 4000
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 9.25287356321839,
|
| 1233 |
+
"grad_norm": 5.217459678649902,
|
| 1234 |
+
"learning_rate": 1.1943805874840359e-05,
|
| 1235 |
+
"loss": 0.189,
|
| 1236 |
+
"step": 4025
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 9.310344827586206,
|
| 1240 |
+
"grad_norm": 84.28585815429688,
|
| 1241 |
+
"learning_rate": 1.1879948914431673e-05,
|
| 1242 |
+
"loss": 0.1744,
|
| 1243 |
+
"step": 4050
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 9.367816091954023,
|
| 1247 |
+
"grad_norm": 0.07747934758663177,
|
| 1248 |
+
"learning_rate": 1.181609195402299e-05,
|
| 1249 |
+
"loss": 0.2831,
|
| 1250 |
+
"step": 4075
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 9.425287356321839,
|
| 1254 |
+
"grad_norm": 125.41938781738281,
|
| 1255 |
+
"learning_rate": 1.1752234993614305e-05,
|
| 1256 |
+
"loss": 0.2265,
|
| 1257 |
+
"step": 4100
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 9.482758620689655,
|
| 1261 |
+
"grad_norm": 36.66695785522461,
|
| 1262 |
+
"learning_rate": 1.168837803320562e-05,
|
| 1263 |
+
"loss": 0.3566,
|
| 1264 |
+
"step": 4125
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 9.540229885057471,
|
| 1268 |
+
"grad_norm": 0.1046290472149849,
|
| 1269 |
+
"learning_rate": 1.1624521072796935e-05,
|
| 1270 |
+
"loss": 0.2652,
|
| 1271 |
+
"step": 4150
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 9.597701149425287,
|
| 1275 |
+
"grad_norm": 66.43321990966797,
|
| 1276 |
+
"learning_rate": 1.156066411238825e-05,
|
| 1277 |
+
"loss": 0.3305,
|
| 1278 |
+
"step": 4175
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 9.655172413793103,
|
| 1282 |
+
"grad_norm": 134.70994567871094,
|
| 1283 |
+
"learning_rate": 1.1496807151979568e-05,
|
| 1284 |
+
"loss": 0.196,
|
| 1285 |
+
"step": 4200
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 9.71264367816092,
|
| 1289 |
+
"grad_norm": 6.355610370635986,
|
| 1290 |
+
"learning_rate": 1.1432950191570882e-05,
|
| 1291 |
+
"loss": 0.3165,
|
| 1292 |
+
"step": 4225
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 9.770114942528735,
|
| 1296 |
+
"grad_norm": 14.816288948059082,
|
| 1297 |
+
"learning_rate": 1.1369093231162198e-05,
|
| 1298 |
+
"loss": 0.509,
|
| 1299 |
+
"step": 4250
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 9.827586206896552,
|
| 1303 |
+
"grad_norm": 0.14707352221012115,
|
| 1304 |
+
"learning_rate": 1.1305236270753512e-05,
|
| 1305 |
+
"loss": 0.1753,
|
| 1306 |
+
"step": 4275
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 9.885057471264368,
|
| 1310 |
+
"grad_norm": 0.12530837953090668,
|
| 1311 |
+
"learning_rate": 1.1241379310344828e-05,
|
| 1312 |
+
"loss": 0.2454,
|
| 1313 |
+
"step": 4300
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 9.942528735632184,
|
| 1317 |
+
"grad_norm": 133.88351440429688,
|
| 1318 |
+
"learning_rate": 1.1177522349936142e-05,
|
| 1319 |
+
"loss": 0.4458,
|
| 1320 |
+
"step": 4325
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 10.0,
|
| 1324 |
+
"grad_norm": 2.5613210201263428,
|
| 1325 |
+
"learning_rate": 1.111366538952746e-05,
|
| 1326 |
+
"loss": 0.1565,
|
| 1327 |
+
"step": 4350
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 10.0,
|
| 1331 |
+
"eval_accuracy": 0.9411764705882353,
|
| 1332 |
+
"eval_loss": 0.34651657938957214,
|
| 1333 |
+
"eval_macro_f1": 0.845728676575657,
|
| 1334 |
+
"eval_runtime": 0.9212,
|
| 1335 |
+
"eval_samples_per_second": 1679.294,
|
| 1336 |
+
"eval_steps_per_second": 105.295,
|
| 1337 |
+
"eval_weighted_f1": 0.9407380369896519,
|
| 1338 |
+
"step": 4350
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 10.057471264367816,
|
| 1342 |
+
"grad_norm": 0.12360727787017822,
|
| 1343 |
+
"learning_rate": 1.1049808429118776e-05,
|
| 1344 |
+
"loss": 0.0788,
|
| 1345 |
+
"step": 4375
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 10.114942528735632,
|
| 1349 |
+
"grad_norm": 0.5807376503944397,
|
| 1350 |
+
"learning_rate": 1.098595146871009e-05,
|
| 1351 |
+
"loss": 0.2313,
|
| 1352 |
+
"step": 4400
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 10.172413793103448,
|
| 1356 |
+
"grad_norm": 169.80821228027344,
|
| 1357 |
+
"learning_rate": 1.0922094508301406e-05,
|
| 1358 |
+
"loss": 0.2292,
|
| 1359 |
+
"step": 4425
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 10.229885057471265,
|
| 1363 |
+
"grad_norm": 0.3469511568546295,
|
| 1364 |
+
"learning_rate": 1.085823754789272e-05,
|
| 1365 |
+
"loss": 0.1256,
|
| 1366 |
+
"step": 4450
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 10.28735632183908,
|
| 1370 |
+
"grad_norm": 16.0269832611084,
|
| 1371 |
+
"learning_rate": 1.0794380587484037e-05,
|
| 1372 |
+
"loss": 0.3115,
|
| 1373 |
+
"step": 4475
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 10.344827586206897,
|
| 1377 |
+
"grad_norm": 31.39403533935547,
|
| 1378 |
+
"learning_rate": 1.0730523627075351e-05,
|
| 1379 |
+
"loss": 0.2883,
|
| 1380 |
+
"step": 4500
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 10.402298850574713,
|
| 1384 |
+
"grad_norm": 0.2734578251838684,
|
| 1385 |
+
"learning_rate": 1.0666666666666667e-05,
|
| 1386 |
+
"loss": 0.2558,
|
| 1387 |
+
"step": 4525
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 10.459770114942529,
|
| 1391 |
+
"grad_norm": 1.9513689279556274,
|
| 1392 |
+
"learning_rate": 1.0602809706257983e-05,
|
| 1393 |
+
"loss": 0.3471,
|
| 1394 |
+
"step": 4550
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 10.517241379310345,
|
| 1398 |
+
"grad_norm": 0.062036216259002686,
|
| 1399 |
+
"learning_rate": 1.0538952745849297e-05,
|
| 1400 |
+
"loss": 0.2528,
|
| 1401 |
+
"step": 4575
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 10.574712643678161,
|
| 1405 |
+
"grad_norm": 6.157421588897705,
|
| 1406 |
+
"learning_rate": 1.0475095785440615e-05,
|
| 1407 |
+
"loss": 0.315,
|
| 1408 |
+
"step": 4600
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 10.632183908045977,
|
| 1412 |
+
"grad_norm": 10.171277046203613,
|
| 1413 |
+
"learning_rate": 1.0411238825031929e-05,
|
| 1414 |
+
"loss": 0.2664,
|
| 1415 |
+
"step": 4625
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 10.689655172413794,
|
| 1419 |
+
"grad_norm": 0.13759061694145203,
|
| 1420 |
+
"learning_rate": 1.0347381864623245e-05,
|
| 1421 |
+
"loss": 0.148,
|
| 1422 |
+
"step": 4650
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 10.74712643678161,
|
| 1426 |
+
"grad_norm": 5.043412208557129,
|
| 1427 |
+
"learning_rate": 1.028352490421456e-05,
|
| 1428 |
+
"loss": 0.102,
|
| 1429 |
+
"step": 4675
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 10.804597701149426,
|
| 1433 |
+
"grad_norm": 129.40708923339844,
|
| 1434 |
+
"learning_rate": 1.0219667943805875e-05,
|
| 1435 |
+
"loss": 0.2517,
|
| 1436 |
+
"step": 4700
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 10.862068965517242,
|
| 1440 |
+
"grad_norm": 10.171950340270996,
|
| 1441 |
+
"learning_rate": 1.0155810983397192e-05,
|
| 1442 |
+
"loss": 0.2992,
|
| 1443 |
+
"step": 4725
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 10.919540229885058,
|
| 1447 |
+
"grad_norm": 20.464923858642578,
|
| 1448 |
+
"learning_rate": 1.0091954022988506e-05,
|
| 1449 |
+
"loss": 0.3593,
|
| 1450 |
+
"step": 4750
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 10.977011494252874,
|
| 1454 |
+
"grad_norm": 316.33941650390625,
|
| 1455 |
+
"learning_rate": 1.0028097062579822e-05,
|
| 1456 |
+
"loss": 0.2407,
|
| 1457 |
+
"step": 4775
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 11.0,
|
| 1461 |
+
"eval_accuracy": 0.9385908209437621,
|
| 1462 |
+
"eval_loss": 0.382472962141037,
|
| 1463 |
+
"eval_macro_f1": 0.8208813217485208,
|
| 1464 |
+
"eval_runtime": 0.9563,
|
| 1465 |
+
"eval_samples_per_second": 1617.637,
|
| 1466 |
+
"eval_steps_per_second": 101.429,
|
| 1467 |
+
"eval_weighted_f1": 0.9356086653328916,
|
| 1468 |
+
"step": 4785
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 11.03448275862069,
|
| 1472 |
+
"grad_norm": 0.06662601977586746,
|
| 1473 |
+
"learning_rate": 9.964240102171138e-06,
|
| 1474 |
+
"loss": 0.2607,
|
| 1475 |
+
"step": 4800
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 11.091954022988507,
|
| 1479 |
+
"grad_norm": 0.08112962543964386,
|
| 1480 |
+
"learning_rate": 9.900383141762452e-06,
|
| 1481 |
+
"loss": 0.1062,
|
| 1482 |
+
"step": 4825
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 11.149425287356323,
|
| 1486 |
+
"grad_norm": 0.11819115281105042,
|
| 1487 |
+
"learning_rate": 9.836526181353768e-06,
|
| 1488 |
+
"loss": 0.1984,
|
| 1489 |
+
"step": 4850
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 11.206896551724139,
|
| 1493 |
+
"grad_norm": 0.060591623187065125,
|
| 1494 |
+
"learning_rate": 9.772669220945084e-06,
|
| 1495 |
+
"loss": 0.2375,
|
| 1496 |
+
"step": 4875
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 11.264367816091955,
|
| 1500 |
+
"grad_norm": 0.839734673500061,
|
| 1501 |
+
"learning_rate": 9.708812260536398e-06,
|
| 1502 |
+
"loss": 0.1414,
|
| 1503 |
+
"step": 4900
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 11.32183908045977,
|
| 1507 |
+
"grad_norm": 13.846209526062012,
|
| 1508 |
+
"learning_rate": 9.644955300127716e-06,
|
| 1509 |
+
"loss": 0.1955,
|
| 1510 |
+
"step": 4925
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 11.379310344827585,
|
| 1514 |
+
"grad_norm": 1.1597739458084106,
|
| 1515 |
+
"learning_rate": 9.58109833971903e-06,
|
| 1516 |
+
"loss": 0.1872,
|
| 1517 |
+
"step": 4950
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 11.436781609195402,
|
| 1521 |
+
"grad_norm": 1.2385079860687256,
|
| 1522 |
+
"learning_rate": 9.517241379310346e-06,
|
| 1523 |
+
"loss": 0.2798,
|
| 1524 |
+
"step": 4975
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 11.494252873563218,
|
| 1528 |
+
"grad_norm": 0.7285603284835815,
|
| 1529 |
+
"learning_rate": 9.453384418901661e-06,
|
| 1530 |
+
"loss": 0.2741,
|
| 1531 |
+
"step": 5000
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 11.551724137931034,
|
| 1535 |
+
"grad_norm": 152.7362060546875,
|
| 1536 |
+
"learning_rate": 9.389527458492976e-06,
|
| 1537 |
+
"loss": 0.1822,
|
| 1538 |
+
"step": 5025
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 11.60919540229885,
|
| 1542 |
+
"grad_norm": 0.0699530690908432,
|
| 1543 |
+
"learning_rate": 9.325670498084291e-06,
|
| 1544 |
+
"loss": 0.1585,
|
| 1545 |
+
"step": 5050
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 11.666666666666666,
|
| 1549 |
+
"grad_norm": 0.15238483250141144,
|
| 1550 |
+
"learning_rate": 9.261813537675607e-06,
|
| 1551 |
+
"loss": 0.1563,
|
| 1552 |
+
"step": 5075
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 11.724137931034482,
|
| 1556 |
+
"grad_norm": 65.6576919555664,
|
| 1557 |
+
"learning_rate": 9.197956577266923e-06,
|
| 1558 |
+
"loss": 0.1491,
|
| 1559 |
+
"step": 5100
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 11.781609195402298,
|
| 1563 |
+
"grad_norm": 0.03878547623753548,
|
| 1564 |
+
"learning_rate": 9.134099616858239e-06,
|
| 1565 |
+
"loss": 0.1046,
|
| 1566 |
+
"step": 5125
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 11.839080459770114,
|
| 1570 |
+
"grad_norm": 0.12864351272583008,
|
| 1571 |
+
"learning_rate": 9.070242656449553e-06,
|
| 1572 |
+
"loss": 0.1495,
|
| 1573 |
+
"step": 5150
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 11.89655172413793,
|
| 1577 |
+
"grad_norm": 13.148490905761719,
|
| 1578 |
+
"learning_rate": 9.006385696040869e-06,
|
| 1579 |
+
"loss": 0.2368,
|
| 1580 |
+
"step": 5175
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 11.954022988505747,
|
| 1584 |
+
"grad_norm": 93.63626098632812,
|
| 1585 |
+
"learning_rate": 8.942528735632185e-06,
|
| 1586 |
+
"loss": 0.2774,
|
| 1587 |
+
"step": 5200
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 12.0,
|
| 1591 |
+
"eval_accuracy": 0.9385908209437621,
|
| 1592 |
+
"eval_loss": 0.4001652002334595,
|
| 1593 |
+
"eval_macro_f1": 0.8140652087091239,
|
| 1594 |
+
"eval_runtime": 0.9344,
|
| 1595 |
+
"eval_samples_per_second": 1655.634,
|
| 1596 |
+
"eval_steps_per_second": 103.812,
|
| 1597 |
+
"eval_weighted_f1": 0.9355913868429291,
|
| 1598 |
+
"step": 5220
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 12.011494252873563,
|
| 1602 |
+
"grad_norm": 0.19162984192371368,
|
| 1603 |
+
"learning_rate": 8.878671775223499e-06,
|
| 1604 |
+
"loss": 0.1878,
|
| 1605 |
+
"step": 5225
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 12.068965517241379,
|
| 1609 |
+
"grad_norm": 0.09594583511352539,
|
| 1610 |
+
"learning_rate": 8.814814814814817e-06,
|
| 1611 |
+
"loss": 0.1788,
|
| 1612 |
+
"step": 5250
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 12.126436781609195,
|
| 1616 |
+
"grad_norm": 0.03951456397771835,
|
| 1617 |
+
"learning_rate": 8.75095785440613e-06,
|
| 1618 |
+
"loss": 0.0575,
|
| 1619 |
+
"step": 5275
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 12.183908045977011,
|
| 1623 |
+
"grad_norm": 0.04577677324414253,
|
| 1624 |
+
"learning_rate": 8.687100893997446e-06,
|
| 1625 |
+
"loss": 0.2344,
|
| 1626 |
+
"step": 5300
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 12.241379310344827,
|
| 1630 |
+
"grad_norm": 0.043315041810274124,
|
| 1631 |
+
"learning_rate": 8.623243933588762e-06,
|
| 1632 |
+
"loss": 0.2377,
|
| 1633 |
+
"step": 5325
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 12.298850574712644,
|
| 1637 |
+
"grad_norm": 163.26429748535156,
|
| 1638 |
+
"learning_rate": 8.559386973180076e-06,
|
| 1639 |
+
"loss": 0.2844,
|
| 1640 |
+
"step": 5350
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 12.35632183908046,
|
| 1644 |
+
"grad_norm": 0.05767310783267021,
|
| 1645 |
+
"learning_rate": 8.495530012771392e-06,
|
| 1646 |
+
"loss": 0.2244,
|
| 1647 |
+
"step": 5375
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 12.413793103448276,
|
| 1651 |
+
"grad_norm": 0.04089861363172531,
|
| 1652 |
+
"learning_rate": 8.431673052362708e-06,
|
| 1653 |
+
"loss": 0.0487,
|
| 1654 |
+
"step": 5400
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 12.471264367816092,
|
| 1658 |
+
"grad_norm": 0.06572862714529037,
|
| 1659 |
+
"learning_rate": 8.367816091954024e-06,
|
| 1660 |
+
"loss": 0.2214,
|
| 1661 |
+
"step": 5425
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 12.528735632183908,
|
| 1665 |
+
"grad_norm": 0.03793521225452423,
|
| 1666 |
+
"learning_rate": 8.30395913154534e-06,
|
| 1667 |
+
"loss": 0.0214,
|
| 1668 |
+
"step": 5450
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 12.586206896551724,
|
| 1672 |
+
"grad_norm": 0.05792602151632309,
|
| 1673 |
+
"learning_rate": 8.240102171136654e-06,
|
| 1674 |
+
"loss": 0.1919,
|
| 1675 |
+
"step": 5475
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 12.64367816091954,
|
| 1679 |
+
"grad_norm": 0.048064567148685455,
|
| 1680 |
+
"learning_rate": 8.17624521072797e-06,
|
| 1681 |
+
"loss": 0.1269,
|
| 1682 |
+
"step": 5500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 12.701149425287356,
|
| 1686 |
+
"grad_norm": 0.06517326086759567,
|
| 1687 |
+
"learning_rate": 8.112388250319286e-06,
|
| 1688 |
+
"loss": 0.2849,
|
| 1689 |
+
"step": 5525
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 12.758620689655173,
|
| 1693 |
+
"grad_norm": 0.20683197677135468,
|
| 1694 |
+
"learning_rate": 8.0485312899106e-06,
|
| 1695 |
+
"loss": 0.1888,
|
| 1696 |
+
"step": 5550
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 12.816091954022989,
|
| 1700 |
+
"grad_norm": 0.06538327783346176,
|
| 1701 |
+
"learning_rate": 7.984674329501917e-06,
|
| 1702 |
+
"loss": 0.2246,
|
| 1703 |
+
"step": 5575
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 12.873563218390805,
|
| 1707 |
+
"grad_norm": 0.0690942108631134,
|
| 1708 |
+
"learning_rate": 7.920817369093231e-06,
|
| 1709 |
+
"loss": 0.1015,
|
| 1710 |
+
"step": 5600
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 12.931034482758621,
|
| 1714 |
+
"grad_norm": 0.05659184604883194,
|
| 1715 |
+
"learning_rate": 7.856960408684547e-06,
|
| 1716 |
+
"loss": 0.1402,
|
| 1717 |
+
"step": 5625
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 12.988505747126437,
|
| 1721 |
+
"grad_norm": 2.6654036045074463,
|
| 1722 |
+
"learning_rate": 7.793103448275863e-06,
|
| 1723 |
+
"loss": 0.1527,
|
| 1724 |
+
"step": 5650
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 13.0,
|
| 1728 |
+
"eval_accuracy": 0.9392372333548804,
|
| 1729 |
+
"eval_loss": 0.4133770167827606,
|
| 1730 |
+
"eval_macro_f1": 0.8287596274036139,
|
| 1731 |
+
"eval_runtime": 0.9535,
|
| 1732 |
+
"eval_samples_per_second": 1622.389,
|
| 1733 |
+
"eval_steps_per_second": 101.727,
|
| 1734 |
+
"eval_weighted_f1": 0.9384936776257768,
|
| 1735 |
+
"step": 5655
|
| 1736 |
+
}
|
| 1737 |
+
],
|
| 1738 |
+
"logging_steps": 25,
|
| 1739 |
+
"max_steps": 8700,
|
| 1740 |
+
"num_input_tokens_seen": 0,
|
| 1741 |
+
"num_train_epochs": 20,
|
| 1742 |
+
"save_steps": 500,
|
| 1743 |
+
"stateful_callbacks": {
|
| 1744 |
+
"EarlyStoppingCallback": {
|
| 1745 |
+
"args": {
|
| 1746 |
+
"early_stopping_patience": 3,
|
| 1747 |
+
"early_stopping_threshold": 0.0
|
| 1748 |
+
},
|
| 1749 |
+
"attributes": {
|
| 1750 |
+
"early_stopping_patience_counter": 3
|
| 1751 |
+
}
|
| 1752 |
+
},
|
| 1753 |
+
"TrainerControl": {
|
| 1754 |
+
"args": {
|
| 1755 |
+
"should_epoch_stop": false,
|
| 1756 |
+
"should_evaluate": false,
|
| 1757 |
+
"should_log": false,
|
| 1758 |
+
"should_save": true,
|
| 1759 |
+
"should_training_stop": true
|
| 1760 |
+
},
|
| 1761 |
+
"attributes": {}
|
| 1762 |
+
}
|
| 1763 |
+
},
|
| 1764 |
+
"total_flos": 0.0,
|
| 1765 |
+
"train_batch_size": 8,
|
| 1766 |
+
"trial_name": null,
|
| 1767 |
+
"trial_params": null
|
| 1768 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-2175/trainer_state.json
ADDED
|
@@ -0,0 +1,707 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 870,
|
| 3 |
+
"best_metric": 0.8186573914816089,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2175,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05747126436781609,
|
| 14 |
+
"grad_norm": 337.1722106933594,
|
| 15 |
+
"learning_rate": 5.517241379310345e-07,
|
| 16 |
+
"loss": 4.6079,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11494252873563218,
|
| 21 |
+
"grad_norm": 67.69326782226562,
|
| 22 |
+
"learning_rate": 1.1264367816091955e-06,
|
| 23 |
+
"loss": 4.374,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1724137931034483,
|
| 28 |
+
"grad_norm": 41.478450775146484,
|
| 29 |
+
"learning_rate": 1.7011494252873565e-06,
|
| 30 |
+
"loss": 3.9368,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.22988505747126436,
|
| 35 |
+
"grad_norm": 39.7985725402832,
|
| 36 |
+
"learning_rate": 2.2758620689655173e-06,
|
| 37 |
+
"loss": 3.6466,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.28735632183908044,
|
| 42 |
+
"grad_norm": 41.95061492919922,
|
| 43 |
+
"learning_rate": 2.850574712643678e-06,
|
| 44 |
+
"loss": 3.4921,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3448275862068966,
|
| 49 |
+
"grad_norm": 26.691255569458008,
|
| 50 |
+
"learning_rate": 3.425287356321839e-06,
|
| 51 |
+
"loss": 3.2444,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.40229885057471265,
|
| 56 |
+
"grad_norm": 23.52078628540039,
|
| 57 |
+
"learning_rate": 4.000000000000001e-06,
|
| 58 |
+
"loss": 3.0767,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.45977011494252873,
|
| 63 |
+
"grad_norm": 25.946090698242188,
|
| 64 |
+
"learning_rate": 4.5747126436781615e-06,
|
| 65 |
+
"loss": 2.6167,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5172413793103449,
|
| 70 |
+
"grad_norm": 36.82063674926758,
|
| 71 |
+
"learning_rate": 5.149425287356322e-06,
|
| 72 |
+
"loss": 2.0207,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5747126436781609,
|
| 77 |
+
"grad_norm": 45.13892364501953,
|
| 78 |
+
"learning_rate": 5.724137931034483e-06,
|
| 79 |
+
"loss": 1.746,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.632183908045977,
|
| 84 |
+
"grad_norm": 38.70794677734375,
|
| 85 |
+
"learning_rate": 6.298850574712644e-06,
|
| 86 |
+
"loss": 1.6106,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6896551724137931,
|
| 91 |
+
"grad_norm": 31.3992919921875,
|
| 92 |
+
"learning_rate": 6.873563218390805e-06,
|
| 93 |
+
"loss": 1.4959,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7471264367816092,
|
| 98 |
+
"grad_norm": 58.24619674682617,
|
| 99 |
+
"learning_rate": 7.4482758620689665e-06,
|
| 100 |
+
"loss": 1.2419,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8045977011494253,
|
| 105 |
+
"grad_norm": 37.67776107788086,
|
| 106 |
+
"learning_rate": 8.022988505747126e-06,
|
| 107 |
+
"loss": 1.2536,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8620689655172413,
|
| 112 |
+
"grad_norm": 50.31374740600586,
|
| 113 |
+
"learning_rate": 8.597701149425289e-06,
|
| 114 |
+
"loss": 1.4452,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9195402298850575,
|
| 119 |
+
"grad_norm": 5.2023186683654785,
|
| 120 |
+
"learning_rate": 9.172413793103448e-06,
|
| 121 |
+
"loss": 1.0064,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9770114942528736,
|
| 126 |
+
"grad_norm": 40.5797004699707,
|
| 127 |
+
"learning_rate": 9.74712643678161e-06,
|
| 128 |
+
"loss": 1.3232,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9256625727213963,
|
| 134 |
+
"eval_loss": 0.23390012979507446,
|
| 135 |
+
"eval_macro_f1": 0.7229957877009646,
|
| 136 |
+
"eval_runtime": 0.9423,
|
| 137 |
+
"eval_samples_per_second": 1641.713,
|
| 138 |
+
"eval_steps_per_second": 102.939,
|
| 139 |
+
"eval_weighted_f1": 0.9144162049150534,
|
| 140 |
+
"step": 435
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.0344827586206897,
|
| 144 |
+
"grad_norm": 25.124582290649414,
|
| 145 |
+
"learning_rate": 1.032183908045977e-05,
|
| 146 |
+
"loss": 1.1516,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.0919540229885056,
|
| 151 |
+
"grad_norm": 91.09334564208984,
|
| 152 |
+
