Add files using upload-large-folder tool
Browse files- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/trainer_state.json +484 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/trainer_state.json +671 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/trainer_state.json +558 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/trainer_state.json +752 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/trainer_state.json +971 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/trainer_state.json +1158 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/trainer_state.json +498 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/trainer_state.json +304 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/trainer_state.json +678 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/trainer_state.json +865 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/trainer_state.json +463 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/trainer_state.json +643 -0
- ood-vifactcheck-bamibert-topic-domain-06-s44-20260828T210529Z-496d6871/trainer/checkpoint-920/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-topic-domain-06-s44-20260828T210529Z-496d6871/trainer/checkpoint-920/tokenizer_config.json +55 -0
ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/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-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/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-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309/trainer_state.json
ADDED
|
@@ -0,0 +1,484 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1309,
|
| 3 |
+
"best_metric": 0.8556166390203034,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1309,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13368983957219252,
|
| 14 |
+
"grad_norm": 14.998248100280762,
|
| 15 |
+
"learning_rate": 1.2834224598930483e-06,
|
| 16 |
+
"loss": 4.4082,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.26737967914438504,
|
| 21 |
+
"grad_norm": 26.6628360748291,
|
| 22 |
+
"learning_rate": 2.6203208556149735e-06,
|
| 23 |
+
"loss": 4.4278,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.40106951871657753,
|
| 28 |
+
"grad_norm": 21.630096435546875,
|
| 29 |
+
"learning_rate": 3.957219251336899e-06,
|
| 30 |
+
"loss": 4.4064,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5347593582887701,
|
| 35 |
+
"grad_norm": 9.354437828063965,
|
| 36 |
+
"learning_rate": 5.294117647058824e-06,
|
| 37 |
+
"loss": 4.3932,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6684491978609626,
|
| 42 |
+
"grad_norm": 16.527864456176758,
|
| 43 |
+
"learning_rate": 6.631016042780749e-06,
|
| 44 |
+
"loss": 4.3986,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8021390374331551,
|
| 49 |
+
"grad_norm": 8.238319396972656,
|
| 50 |
+
"learning_rate": 7.967914438502674e-06,
|
| 51 |
+
"loss": 4.3669,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9358288770053476,
|
| 56 |
+
"grad_norm": 13.496139526367188,
|
| 57 |
+
"learning_rate": 9.3048128342246e-06,
|
| 58 |
+
"loss": 4.2945,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.46153846153846156,
|
| 64 |
+
"eval_loss": 1.048524260520935,
|
| 65 |
+
"eval_macro_f1": 0.36981564100748293,
|
| 66 |
+
"eval_runtime": 1.0626,
|
| 67 |
+
"eval_samples_per_second": 538.288,
|
| 68 |
+
"eval_steps_per_second": 33.878,
|
| 69 |
+
"eval_weighted_f1": 0.37294304024705555,
|
| 70 |
+
"step": 187
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0695187165775402,
|
| 74 |
+
"grad_norm": 13.78651237487793,
|
| 75 |
+
"learning_rate": 1.0641711229946525e-05,
|
| 76 |
+
"loss": 4.2446,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2032085561497325,
|
| 81 |
+
"grad_norm": 20.475452423095703,
|
| 82 |
+
"learning_rate": 1.197860962566845e-05,
|
| 83 |
+
"loss": 4.0271,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3368983957219251,
|
| 88 |
+
"grad_norm": 35.337398529052734,
|
| 89 |
+
"learning_rate": 1.3315508021390375e-05,
|
| 90 |
+
"loss": 3.6094,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4705882352941178,
|
| 95 |
+
"grad_norm": 50.861026763916016,
|
| 96 |
+
"learning_rate": 1.4652406417112302e-05,
|
| 97 |
+
"loss": 3.2553,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6042780748663101,
|
| 102 |
+
"grad_norm": 48.600669860839844,
|
| 103 |
+
"learning_rate": 1.5989304812834226e-05,
|
| 104 |
+
"loss": 2.8483,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7379679144385025,
|
| 109 |
+
"grad_norm": 40.28901290893555,
|
| 110 |
+
"learning_rate": 1.7326203208556153e-05,
|
| 111 |
+
"loss": 2.8516,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.8716577540106951,
|
| 116 |
+
"grad_norm": 44.710628509521484,
|
| 117 |
+
"learning_rate": 1.8663101604278075e-05,
|
| 118 |
+
"loss": 2.7024,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7762237762237763,
|
| 124 |
+
"eval_loss": 0.5299336910247803,
|
| 125 |
+
"eval_macro_f1": 0.7808698231207254,
|
| 126 |
+
"eval_runtime": 1.0373,
|
| 127 |
+
"eval_samples_per_second": 551.44,
|
| 128 |
+
"eval_steps_per_second": 34.706,
|
| 129 |
+
"eval_weighted_f1": 0.7809045851263942,
|
| 130 |
+
"step": 374
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0053475935828877,
|
| 134 |
+
"grad_norm": 45.8042106628418,
|
| 135 |
+
"learning_rate": 2e-05,
|
| 136 |
+
"loss": 2.4003,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.1390374331550803,
|
| 141 |
+
"grad_norm": 46.33683395385742,
|
| 142 |
+
"learning_rate": 1.9851455733808676e-05,
|
| 143 |
+
"loss": 1.8141,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.2727272727272725,
|
| 148 |
+
"grad_norm": 44.520469665527344,
|
| 149 |
+
"learning_rate": 1.9702911467617353e-05,
|
| 150 |
+
"loss": 2.1766,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.406417112299465,
|
| 155 |
+
"grad_norm": 64.53562927246094,
|
| 156 |
+
"learning_rate": 1.9554367201426024e-05,
|
| 157 |
+
"loss": 2.2113,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.5401069518716577,
|
| 162 |
+
"grad_norm": 53.96901321411133,
|
| 163 |
+
"learning_rate": 1.94058229352347e-05,
|
| 164 |
+
"loss": 2.0653,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.6737967914438503,
|
| 169 |
+
"grad_norm": 78.2419662475586,
|
| 170 |
+
"learning_rate": 1.9257278669043376e-05,
|
| 171 |
+
"loss": 2.0692,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.807486631016043,
|
| 176 |
+
"grad_norm": 33.31931686401367,
|
| 177 |
+
"learning_rate": 1.9108734402852053e-05,
|
| 178 |
+
"loss": 1.8708,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.9411764705882355,
|
| 183 |
+
"grad_norm": 23.48499298095703,
|
| 184 |
+
"learning_rate": 1.8960190136660727e-05,
|
| 185 |
+
"loss": 1.7389,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.0,
|
| 190 |
+
"eval_accuracy": 0.8409090909090909,
|
| 191 |
+
"eval_loss": 0.42494767904281616,
|
| 192 |
+
"eval_macro_f1": 0.8414264010732074,
|
| 193 |
+
"eval_runtime": 1.0734,
|
| 194 |
+
"eval_samples_per_second": 532.885,
|
| 195 |
+
"eval_steps_per_second": 33.538,
|
| 196 |
+
"eval_weighted_f1": 0.8408405475971842,
|
| 197 |
+
"step": 561
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0748663101604277,
|
| 201 |
+
"grad_norm": 44.344139099121094,
|
| 202 |
+
"learning_rate": 1.88116458704694e-05,
|
| 203 |
+
"loss": 1.7572,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.2085561497326203,
|
| 208 |
+
"grad_norm": 29.0012264251709,
|
| 209 |
+
"learning_rate": 1.8663101604278075e-05,
|
| 210 |
+
"loss": 1.5159,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.342245989304813,
|
| 215 |
+
"grad_norm": 38.60652542114258,
|
| 216 |
+
"learning_rate": 1.8514557338086753e-05,
|
| 217 |
+
"loss": 1.2328,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4759358288770055,
|
| 222 |
+
"grad_norm": 29.8499698638916,
|
| 223 |
+
"learning_rate": 1.8366013071895427e-05,
|
| 224 |
+
"loss": 1.4623,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.6096256684491976,
|
| 229 |
+
"grad_norm": 56.83005142211914,
|
| 230 |
+
"learning_rate": 1.82174688057041e-05,
|
| 231 |
+
"loss": 1.5152,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.7433155080213902,
|
| 236 |
+
"grad_norm": 25.463768005371094,
|
| 237 |
+
"learning_rate": 1.8068924539512775e-05,
|
| 238 |
+
"loss": 1.4098,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.877005347593583,
|
| 243 |
+
"grad_norm": 40.18172073364258,
|
| 244 |
+
"learning_rate": 1.7920380273321453e-05,
|
| 245 |
+
"loss": 1.5085,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8426573426573427,
|
| 251 |
+
"eval_loss": 0.43529632687568665,
|
| 252 |
+
"eval_macro_f1": 0.8426902149674427,
|
| 253 |
+
"eval_runtime": 0.9898,
|
| 254 |
+
"eval_samples_per_second": 577.908,
|
| 255 |
+
"eval_steps_per_second": 36.372,
|
| 256 |
+
"eval_weighted_f1": 0.8423229433130422,
|
| 257 |
+
"step": 748
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.010695187165775,
|
| 261 |
+
"grad_norm": 34.088924407958984,
|
| 262 |
+
"learning_rate": 1.7771836007130127e-05,
|
| 263 |
+
"loss": 1.3705,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.144385026737968,
|
| 268 |
+
"grad_norm": 54.056461334228516,
|
| 269 |
+
"learning_rate": 1.76232917409388e-05,
|
| 270 |
+
"loss": 1.1595,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.278074866310161,
|
| 275 |
+
"grad_norm": 40.378910064697266,
|
| 276 |
+
"learning_rate": 1.7474747474747475e-05,
|
| 277 |
+
"loss": 0.954,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.411764705882353,
|
| 282 |
+
"grad_norm": 90.93668365478516,
|
| 283 |
+
"learning_rate": 1.7326203208556153e-05,
|
| 284 |
+
"loss": 0.9105,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.545454545454545,
|
| 289 |
+
"grad_norm": 62.43103790283203,
|
| 290 |
+
"learning_rate": 1.7177658942364827e-05,
|
| 291 |
+
"loss": 0.771,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.6791443850267385,
|
| 296 |
+
"grad_norm": 36.349788665771484,
|
| 297 |
+
"learning_rate": 1.70291146761735e-05,
|
| 298 |
+
"loss": 0.9634,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.81283422459893,
|
| 303 |
+
"grad_norm": 88.49812316894531,
|
| 304 |
+
"learning_rate": 1.6880570409982175e-05,
|
| 305 |
+
"loss": 0.8089,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.946524064171123,
|
| 310 |
+
"grad_norm": 132.47991943359375,
|
| 311 |
+
"learning_rate": 1.6732026143790852e-05,
|
| 312 |
+
"loss": 1.178,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.0,
|
| 317 |
+
"eval_accuracy": 0.8041958041958042,
|
| 318 |
+
"eval_loss": 0.6062424182891846,
|
| 319 |
+
"eval_macro_f1": 0.8062141277126513,
|
| 320 |
+
"eval_runtime": 1.0356,
|
| 321 |
+
"eval_samples_per_second": 552.318,
|
| 322 |
+
"eval_steps_per_second": 34.761,
|
| 323 |
+
"eval_weighted_f1": 0.8067557022741634,
|
| 324 |
+
"step": 935
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.080213903743315,
|
| 328 |
+
"grad_norm": 15.832716941833496,
|
| 329 |
+
"learning_rate": 1.6583481877599526e-05,
|
| 330 |
+
"loss": 0.887,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.213903743315508,
|
| 335 |
+
"grad_norm": 32.42772674560547,
|
| 336 |
+
"learning_rate": 1.64349376114082e-05,
|
| 337 |
+
"loss": 0.631,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.347593582887701,
|
| 342 |
+
"grad_norm": 125.02656555175781,
|
| 343 |
+
"learning_rate": 1.6286393345216875e-05,
|
| 344 |
+
"loss": 0.6536,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.481283422459893,
|
| 349 |
+
"grad_norm": 84.05538177490234,
|
| 350 |
+
"learning_rate": 1.6137849079025552e-05,
|
| 351 |
+
"loss": 0.7163,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.614973262032086,
|
| 356 |
+
"grad_norm": 46.46393585205078,
|
| 357 |
+
"learning_rate": 1.5989304812834226e-05,
|
| 358 |
+
"loss": 0.841,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.748663101604278,
|
| 363 |
+
"grad_norm": 52.25996398925781,
|
| 364 |
+
"learning_rate": 1.58407605466429e-05,
|
| 365 |
+
"loss": 0.6103,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.882352941176471,
|
| 370 |
+
"grad_norm": 70.38349914550781,
|
| 371 |
+
"learning_rate": 1.5692216280451574e-05,
|
| 372 |
+
"loss": 0.4908,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.0,
|
| 377 |
+
"eval_accuracy": 0.8356643356643356,
|
| 378 |
+
"eval_loss": 0.597773551940918,
|
| 379 |
+
"eval_macro_f1": 0.8367203204766657,
|
| 380 |
+
"eval_runtime": 1.0873,
|
| 381 |
+
"eval_samples_per_second": 526.057,
|
| 382 |
+
"eval_steps_per_second": 33.108,
|
| 383 |
+
"eval_weighted_f1": 0.8365573859521537,
|
| 384 |
+
"step": 1122
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.016042780748663,
|
| 388 |
+
"grad_norm": 32.182701110839844,
|
| 389 |
+
"learning_rate": 1.5543672014260252e-05,
|
| 390 |
+
"loss": 0.7634,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.149732620320855,
|
| 395 |
+
"grad_norm": 30.86662483215332,
|
| 396 |
+
"learning_rate": 1.5395127748068926e-05,
|
| 397 |
+
"loss": 0.3263,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.283422459893048,
|
| 402 |
+
"grad_norm": 86.34818267822266,
|
| 403 |
+
"learning_rate": 1.5246583481877602e-05,
|
| 404 |
+
"loss": 0.5872,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.4171122994652405,
|
| 409 |
+
"grad_norm": 17.692935943603516,
|
| 410 |
+
"learning_rate": 1.5098039215686276e-05,
|
| 411 |
+
"loss": 0.4611,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.550802139037433,
|
| 416 |
+
"grad_norm": 53.08113479614258,
|
| 417 |
+
"learning_rate": 1.4949494949494952e-05,
|
| 418 |
+
"loss": 0.5075,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.684491978609626,
|
| 423 |
+
"grad_norm": 63.51148986816406,
|
| 424 |
+
"learning_rate": 1.4800950683303626e-05,
|
| 425 |
+
"loss": 0.5439,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.818181818181818,
|
| 430 |
+
"grad_norm": 71.70096588134766,
|
| 431 |
+
"learning_rate": 1.4652406417112302e-05,
|
| 432 |
+
"loss": 0.6084,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.951871657754011,
|
| 437 |
+
"grad_norm": 44.453033447265625,
|
| 438 |
+
"learning_rate": 1.4503862150920976e-05,
|
| 439 |
+
"loss": 0.4496,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 7.0,
|
| 444 |
+
"eval_accuracy": 0.8548951048951049,
|
| 445 |
+
"eval_loss": 0.6412630081176758,
|
| 446 |
+
"eval_macro_f1": 0.8556166390203034,
|
| 447 |
+
"eval_runtime": 1.0748,
|
| 448 |
+
"eval_samples_per_second": 532.171,
|
| 449 |
+
"eval_steps_per_second": 33.493,
|
| 450 |
+
"eval_weighted_f1": 0.8550553996897675,
|
| 451 |
+
"step": 1309
|
| 452 |
+
}
|
| 453 |
+
],
|
| 454 |
+
"logging_steps": 25,
|
| 455 |
+
"max_steps": 3740,
|
| 456 |
+
"num_input_tokens_seen": 0,
|
| 457 |
+
"num_train_epochs": 20,
|
| 458 |
+
"save_steps": 500,
|
| 459 |
+
"stateful_callbacks": {
|
| 460 |
+
"EarlyStoppingCallback": {
|
| 461 |
+
"args": {
|
| 462 |
+
"early_stopping_patience": 3,
|
| 463 |
+
"early_stopping_threshold": 0.0
|
| 464 |
+
},
|
| 465 |
+
"attributes": {
|
| 466 |
+
"early_stopping_patience_counter": 0
|
| 467 |
+
}
|
| 468 |
+
},
|
| 469 |
+
"TrainerControl": {
|
| 470 |
+
"args": {
|
| 471 |
+
"should_epoch_stop": false,
|
| 472 |
+
"should_evaluate": false,
|
| 473 |
+
"should_log": false,
|
| 474 |
+
"should_save": true,
|
| 475 |
+
"should_training_stop": false
|
| 476 |
+
},
|
| 477 |
+
"attributes": {}
|
| 478 |
+
}
|
| 479 |
+
},
|
| 480 |
+
"total_flos": 0.0,
|
| 481 |
+
"train_batch_size": 8,
|
| 482 |
+
"trial_name": null,
|
| 483 |
+
"trial_params": null
|
| 484 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/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-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/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-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1870/trainer_state.json
ADDED
|
@@ -0,0 +1,671 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1309,
|
| 3 |
+
"best_metric": 0.8556166390203034,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-00-s44-20260828T200115Z-496d6871/trainer/checkpoint-1309",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1870,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13368983957219252,
|
| 14 |
+
"grad_norm": 14.998248100280762,
|
| 15 |
+
"learning_rate": 1.2834224598930483e-06,
|
| 16 |
+
"loss": 4.4082,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.26737967914438504,
|
| 21 |
+
"grad_norm": 26.6628360748291,
|
| 22 |
+
"learning_rate": 2.6203208556149735e-06,
|
| 23 |
+
"loss": 4.4278,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.40106951871657753,
|
| 28 |
+
"grad_norm": 21.630096435546875,
|
| 29 |
+
"learning_rate": 3.957219251336899e-06,
|
| 30 |
+
"loss": 4.4064,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5347593582887701,
|
| 35 |
+
"grad_norm": 9.354437828063965,
|
| 36 |
+
"learning_rate": 5.294117647058824e-06,
|
| 37 |
+
"loss": 4.3932,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6684491978609626,
|
| 42 |
+
"grad_norm": 16.527864456176758,
|
| 43 |
+
"learning_rate": 6.631016042780749e-06,
|
| 44 |
+
"loss": 4.3986,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8021390374331551,
|
| 49 |
+
"grad_norm": 8.238319396972656,
|
| 50 |
+
"learning_rate": 7.967914438502674e-06,
|
| 51 |
+
"loss": 4.3669,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9358288770053476,
|
| 56 |
+
"grad_norm": 13.496139526367188,
|
| 57 |
+
"learning_rate": 9.3048128342246e-06,
|
| 58 |
+
"loss": 4.2945,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.46153846153846156,
|
| 64 |
+
"eval_loss": 1.048524260520935,
|
| 65 |
+
"eval_macro_f1": 0.36981564100748293,
|
| 66 |
+
"eval_runtime": 1.0626,
|
| 67 |
+
"eval_samples_per_second": 538.288,
|
| 68 |
+
"eval_steps_per_second": 33.878,
|
| 69 |
+
"eval_weighted_f1": 0.37294304024705555,
|
| 70 |
+
"step": 187
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0695187165775402,
|
| 74 |
+
"grad_norm": 13.78651237487793,
|
| 75 |
+
"learning_rate": 1.0641711229946525e-05,
|
| 76 |
+
"loss": 4.2446,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2032085561497325,
|
| 81 |
+
"grad_norm": 20.475452423095703,
|
| 82 |
+
"learning_rate": 1.197860962566845e-05,
|
| 83 |
+
"loss": 4.0271,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3368983957219251,
|
| 88 |
+
"grad_norm": 35.337398529052734,
|
| 89 |
+
"learning_rate": 1.3315508021390375e-05,
|
| 90 |
+
"loss": 3.6094,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4705882352941178,
|
| 95 |
+
"grad_norm": 50.861026763916016,
|
| 96 |
+
"learning_rate": 1.4652406417112302e-05,
|
| 97 |
+
"loss": 3.2553,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6042780748663101,
|
| 102 |
+
"grad_norm": 48.600669860839844,
|
| 103 |
+
"learning_rate": 1.5989304812834226e-05,
|
| 104 |
+
"loss": 2.8483,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7379679144385025,
|
| 109 |
+
"grad_norm": 40.28901290893555,
|
| 110 |
+
"learning_rate": 1.7326203208556153e-05,
|
| 111 |
+
"loss": 2.8516,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.8716577540106951,
|
| 116 |
+
"grad_norm": 44.710628509521484,
|
| 117 |
+
"learning_rate": 1.8663101604278075e-05,
|
| 118 |
+
"loss": 2.7024,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7762237762237763,
|
| 124 |
+
"eval_loss": 0.5299336910247803,
|
| 125 |
+
"eval_macro_f1": 0.7808698231207254,
|
| 126 |
+
"eval_runtime": 1.0373,
|
| 127 |
+
"eval_samples_per_second": 551.44,
|
| 128 |
+
"eval_steps_per_second": 34.706,
|
| 129 |
+
"eval_weighted_f1": 0.7809045851263942,
|
| 130 |
+
"step": 374
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0053475935828877,
|
| 134 |
+
"grad_norm": 45.8042106628418,
|
| 135 |
+
"learning_rate": 2e-05,
|
| 136 |
+
"loss": 2.4003,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.1390374331550803,
|
| 141 |
+
"grad_norm": 46.33683395385742,
|
| 142 |
+
"learning_rate": 1.9851455733808676e-05,
|
| 143 |
+
"loss": 1.8141,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.2727272727272725,
|
| 148 |
+
"grad_norm": 44.520469665527344,
|
| 149 |
+
"learning_rate": 1.9702911467617353e-05,
|
| 150 |
+
"loss": 2.1766,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.406417112299465,
|
| 155 |
+
"grad_norm": 64.53562927246094,
|
| 156 |
+
"learning_rate": 1.9554367201426024e-05,
|
| 157 |
+
"loss": 2.2113,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.5401069518716577,
|
| 162 |
+
"grad_norm": 53.96901321411133,
|
| 163 |
+
"learning_rate": 1.94058229352347e-05,
|
| 164 |
+
"loss": 2.0653,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.6737967914438503,
|
| 169 |
+
"grad_norm": 78.2419662475586,
|
| 170 |
+
"learning_rate": 1.9257278669043376e-05,
|
| 171 |
+
"loss": 2.0692,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.807486631016043,
|
| 176 |
+
"grad_norm": 33.31931686401367,
|
| 177 |
+
"learning_rate": 1.9108734402852053e-05,
|
| 178 |
+
"loss": 1.8708,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.9411764705882355,
|
| 183 |
+
"grad_norm": 23.48499298095703,
|
| 184 |
+
"learning_rate": 1.8960190136660727e-05,
|
| 185 |
+
"loss": 1.7389,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.0,
|
| 190 |
+
"eval_accuracy": 0.8409090909090909,
|
| 191 |
+
"eval_loss": 0.42494767904281616,
|
| 192 |
+
"eval_macro_f1": 0.8414264010732074,
|
| 193 |
+
"eval_runtime": 1.0734,
|
| 194 |
+
"eval_samples_per_second": 532.885,
|
| 195 |
+
"eval_steps_per_second": 33.538,
|
| 196 |
+
"eval_weighted_f1": 0.8408405475971842,
|
| 197 |
+
"step": 561
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0748663101604277,
|
| 201 |
+
"grad_norm": 44.344139099121094,
|
| 202 |
+
"learning_rate": 1.88116458704694e-05,
|
| 203 |
+
"loss": 1.7572,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.2085561497326203,
|
| 208 |
+
"grad_norm": 29.0012264251709,
|
| 209 |
+
"learning_rate": 1.8663101604278075e-05,
|
| 210 |
+
"loss": 1.5159,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.342245989304813,
|
| 215 |
+
"grad_norm": 38.60652542114258,
|
| 216 |
+
"learning_rate": 1.8514557338086753e-05,
|
| 217 |
+
"loss": 1.2328,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4759358288770055,
|
| 222 |
+
"grad_norm": 29.8499698638916,
|
| 223 |
+
"learning_rate": 1.8366013071895427e-05,
|
| 224 |
+
"loss": 1.4623,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.6096256684491976,
|
| 229 |
+
"grad_norm": 56.83005142211914,
|
| 230 |
+
"learning_rate": 1.82174688057041e-05,
|
| 231 |
+
"loss": 1.5152,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.7433155080213902,
|
| 236 |
+
"grad_norm": 25.463768005371094,
|
| 237 |
+
"learning_rate": 1.8068924539512775e-05,
|
| 238 |
+
"loss": 1.4098,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.877005347593583,
|
| 243 |
+
"grad_norm": 40.18172073364258,
|
| 244 |
+
"learning_rate": 1.7920380273321453e-05,
|
| 245 |
+
"loss": 1.5085,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8426573426573427,
|
| 251 |
+
"eval_loss": 0.43529632687568665,
|
| 252 |
+
"eval_macro_f1": 0.8426902149674427,
|
| 253 |
+
"eval_runtime": 0.9898,
|
| 254 |
+
"eval_samples_per_second": 577.908,
|
| 255 |
+
"eval_steps_per_second": 36.372,
|
| 256 |
+
"eval_weighted_f1": 0.8423229433130422,
|
| 257 |
+
"step": 748
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.010695187165775,
|
| 261 |
+
"grad_norm": 34.088924407958984,
|
| 262 |
+
"learning_rate": 1.7771836007130127e-05,
|
| 263 |
+
"loss": 1.3705,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.144385026737968,
|
| 268 |
+
"grad_norm": 54.056461334228516,
|
| 269 |
+
"learning_rate": 1.76232917409388e-05,
|
| 270 |
+
"loss": 1.1595,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.278074866310161,
|
| 275 |
+
"grad_norm": 40.378910064697266,
|
| 276 |
+
"learning_rate": 1.7474747474747475e-05,
|
| 277 |
+
"loss": 0.954,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.411764705882353,
|
| 282 |
+
"grad_norm": 90.93668365478516,
|
| 283 |
+
"learning_rate": 1.7326203208556153e-05,
|
| 284 |
+
"loss": 0.9105,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.545454545454545,
|
| 289 |
+
"grad_norm": 62.43103790283203,
|
| 290 |
+
"learning_rate": 1.7177658942364827e-05,
|
| 291 |
+
"loss": 0.771,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.6791443850267385,
|
| 296 |
+
"grad_norm": 36.349788665771484,
|
| 297 |
+
"learning_rate": 1.70291146761735e-05,
|
| 298 |
+
"loss": 0.9634,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.81283422459893,
|
| 303 |
+
"grad_norm": 88.49812316894531,
|
| 304 |
+
"learning_rate": 1.6880570409982175e-05,
|
| 305 |
+
"loss": 0.8089,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.946524064171123,
|
| 310 |
+
"grad_norm": 132.47991943359375,
|
| 311 |
+
"learning_rate": 1.6732026143790852e-05,
|
| 312 |
+
"loss": 1.178,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.0,
|
| 317 |
+
"eval_accuracy": 0.8041958041958042,
|
| 318 |
+
"eval_loss": 0.6062424182891846,
|
| 319 |
+
"eval_macro_f1": 0.8062141277126513,
|
| 320 |
+
"eval_runtime": 1.0356,
|
| 321 |
+
"eval_samples_per_second": 552.318,
|
| 322 |
+
"eval_steps_per_second": 34.761,
|
| 323 |
+
"eval_weighted_f1": 0.8067557022741634,
|
| 324 |
+
"step": 935
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.080213903743315,
|
| 328 |
+
"grad_norm": 15.832716941833496,
|
| 329 |
+
"learning_rate": 1.6583481877599526e-05,
|
| 330 |
+
"loss": 0.887,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.213903743315508,
|
| 335 |
+
"grad_norm": 32.42772674560547,
|
| 336 |
+
"learning_rate": 1.64349376114082e-05,
|
| 337 |
+
"loss": 0.631,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.347593582887701,
|
| 342 |
+
"grad_norm": 125.02656555175781,
|
| 343 |
+
"learning_rate": 1.6286393345216875e-05,
|
| 344 |
+
"loss": 0.6536,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.481283422459893,
|
| 349 |
+
"grad_norm": 84.05538177490234,
|
| 350 |
+
"learning_rate": 1.6137849079025552e-05,
|
| 351 |
+
"loss": 0.7163,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.614973262032086,
|
| 356 |
+
"grad_norm": 46.46393585205078,
|
| 357 |
+
"learning_rate": 1.5989304812834226e-05,
|
| 358 |
+
"loss": 0.841,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.748663101604278,
|
| 363 |
+
"grad_norm": 52.25996398925781,
|
| 364 |
+
"learning_rate": 1.58407605466429e-05,
|
| 365 |
+
"loss": 0.6103,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.882352941176471,
|
| 370 |
+
"grad_norm": 70.38349914550781,
|
| 371 |
+
"learning_rate": 1.5692216280451574e-05,
|
| 372 |
+
"loss": 0.4908,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.0,
|
| 377 |
+
"eval_accuracy": 0.8356643356643356,
|
| 378 |
+
"eval_loss": 0.597773551940918,
|
| 379 |
+
"eval_macro_f1": 0.8367203204766657,
|
| 380 |
+
"eval_runtime": 1.0873,
|
| 381 |
+
"eval_samples_per_second": 526.057,
|
| 382 |
+
"eval_steps_per_second": 33.108,
|
| 383 |
+
"eval_weighted_f1": 0.8365573859521537,
|
| 384 |
+
"step": 1122
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.016042780748663,
|
| 388 |
+
"grad_norm": 32.182701110839844,
|
| 389 |
+
"learning_rate": 1.5543672014260252e-05,
|
| 390 |
+
"loss": 0.7634,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.149732620320855,
|
| 395 |
+
"grad_norm": 30.86662483215332,
|
| 396 |
+
"learning_rate": 1.5395127748068926e-05,
|
| 397 |
+
"loss": 0.3263,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.283422459893048,
|
| 402 |
+
"grad_norm": 86.34818267822266,
|
| 403 |
+
"learning_rate": 1.5246583481877602e-05,
|
| 404 |
+
"loss": 0.5872,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.4171122994652405,
|
| 409 |
+
"grad_norm": 17.692935943603516,
|
| 410 |
+
"learning_rate": 1.5098039215686276e-05,
|
| 411 |
+
"loss": 0.4611,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.550802139037433,
|
| 416 |
+
"grad_norm": 53.08113479614258,
|
| 417 |
+
"learning_rate": 1.4949494949494952e-05,
|
| 418 |
+
"loss": 0.5075,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.684491978609626,
|
| 423 |
+
"grad_norm": 63.51148986816406,
|
| 424 |
+
"learning_rate": 1.4800950683303626e-05,
|
| 425 |
+
"loss": 0.5439,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.818181818181818,
|
| 430 |
+
"grad_norm": 71.70096588134766,
|
| 431 |
+
"learning_rate": 1.4652406417112302e-05,
|
| 432 |
+
"loss": 0.6084,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.951871657754011,
|
| 437 |
+
"grad_norm": 44.453033447265625,
|
| 438 |
+
"learning_rate": 1.4503862150920976e-05,
|
| 439 |
+
"loss": 0.4496,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 7.0,
|
| 444 |
+
"eval_accuracy": 0.8548951048951049,
|
| 445 |
+
"eval_loss": 0.6412630081176758,
|
| 446 |
+
"eval_macro_f1": 0.8556166390203034,
|
| 447 |
+
"eval_runtime": 1.0748,
|
| 448 |
+
"eval_samples_per_second": 532.171,
|
| 449 |
+
"eval_steps_per_second": 33.493,
|
| 450 |
+
"eval_weighted_f1": 0.8550553996897675,
|
| 451 |
+
"step": 1309
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.0855614973262036,
|
| 455 |
+
"grad_norm": 0.3045983612537384,
|
| 456 |
+
"learning_rate": 1.4355317884729652e-05,
|
| 457 |
+
"loss": 0.376,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.219251336898395,
|
| 462 |
+
"grad_norm": 29.222652435302734,
|
| 463 |
+
"learning_rate": 1.4206773618538326e-05,
|
| 464 |
+
"loss": 0.3552,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.352941176470588,
|
| 469 |
+
"grad_norm": 46.05845260620117,
|
| 470 |
+
"learning_rate": 1.4058229352347002e-05,
|
| 471 |
+
"loss": 0.3004,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.4866310160427805,
|
| 476 |
+
"grad_norm": 72.29778289794922,
|
| 477 |
+
"learning_rate": 1.3909685086155676e-05,
|
| 478 |
+
"loss": 0.471,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.620320855614973,
|
| 483 |
+
"grad_norm": 3.986952304840088,
|
| 484 |
+
"learning_rate": 1.3761140819964351e-05,
|
| 485 |
+
"loss": 0.392,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.754010695187166,
|
| 490 |
+
"grad_norm": 9.997687339782715,
|
| 491 |
+
"learning_rate": 1.3612596553773026e-05,
|
| 492 |
+
"loss": 0.3483,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 7.887700534759358,
|
| 497 |
+
"grad_norm": 7.379712104797363,
|
| 498 |
+
"learning_rate": 1.3464052287581701e-05,
|
| 499 |
+
"loss": 0.3257,
|
| 500 |
+
"step": 1475
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 8.0,
|
| 504 |
+
"eval_accuracy": 0.8409090909090909,
|
| 505 |
+
"eval_loss": 0.8687260150909424,
|
| 506 |
+
"eval_macro_f1": 0.8417380602609604,
|
| 507 |
+
"eval_runtime": 1.0027,
|
| 508 |
+
"eval_samples_per_second": 570.476,
|
| 509 |
+
"eval_steps_per_second": 35.904,
|
| 510 |
+
"eval_weighted_f1": 0.8415264484868364,
|
| 511 |
+
"step": 1496
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.02139037433155,
|
| 515 |
+
"grad_norm": 0.2209583967924118,
|
| 516 |
+
"learning_rate": 1.3315508021390375e-05,
|
| 517 |
+
"loss": 0.3858,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.155080213903743,
|
| 522 |
+
"grad_norm": 131.19735717773438,
|
| 523 |
+
"learning_rate": 1.3166963755199051e-05,
|
| 524 |
+
"loss": 0.4365,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.288770053475936,
|
| 529 |
+
"grad_norm": 20.309551239013672,
|
| 530 |
+
"learning_rate": 1.3018419489007725e-05,
|
| 531 |
+
"loss": 0.314,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.422459893048128,
|
| 536 |
+
"grad_norm": 0.14381442964076996,
|
| 537 |
+
"learning_rate": 1.2869875222816401e-05,
|
| 538 |
+
"loss": 0.1513,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.556149732620321,
|
| 543 |
+
"grad_norm": 57.89705276489258,
|
| 544 |
+
"learning_rate": 1.2721330956625075e-05,
|
| 545 |
+
"loss": 0.4231,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.689839572192513,
|
| 550 |
+
"grad_norm": 76.58484649658203,
|
| 551 |
+
"learning_rate": 1.257278669043375e-05,
|
| 552 |
+
"loss": 0.2566,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 8.823529411764707,
|
| 557 |
+
"grad_norm": 4.341352939605713,
|
| 558 |
+
"learning_rate": 1.2424242424242425e-05,
|
| 559 |
+
"loss": 0.3454,
|
| 560 |
+
"step": 1650
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 8.957219251336898,
|
| 564 |
+
"grad_norm": 259.73809814453125,
|
| 565 |
+
"learning_rate": 1.22756981580511e-05,
|
| 566 |
+
"loss": 0.5365,
|
| 567 |
+
"step": 1675
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 9.0,
|
| 571 |
+
"eval_accuracy": 0.8374125874125874,
|
| 572 |
+
"eval_loss": 0.9592656493186951,
|
| 573 |
+
"eval_macro_f1": 0.8381007776505479,
|
| 574 |
+
"eval_runtime": 1.0459,
|
| 575 |
+
"eval_samples_per_second": 546.885,
|
| 576 |
+
"eval_steps_per_second": 34.419,
|
| 577 |
+
"eval_weighted_f1": 0.8376940935937394,
|
| 578 |
+
"step": 1683
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.090909090909092,
|
| 582 |
+
"grad_norm": 252.94906616210938,
|
| 583 |
+
"learning_rate": 1.2127153891859775e-05,
|
| 584 |
+
"loss": 0.3456,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.224598930481283,
|
| 589 |
+
"grad_norm": 29.55437469482422,
|
| 590 |
+
"learning_rate": 1.197860962566845e-05,
|
| 591 |
+
"loss": 0.3441,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.358288770053475,
|
| 596 |
+
"grad_norm": 0.10047707706689835,
|
| 597 |
+
"learning_rate": 1.1830065359477125e-05,
|
| 598 |
+
"loss": 0.164,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.491978609625669,
|
| 603 |
+
"grad_norm": 12.29434585571289,
|
| 604 |
+
"learning_rate": 1.1681521093285799e-05,
|
| 605 |
+
"loss": 0.2852,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.62566844919786,
|
| 610 |
+
"grad_norm": 0.1449579894542694,
|
| 611 |
+
"learning_rate": 1.1532976827094475e-05,
|
| 612 |
+
"loss": 0.2883,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 9.759358288770054,
|
| 617 |
+
"grad_norm": 0.610160231590271,
|
| 618 |
+
"learning_rate": 1.1384432560903149e-05,
|
| 619 |
+
"loss": 0.2499,
|
| 620 |
+
"step": 1825
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 9.893048128342246,
|
| 624 |
+
"grad_norm": 1.6577903032302856,
|
| 625 |
+
"learning_rate": 1.1235888294711825e-05,
|
| 626 |
+
"loss": 0.2437,
|
| 627 |
+
"step": 1850
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 10.0,
|
| 631 |
+
"eval_accuracy": 0.833916083916084,
|
| 632 |
+
"eval_loss": 1.0277149677276611,
|
| 633 |
+
"eval_macro_f1": 0.8346846295162278,
|
| 634 |
+
"eval_runtime": 1.0577,
|
| 635 |
+
"eval_samples_per_second": 540.796,
|
| 636 |
+
"eval_steps_per_second": 34.036,
|
| 637 |
+
"eval_weighted_f1": 0.8344204968332342,
|
| 638 |
+
"step": 1870
|
| 639 |
+
}
|
| 640 |
+
],
|
| 641 |
+
"logging_steps": 25,
|
| 642 |
+
"max_steps": 3740,
|
| 643 |
+
"num_input_tokens_seen": 0,
|
| 644 |
+
"num_train_epochs": 20,
|
| 645 |
+
"save_steps": 500,
|
| 646 |
+
"stateful_callbacks": {
|
| 647 |
+
"EarlyStoppingCallback": {
|
| 648 |
+
"args": {
|
| 649 |
+
"early_stopping_patience": 3,
|
| 650 |
+
"early_stopping_threshold": 0.0
|
| 651 |
+
},
|
| 652 |
+
"attributes": {
|
| 653 |
+
"early_stopping_patience_counter": 3
|
| 654 |
+
}
|
| 655 |
+
},
|
| 656 |
+
"TrainerControl": {
|
| 657 |
+
"args": {
|
| 658 |
+
"should_epoch_stop": false,
|
| 659 |
+
"should_evaluate": false,
|
| 660 |
+
"should_log": false,
|
| 661 |
+
"should_save": true,
|
| 662 |
+
"should_training_stop": true
|
| 663 |
+
},
|
| 664 |
+
"attributes": {}
|
| 665 |
+
}
|
| 666 |
+
},
|
| 667 |
+
"total_flos": 0.0,
|
| 668 |
+
"train_batch_size": 8,
|
| 669 |
+
"trial_name": null,
|
| 670 |
+
"trial_params": null
|
| 671 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/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-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/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-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528/trainer_state.json
ADDED
|
@@ -0,0 +1,558 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1528,
|
| 3 |
+
"best_metric": 0.8506965399706874,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1528,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1310615989515072,
|
| 14 |
+
"grad_norm": 7.0655951499938965,
|
| 15 |
+
"learning_rate": 1.256544502617801e-06,
|
| 16 |
+
"loss": 4.4096,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2621231979030144,
|
| 21 |
+
"grad_norm": 10.394327163696289,
|
| 22 |
+
"learning_rate": 2.5654450261780104e-06,
|
| 23 |
+
"loss": 4.4022,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3931847968545216,
|
| 28 |
+
"grad_norm": 9.305167198181152,
|
| 29 |
+
"learning_rate": 3.874345549738221e-06,
|
| 30 |
+
"loss": 4.4126,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5242463958060288,
|
| 35 |
+
"grad_norm": 13.898597717285156,
|
| 36 |
+
"learning_rate": 5.18324607329843e-06,
|
| 37 |
+
"loss": 4.4164,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6553079947575361,
|
| 42 |
+
"grad_norm": 18.38749122619629,
|
| 43 |
+
"learning_rate": 6.492146596858639e-06,
|
| 44 |
+
"loss": 4.3645,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7863695937090432,
|
| 49 |
+
"grad_norm": 21.86370849609375,
|
| 50 |
+
"learning_rate": 7.80104712041885e-06,
|
| 51 |
+
"loss": 4.3804,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9174311926605505,
|
| 56 |
+
"grad_norm": 15.066303253173828,
|
| 57 |
+
"learning_rate": 9.109947643979057e-06,
|
| 58 |
+
"loss": 4.3774,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.5188356164383562,
|
| 64 |
+
"eval_loss": 0.997204065322876,
|
| 65 |
+
"eval_macro_f1": 0.5131051132212359,
|
| 66 |
+
"eval_runtime": 1.0548,
|
| 67 |
+
"eval_samples_per_second": 553.679,
|
| 68 |
+
"eval_steps_per_second": 35.079,
|
| 69 |
+
"eval_weighted_f1": 0.5139636468867662,
|
| 70 |
+
"step": 191
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0471821756225426,
|
| 74 |
+
"grad_norm": 23.784353256225586,
|
| 75 |
+
"learning_rate": 1.0418848167539269e-05,
|
| 76 |
+
"loss": 4.1102,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.1782437745740497,
|
| 81 |
+
"grad_norm": 30.204675674438477,
|
| 82 |
+
"learning_rate": 1.1727748691099476e-05,
|
| 83 |
+
"loss": 3.8862,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.309305373525557,
|
| 88 |
+
"grad_norm": 48.28770065307617,
|
| 89 |
+
"learning_rate": 1.3036649214659686e-05,
|
| 90 |
+
"loss": 3.3161,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4403669724770642,
|
| 95 |
+
"grad_norm": 33.28887176513672,
|
| 96 |
+
"learning_rate": 1.4345549738219897e-05,
|
| 97 |
+
"loss": 2.8112,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.5714285714285714,
|
| 102 |
+
"grad_norm": 54.89958953857422,
|
| 103 |
+
"learning_rate": 1.5654450261780107e-05,
|
| 104 |
+
"loss": 2.7578,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7024901703800785,
|
| 109 |
+
"grad_norm": 38.093074798583984,
|
| 110 |
+
"learning_rate": 1.6963350785340316e-05,
|
| 111 |
+
"loss": 2.5168,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.833551769331586,
|
| 116 |
+
"grad_norm": 38.326072692871094,
|
| 117 |
+
"learning_rate": 1.8272251308900526e-05,
|
| 118 |
+
"loss": 2.3348,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.964613368283093,
|
| 123 |
+
"grad_norm": 31.03067970275879,
|
| 124 |
+
"learning_rate": 1.9581151832460736e-05,
|
| 125 |
+
"loss": 2.4286,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.0,
|
| 130 |
+
"eval_accuracy": 0.8013698630136986,
|
| 131 |
+
"eval_loss": 0.4932255148887634,
|
| 132 |
+
"eval_macro_f1": 0.7969465425692124,
|
| 133 |
+
"eval_runtime": 0.9762,
|
| 134 |
+
"eval_samples_per_second": 598.236,
|
| 135 |
+
"eval_steps_per_second": 37.902,
|
| 136 |
+
"eval_weighted_f1": 0.7971179240286179,
|
| 137 |
+
"step": 382
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.094364351245085,
|
| 141 |
+
"grad_norm": 53.79396438598633,
|
| 142 |
+
"learning_rate": 1.9901105293775453e-05,
|
| 143 |
+
"loss": 1.9474,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.2254259501965925,
|
| 148 |
+
"grad_norm": 29.14629364013672,
|
| 149 |
+
"learning_rate": 1.9755671902268764e-05,
|
| 150 |
+
"loss": 2.2147,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.3564875491480994,
|
| 155 |
+
"grad_norm": 72.80538940429688,
|
| 156 |
+
"learning_rate": 1.961023851076207e-05,
|
| 157 |
+
"loss": 1.8554,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.487549148099607,
|
| 162 |
+
"grad_norm": 54.436893463134766,
|
| 163 |
+
"learning_rate": 1.9464805119255382e-05,
|
| 164 |
+
"loss": 1.8888,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.618610747051114,
|
| 169 |
+
"grad_norm": 25.074434280395508,
|
| 170 |
+
"learning_rate": 1.9319371727748693e-05,
|
| 171 |
+
"loss": 1.7695,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.749672346002621,
|
| 176 |
+
"grad_norm": 52.36857223510742,
|
| 177 |
+
"learning_rate": 1.9173938336242e-05,
|
| 178 |
+
"loss": 1.8699,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.8807339449541285,
|
| 183 |
+
"grad_norm": 52.16041564941406,
|
| 184 |
+
"learning_rate": 1.9028504944735315e-05,
|
| 185 |
+
"loss": 1.8778,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.0,
|
| 190 |
+
"eval_accuracy": 0.833904109589041,
|
| 191 |
+
"eval_loss": 0.43060269951820374,
|
| 192 |
+
"eval_macro_f1": 0.8332285980130937,
|
| 193 |
+
"eval_runtime": 1.0111,
|
| 194 |
+
"eval_samples_per_second": 577.567,
|
| 195 |
+
"eval_steps_per_second": 36.592,
|
| 196 |
+
"eval_weighted_f1": 0.8332468364120059,
|
| 197 |
+
"step": 573
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0104849279161208,
|
| 201 |
+
"grad_norm": 23.629840850830078,
|
| 202 |
+
"learning_rate": 1.8883071553228622e-05,
|
| 203 |
+
"loss": 1.5131,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.1415465268676277,
|
| 208 |
+
"grad_norm": 47.1800422668457,
|
| 209 |
+
"learning_rate": 1.8737638161721933e-05,
|
| 210 |
+
"loss": 1.3236,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.272608125819135,
|
| 215 |
+
"grad_norm": 82.11866760253906,
|
| 216 |
+
"learning_rate": 1.8592204770215244e-05,
|
| 217 |
+
"loss": 1.1795,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4036697247706424,
|
| 222 |
+
"grad_norm": 27.585718154907227,
|
| 223 |
+
"learning_rate": 1.844677137870855e-05,
|
| 224 |
+
"loss": 1.2008,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.5347313237221494,
|
| 229 |
+
"grad_norm": 64.93595123291016,
|
| 230 |
+
"learning_rate": 1.8301337987201865e-05,
|
| 231 |
+
"loss": 1.1102,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.6657929226736568,
|
| 236 |
+
"grad_norm": 60.96466827392578,
|
| 237 |
+
"learning_rate": 1.8155904595695173e-05,
|
| 238 |
+
"loss": 1.1672,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.7968545216251637,
|
| 243 |
+
"grad_norm": 43.670074462890625,
|
| 244 |
+
"learning_rate": 1.8010471204188483e-05,
|
| 245 |
+
"loss": 1.2541,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.927916120576671,
|
| 250 |
+
"grad_norm": 30.041912078857422,
|
| 251 |
+
"learning_rate": 1.7865037812681794e-05,
|
| 252 |
+
"loss": 1.1274,
|
| 253 |
+
"step": 750
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 4.0,
|
| 257 |
+
"eval_accuracy": 0.8184931506849316,
|
| 258 |
+
"eval_loss": 0.4712919592857361,
|
| 259 |
+
"eval_macro_f1": 0.8189065851360052,
|
| 260 |
+
"eval_runtime": 1.0243,
|
| 261 |
+
"eval_samples_per_second": 570.168,
|
| 262 |
+
"eval_steps_per_second": 36.124,
|
| 263 |
+
"eval_weighted_f1": 0.8188688163970698,
|
| 264 |
+
"step": 764
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.057667103538663,
|
| 268 |
+
"grad_norm": 45.4857292175293,
|
| 269 |
+
"learning_rate": 1.77196044211751e-05,
|
| 270 |
+
"loss": 1.0849,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.18872870249017,
|
| 275 |
+
"grad_norm": 40.71622085571289,
|
| 276 |
+
"learning_rate": 1.7574171029668416e-05,
|
| 277 |
+
"loss": 0.7316,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.319790301441677,
|
| 282 |
+
"grad_norm": 62.60859298706055,
|
| 283 |
+
"learning_rate": 1.7428737638161723e-05,
|
| 284 |
+
"loss": 0.825,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.450851900393185,
|
| 289 |
+
"grad_norm": 27.907306671142578,
|
| 290 |
+
"learning_rate": 1.7283304246655034e-05,
|
| 291 |
+
"loss": 0.8251,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.581913499344692,
|
| 296 |
+
"grad_norm": 36.16027069091797,
|
| 297 |
+
"learning_rate": 1.7137870855148345e-05,
|
| 298 |
+
"loss": 0.846,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.712975098296199,
|
| 303 |
+
"grad_norm": 59.31807327270508,
|
| 304 |
+
"learning_rate": 1.6992437463641652e-05,
|
| 305 |
+
"loss": 0.8842,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.844036697247707,
|
| 310 |
+
"grad_norm": 115.43113708496094,
|
| 311 |
+
"learning_rate": 1.6847004072134963e-05,
|
| 312 |
+
"loss": 0.8585,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 4.975098296199214,
|
| 317 |
+
"grad_norm": 33.352783203125,
|
| 318 |
+
"learning_rate": 1.6701570680628274e-05,
|
| 319 |
+
"loss": 0.8284,
|
| 320 |
+
"step": 950
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.0,
|
| 324 |
+
"eval_accuracy": 0.839041095890411,
|
| 325 |
+
"eval_loss": 0.5166526436805725,
|
| 326 |
+
"eval_macro_f1": 0.838960125962807,
|
| 327 |
+
"eval_runtime": 1.0257,
|
| 328 |
+
"eval_samples_per_second": 569.367,
|
| 329 |
+
"eval_steps_per_second": 36.073,
|
| 330 |
+
"eval_weighted_f1": 0.8390335286782824,
|
| 331 |
+
"step": 955
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.104849279161206,
|
| 335 |
+
"grad_norm": 115.93670654296875,
|
| 336 |
+
"learning_rate": 1.6556137289121585e-05,
|
| 337 |
+
"loss": 0.6029,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.235910878112713,
|
| 342 |
+
"grad_norm": 35.721553802490234,
|
| 343 |
+
"learning_rate": 1.6410703897614895e-05,
|
| 344 |
+
"loss": 0.5398,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.36697247706422,
|
| 349 |
+
"grad_norm": 77.26508331298828,
|
| 350 |
+
"learning_rate": 1.6265270506108203e-05,
|
| 351 |
+
"loss": 0.6449,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.498034076015728,
|
| 356 |
+
"grad_norm": 88.83381652832031,
|
| 357 |
+
"learning_rate": 1.6119837114601514e-05,
|
| 358 |
+
"loss": 0.788,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.6290956749672345,
|
| 363 |
+
"grad_norm": 18.69657325744629,
|
| 364 |
+
"learning_rate": 1.5974403723094824e-05,
|
| 365 |
+
"loss": 0.6488,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.760157273918741,
|
| 370 |
+
"grad_norm": 51.9311637878418,
|
| 371 |
+
"learning_rate": 1.5828970331588135e-05,
|
| 372 |
+
"loss": 0.7094,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.891218872870249,
|
| 377 |
+
"grad_norm": 29.13077163696289,
|
| 378 |
+
"learning_rate": 1.5683536940081443e-05,
|
| 379 |
+
"loss": 0.5296,
|
| 380 |
+
"step": 1125
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.0,
|
| 384 |
+
"eval_accuracy": 0.8476027397260274,
|
| 385 |
+
"eval_loss": 0.5409419536590576,
|
| 386 |
+
"eval_macro_f1": 0.8473890353589603,
|
| 387 |
+
"eval_runtime": 1.0337,
|
| 388 |
+
"eval_samples_per_second": 564.977,
|
| 389 |
+
"eval_steps_per_second": 35.795,
|
| 390 |
+
"eval_weighted_f1": 0.8472895122524333,
|
| 391 |
+
"step": 1146
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.0209698558322415,
|
| 395 |
+
"grad_norm": 28.083419799804688,
|
| 396 |
+
"learning_rate": 1.5538103548574753e-05,
|
| 397 |
+
"loss": 0.6039,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.1520314547837485,
|
| 402 |
+
"grad_norm": 57.95465087890625,
|
| 403 |
+
"learning_rate": 1.5392670157068064e-05,
|
| 404 |
+
"loss": 0.4103,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.283093053735255,
|
| 409 |
+
"grad_norm": 90.1124267578125,
|
| 410 |
+
"learning_rate": 1.5247236765561375e-05,
|
| 411 |
+
"loss": 0.3741,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.414154652686763,
|
| 416 |
+
"grad_norm": 73.2316665649414,
|
| 417 |
+
"learning_rate": 1.5101803374054684e-05,
|
| 418 |
+
"loss": 0.5888,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.54521625163827,
|
| 423 |
+
"grad_norm": 19.490676879882812,
|
| 424 |
+
"learning_rate": 1.4956369982547993e-05,
|
| 425 |
+
"loss": 0.5103,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.676277850589777,
|
| 430 |
+
"grad_norm": 21.10092544555664,
|
| 431 |
+
"learning_rate": 1.4810936591041306e-05,
|
| 432 |
+
"loss": 0.6769,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.807339449541285,
|
| 437 |
+
"grad_norm": 1.7164210081100464,
|
| 438 |
+
"learning_rate": 1.4665503199534615e-05,
|
| 439 |
+
"loss": 0.3104,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 6.938401048492792,
|
| 444 |
+
"grad_norm": 66.02993774414062,
|
| 445 |
+
"learning_rate": 1.4520069808027924e-05,
|
| 446 |
+
"loss": 0.6902,
|
| 447 |
+
"step": 1325
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 7.0,
|
| 451 |
+
"eval_accuracy": 0.8476027397260274,
|
| 452 |
+
"eval_loss": 0.7329329252243042,
|
| 453 |
+
"eval_macro_f1": 0.8482001147311004,
|
| 454 |
+
"eval_runtime": 1.0016,
|
| 455 |
+
"eval_samples_per_second": 583.055,
|
| 456 |
+
"eval_steps_per_second": 36.94,
|
| 457 |
+
"eval_weighted_f1": 0.8479630934295177,
|
| 458 |
+
"step": 1337
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.068152031454784,
|
| 462 |
+
"grad_norm": 1.1677628755569458,
|
| 463 |
+
"learning_rate": 1.4374636416521235e-05,
|
| 464 |
+
"loss": 0.4478,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.199213630406291,
|
| 469 |
+
"grad_norm": 218.09445190429688,
|
| 470 |
+
"learning_rate": 1.4229203025014544e-05,
|
| 471 |
+
"loss": 0.3396,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.330275229357798,
|
| 476 |
+
"grad_norm": 147.76715087890625,
|
| 477 |
+
"learning_rate": 1.4083769633507855e-05,
|
| 478 |
+
"loss": 0.3215,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.461336828309306,
|
| 483 |
+
"grad_norm": 145.50807189941406,
|
| 484 |
+
"learning_rate": 1.3938336242001166e-05,
|
| 485 |
+
"loss": 0.305,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.592398427260813,
|
| 490 |
+
"grad_norm": 3.1571762561798096,
|
| 491 |
+
"learning_rate": 1.3792902850494475e-05,
|
| 492 |
+
"loss": 0.2453,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 7.72346002621232,
|
| 497 |
+
"grad_norm": 3.3650453090667725,
|
| 498 |
+
"learning_rate": 1.3647469458987785e-05,
|
| 499 |
+
"loss": 0.3413,
|
| 500 |
+
"step": 1475
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 7.854521625163827,
|
| 504 |
+
"grad_norm": 0.1421426385641098,
|
| 505 |
+
"learning_rate": 1.3502036067481095e-05,
|
| 506 |
+
"loss": 0.3495,
|
| 507 |
+
"step": 1500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 7.985583224115334,
|
| 511 |
+
"grad_norm": 109.3970718383789,
|
| 512 |
+
"learning_rate": 1.3356602675974404e-05,
|
| 513 |
+
"loss": 0.3587,
|
| 514 |
+
"step": 1525
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 8.0,
|
| 518 |
+
"eval_accuracy": 0.851027397260274,
|
| 519 |
+
"eval_loss": 0.8212750554084778,
|
| 520 |
+
"eval_macro_f1": 0.8506965399706874,
|
| 521 |
+
"eval_runtime": 1.0664,
|
| 522 |
+
"eval_samples_per_second": 547.622,
|
| 523 |
+
"eval_steps_per_second": 34.695,
|
| 524 |
+
"eval_weighted_f1": 0.8505670696665327,
|
| 525 |
+
"step": 1528
|
| 526 |
+
}
|
| 527 |
+
],
|
| 528 |
+
"logging_steps": 25,
|
| 529 |
+
"max_steps": 3820,
|
| 530 |
+
"num_input_tokens_seen": 0,
|
| 531 |
+
"num_train_epochs": 20,
|
| 532 |
+
"save_steps": 500,
|
| 533 |
+
"stateful_callbacks": {
|
| 534 |
+
"EarlyStoppingCallback": {
|
| 535 |
+
"args": {
|
| 536 |
+
"early_stopping_patience": 3,
|
| 537 |
+
"early_stopping_threshold": 0.0
|
| 538 |
+
},
|
| 539 |
+
"attributes": {
|
| 540 |
+
"early_stopping_patience_counter": 0
|
| 541 |
+
}
|
| 542 |
+
},
|
| 543 |
+
"TrainerControl": {
|
| 544 |
+
"args": {
|
| 545 |
+
"should_epoch_stop": false,
|
| 546 |
+
"should_evaluate": false,
|
| 547 |
+
"should_log": false,
|
| 548 |
+
"should_save": true,
|
| 549 |
+
"should_training_stop": false
|
| 550 |
+
},
|
| 551 |
+
"attributes": {}
|
| 552 |
+
}
|
| 553 |
+
},
|
| 554 |
+
"total_flos": 0.0,
|
| 555 |
+
"train_batch_size": 8,
|
| 556 |
+
"trial_name": null,
|
| 557 |
+
"trial_params": null
|
| 558 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/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-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/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-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-2101/trainer_state.json
ADDED
|
@@ -0,0 +1,752 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1528,
|
| 3 |
+
"best_metric": 0.8506965399706874,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-01-s44-20260828T201044Z-496d6871/trainer/checkpoint-1528",
|
| 5 |
+
"epoch": 11.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2101,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1310615989515072,
|
| 14 |
+
"grad_norm": 7.0655951499938965,
|
| 15 |
+
"learning_rate": 1.256544502617801e-06,
|
| 16 |
+
"loss": 4.4096,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2621231979030144,
|
| 21 |
+
"grad_norm": 10.394327163696289,
|
| 22 |
+
"learning_rate": 2.5654450261780104e-06,
|
| 23 |
+
"loss": 4.4022,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3931847968545216,
|
| 28 |
+
"grad_norm": 9.305167198181152,
|
| 29 |
+
"learning_rate": 3.874345549738221e-06,
|
| 30 |
+
"loss": 4.4126,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5242463958060288,
|
| 35 |
+
"grad_norm": 13.898597717285156,
|
| 36 |
+
"learning_rate": 5.18324607329843e-06,
|
| 37 |
+
"loss": 4.4164,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6553079947575361,
|
| 42 |
+
"grad_norm": 18.38749122619629,
|
| 43 |
+
"learning_rate": 6.492146596858639e-06,
|
| 44 |
+
"loss": 4.3645,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7863695937090432,
|
| 49 |
+
"grad_norm": 21.86370849609375,
|
| 50 |
+
"learning_rate": 7.80104712041885e-06,
|
| 51 |
+
"loss": 4.3804,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9174311926605505,
|
| 56 |
+
"grad_norm": 15.066303253173828,
|
| 57 |
+
"learning_rate": 9.109947643979057e-06,
|
| 58 |
+
"loss": 4.3774,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.5188356164383562,
|
| 64 |
+
"eval_loss": 0.997204065322876,
|
| 65 |
+
"eval_macro_f1": 0.5131051132212359,
|
| 66 |
+
"eval_runtime": 1.0548,
|
| 67 |
+
"eval_samples_per_second": 553.679,
|
| 68 |
+
"eval_steps_per_second": 35.079,
|
| 69 |
+
"eval_weighted_f1": 0.5139636468867662,
|
| 70 |
+
"step": 191
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0471821756225426,
|
| 74 |
+
"grad_norm": 23.784353256225586,
|
| 75 |
+
"learning_rate": 1.0418848167539269e-05,
|
| 76 |
+
"loss": 4.1102,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.1782437745740497,
|
| 81 |
+
"grad_norm": 30.204675674438477,
|
| 82 |
+
"learning_rate": 1.1727748691099476e-05,
|
| 83 |
+
"loss": 3.8862,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.309305373525557,
|
| 88 |
+
"grad_norm": 48.28770065307617,
|
| 89 |
+
"learning_rate": 1.3036649214659686e-05,
|
| 90 |
+
"loss": 3.3161,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4403669724770642,
|
| 95 |
+
"grad_norm": 33.28887176513672,
|
| 96 |
+
"learning_rate": 1.4345549738219897e-05,
|
| 97 |
+
"loss": 2.8112,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.5714285714285714,
|
| 102 |
+
"grad_norm": 54.89958953857422,
|
| 103 |
+
"learning_rate": 1.5654450261780107e-05,
|
| 104 |
+