"learning_rate": 1.0896551724137932e-05,
|
| 153 |
+
"loss": 1.1513,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.1494252873563218,
|
| 158 |
+
"grad_norm": 16.98522186279297,
|
| 159 |
+
"learning_rate": 1.1471264367816093e-05,
|
| 160 |
+
"loss": 1.1915,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.206896551724138,
|
| 165 |
+
"grad_norm": 58.764469146728516,
|
| 166 |
+
"learning_rate": 1.2045977011494254e-05,
|
| 167 |
+
"loss": 0.906,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.264367816091954,
|
| 172 |
+
"grad_norm": 17.7250919342041,
|
| 173 |
+
"learning_rate": 1.2620689655172415e-05,
|
| 174 |
+
"loss": 1.1088,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.3218390804597702,
|
| 179 |
+
"grad_norm": 13.348613739013672,
|
| 180 |
+
"learning_rate": 1.3195402298850574e-05,
|
| 181 |
+
"loss": 1.185,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3793103448275863,
|
| 186 |
+
"grad_norm": 11.975234031677246,
|
| 187 |
+
"learning_rate": 1.3770114942528736e-05,
|
| 188 |
+
"loss": 1.0208,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4367816091954024,
|
| 193 |
+
"grad_norm": 11.076878547668457,
|
| 194 |
+
"learning_rate": 1.4344827586206897e-05,
|
| 195 |
+
"loss": 1.0298,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4942528735632183,
|
| 200 |
+
"grad_norm": 56.335777282714844,
|
| 201 |
+
"learning_rate": 1.4919540229885058e-05,
|
| 202 |
+
"loss": 1.0796,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.5517241379310345,
|
| 207 |
+
"grad_norm": 20.805221557617188,
|
| 208 |
+
"learning_rate": 1.549425287356322e-05,
|
| 209 |
+
"loss": 0.9171,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6091954022988506,
|
| 214 |
+
"grad_norm": 41.189697265625,
|
| 215 |
+
"learning_rate": 1.6068965517241383e-05,
|
| 216 |
+
"loss": 0.8676,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6666666666666665,
|
| 221 |
+
"grad_norm": 75.09175872802734,
|
| 222 |
+
"learning_rate": 1.664367816091954e-05,
|
| 223 |
+
"loss": 1.0423,
|
| 224 |
+
"step": 725
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7241379310344827,
|
| 228 |
+
"grad_norm": 27.449729919433594,
|
| 229 |
+
"learning_rate": 1.72183908045977e-05,
|
| 230 |
+
"loss": 1.1312,
|
| 231 |
+
"step": 750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.7816091954022988,
|
| 235 |
+
"grad_norm": 40.403297424316406,
|
| 236 |
+
"learning_rate": 1.7793103448275862e-05,
|
| 237 |
+
"loss": 0.9983,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.839080459770115,
|
| 242 |
+
"grad_norm": 8.804283142089844,
|
| 243 |
+
"learning_rate": 1.8367816091954027e-05,
|
| 244 |
+
"loss": 0.9187,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.896551724137931,
|
| 249 |
+
"grad_norm": 53.860774993896484,
|
| 250 |
+
"learning_rate": 1.8942528735632187e-05,
|
| 251 |
+
"loss": 1.099,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.9540229885057472,
|
| 256 |
+
"grad_norm": 25.97994041442871,
|
| 257 |
+
"learning_rate": 1.9517241379310345e-05,
|
| 258 |
+
"loss": 1.1001,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_accuracy": 0.940530058177117,
|
| 264 |
+
"eval_loss": 0.2001289278268814,
|
| 265 |
+
"eval_macro_f1": 0.8186573914816089,
|
| 266 |
+
"eval_runtime": 0.9374,
|
| 267 |
+
"eval_samples_per_second": 1650.273,
|
| 268 |
+
"eval_steps_per_second": 103.475,
|
| 269 |
+
"eval_weighted_f1": 0.9358468283011474,
|
| 270 |
+
"step": 870
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.0114942528735633,
|
| 274 |
+
"grad_norm": 32.21914291381836,
|
| 275 |
+
"learning_rate": 1.9989782886334612e-05,
|
| 276 |
+
"loss": 0.786,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0689655172413794,
|
| 281 |
+
"grad_norm": 44.88948059082031,
|
| 282 |
+
"learning_rate": 1.9925925925925928e-05,
|
| 283 |
+
"loss": 0.7581,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.1264367816091956,
|
| 288 |
+
"grad_norm": 19.019996643066406,
|
| 289 |
+
"learning_rate": 1.9862068965517244e-05,
|
| 290 |
+
"loss": 0.8145,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.1839080459770113,
|
| 295 |
+
"grad_norm": 13.914660453796387,
|
| 296 |
+
"learning_rate": 1.979821200510856e-05,
|
| 297 |
+
"loss": 0.9843,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.2413793103448274,
|
| 302 |
+
"grad_norm": 30.964426040649414,
|
| 303 |
+
"learning_rate": 1.9734355044699872e-05,
|
| 304 |
+
"loss": 0.859,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.2988505747126435,
|
| 309 |
+
"grad_norm": 33.673702239990234,
|
| 310 |
+
"learning_rate": 1.9670498084291188e-05,
|
| 311 |
+
"loss": 0.8565,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.3563218390804597,
|
| 316 |
+
"grad_norm": 17.830869674682617,
|
| 317 |
+
"learning_rate": 1.9606641123882504e-05,
|
| 318 |
+
"loss": 0.9404,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.413793103448276,
|
| 323 |
+
"grad_norm": 50.93587875366211,
|
| 324 |
+
"learning_rate": 1.954278416347382e-05,
|
| 325 |
+
"loss": 0.6671,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.471264367816092,
|
| 330 |
+
"grad_norm": 5.843932151794434,
|
| 331 |
+
"learning_rate": 1.9478927203065136e-05,
|
| 332 |
+
"loss": 0.924,
|
| 333 |
+
"step": 1075
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.528735632183908,
|
| 337 |
+
"grad_norm": 23.064428329467773,
|
| 338 |
+
"learning_rate": 1.941507024265645e-05,
|
| 339 |
+
"loss": 1.0166,
|
| 340 |
+
"step": 1100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.586206896551724,
|
| 344 |
+
"grad_norm": 19.971464157104492,
|
| 345 |
+
"learning_rate": 1.9351213282247767e-05,
|
| 346 |
+
"loss": 0.7844,
|
| 347 |
+
"step": 1125
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.6436781609195403,
|
| 351 |
+
"grad_norm": 42.21748733520508,
|
| 352 |
+
"learning_rate": 1.928735632183908e-05,
|
| 353 |
+
"loss": 1.1053,
|
| 354 |
+
"step": 1150
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.7011494252873565,
|
| 358 |
+
"grad_norm": 45.351375579833984,
|
| 359 |
+
"learning_rate": 1.92234993614304e-05,
|
| 360 |
+
"loss": 0.8594,
|
| 361 |
+
"step": 1175
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.7586206896551726,
|
| 365 |
+
"grad_norm": 30.67836570739746,
|
| 366 |
+
"learning_rate": 1.915964240102171e-05,
|
| 367 |
+
"loss": 0.9351,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.8160919540229887,
|
| 372 |
+
"grad_norm": 27.33807373046875,
|
| 373 |
+
"learning_rate": 1.9095785440613027e-05,
|
| 374 |
+
"loss": 0.6243,
|
| 375 |
+
"step": 1225
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.873563218390805,
|
| 379 |
+
"grad_norm": 23.820077896118164,
|
| 380 |
+
"learning_rate": 1.9031928480204343e-05,
|
| 381 |
+
"loss": 0.6981,
|
| 382 |
+
"step": 1250
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.9310344827586206,
|
| 386 |
+
"grad_norm": 350.10888671875,
|
| 387 |
+
"learning_rate": 1.896807151979566e-05,
|
| 388 |
+
"loss": 0.9111,
|
| 389 |
+
"step": 1275
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.9885057471264367,
|
| 393 |
+
"grad_norm": 36.990665435791016,
|
| 394 |
+
"learning_rate": 1.8904214559386975e-05,
|
| 395 |
+
"loss": 0.8506,
|
| 396 |
+
"step": 1300
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.0,
|
| 400 |
+
"eval_accuracy": 0.940530058177117,
|
| 401 |
+
"eval_loss": 0.20723535120487213,
|
| 402 |
+
"eval_macro_f1": 0.8183496294101141,
|
| 403 |
+
"eval_runtime": 0.9324,
|
| 404 |
+
"eval_samples_per_second": 1659.244,
|
| 405 |
+
"eval_steps_per_second": 104.038,
|
| 406 |
+
"eval_weighted_f1": 0.9367881090602921,
|
| 407 |
+
"step": 1305
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.045977011494253,
|
| 411 |
+
"grad_norm": 23.224864959716797,
|
| 412 |
+
"learning_rate": 1.884035759897829e-05,
|
| 413 |
+
"loss": 0.7263,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.103448275862069,
|
| 418 |
+
"grad_norm": 29.1308536529541,
|
| 419 |
+
"learning_rate": 1.8776500638569606e-05,
|
| 420 |
+
"loss": 0.5455,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.160919540229885,
|
| 425 |
+
"grad_norm": 8.131402015686035,
|
| 426 |
+
"learning_rate": 1.871264367816092e-05,
|
| 427 |
+
"loss": 0.7769,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.218390804597701,
|
| 432 |
+
"grad_norm": 14.32136058807373,
|
| 433 |
+
"learning_rate": 1.8648786717752235e-05,
|
| 434 |
+
"loss": 0.6935,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.2758620689655173,
|
| 439 |
+
"grad_norm": 32.472713470458984,
|
| 440 |
+
"learning_rate": 1.8584929757343554e-05,
|
| 441 |
+
"loss": 0.6545,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.3333333333333335,
|
| 446 |
+
"grad_norm": 62.395416259765625,
|
| 447 |
+
"learning_rate": 1.8521072796934866e-05,
|
| 448 |
+
"loss": 0.8149,
|
| 449 |
+
"step": 1450
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.3908045977011496,
|
| 453 |
+
"grad_norm": 40.000431060791016,
|
| 454 |
+
"learning_rate": 1.8457215836526182e-05,
|
| 455 |
+
"loss": 0.624,
|
| 456 |
+
"step": 1475
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.4482758620689653,
|
| 460 |
+
"grad_norm": 53.911251068115234,
|
| 461 |
+
"learning_rate": 1.8393358876117498e-05,
|
| 462 |
+
"loss": 0.6059,
|
| 463 |
+
"step": 1500
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.5057471264367814,
|
| 467 |
+
"grad_norm": 6.269444465637207,
|
| 468 |
+
"learning_rate": 1.8329501915708814e-05,
|
| 469 |
+
"loss": 0.7348,
|
| 470 |
+
"step": 1525
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.5632183908045976,
|
| 474 |
+
"grad_norm": 19.15593147277832,
|
| 475 |
+
"learning_rate": 1.826564495530013e-05,
|
| 476 |
+
"loss": 0.6886,
|
| 477 |
+
"step": 1550
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.6206896551724137,
|
| 481 |
+
"grad_norm": 22.041454315185547,
|
| 482 |
+
"learning_rate": 1.8201787994891446e-05,
|
| 483 |
+
"loss": 0.7096,
|
| 484 |
+
"step": 1575
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.67816091954023,
|
| 488 |
+
"grad_norm": 6.278102397918701,
|
| 489 |
+
"learning_rate": 1.813793103448276e-05,
|
| 490 |
+
"loss": 0.5676,
|
| 491 |
+
"step": 1600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.735632183908046,
|
| 495 |
+
"grad_norm": 4.127840995788574,
|
| 496 |
+
"learning_rate": 1.8074074074074074e-05,
|
| 497 |
+
"loss": 0.625,
|
| 498 |
+
"step": 1625
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.793103448275862,
|
| 502 |
+
"grad_norm": 16.301708221435547,
|
| 503 |
+
"learning_rate": 1.801021711366539e-05,
|
| 504 |
+
"loss": 0.8405,
|
| 505 |
+
"step": 1650
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.8505747126436782,
|
| 509 |
+
"grad_norm": 13.839818000793457,
|
| 510 |
+
"learning_rate": 1.7946360153256706e-05,
|
| 511 |
+
"loss": 0.7366,
|
| 512 |
+
"step": 1675
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.9080459770114944,
|
| 516 |
+
"grad_norm": 11.068490982055664,
|
| 517 |
+
"learning_rate": 1.788250319284802e-05,
|
| 518 |
+
"loss": 0.6346,
|
| 519 |
+
"step": 1700
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.9655172413793105,
|
| 523 |
+
"grad_norm": 8.15661907196045,
|
| 524 |
+
"learning_rate": 1.7818646232439337e-05,
|
| 525 |
+
"loss": 0.682,
|
| 526 |
+
"step": 1725
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.0,
|
| 530 |
+
"eval_accuracy": 0.9411764705882353,
|
| 531 |
+
"eval_loss": 0.22646477818489075,
|
| 532 |
+
"eval_macro_f1": 0.8129739912640016,
|
| 533 |
+
"eval_runtime": 0.9299,
|
| 534 |
+
"eval_samples_per_second": 1663.66,
|
| 535 |
+
"eval_steps_per_second": 104.315,
|
| 536 |
+
"eval_weighted_f1": 0.9361830533436152,
|
| 537 |
+
"step": 1740
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.022988505747127,
|
| 541 |
+
"grad_norm": 6.274733066558838,
|
| 542 |
+
"learning_rate": 1.7754789272030653e-05,
|
| 543 |
+
"loss": 0.7908,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.080459770114943,
|
| 548 |
+
"grad_norm": 1.9429435729980469,
|
| 549 |
+
"learning_rate": 1.769093231162197e-05,
|
| 550 |
+
"loss": 0.4398,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 4.137931034482759,
|
| 555 |
+
"grad_norm": 119.46365356445312,
|
| 556 |
+
"learning_rate": 1.762707535121328e-05,
|
| 557 |
+
"loss": 0.4924,
|
| 558 |
+
"step": 1800
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 4.195402298850575,
|
| 562 |
+
"grad_norm": 70.17437744140625,
|
| 563 |
+
"learning_rate": 1.75632183908046e-05,
|
| 564 |
+
"loss": 0.6097,
|
| 565 |
+
"step": 1825
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 4.252873563218391,
|
| 569 |
+
"grad_norm": 28.127483367919922,
|
| 570 |
+
"learning_rate": 1.7499361430395913e-05,
|
| 571 |
+
"loss": 0.643,
|
| 572 |
+
"step": 1850
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 4.310344827586207,
|
| 576 |
+
"grad_norm": 18.32999038696289,
|
| 577 |
+
"learning_rate": 1.743550446998723e-05,
|
| 578 |
+
"loss": 0.4264,
|
| 579 |
+
"step": 1875
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 4.3678160919540225,
|
| 583 |
+
"grad_norm": 75.96643829345703,
|
| 584 |
+
"learning_rate": 1.7371647509578545e-05,
|
| 585 |
+
"loss": 0.5842,
|
| 586 |
+
"step": 1900
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.425287356321839,
|
| 590 |
+
"grad_norm": 34.672115325927734,
|
| 591 |
+
"learning_rate": 1.730779054916986e-05,
|
| 592 |
+
"loss": 0.6729,
|
| 593 |
+
"step": 1925
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 4.482758620689655,
|
| 597 |
+
"grad_norm": 49.1175537109375,
|
| 598 |
+
"learning_rate": 1.7243933588761176e-05,
|
| 599 |
+
"loss": 0.6224,
|
| 600 |
+
"step": 1950
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 4.540229885057471,
|
| 604 |
+
"grad_norm": 34.644771575927734,
|
| 605 |
+
"learning_rate": 1.7180076628352492e-05,
|
| 606 |
+
"loss": 0.4916,
|
| 607 |
+
"step": 1975
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 4.597701149425287,
|
| 611 |
+
"grad_norm": 25.47177505493164,
|
| 612 |
+
"learning_rate": 1.7116219667943808e-05,
|
| 613 |
+
"loss": 0.7109,
|
| 614 |
+
"step": 2000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.655172413793103,
|
| 618 |
+
"grad_norm": 19.78028106689453,
|
| 619 |
+
"learning_rate": 1.705236270753512e-05,
|
| 620 |
+
"loss": 0.7024,
|
| 621 |
+
"step": 2025
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 4.712643678160919,
|
| 625 |
+
"grad_norm": 36.67063522338867,
|
| 626 |
+
"learning_rate": 1.6988505747126436e-05,
|
| 627 |
+
"loss": 0.5793,
|
| 628 |
+
"step": 2050
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 4.7701149425287355,
|
| 632 |
+
"grad_norm": 42.03983688354492,
|
| 633 |
+
"learning_rate": 1.6924648786717756e-05,
|
| 634 |
+
"loss": 0.767,
|
| 635 |
+
"step": 2075
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 4.827586206896552,
|
| 639 |
+
"grad_norm": 42.721412658691406,
|
| 640 |
+
"learning_rate": 1.6860791826309068e-05,
|
| 641 |
+
"loss": 0.5105,
|
| 642 |
+
"step": 2100
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 4.885057471264368,
|
| 646 |
+
"grad_norm": 17.916948318481445,
|
| 647 |
+
"learning_rate": 1.6796934865900384e-05,
|
| 648 |
+
"loss": 0.5656,
|
| 649 |
+
"step": 2125
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 4.942528735632184,
|
| 653 |
+
"grad_norm": 48.60200500488281,
|
| 654 |
+
"learning_rate": 1.67330779054917e-05,
|
| 655 |
+
"loss": 0.6482,
|
| 656 |
+
"step": 2150
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.0,
|
| 660 |
+
"grad_norm": 27.99745750427246,
|
| 661 |
+
"learning_rate": 1.6669220945083016e-05,
|
| 662 |
+
"loss": 0.4712,
|
| 663 |
+
"step": 2175
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.0,
|
| 667 |
+
"eval_accuracy": 0.9398836457659987,
|
| 668 |
+
"eval_loss": 0.2528269588947296,
|
| 669 |
+
"eval_macro_f1": 0.8075895386884081,
|
| 670 |
+
"eval_runtime": 0.9529,
|
| 671 |
+
"eval_samples_per_second": 1623.47,
|
| 672 |
+
"eval_steps_per_second": 101.795,
|
| 673 |
+
"eval_weighted_f1": 0.9345993400416372,
|
| 674 |
+
"step": 2175
|
| 675 |
+
}
|
| 676 |
+
],
|
| 677 |
+
"logging_steps": 25,
|
| 678 |
+
"max_steps": 8700,
|
| 679 |
+
"num_input_tokens_seen": 0,
|
| 680 |
+
"num_train_epochs": 20,
|
| 681 |
+
"save_steps": 500,
|
| 682 |
+
"stateful_callbacks": {
|
| 683 |
+
"EarlyStoppingCallback": {
|
| 684 |
+
"args": {
|
| 685 |
+
"early_stopping_patience": 3,
|
| 686 |
+
"early_stopping_threshold": 0.0
|
| 687 |
+
},
|
| 688 |
+
"attributes": {
|
| 689 |
+
"early_stopping_patience_counter": 3
|
| 690 |
+
}
|
| 691 |
+
},
|
| 692 |
+
"TrainerControl": {
|
| 693 |
+
"args": {
|
| 694 |
+
"should_epoch_stop": false,
|
| 695 |
+
"should_evaluate": false,
|
| 696 |
+
"should_log": false,
|
| 697 |
+
"should_save": true,
|
| 698 |
+
"should_training_stop": true
|
| 699 |
+
},
|
| 700 |
+
"attributes": {}
|
| 701 |
+
}
|
| 702 |
+
},
|
| 703 |
+
"total_flos": 0.0,
|
| 704 |
+
"train_batch_size": 8,
|
| 705 |
+
"trial_name": null,
|
| 706 |
+
"trial_params": null
|
| 707 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870/trainer_state.json
ADDED
|
@@ -0,0 +1,303 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 870,
|
| 3 |
+
"best_metric": 0.8186573914816089,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-00-s43-20260827T160912Z-85e5496f/trainer/checkpoint-870",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 870,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05747126436781609,
|
| 14 |
+
"grad_norm": 337.1722106933594,
|
| 15 |
+
"learning_rate": 5.517241379310345e-07,
|
| 16 |
+
"loss": 4.6079,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11494252873563218,
|
| 21 |
+
"grad_norm": 67.69326782226562,
|
| 22 |
+
"learning_rate": 1.1264367816091955e-06,
|
| 23 |
+
"loss": 4.374,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1724137931034483,
|
| 28 |
+
"grad_norm": 41.478450775146484,
|
| 29 |
+
"learning_rate": 1.7011494252873565e-06,
|
| 30 |
+
"loss": 3.9368,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.22988505747126436,
|
| 35 |
+
"grad_norm": 39.7985725402832,
|
| 36 |
+
"learning_rate": 2.2758620689655173e-06,
|
| 37 |
+
"loss": 3.6466,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.28735632183908044,
|
| 42 |
+
"grad_norm": 41.95061492919922,
|
| 43 |
+
"learning_rate": 2.850574712643678e-06,
|
| 44 |
+
"loss": 3.4921,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3448275862068966,
|
| 49 |
+
"grad_norm": 26.691255569458008,
|
| 50 |
+
"learning_rate": 3.425287356321839e-06,
|
| 51 |
+
"loss": 3.2444,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.40229885057471265,
|
| 56 |
+
"grad_norm": 23.52078628540039,
|
| 57 |
+
"learning_rate": 4.000000000000001e-06,
|
| 58 |
+
"loss": 3.0767,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.45977011494252873,
|
| 63 |
+
"grad_norm": 25.946090698242188,
|
| 64 |
+
"learning_rate": 4.5747126436781615e-06,
|
| 65 |
+
"loss": 2.6167,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5172413793103449,
|
| 70 |
+
"grad_norm": 36.82063674926758,
|
| 71 |
+
"learning_rate": 5.149425287356322e-06,
|
| 72 |
+
"loss": 2.0207,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5747126436781609,
|
| 77 |
+
"grad_norm": 45.13892364501953,
|
| 78 |
+
"learning_rate": 5.724137931034483e-06,
|
| 79 |
+
"loss": 1.746,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.632183908045977,
|
| 84 |
+
"grad_norm": 38.70794677734375,
|
| 85 |
+
"learning_rate": 6.298850574712644e-06,
|
| 86 |
+
"loss": 1.6106,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6896551724137931,
|
| 91 |
+
"grad_norm": 31.3992919921875,
|
| 92 |
+
"learning_rate": 6.873563218390805e-06,
|
| 93 |
+
"loss": 1.4959,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7471264367816092,
|
| 98 |
+
"grad_norm": 58.24619674682617,
|
| 99 |
+
"learning_rate": 7.4482758620689665e-06,
|
| 100 |
+
"loss": 1.2419,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8045977011494253,
|
| 105 |
+
"grad_norm": 37.67776107788086,
|
| 106 |
+
"learning_rate": 8.022988505747126e-06,
|
| 107 |
+
"loss": 1.2536,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8620689655172413,
|
| 112 |
+
"grad_norm": 50.31374740600586,
|
| 113 |
+
"learning_rate": 8.597701149425289e-06,
|
| 114 |
+
"loss": 1.4452,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9195402298850575,
|
| 119 |
+
"grad_norm": 5.2023186683654785,
|
| 120 |
+
"learning_rate": 9.172413793103448e-06,
|
| 121 |
+
"loss": 1.0064,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9770114942528736,
|
| 126 |
+
"grad_norm": 40.5797004699707,
|
| 127 |
+
"learning_rate": 9.74712643678161e-06,
|
| 128 |
+
"loss": 1.3232,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9256625727213963,
|
| 134 |
+
"eval_loss": 0.23390012979507446,
|
| 135 |
+
"eval_macro_f1": 0.7229957877009646,
|
| 136 |
+
"eval_runtime": 0.9423,
|
| 137 |
+
"eval_samples_per_second": 1641.713,
|
| 138 |
+
"eval_steps_per_second": 102.939,
|
| 139 |
+
"eval_weighted_f1": 0.9144162049150534,
|
| 140 |
+
"step": 435
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.0344827586206897,
|
| 144 |
+
"grad_norm": 25.124582290649414,
|
| 145 |
+
"learning_rate": 1.032183908045977e-05,
|
| 146 |
+
"loss": 1.1516,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.0919540229885056,
|
| 151 |
+
"grad_norm": 91.09334564208984,
|
| 152 |
+
"learning_rate": 1.0896551724137932e-05,
|
| 153 |
+
"loss": 1.1513,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.1494252873563218,
|
| 158 |
+
"grad_norm": 16.98522186279297,
|
| 159 |
+
"learning_rate": 1.1471264367816093e-05,
|
| 160 |
+
"loss": 1.1915,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.206896551724138,
|
| 165 |
+
"grad_norm": 58.764469146728516,
|
| 166 |
+
"learning_rate": 1.2045977011494254e-05,
|
| 167 |
+
"loss": 0.906,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.264367816091954,
|
| 172 |
+
"grad_norm": 17.7250919342041,
|
| 173 |
+
"learning_rate": 1.2620689655172415e-05,
|
| 174 |
+
"loss": 1.1088,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.3218390804597702,