"loss": 2.7578,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7024901703800785,
|
| 109 |
+
"grad_norm": 38.093074798583984,
|
| 110 |
+
"learning_rate": 1.6963350785340316e-05,
|
| 111 |
+
"loss": 2.5168,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.833551769331586,
|
| 116 |
+
"grad_norm": 38.326072692871094,
|
| 117 |
+
"learning_rate": 1.8272251308900526e-05,
|
| 118 |
+
"loss": 2.3348,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.964613368283093,
|
| 123 |
+
"grad_norm": 31.03067970275879,
|
| 124 |
+
"learning_rate": 1.9581151832460736e-05,
|
| 125 |
+
"loss": 2.4286,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.0,
|
| 130 |
+
"eval_accuracy": 0.8013698630136986,
|
| 131 |
+
"eval_loss": 0.4932255148887634,
|
| 132 |
+
"eval_macro_f1": 0.7969465425692124,
|
| 133 |
+
"eval_runtime": 0.9762,
|
| 134 |
+
"eval_samples_per_second": 598.236,
|
| 135 |
+
"eval_steps_per_second": 37.902,
|
| 136 |
+
"eval_weighted_f1": 0.7971179240286179,
|
| 137 |
+
"step": 382
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.094364351245085,
|
| 141 |
+
"grad_norm": 53.79396438598633,
|
| 142 |
+
"learning_rate": 1.9901105293775453e-05,
|
| 143 |
+
"loss": 1.9474,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.2254259501965925,
|
| 148 |
+
"grad_norm": 29.14629364013672,
|
| 149 |
+
"learning_rate": 1.9755671902268764e-05,
|
| 150 |
+
"loss": 2.2147,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.3564875491480994,
|
| 155 |
+
"grad_norm": 72.80538940429688,
|
| 156 |
+
"learning_rate": 1.961023851076207e-05,
|
| 157 |
+
"loss": 1.8554,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.487549148099607,
|
| 162 |
+
"grad_norm": 54.436893463134766,
|
| 163 |
+
"learning_rate": 1.9464805119255382e-05,
|
| 164 |
+
"loss": 1.8888,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.618610747051114,
|
| 169 |
+
"grad_norm": 25.074434280395508,
|
| 170 |
+
"learning_rate": 1.9319371727748693e-05,
|
| 171 |
+
"loss": 1.7695,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.749672346002621,
|
| 176 |
+
"grad_norm": 52.36857223510742,
|
| 177 |
+
"learning_rate": 1.9173938336242e-05,
|
| 178 |
+
"loss": 1.8699,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.8807339449541285,
|
| 183 |
+
"grad_norm": 52.16041564941406,
|
| 184 |
+
"learning_rate": 1.9028504944735315e-05,
|
| 185 |
+
"loss": 1.8778,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.0,
|
| 190 |
+
"eval_accuracy": 0.833904109589041,
|
| 191 |
+
"eval_loss": 0.43060269951820374,
|
| 192 |
+
"eval_macro_f1": 0.8332285980130937,
|
| 193 |
+
"eval_runtime": 1.0111,
|
| 194 |
+
"eval_samples_per_second": 577.567,
|
| 195 |
+
"eval_steps_per_second": 36.592,
|
| 196 |
+
"eval_weighted_f1": 0.8332468364120059,
|
| 197 |
+
"step": 573
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0104849279161208,
|
| 201 |
+
"grad_norm": 23.629840850830078,
|
| 202 |
+
"learning_rate": 1.8883071553228622e-05,
|
| 203 |
+
"loss": 1.5131,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.1415465268676277,
|
| 208 |
+
"grad_norm": 47.1800422668457,
|
| 209 |
+
"learning_rate": 1.8737638161721933e-05,
|
| 210 |
+
"loss": 1.3236,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.272608125819135,
|
| 215 |
+
"grad_norm": 82.11866760253906,
|
| 216 |
+
"learning_rate": 1.8592204770215244e-05,
|
| 217 |
+
"loss": 1.1795,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4036697247706424,
|
| 222 |
+
"grad_norm": 27.585718154907227,
|
| 223 |
+
"learning_rate": 1.844677137870855e-05,
|
| 224 |
+
"loss": 1.2008,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.5347313237221494,
|
| 229 |
+
"grad_norm": 64.93595123291016,
|
| 230 |
+
"learning_rate": 1.8301337987201865e-05,
|
| 231 |
+
"loss": 1.1102,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.6657929226736568,
|
| 236 |
+
"grad_norm": 60.96466827392578,
|
| 237 |
+
"learning_rate": 1.8155904595695173e-05,
|
| 238 |
+
"loss": 1.1672,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.7968545216251637,
|
| 243 |
+
"grad_norm": 43.670074462890625,
|
| 244 |
+
"learning_rate": 1.8010471204188483e-05,
|
| 245 |
+
"loss": 1.2541,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.927916120576671,
|
| 250 |
+
"grad_norm": 30.041912078857422,
|
| 251 |
+
"learning_rate": 1.7865037812681794e-05,
|
| 252 |
+
"loss": 1.1274,
|
| 253 |
+
"step": 750
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 4.0,
|
| 257 |
+
"eval_accuracy": 0.8184931506849316,
|
| 258 |
+
"eval_loss": 0.4712919592857361,
|
| 259 |
+
"eval_macro_f1": 0.8189065851360052,
|
| 260 |
+
"eval_runtime": 1.0243,
|
| 261 |
+
"eval_samples_per_second": 570.168,
|
| 262 |
+
"eval_steps_per_second": 36.124,
|
| 263 |
+
"eval_weighted_f1": 0.8188688163970698,
|
| 264 |
+
"step": 764
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.057667103538663,
|
| 268 |
+
"grad_norm": 45.4857292175293,
|
| 269 |
+
"learning_rate": 1.77196044211751e-05,
|
| 270 |
+
"loss": 1.0849,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.18872870249017,
|
| 275 |
+
"grad_norm": 40.71622085571289,
|
| 276 |
+
"learning_rate": 1.7574171029668416e-05,
|
| 277 |
+
"loss": 0.7316,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.319790301441677,
|
| 282 |
+
"grad_norm": 62.60859298706055,
|
| 283 |
+
"learning_rate": 1.7428737638161723e-05,
|
| 284 |
+
"loss": 0.825,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.450851900393185,
|
| 289 |
+
"grad_norm": 27.907306671142578,
|
| 290 |
+
"learning_rate": 1.7283304246655034e-05,
|
| 291 |
+
"loss": 0.8251,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.581913499344692,
|
| 296 |
+
"grad_norm": 36.16027069091797,
|
| 297 |
+
"learning_rate": 1.7137870855148345e-05,
|
| 298 |
+
"loss": 0.846,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.712975098296199,
|
| 303 |
+
"grad_norm": 59.31807327270508,
|
| 304 |
+
"learning_rate": 1.6992437463641652e-05,
|
| 305 |
+
"loss": 0.8842,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.844036697247707,
|
| 310 |
+
"grad_norm": 115.43113708496094,
|
| 311 |
+
"learning_rate": 1.6847004072134963e-05,
|
| 312 |
+
"loss": 0.8585,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 4.975098296199214,
|
| 317 |
+
"grad_norm": 33.352783203125,
|
| 318 |
+
"learning_rate": 1.6701570680628274e-05,
|
| 319 |
+
"loss": 0.8284,
|
| 320 |
+
"step": 950
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.0,
|
| 324 |
+
"eval_accuracy": 0.839041095890411,
|
| 325 |
+
"eval_loss": 0.5166526436805725,
|
| 326 |
+
"eval_macro_f1": 0.838960125962807,
|
| 327 |
+
"eval_runtime": 1.0257,
|
| 328 |
+
"eval_samples_per_second": 569.367,
|
| 329 |
+
"eval_steps_per_second": 36.073,
|
| 330 |
+
"eval_weighted_f1": 0.8390335286782824,
|
| 331 |
+
"step": 955
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.104849279161206,
|
| 335 |
+
"grad_norm": 115.93670654296875,
|
| 336 |
+
"learning_rate": 1.6556137289121585e-05,
|
| 337 |
+
"loss": 0.6029,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.235910878112713,
|
| 342 |
+
"grad_norm": 35.721553802490234,
|
| 343 |
+
"learning_rate": 1.6410703897614895e-05,
|
| 344 |
+
"loss": 0.5398,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.36697247706422,
|
| 349 |
+
"grad_norm": 77.26508331298828,
|
| 350 |
+
"learning_rate": 1.6265270506108203e-05,
|
| 351 |
+
"loss": 0.6449,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.498034076015728,
|
| 356 |
+
"grad_norm": 88.83381652832031,
|
| 357 |
+
"learning_rate": 1.6119837114601514e-05,
|
| 358 |
+
"loss": 0.788,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.6290956749672345,
|
| 363 |
+
"grad_norm": 18.69657325744629,
|
| 364 |
+
"learning_rate": 1.5974403723094824e-05,
|
| 365 |
+
"loss": 0.6488,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.760157273918741,
|
| 370 |
+
"grad_norm": 51.9311637878418,
|
| 371 |
+
"learning_rate": 1.5828970331588135e-05,
|
| 372 |
+
"loss": 0.7094,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.891218872870249,
|
| 377 |
+
"grad_norm": 29.13077163696289,
|
| 378 |
+
"learning_rate": 1.5683536940081443e-05,
|
| 379 |
+
"loss": 0.5296,
|
| 380 |
+
"step": 1125
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.0,
|
| 384 |
+
"eval_accuracy": 0.8476027397260274,
|
| 385 |
+
"eval_loss": 0.5409419536590576,
|
| 386 |
+
"eval_macro_f1": 0.8473890353589603,
|
| 387 |
+
"eval_runtime": 1.0337,
|
| 388 |
+
"eval_samples_per_second": 564.977,
|
| 389 |
+
"eval_steps_per_second": 35.795,
|
| 390 |
+
"eval_weighted_f1": 0.8472895122524333,
|
| 391 |
+
"step": 1146
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.0209698558322415,
|
| 395 |
+
"grad_norm": 28.083419799804688,
|
| 396 |
+
"learning_rate": 1.5538103548574753e-05,
|
| 397 |
+
"loss": 0.6039,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.1520314547837485,
|
| 402 |
+
"grad_norm": 57.95465087890625,
|
| 403 |
+
"learning_rate": 1.5392670157068064e-05,
|
| 404 |
+
"loss": 0.4103,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.283093053735255,
|
| 409 |
+
"grad_norm": 90.1124267578125,
|
| 410 |
+
"learning_rate": 1.5247236765561375e-05,
|
| 411 |
+
"loss": 0.3741,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.414154652686763,
|
| 416 |
+
"grad_norm": 73.2316665649414,
|
| 417 |
+
"learning_rate": 1.5101803374054684e-05,
|
| 418 |
+
"loss": 0.5888,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.54521625163827,
|
| 423 |
+
"grad_norm": 19.490676879882812,
|
| 424 |
+
"learning_rate": 1.4956369982547993e-05,
|
| 425 |
+
"loss": 0.5103,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.676277850589777,
|
| 430 |
+
"grad_norm": 21.10092544555664,
|
| 431 |
+
"learning_rate": 1.4810936591041306e-05,
|
| 432 |
+
"loss": 0.6769,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.807339449541285,
|
| 437 |
+
"grad_norm": 1.7164210081100464,
|
| 438 |
+
"learning_rate": 1.4665503199534615e-05,
|
| 439 |
+
"loss": 0.3104,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 6.938401048492792,
|
| 444 |
+
"grad_norm": 66.02993774414062,
|
| 445 |
+
"learning_rate": 1.4520069808027924e-05,
|
| 446 |
+
"loss": 0.6902,
|
| 447 |
+
"step": 1325
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 7.0,
|
| 451 |
+
"eval_accuracy": 0.8476027397260274,
|
| 452 |
+
"eval_loss": 0.7329329252243042,
|
| 453 |
+
"eval_macro_f1": 0.8482001147311004,
|
| 454 |
+
"eval_runtime": 1.0016,
|
| 455 |
+
"eval_samples_per_second": 583.055,
|
| 456 |
+
"eval_steps_per_second": 36.94,
|
| 457 |
+
"eval_weighted_f1": 0.8479630934295177,
|
| 458 |
+
"step": 1337
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.068152031454784,
|
| 462 |
+
"grad_norm": 1.1677628755569458,
|
| 463 |
+
"learning_rate": 1.4374636416521235e-05,
|
| 464 |
+
"loss": 0.4478,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.199213630406291,
|
| 469 |
+
"grad_norm": 218.09445190429688,
|
| 470 |
+
"learning_rate": 1.4229203025014544e-05,
|
| 471 |
+
"loss": 0.3396,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.330275229357798,
|
| 476 |
+
"grad_norm": 147.76715087890625,
|
| 477 |
+
"learning_rate": 1.4083769633507855e-05,
|
| 478 |
+
"loss": 0.3215,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.461336828309306,
|
| 483 |
+
"grad_norm": 145.50807189941406,
|
| 484 |
+
"learning_rate": 1.3938336242001166e-05,
|
| 485 |
+
"loss": 0.305,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.592398427260813,
|
| 490 |
+
"grad_norm": 3.1571762561798096,
|
| 491 |
+
"learning_rate": 1.3792902850494475e-05,
|
| 492 |
+
"loss": 0.2453,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 7.72346002621232,
|
| 497 |
+
"grad_norm": 3.3650453090667725,
|
| 498 |
+
"learning_rate": 1.3647469458987785e-05,
|
| 499 |
+
"loss": 0.3413,
|
| 500 |
+
"step": 1475
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 7.854521625163827,
|
| 504 |
+
"grad_norm": 0.1421426385641098,
|
| 505 |
+
"learning_rate": 1.3502036067481095e-05,
|
| 506 |
+
"loss": 0.3495,
|
| 507 |
+
"step": 1500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 7.985583224115334,
|
| 511 |
+
"grad_norm": 109.3970718383789,
|
| 512 |
+
"learning_rate": 1.3356602675974404e-05,
|
| 513 |
+
"loss": 0.3587,
|
| 514 |
+
"step": 1525
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 8.0,
|
| 518 |
+
"eval_accuracy": 0.851027397260274,
|
| 519 |
+
"eval_loss": 0.8212750554084778,
|
| 520 |
+
"eval_macro_f1": 0.8506965399706874,
|
| 521 |
+
"eval_runtime": 1.0664,
|
| 522 |
+
"eval_samples_per_second": 547.622,
|
| 523 |
+
"eval_steps_per_second": 34.695,
|
| 524 |
+
"eval_weighted_f1": 0.8505670696665327,
|
| 525 |
+
"step": 1528
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.115334207077327,
|
| 529 |
+
"grad_norm": 0.0987938866019249,
|
| 530 |
+
"learning_rate": 1.3211169284467716e-05,
|
| 531 |
+
"loss": 0.3193,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.246395806028833,
|
| 536 |
+
"grad_norm": 0.7317501306533813,
|
| 537 |
+
"learning_rate": 1.3065735892961025e-05,
|
| 538 |
+
"loss": 0.2249,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.37745740498034,
|
| 543 |
+
"grad_norm": 0.957346498966217,
|
| 544 |
+
"learning_rate": 1.2920302501454336e-05,
|
| 545 |
+
"loss": 0.2025,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.508519003931848,
|
| 550 |
+
"grad_norm": 0.2890245020389557,
|
| 551 |
+
"learning_rate": 1.2774869109947645e-05,
|
| 552 |
+
"loss": 0.2176,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 8.639580602883354,
|
| 557 |
+
"grad_norm": 142.0948028564453,
|
| 558 |
+
"learning_rate": 1.2629435718440954e-05,
|
| 559 |
+
"loss": 0.6597,
|
| 560 |
+
"step": 1650
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 8.770642201834862,
|
| 564 |
+
"grad_norm": 76.02135467529297,
|
| 565 |
+
"learning_rate": 1.2484002326934265e-05,
|
| 566 |
+
"loss": 0.3491,
|
| 567 |
+
"step": 1675
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 8.90170380078637,
|
| 571 |
+
"grad_norm": 2.2590975761413574,
|
| 572 |
+
"learning_rate": 1.2338568935427576e-05,
|
| 573 |
+
"loss": 0.2138,
|
| 574 |
+
"step": 1700
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 9.0,
|
| 578 |
+
"eval_accuracy": 0.8476027397260274,
|
| 579 |
+
"eval_loss": 0.8997184038162231,
|
| 580 |
+
"eval_macro_f1": 0.8483238673456065,
|
| 581 |
+
"eval_runtime": 0.9979,
|
| 582 |
+
"eval_samples_per_second": 585.204,
|
| 583 |
+
"eval_steps_per_second": 37.076,
|
| 584 |
+
"eval_weighted_f1": 0.8481780013709361,
|
| 585 |
+
"step": 1719
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.031454783748362,
|
| 589 |
+
"grad_norm": 1.409485936164856,
|
| 590 |
+
"learning_rate": 1.2193135543920885e-05,
|
| 591 |
+
"loss": 0.276,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.162516382699868,
|
| 596 |
+
"grad_norm": 0.10782130062580109,
|
| 597 |
+
"learning_rate": 1.2047702152414196e-05,
|
| 598 |
+
"loss": 0.0335,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.293577981651376,
|
| 603 |
+
"grad_norm": 0.5844252109527588,
|
| 604 |
+
"learning_rate": 1.1902268760907505e-05,
|
| 605 |
+
"loss": 0.0894,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.424639580602884,
|
| 610 |
+
"grad_norm": 211.50326538085938,
|
| 611 |
+
"learning_rate": 1.1756835369400816e-05,
|
| 612 |
+
"loss": 0.186,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 9.55570117955439,
|
| 617 |
+
"grad_norm": 0.058763690292835236,
|
| 618 |
+
"learning_rate": 1.1611401977894125e-05,
|
| 619 |
+
"loss": 0.2332,
|
| 620 |
+
"step": 1825
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 9.686762778505898,
|
| 624 |
+
"grad_norm": 15.001838684082031,
|
| 625 |
+
"learning_rate": 1.1465968586387436e-05,
|
| 626 |
+
"loss": 0.274,
|
| 627 |
+
"step": 1850
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 9.817824377457406,
|
| 631 |
+
"grad_norm": 0.09573527425527573,
|
| 632 |
+
"learning_rate": 1.1320535194880746e-05,
|
| 633 |
+
"loss": 0.1568,
|
| 634 |
+
"step": 1875
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 9.948885976408912,
|
| 638 |
+
"grad_norm": 0.5249341130256653,
|
| 639 |
+
"learning_rate": 1.1175101803374056e-05,
|
| 640 |
+
"loss": 0.147,
|
| 641 |
+
"step": 1900
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 10.0,
|
| 645 |
+
"eval_accuracy": 0.8407534246575342,
|
| 646 |
+
"eval_loss": 0.9433169364929199,
|
| 647 |
+
"eval_macro_f1": 0.8409421747799389,
|
| 648 |
+
"eval_runtime": 1.0191,
|
| 649 |
+
"eval_samples_per_second": 573.032,
|
| 650 |
+
"eval_steps_per_second": 36.305,
|
| 651 |
+
"eval_weighted_f1": 0.8407558538704119,
|
| 652 |
+
"step": 1910
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.078636959370904,
|
| 656 |
+
"grad_norm": 0.34368014335632324,
|
| 657 |
+
"learning_rate": 1.1029668411867365e-05,
|
| 658 |
+
"loss": 0.2835,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.209698558322412,
|
| 663 |
+
"grad_norm": 7.597753047943115,
|
| 664 |
+
"learning_rate": 1.0884235020360675e-05,
|
| 665 |
+
"loss": 0.2162,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.34076015727392,
|
| 670 |
+
"grad_norm": 0.053371064364910126,
|
| 671 |
+
"learning_rate": 1.0738801628853985e-05,
|
| 672 |
+
"loss": 0.1818,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.471821756225426,
|
| 677 |
+
"grad_norm": 148.0724334716797,
|
| 678 |
+
"learning_rate": 1.0593368237347297e-05,
|
| 679 |
+
"loss": 0.198,
|
| 680 |
+
"step": 2000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 10.602883355176933,
|
| 684 |
+
"grad_norm": 0.07850133627653122,
|
| 685 |
+
"learning_rate": 1.0447934845840606e-05,
|
| 686 |
+
"loss": 0.0937,
|
| 687 |
+
"step": 2025
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 10.73394495412844,
|
| 691 |
+
"grad_norm": 26.821496963500977,
|
| 692 |
+
"learning_rate": 1.0302501454333915e-05,
|
| 693 |
+
"loss": 0.2139,
|
| 694 |
+
"step": 2050
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 10.865006553079947,
|
| 698 |
+
"grad_norm": 0.04208998382091522,
|
| 699 |
+
"learning_rate": 1.0157068062827226e-05,
|
| 700 |
+
"loss": 0.1247,
|
| 701 |
+
"step": 2075
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 10.996068152031455,
|
| 705 |
+
"grad_norm": 0.18289154767990112,
|
| 706 |
+
"learning_rate": 1.0011634671320535e-05,
|
| 707 |
+
"loss": 0.3352,
|
| 708 |
+
"step": 2100
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 11.0,
|
| 712 |
+
"eval_accuracy": 0.851027397260274,
|
| 713 |
+
"eval_loss": 0.9814741611480713,
|
| 714 |
+
"eval_macro_f1": 0.8495393583628877,
|
| 715 |
+
"eval_runtime": 1.3029,
|
| 716 |
+
"eval_samples_per_second": 448.224,
|
| 717 |
+
"eval_steps_per_second": 28.398,
|
| 718 |
+
"eval_weighted_f1": 0.849475590153471,
|
| 719 |
+
"step": 2101
|
| 720 |
+
}
|
| 721 |
+
],
|
| 722 |
+
"logging_steps": 25,
|
| 723 |
+
"max_steps": 3820,
|
| 724 |
+
"num_input_tokens_seen": 0,
|
| 725 |
+
"num_train_epochs": 20,
|
| 726 |
+
"save_steps": 500,
|
| 727 |
+
"stateful_callbacks": {
|
| 728 |
+
"EarlyStoppingCallback": {
|
| 729 |
+
"args": {
|
| 730 |
+
"early_stopping_patience": 3,
|
| 731 |
+
"early_stopping_threshold": 0.0
|
| 732 |
+
},
|
| 733 |
+
"attributes": {
|
| 734 |
+
"early_stopping_patience_counter": 3
|
| 735 |
+
}
|
| 736 |
+
},
|
| 737 |
+
"TrainerControl": {
|
| 738 |
+
"args": {
|
| 739 |
+
"should_epoch_stop": false,
|
| 740 |
+
"should_evaluate": false,
|
| 741 |
+
"should_log": false,
|
| 742 |
+
"should_save": true,
|
| 743 |
+
"should_training_stop": true
|
| 744 |
+
},
|
| 745 |
+
"attributes": {}
|
| 746 |
+
}
|
| 747 |
+
},
|
| 748 |
+
"total_flos": 0.0,
|
| 749 |
+
"train_batch_size": 8,
|
| 750 |
+
"trial_name": null,
|
| 751 |
+
"trial_params": null
|
| 752 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/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-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/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-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730/trainer_state.json
ADDED
|
@@ -0,0 +1,971 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2730,
|
| 3 |
+
"best_metric": 0.8574577676289322,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730",
|
| 5 |
+
"epoch": 15.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2730,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13793103448275862,
|
| 14 |
+
"grad_norm": 9.43657112121582,
|
| 15 |
+
"learning_rate": 1.3186813186813187e-06,
|
| 16 |
+
"loss": 4.4249,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27586206896551724,
|
| 21 |
+
"grad_norm": 16.028423309326172,
|
| 22 |
+
"learning_rate": 2.6923076923076923e-06,
|
| 23 |
+
"loss": 4.4211,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.41379310344827586,
|
| 28 |
+
"grad_norm": 12.593598365783691,
|
| 29 |
+
"learning_rate": 4.065934065934066e-06,
|
| 30 |
+
"loss": 4.4057,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5517241379310345,
|
| 35 |
+
"grad_norm": 8.144038200378418,
|
| 36 |
+
"learning_rate": 5.43956043956044e-06,
|
| 37 |
+
"loss": 4.4096,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6896551724137931,
|
| 42 |
+
"grad_norm": 9.942596435546875,
|
| 43 |
+
"learning_rate": 6.813186813186814e-06,
|
| 44 |
+
"loss": 4.3861,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8275862068965517,
|
| 49 |
+
"grad_norm": 21.9925479888916,
|
| 50 |
+
"learning_rate": 8.186813186813188e-06,
|
| 51 |
+
"loss": 4.4044,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9655172413793104,
|
| 56 |
+
"grad_norm": 20.884540557861328,
|
| 57 |
+
"learning_rate": 9.560439560439562e-06,
|
| 58 |
+
"loss": 4.338,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.40755467196819084,
|
| 64 |
+
"eval_loss": 1.0724403858184814,
|
| 65 |
+
"eval_macro_f1": 0.29875293319978163,
|
| 66 |
+
"eval_runtime": 0.9139,
|
| 67 |
+
"eval_samples_per_second": 550.386,
|
| 68 |
+
"eval_steps_per_second": 35.015,
|
| 69 |
+
"eval_weighted_f1": 0.30219928462182194,
|
| 70 |
+
"step": 182
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0993103448275863,
|
| 74 |
+
"grad_norm": 11.181312561035156,
|
| 75 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 76 |
+
"loss": 4.1621,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.237241379310345,
|
| 81 |
+
"grad_norm": 17.32267189025879,
|
| 82 |
+
"learning_rate": 1.230769230769231e-05,
|
| 83 |
+
"loss": 4.2027,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3751724137931034,
|
| 88 |
+
"grad_norm": 23.262279510498047,
|
| 89 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 90 |
+
"loss": 4.1425,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.513103448275862,
|
| 95 |
+
"grad_norm": 39.08340835571289,
|
| 96 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 97 |
+
"loss": 3.6646,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6510344827586207,
|
| 102 |
+
"grad_norm": 43.55769348144531,
|
| 103 |
+
"learning_rate": 1.642857142857143e-05,
|
| 104 |
+
"loss": 3.2564,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7889655172413792,
|
| 109 |
+
"grad_norm": 55.561180114746094,
|
| 110 |
+
"learning_rate": 1.78021978021978e-05,
|
| 111 |
+
"loss": 2.9829,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.926896551724138,
|
| 116 |
+
"grad_norm": 51.493289947509766,
|
| 117 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 118 |
+
"loss": 2.6634,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7654075546719682,
|
| 124 |
+
"eval_loss": 0.5650879740715027,
|
| 125 |
+
"eval_macro_f1": 0.7625662618615071,
|
| 126 |
+
"eval_runtime": 0.9259,
|
| 127 |
+
"eval_samples_per_second": 543.249,
|
| 128 |
+
"eval_steps_per_second": 34.561,
|
| 129 |
+
"eval_weighted_f1": 0.763033267982337,
|
| 130 |
+
"step": 364
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0606896551724136,
|
| 134 |
+
"grad_norm": 101.3663558959961,
|
| 135 |
+
"learning_rate": 1.993894993894994e-05,
|
| 136 |
+
"loss": 2.1245,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.1986206896551725,
|
| 141 |
+
"grad_norm": 43.45347213745117,
|
| 142 |
+
"learning_rate": 1.9786324786324788e-05,
|
| 143 |
+
"loss": 2.0671,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.336551724137931,
|
| 148 |
+
"grad_norm": 46.45637512207031,
|
| 149 |
+
"learning_rate": 1.9633699633699634e-05,
|
| 150 |
+
"loss": 2.2904,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.47448275862069,
|
| 155 |
+
"grad_norm": 39.08163833618164,
|
| 156 |
+
"learning_rate": 1.948107448107448e-05,
|
| 157 |
+
"loss": 2.3561,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.6124137931034483,
|
| 162 |
+
"grad_norm": 38.376033782958984,
|
| 163 |
+
"learning_rate": 1.932844932844933e-05,
|
| 164 |
+
"loss": 2.1347,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.750344827586207,
|
| 169 |
+
"grad_norm": 42.2293586730957,
|
| 170 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 171 |
+
"loss": 2.3107,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8882758620689657,
|
| 176 |
+
"grad_norm": 39.29929733276367,
|
| 177 |
+
"learning_rate": 1.9023199023199025e-05,
|
| 178 |
+
"loss": 2.1769,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.827037773359841,
|
| 184 |
+
"eval_loss": 0.4584631025791168,
|
| 185 |
+
"eval_macro_f1": 0.8268769922976227,
|
| 186 |
+
"eval_runtime": 0.9474,
|
| 187 |
+
"eval_samples_per_second": 530.925,
|
| 188 |
+
"eval_steps_per_second": 33.777,
|
| 189 |
+
"eval_weighted_f1": 0.8269840998585147,
|
| 190 |
+
"step": 546
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0220689655172412,
|
| 194 |
+
"grad_norm": 23.395872116088867,
|
| 195 |
+
"learning_rate": 1.887057387057387e-05,
|
| 196 |
+
"loss": 1.8972,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.16,
|
| 201 |
+
"grad_norm": 30.511388778686523,
|
| 202 |
+
"learning_rate": 1.8717948717948718e-05,
|
| 203 |
+
"loss": 1.3141,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.2979310344827586,
|
| 208 |
+
"grad_norm": 64.49130249023438,
|
| 209 |
+
"learning_rate": 1.8565323565323568e-05,
|
| 210 |
+
"loss": 1.6227,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.435862068965517,
|
| 215 |
+
"grad_norm": 54.51080322265625,
|
| 216 |
+
"learning_rate": 1.8412698412698415e-05,
|
| 217 |
+
"loss": 1.8948,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.573793103448276,
|
| 222 |
+
"grad_norm": 35.362667083740234,
|
| 223 |
+
"learning_rate": 1.826007326007326e-05,
|
| 224 |
+
"loss": 1.3856,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.7117241379310344,
|
| 229 |
+
"grad_norm": 181.82293701171875,
|
| 230 |
+
"learning_rate": 1.8107448107448108e-05,
|
| 231 |
+
"loss": 1.6695,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.849655172413793,
|
| 236 |
+
"grad_norm": 76.79501342773438,
|
| 237 |
+
"learning_rate": 1.7954822954822958e-05,
|
| 238 |
+
"loss": 1.5666,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9875862068965517,
|
| 243 |
+
"grad_norm": 46.8802375793457,
|
| 244 |
+
"learning_rate": 1.78021978021978e-05,
|
| 245 |
+
"loss": 1.542,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8210735586481114,
|
| 251 |
+
"eval_loss": 0.49113520979881287,
|
| 252 |
+
"eval_macro_f1": 0.8158191156068834,
|
| 253 |
+
"eval_runtime": 0.9952,
|
| 254 |
+
"eval_samples_per_second": 505.415,
|
| 255 |
+
"eval_steps_per_second": 32.154,
|
| 256 |
+
"eval_weighted_f1": 0.8162911618242914,
|
| 257 |
+
"step": 728
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.121379310344827,
|
| 261 |
+
"grad_norm": 105.35623168945312,
|
| 262 |
+
"learning_rate": 1.764957264957265e-05,
|
| 263 |
+
"loss": 1.0373,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.259310344827586,
|
| 268 |
+
"grad_norm": 60.612892150878906,
|
| 269 |
+
"learning_rate": 1.74969474969475e-05,
|
| 270 |
+
"loss": 1.1573,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.397241379310345,
|
| 275 |
+
"grad_norm": 47.914344787597656,
|
| 276 |
+
"learning_rate": 1.7344322344322345e-05,
|
| 277 |
+
"loss": 1.0237,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.535172413793103,
|
| 282 |
+
"grad_norm": 87.36337280273438,
|
| 283 |
+
"learning_rate": 1.7191697191697195e-05,
|
| 284 |
+
"loss": 0.9113,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.673103448275862,
|
| 289 |
+
"grad_norm": 35.61204528808594,
|
| 290 |
+
"learning_rate": 1.703907203907204e-05,
|
| 291 |
+
"loss": 1.0866,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.811034482758621,
|
| 296 |
+
"grad_norm": 43.72821044921875,
|
| 297 |
+
"learning_rate": 1.688644688644689e-05,
|
| 298 |
+
"loss": 1.031,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.94896551724138,
|
| 303 |
+
"grad_norm": 40.79308319091797,
|
| 304 |
+
"learning_rate": 1.6733821733821735e-05,
|
| 305 |
+
"loss": 1.0079,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.8230616302186878,
|
| 311 |
+
"eval_loss": 0.5469569563865662,
|
| 312 |
+
"eval_macro_f1": 0.824703148332205,
|
| 313 |
+
"eval_runtime": 0.9478,
|
| 314 |
+
"eval_samples_per_second": 530.698,
|
| 315 |
+
"eval_steps_per_second": 33.762,
|
| 316 |
+
"eval_weighted_f1": 0.8248769915975103,
|
| 317 |
+
"step": 910
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.082758620689655,
|
| 321 |
+
"grad_norm": 126.51851654052734,
|
| 322 |
+
"learning_rate": 1.6581196581196585e-05,
|
| 323 |
+
"loss": 0.9714,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.220689655172414,
|
| 328 |
+
"grad_norm": 44.50932312011719,
|
| 329 |
+
"learning_rate": 1.642857142857143e-05,
|
| 330 |
+
"loss": 0.8657,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.358620689655172,
|
| 335 |
+
"grad_norm": 196.72007751464844,
|
| 336 |
+
"learning_rate": 1.627594627594628e-05,
|
| 337 |
+
"loss": 0.6074,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.496551724137931,
|
| 342 |
+
"grad_norm": 50.4857177734375,
|
| 343 |
+
"learning_rate": 1.6123321123321125e-05,
|
| 344 |
+
"loss": 0.7527,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.63448275862069,
|
| 349 |
+
"grad_norm": 51.30732727050781,
|
| 350 |
+
"learning_rate": 1.5970695970695972e-05,
|
| 351 |
+
"loss": 0.9741,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.772413793103448,
|
| 356 |
+
"grad_norm": 82.42403411865234,
|
| 357 |
+
"learning_rate": 1.581807081807082e-05,
|
| 358 |
+
"loss": 0.8336,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.910344827586207,
|
| 363 |
+
"grad_norm": 103.34027099609375,
|
| 364 |
+
"learning_rate": 1.5665445665445665e-05,
|
| 365 |
+
"loss": 0.7503,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8290258449304175,
|
| 371 |
+
"eval_loss": 0.5356364250183105,
|
| 372 |
+
"eval_macro_f1": 0.8297661927833738,
|
| 373 |
+
"eval_runtime": 0.9548,
|
| 374 |
+
"eval_samples_per_second": 526.828,
|
| 375 |
+
"eval_steps_per_second": 33.516,
|
| 376 |
+
"eval_weighted_f1": 0.8301065499138283,
|
| 377 |
+
"step": 1092
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0441379310344825,
|
| 381 |
+
"grad_norm": 58.315521240234375,
|
| 382 |
+
"learning_rate": 1.5512820512820516e-05,
|
| 383 |
+
"loss": 0.5779,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.182068965517241,
|
| 388 |
+
"grad_norm": 12.383352279663086,
|
| 389 |
+
"learning_rate": 1.5360195360195362e-05,
|
| 390 |
+
"loss": 0.5786,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.32,
|
| 395 |
+
"grad_norm": 67.50831604003906,
|
| 396 |
+
"learning_rate": 1.5207570207570207e-05,
|
| 397 |
+
"loss": 0.3959,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.457931034482758,
|
| 402 |
+
"grad_norm": 85.18661499023438,
|
| 403 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 404 |
+
"loss": 0.3711,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.595862068965517,
|
| 409 |
+
"grad_norm": 88.58021545410156,
|
| 410 |
+
"learning_rate": 1.4902319902319904e-05,
|
| 411 |
+
"loss": 0.4629,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.733793103448276,
|
| 416 |
+
"grad_norm": 45.826438903808594,
|
| 417 |
+
"learning_rate": 1.4749694749694752e-05,
|
| 418 |
+
"loss": 0.4317,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.871724137931034,
|
| 423 |
+
"grad_norm": 28.398372650146484,
|
| 424 |
+
"learning_rate": 1.4597069597069597e-05,
|
| 425 |
+
"loss": 0.3339,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.0,
|
| 430 |
+
"eval_accuracy": 0.8349900596421471,
|
| 431 |
+
"eval_loss": 0.6359565854072571,
|
| 432 |
+
"eval_macro_f1": 0.8349158738315178,
|
| 433 |
+
"eval_runtime": 0.9259,
|
| 434 |
+
"eval_samples_per_second": 543.267,
|
| 435 |
+
"eval_steps_per_second": 34.562,
|
| 436 |
+
"eval_weighted_f1": 0.8352236729087633,
|
| 437 |
+
"step": 1274
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.0055172413793105,
|
| 441 |
+
"grad_norm": 100.83720397949219,
|
| 442 |
+
"learning_rate": 1.4444444444444446e-05,
|
| 443 |
+
"loss": 0.5956,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.1434482758620685,
|
| 448 |
+
"grad_norm": 49.7368278503418,
|
| 449 |
+
"learning_rate": 1.4291819291819293e-05,
|
| 450 |
+
"loss": 0.3889,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.281379310344827,
|
| 455 |
+
"grad_norm": 91.60435485839844,
|
| 456 |
+
"learning_rate": 1.4139194139194141e-05,
|
| 457 |
+
"loss": 0.3232,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.419310344827586,
|
| 462 |
+
"grad_norm": 36.7932014465332,
|
| 463 |
+
"learning_rate": 1.3986568986568988e-05,
|
| 464 |
+
"loss": 0.3001,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.557241379310344,
|
| 469 |
+
"grad_norm": 257.08465576171875,
|
| 470 |
+
"learning_rate": 1.3833943833943834e-05,
|
| 471 |
+
"loss": 0.5329,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.695172413793103,
|
| 476 |
+
"grad_norm": 117.8155288696289,
|
| 477 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 478 |
+
"loss": 0.4343,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.833103448275862,
|
| 483 |
+
"grad_norm": 179.53582763671875,
|
| 484 |
+
"learning_rate": 1.3528693528693531e-05,
|
| 485 |
+
"loss": 0.3409,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.971034482758621,
|
| 490 |
+
"grad_norm": 2.515775442123413,
|
| 491 |
+
"learning_rate": 1.3376068376068376e-05,
|
| 492 |
+
"loss": 0.5128,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8190854870775348,
|
| 498 |
+
"eval_loss": 0.8731481432914734,
|
| 499 |
+
"eval_macro_f1": 0.8229860861858401,
|
| 500 |
+
"eval_runtime": 0.9501,
|
| 501 |
+
"eval_samples_per_second": 529.402,
|
| 502 |
+
"eval_steps_per_second": 33.68,
|
| 503 |
+
"eval_weighted_f1": 0.8230893776293666,
|
| 504 |
+
"step": 1456
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.104827586206897,
|
| 508 |
+
"grad_norm": 82.6282730102539,
|
| 509 |
+
"learning_rate": 1.3223443223443224e-05,
|
| 510 |
+
"loss": 0.536,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.242758620689655,
|
| 515 |
+
"grad_norm": 22.45798110961914,
|
| 516 |
+
"learning_rate": 1.3070818070818073e-05,
|
| 517 |
+
"loss": 0.196,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.380689655172414,
|
| 522 |
+
"grad_norm": 23.877168655395508,
|
| 523 |
+
"learning_rate": 1.291819291819292e-05,
|
| 524 |
+
"loss": 0.3102,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.518620689655172,
|
| 529 |
+
"grad_norm": 171.82057189941406,
|
| 530 |
+
"learning_rate": 1.2765567765567766e-05,
|
| 531 |
+
"loss": 0.3352,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.65655172413793,
|
| 536 |
+
"grad_norm": 155.93370056152344,
|
| 537 |
+
"learning_rate": 1.2612942612942613e-05,
|
| 538 |
+
"loss": 0.2375,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.79448275862069,
|
| 543 |
+
"grad_norm": 0.9011493921279907,
|
| 544 |
+
"learning_rate": 1.2460317460317461e-05,
|
| 545 |
+
"loss": 0.359,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.932413793103448,
|
| 550 |
+
"grad_norm": 126.37146759033203,
|
| 551 |
+
"learning_rate": 1.230769230769231e-05,
|
| 552 |
+
"loss": 0.349,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8330019880715706,
|
| 558 |
+
"eval_loss": 0.9226741194725037,
|
| 559 |
+
"eval_macro_f1": 0.8293588610162166,
|
| 560 |
+
"eval_runtime": 0.9239,
|
| 561 |
+
"eval_samples_per_second": 544.421,
|
| 562 |
+
"eval_steps_per_second": 34.635,
|
| 563 |
+
"eval_weighted_f1": 0.8296478430978693,
|
| 564 |
+
"step": 1638
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.066206896551725,
|
| 568 |
+
"grad_norm": 124.03196716308594,
|
| 569 |
+
"learning_rate": 1.2155067155067155e-05,
|
| 570 |
+
"loss": 0.2682,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.204137931034483,
|
| 575 |
+
"grad_norm": 1.5294833183288574,
|
| 576 |
+
"learning_rate": 1.2002442002442003e-05,
|
| 577 |
+
"loss": 0.2656,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.34206896551724,
|
| 582 |
+
"grad_norm": 1.651520848274231,
|
| 583 |
+
"learning_rate": 1.1849816849816852e-05,
|
| 584 |
+
"loss": 0.4269,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.48,
|
| 589 |
+
"grad_norm": 278.5240173339844,
|
| 590 |
+
"learning_rate": 1.1697191697191698e-05,
|
| 591 |
+
"loss": 0.1675,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.617931034482758,
|
| 596 |
+
"grad_norm": 0.18507838249206543,
|
| 597 |
+
"learning_rate": 1.1544566544566545e-05,
|
| 598 |
+
"loss": 0.2896,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.755862068965516,
|
| 603 |
+
"grad_norm": 8.146827697753906,
|
| 604 |
+
"learning_rate": 1.1391941391941392e-05,
|
| 605 |
+
"loss": 0.2364,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.893793103448276,
|
| 610 |
+
"grad_norm": 96.84983825683594,
|
| 611 |
+
"learning_rate": 1.123931623931624e-05,
|
| 612 |
+
"loss": 0.1734,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 10.0,
|
| 617 |
+
"eval_accuracy": 0.8409542743538767,
|
| 618 |
+
"eval_loss": 0.9496588706970215,
|
| 619 |
+
"eval_macro_f1": 0.8387830874881091,
|
| 620 |
+
"eval_runtime": 0.9469,
|
| 621 |
+
"eval_samples_per_second": 531.218,
|
| 622 |
+
"eval_steps_per_second": 33.795,
|
| 623 |
+
"eval_weighted_f1": 0.8390050725328778,
|
| 624 |
+
"step": 1820
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.027586206896551,
|
| 628 |
+
"grad_norm": 0.07375728338956833,
|
| 629 |
+
"learning_rate": 1.1086691086691088e-05,
|
| 630 |
+
"loss": 0.3168,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.16551724137931,
|
| 635 |
+
"grad_norm": 423.2017517089844,
|
| 636 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 637 |
+
"loss": 0.169,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.303448275862069,
|
| 642 |
+
"grad_norm": 0.09648483246564865,
|
| 643 |
+
"learning_rate": 1.0781440781440782e-05,
|
| 644 |
+
"loss": 0.0424,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.441379310344828,
|
| 649 |
+
"grad_norm": 0.17975841462612152,
|
| 650 |
+
"learning_rate": 1.062881562881563e-05,
|
| 651 |
+
"loss": 0.2072,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.579310344827586,
|
| 656 |
+
"grad_norm": 6.902804374694824,
|
| 657 |
+
"learning_rate": 1.0476190476190477e-05,
|
| 658 |
+
"loss": 0.341,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.717241379310344,
|
| 663 |
+
"grad_norm": 54.266353607177734,
|
| 664 |
+
"learning_rate": 1.0323565323565325e-05,
|
| 665 |
+
"loss": 0.3564,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.855172413793104,
|
| 670 |
+
"grad_norm": 172.48313903808594,
|
| 671 |
+
"learning_rate": 1.0170940170940172e-05,
|
| 672 |
+
"loss": 0.1952,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.993103448275862,
|
| 677 |
+
"grad_norm": 0.08419492095708847,
|
| 678 |
+
"learning_rate": 1.0018315018315019e-05,
|
| 679 |
+
"loss": 0.0761,
|
| 680 |
+
"step": 2000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 11.0,
|
| 684 |
+
"eval_accuracy": 0.8469184890656064,
|
| 685 |
+
"eval_loss": 0.9737921357154846,
|
| 686 |
+
"eval_macro_f1": 0.8472094898565486,
|
| 687 |
+
"eval_runtime": 0.9118,
|
| 688 |
+
"eval_samples_per_second": 551.635,
|
| 689 |
+
"eval_steps_per_second": 35.094,
|
| 690 |
+
"eval_weighted_f1": 0.8474748402885464,
|
| 691 |
+
"step": 2002
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.126896551724139,
|
| 695 |
+
"grad_norm": 73.43778228759766,
|
| 696 |
+
"learning_rate": 9.865689865689867e-06,
|
| 697 |
+
"loss": 0.1141,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.264827586206897,
|
| 702 |
+
"grad_norm": 6.327709197998047,
|
| 703 |
+
"learning_rate": 9.713064713064714e-06,
|
| 704 |
+
"loss": 0.1164,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.402758620689655,
|
| 709 |
+
"grad_norm": 0.040171872824430466,
|
| 710 |
+
"learning_rate": 9.560439560439562e-06,
|
| 711 |
+
"loss": 0.0681,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.540689655172415,
|
| 716 |
+
"grad_norm": 311.22369384765625,
|
| 717 |
+
"learning_rate": 9.407814407814409e-06,
|
| 718 |
+
"loss": 0.1006,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 11.678620689655173,
|
| 723 |
+
"grad_norm": 9.650760650634766,
|
| 724 |
+
"learning_rate": 9.255189255189256e-06,
|
| 725 |
+
"loss": 0.1442,
|
| 726 |
+
"step": 2125
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 11.81655172413793,
|
| 730 |
+
"grad_norm": 15.422511100769043,
|
| 731 |
+
"learning_rate": 9.102564102564104e-06,
|
| 732 |
+
"loss": 0.2103,
|
| 733 |
+
"step": 2150
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 11.95448275862069,
|
| 737 |
+
"grad_norm": 0.7053766250610352,
|
| 738 |
+
"learning_rate": 8.94993894993895e-06,
|
| 739 |
+
"loss": 0.1577,
|
| 740 |
+
"step": 2175
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 12.0,
|
| 744 |
+
"eval_accuracy": 0.8429423459244533,
|
| 745 |
+
"eval_loss": 1.0143959522247314,
|
| 746 |
+
"eval_macro_f1": 0.8411135330976793,
|
| 747 |
+
"eval_runtime": 0.915,
|
| 748 |
+
"eval_samples_per_second": 549.752,
|
| 749 |
+
"eval_steps_per_second": 34.974,
|
| 750 |
+
"eval_weighted_f1": 0.8415853859985961,
|
| 751 |
+
"step": 2184
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.088275862068965,
|
| 755 |
+
"grad_norm": 0.038472745567560196,
|
| 756 |
+
"learning_rate": 8.797313797313797e-06,
|
| 757 |
+
"loss": 0.073,
|
| 758 |
+
"step": 2200
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.226206896551725,
|
| 762 |
+
"grad_norm": 0.04019385576248169,
|
| 763 |
+
"learning_rate": 8.644688644688646e-06,
|
| 764 |
+
"loss": 0.1522,
|
| 765 |
+
"step": 2225
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.364137931034483,
|
| 769 |
+
"grad_norm": 0.033529944717884064,
|
| 770 |
+
"learning_rate": 8.492063492063492e-06,
|
| 771 |
+
"loss": 0.0328,
|
| 772 |
+
"step": 2250
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.50206896551724,
|
| 776 |
+
"grad_norm": 0.04767317324876785,
|
| 777 |
+
"learning_rate": 8.33943833943834e-06,
|
| 778 |
+
"loss": 0.0756,
|
| 779 |
+
"step": 2275
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 12.64,
|
| 783 |
+
"grad_norm": 0.03198707476258278,
|
| 784 |
+
"learning_rate": 8.186813186813188e-06,
|
| 785 |
+
"loss": 0.2246,
|
| 786 |
+
"step": 2300
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 12.777931034482759,
|
| 790 |
+
"grad_norm": 0.40969544649124146,
|
| 791 |
+
"learning_rate": 8.034188034188036e-06,
|
| 792 |
+
"loss": 0.0589,
|
| 793 |
+
"step": 2325
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 12.915862068965517,
|
| 797 |
+
"grad_norm": 55.27804183959961,
|
| 798 |
+
"learning_rate": 7.881562881562883e-06,
|
| 799 |
+
"loss": 0.0141,
|
| 800 |
+
"step": 2350
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 13.0,
|
| 804 |
+
"eval_accuracy": 0.8469184890656064,
|
| 805 |
+
"eval_loss": 1.0392986536026,
|
| 806 |
+
"eval_macro_f1": 0.8466356795359288,
|
| 807 |
+
"eval_runtime": 0.9381,
|
| 808 |
+
"eval_samples_per_second": 536.17,
|
| 809 |
+
"eval_steps_per_second": 34.11,
|
| 810 |
+
"eval_weighted_f1": 0.8470035482910769,
|
| 811 |
+
"step": 2366
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 13.049655172413793,
|
| 815 |
+
"grad_norm": 0.030963147059082985,
|
| 816 |
+
"learning_rate": 7.72893772893773e-06,
|
| 817 |
+
"loss": 0.035,
|
| 818 |
+
"step": 2375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 13.187586206896551,
|
| 822 |
+
"grad_norm": 0.026960330083966255,
|
| 823 |
+
"learning_rate": 7.576312576312577e-06,
|
| 824 |
+
"loss": 0.1324,
|
| 825 |
+
"step": 2400
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 13.32551724137931,
|
| 829 |
+
"grad_norm": 0.07455019652843475,
|
| 830 |
+
"learning_rate": 7.423687423687424e-06,
|
| 831 |
+
"loss": 0.1329,
|
| 832 |
+
"step": 2425
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 13.463448275862069,
|
| 836 |
+
"grad_norm": 81.79376220703125,
|
| 837 |
+
"learning_rate": 7.271062271062271e-06,
|
| 838 |
+
"loss": 0.0955,
|
| 839 |
+
"step": 2450
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 13.601379310344827,
|
| 843 |
+
"grad_norm": 0.1130983754992485,
|
| 844 |
+
"learning_rate": 7.1184371184371195e-06,
|
| 845 |
+
"loss": 0.0444,
|
| 846 |
+
"step": 2475
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 13.739310344827587,
|
| 850 |
+
"grad_norm": 0.04253191500902176,
|
| 851 |
+
"learning_rate": 6.965811965811966e-06,
|
| 852 |
+
"loss": 0.138,
|
| 853 |
+
"step": 2500
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 13.877241379310345,
|
| 857 |
+
"grad_norm": 2.6513543128967285,
|
| 858 |
+
"learning_rate": 6.813186813186814e-06,
|
| 859 |
+
"loss": 0.0751,
|
| 860 |
+
"step": 2525
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 14.0,
|
| 864 |
+
"eval_accuracy": 0.8548707753479126,
|
| 865 |
+
"eval_loss": 1.043471336364746,
|
| 866 |
+
"eval_macro_f1": 0.854641935778914,
|
| 867 |
+
"eval_runtime": 0.9362,
|
| 868 |
+
"eval_samples_per_second": 537.261,
|
| 869 |
+
"eval_steps_per_second": 34.18,
|
| 870 |
+
"eval_weighted_f1": 0.8551078401602452,
|
| 871 |
+
"step": 2548
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 14.011034482758621,
|
| 875 |
+
"grad_norm": 0.03401171788573265,
|
| 876 |
+
"learning_rate": 6.66056166056166e-06,
|
| 877 |
+
"loss": 0.1265,
|
| 878 |
+
"step": 2550
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 14.148965517241379,
|
| 882 |
+
"grad_norm": 0.031571391969919205,
|
| 883 |
+
"learning_rate": 6.507936507936509e-06,
|
| 884 |
+
"loss": 0.0758,
|
| 885 |
+
"step": 2575
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 14.286896551724137,
|
| 889 |
+
"grad_norm": 0.03989469259977341,
|
| 890 |
+
"learning_rate": 6.3553113553113555e-06,
|
| 891 |
+
"loss": 0.0818,
|
| 892 |
+
"step": 2600
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 14.424827586206897,
|
| 896 |
+
"grad_norm": 0.03423743695020676,
|
| 897 |
+
"learning_rate": 6.202686202686203e-06,
|
| 898 |
+
"loss": 0.0781,
|
| 899 |
+
"step": 2625
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 14.562758620689655,
|
| 903 |
+
"grad_norm": 278.81634521484375,
|
| 904 |
+
"learning_rate": 6.05006105006105e-06,
|
| 905 |
+
"loss": 0.0458,
|
| 906 |
+
"step": 2650
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 14.700689655172415,
|
| 910 |
+
"grad_norm": 0.025289248675107956,
|
| 911 |
+
"learning_rate": 5.897435897435898e-06,
|
| 912 |
+
"loss": 0.017,
|
| 913 |
+
"step": 2675
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 14.838620689655173,
|
| 917 |
+
"grad_norm": 480.5322265625,
|
| 918 |
+
"learning_rate": 5.744810744810745e-06,
|
| 919 |
+
"loss": 0.0526,
|
| 920 |
+
"step": 2700
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 14.97655172413793,
|
| 924 |
+
"grad_norm": 0.025748319923877716,
|
| 925 |
+
"learning_rate": 5.592185592185592e-06,
|
| 926 |
+
"loss": 0.1215,
|
| 927 |
+
"step": 2725
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 15.0,
|
| 931 |
+
"eval_accuracy": 0.856858846918489,
|
| 932 |
+
"eval_loss": 1.0679168701171875,
|
| 933 |
+
"eval_macro_f1": 0.8574577676289322,
|
| 934 |
+
"eval_runtime": 0.9146,
|
| 935 |
+
"eval_samples_per_second": 549.982,
|
| 936 |
+
"eval_steps_per_second": 34.989,
|
| 937 |
+
"eval_weighted_f1": 0.8578825875514354,
|
| 938 |
+
"step": 2730
|
| 939 |
+
}
|
| 940 |
+
],
|
| 941 |
+
"logging_steps": 25,
|
| 942 |
+
"max_steps": 3640,
|
| 943 |
+
"num_input_tokens_seen": 0,
|
| 944 |
+
"num_train_epochs": 20,
|
| 945 |
+
"save_steps": 500,
|
| 946 |
+
"stateful_callbacks": {
|
| 947 |
+
"EarlyStoppingCallback": {
|
| 948 |
+
"args": {
|
| 949 |
+
"early_stopping_patience": 3,
|
| 950 |
+
"early_stopping_threshold": 0.0
|
| 951 |
+
},
|
| 952 |
+
"attributes": {
|
| 953 |
+
"early_stopping_patience_counter": 0
|
| 954 |
+
}
|
| 955 |
+
},
|
| 956 |
+
"TrainerControl": {
|
| 957 |
+
"args": {
|
| 958 |
+
"should_epoch_stop": false,
|
| 959 |
+
"should_evaluate": false,
|
| 960 |
+
"should_log": false,
|
| 961 |
+
"should_save": true,
|
| 962 |
+
"should_training_stop": false
|
| 963 |
+
},
|
| 964 |
+
"attributes": {}
|
| 965 |
+
}
|
| 966 |
+
},
|
| 967 |
+
"total_flos": 0.0,
|
| 968 |
+
"train_batch_size": 8,
|
| 969 |
+
"trial_name": null,
|
| 970 |
+
"trial_params": null
|
| 971 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/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-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/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-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-3276/trainer_state.json
ADDED
|
@@ -0,0 +1,1158 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2730,
|
| 3 |
+
"best_metric": 0.8574577676289322,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-02-s44-20260828T202123Z-496d6871/trainer/checkpoint-2730",
|
| 5 |
+
"epoch": 18.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3276,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13793103448275862,
|
| 14 |
+
"grad_norm": 9.43657112121582,
|
| 15 |
+
"learning_rate": 1.3186813186813187e-06,
|
| 16 |
+
"loss": 4.4249,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27586206896551724,
|
| 21 |
+
"grad_norm": 16.028423309326172,
|
| 22 |
+
"learning_rate": 2.6923076923076923e-06,
|
| 23 |
+
"loss": 4.4211,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.41379310344827586,
|
| 28 |
+
"grad_norm": 12.593598365783691,
|
| 29 |
+
"learning_rate": 4.065934065934066e-06,
|
| 30 |
+
"loss": 4.4057,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5517241379310345,
|
| 35 |
+
"grad_norm": 8.144038200378418,
|
| 36 |
+
"learning_rate": 5.43956043956044e-06,
|
| 37 |
+
"loss": 4.4096,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6896551724137931,
|
| 42 |
+
"grad_norm": 9.942596435546875,
|
| 43 |
+
"learning_rate": 6.813186813186814e-06,
|
| 44 |
+
"loss": 4.3861,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8275862068965517,
|
| 49 |
+
"grad_norm": 21.9925479888916,
|
| 50 |
+
"learning_rate": 8.186813186813188e-06,
|
| 51 |
+
"loss": 4.4044,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9655172413793104,
|
| 56 |
+
"grad_norm": 20.884540557861328,
|
| 57 |
+
"learning_rate": 9.560439560439562e-06,
|
| 58 |
+
"loss": 4.338,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.40755467196819084,
|
| 64 |
+
"eval_loss": 1.0724403858184814,
|
| 65 |
+
"eval_macro_f1": 0.29875293319978163,
|
| 66 |
+
"eval_runtime": 0.9139,
|
| 67 |
+
"eval_samples_per_second": 550.386,
|
| 68 |
+
"eval_steps_per_second": 35.015,
|
| 69 |
+
"eval_weighted_f1": 0.30219928462182194,
|
| 70 |
+
"step": 182
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0993103448275863,
|
| 74 |
+
"grad_norm": 11.181312561035156,
|
| 75 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 76 |
+
"loss": 4.1621,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.237241379310345,
|
| 81 |
+
"grad_norm": 17.32267189025879,
|
| 82 |
+
"learning_rate": 1.230769230769231e-05,
|
| 83 |
+
"loss": 4.2027,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3751724137931034,
|
| 88 |
+
"grad_norm": 23.262279510498047,
|
| 89 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 90 |
+
"loss": 4.1425,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.513103448275862,
|
| 95 |
+
"grad_norm": 39.08340835571289,
|
| 96 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 97 |
+
"loss": 3.6646,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6510344827586207,
|
| 102 |
+
"grad_norm": 43.55769348144531,
|
| 103 |
+
"learning_rate": 1.642857142857143e-05,
|
| 104 |
+
"loss": 3.2564,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7889655172413792,
|
| 109 |
+
"grad_norm": 55.561180114746094,
|
| 110 |
+
"learning_rate": 1.78021978021978e-05,
|
| 111 |
+
"loss": 2.9829,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.926896551724138,
|
| 116 |
+
"grad_norm": 51.493289947509766,
|
| 117 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 118 |
+
"loss": 2.6634,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7654075546719682,
|
| 124 |
+
"eval_loss": 0.5650879740715027,
|
| 125 |
+
"eval_macro_f1": 0.7625662618615071,
|
| 126 |
+
"eval_runtime": 0.9259,
|
| 127 |
+
"eval_samples_per_second": 543.249,
|
| 128 |
+
"eval_steps_per_second": 34.561,
|
| 129 |
+
"eval_weighted_f1": 0.763033267982337,
|
| 130 |
+
"step": 364
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0606896551724136,
|
| 134 |
+
"grad_norm": 101.3663558959961,
|
| 135 |
+
"learning_rate": 1.993894993894994e-05,
|
| 136 |
+
"loss": 2.1245,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.1986206896551725,
|
| 141 |
+
"grad_norm": 43.45347213745117,
|
| 142 |
+
"learning_rate": 1.9786324786324788e-05,
|
| 143 |
+
"loss": 2.0671,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.336551724137931,
|
| 148 |
+
"grad_norm": 46.45637512207031,
|
| 149 |
+
"learning_rate": 1.9633699633699634e-05,
|
| 150 |
+
"loss": 2.2904,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.47448275862069,
|
| 155 |
+
"grad_norm": 39.08163833618164,
|
| 156 |
+
"learning_rate": 1.948107448107448e-05,
|
| 157 |
+
"loss": 2.3561,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.6124137931034483,
|
| 162 |
+
"grad_norm": 38.376033782958984,
|
| 163 |
+
"learning_rate": 1.932844932844933e-05,
|
| 164 |
+
"loss": 2.1347,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.750344827586207,
|
| 169 |
+
"grad_norm": 42.2293586730957,
|
| 170 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 171 |
+
"loss": 2.3107,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8882758620689657,
|
| 176 |
+
"grad_norm": 39.29929733276367,
|
| 177 |
+
"learning_rate": 1.9023199023199025e-05,
|
| 178 |
+
"loss": 2.1769,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.827037773359841,
|
| 184 |
+
"eval_loss": 0.4584631025791168,
|
| 185 |
+
"eval_macro_f1": 0.8268769922976227,
|
| 186 |
+
"eval_runtime": 0.9474,
|
| 187 |
+
"eval_samples_per_second": 530.925,
|
| 188 |
+
"eval_steps_per_second": 33.777,
|
| 189 |
+
"eval_weighted_f1": 0.8269840998585147,
|
| 190 |
+
"step": 546
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0220689655172412,
|
| 194 |
+
"grad_norm": 23.395872116088867,
|
| 195 |
+
"learning_rate": 1.887057387057387e-05,
|
| 196 |
+
"loss": 1.8972,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.16,
|
| 201 |
+
"grad_norm": 30.511388778686523,
|
| 202 |
+
"learning_rate": 1.8717948717948718e-05,
|
| 203 |
+
"loss": 1.3141,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.2979310344827586,
|
| 208 |
+
"grad_norm": 64.49130249023438,
|
| 209 |
+
"learning_rate": 1.8565323565323568e-05,
|
| 210 |
+
"loss": 1.6227,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.435862068965517,
|
| 215 |
+
"grad_norm": 54.51080322265625,
|
| 216 |
+
"learning_rate": 1.8412698412698415e-05,
|
| 217 |
+
"loss": 1.8948,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.573793103448276,