|
| 179 |
+
"grad_norm": 13.348613739013672,
|
| 180 |
+
"learning_rate": 1.3195402298850574e-05,
|
| 181 |
+
"loss": 1.185,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3793103448275863,
|
| 186 |
+
"grad_norm": 11.975234031677246,
|
| 187 |
+
"learning_rate": 1.3770114942528736e-05,
|
| 188 |
+
"loss": 1.0208,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4367816091954024,
|
| 193 |
+
"grad_norm": 11.076878547668457,
|
| 194 |
+
"learning_rate": 1.4344827586206897e-05,
|
| 195 |
+
"loss": 1.0298,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4942528735632183,
|
| 200 |
+
"grad_norm": 56.335777282714844,
|
| 201 |
+
"learning_rate": 1.4919540229885058e-05,
|
| 202 |
+
"loss": 1.0796,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.5517241379310345,
|
| 207 |
+
"grad_norm": 20.805221557617188,
|
| 208 |
+
"learning_rate": 1.549425287356322e-05,
|
| 209 |
+
"loss": 0.9171,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6091954022988506,
|
| 214 |
+
"grad_norm": 41.189697265625,
|
| 215 |
+
"learning_rate": 1.6068965517241383e-05,
|
| 216 |
+
"loss": 0.8676,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6666666666666665,
|
| 221 |
+
"grad_norm": 75.09175872802734,
|
| 222 |
+
"learning_rate": 1.664367816091954e-05,
|
| 223 |
+
"loss": 1.0423,
|
| 224 |
+
"step": 725
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7241379310344827,
|
| 228 |
+
"grad_norm": 27.449729919433594,
|
| 229 |
+
"learning_rate": 1.72183908045977e-05,
|
| 230 |
+
"loss": 1.1312,
|
| 231 |
+
"step": 750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.7816091954022988,
|
| 235 |
+
"grad_norm": 40.403297424316406,
|
| 236 |
+
"learning_rate": 1.7793103448275862e-05,
|
| 237 |
+
"loss": 0.9983,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.839080459770115,
|
| 242 |
+
"grad_norm": 8.804283142089844,
|
| 243 |
+
"learning_rate": 1.8367816091954027e-05,
|
| 244 |
+
"loss": 0.9187,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.896551724137931,
|
| 249 |
+
"grad_norm": 53.860774993896484,
|
| 250 |
+
"learning_rate": 1.8942528735632187e-05,
|
| 251 |
+
"loss": 1.099,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.9540229885057472,
|
| 256 |
+
"grad_norm": 25.97994041442871,
|
| 257 |
+
"learning_rate": 1.9517241379310345e-05,
|
| 258 |
+
"loss": 1.1001,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_accuracy": 0.940530058177117,
|
| 264 |
+
"eval_loss": 0.2001289278268814,
|
| 265 |
+
"eval_macro_f1": 0.8186573914816089,
|
| 266 |
+
"eval_runtime": 0.9374,
|
| 267 |
+
"eval_samples_per_second": 1650.273,
|
| 268 |
+
"eval_steps_per_second": 103.475,
|
| 269 |
+
"eval_weighted_f1": 0.9358468283011474,
|
| 270 |
+
"step": 870
|
| 271 |
+
}
|
| 272 |
+
],
|
| 273 |
+
"logging_steps": 25,
|
| 274 |
+
"max_steps": 8700,
|
| 275 |
+
"num_input_tokens_seen": 0,
|
| 276 |
+
"num_train_epochs": 20,
|
| 277 |
+
"save_steps": 500,
|
| 278 |
+
"stateful_callbacks": {
|
| 279 |
+
"EarlyStoppingCallback": {
|
| 280 |
+
"args": {
|
| 281 |
+
"early_stopping_patience": 3,
|
| 282 |
+
"early_stopping_threshold": 0.0
|
| 283 |
+
},
|
| 284 |
+
"attributes": {
|
| 285 |
+
"early_stopping_patience_counter": 0
|
| 286 |
+
}
|
| 287 |
+
},
|
| 288 |
+
"TrainerControl": {
|
| 289 |
+
"args": {
|
| 290 |
+
"should_epoch_stop": false,
|
| 291 |
+
"should_evaluate": false,
|
| 292 |
+
"should_log": false,
|
| 293 |
+
"should_save": true,
|
| 294 |
+
"should_training_stop": false
|
| 295 |
+
},
|
| 296 |
+
"attributes": {}
|
| 297 |
+
}
|
| 298 |
+
},
|
| 299 |
+
"total_flos": 0.0,
|
| 300 |
+
"train_batch_size": 8,
|
| 301 |
+
"trial_name": null,
|
| 302 |
+
"trial_params": null
|
| 303 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305/trainer_state.json
ADDED
|
@@ -0,0 +1,440 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1305,
|
| 3 |
+
"best_metric": 0.8347293009015985,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1305,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05747126436781609,
|
| 14 |
+
"grad_norm": 34.55548095703125,
|
| 15 |
+
"learning_rate": 5.517241379310345e-07,
|
| 16 |
+
"loss": 4.3117,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11494252873563218,
|
| 21 |
+
"grad_norm": 18.20611572265625,
|
| 22 |
+
"learning_rate": 1.1264367816091955e-06,
|
| 23 |
+
"loss": 4.0948,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1724137931034483,
|
| 28 |
+
"grad_norm": 18.435766220092773,
|
| 29 |
+
"learning_rate": 1.7011494252873565e-06,
|
| 30 |
+
"loss": 3.7305,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.22988505747126436,
|
| 35 |
+
"grad_norm": 17.956161499023438,
|
| 36 |
+
"learning_rate": 2.2758620689655173e-06,
|
| 37 |
+
"loss": 3.5471,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.28735632183908044,
|
| 42 |
+
"grad_norm": 12.342960357666016,
|
| 43 |
+
"learning_rate": 2.850574712643678e-06,
|
| 44 |
+
"loss": 3.3605,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3448275862068966,
|
| 49 |
+
"grad_norm": 29.274578094482422,
|
| 50 |
+
"learning_rate": 3.425287356321839e-06,
|
| 51 |
+
"loss": 3.3571,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.40229885057471265,
|
| 56 |
+
"grad_norm": 27.83257293701172,
|
| 57 |
+
"learning_rate": 4.000000000000001e-06,
|
| 58 |
+
"loss": 3.17,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.45977011494252873,
|
| 63 |
+
"grad_norm": 28.005407333374023,
|
| 64 |
+
"learning_rate": 4.5747126436781615e-06,
|
| 65 |
+
"loss": 2.8915,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5172413793103449,
|
| 70 |
+
"grad_norm": 51.48032760620117,
|
| 71 |
+
"learning_rate": 5.149425287356322e-06,
|
| 72 |
+
"loss": 2.2248,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5747126436781609,
|
| 77 |
+
"grad_norm": 55.9351806640625,
|
| 78 |
+
"learning_rate": 5.724137931034483e-06,
|
| 79 |
+
"loss": 1.7729,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.632183908045977,
|
| 84 |
+
"grad_norm": 55.01822280883789,
|
| 85 |
+
"learning_rate": 6.298850574712644e-06,
|
| 86 |
+
"loss": 1.5876,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6896551724137931,
|
| 91 |
+
"grad_norm": 31.640888214111328,
|
| 92 |
+
"learning_rate": 6.873563218390805e-06,
|
| 93 |
+
"loss": 1.5514,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7471264367816092,
|
| 98 |
+
"grad_norm": 15.370311737060547,
|
| 99 |
+
"learning_rate": 7.4482758620689665e-06,
|
| 100 |
+
"loss": 1.4601,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8045977011494253,
|
| 105 |
+
"grad_norm": 42.0816535949707,
|
| 106 |
+
"learning_rate": 8.022988505747126e-06,
|
| 107 |
+
"loss": 1.5308,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8620689655172413,
|
| 112 |
+
"grad_norm": 28.904016494750977,
|
| 113 |
+
"learning_rate": 8.597701149425289e-06,
|
| 114 |
+
"loss": 1.4649,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9195402298850575,
|
| 119 |
+
"grad_norm": 66.95686340332031,
|
| 120 |
+
"learning_rate": 9.172413793103448e-06,
|
| 121 |
+
"loss": 1.272,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9770114942528736,
|
| 126 |
+
"grad_norm": 47.044410705566406,
|
| 127 |
+
"learning_rate": 9.74712643678161e-06,
|
| 128 |
+
"loss": 1.2994,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9153199741435035,
|
| 134 |
+
"eval_loss": 0.27683934569358826,
|
| 135 |
+
"eval_macro_f1": 0.6240917316304418,
|
| 136 |
+
"eval_runtime": 0.9525,
|
| 137 |
+
"eval_samples_per_second": 1624.098,
|
| 138 |
+
"eval_steps_per_second": 101.834,
|
| 139 |
+
"eval_weighted_f1": 0.8947070150321287,
|
| 140 |
+
"step": 435
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.0344827586206897,
|
| 144 |
+
"grad_norm": 50.5892219543457,
|
| 145 |
+
"learning_rate": 1.032183908045977e-05,
|
| 146 |
+
"loss": 1.1264,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.0919540229885056,
|
| 151 |
+
"grad_norm": 75.66035461425781,
|
| 152 |
+
"learning_rate": 1.0896551724137932e-05,
|
| 153 |
+
"loss": 1.059,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.1494252873563218,
|
| 158 |
+
"grad_norm": 44.54843521118164,
|
| 159 |
+
"learning_rate": 1.1471264367816093e-05,
|
| 160 |
+
"loss": 0.8961,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.206896551724138,
|
| 165 |
+
"grad_norm": 64.8303451538086,
|
| 166 |
+
"learning_rate": 1.2045977011494254e-05,
|
| 167 |
+
"loss": 1.2017,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.264367816091954,
|
| 172 |
+
"grad_norm": 38.89005661010742,
|
| 173 |
+
"learning_rate": 1.2620689655172415e-05,
|
| 174 |
+
"loss": 1.1438,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.3218390804597702,
|
| 179 |
+
"grad_norm": 59.12514877319336,
|
| 180 |
+
"learning_rate": 1.3195402298850574e-05,
|
| 181 |
+
"loss": 0.9477,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3793103448275863,
|
| 186 |
+
"grad_norm": 12.272050857543945,
|
| 187 |
+
"learning_rate": 1.3770114942528736e-05,
|
| 188 |
+
"loss": 1.0689,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4367816091954024,
|
| 193 |
+
"grad_norm": 33.95408630371094,
|
| 194 |
+
"learning_rate": 1.4344827586206897e-05,
|
| 195 |
+
"loss": 1.0294,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4942528735632183,
|
| 200 |
+
"grad_norm": 16.79743003845215,
|
| 201 |
+
"learning_rate": 1.4919540229885058e-05,
|
| 202 |
+
"loss": 1.0753,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.5517241379310345,
|
| 207 |
+
"grad_norm": 44.943397521972656,
|
| 208 |
+
"learning_rate": 1.549425287356322e-05,
|
| 209 |
+
"loss": 1.0537,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6091954022988506,
|
| 214 |
+
"grad_norm": 17.969419479370117,
|
| 215 |
+
"learning_rate": 1.6068965517241383e-05,
|
| 216 |
+
"loss": 1.1684,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6666666666666665,
|
| 221 |
+
"grad_norm": 26.772727966308594,
|
| 222 |
+
"learning_rate": 1.664367816091954e-05,
|
| 223 |
+
"loss": 1.0037,
|
| 224 |
+
"step": 725
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7241379310344827,
|
| 228 |
+
"grad_norm": 15.204939842224121,
|
| 229 |
+
"learning_rate": 1.72183908045977e-05,
|
| 230 |
+
"loss": 0.962,
|
| 231 |
+
"step": 750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.7816091954022988,
|
| 235 |
+
"grad_norm": 16.9458065032959,
|
| 236 |
+
"learning_rate": 1.7793103448275862e-05,
|
| 237 |
+
"loss": 1.0536,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.839080459770115,
|
| 242 |
+
"grad_norm": 16.120819091796875,
|
| 243 |
+
"learning_rate": 1.8367816091954027e-05,
|
| 244 |
+
"loss": 0.8744,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.896551724137931,
|
| 249 |
+
"grad_norm": 21.2254581451416,
|
| 250 |
+
"learning_rate": 1.8942528735632187e-05,
|
| 251 |
+
"loss": 0.9102,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.9540229885057472,
|
| 256 |
+
"grad_norm": 24.79235076904297,
|
| 257 |
+
"learning_rate": 1.9517241379310345e-05,
|
| 258 |
+
"loss": 1.0437,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_accuracy": 0.9372979961215255,
|
| 264 |
+
"eval_loss": 0.19322560727596283,
|
| 265 |
+
"eval_macro_f1": 0.8105083400344721,
|
| 266 |
+
"eval_runtime": 0.9618,
|
| 267 |
+
"eval_samples_per_second": 1608.507,
|
| 268 |
+
"eval_steps_per_second": 100.857,
|
| 269 |
+
"eval_weighted_f1": 0.9340113059139937,
|
| 270 |
+
"step": 870
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.0114942528735633,
|
| 274 |
+
"grad_norm": 44.39748001098633,
|
| 275 |
+
"learning_rate": 1.9989782886334612e-05,
|
| 276 |
+
"loss": 1.0047,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0689655172413794,
|
| 281 |
+
"grad_norm": 6.5618109703063965,
|
| 282 |
+
"learning_rate": 1.9925925925925928e-05,
|
| 283 |
+
"loss": 0.7388,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.1264367816091956,
|
| 288 |
+
"grad_norm": 57.590335845947266,
|
| 289 |
+
"learning_rate": 1.9862068965517244e-05,
|
| 290 |
+
"loss": 0.7571,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.1839080459770113,
|
| 295 |
+
"grad_norm": 15.785080909729004,
|
| 296 |
+
"learning_rate": 1.979821200510856e-05,
|
| 297 |
+
"loss": 1.0507,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.2413793103448274,
|
| 302 |
+
"grad_norm": 52.038204193115234,
|
| 303 |
+
"learning_rate": 1.9734355044699872e-05,
|
| 304 |
+
"loss": 0.7791,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.2988505747126435,
|
| 309 |
+
"grad_norm": 12.01987361907959,
|
| 310 |
+
"learning_rate": 1.9670498084291188e-05,
|
| 311 |
+
"loss": 0.8555,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.3563218390804597,
|
| 316 |
+
"grad_norm": 50.869869232177734,
|
| 317 |
+
"learning_rate": 1.9606641123882504e-05,
|
| 318 |
+
"loss": 0.8696,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.413793103448276,
|
| 323 |
+
"grad_norm": 40.58417510986328,
|
| 324 |
+
"learning_rate": 1.954278416347382e-05,
|
| 325 |
+
"loss": 0.8047,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.471264367816092,
|
| 330 |
+
"grad_norm": 40.40437698364258,
|
| 331 |
+
"learning_rate": 1.9478927203065136e-05,
|
| 332 |
+
"loss": 0.9368,
|
| 333 |
+
"step": 1075
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.528735632183908,
|
| 337 |
+
"grad_norm": 23.701034545898438,
|
| 338 |
+
"learning_rate": 1.941507024265645e-05,
|
| 339 |
+
"loss": 0.772,
|
| 340 |
+
"step": 1100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.586206896551724,
|
| 344 |
+
"grad_norm": 32.53304672241211,
|
| 345 |
+
"learning_rate": 1.9351213282247767e-05,
|
| 346 |
+
"loss": 0.7373,
|
| 347 |
+
"step": 1125
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.6436781609195403,
|
| 351 |
+
"grad_norm": 18.946821212768555,
|
| 352 |
+
"learning_rate": 1.928735632183908e-05,
|
| 353 |
+
"loss": 0.9495,
|
| 354 |
+
"step": 1150
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.7011494252873565,
|
| 358 |
+
"grad_norm": 62.86573028564453,
|
| 359 |
+
"learning_rate": 1.92234993614304e-05,
|
| 360 |
+
"loss": 0.7644,
|
| 361 |
+
"step": 1175
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.7586206896551726,
|
| 365 |
+
"grad_norm": 8.561220169067383,
|
| 366 |
+
"learning_rate": 1.915964240102171e-05,
|
| 367 |
+
"loss": 0.7611,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.8160919540229887,
|
| 372 |
+
"grad_norm": 31.526269912719727,
|
| 373 |
+
"learning_rate": 1.9095785440613027e-05,
|
| 374 |
+
"loss": 0.9984,
|
| 375 |
+
"step": 1225
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.873563218390805,
|
| 379 |
+
"grad_norm": 57.629737854003906,
|
| 380 |
+
"learning_rate": 1.9031928480204343e-05,
|
| 381 |
+
"loss": 0.7771,
|
| 382 |
+
"step": 1250
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.9310344827586206,
|
| 386 |
+
"grad_norm": 6.244096279144287,
|
| 387 |
+
"learning_rate": 1.896807151979566e-05,
|
| 388 |
+
"loss": 0.7165,
|
| 389 |
+
"step": 1275
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.9885057471264367,
|
| 393 |
+
"grad_norm": 24.01633071899414,
|
| 394 |
+
"learning_rate": 1.8904214559386975e-05,
|
| 395 |
+
"loss": 1.0622,
|
| 396 |
+
"step": 1300
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.0,
|
| 400 |
+
"eval_accuracy": 0.9431157078215902,
|
| 401 |
+
"eval_loss": 0.2161780446767807,
|
| 402 |
+
"eval_macro_f1": 0.8347293009015985,
|
| 403 |
+
"eval_runtime": 0.9243,
|
| 404 |
+
"eval_samples_per_second": 1673.638,
|
| 405 |
+
"eval_steps_per_second": 104.94,
|
| 406 |
+
"eval_weighted_f1": 0.9403893208783903,
|
| 407 |
+
"step": 1305
|
| 408 |
+
}
|
| 409 |
+
],
|
| 410 |
+
"logging_steps": 25,
|
| 411 |
+
"max_steps": 8700,
|
| 412 |
+
"num_input_tokens_seen": 0,
|
| 413 |
+
"num_train_epochs": 20,
|
| 414 |
+
"save_steps": 500,
|
| 415 |
+
"stateful_callbacks": {
|
| 416 |
+
"EarlyStoppingCallback": {
|
| 417 |
+
"args": {
|
| 418 |
+
"early_stopping_patience": 3,
|
| 419 |
+
"early_stopping_threshold": 0.0
|
| 420 |
+
},
|
| 421 |
+
"attributes": {
|
| 422 |
+
"early_stopping_patience_counter": 0
|
| 423 |
+
}
|
| 424 |
+
},
|
| 425 |
+
"TrainerControl": {
|
| 426 |
+
"args": {
|
| 427 |
+
"should_epoch_stop": false,
|
| 428 |
+
"should_evaluate": false,
|
| 429 |
+
"should_log": false,
|
| 430 |
+
"should_save": true,
|
| 431 |
+
"should_training_stop": false
|
| 432 |
+
},
|
| 433 |
+
"attributes": {}
|
| 434 |
+
}
|
| 435 |
+
},
|
| 436 |
+
"total_flos": 0.0,
|
| 437 |
+
"train_batch_size": 8,
|
| 438 |
+
"trial_name": null,
|
| 439 |
+
"trial_params": null
|
| 440 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-2610/trainer_state.json
ADDED
|
@@ -0,0 +1,837 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1305,
|
| 3 |
+
"best_metric": 0.8347293009015985,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-00-s44-20260827T163014Z-b6645c77/trainer/checkpoint-1305",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2610,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05747126436781609,
|
| 14 |
+
"grad_norm": 34.55548095703125,
|
| 15 |
+
"learning_rate": 5.517241379310345e-07,
|
| 16 |
+
"loss": 4.3117,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11494252873563218,
|
| 21 |
+
"grad_norm": 18.20611572265625,
|
| 22 |
+
"learning_rate": 1.1264367816091955e-06,
|
| 23 |
+
"loss": 4.0948,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1724137931034483,
|
| 28 |
+
"grad_norm": 18.435766220092773,
|
| 29 |
+
"learning_rate": 1.7011494252873565e-06,
|
| 30 |
+
"loss": 3.7305,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.22988505747126436,
|
| 35 |
+
"grad_norm": 17.956161499023438,
|
| 36 |
+
"learning_rate": 2.2758620689655173e-06,
|
| 37 |
+
"loss": 3.5471,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.28735632183908044,
|
| 42 |
+
"grad_norm": 12.342960357666016,
|
| 43 |
+
"learning_rate": 2.850574712643678e-06,
|
| 44 |
+
"loss": 3.3605,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3448275862068966,
|
| 49 |
+
"grad_norm": 29.274578094482422,
|
| 50 |
+
"learning_rate": 3.425287356321839e-06,
|
| 51 |
+
"loss": 3.3571,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.40229885057471265,
|
| 56 |
+
"grad_norm": 27.83257293701172,
|
| 57 |
+
"learning_rate": 4.000000000000001e-06,
|
| 58 |
+
"loss": 3.17,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.45977011494252873,
|
| 63 |
+
"grad_norm": 28.005407333374023,
|
| 64 |
+
"learning_rate": 4.5747126436781615e-06,
|
| 65 |
+
"loss": 2.8915,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5172413793103449,
|
| 70 |
+
"grad_norm": 51.48032760620117,
|
| 71 |
+
"learning_rate": 5.149425287356322e-06,
|
| 72 |
+
"loss": 2.2248,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5747126436781609,
|
| 77 |
+
"grad_norm": 55.9351806640625,
|
| 78 |
+
"learning_rate": 5.724137931034483e-06,
|
| 79 |
+
"loss": 1.7729,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.632183908045977,
|
| 84 |
+
"grad_norm": 55.01822280883789,
|
| 85 |
+
"learning_rate": 6.298850574712644e-06,
|
| 86 |
+
"loss": 1.5876,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6896551724137931,
|
| 91 |
+
"grad_norm": 31.640888214111328,
|
| 92 |
+
"learning_rate": 6.873563218390805e-06,
|
| 93 |
+
"loss": 1.5514,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.7471264367816092,
|
| 98 |
+
"grad_norm": 15.370311737060547,
|
| 99 |
+
"learning_rate": 7.4482758620689665e-06,
|
| 100 |
+
"loss": 1.4601,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.8045977011494253,
|
| 105 |
+
"grad_norm": 42.0816535949707,
|
| 106 |
+
"learning_rate": 8.022988505747126e-06,
|
| 107 |
+
"loss": 1.5308,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8620689655172413,
|
| 112 |
+
"grad_norm": 28.904016494750977,
|
| 113 |
+
"learning_rate": 8.597701149425289e-06,
|
| 114 |
+
"loss": 1.4649,
|
| 115 |
+
"step": 375
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.9195402298850575,
|
| 119 |
+
"grad_norm": 66.95686340332031,
|
| 120 |
+
"learning_rate": 9.172413793103448e-06,
|
| 121 |
+
"loss": 1.272,
|
| 122 |
+
"step": 400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9770114942528736,
|
| 126 |
+
"grad_norm": 47.044410705566406,
|
| 127 |
+
"learning_rate": 9.74712643678161e-06,
|
| 128 |
+
"loss": 1.2994,
|
| 129 |
+
"step": 425
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.9153199741435035,
|
| 134 |
+
"eval_loss": 0.27683934569358826,
|
| 135 |
+
"eval_macro_f1": 0.6240917316304418,
|
| 136 |
+
"eval_runtime": 0.9525,
|
| 137 |
+
"eval_samples_per_second": 1624.098,
|
| 138 |
+
"eval_steps_per_second": 101.834,
|
| 139 |
+
"eval_weighted_f1": 0.8947070150321287,
|
| 140 |
+
"step": 435
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.0344827586206897,
|
| 144 |
+
"grad_norm": 50.5892219543457,
|
| 145 |
+
"learning_rate": 1.032183908045977e-05,
|
| 146 |
+
"loss": 1.1264,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.0919540229885056,
|
| 151 |
+
"grad_norm": 75.66035461425781,
|
| 152 |
+
"learning_rate": 1.0896551724137932e-05,
|
| 153 |
+
"loss": 1.059,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.1494252873563218,
|
| 158 |
+
"grad_norm": 44.54843521118164,
|
| 159 |
+
"learning_rate": 1.1471264367816093e-05,
|
| 160 |
+
"loss": 0.8961,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.206896551724138,
|
| 165 |
+
"grad_norm": 64.8303451538086,
|
| 166 |
+
"learning_rate": 1.2045977011494254e-05,
|
| 167 |
+
"loss": 1.2017,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.264367816091954,
|
| 172 |
+
"grad_norm": 38.89005661010742,
|
| 173 |
+
"learning_rate": 1.2620689655172415e-05,
|
| 174 |
+
"loss": 1.1438,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.3218390804597702,
|
| 179 |
+
"grad_norm": 59.12514877319336,
|
| 180 |
+
"learning_rate": 1.3195402298850574e-05,
|
| 181 |
+
"loss": 0.9477,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3793103448275863,
|
| 186 |
+
"grad_norm": 12.272050857543945,
|
| 187 |
+
"learning_rate": 1.3770114942528736e-05,
|
| 188 |
+