|
| 222 |
+
"grad_norm": 35.362667083740234,
|
| 223 |
+
"learning_rate": 1.826007326007326e-05,
|
| 224 |
+
"loss": 1.3856,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.7117241379310344,
|
| 229 |
+
"grad_norm": 181.82293701171875,
|
| 230 |
+
"learning_rate": 1.8107448107448108e-05,
|
| 231 |
+
"loss": 1.6695,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.849655172413793,
|
| 236 |
+
"grad_norm": 76.79501342773438,
|
| 237 |
+
"learning_rate": 1.7954822954822958e-05,
|
| 238 |
+
"loss": 1.5666,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9875862068965517,
|
| 243 |
+
"grad_norm": 46.8802375793457,
|
| 244 |
+
"learning_rate": 1.78021978021978e-05,
|
| 245 |
+
"loss": 1.542,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8210735586481114,
|
| 251 |
+
"eval_loss": 0.49113520979881287,
|
| 252 |
+
"eval_macro_f1": 0.8158191156068834,
|
| 253 |
+
"eval_runtime": 0.9952,
|
| 254 |
+
"eval_samples_per_second": 505.415,
|
| 255 |
+
"eval_steps_per_second": 32.154,
|
| 256 |
+
"eval_weighted_f1": 0.8162911618242914,
|
| 257 |
+
"step": 728
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.121379310344827,
|
| 261 |
+
"grad_norm": 105.35623168945312,
|
| 262 |
+
"learning_rate": 1.764957264957265e-05,
|
| 263 |
+
"loss": 1.0373,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.259310344827586,
|
| 268 |
+
"grad_norm": 60.612892150878906,
|
| 269 |
+
"learning_rate": 1.74969474969475e-05,
|
| 270 |
+
"loss": 1.1573,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.397241379310345,
|
| 275 |
+
"grad_norm": 47.914344787597656,
|
| 276 |
+
"learning_rate": 1.7344322344322345e-05,
|
| 277 |
+
"loss": 1.0237,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.535172413793103,
|
| 282 |
+
"grad_norm": 87.36337280273438,
|
| 283 |
+
"learning_rate": 1.7191697191697195e-05,
|
| 284 |
+
"loss": 0.9113,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.673103448275862,
|
| 289 |
+
"grad_norm": 35.61204528808594,
|
| 290 |
+
"learning_rate": 1.703907203907204e-05,
|
| 291 |
+
"loss": 1.0866,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.811034482758621,
|
| 296 |
+
"grad_norm": 43.72821044921875,
|
| 297 |
+
"learning_rate": 1.688644688644689e-05,
|
| 298 |
+
"loss": 1.031,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.94896551724138,
|
| 303 |
+
"grad_norm": 40.79308319091797,
|
| 304 |
+
"learning_rate": 1.6733821733821735e-05,
|
| 305 |
+
"loss": 1.0079,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.8230616302186878,
|
| 311 |
+
"eval_loss": 0.5469569563865662,
|
| 312 |
+
"eval_macro_f1": 0.824703148332205,
|
| 313 |
+
"eval_runtime": 0.9478,
|
| 314 |
+
"eval_samples_per_second": 530.698,
|
| 315 |
+
"eval_steps_per_second": 33.762,
|
| 316 |
+
"eval_weighted_f1": 0.8248769915975103,
|
| 317 |
+
"step": 910
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.082758620689655,
|
| 321 |
+
"grad_norm": 126.51851654052734,
|
| 322 |
+
"learning_rate": 1.6581196581196585e-05,
|
| 323 |
+
"loss": 0.9714,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.220689655172414,
|
| 328 |
+
"grad_norm": 44.50932312011719,
|
| 329 |
+
"learning_rate": 1.642857142857143e-05,
|
| 330 |
+
"loss": 0.8657,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.358620689655172,
|
| 335 |
+
"grad_norm": 196.72007751464844,
|
| 336 |
+
"learning_rate": 1.627594627594628e-05,
|
| 337 |
+
"loss": 0.6074,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.496551724137931,
|
| 342 |
+
"grad_norm": 50.4857177734375,
|
| 343 |
+
"learning_rate": 1.6123321123321125e-05,
|
| 344 |
+
"loss": 0.7527,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.63448275862069,
|
| 349 |
+
"grad_norm": 51.30732727050781,
|
| 350 |
+
"learning_rate": 1.5970695970695972e-05,
|
| 351 |
+
"loss": 0.9741,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.772413793103448,
|
| 356 |
+
"grad_norm": 82.42403411865234,
|
| 357 |
+
"learning_rate": 1.581807081807082e-05,
|
| 358 |
+
"loss": 0.8336,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.910344827586207,
|
| 363 |
+
"grad_norm": 103.34027099609375,
|
| 364 |
+
"learning_rate": 1.5665445665445665e-05,
|
| 365 |
+
"loss": 0.7503,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8290258449304175,
|
| 371 |
+
"eval_loss": 0.5356364250183105,
|
| 372 |
+
"eval_macro_f1": 0.8297661927833738,
|
| 373 |
+
"eval_runtime": 0.9548,
|
| 374 |
+
"eval_samples_per_second": 526.828,
|
| 375 |
+
"eval_steps_per_second": 33.516,
|
| 376 |
+
"eval_weighted_f1": 0.8301065499138283,
|
| 377 |
+
"step": 1092
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0441379310344825,
|
| 381 |
+
"grad_norm": 58.315521240234375,
|
| 382 |
+
"learning_rate": 1.5512820512820516e-05,
|
| 383 |
+
"loss": 0.5779,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.182068965517241,
|
| 388 |
+
"grad_norm": 12.383352279663086,
|
| 389 |
+
"learning_rate": 1.5360195360195362e-05,
|
| 390 |
+
"loss": 0.5786,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.32,
|
| 395 |
+
"grad_norm": 67.50831604003906,
|
| 396 |
+
"learning_rate": 1.5207570207570207e-05,
|
| 397 |
+
"loss": 0.3959,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.457931034482758,
|
| 402 |
+
"grad_norm": 85.18661499023438,
|
| 403 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 404 |
+
"loss": 0.3711,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.595862068965517,
|
| 409 |
+
"grad_norm": 88.58021545410156,
|
| 410 |
+
"learning_rate": 1.4902319902319904e-05,
|
| 411 |
+
"loss": 0.4629,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.733793103448276,
|
| 416 |
+
"grad_norm": 45.826438903808594,
|
| 417 |
+
"learning_rate": 1.4749694749694752e-05,
|
| 418 |
+
"loss": 0.4317,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.871724137931034,
|
| 423 |
+
"grad_norm": 28.398372650146484,
|
| 424 |
+
"learning_rate": 1.4597069597069597e-05,
|
| 425 |
+
"loss": 0.3339,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.0,
|
| 430 |
+
"eval_accuracy": 0.8349900596421471,
|
| 431 |
+
"eval_loss": 0.6359565854072571,
|
| 432 |
+
"eval_macro_f1": 0.8349158738315178,
|
| 433 |
+
"eval_runtime": 0.9259,
|
| 434 |
+
"eval_samples_per_second": 543.267,
|
| 435 |
+
"eval_steps_per_second": 34.562,
|
| 436 |
+
"eval_weighted_f1": 0.8352236729087633,
|
| 437 |
+
"step": 1274
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.0055172413793105,
|
| 441 |
+
"grad_norm": 100.83720397949219,
|
| 442 |
+
"learning_rate": 1.4444444444444446e-05,
|
| 443 |
+
"loss": 0.5956,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.1434482758620685,
|
| 448 |
+
"grad_norm": 49.7368278503418,
|
| 449 |
+
"learning_rate": 1.4291819291819293e-05,
|
| 450 |
+
"loss": 0.3889,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.281379310344827,
|
| 455 |
+
"grad_norm": 91.60435485839844,
|
| 456 |
+
"learning_rate": 1.4139194139194141e-05,
|
| 457 |
+
"loss": 0.3232,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.419310344827586,
|
| 462 |
+
"grad_norm": 36.7932014465332,
|
| 463 |
+
"learning_rate": 1.3986568986568988e-05,
|
| 464 |
+
"loss": 0.3001,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.557241379310344,
|
| 469 |
+
"grad_norm": 257.08465576171875,
|
| 470 |
+
"learning_rate": 1.3833943833943834e-05,
|
| 471 |
+
"loss": 0.5329,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.695172413793103,
|
| 476 |
+
"grad_norm": 117.8155288696289,
|
| 477 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 478 |
+
"loss": 0.4343,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.833103448275862,
|
| 483 |
+
"grad_norm": 179.53582763671875,
|
| 484 |
+
"learning_rate": 1.3528693528693531e-05,
|
| 485 |
+
"loss": 0.3409,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.971034482758621,
|
| 490 |
+
"grad_norm": 2.515775442123413,
|
| 491 |
+
"learning_rate": 1.3376068376068376e-05,
|
| 492 |
+
"loss": 0.5128,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8190854870775348,
|
| 498 |
+
"eval_loss": 0.8731481432914734,
|
| 499 |
+
"eval_macro_f1": 0.8229860861858401,
|
| 500 |
+
"eval_runtime": 0.9501,
|
| 501 |
+
"eval_samples_per_second": 529.402,
|
| 502 |
+
"eval_steps_per_second": 33.68,
|
| 503 |
+
"eval_weighted_f1": 0.8230893776293666,
|
| 504 |
+
"step": 1456
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.104827586206897,
|
| 508 |
+
"grad_norm": 82.6282730102539,
|
| 509 |
+
"learning_rate": 1.3223443223443224e-05,
|
| 510 |
+
"loss": 0.536,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.242758620689655,
|
| 515 |
+
"grad_norm": 22.45798110961914,
|
| 516 |
+
"learning_rate": 1.3070818070818073e-05,
|
| 517 |
+
"loss": 0.196,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.380689655172414,
|
| 522 |
+
"grad_norm": 23.877168655395508,
|
| 523 |
+
"learning_rate": 1.291819291819292e-05,
|
| 524 |
+
"loss": 0.3102,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.518620689655172,
|
| 529 |
+
"grad_norm": 171.82057189941406,
|
| 530 |
+
"learning_rate": 1.2765567765567766e-05,
|
| 531 |
+
"loss": 0.3352,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.65655172413793,
|
| 536 |
+
"grad_norm": 155.93370056152344,
|
| 537 |
+
"learning_rate": 1.2612942612942613e-05,
|
| 538 |
+
"loss": 0.2375,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.79448275862069,
|
| 543 |
+
"grad_norm": 0.9011493921279907,
|
| 544 |
+
"learning_rate": 1.2460317460317461e-05,
|
| 545 |
+
"loss": 0.359,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.932413793103448,
|
| 550 |
+
"grad_norm": 126.37146759033203,
|
| 551 |
+
"learning_rate": 1.230769230769231e-05,
|
| 552 |
+
"loss": 0.349,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8330019880715706,
|
| 558 |
+
"eval_loss": 0.9226741194725037,
|
| 559 |
+
"eval_macro_f1": 0.8293588610162166,
|
| 560 |
+
"eval_runtime": 0.9239,
|
| 561 |
+
"eval_samples_per_second": 544.421,
|
| 562 |
+
"eval_steps_per_second": 34.635,
|
| 563 |
+
"eval_weighted_f1": 0.8296478430978693,
|
| 564 |
+
"step": 1638
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.066206896551725,
|
| 568 |
+
"grad_norm": 124.03196716308594,
|
| 569 |
+
"learning_rate": 1.2155067155067155e-05,
|
| 570 |
+
"loss": 0.2682,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.204137931034483,
|
| 575 |
+
"grad_norm": 1.5294833183288574,
|
| 576 |
+
"learning_rate": 1.2002442002442003e-05,
|
| 577 |
+
"loss": 0.2656,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.34206896551724,
|
| 582 |
+
"grad_norm": 1.651520848274231,
|
| 583 |
+
"learning_rate": 1.1849816849816852e-05,
|
| 584 |
+
"loss": 0.4269,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.48,
|
| 589 |
+
"grad_norm": 278.5240173339844,
|
| 590 |
+
"learning_rate": 1.1697191697191698e-05,
|
| 591 |
+
"loss": 0.1675,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.617931034482758,
|
| 596 |
+
"grad_norm": 0.18507838249206543,
|
| 597 |
+
"learning_rate": 1.1544566544566545e-05,
|
| 598 |
+
"loss": 0.2896,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.755862068965516,
|
| 603 |
+
"grad_norm": 8.146827697753906,
|
| 604 |
+
"learning_rate": 1.1391941391941392e-05,
|
| 605 |
+
"loss": 0.2364,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.893793103448276,
|
| 610 |
+
"grad_norm": 96.84983825683594,
|
| 611 |
+
"learning_rate": 1.123931623931624e-05,
|
| 612 |
+
"loss": 0.1734,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 10.0,
|
| 617 |
+
"eval_accuracy": 0.8409542743538767,
|
| 618 |
+
"eval_loss": 0.9496588706970215,
|
| 619 |
+
"eval_macro_f1": 0.8387830874881091,
|
| 620 |
+
"eval_runtime": 0.9469,
|
| 621 |
+
"eval_samples_per_second": 531.218,
|
| 622 |
+
"eval_steps_per_second": 33.795,
|
| 623 |
+
"eval_weighted_f1": 0.8390050725328778,
|
| 624 |
+
"step": 1820
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.027586206896551,
|
| 628 |
+
"grad_norm": 0.07375728338956833,
|
| 629 |
+
"learning_rate": 1.1086691086691088e-05,
|
| 630 |
+
"loss": 0.3168,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.16551724137931,
|
| 635 |
+
"grad_norm": 423.2017517089844,
|
| 636 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 637 |
+
"loss": 0.169,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.303448275862069,
|
| 642 |
+
"grad_norm": 0.09648483246564865,
|
| 643 |
+
"learning_rate": 1.0781440781440782e-05,
|
| 644 |
+
"loss": 0.0424,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.441379310344828,
|
| 649 |
+
"grad_norm": 0.17975841462612152,
|
| 650 |
+
"learning_rate": 1.062881562881563e-05,
|
| 651 |
+
"loss": 0.2072,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.579310344827586,
|
| 656 |
+
"grad_norm": 6.902804374694824,
|
| 657 |
+
"learning_rate": 1.0476190476190477e-05,
|
| 658 |
+
"loss": 0.341,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.717241379310344,
|
| 663 |
+
"grad_norm": 54.266353607177734,
|
| 664 |
+
"learning_rate": 1.0323565323565325e-05,
|
| 665 |
+
"loss": 0.3564,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.855172413793104,
|
| 670 |
+
"grad_norm": 172.48313903808594,
|
| 671 |
+
"learning_rate": 1.0170940170940172e-05,
|
| 672 |
+
"loss": 0.1952,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.993103448275862,
|
| 677 |
+
"grad_norm": 0.08419492095708847,
|
| 678 |
+
"learning_rate": 1.0018315018315019e-05,
|
| 679 |
+
"loss": 0.0761,
|
| 680 |
+
"step": 2000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 11.0,
|
| 684 |
+
"eval_accuracy": 0.8469184890656064,
|
| 685 |
+
"eval_loss": 0.9737921357154846,
|
| 686 |
+
"eval_macro_f1": 0.8472094898565486,
|
| 687 |
+
"eval_runtime": 0.9118,
|
| 688 |
+
"eval_samples_per_second": 551.635,
|
| 689 |
+
"eval_steps_per_second": 35.094,
|
| 690 |
+
"eval_weighted_f1": 0.8474748402885464,
|
| 691 |
+
"step": 2002
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.126896551724139,
|
| 695 |
+
"grad_norm": 73.43778228759766,
|
| 696 |
+
"learning_rate": 9.865689865689867e-06,
|
| 697 |
+
"loss": 0.1141,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.264827586206897,
|
| 702 |
+
"grad_norm": 6.327709197998047,
|
| 703 |
+
"learning_rate": 9.713064713064714e-06,
|
| 704 |
+
"loss": 0.1164,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.402758620689655,
|
| 709 |
+
"grad_norm": 0.040171872824430466,
|
| 710 |
+
"learning_rate": 9.560439560439562e-06,
|
| 711 |
+
"loss": 0.0681,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.540689655172415,
|
| 716 |
+
"grad_norm": 311.22369384765625,
|
| 717 |
+
"learning_rate": 9.407814407814409e-06,
|
| 718 |
+
"loss": 0.1006,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 11.678620689655173,
|
| 723 |
+
"grad_norm": 9.650760650634766,
|
| 724 |
+
"learning_rate": 9.255189255189256e-06,
|
| 725 |
+
"loss": 0.1442,
|
| 726 |
+
"step": 2125
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 11.81655172413793,
|
| 730 |
+
"grad_norm": 15.422511100769043,
|
| 731 |
+
"learning_rate": 9.102564102564104e-06,
|
| 732 |
+
"loss": 0.2103,
|
| 733 |
+
"step": 2150
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 11.95448275862069,
|
| 737 |
+
"grad_norm": 0.7053766250610352,
|
| 738 |
+
"learning_rate": 8.94993894993895e-06,
|
| 739 |
+
"loss": 0.1577,
|
| 740 |
+
"step": 2175
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 12.0,
|
| 744 |
+
"eval_accuracy": 0.8429423459244533,
|
| 745 |
+
"eval_loss": 1.0143959522247314,
|
| 746 |
+
"eval_macro_f1": 0.8411135330976793,
|
| 747 |
+
"eval_runtime": 0.915,
|
| 748 |
+
"eval_samples_per_second": 549.752,
|
| 749 |
+
"eval_steps_per_second": 34.974,
|
| 750 |
+
"eval_weighted_f1": 0.8415853859985961,
|
| 751 |
+
"step": 2184
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.088275862068965,
|
| 755 |
+
"grad_norm": 0.038472745567560196,
|
| 756 |
+
"learning_rate": 8.797313797313797e-06,
|
| 757 |
+
"loss": 0.073,
|
| 758 |
+
"step": 2200
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.226206896551725,
|
| 762 |
+
"grad_norm": 0.04019385576248169,
|
| 763 |
+
"learning_rate": 8.644688644688646e-06,
|
| 764 |
+
"loss": 0.1522,
|
| 765 |
+
"step": 2225
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.364137931034483,
|
| 769 |
+
"grad_norm": 0.033529944717884064,
|
| 770 |
+
"learning_rate": 8.492063492063492e-06,
|
| 771 |
+
"loss": 0.0328,
|
| 772 |
+
"step": 2250
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.50206896551724,
|
| 776 |
+
"grad_norm": 0.04767317324876785,
|
| 777 |
+
"learning_rate": 8.33943833943834e-06,
|
| 778 |
+
"loss": 0.0756,
|
| 779 |
+
"step": 2275
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 12.64,
|
| 783 |
+
"grad_norm": 0.03198707476258278,
|
| 784 |
+
"learning_rate": 8.186813186813188e-06,
|
| 785 |
+
"loss": 0.2246,
|
| 786 |
+
"step": 2300
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 12.777931034482759,
|
| 790 |
+
"grad_norm": 0.40969544649124146,
|
| 791 |
+
"learning_rate": 8.034188034188036e-06,
|
| 792 |
+
"loss": 0.0589,
|
| 793 |
+
"step": 2325
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 12.915862068965517,
|
| 797 |
+
"grad_norm": 55.27804183959961,
|
| 798 |
+
"learning_rate": 7.881562881562883e-06,
|
| 799 |
+
"loss": 0.0141,
|
| 800 |
+
"step": 2350
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 13.0,
|
| 804 |
+
"eval_accuracy": 0.8469184890656064,
|
| 805 |
+
"eval_loss": 1.0392986536026,
|
| 806 |
+
"eval_macro_f1": 0.8466356795359288,
|
| 807 |
+
"eval_runtime": 0.9381,
|
| 808 |
+
"eval_samples_per_second": 536.17,
|
| 809 |
+
"eval_steps_per_second": 34.11,
|
| 810 |
+
"eval_weighted_f1": 0.8470035482910769,
|
| 811 |
+
"step": 2366
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 13.049655172413793,
|
| 815 |
+
"grad_norm": 0.030963147059082985,
|
| 816 |
+
"learning_rate": 7.72893772893773e-06,
|
| 817 |
+
"loss": 0.035,
|
| 818 |
+
"step": 2375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 13.187586206896551,
|
| 822 |
+
"grad_norm": 0.026960330083966255,
|
| 823 |
+
"learning_rate": 7.576312576312577e-06,
|
| 824 |
+
"loss": 0.1324,
|
| 825 |
+
"step": 2400
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 13.32551724137931,
|
| 829 |
+
"grad_norm": 0.07455019652843475,
|
| 830 |
+
"learning_rate": 7.423687423687424e-06,
|
| 831 |
+
"loss": 0.1329,
|
| 832 |
+
"step": 2425
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 13.463448275862069,
|
| 836 |
+
"grad_norm": 81.79376220703125,
|
| 837 |
+
"learning_rate": 7.271062271062271e-06,
|
| 838 |
+
"loss": 0.0955,
|
| 839 |
+
"step": 2450
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 13.601379310344827,
|
| 843 |
+
"grad_norm": 0.1130983754992485,
|
| 844 |
+
"learning_rate": 7.1184371184371195e-06,
|
| 845 |
+
"loss": 0.0444,
|
| 846 |
+
"step": 2475
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 13.739310344827587,
|
| 850 |
+
"grad_norm": 0.04253191500902176,
|
| 851 |
+
"learning_rate": 6.965811965811966e-06,
|
| 852 |
+
"loss": 0.138,
|
| 853 |
+
"step": 2500
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 13.877241379310345,
|
| 857 |
+
"grad_norm": 2.6513543128967285,
|
| 858 |
+
"learning_rate": 6.813186813186814e-06,
|
| 859 |
+
"loss": 0.0751,
|
| 860 |
+
"step": 2525
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 14.0,
|
| 864 |
+
"eval_accuracy": 0.8548707753479126,
|
| 865 |
+
"eval_loss": 1.043471336364746,
|
| 866 |
+
"eval_macro_f1": 0.854641935778914,
|
| 867 |
+
"eval_runtime": 0.9362,
|
| 868 |
+
"eval_samples_per_second": 537.261,
|
| 869 |
+
"eval_steps_per_second": 34.18,
|
| 870 |
+
"eval_weighted_f1": 0.8551078401602452,
|
| 871 |
+
"step": 2548
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 14.011034482758621,
|
| 875 |
+
"grad_norm": 0.03401171788573265,
|
| 876 |
+
"learning_rate": 6.66056166056166e-06,
|
| 877 |
+
"loss": 0.1265,
|
| 878 |
+
"step": 2550
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 14.148965517241379,
|
| 882 |
+
"grad_norm": 0.031571391969919205,
|
| 883 |
+
"learning_rate": 6.507936507936509e-06,
|
| 884 |
+
"loss": 0.0758,
|
| 885 |
+
"step": 2575
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 14.286896551724137,
|
| 889 |
+
"grad_norm": 0.03989469259977341,
|
| 890 |
+
"learning_rate": 6.3553113553113555e-06,
|
| 891 |
+
"loss": 0.0818,
|
| 892 |
+
"step": 2600
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 14.424827586206897,
|
| 896 |
+
"grad_norm": 0.03423743695020676,
|
| 897 |
+
"learning_rate": 6.202686202686203e-06,
|
| 898 |
+
"loss": 0.0781,
|
| 899 |
+
"step": 2625
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 14.562758620689655,
|
| 903 |
+
"grad_norm": 278.81634521484375,
|
| 904 |
+
"learning_rate": 6.05006105006105e-06,
|
| 905 |
+
"loss": 0.0458,
|
| 906 |
+
"step": 2650
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 14.700689655172415,
|
| 910 |
+
"grad_norm": 0.025289248675107956,
|
| 911 |
+
"learning_rate": 5.897435897435898e-06,
|
| 912 |
+
"loss": 0.017,
|
| 913 |
+
"step": 2675
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 14.838620689655173,
|
| 917 |
+
"grad_norm": 480.5322265625,
|
| 918 |
+
"learning_rate": 5.744810744810745e-06,
|
| 919 |
+
"loss": 0.0526,
|
| 920 |
+
"step": 2700
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 14.97655172413793,
|
| 924 |
+
"grad_norm": 0.025748319923877716,
|
| 925 |
+
"learning_rate": 5.592185592185592e-06,
|
| 926 |
+
"loss": 0.1215,
|
| 927 |
+
"step": 2725
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 15.0,
|
| 931 |
+
"eval_accuracy": 0.856858846918489,
|
| 932 |
+
"eval_loss": 1.0679168701171875,
|
| 933 |
+
"eval_macro_f1": 0.8574577676289322,
|
| 934 |
+
"eval_runtime": 0.9146,
|
| 935 |
+
"eval_samples_per_second": 549.982,
|
| 936 |
+
"eval_steps_per_second": 34.989,
|
| 937 |
+
"eval_weighted_f1": 0.8578825875514354,
|
| 938 |
+
"step": 2730
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 15.110344827586207,
|
| 942 |
+
"grad_norm": 0.02715623565018177,
|
| 943 |
+
"learning_rate": 5.43956043956044e-06,
|
| 944 |
+
"loss": 0.0462,
|
| 945 |
+
"step": 2750
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 15.248275862068965,
|
| 949 |
+
"grad_norm": 0.033547624945640564,
|
| 950 |
+
"learning_rate": 5.2869352869352875e-06,
|
| 951 |
+
"loss": 0.0525,
|
| 952 |
+
"step": 2775
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 15.386206896551725,
|
| 956 |
+
"grad_norm": 0.0275338776409626,
|
| 957 |
+
"learning_rate": 5.134310134310134e-06,
|
| 958 |
+
"loss": 0.0414,
|
| 959 |
+
"step": 2800
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 15.524137931034483,
|
| 963 |
+
"grad_norm": 0.023391880095005035,
|
| 964 |
+
"learning_rate": 4.9816849816849826e-06,
|
| 965 |
+
"loss": 0.0754,
|
| 966 |
+
"step": 2825
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 15.662068965517241,
|
| 970 |
+
"grad_norm": 384.76544189453125,
|
| 971 |
+
"learning_rate": 4.829059829059829e-06,
|
| 972 |
+
"loss": 0.0261,
|
| 973 |
+
"step": 2850
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 15.8,
|
| 977 |
+
"grad_norm": 0.2929709255695343,
|
| 978 |
+
"learning_rate": 4.676434676434677e-06,
|
| 979 |
+
"loss": 0.0036,
|
| 980 |
+
"step": 2875
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 15.937931034482759,
|
| 984 |
+
"grad_norm": 267.56744384765625,
|
| 985 |
+
"learning_rate": 4.523809523809524e-06,
|
| 986 |
+
"loss": 0.0379,
|
| 987 |
+
"step": 2900
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 16.0,
|
| 991 |
+
"eval_accuracy": 0.852882703777336,
|
| 992 |
+
"eval_loss": 1.0708694458007812,
|
| 993 |
+
"eval_macro_f1": 0.8522950329753924,
|
| 994 |
+
"eval_runtime": 0.9527,
|
| 995 |
+
"eval_samples_per_second": 527.96,
|
| 996 |
+
"eval_steps_per_second": 33.588,
|
| 997 |
+
"eval_weighted_f1": 0.8526407826448404,
|
| 998 |
+
"step": 2912
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 16.071724137931035,
|
| 1002 |
+
"grad_norm": 0.9050227403640747,
|
| 1003 |
+
"learning_rate": 4.371184371184372e-06,
|
| 1004 |
+
"loss": 0.0134,
|
| 1005 |
+
"step": 2925
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 16.209655172413793,
|
| 1009 |
+
"grad_norm": 0.019924191758036613,
|
| 1010 |
+
"learning_rate": 4.218559218559219e-06,
|
| 1011 |
+
"loss": 0.0637,
|
| 1012 |
+
"step": 2950
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 16.34758620689655,
|
| 1016 |
+
"grad_norm": 0.02051049843430519,
|
| 1017 |
+
"learning_rate": 4.065934065934066e-06,
|
| 1018 |
+
"loss": 0.0379,
|
| 1019 |
+
"step": 2975
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 16.48551724137931,
|
| 1023 |
+
"grad_norm": 0.024199359118938446,
|
| 1024 |
+
"learning_rate": 3.913308913308914e-06,
|
| 1025 |
+
"loss": 0.0463,
|
| 1026 |
+
"step": 3000
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 16.623448275862067,
|
| 1030 |
+
"grad_norm": 83.3418960571289,
|
| 1031 |
+
"learning_rate": 3.760683760683761e-06,
|
| 1032 |
+
"loss": 0.1145,
|
| 1033 |
+
"step": 3025
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 16.76137931034483,
|
| 1037 |
+
"grad_norm": 0.020937908440828323,
|
| 1038 |
+
"learning_rate": 3.6080586080586083e-06,
|
| 1039 |
+
"loss": 0.0878,
|
| 1040 |
+
"step": 3050
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 16.899310344827587,
|
| 1044 |
+
"grad_norm": 0.03406038507819176,
|
| 1045 |
+
"learning_rate": 3.4554334554334555e-06,
|
| 1046 |
+
"loss": 0.0424,
|
| 1047 |
+
"step": 3075
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 17.0,
|
| 1051 |
+
"eval_accuracy": 0.8508946322067594,
|
| 1052 |
+
"eval_loss": 1.0796154737472534,
|
| 1053 |
+
"eval_macro_f1": 0.8497670218848441,
|
| 1054 |
+
"eval_runtime": 0.9395,
|
| 1055 |
+
"eval_samples_per_second": 535.364,
|
| 1056 |
+
"eval_steps_per_second": 34.059,
|
| 1057 |
+
"eval_weighted_f1": 0.8500986007789454,
|
| 1058 |
+
"step": 3094
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 17.033103448275863,
|
| 1062 |
+
"grad_norm": 298.5498046875,
|
| 1063 |
+
"learning_rate": 3.302808302808303e-06,
|
| 1064 |
+
"loss": 0.0179,
|
| 1065 |
+
"step": 3100
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 17.17103448275862,
|
| 1069 |
+
"grad_norm": 0.025099461898207664,
|
| 1070 |
+
"learning_rate": 3.1501831501831505e-06,
|
| 1071 |
+
"loss": 0.0041,
|
| 1072 |
+
"step": 3125
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 17.30896551724138,
|
| 1076 |
+
"grad_norm": 0.01884295605123043,
|
| 1077 |
+
"learning_rate": 2.9975579975579977e-06,
|
| 1078 |
+
"loss": 0.007,
|
| 1079 |
+
"step": 3150
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 17.446896551724137,
|
| 1083 |
+
"grad_norm": 0.034199103713035583,
|
| 1084 |
+
"learning_rate": 2.844932844932845e-06,
|
| 1085 |
+
"loss": 0.0388,
|
| 1086 |
+
"step": 3175
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 17.584827586206895,
|
| 1090 |
+
"grad_norm": 0.023717287927865982,
|
| 1091 |
+
"learning_rate": 2.6923076923076923e-06,
|
| 1092 |
+
"loss": 0.0933,
|
| 1093 |
+
"step": 3200
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 17.722758620689657,
|
| 1097 |
+
"grad_norm": 0.02133820578455925,
|
| 1098 |
+
"learning_rate": 2.53968253968254e-06,
|
| 1099 |
+
"loss": 0.0402,
|
| 1100 |
+
"step": 3225
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 17.860689655172415,
|
| 1104 |
+
"grad_norm": 0.020610710605978966,
|
| 1105 |
+
"learning_rate": 2.3870573870573874e-06,
|
| 1106 |
+
"loss": 0.0405,
|
| 1107 |
+
"step": 3250
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 17.998620689655173,
|
| 1111 |
+
"grad_norm": 0.022689277306199074,
|
| 1112 |
+
"learning_rate": 2.2344322344322345e-06,
|
| 1113 |
+
"loss": 0.0777,
|
| 1114 |
+
"step": 3275
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 18.0,
|
| 1118 |
+
"eval_accuracy": 0.852882703777336,
|
| 1119 |
+
"eval_loss": 1.0913971662521362,
|
| 1120 |
+
"eval_macro_f1": 0.8523928553715788,
|
| 1121 |
+
"eval_runtime": 0.9344,
|
| 1122 |
+
"eval_samples_per_second": 538.288,
|
| 1123 |
+
"eval_steps_per_second": 34.245,
|
| 1124 |
+
"eval_weighted_f1": 0.8526433677992411,
|
| 1125 |
+
"step": 3276
|
| 1126 |
+
}
|
| 1127 |
+
],
|
| 1128 |
+
"logging_steps": 25,
|
| 1129 |
+
"max_steps": 3640,
|
| 1130 |
+
"num_input_tokens_seen": 0,
|
| 1131 |
+
"num_train_epochs": 20,
|
| 1132 |
+
"save_steps": 500,
|
| 1133 |
+
"stateful_callbacks": {
|
| 1134 |
+
"EarlyStoppingCallback": {
|
| 1135 |
+
"args": {
|
| 1136 |
+
"early_stopping_patience": 3,
|
| 1137 |
+
"early_stopping_threshold": 0.0
|
| 1138 |
+
},
|
| 1139 |
+
"attributes": {
|
| 1140 |
+
"early_stopping_patience_counter": 3
|
| 1141 |
+
}
|
| 1142 |
+
},
|
| 1143 |
+
"TrainerControl": {
|
| 1144 |
+
"args": {
|
| 1145 |
+
"should_epoch_stop": false,
|
| 1146 |
+
"should_evaluate": false,
|
| 1147 |
+
"should_log": false,
|
| 1148 |
+
"should_save": true,
|
| 1149 |
+
"should_training_stop": true
|
| 1150 |
+
},
|
| 1151 |
+
"attributes": {}
|
| 1152 |
+
}
|
| 1153 |
+
},
|
| 1154 |
+
"total_flos": 0.0,
|
| 1155 |
+
"train_batch_size": 8,
|
| 1156 |
+
"trial_name": null,
|
| 1157 |
+
"trial_params": null
|
| 1158 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/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-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/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-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-1365/trainer_state.json
ADDED
|
@@ -0,0 +1,498 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 780,
|
| 3 |
+
"best_metric": 0.8406674936749367,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1365,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.12836970474967907,
|
| 14 |
+
"grad_norm": 9.430704116821289,
|
| 15 |
+
"learning_rate": 1.230769230769231e-06,
|
| 16 |
+
"loss": 4.3891,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.25673940949935814,
|
| 21 |
+
"grad_norm": 14.289024353027344,
|
| 22 |
+
"learning_rate": 2.512820512820513e-06,
|
| 23 |
+
"loss": 4.4178,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3851091142490372,
|
| 28 |
+
"grad_norm": 17.2978458404541,
|
| 29 |
+
"learning_rate": 3.794871794871795e-06,
|
| 30 |
+
"loss": 4.4374,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5134788189987163,
|
| 35 |
+
"grad_norm": 10.681659698486328,
|
| 36 |
+
"learning_rate": 5.076923076923077e-06,
|
| 37 |
+
"loss": 4.3835,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6418485237483954,
|
| 42 |
+
"grad_norm": 15.685321807861328,
|
| 43 |
+
"learning_rate": 6.358974358974359e-06,
|
| 44 |
+
"loss": 4.4031,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7702182284980744,
|
| 49 |
+
"grad_norm": 35.6545524597168,
|
| 50 |
+
"learning_rate": 7.641025641025641e-06,
|
| 51 |
+
"loss": 4.3319,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.8985879332477535,
|
| 56 |
+
"grad_norm": 69.51312255859375,
|
| 57 |
+
"learning_rate": 8.923076923076925e-06,
|
| 58 |
+
"loss": 4.3044,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.42560553633217996,
|
| 64 |
+
"eval_loss": 1.0555390119552612,
|
| 65 |
+
"eval_macro_f1": 0.33934758984498514,
|
| 66 |
+
"eval_runtime": 1.0741,
|
| 67 |
+
"eval_samples_per_second": 538.129,
|
| 68 |
+
"eval_steps_per_second": 34.448,
|
| 69 |
+
"eval_weighted_f1": 0.34398208228487587,
|
| 70 |
+
"step": 195
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0256739409499358,
|
| 74 |
+
"grad_norm": 22.040842056274414,
|
| 75 |
+
"learning_rate": 1.0205128205128205e-05,
|
| 76 |
+
"loss": 4.2103,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.1540436456996148,
|
| 81 |
+
"grad_norm": 1615.3883056640625,
|
| 82 |
+
"learning_rate": 1.1487179487179487e-05,
|
| 83 |
+
"loss": 4.0052,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.282413350449294,
|
| 88 |
+
"grad_norm": 50.245086669921875,
|
| 89 |
+
"learning_rate": 1.2769230769230769e-05,
|
| 90 |
+
"loss": 3.6948,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.410783055198973,
|
| 95 |
+
"grad_norm": 61.75981521606445,
|
| 96 |
+
"learning_rate": 1.405128205128205e-05,
|
| 97 |
+
"loss": 3.1955,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.539152759948652,
|
| 102 |
+
"grad_norm": 62.491371154785156,
|
| 103 |
+
"learning_rate": 1.5333333333333334e-05,
|
| 104 |
+
"loss": 2.8663,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.6675224646983313,
|
| 109 |
+
"grad_norm": 54.36432647705078,
|
| 110 |
+
"learning_rate": 1.6615384615384618e-05,
|
| 111 |
+
"loss": 2.7885,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.7958921694480101,
|
| 116 |
+
"grad_norm": 43.571815490722656,
|
| 117 |
+
"learning_rate": 1.78974358974359e-05,
|
| 118 |
+
"loss": 2.5408,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.9242618741976893,
|
| 123 |
+
"grad_norm": 73.8813705444336,
|
| 124 |
+
"learning_rate": 1.9179487179487182e-05,
|
| 125 |
+
"loss": 2.5575,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.0,
|
| 130 |
+
"eval_accuracy": 0.7508650519031141,
|
| 131 |
+
"eval_loss": 0.6272709965705872,
|
| 132 |
+
"eval_macro_f1": 0.7333974124187493,
|
| 133 |
+
"eval_runtime": 1.0563,
|
| 134 |
+
"eval_samples_per_second": 547.215,
|
| 135 |
+
"eval_steps_per_second": 35.029,
|
| 136 |
+
"eval_weighted_f1": 0.7359987627729784,
|
| 137 |
+
"step": 390
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.0513478818998716,
|
| 141 |
+
"grad_norm": 34.3968391418457,
|
| 142 |
+
"learning_rate": 1.994871794871795e-05,
|
| 143 |
+
"loss": 2.1955,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.179717586649551,
|
| 148 |
+
"grad_norm": 34.10501480102539,
|
| 149 |
+
"learning_rate": 1.9806267806267806e-05,
|
| 150 |
+
"loss": 1.8204,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.3080872913992296,
|
| 155 |
+
"grad_norm": 22.600313186645508,
|
| 156 |
+
"learning_rate": 1.9663817663817667e-05,
|
| 157 |
+
"loss": 2.2013,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.436456996148909,
|
| 162 |
+
"grad_norm": 38.87358856201172,
|
| 163 |
+
"learning_rate": 1.952136752136752e-05,
|
| 164 |
+
"loss": 2.2847,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.564826700898588,
|
| 169 |
+
"grad_norm": 41.36630630493164,
|
| 170 |
+
"learning_rate": 1.9378917378917382e-05,
|
| 171 |
+
"loss": 1.902,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.693196405648267,
|
| 176 |
+
"grad_norm": 36.666648864746094,
|
| 177 |
+
"learning_rate": 1.923646723646724e-05,
|
| 178 |
+
"loss": 1.6207,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.821566110397946,
|
| 183 |
+
"grad_norm": 43.031070709228516,
|
| 184 |
+
"learning_rate": 1.9094017094017097e-05,
|
| 185 |
+
"loss": 1.971,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.949935815147625,
|
| 190 |
+
"grad_norm": 33.17438507080078,
|
| 191 |
+
"learning_rate": 1.8951566951566954e-05,
|
| 192 |
+
"loss": 2.0082,
|
| 193 |
+
"step": 575
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 3.0,
|
| 197 |
+
"eval_accuracy": 0.8096885813148789,
|
| 198 |
+
"eval_loss": 0.5165865421295166,
|
| 199 |
+
"eval_macro_f1": 0.8093737679781864,
|
| 200 |
+
"eval_runtime": 1.0221,
|
| 201 |
+
"eval_samples_per_second": 565.525,
|
| 202 |
+
"eval_steps_per_second": 36.201,
|
| 203 |
+
"eval_weighted_f1": 0.8095560983702543,
|
| 204 |
+
"step": 585
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.0770218228498076,
|
| 208 |
+
"grad_norm": 58.33926773071289,
|
| 209 |
+
"learning_rate": 1.880911680911681e-05,
|
| 210 |
+
"loss": 1.6024,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.2053915275994864,
|
| 215 |
+
"grad_norm": 38.67941665649414,
|
| 216 |
+
"learning_rate": 1.866666666666667e-05,
|
| 217 |
+
"loss": 1.2744,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.3337612323491657,
|
| 222 |
+
"grad_norm": 45.609169006347656,
|
| 223 |
+
"learning_rate": 1.8524216524216526e-05,
|
| 224 |
+
"loss": 1.368,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.462130937098845,
|
| 229 |
+
"grad_norm": 51.29036331176758,
|
| 230 |
+
"learning_rate": 1.8381766381766383e-05,
|
| 231 |
+
"loss": 1.377,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.5905006418485237,
|
| 236 |
+
"grad_norm": 40.226436614990234,
|
| 237 |
+
"learning_rate": 1.823931623931624e-05,
|
| 238 |
+
"loss": 1.2487,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.718870346598203,
|
| 243 |
+
"grad_norm": 53.55995559692383,
|
| 244 |
+
"learning_rate": 1.8096866096866098e-05,
|
| 245 |
+
"loss": 1.2255,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.8472400513478817,
|
| 250 |
+
"grad_norm": 64.41142272949219,
|
| 251 |
+
"learning_rate": 1.7954415954415955e-05,
|
| 252 |
+
"loss": 1.1941,
|
| 253 |
+
"step": 750
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.975609756097561,
|
| 257 |
+
"grad_norm": 46.73796081542969,
|
| 258 |
+
"learning_rate": 1.7811965811965813e-05,
|
| 259 |
+
"loss": 1.1488,
|
| 260 |
+
"step": 775
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 4.0,
|
| 264 |
+
"eval_accuracy": 0.8408304498269896,
|
| 265 |
+
"eval_loss": 0.46307942271232605,
|
| 266 |
+
"eval_macro_f1": 0.8406674936749367,
|
| 267 |
+
"eval_runtime": 1.0711,
|
| 268 |
+
"eval_samples_per_second": 539.643,
|
| 269 |
+
"eval_steps_per_second": 34.545,
|
| 270 |
+
"eval_weighted_f1": 0.840722341787341,
|
| 271 |
+
"step": 780
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.102695763799743,
|
| 275 |
+
"grad_norm": 81.4758071899414,
|
| 276 |
+
"learning_rate": 1.766951566951567e-05,
|
| 277 |
+
"loss": 0.8744,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.2310654685494224,
|
| 282 |
+
"grad_norm": 48.67966842651367,
|
| 283 |
+
"learning_rate": 1.7527065527065528e-05,
|
| 284 |
+
"loss": 0.9179,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.359435173299102,
|
| 289 |
+
"grad_norm": 66.86219787597656,
|
| 290 |
+
"learning_rate": 1.7384615384615385e-05,
|
| 291 |
+
"loss": 0.6709,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.487804878048781,
|
| 296 |
+
"grad_norm": 47.42359161376953,
|
| 297 |
+
"learning_rate": 1.7242165242165242e-05,
|
| 298 |
+
"loss": 0.7896,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.616174582798459,
|
| 303 |
+
"grad_norm": 23.642698287963867,
|
| 304 |
+
"learning_rate": 1.7099715099715103e-05,
|
| 305 |
+
"loss": 0.8075,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.7445442875481385,
|
| 310 |
+
"grad_norm": 74.13261413574219,
|
| 311 |
+
"learning_rate": 1.6957264957264957e-05,
|
| 312 |
+
"loss": 1.0293,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 4.872913992297818,
|
| 317 |
+
"grad_norm": 47.64870834350586,
|
| 318 |
+
"learning_rate": 1.6814814814814818e-05,
|
| 319 |
+
"loss": 0.9265,
|
| 320 |
+
"step": 950
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 5.0,
|
| 324 |
+
"grad_norm": 45.95494842529297,
|
| 325 |
+
"learning_rate": 1.6672364672364675e-05,
|
| 326 |
+
"loss": 1.0872,
|
| 327 |
+
"step": 975
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.0,
|
| 331 |
+
"eval_accuracy": 0.8339100346020761,
|
| 332 |
+
"eval_loss": 0.5500914454460144,
|
| 333 |
+
"eval_macro_f1": 0.8348154951148782,
|
| 334 |
+
"eval_runtime": 1.0471,
|
| 335 |
+
"eval_samples_per_second": 552.017,
|
| 336 |
+
"eval_steps_per_second": 35.337,
|
| 337 |
+
"eval_weighted_f1": 0.8351279576978542,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.128369704749679,
|
| 342 |
+
"grad_norm": 14.470373153686523,
|
| 343 |
+
"learning_rate": 1.652991452991453e-05,
|
| 344 |
+
"loss": 0.4608,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.2567394094993585,
|
| 349 |
+
"grad_norm": 78.5843276977539,
|
| 350 |
+
"learning_rate": 1.638746438746439e-05,
|
| 351 |
+
"loss": 0.8012,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.385109114249037,
|
| 356 |
+
"grad_norm": 99.31294250488281,
|
| 357 |
+
"learning_rate": 1.6245014245014247e-05,
|
| 358 |
+
"loss": 0.6991,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.513478818998716,
|
| 363 |
+
"grad_norm": 81.35556030273438,
|
| 364 |
+
"learning_rate": 1.6102564102564105e-05,
|
| 365 |
+
"loss": 0.756,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.641848523748395,
|
| 370 |
+
"grad_norm": 18.775577545166016,
|
| 371 |
+
"learning_rate": 1.5960113960113962e-05,
|
| 372 |
+
"loss": 0.6547,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.7702182284980745,
|
| 377 |
+
"grad_norm": 63.5207633972168,
|
| 378 |
+
"learning_rate": 1.581766381766382e-05,
|
| 379 |
+
"loss": 0.6791,
|
| 380 |
+
"step": 1125
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 5.898587933247754,
|
| 384 |
+
"grad_norm": 52.03333282470703,
|
| 385 |
+
"learning_rate": 1.5675213675213677e-05,
|
| 386 |
+
"loss": 0.4433,
|
| 387 |
+
"step": 1150
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.0,
|
| 391 |
+
"eval_accuracy": 0.8269896193771626,
|
| 392 |
+
"eval_loss": 0.741005003452301,
|
| 393 |
+
"eval_macro_f1": 0.8225979713474927,
|
| 394 |
+
"eval_runtime": 1.0073,
|
| 395 |
+
"eval_samples_per_second": 573.825,
|
| 396 |
+
"eval_steps_per_second": 36.733,
|
| 397 |
+
"eval_weighted_f1": 0.8234008580238501,
|
| 398 |
+
"step": 1170
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.025673940949936,
|
| 402 |
+
"grad_norm": 42.387657165527344,
|
| 403 |
+
"learning_rate": 1.5532763532763534e-05,
|
| 404 |
+
"loss": 0.5889,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.154043645699615,
|
| 409 |
+
"grad_norm": 30.781919479370117,
|
| 410 |
+
"learning_rate": 1.539031339031339e-05,
|
| 411 |
+
"loss": 0.4634,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.282413350449294,
|
| 416 |
+
"grad_norm": 64.297607421875,
|
| 417 |
+
"learning_rate": 1.5247863247863249e-05,
|
| 418 |
+
"loss": 0.3153,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.410783055198973,
|
| 423 |
+
"grad_norm": 87.80364990234375,
|
| 424 |
+
"learning_rate": 1.5105413105413106e-05,
|
| 425 |
+
"loss": 0.5873,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.539152759948652,
|
| 430 |
+
"grad_norm": 120.49800872802734,
|
| 431 |
+
"learning_rate": 1.4962962962962964e-05,
|
| 432 |
+
"loss": 0.8297,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.667522464698331,
|
| 437 |
+
"grad_norm": 58.5731315612793,
|
| 438 |
+
"learning_rate": 1.4820512820512823e-05,
|
| 439 |
+
"loss": 0.4958,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 6.7958921694480106,
|
| 444 |
+
"grad_norm": 52.797672271728516,
|
| 445 |
+
"learning_rate": 1.4678062678062678e-05,
|
| 446 |
+
"loss": 0.4947,
|
| 447 |
+
"step": 1325
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 6.92426187419769,
|
| 451 |
+
"grad_norm": 92.18386840820312,
|
| 452 |
+
"learning_rate": 1.4535612535612537e-05,
|
| 453 |
+
"loss": 0.3268,
|
| 454 |
+
"step": 1350
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 7.0,
|
| 458 |
+
"eval_accuracy": 0.8373702422145328,
|
| 459 |
+
"eval_loss": 0.708697497844696,
|
| 460 |
+
"eval_macro_f1": 0.835908544996295,
|
| 461 |
+
"eval_runtime": 1.0431,
|
| 462 |
+
"eval_samples_per_second": 554.112,
|
| 463 |
+
"eval_steps_per_second": 35.471,
|
| 464 |
+
"eval_weighted_f1": 0.836267936226824,
|
| 465 |
+
"step": 1365
|
| 466 |
+
}
|
| 467 |
+
],
|
| 468 |
+
"logging_steps": 25,
|
| 469 |
+
"max_steps": 3900,
|
| 470 |
+
"num_input_tokens_seen": 0,
|
| 471 |
+
"num_train_epochs": 20,
|
| 472 |
+
"save_steps": 500,
|
| 473 |
+
"stateful_callbacks": {
|
| 474 |
+
"EarlyStoppingCallback": {
|
| 475 |
+
"args": {
|
| 476 |
+
"early_stopping_patience": 3,
|
| 477 |
+
"early_stopping_threshold": 0.0
|
| 478 |
+
},
|
| 479 |
+
"attributes": {
|
| 480 |
+
"early_stopping_patience_counter": 3
|
| 481 |
+
}
|
| 482 |
+
},
|
| 483 |
+
"TrainerControl": {
|
| 484 |
+
"args": {
|
| 485 |
+
"should_epoch_stop": false,
|
| 486 |
+
"should_evaluate": false,
|
| 487 |
+
"should_log": false,
|
| 488 |
+
"should_save": true,
|
| 489 |
+
"should_training_stop": true
|
| 490 |
+
},
|
| 491 |
+
"attributes": {}
|
| 492 |
+
}
|
| 493 |
+
},
|
| 494 |
+
"total_flos": 0.0,
|
| 495 |
+
"train_batch_size": 8,
|
| 496 |
+
"trial_name": null,
|
| 497 |
+
"trial_params": null
|
| 498 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/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-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/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-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780/trainer_state.json
ADDED
|
@@ -0,0 +1,304 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 780,
|
| 3 |
+
"best_metric": 0.8406674936749367,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-03-s44-20260828T203718Z-496d6871/trainer/checkpoint-780",
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 780,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.12836970474967907,
|
| 14 |
+
"grad_norm": 9.430704116821289,
|
| 15 |
+
"learning_rate": 1.230769230769231e-06,
|
| 16 |
+
"loss": 4.3891,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.25673940949935814,
|
| 21 |
+
"grad_norm": 14.289024353027344,
|
| 22 |
+
"learning_rate": 2.512820512820513e-06,
|
| 23 |
+
"loss": 4.4178,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3851091142490372,
|
| 28 |
+
"grad_norm": 17.2978458404541,
|
| 29 |
+
"learning_rate": 3.794871794871795e-06,
|
| 30 |
+
"loss": 4.4374,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5134788189987163,
|
| 35 |
+
"grad_norm": 10.681659698486328,
|
| 36 |
+
"learning_rate": 5.076923076923077e-06,
|
| 37 |
+
"loss": 4.3835,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6418485237483954,
|
| 42 |
+
"grad_norm": 15.685321807861328,
|
| 43 |
+
"learning_rate": 6.358974358974359e-06,
|
| 44 |
+
"loss": 4.4031,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7702182284980744,
|
| 49 |
+
"grad_norm": 35.6545524597168,
|
| 50 |
+
"learning_rate": 7.641025641025641e-06,
|
| 51 |
+
"loss": 4.3319,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.8985879332477535,
|
| 56 |
+
"grad_norm": 69.51312255859375,
|
| 57 |
+
"learning_rate": 8.923076923076925e-06,
|
| 58 |
+
"loss": 4.3044,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.42560553633217996,
|
| 64 |
+
"eval_loss": 1.0555390119552612,
|
| 65 |
+
"eval_macro_f1": 0.33934758984498514,
|
| 66 |
+
"eval_runtime": 1.0741,
|
| 67 |
+
"eval_samples_per_second": 538.129,
|
| 68 |
+
"eval_steps_per_second": 34.448,
|
| 69 |
+
"eval_weighted_f1": 0.34398208228487587,
|
| 70 |
+
"step": 195
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0256739409499358,
|
| 74 |
+
"grad_norm": 22.040842056274414,
|
| 75 |
+
"learning_rate": 1.0205128205128205e-05,
|
| 76 |
+
"loss": 4.2103,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.1540436456996148,
|
| 81 |
+
"grad_norm": 1615.3883056640625,
|
| 82 |
+
"learning_rate": 1.1487179487179487e-05,
|
| 83 |
+
"loss": 4.0052,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.282413350449294,
|
| 88 |
+
"grad_norm": 50.245086669921875,
|
| 89 |
+
"learning_rate": 1.2769230769230769e-05,
|
| 90 |
+
"loss": 3.6948,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.410783055198973,
|
| 95 |
+
"grad_norm": 61.75981521606445,
|
| 96 |
+
"learning_rate": 1.405128205128205e-05,
|
| 97 |
+
"loss": 3.1955,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.539152759948652,
|
| 102 |
+
"grad_norm": 62.491371154785156,
|
| 103 |
+
"learning_rate": 1.5333333333333334e-05,
|
| 104 |
+
"loss": 2.8663,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.6675224646983313,
|
| 109 |
+
"grad_norm": 54.36432647705078,
|
| 110 |
+
"learning_rate": 1.6615384615384618e-05,
|
| 111 |
+
"loss": 2.7885,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.7958921694480101,
|
| 116 |
+
"grad_norm": 43.571815490722656,
|
| 117 |
+
"learning_rate": 1.78974358974359e-05,
|
| 118 |
+
"loss": 2.5408,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.9242618741976893,
|
| 123 |
+
"grad_norm": 73.8813705444336,
|
| 124 |
+
"learning_rate": 1.9179487179487182e-05,
|
| 125 |
+
"loss": 2.5575,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.0,
|
| 130 |
+
"eval_accuracy": 0.7508650519031141,
|
| 131 |
+
"eval_loss": 0.6272709965705872,
|
| 132 |
+
"eval_macro_f1": 0.7333974124187493,
|
| 133 |
+
"eval_runtime": 1.0563,
|
| 134 |
+
"eval_samples_per_second": 547.215,
|
| 135 |
+
"eval_steps_per_second": 35.029,
|
| 136 |
+
"eval_weighted_f1": 0.7359987627729784,
|
| 137 |
+
"step": 390
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.0513478818998716,
|
| 141 |
+
"grad_norm": 34.3968391418457,
|
| 142 |
+
"learning_rate": 1.994871794871795e-05,
|
| 143 |
+
"loss": 2.1955,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.179717586649551,
|
| 148 |
+
"grad_norm": 34.10501480102539,
|
| 149 |
+
"learning_rate": 1.9806267806267806e-05,
|
| 150 |
+
"loss": 1.8204,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.3080872913992296,
|
| 155 |
+
"grad_norm": 22.600313186645508,
|
| 156 |
+
"learning_rate": 1.9663817663817667e-05,
|
| 157 |
+
"loss": 2.2013,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.436456996148909,
|
| 162 |
+
"grad_norm": 38.87358856201172,
|
| 163 |
+
"learning_rate": 1.952136752136752e-05,
|
| 164 |
+
"loss": 2.2847,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.564826700898588,
|
| 169 |
+
"grad_norm": 41.36630630493164,
|
| 170 |
+
"learning_rate": 1.9378917378917382e-05,
|
| 171 |
+
"loss": 1.902,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.693196405648267,
|
| 176 |
+
"grad_norm": 36.666648864746094,
|
| 177 |
+
"learning_rate": 1.923646723646724e-05,
|
| 178 |
+
"loss": 1.6207,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.821566110397946,
|
| 183 |
+
"grad_norm": 43.031070709228516,
|
| 184 |
+
"learning_rate": 1.9094017094017097e-05,
|
| 185 |
+
"loss": 1.971,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.949935815147625,
|
| 190 |
+
"grad_norm": 33.17438507080078,
|
| 191 |
+
"learning_rate": 1.8951566951566954e-05,
|
| 192 |
+
"loss": 2.0082,
|
| 193 |
+
"step": 575
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 3.0,
|
| 197 |
+
"eval_accuracy": 0.8096885813148789,
|
| 198 |
+
"eval_loss": 0.5165865421295166,
|
| 199 |
+
"eval_macro_f1": 0.8093737679781864,
|
| 200 |
+
"eval_runtime": 1.0221,
|
| 201 |
+
"eval_samples_per_second": 565.525,
|
| 202 |
+
"eval_steps_per_second": 36.201,
|
| 203 |
+
"eval_weighted_f1": 0.8095560983702543,
|
| 204 |
+
"step": 585
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.0770218228498076,
|
| 208 |
+
"grad_norm": 58.33926773071289,
|
| 209 |
+
"learning_rate": 1.880911680911681e-05,
|
| 210 |
+
"loss": 1.6024,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.2053915275994864,
|
| 215 |
+
"grad_norm": 38.67941665649414,
|
| 216 |
+
"learning_rate": 1.866666666666667e-05,
|
| 217 |
+
"loss": 1.2744,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.3337612323491657,
|
| 222 |
+
"grad_norm": 45.609169006347656,
|
| 223 |
+
"learning_rate": 1.8524216524216526e-05,
|
| 224 |
+
"loss": 1.368,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.462130937098845,
|
| 229 |
+
"grad_norm": 51.29036331176758,
|
| 230 |
+
"learning_rate": 1.8381766381766383e-05,
|
| 231 |
+
"loss": 1.377,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.5905006418485237,
|
| 236 |
+
"grad_norm": 40.226436614990234,
|
| 237 |
+
"learning_rate": 1.823931623931624e-05,
|
| 238 |
+
"loss": 1.2487,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.718870346598203,
|
| 243 |
+
"grad_norm": 53.55995559692383,
|
| 244 |
+
"learning_rate": 1.8096866096866098e-05,
|
| 245 |
+
"loss": 1.2255,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.8472400513478817,
|
| 250 |
+
"grad_norm": 64.41142272949219,
|
| 251 |
+
"learning_rate": 1.7954415954415955e-05,
|
| 252 |
+
"loss": 1.1941,
|
| 253 |
+
"step": 750
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.975609756097561,
|
| 257 |
+
"grad_norm": 46.73796081542969,
|
| 258 |
+
"learning_rate": 1.7811965811965813e-05,
|
| 259 |
+
"loss": 1.1488,
|
| 260 |
+
"step": 775
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 4.0,
|
| 264 |
+
"eval_accuracy": 0.8408304498269896,
|
| 265 |
+
"eval_loss": 0.46307942271232605,
|
| 266 |
+
"eval_macro_f1": 0.8406674936749367,
|
| 267 |
+
"eval_runtime": 1.0711,
|
| 268 |
+
"eval_samples_per_second": 539.643,
|
| 269 |
+
"eval_steps_per_second": 34.545,
|
| 270 |
+
"eval_weighted_f1": 0.840722341787341,
|
| 271 |
+
"step": 780
|
| 272 |
+
}
|
| 273 |
+
],
|
| 274 |
+
"logging_steps": 25,
|
| 275 |
+
"max_steps": 3900,
|
| 276 |
+
"num_input_tokens_seen": 0,
|
| 277 |
+
"num_train_epochs": 20,
|
| 278 |
+
"save_steps": 500,
|
| 279 |
+
"stateful_callbacks": {
|
| 280 |
+
"EarlyStoppingCallback": {
|
| 281 |
+
"args": {
|
| 282 |
+
"early_stopping_patience": 3,
|
| 283 |
+
"early_stopping_threshold": 0.0
|
| 284 |
+
},
|
| 285 |
+
"attributes": {
|
| 286 |
+
"early_stopping_patience_counter": 0
|
| 287 |
+
}
|
| 288 |
+
},
|
| 289 |
+
"TrainerControl": {
|
| 290 |
+
"args": {
|
| 291 |
+
"should_epoch_stop": false,
|
| 292 |
+
"should_evaluate": false,
|
| 293 |
+
"should_log": false,
|
| 294 |
+
"should_save": true,
|
| 295 |
+
"should_training_stop": false
|
| 296 |
+
},
|
| 297 |
+
"attributes": {}
|
| 298 |
+
}
|
| 299 |
+
},
|
| 300 |
+
"total_flos": 0.0,
|
| 301 |
+
"train_batch_size": 8,
|
| 302 |
+
"trial_name": null,
|
| 303 |
+
"trial_params": null
|
| 304 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/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-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/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-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880/trainer_state.json
ADDED
|
@@ -0,0 +1,678 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1880,
|
| 3 |
+
"best_metric": 0.8325364371612837,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1880,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13351134846461948,
|
| 14 |
+
"grad_norm": 20.673627853393555,
|
| 15 |
+
"learning_rate": 1.276595744680851e-06,
|
| 16 |
+
"loss": 4.4179,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.26702269692923897,
|
| 21 |
+
"grad_norm": 6.811794281005859,
|
| 22 |
+
"learning_rate": 2.6063829787234047e-06,
|
| 23 |
+
"loss": 4.4111,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.40053404539385845,
|
| 28 |
+
"grad_norm": 14.931140899658203,
|
| 29 |
+
"learning_rate": 3.936170212765958e-06,
|
| 30 |
+
"loss": 4.4042,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5340453938584779,
|
| 35 |
+
"grad_norm": 7.403614044189453,
|
| 36 |
+
"learning_rate": 5.265957446808511e-06,
|
| 37 |
+
"loss": 4.4075,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6675567423230975,
|
| 42 |
+
"grad_norm": 12.091256141662598,
|
| 43 |
+
"learning_rate": 6.595744680851064e-06,
|
| 44 |
+
"loss": 4.3679,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8010680907877169,
|
| 49 |
+
"grad_norm": 10.019458770751953,
|
| 50 |
+
"learning_rate": 7.925531914893617e-06,
|
| 51 |
+
"loss": 4.3565,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9345794392523364,
|
| 56 |
+
"grad_norm": 14.457362174987793,
|
| 57 |
+
"learning_rate": 9.255319148936171e-06,
|
| 58 |
+
"loss": 4.3684,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.4441687344913151,
|
| 64 |
+
"eval_loss": 1.0651849508285522,
|
| 65 |
+
"eval_macro_f1": 0.3753786024955524,
|
| 66 |
+
"eval_runtime": 1.3533,
|
| 67 |
+
"eval_samples_per_second": 595.59,
|
| 68 |
+
"eval_steps_per_second": 37.686,
|
| 69 |
+
"eval_weighted_f1": 0.37753314368193,
|
| 70 |
+
"step": 188
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0640854472630175,
|
| 74 |
+
"grad_norm": 50.25746154785156,
|
| 75 |
+
"learning_rate": 1.0585106382978725e-05,
|
| 76 |
+