"loss": 1.0689,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4367816091954024,
|
| 193 |
+
"grad_norm": 33.95408630371094,
|
| 194 |
+
"learning_rate": 1.4344827586206897e-05,
|
| 195 |
+
"loss": 1.0294,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.4942528735632183,
|
| 200 |
+
"grad_norm": 16.79743003845215,
|
| 201 |
+
"learning_rate": 1.4919540229885058e-05,
|
| 202 |
+
"loss": 1.0753,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.5517241379310345,
|
| 207 |
+
"grad_norm": 44.943397521972656,
|
| 208 |
+
"learning_rate": 1.549425287356322e-05,
|
| 209 |
+
"loss": 1.0537,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.6091954022988506,
|
| 214 |
+
"grad_norm": 17.969419479370117,
|
| 215 |
+
"learning_rate": 1.6068965517241383e-05,
|
| 216 |
+
"loss": 1.1684,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.6666666666666665,
|
| 221 |
+
"grad_norm": 26.772727966308594,
|
| 222 |
+
"learning_rate": 1.664367816091954e-05,
|
| 223 |
+
"loss": 1.0037,
|
| 224 |
+
"step": 725
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7241379310344827,
|
| 228 |
+
"grad_norm": 15.204939842224121,
|
| 229 |
+
"learning_rate": 1.72183908045977e-05,
|
| 230 |
+
"loss": 0.962,
|
| 231 |
+
"step": 750
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.7816091954022988,
|
| 235 |
+
"grad_norm": 16.9458065032959,
|
| 236 |
+
"learning_rate": 1.7793103448275862e-05,
|
| 237 |
+
"loss": 1.0536,
|
| 238 |
+
"step": 775
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.839080459770115,
|
| 242 |
+
"grad_norm": 16.120819091796875,
|
| 243 |
+
"learning_rate": 1.8367816091954027e-05,
|
| 244 |
+
"loss": 0.8744,
|
| 245 |
+
"step": 800
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.896551724137931,
|
| 249 |
+
"grad_norm": 21.2254581451416,
|
| 250 |
+
"learning_rate": 1.8942528735632187e-05,
|
| 251 |
+
"loss": 0.9102,
|
| 252 |
+
"step": 825
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.9540229885057472,
|
| 256 |
+
"grad_norm": 24.79235076904297,
|
| 257 |
+
"learning_rate": 1.9517241379310345e-05,
|
| 258 |
+
"loss": 1.0437,
|
| 259 |
+
"step": 850
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_accuracy": 0.9372979961215255,
|
| 264 |
+
"eval_loss": 0.19322560727596283,
|
| 265 |
+
"eval_macro_f1": 0.8105083400344721,
|
| 266 |
+
"eval_runtime": 0.9618,
|
| 267 |
+
"eval_samples_per_second": 1608.507,
|
| 268 |
+
"eval_steps_per_second": 100.857,
|
| 269 |
+
"eval_weighted_f1": 0.9340113059139937,
|
| 270 |
+
"step": 870
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.0114942528735633,
|
| 274 |
+
"grad_norm": 44.39748001098633,
|
| 275 |
+
"learning_rate": 1.9989782886334612e-05,
|
| 276 |
+
"loss": 1.0047,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.0689655172413794,
|
| 281 |
+
"grad_norm": 6.5618109703063965,
|
| 282 |
+
"learning_rate": 1.9925925925925928e-05,
|
| 283 |
+
"loss": 0.7388,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.1264367816091956,
|
| 288 |
+
"grad_norm": 57.590335845947266,
|
| 289 |
+
"learning_rate": 1.9862068965517244e-05,
|
| 290 |
+
"loss": 0.7571,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.1839080459770113,
|
| 295 |
+
"grad_norm": 15.785080909729004,
|
| 296 |
+
"learning_rate": 1.979821200510856e-05,
|
| 297 |
+
"loss": 1.0507,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.2413793103448274,
|
| 302 |
+
"grad_norm": 52.038204193115234,
|
| 303 |
+
"learning_rate": 1.9734355044699872e-05,
|
| 304 |
+
"loss": 0.7791,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.2988505747126435,
|
| 309 |
+
"grad_norm": 12.01987361907959,
|
| 310 |
+
"learning_rate": 1.9670498084291188e-05,
|
| 311 |
+
"loss": 0.8555,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.3563218390804597,
|
| 316 |
+
"grad_norm": 50.869869232177734,
|
| 317 |
+
"learning_rate": 1.9606641123882504e-05,
|
| 318 |
+
"loss": 0.8696,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.413793103448276,
|
| 323 |
+
"grad_norm": 40.58417510986328,
|
| 324 |
+
"learning_rate": 1.954278416347382e-05,
|
| 325 |
+
"loss": 0.8047,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.471264367816092,
|
| 330 |
+
"grad_norm": 40.40437698364258,
|
| 331 |
+
"learning_rate": 1.9478927203065136e-05,
|
| 332 |
+
"loss": 0.9368,
|
| 333 |
+
"step": 1075
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.528735632183908,
|
| 337 |
+
"grad_norm": 23.701034545898438,
|
| 338 |
+
"learning_rate": 1.941507024265645e-05,
|
| 339 |
+
"loss": 0.772,
|
| 340 |
+
"step": 1100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.586206896551724,
|
| 344 |
+
"grad_norm": 32.53304672241211,
|
| 345 |
+
"learning_rate": 1.9351213282247767e-05,
|
| 346 |
+
"loss": 0.7373,
|
| 347 |
+
"step": 1125
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.6436781609195403,
|
| 351 |
+
"grad_norm": 18.946821212768555,
|
| 352 |
+
"learning_rate": 1.928735632183908e-05,
|
| 353 |
+
"loss": 0.9495,
|
| 354 |
+
"step": 1150
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.7011494252873565,
|
| 358 |
+
"grad_norm": 62.86573028564453,
|
| 359 |
+
"learning_rate": 1.92234993614304e-05,
|
| 360 |
+
"loss": 0.7644,
|
| 361 |
+
"step": 1175
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.7586206896551726,
|
| 365 |
+
"grad_norm": 8.561220169067383,
|
| 366 |
+
"learning_rate": 1.915964240102171e-05,
|
| 367 |
+
"loss": 0.7611,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.8160919540229887,
|
| 372 |
+
"grad_norm": 31.526269912719727,
|
| 373 |
+
"learning_rate": 1.9095785440613027e-05,
|
| 374 |
+
"loss": 0.9984,
|
| 375 |
+
"step": 1225
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.873563218390805,
|
| 379 |
+
"grad_norm": 57.629737854003906,
|
| 380 |
+
"learning_rate": 1.9031928480204343e-05,
|
| 381 |
+
"loss": 0.7771,
|
| 382 |
+
"step": 1250
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.9310344827586206,
|
| 386 |
+
"grad_norm": 6.244096279144287,
|
| 387 |
+
"learning_rate": 1.896807151979566e-05,
|
| 388 |
+
"loss": 0.7165,
|
| 389 |
+
"step": 1275
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.9885057471264367,
|
| 393 |
+
"grad_norm": 24.01633071899414,
|
| 394 |
+
"learning_rate": 1.8904214559386975e-05,
|
| 395 |
+
"loss": 1.0622,
|
| 396 |
+
"step": 1300
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.0,
|
| 400 |
+
"eval_accuracy": 0.9431157078215902,
|
| 401 |
+
"eval_loss": 0.2161780446767807,
|
| 402 |
+
"eval_macro_f1": 0.8347293009015985,
|
| 403 |
+
"eval_runtime": 0.9243,
|
| 404 |
+
"eval_samples_per_second": 1673.638,
|
| 405 |
+
"eval_steps_per_second": 104.94,
|
| 406 |
+
"eval_weighted_f1": 0.9403893208783903,
|
| 407 |
+
"step": 1305
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.045977011494253,
|
| 411 |
+
"grad_norm": 2.608572244644165,
|
| 412 |
+
"learning_rate": 1.884035759897829e-05,
|
| 413 |
+
"loss": 0.7171,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.103448275862069,
|
| 418 |
+
"grad_norm": 48.258338928222656,
|
| 419 |
+
"learning_rate": 1.8776500638569606e-05,
|
| 420 |
+
"loss": 0.651,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.160919540229885,
|
| 425 |
+
"grad_norm": 21.637271881103516,
|
| 426 |
+
"learning_rate": 1.871264367816092e-05,
|
| 427 |
+
"loss": 0.9389,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.218390804597701,
|
| 432 |
+
"grad_norm": 51.734230041503906,
|
| 433 |
+
"learning_rate": 1.8648786717752235e-05,
|
| 434 |
+
"loss": 0.6224,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.2758620689655173,
|
| 439 |
+
"grad_norm": 88.5042724609375,
|
| 440 |
+
"learning_rate": 1.8584929757343554e-05,
|
| 441 |
+
"loss": 0.6618,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 3.3333333333333335,
|
| 446 |
+
"grad_norm": 37.068416595458984,
|
| 447 |
+
"learning_rate": 1.8521072796934866e-05,
|
| 448 |
+
"loss": 0.6822,
|
| 449 |
+
"step": 1450
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.3908045977011496,
|
| 453 |
+
"grad_norm": 26.91570472717285,
|
| 454 |
+
"learning_rate": 1.8457215836526182e-05,
|
| 455 |
+
"loss": 0.607,
|
| 456 |
+
"step": 1475
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.4482758620689653,
|
| 460 |
+
"grad_norm": 21.589069366455078,
|
| 461 |
+
"learning_rate": 1.8393358876117498e-05,
|
| 462 |
+
"loss": 0.7733,
|
| 463 |
+
"step": 1500
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 3.5057471264367814,
|
| 467 |
+
"grad_norm": 28.250329971313477,
|
| 468 |
+
"learning_rate": 1.8329501915708814e-05,
|
| 469 |
+
"loss": 0.602,
|
| 470 |
+
"step": 1525
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.5632183908045976,
|
| 474 |
+
"grad_norm": 48.20625305175781,
|
| 475 |
+
"learning_rate": 1.826564495530013e-05,
|
| 476 |
+
"loss": 0.8699,
|
| 477 |
+
"step": 1550
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.6206896551724137,
|
| 481 |
+
"grad_norm": 97.2187271118164,
|
| 482 |
+
"learning_rate": 1.8201787994891446e-05,
|
| 483 |
+
"loss": 0.8086,
|
| 484 |
+
"step": 1575
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.67816091954023,
|
| 488 |
+
"grad_norm": 15.602619171142578,
|
| 489 |
+
"learning_rate": 1.813793103448276e-05,
|
| 490 |
+
"loss": 0.8259,
|
| 491 |
+
"step": 1600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.735632183908046,
|
| 495 |
+
"grad_norm": 12.088094711303711,
|
| 496 |
+
"learning_rate": 1.8074074074074074e-05,
|
| 497 |
+
"loss": 0.6784,
|
| 498 |
+
"step": 1625
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.793103448275862,
|
| 502 |
+
"grad_norm": 17.619722366333008,
|
| 503 |
+
"learning_rate": 1.801021711366539e-05,
|
| 504 |
+
"loss": 0.8504,
|
| 505 |
+
"step": 1650
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.8505747126436782,
|
| 509 |
+
"grad_norm": 34.27809143066406,
|
| 510 |
+
"learning_rate": 1.7946360153256706e-05,
|
| 511 |
+
"loss": 0.646,
|
| 512 |
+
"step": 1675
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.9080459770114944,
|
| 516 |
+
"grad_norm": 11.615314483642578,
|
| 517 |
+
"learning_rate": 1.788250319284802e-05,
|
| 518 |
+
"loss": 0.8702,
|
| 519 |
+
"step": 1700
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.9655172413793105,
|
| 523 |
+
"grad_norm": 41.039283752441406,
|
| 524 |
+
"learning_rate": 1.7818646232439337e-05,
|
| 525 |
+
"loss": 0.7134,
|
| 526 |
+
"step": 1725
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.0,
|
| 530 |
+
"eval_accuracy": 0.9385908209437621,
|
| 531 |
+
"eval_loss": 0.2351500689983368,
|
| 532 |
+
"eval_macro_f1": 0.7913105304394721,
|
| 533 |
+
"eval_runtime": 0.9213,
|
| 534 |
+
"eval_samples_per_second": 1679.119,
|
| 535 |
+
"eval_steps_per_second": 105.284,
|
| 536 |
+
"eval_weighted_f1": 0.931656139221909,
|
| 537 |
+
"step": 1740
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.022988505747127,
|
| 541 |
+
"grad_norm": 2.42942214012146,
|
| 542 |
+
"learning_rate": 1.7754789272030653e-05,
|
| 543 |
+
"loss": 0.6156,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.080459770114943,
|
| 548 |
+
"grad_norm": 20.093666076660156,
|
| 549 |
+
"learning_rate": 1.769093231162197e-05,
|
| 550 |
+
"loss": 0.4679,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 4.137931034482759,
|
| 555 |
+
"grad_norm": 55.2880744934082,
|
| 556 |
+
"learning_rate": 1.762707535121328e-05,
|
| 557 |
+
"loss": 0.3482,
|
| 558 |
+
"step": 1800
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 4.195402298850575,
|
| 562 |
+
"grad_norm": 19.47980499267578,
|
| 563 |
+
"learning_rate": 1.75632183908046e-05,
|
| 564 |
+
"loss": 0.7355,
|
| 565 |
+
"step": 1825
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 4.252873563218391,
|
| 569 |
+
"grad_norm": 60.517208099365234,
|
| 570 |
+
"learning_rate": 1.7499361430395913e-05,
|
| 571 |
+
"loss": 0.5104,
|
| 572 |
+
"step": 1850
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 4.310344827586207,
|
| 576 |
+
"grad_norm": 3.793043851852417,
|
| 577 |
+
"learning_rate": 1.743550446998723e-05,
|
| 578 |
+
"loss": 0.5477,
|
| 579 |
+
"step": 1875
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 4.3678160919540225,
|
| 583 |
+
"grad_norm": 10.501693725585938,
|
| 584 |
+
"learning_rate": 1.7371647509578545e-05,
|
| 585 |
+
"loss": 0.893,
|
| 586 |
+
"step": 1900
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.425287356321839,
|
| 590 |
+
"grad_norm": 12.592524528503418,
|
| 591 |
+
"learning_rate": 1.730779054916986e-05,
|
| 592 |
+
"loss": 0.526,
|
| 593 |
+
"step": 1925
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 4.482758620689655,
|
| 597 |
+
"grad_norm": 102.22213745117188,
|
| 598 |
+
"learning_rate": 1.7243933588761176e-05,
|
| 599 |
+
"loss": 0.729,
|
| 600 |
+
"step": 1950
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 4.540229885057471,
|
| 604 |
+
"grad_norm": 82.27591705322266,
|
| 605 |
+
"learning_rate": 1.7180076628352492e-05,
|
| 606 |
+
"loss": 0.5303,
|
| 607 |
+
"step": 1975
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 4.597701149425287,
|
| 611 |
+
"grad_norm": 8.354979515075684,
|
| 612 |
+
"learning_rate": 1.7116219667943808e-05,
|
| 613 |
+
"loss": 0.5196,
|
| 614 |
+
"step": 2000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.655172413793103,
|
| 618 |
+
"grad_norm": 2.7990167140960693,
|
| 619 |
+
"learning_rate": 1.705236270753512e-05,
|
| 620 |
+
"loss": 0.6641,
|
| 621 |
+
"step": 2025
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 4.712643678160919,
|
| 625 |
+
"grad_norm": 18.160293579101562,
|
| 626 |
+
"learning_rate": 1.6988505747126436e-05,
|
| 627 |
+
"loss": 0.4141,
|
| 628 |
+
"step": 2050
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 4.7701149425287355,
|
| 632 |
+
"grad_norm": 9.09523868560791,
|
| 633 |
+
"learning_rate": 1.6924648786717756e-05,
|
| 634 |
+
"loss": 0.6617,
|
| 635 |
+
"step": 2075
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 4.827586206896552,
|
| 639 |
+
"grad_norm": 36.97121810913086,
|
| 640 |
+
"learning_rate": 1.6860791826309068e-05,
|
| 641 |
+
"loss": 0.7262,
|
| 642 |
+
"step": 2100
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 4.885057471264368,
|
| 646 |
+
"grad_norm": 41.245147705078125,
|
| 647 |
+
"learning_rate": 1.6796934865900384e-05,
|
| 648 |
+
"loss": 0.6243,
|
| 649 |
+
"step": 2125
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 4.942528735632184,
|
| 653 |
+
"grad_norm": 62.95018005371094,
|
| 654 |
+
"learning_rate": 1.67330779054917e-05,
|
| 655 |
+
"loss": 0.6,
|
| 656 |
+
"step": 2150
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.0,
|
| 660 |
+
"grad_norm": 43.248779296875,
|
| 661 |
+
"learning_rate": 1.6669220945083016e-05,
|
| 662 |
+
"loss": 0.7969,
|
| 663 |
+
"step": 2175
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.0,
|
| 667 |
+
"eval_accuracy": 0.9379444085326438,
|
| 668 |
+
"eval_loss": 0.23701989650726318,
|
| 669 |
+
"eval_macro_f1": 0.8274026602802876,
|
| 670 |
+
"eval_runtime": 0.966,
|
| 671 |
+
"eval_samples_per_second": 1601.507,
|
| 672 |
+
"eval_steps_per_second": 100.418,
|
| 673 |
+
"eval_weighted_f1": 0.9363602396254669,
|
| 674 |
+
"step": 2175
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 5.057471264367816,
|
| 678 |
+
"grad_norm": 76.33407592773438,
|
| 679 |
+
"learning_rate": 1.6605363984674328e-05,
|
| 680 |
+
"loss": 0.4662,
|
| 681 |
+
"step": 2200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 5.114942528735632,
|
| 685 |
+
"grad_norm": 61.55803298950195,
|
| 686 |
+
"learning_rate": 1.6541507024265647e-05,
|
| 687 |
+
"loss": 0.4666,
|
| 688 |
+
"step": 2225
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 5.172413793103448,
|
| 692 |
+
"grad_norm": 0.5578628778457642,
|
| 693 |
+
"learning_rate": 1.6477650063856963e-05,
|
| 694 |
+
"loss": 0.2904,
|
| 695 |
+
"step": 2250
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 5.2298850574712645,
|
| 699 |
+
"grad_norm": 6.503316402435303,
|
| 700 |
+
"learning_rate": 1.6413793103448276e-05,
|
| 701 |
+
"loss": 0.5668,
|
| 702 |
+
"step": 2275
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.287356321839081,
|
| 706 |
+
"grad_norm": 5.802130699157715,
|
| 707 |
+
"learning_rate": 1.634993614303959e-05,
|
| 708 |
+
"loss": 0.7934,
|
| 709 |
+
"step": 2300
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 5.344827586206897,
|
| 713 |
+
"grad_norm": 27.219497680664062,
|
| 714 |
+
"learning_rate": 1.6286079182630907e-05,
|
| 715 |
+
"loss": 0.5566,
|
| 716 |
+
"step": 2325
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 5.402298850574713,
|
| 720 |
+
"grad_norm": 3.6371712684631348,
|
| 721 |
+
"learning_rate": 1.6222222222222223e-05,
|
| 722 |
+
"loss": 0.523,
|
| 723 |
+
"step": 2350
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 5.459770114942529,
|
| 727 |
+
"grad_norm": 31.591482162475586,
|
| 728 |
+
"learning_rate": 1.615836526181354e-05,
|
| 729 |
+
"loss": 0.7941,
|
| 730 |
+
"step": 2375
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 5.517241379310345,
|
| 734 |
+
"grad_norm": 0.7491744756698608,
|
| 735 |
+
"learning_rate": 1.6094508301404855e-05,
|
| 736 |
+
"loss": 0.4109,
|
| 737 |
+
"step": 2400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 5.574712643678161,
|
| 741 |
+
"grad_norm": 34.23863220214844,
|
| 742 |
+
"learning_rate": 1.603065134099617e-05,
|
| 743 |
+
"loss": 0.7335,
|
| 744 |
+
"step": 2425
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.6321839080459775,
|
| 748 |
+
"grad_norm": 32.29622268676758,
|
| 749 |
+
"learning_rate": 1.5966794380587483e-05,
|
| 750 |
+
"loss": 0.4088,
|
| 751 |
+
"step": 2450
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 5.689655172413794,
|
| 755 |
+
"grad_norm": 39.42219924926758,
|
| 756 |
+
"learning_rate": 1.5902937420178802e-05,
|
| 757 |
+
"loss": 0.6092,
|
| 758 |
+
"step": 2475
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.747126436781609,
|
| 762 |
+
"grad_norm": 46.9281120300293,
|
| 763 |
+
"learning_rate": 1.5839080459770115e-05,
|
| 764 |
+
"loss": 0.5459,
|
| 765 |
+
"step": 2500
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 5.804597701149425,
|
| 769 |
+
"grad_norm": 1.7171404361724854,
|
| 770 |
+
"learning_rate": 1.577522349936143e-05,
|
| 771 |
+
"loss": 0.5467,
|
| 772 |
+
"step": 2525
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 5.862068965517241,
|
| 776 |
+
"grad_norm": 36.909358978271484,
|
| 777 |
+
"learning_rate": 1.5711366538952746e-05,
|
| 778 |
+
"loss": 0.6461,
|
| 779 |
+
"step": 2550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 5.919540229885057,
|
| 783 |
+
"grad_norm": 28.31206512451172,
|
| 784 |
+
"learning_rate": 1.5647509578544062e-05,
|
| 785 |
+
"loss": 0.5061,
|
| 786 |
+
"step": 2575
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 5.977011494252873,
|
| 790 |
+
"grad_norm": 184.04754638671875,
|
| 791 |
+
"learning_rate": 1.5583652618135378e-05,
|
| 792 |
+
"loss": 0.6764,
|
| 793 |
+
"step": 2600
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 6.0,
|
| 797 |
+
"eval_accuracy": 0.9398836457659987,
|
| 798 |
+
"eval_loss": 0.2670440375804901,
|
| 799 |
+
"eval_macro_f1": 0.825393379499114,
|
| 800 |
+
"eval_runtime": 0.934,
|
| 801 |
+
"eval_samples_per_second": 1656.39,
|
| 802 |
+
"eval_steps_per_second": 103.859,
|
| 803 |
+
"eval_weighted_f1": 0.9376347938341673,
|
| 804 |
+
"step": 2610
|
| 805 |
+
}
|
| 806 |
+
],
|
| 807 |
+
"logging_steps": 25,
|
| 808 |
+
"max_steps": 8700,
|
| 809 |
+
"num_input_tokens_seen": 0,
|
| 810 |
+
"num_train_epochs": 20,
|
| 811 |
+
"save_steps": 500,
|
| 812 |
+
"stateful_callbacks": {
|
| 813 |
+
"EarlyStoppingCallback": {
|
| 814 |
+
"args": {
|
| 815 |
+
"early_stopping_patience": 3,
|
| 816 |
+
"early_stopping_threshold": 0.0
|
| 817 |
+
},
|
| 818 |
+
"attributes": {
|
| 819 |
+
"early_stopping_patience_counter": 3
|
| 820 |
+
}
|
| 821 |
+
},
|
| 822 |
+
"TrainerControl": {
|
| 823 |
+
"args": {
|
| 824 |
+
"should_epoch_stop": false,
|
| 825 |
+
"should_evaluate": false,
|
| 826 |
+
"should_log": false,
|
| 827 |
+
"should_save": true,
|
| 828 |
+
"should_training_stop": true
|
| 829 |
+
},
|
| 830 |
+
"attributes": {}
|
| 831 |
+
}
|
| 832 |
+
},
|
| 833 |
+
"total_flos": 0.0,
|
| 834 |
+
"train_batch_size": 8,
|
| 835 |
+
"trial_name": null,
|
| 836 |
+
"trial_params": null
|
| 837 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161/trainer_state.json
ADDED
|
@@ -0,0 +1,464 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1161,
|
| 3 |
+
"best_metric": 0.8933721587205007,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161",
|
| 5 |
+
"epoch": 9.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1161,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.19455252918287938,
|
| 14 |
+
"grad_norm": 43.42185592651367,
|
| 15 |
+
"learning_rate": 1.86046511627907e-06,
|
| 16 |
+
"loss": 4.0365,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.38910505836575876,
|
| 21 |
+
"grad_norm": 33.00918960571289,
|
| 22 |
+
"learning_rate": 3.798449612403101e-06,
|
| 23 |
+
"loss": 3.2662,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5836575875486382,
|
| 28 |
+
"grad_norm": 31.25638771057129,
|
| 29 |
+
"learning_rate": 5.736434108527133e-06,
|
| 30 |
+
"loss": 2.9813,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7782101167315175,
|
| 35 |
+
"grad_norm": 24.589290618896484,
|
| 36 |
+
"learning_rate": 7.674418604651164e-06,
|
| 37 |
+
"loss": 3.0457,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.9727626459143969,
|
| 42 |
+
"grad_norm": 19.884830474853516,
|
| 43 |
+
"learning_rate": 9.612403100775196e-06,
|
| 44 |
+
"loss": 2.9141,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.0,
|
| 49 |
+
"eval_accuracy": 0.7286652078774617,
|
| 50 |
+
"eval_loss": 0.6383342742919922,
|
| 51 |
+
"eval_macro_f1": 0.2810126582278481,
|
| 52 |
+
"eval_runtime": 0.3548,
|
| 53 |
+
"eval_samples_per_second": 1288.048,
|
| 54 |
+
"eval_steps_per_second": 81.736,
|
| 55 |
+
"eval_weighted_f1": 0.6142924410713793,
|
| 56 |
+
"step": 129
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.1634241245136188,
|
| 60 |
+
"grad_norm": 23.574623107910156,
|
| 61 |
+
"learning_rate": 1.1550387596899227e-05,
|
| 62 |
+
"loss": 2.565,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.3579766536964981,
|
| 67 |
+
"grad_norm": 82.8171615600586,
|
| 68 |
+
"learning_rate": 1.3488372093023257e-05,
|
| 69 |
+
"loss": 2.1606,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.5525291828793775,
|
| 74 |
+
"grad_norm": 47.75974655151367,
|
| 75 |
+
"learning_rate": 1.542635658914729e-05,
|
| 76 |
+
"loss": 2.1759,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.7470817120622568,
|
| 81 |
+
"grad_norm": 47.37407684326172,
|
| 82 |
+
"learning_rate": 1.736434108527132e-05,
|
| 83 |
+
"loss": 1.5946,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.9416342412451362,
|
| 88 |
+
"grad_norm": 65.9006576538086,
|
| 89 |
+
"learning_rate": 1.9302325581395353e-05,
|
| 90 |
+
"loss": 1.5821,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_accuracy": 0.9015317286652079,
|
| 96 |
+
"eval_loss": 0.30102577805519104,
|
| 97 |
+
"eval_macro_f1": 0.7960300418501056,
|
| 98 |
+
"eval_runtime": 0.3466,
|
| 99 |
+
"eval_samples_per_second": 1318.439,
|
| 100 |
+
"eval_steps_per_second": 83.665,
|
| 101 |
+
"eval_weighted_f1": 0.8938363580523228,
|
| 102 |
+
"step": 258
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.132295719844358,
|
| 106 |
+
"grad_norm": 45.2689323425293,
|
| 107 |
+
"learning_rate": 1.9862187769164516e-05,
|
| 108 |
+
"loss": 1.3294,
|
| 109 |
+
"step": 275
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3268482490272375,
|
| 113 |
+
"grad_norm": 27.64436912536621,
|
| 114 |
+
"learning_rate": 1.9646856158484065e-05,
|
| 115 |
+
"loss": 1.2311,
|
| 116 |
+
"step": 300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.5214007782101167,
|
| 120 |
+
"grad_norm": 41.63043212890625,
|
| 121 |
+
"learning_rate": 1.943152454780362e-05,
|
| 122 |
+
"loss": 1.1992,
|
| 123 |
+
"step": 325
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.7159533073929962,
|
| 127 |
+
"grad_norm": 22.198444366455078,
|
| 128 |
+
"learning_rate": 1.921619293712317e-05,
|
| 129 |
+
"loss": 1.1916,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.9105058365758754,
|
| 134 |
+
"grad_norm": 46.360687255859375,
|
| 135 |
+
"learning_rate": 1.9000861326442723e-05,
|
| 136 |
+
"loss": 1.3336,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.0,
|
| 141 |
+
"eval_accuracy": 0.9212253829321663,
|
| 142 |
+
"eval_loss": 0.2455613911151886,
|
| 143 |
+
"eval_macro_f1": 0.853989433678092,
|
| 144 |
+
"eval_runtime": 0.3383,
|
| 145 |
+
"eval_samples_per_second": 1350.935,
|
| 146 |
+
"eval_steps_per_second": 85.727,
|
| 147 |
+
"eval_weighted_f1": 0.9189891547063483,
|
| 148 |
+
"step": 387
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.1011673151750974,
|
| 152 |
+
"grad_norm": 41.747093200683594,
|
| 153 |
+
"learning_rate": 1.8785529715762276e-05,
|
| 154 |
+
"loss": 1.0626,
|
| 155 |
+
"step": 400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.2957198443579765,
|
| 159 |
+
"grad_norm": 26.631877899169922,
|
| 160 |
+
"learning_rate": 1.857019810508183e-05,
|
| 161 |
+
"loss": 1.1279,
|
| 162 |
+
"step": 425
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.490272373540856,
|
| 166 |
+
"grad_norm": 54.40272521972656,
|
| 167 |
+
"learning_rate": 1.835486649440138e-05,
|
| 168 |
+
"loss": 1.1985,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.6848249027237356,
|
| 173 |
+
"grad_norm": 15.547079086303711,
|
| 174 |
+
"learning_rate": 1.813953488372093e-05,
|
| 175 |
+
"loss": 1.0705,
|
| 176 |
+
"step": 475
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.8793774319066148,
|
| 180 |
+
"grad_norm": 59.61214065551758,
|
| 181 |
+
"learning_rate": 1.7924203273040483e-05,
|
| 182 |
+
"loss": 0.8552,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.912472647702407,
|
| 188 |
+
"eval_loss": 0.30328890681266785,
|
| 189 |
+
"eval_macro_f1": 0.8322500321677185,
|
| 190 |
+
"eval_runtime": 0.3621,
|
| 191 |
+
"eval_samples_per_second": 1261.98,
|
| 192 |
+
"eval_steps_per_second": 80.082,
|
| 193 |
+
"eval_weighted_f1": 0.9078573030179498,
|
| 194 |
+
"step": 516
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.070038910505836,
|
| 198 |
+
"grad_norm": 21.746002197265625,
|
| 199 |
+
"learning_rate": 1.7708871662360035e-05,
|
| 200 |
+
"loss": 0.8159,
|
| 201 |
+
"step": 525
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 4.264591439688716,
|
| 205 |
+
"grad_norm": 25.9829158782959,
|
| 206 |
+
"learning_rate": 1.7493540051679588e-05,
|
| 207 |
+
"loss": 0.8029,
|
| 208 |
+
"step": 550
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 4.4591439688715955,
|
| 212 |
+
"grad_norm": 18.974702835083008,
|
| 213 |
+
"learning_rate": 1.727820844099914e-05,
|
| 214 |
+
"loss": 0.7237,
|
| 215 |
+
"step": 575
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.653696498054475,
|
| 219 |
+
"grad_norm": 8.793298721313477,
|
| 220 |
+
"learning_rate": 1.7062876830318693e-05,
|
| 221 |
+
"loss": 0.7558,
|
| 222 |
+
"step": 600
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.848249027237354,
|
| 226 |
+
"grad_norm": 23.602251052856445,
|
| 227 |
+
"learning_rate": 1.6847545219638243e-05,
|
| 228 |
+
"loss": 0.9415,
|
| 229 |
+
"step": 625
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.9168490153172867,
|
| 234 |
+
"eval_loss": 0.28598931431770325,
|
| 235 |
+
"eval_macro_f1": 0.8564511542784734,
|
| 236 |
+
"eval_runtime": 0.3704,
|
| 237 |
+
"eval_samples_per_second": 1233.742,
|
| 238 |
+
"eval_steps_per_second": 78.29,
|
| 239 |
+
"eval_weighted_f1": 0.9161291817810625,
|
| 240 |
+
"step": 645
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.038910505836576,
|
| 244 |
+
"grad_norm": 10.088534355163574,
|
| 245 |
+
"learning_rate": 1.6632213608957795e-05,
|
| 246 |
+
"loss": 0.8484,
|
| 247 |
+
"step": 650
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.233463035019455,
|
| 251 |
+
"grad_norm": 54.11484146118164,
|
| 252 |
+
"learning_rate": 1.6416881998277348e-05,
|
| 253 |
+
"loss": 0.7538,
|
| 254 |
+
"step": 675
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.428015564202335,
|
| 258 |
+
"grad_norm": 50.45320510864258,
|
| 259 |
+
"learning_rate": 1.62015503875969e-05,
|
| 260 |
+
"loss": 0.4758,
|
| 261 |
+
"step": 700
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.622568093385214,
|
| 265 |
+
"grad_norm": 61.879600524902344,
|
| 266 |
+
"learning_rate": 1.5986218776916453e-05,
|
| 267 |
+
"loss": 0.8378,
|
| 268 |
+
"step": 725
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.817120622568093,
|
| 272 |
+
"grad_norm": 16.213138580322266,
|
| 273 |
+
"learning_rate": 1.5770887166236006e-05,
|
| 274 |
+
"loss": 0.6686,
|
| 275 |
+
"step": 750
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.0,
|
| 279 |
+
"eval_accuracy": 0.9212253829321663,
|
| 280 |
+
"eval_loss": 0.31217917799949646,
|
| 281 |
+
"eval_macro_f1": 0.8698320583838917,
|
| 282 |
+
"eval_runtime": 0.3513,
|
| 283 |
+
"eval_samples_per_second": 1300.726,
|
| 284 |
+
"eval_steps_per_second": 82.541,
|
| 285 |
+
"eval_weighted_f1": 0.9204734003776074,
|
| 286 |
+
"step": 774
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 6.007782101167315,
|
| 290 |
+
"grad_norm": 11.760616302490234,
|
| 291 |
+
"learning_rate": 1.555555555555556e-05,
|
| 292 |
+
"loss": 0.4805,
|
| 293 |
+
"step": 775
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 6.202334630350195,
|
| 297 |
+
"grad_norm": 9.390809059143066,
|
| 298 |
+
"learning_rate": 1.534022394487511e-05,
|
| 299 |
+
"loss": 0.3366,
|
| 300 |
+
"step": 800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 6.396887159533074,
|
| 304 |
+
"grad_norm": 0.6686726212501526,
|
| 305 |
+
"learning_rate": 1.512489233419466e-05,
|
| 306 |
+
"loss": 0.4272,
|
| 307 |
+
"step": 825
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 6.591439688715953,
|
| 311 |
+
"grad_norm": 57.247215270996094,
|
| 312 |
+
"learning_rate": 1.4909560723514215e-05,
|
| 313 |
+
"loss": 0.6792,
|
| 314 |
+
"step": 850
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 6.785992217898833,
|
| 318 |
+
"grad_norm": 28.846025466918945,
|
| 319 |
+
"learning_rate": 1.4694229112833764e-05,
|
| 320 |
+
"loss": 0.5075,
|
| 321 |
+
"step": 875
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.980544747081712,
|
| 325 |
+
"grad_norm": 30.43777847290039,
|
| 326 |
+
"learning_rate": 1.4478897502153318e-05,
|
| 327 |
+
"loss": 0.6462,
|
| 328 |
+
"step": 900
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 7.0,
|
| 332 |
+
"eval_accuracy": 0.9234135667396062,
|
| 333 |
+
"eval_loss": 0.3408062160015106,
|
| 334 |
+
"eval_macro_f1": 0.8754667450288961,
|
| 335 |
+
"eval_runtime": 0.3487,
|
| 336 |
+
"eval_samples_per_second": 1310.444,
|
| 337 |
+
"eval_steps_per_second": 83.157,
|
| 338 |
+
"eval_weighted_f1": 0.9242855737407655,
|
| 339 |
+
"step": 903
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 7.171206225680934,
|
| 343 |
+
"grad_norm": 44.301124572753906,
|
| 344 |
+
"learning_rate": 1.426356589147287e-05,
|
| 345 |
+
"loss": 0.4255,
|
| 346 |
+
"step": 925
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.365758754863813,
|
| 350 |
+
"grad_norm": 38.106292724609375,
|
| 351 |
+
"learning_rate": 1.4048234280792422e-05,
|
| 352 |
+
"loss": 0.4863,
|
| 353 |
+
"step": 950
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 7.560311284046692,
|
| 357 |
+
"grad_norm": 13.320198059082031,
|
| 358 |
+
"learning_rate": 1.3832902670111974e-05,
|
| 359 |
+
"loss": 0.3598,
|
| 360 |
+
"step": 975
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 7.754863813229572,
|
| 364 |
+
"grad_norm": 0.39185717701911926,
|
| 365 |
+
"learning_rate": 1.3617571059431525e-05,
|
| 366 |
+
"loss": 0.3488,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.9494163424124515,
|
| 371 |
+
"grad_norm": 65.87716674804688,
|
| 372 |
+
"learning_rate": 1.3402239448751078e-05,
|
| 373 |
+
"loss": 0.6857,
|
| 374 |
+
"step": 1025
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.0,
|
| 378 |
+
"eval_accuracy": 0.9234135667396062,
|
| 379 |
+
"eval_loss": 0.3978811800479889,
|
| 380 |
+
"eval_macro_f1": 0.8724444593347033,
|
| 381 |
+
"eval_runtime": 0.3536,
|
| 382 |
+
"eval_samples_per_second": 1292.375,
|
| 383 |
+
"eval_steps_per_second": 82.011,
|
| 384 |
+
"eval_weighted_f1": 0.9225742180023816,
|
| 385 |
+
"step": 1032
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 8.140077821011673,
|
| 389 |
+
"grad_norm": 50.72336196899414,
|
| 390 |
+
"learning_rate": 1.318690783807063e-05,
|
| 391 |
+
"loss": 0.3311,
|
| 392 |
+
"step": 1050
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 8.334630350194553,
|
| 396 |
+
"grad_norm": 29.121295928955078,
|
| 397 |
+
"learning_rate": 1.2971576227390181e-05,
|
| 398 |
+
"loss": 0.4558,
|
| 399 |
+
"step": 1075
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 8.529182879377432,
|
| 403 |
+
"grad_norm": 29.663639068603516,
|
| 404 |
+
"learning_rate": 1.2756244616709734e-05,
|
| 405 |
+
"loss": 0.3734,
|
| 406 |
+
"step": 1100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 8.72373540856031,
|
| 410 |
+
"grad_norm": 95.52642059326172,
|
| 411 |
+
"learning_rate": 1.2540913006029287e-05,
|
| 412 |
+
"loss": 0.3917,
|
| 413 |
+
"step": 1125
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 8.918287937743191,
|
| 417 |
+
"grad_norm": 0.6678557991981506,
|
| 418 |
+
"learning_rate": 1.2325581395348838e-05,
|
| 419 |
+
"loss": 0.2513,
|
| 420 |
+
"step": 1150
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 9.0,
|
| 424 |
+
"eval_accuracy": 0.9343544857768052,
|
| 425 |
+
"eval_loss": 0.353714257478714,
|
| 426 |
+
"eval_macro_f1": 0.8933721587205007,
|
| 427 |
+
"eval_runtime": 0.3625,
|
| 428 |
+
"eval_samples_per_second": 1260.797,
|
| 429 |
+
"eval_steps_per_second": 80.007,
|
| 430 |
+
"eval_weighted_f1": 0.9342032177875788,
|
| 431 |
+
"step": 1161
|
| 432 |
+
}
|
| 433 |
+
],
|
| 434 |
+
"logging_steps": 25,
|
| 435 |
+
"max_steps": 2580,
|
| 436 |
+
"num_input_tokens_seen": 0,
|
| 437 |
+
"num_train_epochs": 20,
|
| 438 |
+
"save_steps": 500,
|
| 439 |
+
"stateful_callbacks": {
|
| 440 |
+
"EarlyStoppingCallback": {
|
| 441 |
+
"args": {
|
| 442 |
+
"early_stopping_patience": 3,
|
| 443 |
+
"early_stopping_threshold": 0.0
|
| 444 |
+
},
|
| 445 |
+
"attributes": {
|
| 446 |
+
"early_stopping_patience_counter": 0
|
| 447 |
+
}
|
| 448 |
+
},
|
| 449 |
+
"TrainerControl": {
|
| 450 |
+
"args": {
|
| 451 |
+
"should_epoch_stop": false,
|
| 452 |
+
"should_evaluate": false,
|
| 453 |
+
"should_log": false,
|
| 454 |
+
"should_save": true,
|
| 455 |
+
"should_training_stop": false
|
| 456 |
+
},
|
| 457 |
+
"attributes": {}
|
| 458 |
+
}
|
| 459 |
+
},
|
| 460 |
+
"total_flos": 0.0,
|
| 461 |
+
"train_batch_size": 8,
|
| 462 |
+
"trial_name": null,
|
| 463 |
+
"trial_params": null
|
| 464 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1548/trainer_state.json
ADDED
|
@@ -0,0 +1,602 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1161,
|
| 3 |
+
"best_metric": 0.8933721587205007,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-01-s42-20260827T155348Z-f97f2036/trainer/checkpoint-1161",
|
| 5 |
+
"epoch": 12.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1548,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.19455252918287938,
|
| 14 |
+
"grad_norm": 43.42185592651367,
|
| 15 |
+
"learning_rate": 1.86046511627907e-06,
|
| 16 |
+
"loss": 4.0365,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.38910505836575876,
|
| 21 |
+
"grad_norm": 33.00918960571289,
|
| 22 |
+
"learning_rate": 3.798449612403101e-06,
|
| 23 |
+
"loss": 3.2662,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5836575875486382,
|
| 28 |
+
"grad_norm": 31.25638771057129,
|
| 29 |
+
"learning_rate": 5.736434108527133e-06,
|
| 30 |
+
"loss": 2.9813,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7782101167315175,
|
| 35 |
+
"grad_norm": 24.589290618896484,
|
| 36 |
+
"learning_rate": 7.674418604651164e-06,
|
| 37 |
+
"loss": 3.0457,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.9727626459143969,
|
| 42 |
+
"grad_norm": 19.884830474853516,
|
| 43 |
+
"learning_rate": 9.612403100775196e-06,
|
| 44 |
+
"loss": 2.9141,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.0,
|
| 49 |
+
"eval_accuracy": 0.7286652078774617,
|
| 50 |
+
"eval_loss": 0.6383342742919922,
|
| 51 |
+
"eval_macro_f1": 0.2810126582278481,
|
| 52 |
+
"eval_runtime": 0.3548,
|
| 53 |
+
"eval_samples_per_second": 1288.048,
|
| 54 |
+
"eval_steps_per_second": 81.736,
|
| 55 |
+
"eval_weighted_f1": 0.6142924410713793,
|
| 56 |
+
"step": 129
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.1634241245136188,
|
| 60 |
+
"grad_norm": 23.574623107910156,
|
| 61 |
+
"learning_rate": 1.1550387596899227e-05,
|
| 62 |
+
"loss": 2.565,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.3579766536964981,
|
| 67 |
+
"grad_norm": 82.8171615600586,
|
| 68 |
+
"learning_rate": 1.3488372093023257e-05,
|
| 69 |
+
"loss": 2.1606,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.5525291828793775,
|
| 74 |
+
"grad_norm": 47.75974655151367,
|
| 75 |
+
"learning_rate": 1.542635658914729e-05,
|
| 76 |
+
"loss": 2.1759,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.7470817120622568,
|
| 81 |
+
"grad_norm": 47.37407684326172,
|
| 82 |
+
"learning_rate": 1.736434108527132e-05,
|
| 83 |
+
"loss": 1.5946,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.9416342412451362,
|
| 88 |
+
"grad_norm": 65.9006576538086,
|
| 89 |
+
"learning_rate": 1.9302325581395353e-05,
|
| 90 |
+
"loss": 1.5821,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_accuracy": 0.9015317286652079,
|
| 96 |
+
"eval_loss": 0.30102577805519104,
|
| 97 |
+
"eval_macro_f1": 0.7960300418501056,
|
| 98 |
+
"eval_runtime": 0.3466,
|
| 99 |
+
"eval_samples_per_second": 1318.439,
|
| 100 |
+
"eval_steps_per_second": 83.665,
|
| 101 |
+
"eval_weighted_f1": 0.8938363580523228,
|
| 102 |
+
"step": 258
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.132295719844358,
|
| 106 |
+
"grad_norm": 45.2689323425293,
|
| 107 |
+
"learning_rate": 1.9862187769164516e-05,
|
| 108 |
+
"loss": 1.3294,
|
| 109 |
+
"step": 275
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3268482490272375,
|
| 113 |
+
"grad_norm": 27.64436912536621,
|
| 114 |
+
"learning_rate": 1.9646856158484065e-05,
|
| 115 |
+
"loss": 1.2311,
|
| 116 |
+
"step": 300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.5214007782101167,
|
| 120 |
+
"grad_norm": 41.63043212890625,
|
| 121 |
+
"learning_rate": 1.943152454780362e-05,
|
| 122 |
+
"loss": 1.1992,
|
| 123 |
+
"step": 325
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.7159533073929962,
|
| 127 |
+
"grad_norm": 22.198444366455078,
|
| 128 |
+
"learning_rate": 1.921619293712317e-05,
|
| 129 |
+
"loss": 1.1916,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.9105058365758754,
|
| 134 |
+
"grad_norm": 46.360687255859375,
|
| 135 |
+
"learning_rate": 1.9000861326442723e-05,
|
| 136 |
+
"loss": 1.3336,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.0,
|
| 141 |
+
"eval_accuracy": 0.9212253829321663,
|
| 142 |
+
"eval_loss": 0.2455613911151886,
|
| 143 |
+
"eval_macro_f1": 0.853989433678092,
|
| 144 |
+
"eval_runtime": 0.3383,
|
| 145 |
+
"eval_samples_per_second": 1350.935,
|
| 146 |
+
"eval_steps_per_second": 85.727,
|
| 147 |
+
"eval_weighted_f1": 0.9189891547063483,
|
| 148 |
+
"step": 387
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.1011673151750974,
|
| 152 |
+
"grad_norm": 41.747093200683594,
|
| 153 |
+
"learning_rate": 1.8785529715762276e-05,
|
| 154 |
+
"loss": 1.0626,
|
| 155 |
+
"step": 400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.2957198443579765,
|
| 159 |
+
"grad_norm": 26.631877899169922,
|
| 160 |
+
"learning_rate": 1.857019810508183e-05,
|
| 161 |
+
"loss": 1.1279,
|
| 162 |
+
"step": 425
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.490272373540856,
|
| 166 |
+
"grad_norm": 54.40272521972656,
|
| 167 |
+
"learning_rate": 1.835486649440138e-05,
|
| 168 |
+
"loss": 1.1985,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.6848249027237356,
|
| 173 |
+
"grad_norm": 15.547079086303711,
|
| 174 |
+
"learning_rate": 1.813953488372093e-05,
|
| 175 |
+
"loss": 1.0705,
|
| 176 |
+
"step": 475
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.8793774319066148,
|
| 180 |
+
"grad_norm": 59.61214065551758,
|
| 181 |
+
"learning_rate": 1.7924203273040483e-05,
|
| 182 |
+
"loss": 0.8552,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.912472647702407,
|
| 188 |
+
"eval_loss": 0.30328890681266785,
|
| 189 |
+
"eval_macro_f1": 0.8322500321677185,
|
| 190 |
+
"eval_runtime": 0.3621,
|
| 191 |
+
"eval_samples_per_second": 1261.98,
|
| 192 |
+
"eval_steps_per_second": 80.082,
|
| 193 |
+
"eval_weighted_f1": 0.9078573030179498,
|
| 194 |
+
"step": 516
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.070038910505836,
|
| 198 |
+
"grad_norm": 21.746002197265625,
|
| 199 |
+
"learning_rate": 1.7708871662360035e-05,
|
| 200 |
+
"loss": 0.8159,
|
| 201 |
+
"step": 525
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 4.264591439688716,
|
| 205 |
+
"grad_norm": 25.9829158782959,
|
| 206 |
+
"learning_rate": 1.7493540051679588e-05,
|
| 207 |
+
"loss": 0.8029,
|
| 208 |
+
"step": 550
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 4.4591439688715955,
|
| 212 |
+
"grad_norm": 18.974702835083008,
|
| 213 |
+
"learning_rate": 1.727820844099914e-05,
|
| 214 |
+
"loss": 0.7237,
|
| 215 |
+
"step": 575
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.653696498054475,
|
| 219 |
+
"grad_norm": 8.793298721313477,
|
| 220 |
+
"learning_rate": 1.7062876830318693e-05,
|
| 221 |
+
"loss": 0.7558,
|
| 222 |
+
"step": 600
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.848249027237354,
|
| 226 |
+
"grad_norm": 23.602251052856445,
|
| 227 |
+
"learning_rate": 1.6847545219638243e-05,
|
| 228 |
+
"loss": 0.9415,
|
| 229 |
+
"step": 625
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.9168490153172867,
|
| 234 |
+
"eval_loss": 0.28598931431770325,
|
| 235 |
+
"eval_macro_f1": 0.8564511542784734,
|
| 236 |
+
"eval_runtime": 0.3704,
|
| 237 |
+
"eval_samples_per_second": 1233.742,
|
| 238 |
+
"eval_steps_per_second": 78.29,
|
| 239 |
+
"eval_weighted_f1": 0.9161291817810625,
|
| 240 |
+
"step": 645
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.038910505836576,
|
| 244 |
+
"grad_norm": 10.088534355163574,
|
| 245 |
+
"learning_rate": 1.6632213608957795e-05,
|
| 246 |
+
"loss": 0.8484,
|
| 247 |
+
"step": 650
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.233463035019455,
|
| 251 |
+
"grad_norm": 54.11484146118164,
|
| 252 |
+
"learning_rate": 1.6416881998277348e-05,
|
| 253 |
+
"loss": 0.7538,
|
| 254 |
+
"step": 675
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.428015564202335,
|
| 258 |
+
"grad_norm": 50.45320510864258,
|
| 259 |
+
"learning_rate": 1.62015503875969e-05,
|
| 260 |
+
"loss": 0.4758,
|
| 261 |
+
"step": 700
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.622568093385214,
|
| 265 |
+
"grad_norm": 61.879600524902344,
|
| 266 |
+
"learning_rate": 1.5986218776916453e-05,
|
| 267 |
+
"loss": 0.8378,
|
| 268 |
+
"step": 725
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.817120622568093,
|
| 272 |
+
"grad_norm": 16.213138580322266,
|
| 273 |
+
"learning_rate": 1.5770887166236006e-05,
|
| 274 |
+
"loss": 0.6686,
|
| 275 |
+
"step": 750
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.0,
|
| 279 |
+
"eval_accuracy": 0.9212253829321663,
|
| 280 |
+
"eval_loss": 0.31217917799949646,
|
| 281 |
+
"eval_macro_f1": 0.8698320583838917,
|
| 282 |
+
"eval_runtime": 0.3513,
|
| 283 |
+
"eval_samples_per_second": 1300.726,
|
| 284 |
+
"eval_steps_per_second": 82.541,
|
| 285 |
+
"eval_weighted_f1": 0.9204734003776074,
|
| 286 |
+
"step": 774
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 6.007782101167315,
|
| 290 |
+
"grad_norm": 11.760616302490234,
|
| 291 |
+
"learning_rate": 1.555555555555556e-05,