"loss": 4.0331,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.1975967957276368,
|
| 81 |
+
"grad_norm": 28.132944107055664,
|
| 82 |
+
"learning_rate": 1.1914893617021277e-05,
|
| 83 |
+
"loss": 3.9971,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3311081441922563,
|
| 88 |
+
"grad_norm": 26.1246337890625,
|
| 89 |
+
"learning_rate": 1.3244680851063832e-05,
|
| 90 |
+
"loss": 3.6115,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4646194926568759,
|
| 95 |
+
"grad_norm": 35.02851867675781,
|
| 96 |
+
"learning_rate": 1.4574468085106384e-05,
|
| 97 |
+
"loss": 3.2223,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.5981308411214954,
|
| 102 |
+
"grad_norm": 63.445247650146484,
|
| 103 |
+
"learning_rate": 1.590425531914894e-05,
|
| 104 |
+
"loss": 2.9527,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7316421895861147,
|
| 109 |
+
"grad_norm": 35.380924224853516,
|
| 110 |
+
"learning_rate": 1.723404255319149e-05,
|
| 111 |
+
"loss": 2.8674,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.8651535380507345,
|
| 116 |
+
"grad_norm": 64.34720611572266,
|
| 117 |
+
"learning_rate": 1.8563829787234043e-05,
|
| 118 |
+
"loss": 2.4821,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.9986648865153538,
|
| 123 |
+
"grad_norm": 48.913997650146484,
|
| 124 |
+
"learning_rate": 1.9893617021276595e-05,
|
| 125 |
+
"loss": 2.7957,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.0,
|
| 130 |
+
"eval_accuracy": 0.7419354838709677,
|
| 131 |
+
"eval_loss": 0.6129544377326965,
|
| 132 |
+
"eval_macro_f1": 0.7391019638003171,
|
| 133 |
+
"eval_runtime": 1.4116,
|
| 134 |
+
"eval_samples_per_second": 570.974,
|
| 135 |
+
"eval_steps_per_second": 36.129,
|
| 136 |
+
"eval_weighted_f1": 0.7390724544145634,
|
| 137 |
+
"step": 376
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.128170894526035,
|
| 141 |
+
"grad_norm": 39.42140197753906,
|
| 142 |
+
"learning_rate": 1.986406619385343e-05,
|
| 143 |
+
"loss": 2.1604,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.2616822429906542,
|
| 148 |
+
"grad_norm": 156.8961944580078,
|
| 149 |
+
"learning_rate": 1.971631205673759e-05,
|
| 150 |
+
"loss": 1.9432,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.3951935914552736,
|
| 155 |
+
"grad_norm": 44.37391662597656,
|
| 156 |
+
"learning_rate": 1.956855791962175e-05,
|
| 157 |
+
"loss": 2.2598,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.5287049399198933,
|
| 162 |
+
"grad_norm": 27.17721939086914,
|
| 163 |
+
"learning_rate": 1.9420803782505913e-05,
|
| 164 |
+
"loss": 1.9138,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.6622162883845126,
|
| 169 |
+
"grad_norm": 54.70370864868164,
|
| 170 |
+
"learning_rate": 1.927304964539007e-05,
|
| 171 |
+
"loss": 2.0825,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.795727636849132,
|
| 176 |
+
"grad_norm": 40.60383224487305,
|
| 177 |
+
"learning_rate": 1.9125295508274234e-05,
|
| 178 |
+
"loss": 1.9968,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.9292389853137517,
|
| 183 |
+
"grad_norm": 68.17862701416016,
|
| 184 |
+
"learning_rate": 1.8977541371158394e-05,
|
| 185 |
+
"loss": 2.1597,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.0,
|
| 190 |
+
"eval_accuracy": 0.8002481389578163,
|
| 191 |
+
"eval_loss": 0.48578470945358276,
|
| 192 |
+
"eval_macro_f1": 0.797633914123045,
|
| 193 |
+
"eval_runtime": 1.3433,
|
| 194 |
+
"eval_samples_per_second": 600.001,
|
| 195 |
+
"eval_steps_per_second": 37.965,
|
| 196 |
+
"eval_weighted_f1": 0.7976918008957689,
|
| 197 |
+
"step": 564
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0587449933244324,
|
| 201 |
+
"grad_norm": 66.7054214477539,
|
| 202 |
+
"learning_rate": 1.8829787234042557e-05,
|
| 203 |
+
"loss": 1.8278,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.192256341789052,
|
| 208 |
+
"grad_norm": 79.62871551513672,
|
| 209 |
+
"learning_rate": 1.8682033096926714e-05,
|
| 210 |
+
"loss": 1.2467,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.3257676902536715,
|
| 215 |
+
"grad_norm": 44.97430419921875,
|
| 216 |
+
"learning_rate": 1.8534278959810878e-05,
|
| 217 |
+
"loss": 1.3405,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4592790387182912,
|
| 222 |
+
"grad_norm": 42.74566650390625,
|
| 223 |
+
"learning_rate": 1.8386524822695038e-05,
|
| 224 |
+
"loss": 1.6122,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.5927903871829105,
|
| 229 |
+
"grad_norm": 21.341346740722656,
|
| 230 |
+
"learning_rate": 1.8238770685579198e-05,
|
| 231 |
+
"loss": 1.6088,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.72630173564753,
|
| 236 |
+
"grad_norm": 41.35243606567383,
|
| 237 |
+
"learning_rate": 1.8091016548463358e-05,
|
| 238 |
+
"loss": 1.3344,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.8598130841121496,
|
| 243 |
+
"grad_norm": 87.04261016845703,
|
| 244 |
+
"learning_rate": 1.7943262411347518e-05,
|
| 245 |
+
"loss": 1.3879,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.993324432576769,
|
| 250 |
+
"grad_norm": 35.48392105102539,
|
| 251 |
+
"learning_rate": 1.779550827423168e-05,
|
| 252 |
+
"loss": 1.3266,
|
| 253 |
+
"step": 750
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 4.0,
|
| 257 |
+
"eval_accuracy": 0.8114143920595533,
|
| 258 |
+
"eval_loss": 0.5030881762504578,
|
| 259 |
+
"eval_macro_f1": 0.8106581838763844,
|
| 260 |
+
"eval_runtime": 1.3801,
|
| 261 |
+
"eval_samples_per_second": 584.003,
|
| 262 |
+
"eval_steps_per_second": 36.953,
|
| 263 |
+
"eval_weighted_f1": 0.810642469392529,
|
| 264 |
+
"step": 752
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.12283044058745,
|
| 268 |
+
"grad_norm": 22.939712524414062,
|
| 269 |
+
"learning_rate": 1.7647754137115842e-05,
|
| 270 |
+
"loss": 0.8932,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.25634178905207,
|
| 275 |
+
"grad_norm": 45.860713958740234,
|
| 276 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 277 |
+
"loss": 1.0112,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.389853137516689,
|
| 282 |
+
"grad_norm": 77.18540954589844,
|
| 283 |
+
"learning_rate": 1.7352245862884162e-05,
|
| 284 |
+
"loss": 1.06,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.5233644859813085,
|
| 289 |
+
"grad_norm": 34.39630889892578,
|
| 290 |
+
"learning_rate": 1.7204491725768322e-05,
|
| 291 |
+
"loss": 0.9128,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.656875834445928,
|
| 296 |
+
"grad_norm": 54.38111877441406,
|
| 297 |
+
"learning_rate": 1.7056737588652482e-05,
|
| 298 |
+
"loss": 0.7028,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.790387182910547,
|
| 303 |
+
"grad_norm": 86.46287536621094,
|
| 304 |
+
"learning_rate": 1.6908983451536646e-05,
|
| 305 |
+
"loss": 1.0283,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.923898531375167,
|
| 310 |
+
"grad_norm": 39.945159912109375,
|
| 311 |
+
"learning_rate": 1.6761229314420806e-05,
|
| 312 |
+
"loss": 0.9587,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.0,
|
| 317 |
+
"eval_accuracy": 0.815136476426799,
|
| 318 |
+
"eval_loss": 0.5337904691696167,
|
| 319 |
+
"eval_macro_f1": 0.8152641683112084,
|
| 320 |
+
"eval_runtime": 1.3407,
|
| 321 |
+
"eval_samples_per_second": 601.193,
|
| 322 |
+
"eval_steps_per_second": 38.041,
|
| 323 |
+
"eval_weighted_f1": 0.8152093741213681,
|
| 324 |
+
"step": 940
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.0534045393858475,
|
| 328 |
+
"grad_norm": 58.76649475097656,
|
| 329 |
+
"learning_rate": 1.6613475177304966e-05,
|
| 330 |
+
"loss": 0.9582,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.186915887850467,
|
| 335 |
+
"grad_norm": 17.8238525390625,
|
| 336 |
+
"learning_rate": 1.6465721040189126e-05,
|
| 337 |
+
"loss": 0.6535,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.320427236315087,
|
| 342 |
+
"grad_norm": 33.73042678833008,
|
| 343 |
+
"learning_rate": 1.6317966903073286e-05,
|
| 344 |
+
"loss": 0.4935,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.453938584779706,
|
| 349 |
+
"grad_norm": 25.605831146240234,
|
| 350 |
+
"learning_rate": 1.6170212765957446e-05,
|
| 351 |
+
"loss": 0.669,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.587449933244326,
|
| 356 |
+
"grad_norm": 36.81806564331055,
|
| 357 |
+
"learning_rate": 1.602245862884161e-05,
|
| 358 |
+
"loss": 0.7954,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.720961281708945,
|
| 363 |
+
"grad_norm": 62.5572509765625,
|
| 364 |
+
"learning_rate": 1.587470449172577e-05,
|
| 365 |
+
"loss": 0.5398,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.854472630173564,
|
| 370 |
+
"grad_norm": 37.748619079589844,
|
| 371 |
+
"learning_rate": 1.572695035460993e-05,
|
| 372 |
+
"loss": 0.6565,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.9879839786381845,
|
| 377 |
+
"grad_norm": 64.0909423828125,
|
| 378 |
+
"learning_rate": 1.557919621749409e-05,
|
| 379 |
+
"loss": 0.7397,
|
| 380 |
+
"step": 1125
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.0,
|
| 384 |
+
"eval_accuracy": 0.7990074441687345,
|
| 385 |
+
"eval_loss": 0.6556637287139893,
|
| 386 |
+
"eval_macro_f1": 0.7999806740506581,
|
| 387 |
+
"eval_runtime": 1.3755,
|
| 388 |
+
"eval_samples_per_second": 585.968,
|
| 389 |
+
"eval_steps_per_second": 37.077,
|
| 390 |
+
"eval_weighted_f1": 0.7995959163150099,
|
| 391 |
+
"step": 1128
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.117489986648865,
|
| 395 |
+
"grad_norm": 30.407459259033203,
|
| 396 |
+
"learning_rate": 1.543144208037825e-05,
|
| 397 |
+
"loss": 0.3339,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.251001335113485,
|
| 402 |
+
"grad_norm": 5.074549674987793,
|
| 403 |
+
"learning_rate": 1.5283687943262414e-05,
|
| 404 |
+
"loss": 0.5162,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.384512683578104,
|
| 409 |
+
"grad_norm": 3.213144302368164,
|
| 410 |
+
"learning_rate": 1.5135933806146574e-05,
|
| 411 |
+
"loss": 0.5794,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.518024032042724,
|
| 416 |
+
"grad_norm": 148.7412872314453,
|
| 417 |
+
"learning_rate": 1.4988179669030734e-05,
|
| 418 |
+
"loss": 0.8456,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.651535380507343,
|
| 423 |
+
"grad_norm": 78.37199401855469,
|
| 424 |
+
"learning_rate": 1.4840425531914894e-05,
|
| 425 |
+
"loss": 0.4425,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.785046728971962,
|
| 430 |
+
"grad_norm": 31.29233741760254,
|
| 431 |
+
"learning_rate": 1.4692671394799056e-05,
|
| 432 |
+
"loss": 0.5563,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.9185580774365825,
|
| 437 |
+
"grad_norm": 69.88211059570312,
|
| 438 |
+
"learning_rate": 1.4544917257683215e-05,
|
| 439 |
+
"loss": 0.4397,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 7.0,
|
| 444 |
+
"eval_accuracy": 0.8176178660049628,
|
| 445 |
+
"eval_loss": 0.8486976623535156,
|
| 446 |
+
"eval_macro_f1": 0.8166522074416811,
|
| 447 |
+
"eval_runtime": 1.3509,
|
| 448 |
+
"eval_samples_per_second": 596.62,
|
| 449 |
+
"eval_steps_per_second": 37.751,
|
| 450 |
+
"eval_weighted_f1": 0.8166431779008507,
|
| 451 |
+
"step": 1316
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.048064085447263,
|
| 455 |
+
"grad_norm": 3.868504762649536,
|
| 456 |
+
"learning_rate": 1.4397163120567377e-05,
|
| 457 |
+
"loss": 0.2631,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.181575433911883,
|
| 462 |
+
"grad_norm": 1.3365354537963867,
|
| 463 |
+
"learning_rate": 1.4249408983451538e-05,
|
| 464 |
+
"loss": 0.3103,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.315086782376502,
|
| 469 |
+
"grad_norm": 123.04847717285156,
|
| 470 |
+
"learning_rate": 1.41016548463357e-05,
|
| 471 |
+
"loss": 0.2794,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.4485981308411215,
|
| 476 |
+
"grad_norm": 79.79776763916016,
|
| 477 |
+
"learning_rate": 1.3953900709219859e-05,
|
| 478 |
+
"loss": 0.2952,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.582109479305741,
|
| 483 |
+
"grad_norm": 54.26392364501953,
|
| 484 |
+
"learning_rate": 1.380614657210402e-05,
|
| 485 |
+
"loss": 0.531,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.71562082777036,
|
| 490 |
+
"grad_norm": 168.4088134765625,
|
| 491 |
+
"learning_rate": 1.365839243498818e-05,
|
| 492 |
+
"loss": 0.599,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 7.8491321762349795,
|
| 497 |
+
"grad_norm": 128.00387573242188,
|
| 498 |
+
"learning_rate": 1.351063829787234e-05,
|
| 499 |
+
"loss": 0.5645,
|
| 500 |
+
"step": 1475
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 7.9826435246996,
|
| 504 |
+
"grad_norm": 8.781241416931152,
|
| 505 |
+
"learning_rate": 1.3362884160756503e-05,
|
| 506 |
+
"loss": 0.3614,
|
| 507 |
+
"step": 1500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 8.0,
|
| 511 |
+
"eval_accuracy": 0.826302729528536,
|
| 512 |
+
"eval_loss": 0.9394305348396301,
|
| 513 |
+
"eval_macro_f1": 0.8262771676591104,
|
| 514 |
+
"eval_runtime": 1.3398,
|
| 515 |
+
"eval_samples_per_second": 601.583,
|
| 516 |
+
"eval_steps_per_second": 38.065,
|
| 517 |
+
"eval_weighted_f1": 0.8261771474759725,
|
| 518 |
+
"step": 1504
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.11214953271028,
|
| 522 |
+
"grad_norm": 131.1908721923828,
|
| 523 |
+
"learning_rate": 1.3215130023640663e-05,
|
| 524 |
+
"loss": 0.4791,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.2456608811749,
|
| 529 |
+
"grad_norm": 4.451777935028076,
|
| 530 |
+
"learning_rate": 1.3067375886524825e-05,
|
| 531 |
+
"loss": 0.2627,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.379172229639519,
|
| 536 |
+
"grad_norm": 8.379265785217285,
|
| 537 |
+
"learning_rate": 1.2919621749408983e-05,
|
| 538 |
+
"loss": 0.5037,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.51268357810414,
|
| 543 |
+
"grad_norm": 254.3783416748047,
|
| 544 |
+
"learning_rate": 1.2771867612293145e-05,
|
| 545 |
+
"loss": 0.3598,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.646194926568759,
|
| 550 |
+
"grad_norm": 1.897127389907837,
|
| 551 |
+
"learning_rate": 1.2624113475177307e-05,
|
| 552 |
+
"loss": 0.2337,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 8.779706275033378,
|
| 557 |
+
"grad_norm": 5.200291633605957,
|
| 558 |
+
"learning_rate": 1.2476359338061469e-05,
|
| 559 |
+
"loss": 0.226,
|
| 560 |
+
"step": 1650
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 8.913217623497998,
|
| 564 |
+
"grad_norm": 0.13330727815628052,
|
| 565 |
+
"learning_rate": 1.2328605200945627e-05,
|
| 566 |
+
"loss": 0.2286,
|
| 567 |
+
"step": 1675
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 9.0,
|
| 571 |
+
"eval_accuracy": 0.8225806451612904,
|
| 572 |
+
"eval_loss": 1.0447702407836914,
|
| 573 |
+
"eval_macro_f1": 0.8239610502521302,
|
| 574 |
+
"eval_runtime": 1.3237,
|
| 575 |
+
"eval_samples_per_second": 608.91,
|
| 576 |
+
"eval_steps_per_second": 38.529,
|
| 577 |
+
"eval_weighted_f1": 0.8238013412706735,
|
| 578 |
+
"step": 1692
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.042723631508679,
|
| 582 |
+
"grad_norm": 1.7377159595489502,
|
| 583 |
+
"learning_rate": 1.2180851063829789e-05,
|
| 584 |
+
"loss": 0.1298,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.176234979973298,
|
| 589 |
+
"grad_norm": 0.25821366906166077,
|
| 590 |
+
"learning_rate": 1.2033096926713949e-05,
|
| 591 |
+
"loss": 0.1932,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.309746328437917,
|
| 596 |
+
"grad_norm": 126.08575439453125,
|
| 597 |
+
"learning_rate": 1.1885342789598109e-05,
|
| 598 |
+
"loss": 0.1659,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.443257676902537,
|
| 603 |
+
"grad_norm": 15.046143531799316,
|
| 604 |
+
"learning_rate": 1.1737588652482271e-05,
|
| 605 |
+
"loss": 0.2783,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.576769025367156,
|
| 610 |
+
"grad_norm": 3.01290225982666,
|
| 611 |
+
"learning_rate": 1.1589834515366431e-05,
|
| 612 |
+
"loss": 0.1438,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 9.710280373831775,
|
| 617 |
+
"grad_norm": 0.2535240650177002,
|
| 618 |
+
"learning_rate": 1.1442080378250593e-05,
|
| 619 |
+
"loss": 0.292,
|
| 620 |
+
"step": 1825
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 9.843791722296395,
|
| 624 |
+
"grad_norm": 0.08549287170171738,
|
| 625 |
+
"learning_rate": 1.1294326241134751e-05,
|
| 626 |
+
"loss": 0.2151,
|
| 627 |
+
"step": 1850
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 9.977303070761014,
|
| 631 |
+
"grad_norm": 1.3324421644210815,
|
| 632 |
+
"learning_rate": 1.1146572104018913e-05,
|
| 633 |
+
"loss": 0.2083,
|
| 634 |
+
"step": 1875
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 10.0,
|
| 638 |
+
"eval_accuracy": 0.8325062034739454,
|
| 639 |
+
"eval_loss": 1.0802310705184937,
|
| 640 |
+
"eval_macro_f1": 0.8325364371612837,
|
| 641 |
+
"eval_runtime": 1.2717,
|
| 642 |
+
"eval_samples_per_second": 633.792,
|
| 643 |
+
"eval_steps_per_second": 40.103,
|
| 644 |
+
"eval_weighted_f1": 0.8325432770935196,
|
| 645 |
+
"step": 1880
|
| 646 |
+
}
|
| 647 |
+
],
|
| 648 |
+
"logging_steps": 25,
|
| 649 |
+
"max_steps": 3760,
|
| 650 |
+
"num_input_tokens_seen": 0,
|
| 651 |
+
"num_train_epochs": 20,
|
| 652 |
+
"save_steps": 500,
|
| 653 |
+
"stateful_callbacks": {
|
| 654 |
+
"EarlyStoppingCallback": {
|
| 655 |
+
"args": {
|
| 656 |
+
"early_stopping_patience": 3,
|
| 657 |
+
"early_stopping_threshold": 0.0
|
| 658 |
+
},
|
| 659 |
+
"attributes": {
|
| 660 |
+
"early_stopping_patience_counter": 0
|
| 661 |
+
}
|
| 662 |
+
},
|
| 663 |
+
"TrainerControl": {
|
| 664 |
+
"args": {
|
| 665 |
+
"should_epoch_stop": false,
|
| 666 |
+
"should_evaluate": false,
|
| 667 |
+
"should_log": false,
|
| 668 |
+
"should_save": true,
|
| 669 |
+
"should_training_stop": false
|
| 670 |
+
},
|
| 671 |
+
"attributes": {}
|
| 672 |
+
}
|
| 673 |
+
},
|
| 674 |
+
"total_flos": 0.0,
|
| 675 |
+
"train_batch_size": 8,
|
| 676 |
+
"trial_name": null,
|
| 677 |
+
"trial_params": null
|
| 678 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/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-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/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-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-2444/trainer_state.json
ADDED
|
@@ -0,0 +1,865 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1880,
|
| 3 |
+
"best_metric": 0.8325364371612837,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-04-s44-20260828T204407Z-496d6871/trainer/checkpoint-1880",
|
| 5 |
+
"epoch": 13.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2444,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13351134846461948,
|
| 14 |
+
"grad_norm": 20.673627853393555,
|
| 15 |
+
"learning_rate": 1.276595744680851e-06,
|
| 16 |
+
"loss": 4.4179,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.26702269692923897,
|
| 21 |
+
"grad_norm": 6.811794281005859,
|
| 22 |
+
"learning_rate": 2.6063829787234047e-06,
|
| 23 |
+
"loss": 4.4111,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.40053404539385845,
|
| 28 |
+
"grad_norm": 14.931140899658203,
|
| 29 |
+
"learning_rate": 3.936170212765958e-06,
|
| 30 |
+
"loss": 4.4042,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5340453938584779,
|
| 35 |
+
"grad_norm": 7.403614044189453,
|
| 36 |
+
"learning_rate": 5.265957446808511e-06,
|
| 37 |
+
"loss": 4.4075,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6675567423230975,
|
| 42 |
+
"grad_norm": 12.091256141662598,
|
| 43 |
+
"learning_rate": 6.595744680851064e-06,
|
| 44 |
+
"loss": 4.3679,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8010680907877169,
|
| 49 |
+
"grad_norm": 10.019458770751953,
|
| 50 |
+
"learning_rate": 7.925531914893617e-06,
|
| 51 |
+
"loss": 4.3565,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9345794392523364,
|
| 56 |
+
"grad_norm": 14.457362174987793,
|
| 57 |
+
"learning_rate": 9.255319148936171e-06,
|
| 58 |
+
"loss": 4.3684,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.4441687344913151,
|
| 64 |
+
"eval_loss": 1.0651849508285522,
|
| 65 |
+
"eval_macro_f1": 0.3753786024955524,
|
| 66 |
+
"eval_runtime": 1.3533,
|
| 67 |
+
"eval_samples_per_second": 595.59,
|
| 68 |
+
"eval_steps_per_second": 37.686,
|
| 69 |
+
"eval_weighted_f1": 0.37753314368193,
|
| 70 |
+
"step": 188
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0640854472630175,
|
| 74 |
+
"grad_norm": 50.25746154785156,
|
| 75 |
+
"learning_rate": 1.0585106382978725e-05,
|
| 76 |
+
"loss": 4.0331,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.1975967957276368,
|
| 81 |
+
"grad_norm": 28.132944107055664,
|
| 82 |
+
"learning_rate": 1.1914893617021277e-05,
|
| 83 |
+
"loss": 3.9971,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3311081441922563,
|
| 88 |
+
"grad_norm": 26.1246337890625,
|
| 89 |
+
"learning_rate": 1.3244680851063832e-05,
|
| 90 |
+
"loss": 3.6115,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4646194926568759,
|
| 95 |
+
"grad_norm": 35.02851867675781,
|
| 96 |
+
"learning_rate": 1.4574468085106384e-05,
|
| 97 |
+
"loss": 3.2223,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.5981308411214954,
|
| 102 |
+
"grad_norm": 63.445247650146484,
|
| 103 |
+
"learning_rate": 1.590425531914894e-05,
|
| 104 |
+
"loss": 2.9527,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7316421895861147,
|
| 109 |
+
"grad_norm": 35.380924224853516,
|
| 110 |
+
"learning_rate": 1.723404255319149e-05,
|
| 111 |
+
"loss": 2.8674,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.8651535380507345,
|
| 116 |
+
"grad_norm": 64.34720611572266,
|
| 117 |
+
"learning_rate": 1.8563829787234043e-05,
|
| 118 |
+
"loss": 2.4821,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.9986648865153538,
|
| 123 |
+
"grad_norm": 48.913997650146484,
|
| 124 |
+
"learning_rate": 1.9893617021276595e-05,
|
| 125 |
+
"loss": 2.7957,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.0,
|
| 130 |
+
"eval_accuracy": 0.7419354838709677,
|
| 131 |
+
"eval_loss": 0.6129544377326965,
|
| 132 |
+
"eval_macro_f1": 0.7391019638003171,
|
| 133 |
+
"eval_runtime": 1.4116,
|
| 134 |
+
"eval_samples_per_second": 570.974,
|
| 135 |
+
"eval_steps_per_second": 36.129,
|
| 136 |
+
"eval_weighted_f1": 0.7390724544145634,
|
| 137 |
+
"step": 376
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.128170894526035,
|
| 141 |
+
"grad_norm": 39.42140197753906,
|
| 142 |
+
"learning_rate": 1.986406619385343e-05,
|
| 143 |
+
"loss": 2.1604,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.2616822429906542,
|
| 148 |
+
"grad_norm": 156.8961944580078,
|
| 149 |
+
"learning_rate": 1.971631205673759e-05,
|
| 150 |
+
"loss": 1.9432,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.3951935914552736,
|
| 155 |
+
"grad_norm": 44.37391662597656,
|
| 156 |
+
"learning_rate": 1.956855791962175e-05,
|
| 157 |
+
"loss": 2.2598,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.5287049399198933,
|
| 162 |
+
"grad_norm": 27.17721939086914,
|
| 163 |
+
"learning_rate": 1.9420803782505913e-05,
|
| 164 |
+
"loss": 1.9138,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.6622162883845126,
|
| 169 |
+
"grad_norm": 54.70370864868164,
|
| 170 |
+
"learning_rate": 1.927304964539007e-05,
|
| 171 |
+
"loss": 2.0825,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.795727636849132,
|
| 176 |
+
"grad_norm": 40.60383224487305,
|
| 177 |
+
"learning_rate": 1.9125295508274234e-05,
|
| 178 |
+
"loss": 1.9968,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.9292389853137517,
|
| 183 |
+
"grad_norm": 68.17862701416016,
|
| 184 |
+
"learning_rate": 1.8977541371158394e-05,
|
| 185 |
+
"loss": 2.1597,
|
| 186 |
+
"step": 550
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 3.0,
|
| 190 |
+
"eval_accuracy": 0.8002481389578163,
|
| 191 |
+
"eval_loss": 0.48578470945358276,
|
| 192 |
+
"eval_macro_f1": 0.797633914123045,
|
| 193 |
+
"eval_runtime": 1.3433,
|
| 194 |
+
"eval_samples_per_second": 600.001,
|
| 195 |
+
"eval_steps_per_second": 37.965,
|
| 196 |
+
"eval_weighted_f1": 0.7976918008957689,
|
| 197 |
+
"step": 564
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.0587449933244324,
|
| 201 |
+
"grad_norm": 66.7054214477539,
|
| 202 |
+
"learning_rate": 1.8829787234042557e-05,
|
| 203 |
+
"loss": 1.8278,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.192256341789052,
|
| 208 |
+
"grad_norm": 79.62871551513672,
|
| 209 |
+
"learning_rate": 1.8682033096926714e-05,
|
| 210 |
+
"loss": 1.2467,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.3257676902536715,
|
| 215 |
+
"grad_norm": 44.97430419921875,
|
| 216 |
+
"learning_rate": 1.8534278959810878e-05,
|
| 217 |
+
"loss": 1.3405,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4592790387182912,
|
| 222 |
+
"grad_norm": 42.74566650390625,
|
| 223 |
+
"learning_rate": 1.8386524822695038e-05,
|
| 224 |
+
"loss": 1.6122,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.5927903871829105,
|
| 229 |
+
"grad_norm": 21.341346740722656,
|
| 230 |
+
"learning_rate": 1.8238770685579198e-05,
|
| 231 |
+
"loss": 1.6088,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.72630173564753,
|
| 236 |
+
"grad_norm": 41.35243606567383,
|
| 237 |
+
"learning_rate": 1.8091016548463358e-05,
|
| 238 |
+
"loss": 1.3344,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.8598130841121496,
|
| 243 |
+
"grad_norm": 87.04261016845703,
|
| 244 |
+
"learning_rate": 1.7943262411347518e-05,
|
| 245 |
+
"loss": 1.3879,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.993324432576769,
|
| 250 |
+
"grad_norm": 35.48392105102539,
|
| 251 |
+
"learning_rate": 1.779550827423168e-05,
|
| 252 |
+
"loss": 1.3266,
|
| 253 |
+
"step": 750
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 4.0,
|
| 257 |
+
"eval_accuracy": 0.8114143920595533,
|
| 258 |
+
"eval_loss": 0.5030881762504578,
|
| 259 |
+
"eval_macro_f1": 0.8106581838763844,
|
| 260 |
+
"eval_runtime": 1.3801,
|
| 261 |
+
"eval_samples_per_second": 584.003,
|
| 262 |
+
"eval_steps_per_second": 36.953,
|
| 263 |
+
"eval_weighted_f1": 0.810642469392529,
|
| 264 |
+
"step": 752
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.12283044058745,
|
| 268 |
+
"grad_norm": 22.939712524414062,
|
| 269 |
+
"learning_rate": 1.7647754137115842e-05,
|
| 270 |
+
"loss": 0.8932,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.25634178905207,
|
| 275 |
+
"grad_norm": 45.860713958740234,
|
| 276 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 277 |
+
"loss": 1.0112,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.389853137516689,
|
| 282 |
+
"grad_norm": 77.18540954589844,
|
| 283 |
+
"learning_rate": 1.7352245862884162e-05,
|
| 284 |
+
"loss": 1.06,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.5233644859813085,
|
| 289 |
+
"grad_norm": 34.39630889892578,
|
| 290 |
+
"learning_rate": 1.7204491725768322e-05,
|
| 291 |
+
"loss": 0.9128,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.656875834445928,
|
| 296 |
+
"grad_norm": 54.38111877441406,
|
| 297 |
+
"learning_rate": 1.7056737588652482e-05,
|
| 298 |
+
"loss": 0.7028,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.790387182910547,
|
| 303 |
+
"grad_norm": 86.46287536621094,
|
| 304 |
+
"learning_rate": 1.6908983451536646e-05,
|
| 305 |
+
"loss": 1.0283,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.923898531375167,
|
| 310 |
+
"grad_norm": 39.945159912109375,
|
| 311 |
+
"learning_rate": 1.6761229314420806e-05,
|
| 312 |
+
"loss": 0.9587,
|
| 313 |
+
"step": 925
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 5.0,
|
| 317 |
+
"eval_accuracy": 0.815136476426799,
|
| 318 |
+
"eval_loss": 0.5337904691696167,
|
| 319 |
+
"eval_macro_f1": 0.8152641683112084,
|
| 320 |
+
"eval_runtime": 1.3407,
|
| 321 |
+
"eval_samples_per_second": 601.193,
|
| 322 |
+
"eval_steps_per_second": 38.041,
|
| 323 |
+