|
| 292 |
+
"loss": 0.4805,
|
| 293 |
+
"step": 775
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 6.202334630350195,
|
| 297 |
+
"grad_norm": 9.390809059143066,
|
| 298 |
+
"learning_rate": 1.534022394487511e-05,
|
| 299 |
+
"loss": 0.3366,
|
| 300 |
+
"step": 800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 6.396887159533074,
|
| 304 |
+
"grad_norm": 0.6686726212501526,
|
| 305 |
+
"learning_rate": 1.512489233419466e-05,
|
| 306 |
+
"loss": 0.4272,
|
| 307 |
+
"step": 825
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 6.591439688715953,
|
| 311 |
+
"grad_norm": 57.247215270996094,
|
| 312 |
+
"learning_rate": 1.4909560723514215e-05,
|
| 313 |
+
"loss": 0.6792,
|
| 314 |
+
"step": 850
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 6.785992217898833,
|
| 318 |
+
"grad_norm": 28.846025466918945,
|
| 319 |
+
"learning_rate": 1.4694229112833764e-05,
|
| 320 |
+
"loss": 0.5075,
|
| 321 |
+
"step": 875
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.980544747081712,
|
| 325 |
+
"grad_norm": 30.43777847290039,
|
| 326 |
+
"learning_rate": 1.4478897502153318e-05,
|
| 327 |
+
"loss": 0.6462,
|
| 328 |
+
"step": 900
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 7.0,
|
| 332 |
+
"eval_accuracy": 0.9234135667396062,
|
| 333 |
+
"eval_loss": 0.3408062160015106,
|
| 334 |
+
"eval_macro_f1": 0.8754667450288961,
|
| 335 |
+
"eval_runtime": 0.3487,
|
| 336 |
+
"eval_samples_per_second": 1310.444,
|
| 337 |
+
"eval_steps_per_second": 83.157,
|
| 338 |
+
"eval_weighted_f1": 0.9242855737407655,
|
| 339 |
+
"step": 903
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 7.171206225680934,
|
| 343 |
+
"grad_norm": 44.301124572753906,
|
| 344 |
+
"learning_rate": 1.426356589147287e-05,
|
| 345 |
+
"loss": 0.4255,
|
| 346 |
+
"step": 925
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.365758754863813,
|
| 350 |
+
"grad_norm": 38.106292724609375,
|
| 351 |
+
"learning_rate": 1.4048234280792422e-05,
|
| 352 |
+
"loss": 0.4863,
|
| 353 |
+
"step": 950
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 7.560311284046692,
|
| 357 |
+
"grad_norm": 13.320198059082031,
|
| 358 |
+
"learning_rate": 1.3832902670111974e-05,
|
| 359 |
+
"loss": 0.3598,
|
| 360 |
+
"step": 975
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 7.754863813229572,
|
| 364 |
+
"grad_norm": 0.39185717701911926,
|
| 365 |
+
"learning_rate": 1.3617571059431525e-05,
|
| 366 |
+
"loss": 0.3488,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.9494163424124515,
|
| 371 |
+
"grad_norm": 65.87716674804688,
|
| 372 |
+
"learning_rate": 1.3402239448751078e-05,
|
| 373 |
+
"loss": 0.6857,
|
| 374 |
+
"step": 1025
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.0,
|
| 378 |
+
"eval_accuracy": 0.9234135667396062,
|
| 379 |
+
"eval_loss": 0.3978811800479889,
|
| 380 |
+
"eval_macro_f1": 0.8724444593347033,
|
| 381 |
+
"eval_runtime": 0.3536,
|
| 382 |
+
"eval_samples_per_second": 1292.375,
|
| 383 |
+
"eval_steps_per_second": 82.011,
|
| 384 |
+
"eval_weighted_f1": 0.9225742180023816,
|
| 385 |
+
"step": 1032
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 8.140077821011673,
|
| 389 |
+
"grad_norm": 50.72336196899414,
|
| 390 |
+
"learning_rate": 1.318690783807063e-05,
|
| 391 |
+
"loss": 0.3311,
|
| 392 |
+
"step": 1050
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 8.334630350194553,
|
| 396 |
+
"grad_norm": 29.121295928955078,
|
| 397 |
+
"learning_rate": 1.2971576227390181e-05,
|
| 398 |
+
"loss": 0.4558,
|
| 399 |
+
"step": 1075
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 8.529182879377432,
|
| 403 |
+
"grad_norm": 29.663639068603516,
|
| 404 |
+
"learning_rate": 1.2756244616709734e-05,
|
| 405 |
+
"loss": 0.3734,
|
| 406 |
+
"step": 1100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 8.72373540856031,
|
| 410 |
+
"grad_norm": 95.52642059326172,
|
| 411 |
+
"learning_rate": 1.2540913006029287e-05,
|
| 412 |
+
"loss": 0.3917,
|
| 413 |
+
"step": 1125
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 8.918287937743191,
|
| 417 |
+
"grad_norm": 0.6678557991981506,
|
| 418 |
+
"learning_rate": 1.2325581395348838e-05,
|
| 419 |
+
"loss": 0.2513,
|
| 420 |
+
"step": 1150
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 9.0,
|
| 424 |
+
"eval_accuracy": 0.9343544857768052,
|
| 425 |
+
"eval_loss": 0.353714257478714,
|
| 426 |
+
"eval_macro_f1": 0.8933721587205007,
|
| 427 |
+
"eval_runtime": 0.3625,
|
| 428 |
+
"eval_samples_per_second": 1260.797,
|
| 429 |
+
"eval_steps_per_second": 80.007,
|
| 430 |
+
"eval_weighted_f1": 0.9342032177875788,
|
| 431 |
+
"step": 1161
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 9.108949416342412,
|
| 435 |
+
"grad_norm": 62.43782043457031,
|
| 436 |
+
"learning_rate": 1.211024978466839e-05,
|
| 437 |
+
"loss": 0.5743,
|
| 438 |
+
"step": 1175
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 9.303501945525293,
|
| 442 |
+
"grad_norm": 2.6969964504241943,
|
| 443 |
+
"learning_rate": 1.1894918173987941e-05,
|
| 444 |
+
"loss": 0.358,
|
| 445 |
+
"step": 1200
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 9.498054474708171,
|
| 449 |
+
"grad_norm": 0.214388906955719,
|
| 450 |
+
"learning_rate": 1.1679586563307494e-05,
|
| 451 |
+
"loss": 0.4152,
|
| 452 |
+
"step": 1225
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 9.69260700389105,
|
| 456 |
+
"grad_norm": 0.138688325881958,
|
| 457 |
+
"learning_rate": 1.1464254952627048e-05,
|
| 458 |
+
"loss": 0.3211,
|
| 459 |
+
"step": 1250
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 9.88715953307393,
|
| 463 |
+
"grad_norm": 179.92041015625,
|
| 464 |
+
"learning_rate": 1.1248923341946599e-05,
|
| 465 |
+
"loss": 0.3288,
|
| 466 |
+
"step": 1275
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.0,
|
| 470 |
+
"eval_accuracy": 0.936542669584245,
|
| 471 |
+
"eval_loss": 0.377521276473999,
|
| 472 |
+
"eval_macro_f1": 0.8888152931868779,
|
| 473 |
+
"eval_runtime": 0.3346,
|
| 474 |
+
"eval_samples_per_second": 1365.739,
|
| 475 |
+
"eval_steps_per_second": 86.666,
|
| 476 |
+
"eval_weighted_f1": 0.936057614322851,
|
| 477 |
+
"step": 1290
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 10.077821011673151,
|
| 481 |
+
"grad_norm": 141.60292053222656,
|
| 482 |
+
"learning_rate": 1.1033591731266152e-05,
|
| 483 |
+
"loss": 0.2107,
|
| 484 |
+
"step": 1300
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 10.272373540856032,
|
| 488 |
+
"grad_norm": 218.48448181152344,
|
| 489 |
+
"learning_rate": 1.0818260120585703e-05,
|
| 490 |
+
"loss": 0.2774,
|
| 491 |
+
"step": 1325
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 10.46692607003891,
|
| 495 |
+
"grad_norm": 0.3349389433860779,
|
| 496 |
+
"learning_rate": 1.0602928509905255e-05,
|
| 497 |
+
"loss": 0.1595,
|
| 498 |
+
"step": 1350
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 10.66147859922179,
|
| 502 |
+
"grad_norm": 5.962428092956543,
|
| 503 |
+
"learning_rate": 1.0387596899224808e-05,
|
| 504 |
+
"loss": 0.3567,
|
| 505 |
+
"step": 1375
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 10.85603112840467,
|
| 509 |
+
"grad_norm": 88.04630279541016,
|
| 510 |
+
"learning_rate": 1.0172265288544359e-05,
|
| 511 |
+
"loss": 0.4024,
|
| 512 |
+
"step": 1400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 11.0,
|
| 516 |
+
"eval_accuracy": 0.925601750547046,
|
| 517 |
+
"eval_loss": 0.45094430446624756,
|
| 518 |
+
"eval_macro_f1": 0.8687936819858734,
|
| 519 |
+
"eval_runtime": 0.4231,
|
| 520 |
+
"eval_samples_per_second": 1080.11,
|
| 521 |
+
"eval_steps_per_second": 68.541,
|
| 522 |
+
"eval_weighted_f1": 0.9249293001345136,
|
| 523 |
+
"step": 1419
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 11.04669260700389,
|
| 527 |
+
"grad_norm": 45.47275161743164,
|
| 528 |
+
"learning_rate": 9.956933677863911e-06,
|
| 529 |
+
"loss": 0.2046,
|
| 530 |
+
"step": 1425
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 11.24124513618677,
|
| 534 |
+
"grad_norm": 78.47571563720703,
|
| 535 |
+
"learning_rate": 9.741602067183464e-06,
|
| 536 |
+
"loss": 0.2269,
|
| 537 |
+
"step": 1450
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 11.43579766536965,
|
| 541 |
+
"grad_norm": 44.41475296020508,
|
| 542 |
+
"learning_rate": 9.526270456503015e-06,
|
| 543 |
+
"loss": 0.2691,
|
| 544 |
+
"step": 1475
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 11.630350194552529,
|
| 548 |
+
"grad_norm": 96.01416778564453,
|
| 549 |
+
"learning_rate": 9.310938845822568e-06,
|
| 550 |
+
"loss": 0.0855,
|
| 551 |
+
"step": 1500
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 11.82490272373541,
|
| 555 |
+
"grad_norm": 0.2217947393655777,
|
| 556 |
+
"learning_rate": 9.09560723514212e-06,
|
| 557 |
+
"loss": 0.1822,
|
| 558 |
+
"step": 1525
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 12.0,
|
| 562 |
+
"eval_accuracy": 0.9234135667396062,
|
| 563 |
+
"eval_loss": 0.4788913130760193,
|
| 564 |
+
"eval_macro_f1": 0.8695500930795048,
|
| 565 |
+
"eval_runtime": 0.364,
|
| 566 |
+
"eval_samples_per_second": 1255.514,
|
| 567 |
+
"eval_steps_per_second": 79.672,
|
| 568 |
+
"eval_weighted_f1": 0.9231336904396501,
|
| 569 |
+
"step": 1548
|
| 570 |
+
}
|
| 571 |
+
],
|
| 572 |
+
"logging_steps": 25,
|
| 573 |
+
"max_steps": 2580,
|
| 574 |
+
"num_input_tokens_seen": 0,
|
| 575 |
+
"num_train_epochs": 20,
|
| 576 |
+
"save_steps": 500,
|
| 577 |
+
"stateful_callbacks": {
|
| 578 |
+
"EarlyStoppingCallback": {
|
| 579 |
+
"args": {
|
| 580 |
+
"early_stopping_patience": 3,
|
| 581 |
+
"early_stopping_threshold": 0.0
|
| 582 |
+
},
|
| 583 |
+
"attributes": {
|
| 584 |
+
"early_stopping_patience_counter": 3
|
| 585 |
+
}
|
| 586 |
+
},
|
| 587 |
+
"TrainerControl": {
|
| 588 |
+
"args": {
|
| 589 |
+
"should_epoch_stop": false,
|
| 590 |
+
"should_evaluate": false,
|
| 591 |
+
"should_log": false,
|
| 592 |
+
"should_save": true,
|
| 593 |
+
"should_training_stop": true
|
| 594 |
+
},
|
| 595 |
+
"attributes": {}
|
| 596 |
+
}
|
| 597 |
+
},
|
| 598 |
+
"total_flos": 0.0,
|
| 599 |
+
"train_batch_size": 8,
|
| 600 |
+
"trial_name": null,
|
| 601 |
+
"trial_params": null
|
| 602 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-1161/trainer_state.json
ADDED
|
@@ -0,0 +1,464 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 774,
|
| 3 |
+
"best_metric": 0.8782036171509855,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774",
|
| 5 |
+
"epoch": 9.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1161,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.19455252918287938,
|
| 14 |
+
"grad_norm": 42.437381744384766,
|
| 15 |
+
"learning_rate": 1.86046511627907e-06,
|
| 16 |
+
"loss": 4.3146,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.38910505836575876,
|
| 21 |
+
"grad_norm": 68.05741119384766,
|
| 22 |
+
"learning_rate": 3.798449612403101e-06,
|
| 23 |
+
"loss": 3.2451,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5836575875486382,
|
| 28 |
+
"grad_norm": 93.28146362304688,
|
| 29 |
+
"learning_rate": 5.736434108527133e-06,
|
| 30 |
+
"loss": 2.9301,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7782101167315175,
|
| 35 |
+
"grad_norm": 60.758113861083984,
|
| 36 |
+
"learning_rate": 7.674418604651164e-06,
|
| 37 |
+
"loss": 2.9008,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.9727626459143969,
|
| 42 |
+
"grad_norm": 31.677011489868164,
|
| 43 |
+
"learning_rate": 9.612403100775196e-06,
|
| 44 |
+
"loss": 2.8955,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.0,
|
| 49 |
+
"eval_accuracy": 0.75054704595186,
|
| 50 |
+
"eval_loss": 0.6465314030647278,
|
| 51 |
+
"eval_macro_f1": 0.370311029729661,
|
| 52 |
+
"eval_runtime": 0.3964,
|
| 53 |
+
"eval_samples_per_second": 1152.764,
|
| 54 |
+
"eval_steps_per_second": 73.151,
|
| 55 |
+
"eval_weighted_f1": 0.6700552046036169,
|
| 56 |
+
"step": 129
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.1634241245136188,
|
| 60 |
+
"grad_norm": 19.683895111083984,
|
| 61 |
+
"learning_rate": 1.1550387596899227e-05,
|
| 62 |
+
"loss": 2.6277,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.3579766536964981,
|
| 67 |
+
"grad_norm": 33.9718132019043,
|
| 68 |
+
"learning_rate": 1.3488372093023257e-05,
|
| 69 |
+
"loss": 2.6868,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.5525291828793775,
|
| 74 |
+
"grad_norm": 57.723384857177734,
|
| 75 |
+
"learning_rate": 1.542635658914729e-05,
|
| 76 |
+
"loss": 2.1346,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.7470817120622568,
|
| 81 |
+
"grad_norm": 42.020530700683594,
|
| 82 |
+
"learning_rate": 1.736434108527132e-05,
|
| 83 |
+
"loss": 1.8879,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.9416342412451362,
|
| 88 |
+
"grad_norm": 31.28432273864746,
|
| 89 |
+
"learning_rate": 1.9302325581395353e-05,
|
| 90 |
+
"loss": 1.7204,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_accuracy": 0.87527352297593,
|
| 96 |
+
"eval_loss": 0.3245304822921753,
|
| 97 |
+
"eval_macro_f1": 0.7129172985692023,
|
| 98 |
+
"eval_runtime": 0.3725,
|
| 99 |
+
"eval_samples_per_second": 1226.801,
|
| 100 |
+
"eval_steps_per_second": 77.85,
|
| 101 |
+
"eval_weighted_f1": 0.8616802328120389,
|
| 102 |
+
"step": 258
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.132295719844358,
|
| 106 |
+
"grad_norm": 36.273033142089844,
|
| 107 |
+
"learning_rate": 1.9862187769164516e-05,
|
| 108 |
+
"loss": 1.4204,
|
| 109 |
+
"step": 275
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3268482490272375,
|
| 113 |
+
"grad_norm": 35.57266616821289,
|
| 114 |
+
"learning_rate": 1.9646856158484065e-05,
|
| 115 |
+
"loss": 1.4879,
|
| 116 |
+
"step": 300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.5214007782101167,
|
| 120 |
+
"grad_norm": 57.251346588134766,
|
| 121 |
+
"learning_rate": 1.943152454780362e-05,
|
| 122 |
+
"loss": 1.6221,
|
| 123 |
+
"step": 325
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.7159533073929962,
|
| 127 |
+
"grad_norm": 41.13840866088867,
|
| 128 |
+
"learning_rate": 1.921619293712317e-05,
|
| 129 |
+
"loss": 1.3329,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.9105058365758754,
|
| 134 |
+
"grad_norm": 71.34947967529297,
|
| 135 |
+
"learning_rate": 1.9000861326442723e-05,
|
| 136 |
+
"loss": 1.2789,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.0,
|
| 141 |
+
"eval_accuracy": 0.9080962800875274,
|
| 142 |
+
"eval_loss": 0.2612227201461792,
|
| 143 |
+
"eval_macro_f1": 0.828922565764671,
|
| 144 |
+
"eval_runtime": 0.3702,
|
| 145 |
+
"eval_samples_per_second": 1234.603,
|
| 146 |
+
"eval_steps_per_second": 78.345,
|
| 147 |
+
"eval_weighted_f1": 0.9070559496679501,
|
| 148 |
+
"step": 387
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.1011673151750974,
|
| 152 |
+
"grad_norm": 27.597007751464844,
|
| 153 |
+
"learning_rate": 1.8785529715762276e-05,
|
| 154 |
+
"loss": 1.0501,
|
| 155 |
+
"step": 400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.2957198443579765,
|
| 159 |
+
"grad_norm": 41.9455680847168,
|
| 160 |
+
"learning_rate": 1.857019810508183e-05,
|
| 161 |
+
"loss": 1.0763,
|
| 162 |
+
"step": 425
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.490272373540856,
|
| 166 |
+
"grad_norm": 26.788463592529297,
|
| 167 |
+
"learning_rate": 1.835486649440138e-05,
|
| 168 |
+
"loss": 0.95,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.6848249027237356,
|
| 173 |
+
"grad_norm": 93.82257843017578,
|
| 174 |
+
"learning_rate": 1.813953488372093e-05,
|
| 175 |
+
"loss": 1.1273,
|
| 176 |
+
"step": 475
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.8793774319066148,
|
| 180 |
+
"grad_norm": 41.75449752807617,
|
| 181 |
+
"learning_rate": 1.7924203273040483e-05,
|
| 182 |
+
"loss": 1.1741,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.9321663019693655,
|
| 188 |
+
"eval_loss": 0.2191532701253891,
|
| 189 |
+
"eval_macro_f1": 0.8629308173205711,
|
| 190 |
+
"eval_runtime": 0.3607,
|
| 191 |
+
"eval_samples_per_second": 1266.859,
|
| 192 |
+
"eval_steps_per_second": 80.391,
|
| 193 |
+
"eval_weighted_f1": 0.930750777984384,
|
| 194 |
+
"step": 516
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.070038910505836,
|
| 198 |
+
"grad_norm": 68.79255676269531,
|
| 199 |
+
"learning_rate": 1.7708871662360035e-05,
|
| 200 |
+
"loss": 1.1744,
|
| 201 |
+
"step": 525
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 4.264591439688716,
|
| 205 |
+
"grad_norm": 41.710453033447266,
|
| 206 |
+
"learning_rate": 1.7493540051679588e-05,
|
| 207 |
+
"loss": 0.7889,
|
| 208 |
+
"step": 550
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 4.4591439688715955,
|
| 212 |
+
"grad_norm": 14.018392562866211,
|
| 213 |
+
"learning_rate": 1.727820844099914e-05,
|
| 214 |
+
"loss": 0.6165,
|
| 215 |
+
"step": 575
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.653696498054475,
|
| 219 |
+
"grad_norm": 130.08865356445312,
|
| 220 |
+
"learning_rate": 1.7062876830318693e-05,
|
| 221 |
+
"loss": 1.0093,
|
| 222 |
+
"step": 600
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.848249027237354,
|
| 226 |
+
"grad_norm": 40.896446228027344,
|
| 227 |
+
"learning_rate": 1.6847545219638243e-05,
|
| 228 |
+
"loss": 0.8569,
|
| 229 |
+
"step": 625
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.9299781181619255,
|
| 234 |
+
"eval_loss": 0.23258835077285767,
|
| 235 |
+
"eval_macro_f1": 0.8734708153094645,
|
| 236 |
+
"eval_runtime": 0.3686,
|
| 237 |
+
"eval_samples_per_second": 1239.667,
|
| 238 |
+
"eval_steps_per_second": 78.666,
|
| 239 |
+
"eval_weighted_f1": 0.9282719573464747,
|
| 240 |
+
"step": 645
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.038910505836576,
|
| 244 |
+
"grad_norm": 21.36417007446289,
|
| 245 |
+
"learning_rate": 1.6632213608957795e-05,
|
| 246 |
+
"loss": 0.7699,
|
| 247 |
+
"step": 650
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.233463035019455,
|
| 251 |
+
"grad_norm": 46.736480712890625,
|
| 252 |
+
"learning_rate": 1.6416881998277348e-05,
|
| 253 |
+
"loss": 0.5599,
|
| 254 |
+
"step": 675
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.428015564202335,
|
| 258 |
+
"grad_norm": 74.67058563232422,
|
| 259 |
+
"learning_rate": 1.62015503875969e-05,
|
| 260 |
+
"loss": 0.6069,
|
| 261 |
+
"step": 700
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.622568093385214,
|
| 265 |
+
"grad_norm": 26.96391487121582,
|
| 266 |
+
"learning_rate": 1.5986218776916453e-05,
|
| 267 |
+
"loss": 0.7992,
|
| 268 |
+
"step": 725
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.817120622568093,
|
| 272 |
+
"grad_norm": 58.64337921142578,
|
| 273 |
+
"learning_rate": 1.5770887166236006e-05,
|
| 274 |
+
"loss": 0.7714,
|
| 275 |
+
"step": 750
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.0,
|
| 279 |
+
"eval_accuracy": 0.9321663019693655,
|
| 280 |
+
"eval_loss": 0.25988802313804626,
|
| 281 |
+
"eval_macro_f1": 0.8782036171509855,
|
| 282 |
+
"eval_runtime": 0.3627,
|
| 283 |
+
"eval_samples_per_second": 1260.106,
|
| 284 |
+
"eval_steps_per_second": 79.963,
|
| 285 |
+
"eval_weighted_f1": 0.9312984809886804,
|
| 286 |
+
"step": 774
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 6.007782101167315,
|
| 290 |
+
"grad_norm": 14.113422393798828,
|
| 291 |
+
"learning_rate": 1.555555555555556e-05,
|
| 292 |
+
"loss": 0.7661,
|
| 293 |
+
"step": 775
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 6.202334630350195,
|
| 297 |
+
"grad_norm": 39.16048812866211,
|
| 298 |
+
"learning_rate": 1.534022394487511e-05,
|
| 299 |
+
"loss": 0.5945,
|
| 300 |
+
"step": 800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 6.396887159533074,
|
| 304 |
+
"grad_norm": 7.581872463226318,
|
| 305 |
+
"learning_rate": 1.512489233419466e-05,
|
| 306 |
+
"loss": 0.5984,
|
| 307 |
+
"step": 825
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 6.591439688715953,
|
| 311 |
+
"grad_norm": 17.86186981201172,
|
| 312 |
+
"learning_rate": 1.4909560723514215e-05,
|
| 313 |
+
"loss": 0.4133,
|
| 314 |
+
"step": 850
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 6.785992217898833,
|
| 318 |
+
"grad_norm": 56.85751724243164,
|
| 319 |
+
"learning_rate": 1.4694229112833764e-05,
|
| 320 |
+
"loss": 0.488,
|
| 321 |
+
"step": 875
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.980544747081712,
|
| 325 |
+
"grad_norm": 4.516177177429199,
|
| 326 |
+
"learning_rate": 1.4478897502153318e-05,
|
| 327 |
+
"loss": 0.6025,
|
| 328 |
+
"step": 900
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 7.0,
|
| 332 |
+
"eval_accuracy": 0.9234135667396062,
|
| 333 |
+
"eval_loss": 0.35557329654693604,
|
| 334 |
+
"eval_macro_f1": 0.8623323255239518,
|
| 335 |
+
"eval_runtime": 0.3551,
|
| 336 |
+
"eval_samples_per_second": 1286.999,
|
| 337 |
+
"eval_steps_per_second": 81.67,
|
| 338 |
+
"eval_weighted_f1": 0.9238155733170861,
|
| 339 |
+
"step": 903
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 7.171206225680934,
|
| 343 |
+
"grad_norm": 119.83162689208984,
|
| 344 |
+
"learning_rate": 1.426356589147287e-05,
|
| 345 |
+
"loss": 0.4051,
|
| 346 |
+
"step": 925
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.365758754863813,
|
| 350 |
+
"grad_norm": 69.04756164550781,
|
| 351 |
+
"learning_rate": 1.4048234280792422e-05,
|
| 352 |
+
"loss": 0.6472,
|
| 353 |
+
"step": 950
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 7.560311284046692,
|
| 357 |
+
"grad_norm": 0.9695833325386047,
|
| 358 |
+
"learning_rate": 1.3832902670111974e-05,
|
| 359 |
+
"loss": 0.5742,
|
| 360 |
+
"step": 975
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 7.754863813229572,
|
| 364 |
+
"grad_norm": 135.23321533203125,
|
| 365 |
+
"learning_rate": 1.3617571059431525e-05,
|
| 366 |
+
"loss": 0.6264,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.9494163424124515,
|
| 371 |
+
"grad_norm": 53.247135162353516,
|
| 372 |
+
"learning_rate": 1.3402239448751078e-05,
|
| 373 |
+
"loss": 0.4108,
|
| 374 |
+
"step": 1025
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.0,
|
| 378 |
+
"eval_accuracy": 0.9146608315098468,
|
| 379 |
+
"eval_loss": 0.40400078892707825,
|
| 380 |
+
"eval_macro_f1": 0.8359925127467779,
|
| 381 |
+
"eval_runtime": 0.3978,
|
| 382 |
+
"eval_samples_per_second": 1148.862,
|
| 383 |
+
"eval_steps_per_second": 72.904,
|
| 384 |
+
"eval_weighted_f1": 0.9115045291820524,
|
| 385 |
+
"step": 1032
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 8.140077821011673,
|
| 389 |
+
"grad_norm": 68.49388122558594,
|
| 390 |
+
"learning_rate": 1.318690783807063e-05,
|
| 391 |
+
"loss": 0.5413,
|
| 392 |
+
"step": 1050
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 8.334630350194553,
|
| 396 |
+
"grad_norm": 0.12987364828586578,
|
| 397 |
+
"learning_rate": 1.2971576227390181e-05,
|
| 398 |
+
"loss": 0.3313,
|
| 399 |
+
"step": 1075
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 8.529182879377432,
|
| 403 |
+
"grad_norm": 72.9069595336914,
|
| 404 |
+
"learning_rate": 1.2756244616709734e-05,
|
| 405 |
+
"loss": 0.6094,
|
| 406 |
+
"step": 1100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 8.72373540856031,
|
| 410 |
+
"grad_norm": 1.0227947235107422,
|
| 411 |
+
"learning_rate": 1.2540913006029287e-05,
|
| 412 |
+
"loss": 0.4517,
|
| 413 |
+
"step": 1125
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 8.918287937743191,
|
| 417 |
+
"grad_norm": 129.0871124267578,
|
| 418 |
+
"learning_rate": 1.2325581395348838e-05,
|
| 419 |
+
"loss": 0.5077,
|
| 420 |
+
"step": 1150
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 9.0,
|
| 424 |
+
"eval_accuracy": 0.9168490153172867,
|
| 425 |
+
"eval_loss": 0.4403058886528015,
|
| 426 |
+
"eval_macro_f1": 0.8547903079174889,
|
| 427 |
+
"eval_runtime": 0.363,
|
| 428 |
+
"eval_samples_per_second": 1258.847,
|
| 429 |
+
"eval_steps_per_second": 79.883,
|
| 430 |
+
"eval_weighted_f1": 0.9159470641159729,
|
| 431 |
+
"step": 1161
|
| 432 |
+
}
|
| 433 |
+
],
|
| 434 |
+
"logging_steps": 25,
|
| 435 |
+
"max_steps": 2580,
|
| 436 |
+
"num_input_tokens_seen": 0,
|
| 437 |
+
"num_train_epochs": 20,
|
| 438 |
+
"save_steps": 500,
|
| 439 |
+
"stateful_callbacks": {
|
| 440 |
+
"EarlyStoppingCallback": {
|
| 441 |
+
"args": {
|
| 442 |
+
"early_stopping_patience": 3,
|
| 443 |
+
"early_stopping_threshold": 0.0
|
| 444 |
+
},
|
| 445 |
+
"attributes": {
|
| 446 |
+
"early_stopping_patience_counter": 3
|
| 447 |
+
}
|
| 448 |
+
},
|
| 449 |
+
"TrainerControl": {
|
| 450 |
+
"args": {
|
| 451 |
+
"should_epoch_stop": false,
|
| 452 |
+
"should_evaluate": false,
|
| 453 |
+
"should_log": false,
|
| 454 |
+
"should_save": true,
|
| 455 |
+
"should_training_stop": true
|
| 456 |
+
},
|
| 457 |
+
"attributes": {}
|
| 458 |
+
}
|
| 459 |
+
},
|
| 460 |
+
"total_flos": 0.0,
|
| 461 |
+
"train_batch_size": 8,
|
| 462 |
+
"trial_name": null,
|
| 463 |
+
"trial_params": null
|
| 464 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774/trainer_state.json
ADDED
|
@@ -0,0 +1,319 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 774,
|
| 3 |
+
"best_metric": 0.8782036171509855,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-01-s43-20260827T161312Z-85e5496f/trainer/checkpoint-774",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 774,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.19455252918287938,
|
| 14 |
+
"grad_norm": 42.437381744384766,
|
| 15 |
+
"learning_rate": 1.86046511627907e-06,
|
| 16 |
+
"loss": 4.3146,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.38910505836575876,
|
| 21 |
+
"grad_norm": 68.05741119384766,
|
| 22 |
+
"learning_rate": 3.798449612403101e-06,
|
| 23 |
+
"loss": 3.2451,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5836575875486382,
|
| 28 |
+
"grad_norm": 93.28146362304688,
|
| 29 |
+
"learning_rate": 5.736434108527133e-06,
|
| 30 |
+
"loss": 2.9301,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7782101167315175,
|
| 35 |
+
"grad_norm": 60.758113861083984,
|
| 36 |
+
"learning_rate": 7.674418604651164e-06,
|
| 37 |
+
"loss": 2.9008,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.9727626459143969,
|
| 42 |
+
"grad_norm": 31.677011489868164,
|
| 43 |
+
"learning_rate": 9.612403100775196e-06,
|
| 44 |
+
"loss": 2.8955,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.0,
|
| 49 |
+
"eval_accuracy": 0.75054704595186,
|
| 50 |
+
"eval_loss": 0.6465314030647278,
|
| 51 |
+
"eval_macro_f1": 0.370311029729661,
|
| 52 |
+
"eval_runtime": 0.3964,
|
| 53 |
+
"eval_samples_per_second": 1152.764,
|
| 54 |
+
"eval_steps_per_second": 73.151,
|
| 55 |
+
"eval_weighted_f1": 0.6700552046036169,
|
| 56 |
+
"step": 129
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.1634241245136188,
|
| 60 |
+
"grad_norm": 19.683895111083984,
|
| 61 |
+
"learning_rate": 1.1550387596899227e-05,
|
| 62 |
+
"loss": 2.6277,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.3579766536964981,
|
| 67 |
+
"grad_norm": 33.9718132019043,
|
| 68 |
+
"learning_rate": 1.3488372093023257e-05,
|
| 69 |
+
"loss": 2.6868,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.5525291828793775,
|
| 74 |
+
"grad_norm": 57.723384857177734,
|
| 75 |
+
"learning_rate": 1.542635658914729e-05,
|
| 76 |
+
"loss": 2.1346,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.7470817120622568,
|
| 81 |
+
"grad_norm": 42.020530700683594,
|
| 82 |
+
"learning_rate": 1.736434108527132e-05,
|
| 83 |
+
"loss": 1.8879,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.9416342412451362,
|
| 88 |
+
"grad_norm": 31.28432273864746,
|
| 89 |
+
"learning_rate": 1.9302325581395353e-05,
|
| 90 |
+
"loss": 1.7204,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_accuracy": 0.87527352297593,
|
| 96 |
+
"eval_loss": 0.3245304822921753,
|
| 97 |
+
"eval_macro_f1": 0.7129172985692023,
|
| 98 |
+
"eval_runtime": 0.3725,
|
| 99 |
+
"eval_samples_per_second": 1226.801,
|
| 100 |
+
"eval_steps_per_second": 77.85,
|
| 101 |
+
"eval_weighted_f1": 0.8616802328120389,
|
| 102 |
+
"step": 258
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.132295719844358,
|
| 106 |
+
"grad_norm": 36.273033142089844,
|
| 107 |
+
"learning_rate": 1.9862187769164516e-05,
|
| 108 |
+
"loss": 1.4204,
|
| 109 |
+
"step": 275
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3268482490272375,
|
| 113 |
+
"grad_norm": 35.57266616821289,
|
| 114 |
+
"learning_rate": 1.9646856158484065e-05,
|
| 115 |
+
"loss": 1.4879,
|
| 116 |
+
"step": 300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.5214007782101167,
|
| 120 |
+
"grad_norm": 57.251346588134766,
|
| 121 |
+
"learning_rate": 1.943152454780362e-05,
|
| 122 |
+
"loss": 1.6221,
|
| 123 |
+
"step": 325
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.7159533073929962,
|
| 127 |
+
"grad_norm": 41.13840866088867,
|
| 128 |
+
"learning_rate": 1.921619293712317e-05,
|
| 129 |
+
"loss": 1.3329,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.9105058365758754,
|
| 134 |
+
"grad_norm": 71.34947967529297,
|
| 135 |
+
"learning_rate": 1.9000861326442723e-05,
|
| 136 |
+
"loss": 1.2789,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.0,
|
| 141 |
+
"eval_accuracy": 0.9080962800875274,
|
| 142 |
+
"eval_loss": 0.2612227201461792,
|
| 143 |
+
"eval_macro_f1": 0.828922565764671,
|
| 144 |
+
"eval_runtime": 0.3702,
|
| 145 |
+
"eval_samples_per_second": 1234.603,
|
| 146 |
+
"eval_steps_per_second": 78.345,
|
| 147 |
+
"eval_weighted_f1": 0.9070559496679501,
|
| 148 |
+
"step": 387
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.1011673151750974,
|
| 152 |
+
"grad_norm": 27.597007751464844,
|
| 153 |
+
"learning_rate": 1.8785529715762276e-05,
|
| 154 |
+
"loss": 1.0501,
|
| 155 |
+
"step": 400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.2957198443579765,
|
| 159 |
+
"grad_norm": 41.9455680847168,
|
| 160 |
+
"learning_rate": 1.857019810508183e-05,
|
| 161 |
+
"loss": 1.0763,
|
| 162 |
+
"step": 425
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.490272373540856,
|
| 166 |
+
"grad_norm": 26.788463592529297,
|
| 167 |
+
"learning_rate": 1.835486649440138e-05,
|
| 168 |
+
"loss": 0.95,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.6848249027237356,
|
| 173 |
+
"grad_norm": 93.82257843017578,
|
| 174 |
+
"learning_rate": 1.813953488372093e-05,
|
| 175 |
+
"loss": 1.1273,
|
| 176 |
+
"step": 475
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.8793774319066148,
|
| 180 |
+
"grad_norm": 41.75449752807617,
|
| 181 |
+
"learning_rate": 1.7924203273040483e-05,
|
| 182 |
+
"loss": 1.1741,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.9321663019693655,
|
| 188 |
+
"eval_loss": 0.2191532701253891,
|
| 189 |
+
"eval_macro_f1": 0.8629308173205711,
|
| 190 |
+
"eval_runtime": 0.3607,
|
| 191 |
+
"eval_samples_per_second": 1266.859,
|
| 192 |
+
"eval_steps_per_second": 80.391,
|
| 193 |
+
"eval_weighted_f1": 0.930750777984384,
|
| 194 |
+
"step": 516
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.070038910505836,
|
| 198 |
+
"grad_norm": 68.79255676269531,
|
| 199 |
+
"learning_rate": 1.7708871662360035e-05,
|
| 200 |
+
"loss": 1.1744,
|
| 201 |
+
"step": 525
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 4.264591439688716,
|
| 205 |
+
"grad_norm": 41.710453033447266,
|
| 206 |
+
"learning_rate": 1.7493540051679588e-05,
|
| 207 |
+
"loss": 0.7889,
|
| 208 |
+
"step": 550
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 4.4591439688715955,
|
| 212 |
+
"grad_norm": 14.018392562866211,
|
| 213 |
+
"learning_rate": 1.727820844099914e-05,
|
| 214 |
+
"loss": 0.6165,
|
| 215 |
+
"step": 575
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.653696498054475,
|
| 219 |
+
"grad_norm": 130.08865356445312,
|
| 220 |
+
"learning_rate": 1.7062876830318693e-05,
|
| 221 |
+
"loss": 1.0093,
|
| 222 |
+
"step": 600
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.848249027237354,
|
| 226 |
+
"grad_norm": 40.896446228027344,
|
| 227 |
+
"learning_rate": 1.6847545219638243e-05,
|
| 228 |
+
"loss": 0.8569,
|
| 229 |
+
"step": 625
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.9299781181619255,
|
| 234 |
+
"eval_loss": 0.23258835077285767,
|
| 235 |
+
"eval_macro_f1": 0.8734708153094645,
|
| 236 |
+
"eval_runtime": 0.3686,
|
| 237 |
+
"eval_samples_per_second": 1239.667,
|
| 238 |
+
"eval_steps_per_second": 78.666,
|
| 239 |
+
"eval_weighted_f1": 0.9282719573464747,
|
| 240 |
+
"step": 645
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.038910505836576,
|
| 244 |
+
"grad_norm": 21.36417007446289,
|
| 245 |
+
"learning_rate": 1.6632213608957795e-05,
|
| 246 |
+
"loss": 0.7699,
|
| 247 |
+
"step": 650
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.233463035019455,
|
| 251 |
+
"grad_norm": 46.736480712890625,
|
| 252 |
+
"learning_rate": 1.6416881998277348e-05,
|
| 253 |
+
"loss": 0.5599,
|
| 254 |
+
"step": 675
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.428015564202335,
|
| 258 |
+
"grad_norm": 74.67058563232422,
|
| 259 |
+
"learning_rate": 1.62015503875969e-05,
|
| 260 |
+
"loss": 0.6069,
|
| 261 |
+
"step": 700
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.622568093385214,
|
| 265 |
+
"grad_norm": 26.96391487121582,
|
| 266 |
+
"learning_rate": 1.5986218776916453e-05,
|
| 267 |
+
"loss": 0.7992,
|
| 268 |
+
"step": 725
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.817120622568093,
|
| 272 |
+
"grad_norm": 58.64337921142578,
|
| 273 |
+
"learning_rate": 1.5770887166236006e-05,
|
| 274 |
+
"loss": 0.7714,
|
| 275 |
+
"step": 750
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.0,
|
| 279 |
+
"eval_accuracy": 0.9321663019693655,
|
| 280 |
+
"eval_loss": 0.25988802313804626,
|
| 281 |
+
"eval_macro_f1": 0.8782036171509855,
|
| 282 |
+
"eval_runtime": 0.3627,
|
| 283 |
+
"eval_samples_per_second": 1260.106,
|
| 284 |
+
"eval_steps_per_second": 79.963,
|
| 285 |
+
"eval_weighted_f1": 0.9312984809886804,
|
| 286 |
+
"step": 774
|
| 287 |
+
}
|
| 288 |
+
],
|
| 289 |
+
"logging_steps": 25,
|
| 290 |
+
"max_steps": 2580,
|
| 291 |
+
"num_input_tokens_seen": 0,
|
| 292 |
+
"num_train_epochs": 20,
|
| 293 |
+
"save_steps": 500,
|
| 294 |
+
"stateful_callbacks": {
|
| 295 |
+
"EarlyStoppingCallback": {
|
| 296 |
+
"args": {
|
| 297 |
+
"early_stopping_patience": 3,
|
| 298 |
+
"early_stopping_threshold": 0.0
|
| 299 |
+
},
|
| 300 |
+
"attributes": {
|
| 301 |
+
"early_stopping_patience_counter": 0
|
| 302 |
+
}
|
| 303 |
+
},
|
| 304 |
+
"TrainerControl": {
|
| 305 |
+
"args": {
|
| 306 |
+
"should_epoch_stop": false,
|
| 307 |
+
"should_evaluate": false,
|
| 308 |
+
"should_log": false,
|
| 309 |
+
"should_save": true,
|
| 310 |
+
"should_training_stop": false
|
| 311 |
+
},
|
| 312 |
+
"attributes": {}
|
| 313 |
+
}
|
| 314 |
+
},
|
| 315 |
+
"total_flos": 0.0,
|
| 316 |
+
"train_batch_size": 8,
|
| 317 |
+
"trial_name": null,
|
| 318 |
+
"trial_params": null
|
| 319 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032/trainer_state.json
ADDED
|
@@ -0,0 +1,418 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1032,
|
| 3 |
+
"best_metric": 0.8541751239520002,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1032,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.19455252918287938,
|
| 14 |
+
"grad_norm": 67.18620300292969,
|
| 15 |
+
"learning_rate": 1.86046511627907e-06,
|
| 16 |
+
"loss": 4.1527,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.38910505836575876,
|
| 21 |
+
"grad_norm": 20.784238815307617,
|
| 22 |
+
"learning_rate": 3.798449612403101e-06,
|
| 23 |
+
"loss": 3.2703,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5836575875486382,
|
| 28 |
+
"grad_norm": 36.513999938964844,
|
| 29 |
+
"learning_rate": 5.736434108527133e-06,
|
| 30 |
+
"loss": 2.9434,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7782101167315175,
|
| 35 |
+
"grad_norm": 30.496740341186523,
|
| 36 |
+
"learning_rate": 7.674418604651164e-06,
|
| 37 |
+
"loss": 2.9142,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.9727626459143969,
|
| 42 |
+
"grad_norm": 21.509984970092773,
|
| 43 |
+
"learning_rate": 9.612403100775196e-06,
|
| 44 |
+
"loss": 2.8264,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.0,
|
| 49 |
+
"eval_accuracy": 0.7439824945295405,
|
| 50 |
+
"eval_loss": 0.6406143307685852,
|
| 51 |
+
"eval_macro_f1": 0.3404118404118404,
|
| 52 |
+
"eval_runtime": 0.3569,
|
| 53 |
+
"eval_samples_per_second": 1280.615,
|
| 54 |
+
"eval_steps_per_second": 81.264,
|
| 55 |
+
"eval_weighted_f1": 0.6529644680629363,
|
| 56 |
+
"step": 129
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.1634241245136188,
|
| 60 |
+
"grad_norm": 30.22808837890625,
|
| 61 |
+
"learning_rate": 1.1550387596899227e-05,
|
| 62 |
+
"loss": 2.5987,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.3579766536964981,
|
| 67 |
+
"grad_norm": 49.629310607910156,
|
| 68 |
+
"learning_rate": 1.3488372093023257e-05,
|
| 69 |
+
"loss": 2.2793,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.5525291828793775,
|
| 74 |
+
"grad_norm": 37.5396728515625,
|
| 75 |
+
"learning_rate": 1.542635658914729e-05,
|
| 76 |
+
"loss": 2.1671,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.7470817120622568,
|
| 81 |
+
"grad_norm": 43.743614196777344,
|
| 82 |
+
"learning_rate": 1.736434108527132e-05,
|
| 83 |
+
"loss": 1.7104,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.9416342412451362,
|
| 88 |
+
"grad_norm": 30.66181182861328,
|
| 89 |
+
"learning_rate": 1.9302325581395353e-05,
|
| 90 |
+
"loss": 1.5832,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_accuracy": 0.8971553610503282,
|
| 96 |
+
"eval_loss": 0.3080006241798401,
|
| 97 |
+
"eval_macro_f1": 0.752348182193891,
|
| 98 |
+
"eval_runtime": 0.3385,
|
| 99 |
+
"eval_samples_per_second": 1350.148,
|
| 100 |
+
"eval_steps_per_second": 85.677,
|
| 101 |
+
"eval_weighted_f1": 0.8857646803418038,
|
| 102 |
+
"step": 258
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.132295719844358,
|
| 106 |
+
"grad_norm": 58.23406982421875,
|
| 107 |
+
"learning_rate": 1.9862187769164516e-05,
|
| 108 |
+
"loss": 1.2724,
|
| 109 |
+
"step": 275
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3268482490272375,
|
| 113 |
+
"grad_norm": 72.3504638671875,
|
| 114 |
+
"learning_rate": 1.9646856158484065e-05,
|
| 115 |
+
"loss": 1.3294,
|
| 116 |
+
"step": 300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.5214007782101167,
|
| 120 |
+
"grad_norm": 19.157331466674805,
|
| 121 |
+
"learning_rate": 1.943152454780362e-05,
|
| 122 |
+
"loss": 1.13,
|
| 123 |
+
"step": 325
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.7159533073929962,
|
| 127 |
+
"grad_norm": 18.174213409423828,
|
| 128 |
+
"learning_rate": 1.921619293712317e-05,
|
| 129 |
+
"loss": 1.2698,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.9105058365758754,
|
| 134 |
+
"grad_norm": 27.829322814941406,
|
| 135 |
+
"learning_rate": 1.9000861326442723e-05,
|
| 136 |
+
"loss": 1.3123,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.0,
|
| 141 |
+
"eval_accuracy": 0.9015317286652079,
|
| 142 |
+
"eval_loss": 0.2923772931098938,
|
| 143 |
+
"eval_macro_f1": 0.8243207601705326,
|
| 144 |
+
"eval_runtime": 0.3518,
|
| 145 |
+
"eval_samples_per_second": 1298.885,
|
| 146 |
+
"eval_steps_per_second": 82.424,
|
| 147 |
+
"eval_weighted_f1": 0.9017740155900388,
|
| 148 |
+
"step": 387
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.1011673151750974,
|
| 152 |
+
"grad_norm": 26.044574737548828,
|
| 153 |
+
"learning_rate": 1.8785529715762276e-05,
|
| 154 |
+
"loss": 1.2595,
|
| 155 |
+
"step": 400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.2957198443579765,
|
| 159 |
+
"grad_norm": 34.246055603027344,
|
| 160 |
+
"learning_rate": 1.857019810508183e-05,
|
| 161 |
+
"loss": 1.0944,
|
| 162 |
+
"step": 425
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.490272373540856,
|
| 166 |
+
"grad_norm": 21.371212005615234,
|
| 167 |
+
"learning_rate": 1.835486649440138e-05,
|
| 168 |
+
"loss": 0.9907,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.6848249027237356,
|
| 173 |
+
"grad_norm": 39.82756042480469,
|
| 174 |
+
"learning_rate": 1.813953488372093e-05,
|
| 175 |
+
"loss": 1.0339,
|
| 176 |
+
"step": 475
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.8793774319066148,
|
| 180 |
+
"grad_norm": 39.80973434448242,
|
| 181 |
+
"learning_rate": 1.7924203273040483e-05,
|
| 182 |
+
"loss": 1.0511,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.9102844638949672,
|
| 188 |
+
"eval_loss": 0.2663729190826416,
|
| 189 |
+
"eval_macro_f1": 0.836096389908049,
|
| 190 |
+
"eval_runtime": 0.3575,
|
| 191 |
+
"eval_samples_per_second": 1278.369,
|
| 192 |
+
"eval_steps_per_second": 81.122,
|
| 193 |
+
"eval_weighted_f1": 0.9096472174324418,
|
| 194 |
+
"step": 516
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.070038910505836,
|
| 198 |
+
"grad_norm": 5.0994486808776855,
|
| 199 |
+
"learning_rate": 1.7708871662360035e-05,
|
| 200 |
+
"loss": 0.7736,
|
| 201 |
+
"step": 525
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 4.264591439688716,
|
| 205 |
+
"grad_norm": 7.552597522735596,
|
| 206 |
+
"learning_rate": 1.7493540051679588e-05,
|
| 207 |
+
"loss": 0.7887,
|
| 208 |
+
"step": 550
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 4.4591439688715955,
|
| 212 |
+
"grad_norm": 74.06793212890625,
|
| 213 |
+
"learning_rate": 1.727820844099914e-05,
|
| 214 |
+
"loss": 0.7283,
|
| 215 |
+
"step": 575
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.653696498054475,
|
| 219 |
+
"grad_norm": 36.153751373291016,
|
| 220 |
+
"learning_rate": 1.7062876830318693e-05,
|
| 221 |
+
"loss": 0.8634,
|
| 222 |
+
"step": 600
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.848249027237354,
|
| 226 |
+
"grad_norm": 18.001266479492188,
|
| 227 |
+
"learning_rate": 1.6847545219638243e-05,
|
| 228 |
+
"loss": 0.7718,
|
| 229 |
+
"step": 625
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.899343544857768,
|
| 234 |
+
"eval_loss": 0.3496442437171936,
|
| 235 |
+
"eval_macro_f1": 0.8211461567966035,
|
| 236 |
+
"eval_runtime": 0.3452,
|
| 237 |
+
"eval_samples_per_second": 1324.055,
|
| 238 |
+
"eval_steps_per_second": 84.021,
|
| 239 |
+
"eval_weighted_f1": 0.9007052194156034,
|
| 240 |
+
"step": 645
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.038910505836576,
|
| 244 |
+
"grad_norm": 23.73420524597168,
|
| 245 |
+
"learning_rate": 1.6632213608957795e-05,
|
| 246 |
+
"loss": 0.9499,
|
| 247 |
+
"step": 650
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.233463035019455,
|
| 251 |
+
"grad_norm": 2.291680335998535,
|
| 252 |
+
"learning_rate": 1.6416881998277348e-05,
|
| 253 |
+
"loss": 0.6049,
|
| 254 |
+
"step": 675
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.428015564202335,
|
| 258 |
+
"grad_norm": 35.31270217895508,
|
| 259 |
+
"learning_rate": 1.62015503875969e-05,
|
| 260 |
+
"loss": 0.6768,
|
| 261 |
+
"step": 700
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.622568093385214,
|
| 265 |
+
"grad_norm": 77.3118667602539,
|
| 266 |
+
"learning_rate": 1.5986218776916453e-05,
|
| 267 |
+
"loss": 0.5712,
|
| 268 |
+
"step": 725
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.817120622568093,
|
| 272 |
+
"grad_norm": 44.21165466308594,
|
| 273 |
+
"learning_rate": 1.5770887166236006e-05,
|
| 274 |
+
"loss": 0.4677,
|
| 275 |
+
"step": 750
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.0,
|
| 279 |
+
"eval_accuracy": 0.9080962800875274,
|
| 280 |
+
"eval_loss": 0.36534589529037476,
|
| 281 |
+
"eval_macro_f1": 0.841231858511076,
|
| 282 |
+
"eval_runtime": 0.3687,
|
| 283 |
+
"eval_samples_per_second": 1239.554,
|
| 284 |
+
"eval_steps_per_second": 78.659,
|
| 285 |
+
"eval_weighted_f1": 0.9093563950857859,
|
| 286 |
+
"step": 774
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 6.007782101167315,
|
| 290 |
+
"grad_norm": 25.171833038330078,
|
| 291 |
+
"learning_rate": 1.555555555555556e-05,
|
| 292 |
+
"loss": 0.8539,
|
| 293 |
+
"step": 775
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 6.202334630350195,
|
| 297 |
+
"grad_norm": 65.7700424194336,
|
| 298 |
+
"learning_rate": 1.534022394487511e-05,
|
| 299 |
+
"loss": 0.4301,
|
| 300 |
+
"step": 800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 6.396887159533074,
|
| 304 |
+
"grad_norm": 31.40825653076172,
|
| 305 |
+
"learning_rate": 1.512489233419466e-05,
|
| 306 |
+
"loss": 0.5198,
|
| 307 |
+
"step": 825
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 6.591439688715953,
|
| 311 |
+
"grad_norm": 52.697208404541016,
|
| 312 |
+
"learning_rate": 1.4909560723514215e-05,
|
| 313 |
+
"loss": 0.5081,
|
| 314 |
+
"step": 850
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 6.785992217898833,
|
| 318 |
+
"grad_norm": 1.0973656177520752,
|
| 319 |
+
"learning_rate": 1.4694229112833764e-05,
|
| 320 |
+
"loss": 0.6172,
|
| 321 |
+
"step": 875
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.980544747081712,
|
| 325 |
+
"grad_norm": 19.881786346435547,
|
| 326 |
+
"learning_rate": 1.4478897502153318e-05,
|
| 327 |
+
"loss": 0.5367,
|
| 328 |
+
"step": 900
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 7.0,
|
| 332 |
+
"eval_accuracy": 0.9080962800875274,
|
| 333 |
+
"eval_loss": 0.41105344891548157,
|
| 334 |
+
"eval_macro_f1": 0.8381604261356804,
|
| 335 |
+
"eval_runtime": 0.3444,
|
| 336 |
+
"eval_samples_per_second": 1326.814,
|
| 337 |
+
"eval_steps_per_second": 84.196,
|
| 338 |
+
"eval_weighted_f1": 0.9076124556139362,
|
| 339 |
+
"step": 903
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 7.171206225680934,
|
| 343 |
+
"grad_norm": 102.81494903564453,
|
| 344 |
+
"learning_rate": 1.426356589147287e-05,
|
| 345 |
+
"loss": 0.4679,
|
| 346 |
+
"step": 925
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.365758754863813,
|
| 350 |
+
"grad_norm": 13.090529441833496,
|
| 351 |
+
"learning_rate": 1.4048234280792422e-05,
|
| 352 |
+
"loss": 0.3767,
|
| 353 |
+
"step": 950
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 7.560311284046692,
|
| 357 |
+
"grad_norm": 20.91201400756836,
|
| 358 |
+
"learning_rate": 1.3832902670111974e-05,
|
| 359 |
+
"loss": 0.6291,
|
| 360 |
+
"step": 975
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 7.754863813229572,
|
| 364 |
+
"grad_norm": 40.23540496826172,
|
| 365 |
+
"learning_rate": 1.3617571059431525e-05,
|
| 366 |
+
"loss": 0.3955,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.9494163424124515,
|
| 371 |
+
"grad_norm": 46.20268249511719,
|
| 372 |
+
"learning_rate": 1.3402239448751078e-05,
|
| 373 |
+
"loss": 0.5452,
|
| 374 |
+
"step": 1025
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.0,
|
| 378 |
+
"eval_accuracy": 0.9146608315098468,
|
| 379 |
+
"eval_loss": 0.45139989256858826,
|
| 380 |
+
"eval_macro_f1": 0.8541751239520002,
|
| 381 |
+
"eval_runtime": 0.3432,
|
| 382 |
+
"eval_samples_per_second": 1331.622,
|
| 383 |
+
"eval_steps_per_second": 84.501,
|
| 384 |
+
"eval_weighted_f1": 0.9143449792137548,
|
| 385 |
+
"step": 1032
|
| 386 |
+
}
|
| 387 |
+
],
|
| 388 |
+
"logging_steps": 25,
|
| 389 |
+
"max_steps": 2580,
|
| 390 |
+
"num_input_tokens_seen": 0,
|
| 391 |
+
"num_train_epochs": 20,
|
| 392 |
+
"save_steps": 500,
|
| 393 |
+
"stateful_callbacks": {
|
| 394 |
+
"EarlyStoppingCallback": {
|
| 395 |
+
"args": {
|
| 396 |
+
"early_stopping_patience": 3,
|
| 397 |
+
"early_stopping_threshold": 0.0
|
| 398 |
+
},
|
| 399 |
+
"attributes": {
|
| 400 |
+
"early_stopping_patience_counter": 0
|
| 401 |
+
}
|
| 402 |
+
},
|
| 403 |
+
"TrainerControl": {
|
| 404 |
+
"args": {
|
| 405 |
+
"should_epoch_stop": false,
|
| 406 |
+
"should_evaluate": false,
|
| 407 |
+
"should_log": false,
|
| 408 |
+
"should_save": true,
|
| 409 |
+
"should_training_stop": false
|
| 410 |
+
},
|
| 411 |
+
"attributes": {}
|
| 412 |
+
}
|
| 413 |
+
},
|
| 414 |
+
"total_flos": 0.0,
|
| 415 |
+
"train_batch_size": 8,
|
| 416 |
+
"trial_name": null,
|
| 417 |
+
"trial_params": null
|
| 418 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1419/trainer_state.json
ADDED
|
@@ -0,0 +1,556 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1032,
|
| 3 |
+
"best_metric": 0.8541751239520002,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-uit_vsfc-bamibert-topic-domain-01-s44-20260827T163455Z-b6645c77/trainer/checkpoint-1032",
|
| 5 |
+
"epoch": 11.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1419,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.19455252918287938,
|
| 14 |
+
"grad_norm": 67.18620300292969,
|
| 15 |
+
"learning_rate": 1.86046511627907e-06,
|
| 16 |
+
"loss": 4.1527,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.38910505836575876,
|
| 21 |
+
"grad_norm": 20.784238815307617,
|
| 22 |
+
"learning_rate": 3.798449612403101e-06,
|
| 23 |
+
"loss": 3.2703,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.5836575875486382,
|
| 28 |
+
"grad_norm": 36.513999938964844,
|
| 29 |
+
"learning_rate": 5.736434108527133e-06,
|
| 30 |
+
"loss": 2.9434,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.7782101167315175,
|
| 35 |
+
"grad_norm": 30.496740341186523,
|
| 36 |
+
"learning_rate": 7.674418604651164e-06,
|
| 37 |
+
"loss": 2.9142,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.9727626459143969,
|
| 42 |
+
"grad_norm": 21.509984970092773,
|
| 43 |
+
"learning_rate": 9.612403100775196e-06,
|
| 44 |
+
"loss": 2.8264,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.0,
|
| 49 |
+
"eval_accuracy": 0.7439824945295405,
|
| 50 |
+
"eval_loss": 0.6406143307685852,
|
| 51 |
+
"eval_macro_f1": 0.3404118404118404,
|
| 52 |
+
"eval_runtime": 0.3569,
|
| 53 |
+
"eval_samples_per_second": 1280.615,
|
| 54 |
+
"eval_steps_per_second": 81.264,
|
| 55 |
+
"eval_weighted_f1": 0.6529644680629363,
|
| 56 |
+
"step": 129
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.1634241245136188,
|
| 60 |
+
"grad_norm": 30.22808837890625,
|
| 61 |
+
"learning_rate": 1.1550387596899227e-05,
|
| 62 |
+
"loss": 2.5987,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.3579766536964981,
|
| 67 |
+
"grad_norm": 49.629310607910156,
|
| 68 |
+
"learning_rate": 1.3488372093023257e-05,
|
| 69 |
+
"loss": 2.2793,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.5525291828793775,
|
| 74 |
+
"grad_norm": 37.5396728515625,
|
| 75 |
+
"learning_rate": 1.542635658914729e-05,
|
| 76 |
+
"loss": 2.1671,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.7470817120622568,
|
| 81 |
+
"grad_norm": 43.743614196777344,
|
| 82 |
+
"learning_rate": 1.736434108527132e-05,
|
| 83 |
+
"loss": 1.7104,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.9416342412451362,
|
| 88 |
+
"grad_norm": 30.66181182861328,
|
| 89 |
+
"learning_rate": 1.9302325581395353e-05,
|
| 90 |
+
"loss": 1.5832,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 2.0,
|
| 95 |
+
"eval_accuracy": 0.8971553610503282,
|
| 96 |
+
"eval_loss": 0.3080006241798401,
|
| 97 |
+
"eval_macro_f1": 0.752348182193891,
|
| 98 |
+
"eval_runtime": 0.3385,
|
| 99 |
+
"eval_samples_per_second": 1350.148,
|
| 100 |
+
"eval_steps_per_second": 85.677,
|
| 101 |
+
"eval_weighted_f1": 0.8857646803418038,
|
| 102 |
+
"step": 258
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.132295719844358,
|
| 106 |
+
"grad_norm": 58.23406982421875,
|
| 107 |
+
"learning_rate": 1.9862187769164516e-05,
|
| 108 |
+
"loss": 1.2724,
|
| 109 |
+
"step": 275
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.3268482490272375,
|
| 113 |
+
"grad_norm": 72.3504638671875,
|
| 114 |
+
"learning_rate": 1.9646856158484065e-05,
|
| 115 |
+
"loss": 1.3294,
|
| 116 |
+
"step": 300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.5214007782101167,
|
| 120 |
+
"grad_norm": 19.157331466674805,
|
| 121 |
+
"learning_rate": 1.943152454780362e-05,
|
| 122 |
+
"loss": 1.13,
|
| 123 |
+
"step": 325
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.7159533073929962,
|
| 127 |
+
"grad_norm": 18.174213409423828,
|
| 128 |
+
"learning_rate": 1.921619293712317e-05,
|
| 129 |
+
"loss": 1.2698,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.9105058365758754,
|
| 134 |
+
"grad_norm": 27.829322814941406,
|
| 135 |
+
"learning_rate": 1.9000861326442723e-05,
|
| 136 |
+
"loss": 1.3123,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.0,
|
| 141 |
+
"eval_accuracy": 0.9015317286652079,
|
| 142 |
+
"eval_loss": 0.2923772931098938,
|
| 143 |
+
"eval_macro_f1": 0.8243207601705326,
|
| 144 |
+
"eval_runtime": 0.3518,
|
| 145 |
+
"eval_samples_per_second": 1298.885,
|
| 146 |
+
"eval_steps_per_second": 82.424,
|
| 147 |
+
"eval_weighted_f1": 0.9017740155900388,
|
| 148 |
+
"step": 387
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 3.1011673151750974,
|
| 152 |
+
"grad_norm": 26.044574737548828,
|
| 153 |
+
"learning_rate": 1.8785529715762276e-05,
|
| 154 |
+
"loss": 1.2595,
|
| 155 |
+
"step": 400
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.2957198443579765,
|
| 159 |
+
"grad_norm": 34.246055603027344,
|
| 160 |
+
"learning_rate": 1.857019810508183e-05,
|
| 161 |
+
"loss": 1.0944,
|
| 162 |
+
"step": 425
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.490272373540856,
|
| 166 |
+
"grad_norm": 21.371212005615234,
|
| 167 |
+
"learning_rate": 1.835486649440138e-05,
|
| 168 |
+
"loss": 0.9907,
|
| 169 |
+
"step": 450
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.6848249027237356,
|
| 173 |
+
"grad_norm": 39.82756042480469,
|
| 174 |
+
"learning_rate": 1.813953488372093e-05,
|
| 175 |
+
"loss": 1.0339,
|
| 176 |
+
"step": 475
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.8793774319066148,
|
| 180 |
+
"grad_norm": 39.80973434448242,
|
| 181 |
+
"learning_rate": 1.7924203273040483e-05,
|
| 182 |
+
"loss": 1.0511,
|
| 183 |
+
"step": 500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.0,
|
| 187 |
+
"eval_accuracy": 0.9102844638949672,
|
| 188 |
+
"eval_loss": 0.2663729190826416,
|
| 189 |
+
"eval_macro_f1": 0.836096389908049,
|
| 190 |
+
"eval_runtime": 0.3575,
|
| 191 |
+
"eval_samples_per_second": 1278.369,
|
| 192 |
+
"eval_steps_per_second": 81.122,
|
| 193 |
+
"eval_weighted_f1": 0.9096472174324418,
|
| 194 |
+
"step": 516
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 4.070038910505836,
|
| 198 |
+
"grad_norm": 5.0994486808776855,
|
| 199 |
+
"learning_rate": 1.7708871662360035e-05,
|
| 200 |
+
"loss": 0.7736,
|
| 201 |
+
"step": 525
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 4.264591439688716,
|
| 205 |
+
"grad_norm": 7.552597522735596,
|
| 206 |
+
"learning_rate": 1.7493540051679588e-05,
|
| 207 |
+
"loss": 0.7887,
|
| 208 |
+
"step": 550
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 4.4591439688715955,
|
| 212 |
+
"grad_norm": 74.06793212890625,
|
| 213 |
+
"learning_rate": 1.727820844099914e-05,
|
| 214 |
+
"loss": 0.7283,
|
| 215 |
+
"step": 575
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.653696498054475,
|
| 219 |
+
"grad_norm": 36.153751373291016,
|
| 220 |
+
"learning_rate": 1.7062876830318693e-05,
|
| 221 |
+
"loss": 0.8634,
|
| 222 |
+
"step": 600
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.848249027237354,
|
| 226 |
+
"grad_norm": 18.001266479492188,
|
| 227 |
+
"learning_rate": 1.6847545219638243e-05,
|
| 228 |
+
"loss": 0.7718,
|
| 229 |
+
"step": 625
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 5.0,
|
| 233 |
+
"eval_accuracy": 0.899343544857768,
|
| 234 |
+
"eval_loss": 0.3496442437171936,
|
| 235 |
+
"eval_macro_f1": 0.8211461567966035,
|
| 236 |
+
"eval_runtime": 0.3452,
|
| 237 |
+
"eval_samples_per_second": 1324.055,
|
| 238 |
+
"eval_steps_per_second": 84.021,
|
| 239 |
+
"eval_weighted_f1": 0.9007052194156034,
|
| 240 |
+
"step": 645
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.038910505836576,
|
| 244 |
+
"grad_norm": 23.73420524597168,
|
| 245 |
+
"learning_rate": 1.6632213608957795e-05,
|
| 246 |
+
"loss": 0.9499,
|
| 247 |
+
"step": 650
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.233463035019455,
|
| 251 |
+
"grad_norm": 2.291680335998535,
|
| 252 |
+
"learning_rate": 1.6416881998277348e-05,
|
| 253 |
+
"loss": 0.6049,
|
| 254 |
+
"step": 675
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.428015564202335,
|
| 258 |
+
"grad_norm": 35.31270217895508,
|
| 259 |
+
"learning_rate": 1.62015503875969e-05,
|
| 260 |
+
"loss": 0.6768,
|
| 261 |
+
"step": 700
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.622568093385214,
|
| 265 |
+
"grad_norm": 77.3118667602539,
|
| 266 |
+
"learning_rate": 1.5986218776916453e-05,
|
| 267 |
+
"loss": 0.5712,
|
| 268 |
+
"step": 725
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.817120622568093,
|
| 272 |
+
"grad_norm": 44.21165466308594,
|
| 273 |
+
"learning_rate": 1.5770887166236006e-05,
|
| 274 |
+
"loss": 0.4677,
|
| 275 |
+
"step": 750
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.0,
|
| 279 |
+
"eval_accuracy": 0.9080962800875274,
|
| 280 |
+
"eval_loss": 0.36534589529037476,
|
| 281 |
+
"eval_macro_f1": 0.841231858511076,
|
| 282 |
+
"eval_runtime": 0.3687,
|
| 283 |
+
"eval_samples_per_second": 1239.554,
|
| 284 |
+
"eval_steps_per_second": 78.659,
|
| 285 |
+
"eval_weighted_f1": 0.9093563950857859,
|
| 286 |
+
"step": 774
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 6.007782101167315,
|
| 290 |
+
"grad_norm": 25.171833038330078,
|
| 291 |
+
"learning_rate": 1.555555555555556e-05,
|
| 292 |
+
"loss": 0.8539,
|
| 293 |
+
"step": 775
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 6.202334630350195,
|
| 297 |
+
"grad_norm": 65.7700424194336,
|
| 298 |
+
"learning_rate": 1.534022394487511e-05,
|
| 299 |
+
"loss": 0.4301,
|
| 300 |
+
"step": 800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 6.396887159533074,
|
| 304 |
+
"grad_norm": 31.40825653076172,
|
| 305 |
+
"learning_rate": 1.512489233419466e-05,
|
| 306 |
+
"loss": 0.5198,
|
| 307 |
+
"step": 825
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 6.591439688715953,
|
| 311 |
+
"grad_norm": 52.697208404541016,
|
| 312 |
+
"learning_rate": 1.4909560723514215e-05,
|
| 313 |
+
"loss": 0.5081,
|
| 314 |
+
"step": 850
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 6.785992217898833,
|
| 318 |
+
"grad_norm": 1.0973656177520752,
|
| 319 |
+
"learning_rate": 1.4694229112833764e-05,
|
| 320 |
+
"loss": 0.6172,
|
| 321 |
+
"step": 875
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 6.980544747081712,
|
| 325 |
+
"grad_norm": 19.881786346435547,
|
| 326 |
+
"learning_rate": 1.4478897502153318e-05,
|
| 327 |
+
"loss": 0.5367,
|
| 328 |
+
"step": 900
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 7.0,
|
| 332 |
+
"eval_accuracy": 0.9080962800875274,
|
| 333 |
+
"eval_loss": 0.41105344891548157,
|
| 334 |
+
"eval_macro_f1": 0.8381604261356804,
|
| 335 |
+
"eval_runtime": 0.3444,
|
| 336 |
+
"eval_samples_per_second": 1326.814,
|
| 337 |
+
"eval_steps_per_second": 84.196,
|
| 338 |
+
"eval_weighted_f1": 0.9076124556139362,
|
| 339 |
+
"step": 903
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 7.171206225680934,
|
| 343 |
+
"grad_norm": 102.81494903564453,
|
| 344 |
+
"learning_rate": 1.426356589147287e-05,
|
| 345 |
+
"loss": 0.4679,
|
| 346 |
+
"step": 925
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.365758754863813,
|
| 350 |
+
"grad_norm": 13.090529441833496,
|
| 351 |
+
"learning_rate": 1.4048234280792422e-05,
|
| 352 |
+
"loss": 0.3767,
|
| 353 |
+
"step": 950
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 7.560311284046692,
|
| 357 |
+
"grad_norm": 20.91201400756836,
|
| 358 |
+
"learning_rate": 1.3832902670111974e-05,
|
| 359 |
+
"loss": 0.6291,
|
| 360 |
+
"step": 975
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 7.754863813229572,
|
| 364 |
+
"grad_norm": 40.23540496826172,
|
| 365 |
+
"learning_rate": 1.3617571059431525e-05,
|
| 366 |
+
"loss": 0.3955,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.9494163424124515,
|
| 371 |
+
"grad_norm": 46.20268249511719,
|
| 372 |
+
"learning_rate": 1.3402239448751078e-05,
|
| 373 |
+
"loss": 0.5452,
|
| 374 |
+
"step": 1025
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.0,
|
| 378 |
+
"eval_accuracy": 0.9146608315098468,
|
| 379 |
+
"eval_loss": 0.45139989256858826,
|
| 380 |
+
"eval_macro_f1": 0.8541751239520002,
|
| 381 |
+
"eval_runtime": 0.3432,
|
| 382 |
+
"eval_samples_per_second": 1331.622,
|
| 383 |
+
"eval_steps_per_second": 84.501,
|
| 384 |
+
"eval_weighted_f1": 0.9143449792137548,
|
| 385 |
+
"step": 1032
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 8.140077821011673,
|
| 389 |
+
"grad_norm": 35.48772048950195,
|
| 390 |
+
"learning_rate": 1.318690783807063e-05,
|
| 391 |
+
"loss": 0.317,
|
| 392 |
+
"step": 1050
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 8.334630350194553,
|
| 396 |
+
"grad_norm": 0.7826412916183472,
|
| 397 |
+
"learning_rate": 1.2971576227390181e-05,
|
| 398 |
+
"loss": 0.3626,
|
| 399 |
+
"step": 1075
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 8.529182879377432,
|
| 403 |
+
"grad_norm": 17.235637664794922,
|
| 404 |
+
"learning_rate": 1.2756244616709734e-05,
|
| 405 |
+
"loss": 0.3748,
|
| 406 |
+
"step": 1100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 8.72373540856031,
|
| 410 |
+
"grad_norm": 32.62348175048828,
|
| 411 |
+
"learning_rate": 1.2540913006029287e-05,
|
| 412 |
+
"loss": 0.4235,
|
| 413 |
+
"step": 1125
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 8.918287937743191,
|
| 417 |
+
"grad_norm": 2.9188437461853027,
|
| 418 |
+
"learning_rate": 1.2325581395348838e-05,
|
| 419 |
+
"loss": 0.616,
|
| 420 |
+
"step": 1150
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 9.0,
|
| 424 |
+
"eval_accuracy": 0.9080962800875274,
|
| 425 |
+
"eval_loss": 0.49155890941619873,
|
| 426 |
+
"eval_macro_f1": 0.8320707070707071,
|
| 427 |
+
"eval_runtime": 0.3494,
|
| 428 |
+
"eval_samples_per_second": 1307.831,
|
| 429 |
+
"eval_steps_per_second": 82.991,
|
| 430 |
+
"eval_weighted_f1": 0.9066481476218894,
|
| 431 |
+
"step": 1161
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 9.108949416342412,
|
| 435 |
+
"grad_norm": 31.342397689819336,
|
| 436 |
+
"learning_rate": 1.211024978466839e-05,
|
| 437 |
+
"loss": 0.5013,
|
| 438 |
+
"step": 1175
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 9.303501945525293,
|
| 442 |
+
"grad_norm": 2.2257773876190186,
|
| 443 |
+
"learning_rate": 1.1894918173987941e-05,
|
| 444 |
+
"loss": 0.2763,
|
| 445 |
+
"step": 1200
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 9.498054474708171,
|
| 449 |
+
"grad_norm": 11.369057655334473,
|
| 450 |
+
"learning_rate": 1.1679586563307494e-05,
|
| 451 |
+
"loss": 0.2969,
|
| 452 |
+
"step": 1225
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 9.69260700389105,
|
| 456 |
+
"grad_norm": 0.32061201333999634,
|
| 457 |
+
"learning_rate": 1.1464254952627048e-05,
|
| 458 |
+
"loss": 0.3263,
|
| 459 |
+
"step": 1250
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 9.88715953307393,
|
| 463 |
+
"grad_norm": 46.53280258178711,
|
| 464 |
+
"learning_rate": 1.1248923341946599e-05,
|
| 465 |
+
"loss": 0.3745,
|
| 466 |
+
"step": 1275
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 10.0,
|
| 470 |
+
"eval_accuracy": 0.9102844638949672,
|
| 471 |
+
"eval_loss": 0.5093547105789185,
|
| 472 |
+
"eval_macro_f1": 0.8360237232619854,
|
| 473 |
+
"eval_runtime": 0.349,
|
| 474 |
+
"eval_samples_per_second": 1309.414,
|
| 475 |
+
"eval_steps_per_second": 83.092,
|
| 476 |
+
"eval_weighted_f1": 0.9077637075905489,
|
| 477 |
+
"step": 1290
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 10.077821011673151,
|
| 481 |
+
"grad_norm": 0.25982579588890076,
|
| 482 |
+
"learning_rate": 1.1033591731266152e-05,
|
| 483 |
+
"loss": 0.2259,
|
| 484 |
+
"step": 1300
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 10.272373540856032,
|
| 488 |
+
"grad_norm": 38.89549255371094,
|
| 489 |
+
"learning_rate": 1.0818260120585703e-05,
|
| 490 |
+
"loss": 0.1822,
|
| 491 |
+
"step": 1325
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 10.46692607003891,
|
| 495 |
+
"grad_norm": 0.11628062278032303,
|
| 496 |
+
"learning_rate": 1.0602928509905255e-05,
|
| 497 |
+
"loss": 0.2292,
|
| 498 |
+
"step": 1350
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 10.66147859922179,
|
| 502 |
+
"grad_norm": 6.040037631988525,
|
| 503 |
+
"learning_rate": 1.0387596899224808e-05,
|
| 504 |
+
"loss": 0.3292,
|
| 505 |
+
"step": 1375
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 10.85603112840467,
|
| 509 |
+
"grad_norm": 5.655516147613525,
|
| 510 |
+
"learning_rate": 1.0172265288544359e-05,
|
| 511 |
+
"loss": 0.2277,
|
| 512 |
+
"step": 1400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 11.0,
|
| 516 |
+
"eval_accuracy": 0.9080962800875274,
|
| 517 |
+
"eval_loss": 0.5406879186630249,
|
| 518 |
+
"eval_macro_f1": 0.8486245698454643,
|
| 519 |
+
"eval_runtime": 0.3954,
|
| 520 |
+
"eval_samples_per_second": 1155.854,
|
| 521 |
+
"eval_steps_per_second": 73.347,
|
| 522 |
+
"eval_weighted_f1": 0.9088175844059387,
|
| 523 |
+
"step": 1419
|
| 524 |
+
}
|
| 525 |
+
],
|
| 526 |
+
"logging_steps": 25,
|
| 527 |
+
"max_steps": 2580,
|
| 528 |
+
"num_input_tokens_seen": 0,
|
| 529 |
+
"num_train_epochs": 20,
|
| 530 |
+
"save_steps": 500,
|
| 531 |
+
"stateful_callbacks": {
|
| 532 |
+
"EarlyStoppingCallback": {
|
| 533 |
+
"args": {
|
| 534 |
+
"early_stopping_patience": 3,
|
| 535 |
+
"early_stopping_threshold": 0.0
|
| 536 |
+
},
|
| 537 |
+
"attributes": {
|
| 538 |
+
"early_stopping_patience_counter": 3
|
| 539 |
+
}
|
| 540 |
+
},
|
| 541 |
+
"TrainerControl": {
|
| 542 |
+
"args": {
|
| 543 |
+
"should_epoch_stop": false,
|
| 544 |
+
"should_evaluate": false,
|
| 545 |
+
"should_log": false,
|
| 546 |
+
"should_save": true,
|
| 547 |
+
"should_training_stop": true
|
| 548 |
+
},
|
| 549 |
+
"attributes": {}
|
| 550 |
+
}
|
| 551 |
+
},
|
| 552 |
+
"total_flos": 0.0,
|
| 553 |
+
"train_batch_size": 8,
|
| 554 |
+
"trial_name": null,
|
| 555 |
+
"trial_params": null
|
| 556 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-02-s42-20260827T155652Z-f97f2036/trainer/checkpoint-2160/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 |
+
}
|
ood-uit_vsfc-bamibert-topic-domain-02-s42-20260827T155652Z-f97f2036/trainer/checkpoint-2160/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|