"eval_weighted_f1": 0.8152093741213681,
|
| 324 |
+
"step": 940
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.0534045393858475,
|
| 328 |
+
"grad_norm": 58.76649475097656,
|
| 329 |
+
"learning_rate": 1.6613475177304966e-05,
|
| 330 |
+
"loss": 0.9582,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.186915887850467,
|
| 335 |
+
"grad_norm": 17.8238525390625,
|
| 336 |
+
"learning_rate": 1.6465721040189126e-05,
|
| 337 |
+
"loss": 0.6535,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.320427236315087,
|
| 342 |
+
"grad_norm": 33.73042678833008,
|
| 343 |
+
"learning_rate": 1.6317966903073286e-05,
|
| 344 |
+
"loss": 0.4935,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.453938584779706,
|
| 349 |
+
"grad_norm": 25.605831146240234,
|
| 350 |
+
"learning_rate": 1.6170212765957446e-05,
|
| 351 |
+
"loss": 0.669,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.587449933244326,
|
| 356 |
+
"grad_norm": 36.81806564331055,
|
| 357 |
+
"learning_rate": 1.602245862884161e-05,
|
| 358 |
+
"loss": 0.7954,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.720961281708945,
|
| 363 |
+
"grad_norm": 62.5572509765625,
|
| 364 |
+
"learning_rate": 1.587470449172577e-05,
|
| 365 |
+
"loss": 0.5398,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.854472630173564,
|
| 370 |
+
"grad_norm": 37.748619079589844,
|
| 371 |
+
"learning_rate": 1.572695035460993e-05,
|
| 372 |
+
"loss": 0.6565,
|
| 373 |
+
"step": 1100
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.9879839786381845,
|
| 377 |
+
"grad_norm": 64.0909423828125,
|
| 378 |
+
"learning_rate": 1.557919621749409e-05,
|
| 379 |
+
"loss": 0.7397,
|
| 380 |
+
"step": 1125
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.0,
|
| 384 |
+
"eval_accuracy": 0.7990074441687345,
|
| 385 |
+
"eval_loss": 0.6556637287139893,
|
| 386 |
+
"eval_macro_f1": 0.7999806740506581,
|
| 387 |
+
"eval_runtime": 1.3755,
|
| 388 |
+
"eval_samples_per_second": 585.968,
|
| 389 |
+
"eval_steps_per_second": 37.077,
|
| 390 |
+
"eval_weighted_f1": 0.7995959163150099,
|
| 391 |
+
"step": 1128
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.117489986648865,
|
| 395 |
+
"grad_norm": 30.407459259033203,
|
| 396 |
+
"learning_rate": 1.543144208037825e-05,
|
| 397 |
+
"loss": 0.3339,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.251001335113485,
|
| 402 |
+
"grad_norm": 5.074549674987793,
|
| 403 |
+
"learning_rate": 1.5283687943262414e-05,
|
| 404 |
+
"loss": 0.5162,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.384512683578104,
|
| 409 |
+
"grad_norm": 3.213144302368164,
|
| 410 |
+
"learning_rate": 1.5135933806146574e-05,
|
| 411 |
+
"loss": 0.5794,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.518024032042724,
|
| 416 |
+
"grad_norm": 148.7412872314453,
|
| 417 |
+
"learning_rate": 1.4988179669030734e-05,
|
| 418 |
+
"loss": 0.8456,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.651535380507343,
|
| 423 |
+
"grad_norm": 78.37199401855469,
|
| 424 |
+
"learning_rate": 1.4840425531914894e-05,
|
| 425 |
+
"loss": 0.4425,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.785046728971962,
|
| 430 |
+
"grad_norm": 31.29233741760254,
|
| 431 |
+
"learning_rate": 1.4692671394799056e-05,
|
| 432 |
+
"loss": 0.5563,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 6.9185580774365825,
|
| 437 |
+
"grad_norm": 69.88211059570312,
|
| 438 |
+
"learning_rate": 1.4544917257683215e-05,
|
| 439 |
+
"loss": 0.4397,
|
| 440 |
+
"step": 1300
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 7.0,
|
| 444 |
+
"eval_accuracy": 0.8176178660049628,
|
| 445 |
+
"eval_loss": 0.8486976623535156,
|
| 446 |
+
"eval_macro_f1": 0.8166522074416811,
|
| 447 |
+
"eval_runtime": 1.3509,
|
| 448 |
+
"eval_samples_per_second": 596.62,
|
| 449 |
+
"eval_steps_per_second": 37.751,
|
| 450 |
+
"eval_weighted_f1": 0.8166431779008507,
|
| 451 |
+
"step": 1316
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.048064085447263,
|
| 455 |
+
"grad_norm": 3.868504762649536,
|
| 456 |
+
"learning_rate": 1.4397163120567377e-05,
|
| 457 |
+
"loss": 0.2631,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.181575433911883,
|
| 462 |
+
"grad_norm": 1.3365354537963867,
|
| 463 |
+
"learning_rate": 1.4249408983451538e-05,
|
| 464 |
+
"loss": 0.3103,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.315086782376502,
|
| 469 |
+
"grad_norm": 123.04847717285156,
|
| 470 |
+
"learning_rate": 1.41016548463357e-05,
|
| 471 |
+
"loss": 0.2794,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.4485981308411215,
|
| 476 |
+
"grad_norm": 79.79776763916016,
|
| 477 |
+
"learning_rate": 1.3953900709219859e-05,
|
| 478 |
+
"loss": 0.2952,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.582109479305741,
|
| 483 |
+
"grad_norm": 54.26392364501953,
|
| 484 |
+
"learning_rate": 1.380614657210402e-05,
|
| 485 |
+
"loss": 0.531,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.71562082777036,
|
| 490 |
+
"grad_norm": 168.4088134765625,
|
| 491 |
+
"learning_rate": 1.365839243498818e-05,
|
| 492 |
+
"loss": 0.599,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 7.8491321762349795,
|
| 497 |
+
"grad_norm": 128.00387573242188,
|
| 498 |
+
"learning_rate": 1.351063829787234e-05,
|
| 499 |
+
"loss": 0.5645,
|
| 500 |
+
"step": 1475
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 7.9826435246996,
|
| 504 |
+
"grad_norm": 8.781241416931152,
|
| 505 |
+
"learning_rate": 1.3362884160756503e-05,
|
| 506 |
+
"loss": 0.3614,
|
| 507 |
+
"step": 1500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 8.0,
|
| 511 |
+
"eval_accuracy": 0.826302729528536,
|
| 512 |
+
"eval_loss": 0.9394305348396301,
|
| 513 |
+
"eval_macro_f1": 0.8262771676591104,
|
| 514 |
+
"eval_runtime": 1.3398,
|
| 515 |
+
"eval_samples_per_second": 601.583,
|
| 516 |
+
"eval_steps_per_second": 38.065,
|
| 517 |
+
"eval_weighted_f1": 0.8261771474759725,
|
| 518 |
+
"step": 1504
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.11214953271028,
|
| 522 |
+
"grad_norm": 131.1908721923828,
|
| 523 |
+
"learning_rate": 1.3215130023640663e-05,
|
| 524 |
+
"loss": 0.4791,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.2456608811749,
|
| 529 |
+
"grad_norm": 4.451777935028076,
|
| 530 |
+
"learning_rate": 1.3067375886524825e-05,
|
| 531 |
+
"loss": 0.2627,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.379172229639519,
|
| 536 |
+
"grad_norm": 8.379265785217285,
|
| 537 |
+
"learning_rate": 1.2919621749408983e-05,
|
| 538 |
+
"loss": 0.5037,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.51268357810414,
|
| 543 |
+
"grad_norm": 254.3783416748047,
|
| 544 |
+
"learning_rate": 1.2771867612293145e-05,
|
| 545 |
+
"loss": 0.3598,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.646194926568759,
|
| 550 |
+
"grad_norm": 1.897127389907837,
|
| 551 |
+
"learning_rate": 1.2624113475177307e-05,
|
| 552 |
+
"loss": 0.2337,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 8.779706275033378,
|
| 557 |
+
"grad_norm": 5.200291633605957,
|
| 558 |
+
"learning_rate": 1.2476359338061469e-05,
|
| 559 |
+
"loss": 0.226,
|
| 560 |
+
"step": 1650
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 8.913217623497998,
|
| 564 |
+
"grad_norm": 0.13330727815628052,
|
| 565 |
+
"learning_rate": 1.2328605200945627e-05,
|
| 566 |
+
"loss": 0.2286,
|
| 567 |
+
"step": 1675
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 9.0,
|
| 571 |
+
"eval_accuracy": 0.8225806451612904,
|
| 572 |
+
"eval_loss": 1.0447702407836914,
|
| 573 |
+
"eval_macro_f1": 0.8239610502521302,
|
| 574 |
+
"eval_runtime": 1.3237,
|
| 575 |
+
"eval_samples_per_second": 608.91,
|
| 576 |
+
"eval_steps_per_second": 38.529,
|
| 577 |
+
"eval_weighted_f1": 0.8238013412706735,
|
| 578 |
+
"step": 1692
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.042723631508679,
|
| 582 |
+
"grad_norm": 1.7377159595489502,
|
| 583 |
+
"learning_rate": 1.2180851063829789e-05,
|
| 584 |
+
"loss": 0.1298,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.176234979973298,
|
| 589 |
+
"grad_norm": 0.25821366906166077,
|
| 590 |
+
"learning_rate": 1.2033096926713949e-05,
|
| 591 |
+
"loss": 0.1932,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.309746328437917,
|
| 596 |
+
"grad_norm": 126.08575439453125,
|
| 597 |
+
"learning_rate": 1.1885342789598109e-05,
|
| 598 |
+
"loss": 0.1659,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.443257676902537,
|
| 603 |
+
"grad_norm": 15.046143531799316,
|
| 604 |
+
"learning_rate": 1.1737588652482271e-05,
|
| 605 |
+
"loss": 0.2783,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.576769025367156,
|
| 610 |
+
"grad_norm": 3.01290225982666,
|
| 611 |
+
"learning_rate": 1.1589834515366431e-05,
|
| 612 |
+
"loss": 0.1438,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 9.710280373831775,
|
| 617 |
+
"grad_norm": 0.2535240650177002,
|
| 618 |
+
"learning_rate": 1.1442080378250593e-05,
|
| 619 |
+
"loss": 0.292,
|
| 620 |
+
"step": 1825
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 9.843791722296395,
|
| 624 |
+
"grad_norm": 0.08549287170171738,
|
| 625 |
+
"learning_rate": 1.1294326241134751e-05,
|
| 626 |
+
"loss": 0.2151,
|
| 627 |
+
"step": 1850
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 9.977303070761014,
|
| 631 |
+
"grad_norm": 1.3324421644210815,
|
| 632 |
+
"learning_rate": 1.1146572104018913e-05,
|
| 633 |
+
"loss": 0.2083,
|
| 634 |
+
"step": 1875
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 10.0,
|
| 638 |
+
"eval_accuracy": 0.8325062034739454,
|
| 639 |
+
"eval_loss": 1.0802310705184937,
|
| 640 |
+
"eval_macro_f1": 0.8325364371612837,
|
| 641 |
+
"eval_runtime": 1.2717,
|
| 642 |
+
"eval_samples_per_second": 633.792,
|
| 643 |
+
"eval_steps_per_second": 40.103,
|
| 644 |
+
"eval_weighted_f1": 0.8325432770935196,
|
| 645 |
+
"step": 1880
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.106809078771695,
|
| 649 |
+
"grad_norm": 2.575378894805908,
|
| 650 |
+
"learning_rate": 1.0998817966903075e-05,
|
| 651 |
+
"loss": 0.2403,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.240320427236314,
|
| 656 |
+
"grad_norm": 0.07811494171619415,
|
| 657 |
+
"learning_rate": 1.0851063829787233e-05,
|
| 658 |
+
"loss": 0.2612,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.373831775700934,
|
| 663 |
+
"grad_norm": 133.27186584472656,
|
| 664 |
+
"learning_rate": 1.0703309692671395e-05,
|
| 665 |
+
"loss": 0.278,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.507343124165555,
|
| 670 |
+
"grad_norm": 15.9158296585083,
|
| 671 |
+
"learning_rate": 1.0555555555555557e-05,
|
| 672 |
+
"loss": 0.1647,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.640854472630174,
|
| 677 |
+
"grad_norm": 44.299591064453125,
|
| 678 |
+
"learning_rate": 1.0407801418439717e-05,
|
| 679 |
+
"loss": 0.1822,
|
| 680 |
+
"step": 2000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 10.774365821094793,
|
| 684 |
+
"grad_norm": 15.27917194366455,
|
| 685 |
+
"learning_rate": 1.0260047281323877e-05,
|
| 686 |
+
"loss": 0.1564,
|
| 687 |
+
"step": 2025
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 10.907877169559413,
|
| 691 |
+
"grad_norm": 0.06366876512765884,
|
| 692 |
+
"learning_rate": 1.011229314420804e-05,
|
| 693 |
+
"loss": 0.1086,
|
| 694 |
+
"step": 2050
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 11.0,
|
| 698 |
+
"eval_accuracy": 0.8163771712158809,
|
| 699 |
+
"eval_loss": 1.1790789365768433,
|
| 700 |
+
"eval_macro_f1": 0.8165124624461798,
|
| 701 |
+
"eval_runtime": 1.4019,
|
| 702 |
+
"eval_samples_per_second": 574.935,
|
| 703 |
+
"eval_steps_per_second": 36.379,
|
| 704 |
+
"eval_weighted_f1": 0.8165232548356002,
|
| 705 |
+
"step": 2068
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.037383177570094,
|
| 709 |
+
"grad_norm": 0.42240583896636963,
|
| 710 |
+
"learning_rate": 9.9645390070922e-06,
|
| 711 |
+
"loss": 0.1795,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.170894526034713,
|
| 716 |
+
"grad_norm": 0.15088433027267456,
|
| 717 |
+
"learning_rate": 9.81678486997636e-06,
|
| 718 |
+
"loss": 0.1384,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 11.304405874499333,
|
| 723 |
+
"grad_norm": 443.4081726074219,
|
| 724 |
+
"learning_rate": 9.66903073286052e-06,
|
| 725 |
+
"loss": 0.1449,
|
| 726 |
+
"step": 2125
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 11.437917222963952,
|
| 730 |
+
"grad_norm": 36.89760971069336,
|
| 731 |
+
"learning_rate": 9.521276595744681e-06,
|
| 732 |
+
"loss": 0.1484,
|
| 733 |
+
"step": 2150
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 11.571428571428571,
|
| 737 |
+
"grad_norm": 0.04213828220963478,
|
| 738 |
+
"learning_rate": 9.373522458628842e-06,
|
| 739 |
+
"loss": 0.1988,
|
| 740 |
+
"step": 2175
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 11.70493991989319,
|
| 744 |
+
"grad_norm": 30.746477127075195,
|
| 745 |
+
"learning_rate": 9.225768321513003e-06,
|
| 746 |
+
"loss": 0.1649,
|
| 747 |
+
"step": 2200
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 11.83845126835781,
|
| 751 |
+
"grad_norm": 0.22835691273212433,
|
| 752 |
+
"learning_rate": 9.078014184397164e-06,
|
| 753 |
+
"loss": 0.1346,
|
| 754 |
+
"step": 2225
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 11.97196261682243,
|
| 758 |
+
"grad_norm": 0.06718920171260834,
|
| 759 |
+
"learning_rate": 8.930260047281325e-06,
|
| 760 |
+
"loss": 0.0356,
|
| 761 |
+
"step": 2250
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 12.0,
|
| 765 |
+
"eval_accuracy": 0.8312655086848635,
|
| 766 |
+
"eval_loss": 1.1762371063232422,
|
| 767 |
+
"eval_macro_f1": 0.8313075534951636,
|
| 768 |
+
"eval_runtime": 1.3797,
|
| 769 |
+
"eval_samples_per_second": 584.195,
|
| 770 |
+
"eval_steps_per_second": 36.965,
|
| 771 |
+
"eval_weighted_f1": 0.8314224469566837,
|
| 772 |
+
"step": 2256
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.10146862483311,
|
| 776 |
+
"grad_norm": 0.20685747265815735,
|
| 777 |
+
"learning_rate": 8.782505910165485e-06,
|
| 778 |
+
"loss": 0.1086,
|
| 779 |
+
"step": 2275
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 12.23497997329773,
|
| 783 |
+
"grad_norm": 234.74868774414062,
|
| 784 |
+
"learning_rate": 8.634751773049647e-06,
|
| 785 |
+
"loss": 0.1674,
|
| 786 |
+
"step": 2300
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 12.368491321762349,
|
| 790 |
+
"grad_norm": 0.03646189719438553,
|
| 791 |
+
"learning_rate": 8.486997635933807e-06,
|
| 792 |
+
"loss": 0.1373,
|
| 793 |
+
"step": 2325
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 12.50200267022697,
|
| 797 |
+
"grad_norm": 0.16963981091976166,
|
| 798 |
+
"learning_rate": 8.339243498817968e-06,
|
| 799 |
+
"loss": 0.0834,
|
| 800 |
+
"step": 2350
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 12.63551401869159,
|
| 804 |
+
"grad_norm": 0.053275998681783676,
|
| 805 |
+
"learning_rate": 8.191489361702128e-06,
|
| 806 |
+
"loss": 0.0703,
|
| 807 |
+
"step": 2375
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 12.769025367156209,
|
| 811 |
+
"grad_norm": 0.034139759838581085,
|
| 812 |
+
"learning_rate": 8.04373522458629e-06,
|
| 813 |
+
"loss": 0.011,
|
| 814 |
+
"step": 2400
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 12.902536715620828,
|
| 818 |
+
"grad_norm": 0.09221059828996658,
|
| 819 |
+
"learning_rate": 7.89598108747045e-06,
|
| 820 |
+
"loss": 0.1343,
|
| 821 |
+
"step": 2425
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 13.0,
|
| 825 |
+
"eval_accuracy": 0.8176178660049628,
|
| 826 |
+
"eval_loss": 1.2706037759780884,
|
| 827 |
+
"eval_macro_f1": 0.817624857215489,
|
| 828 |
+
"eval_runtime": 1.4343,
|
| 829 |
+
"eval_samples_per_second": 561.944,
|
| 830 |
+
"eval_steps_per_second": 35.557,
|
| 831 |
+
"eval_weighted_f1": 0.8176586923110227,
|
| 832 |
+
"step": 2444
|
| 833 |
+
}
|
| 834 |
+
],
|
| 835 |
+
"logging_steps": 25,
|
| 836 |
+
"max_steps": 3760,
|
| 837 |
+
"num_input_tokens_seen": 0,
|
| 838 |
+
"num_train_epochs": 20,
|
| 839 |
+
"save_steps": 500,
|
| 840 |
+
"stateful_callbacks": {
|
| 841 |
+
"EarlyStoppingCallback": {
|
| 842 |
+
"args": {
|
| 843 |
+
"early_stopping_patience": 3,
|
| 844 |
+
"early_stopping_threshold": 0.0
|
| 845 |
+
},
|
| 846 |
+
"attributes": {
|
| 847 |
+
"early_stopping_patience_counter": 3
|
| 848 |
+
}
|
| 849 |
+
},
|
| 850 |
+
"TrainerControl": {
|
| 851 |
+
"args": {
|
| 852 |
+
"should_epoch_stop": false,
|
| 853 |
+
"should_evaluate": false,
|
| 854 |
+
"should_log": false,
|
| 855 |
+
"should_save": true,
|
| 856 |
+
"should_training_stop": true
|
| 857 |
+
},
|
| 858 |
+
"attributes": {}
|
| 859 |
+
}
|
| 860 |
+
},
|
| 861 |
+
"total_flos": 0.0,
|
| 862 |
+
"train_batch_size": 8,
|
| 863 |
+
"trial_name": null,
|
| 864 |
+
"trial_params": null
|
| 865 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/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-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/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-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225/trainer_state.json
ADDED
|
@@ -0,0 +1,463 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1225,
|
| 3 |
+
"best_metric": 0.823499495513782,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1225,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1430615164520744,
|
| 14 |
+
"grad_norm": 10.956339836120605,
|
| 15 |
+
"learning_rate": 1.3714285714285717e-06,
|
| 16 |
+
"loss": 4.4118,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2861230329041488,
|
| 21 |
+
"grad_norm": 10.109342575073242,
|
| 22 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 23 |
+
"loss": 4.4152,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4291845493562232,
|
| 28 |
+
"grad_norm": 10.964703559875488,
|
| 29 |
+
"learning_rate": 4.228571428571429e-06,
|
| 30 |
+
"loss": 4.4096,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5722460658082976,
|
| 35 |
+
"grad_norm": 11.63598346710205,
|
| 36 |
+
"learning_rate": 5.6571428571428576e-06,
|
| 37 |
+
"loss": 4.4094,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7153075822603719,
|
| 42 |
+
"grad_norm": 14.79238510131836,
|
| 43 |
+
"learning_rate": 7.085714285714286e-06,
|
| 44 |
+
"loss": 4.4251,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8583690987124464,
|
| 49 |
+
"grad_norm": 15.961112976074219,
|
| 50 |
+
"learning_rate": 8.514285714285715e-06,
|
| 51 |
+
"loss": 4.3877,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"grad_norm": 8.027030944824219,
|
| 57 |
+
"learning_rate": 9.942857142857145e-06,
|
| 58 |
+
"loss": 4.3078,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.45417789757412397,
|
| 64 |
+
"eval_loss": 1.0636950731277466,
|
| 65 |
+
"eval_macro_f1": 0.34917420611286004,
|
| 66 |
+
"eval_runtime": 1.5377,
|
| 67 |
+
"eval_samples_per_second": 482.55,
|
| 68 |
+
"eval_steps_per_second": 30.566,
|
| 69 |
+
"eval_weighted_f1": 0.3559217202787933,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1430615164520743,
|
| 74 |
+
"grad_norm": 16.617094039916992,
|
| 75 |
+
"learning_rate": 1.1371428571428573e-05,
|
| 76 |
+
"loss": 4.1219,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2861230329041489,
|
| 81 |
+
"grad_norm": 30.883926391601562,
|
| 82 |
+
"learning_rate": 1.2800000000000001e-05,
|
| 83 |
+
"loss": 4.0633,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4291845493562232,
|
| 88 |
+
"grad_norm": 30.19638442993164,
|
| 89 |
+
"learning_rate": 1.422857142857143e-05,
|
| 90 |
+
"loss": 3.7612,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5722460658082977,
|
| 95 |
+
"grad_norm": 25.96612548828125,
|
| 96 |
+
"learning_rate": 1.5657142857142856e-05,
|
| 97 |
+
"loss": 3.1714,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.715307582260372,
|
| 102 |
+
"grad_norm": 42.82233428955078,
|
| 103 |
+
"learning_rate": 1.708571428571429e-05,
|
| 104 |
+
"loss": 2.8573,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.8583690987124464,
|
| 109 |
+
"grad_norm": 69.23664855957031,
|
| 110 |
+
"learning_rate": 1.8514285714285716e-05,
|
| 111 |
+
"loss": 2.524,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"grad_norm": 52.727821350097656,
|
| 117 |
+
"learning_rate": 1.9942857142857142e-05,
|
| 118 |
+
"loss": 2.4357,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7318059299191375,
|
| 124 |
+
"eval_loss": 0.6728650331497192,
|
| 125 |
+
"eval_macro_f1": 0.7215115431703327,
|
| 126 |
+
"eval_runtime": 1.3037,
|
| 127 |
+
"eval_samples_per_second": 569.151,
|
| 128 |
+
"eval_steps_per_second": 36.051,
|
| 129 |
+
"eval_weighted_f1": 0.7228918433060562,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.1430615164520743,
|
| 134 |
+
"grad_norm": 27.7490234375,
|
| 135 |
+
"learning_rate": 1.984761904761905e-05,
|
| 136 |
+
"loss": 1.8818,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2861230329041486,
|
| 141 |
+
"grad_norm": 41.11962890625,
|
| 142 |
+
"learning_rate": 1.968888888888889e-05,
|
| 143 |
+
"loss": 1.9635,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.429184549356223,
|
| 148 |
+
"grad_norm": 41.880027770996094,
|
| 149 |
+
"learning_rate": 1.953015873015873e-05,
|
| 150 |
+
"loss": 1.824,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5722460658082977,
|
| 155 |
+
"grad_norm": 35.391849517822266,
|
| 156 |
+
"learning_rate": 1.937142857142857e-05,
|
| 157 |
+
"loss": 1.975,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.715307582260372,
|
| 162 |
+
"grad_norm": 47.064674377441406,
|
| 163 |
+
"learning_rate": 1.9212698412698414e-05,
|
| 164 |
+
"loss": 1.8885,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8583690987124464,
|
| 169 |
+
"grad_norm": 44.9041748046875,
|
| 170 |
+
"learning_rate": 1.9053968253968257e-05,
|
| 171 |
+
"loss": 1.871,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"grad_norm": 24.269550323486328,
|
| 177 |
+
"learning_rate": 1.8895238095238096e-05,
|
| 178 |
+
"loss": 1.8795,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.7722371967654986,
|
| 184 |
+
"eval_loss": 0.6149473190307617,
|
| 185 |
+
"eval_macro_f1": 0.7639556933946593,
|
| 186 |
+
"eval_runtime": 1.2075,
|
| 187 |
+
"eval_samples_per_second": 614.511,
|
| 188 |
+
"eval_steps_per_second": 38.925,
|
| 189 |
+
"eval_weighted_f1": 0.7647290763983028,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.1430615164520743,
|
| 194 |
+
"grad_norm": 35.27069091796875,
|
| 195 |
+
"learning_rate": 1.8736507936507936e-05,
|
| 196 |
+
"loss": 1.5831,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.2861230329041486,
|
| 201 |
+
"grad_norm": 75.83830261230469,
|
| 202 |
+
"learning_rate": 1.857777777777778e-05,
|
| 203 |
+
"loss": 1.3505,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.429184549356223,
|
| 208 |
+
"grad_norm": 26.643152236938477,
|
| 209 |
+
"learning_rate": 1.8419047619047622e-05,
|
| 210 |
+
"loss": 1.2001,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5722460658082977,
|
| 215 |
+
"grad_norm": 80.0782699584961,
|
| 216 |
+
"learning_rate": 1.8260317460317462e-05,
|
| 217 |
+
"loss": 1.2431,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.715307582260372,
|
| 222 |
+
"grad_norm": 68.80171966552734,
|
| 223 |
+
"learning_rate": 1.81015873015873e-05,
|
| 224 |
+
"loss": 1.1344,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.8583690987124464,
|
| 229 |
+
"grad_norm": 44.56772232055664,
|
| 230 |
+
"learning_rate": 1.7942857142857144e-05,
|
| 231 |
+
"loss": 1.0834,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"grad_norm": 54.323429107666016,
|
| 237 |
+
"learning_rate": 1.7784126984126987e-05,
|
| 238 |
+
"loss": 1.2617,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.8018867924528302,
|
| 244 |
+
"eval_loss": 0.5251675248146057,
|
| 245 |
+
"eval_macro_f1": 0.7965134593890665,
|
| 246 |
+
"eval_runtime": 1.4242,
|
| 247 |
+
"eval_samples_per_second": 521.007,
|
| 248 |
+
"eval_steps_per_second": 33.002,
|
| 249 |
+
"eval_weighted_f1": 0.7968561401034333,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.143061516452074,
|
| 254 |
+
"grad_norm": 51.52719497680664,
|
| 255 |
+
"learning_rate": 1.7625396825396827e-05,
|
| 256 |
+
"loss": 0.7874,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.286123032904149,
|
| 261 |
+
"grad_norm": 40.51311492919922,
|
| 262 |
+
"learning_rate": 1.7466666666666667e-05,
|
| 263 |
+
"loss": 0.9152,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.429184549356223,
|
| 268 |
+
"grad_norm": 31.555131912231445,
|
| 269 |
+
"learning_rate": 1.730793650793651e-05,
|
| 270 |
+
"loss": 0.7222,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.572246065808297,
|
| 275 |
+
"grad_norm": 76.65121459960938,
|
| 276 |
+
"learning_rate": 1.7149206349206353e-05,
|
| 277 |
+
"loss": 0.8278,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.715307582260372,
|
| 282 |
+
"grad_norm": 60.05152893066406,
|
| 283 |
+
"learning_rate": 1.6990476190476192e-05,
|
| 284 |
+
"loss": 0.7635,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.858369098712446,
|
| 289 |
+
"grad_norm": 71.91492462158203,
|
| 290 |
+
"learning_rate": 1.6831746031746032e-05,
|
| 291 |
+
"loss": 0.9925,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"grad_norm": 40.52027893066406,
|
| 297 |
+
"learning_rate": 1.6673015873015875e-05,
|
| 298 |
+
"loss": 0.983,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.0,
|
| 303 |
+
"eval_accuracy": 0.807277628032345,
|
| 304 |
+
"eval_loss": 0.5738729238510132,
|
| 305 |
+
"eval_macro_f1": 0.8050125481376144,
|
| 306 |
+
"eval_runtime": 1.265,
|
| 307 |
+
"eval_samples_per_second": 586.573,
|
| 308 |
+
"eval_steps_per_second": 37.155,
|
| 309 |
+
"eval_weighted_f1": 0.8052675572176005,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.143061516452074,
|
| 314 |
+
"grad_norm": 83.3038101196289,
|
| 315 |
+
"learning_rate": 1.6514285714285714e-05,
|
| 316 |
+
"loss": 0.5309,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.286123032904149,
|
| 321 |
+
"grad_norm": 27.02752685546875,
|
| 322 |
+
"learning_rate": 1.6355555555555557e-05,
|
| 323 |
+
"loss": 0.5224,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.429184549356223,
|
| 328 |
+
"grad_norm": 13.689112663269043,
|
| 329 |
+
"learning_rate": 1.6196825396825397e-05,
|
| 330 |
+
"loss": 0.5628,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.572246065808297,
|
| 335 |
+
"grad_norm": 12.405098915100098,
|
| 336 |
+
"learning_rate": 1.603809523809524e-05,
|
| 337 |
+
"loss": 0.4872,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.715307582260372,
|
| 342 |
+
"grad_norm": 14.685806274414062,
|
| 343 |
+
"learning_rate": 1.587936507936508e-05,
|
| 344 |
+
"loss": 0.3895,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.858369098712446,
|
| 349 |
+
"grad_norm": 27.080974578857422,
|
| 350 |
+
"learning_rate": 1.5720634920634923e-05,
|
| 351 |
+
"loss": 0.5834,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"grad_norm": 20.062631607055664,
|
| 357 |
+
"learning_rate": 1.5561904761904762e-05,
|
| 358 |
+
"loss": 0.5135,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.0,
|
| 363 |
+
"eval_accuracy": 0.8113207547169812,
|
| 364 |
+
"eval_loss": 0.8136313557624817,
|
| 365 |
+
"eval_macro_f1": 0.8100761397749455,
|
| 366 |
+
"eval_runtime": 1.2443,
|
| 367 |
+
"eval_samples_per_second": 596.324,
|
| 368 |
+
"eval_steps_per_second": 37.773,
|
| 369 |
+
"eval_weighted_f1": 0.8103175567646799,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.143061516452074,
|
| 374 |
+
"grad_norm": 75.22364807128906,
|
| 375 |
+
"learning_rate": 1.5403174603174605e-05,
|
| 376 |
+
"loss": 0.3682,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.286123032904149,
|
| 381 |
+
"grad_norm": 80.3740463256836,
|
| 382 |
+
"learning_rate": 1.5244444444444447e-05,
|
| 383 |
+
"loss": 0.4421,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.429184549356223,
|
| 388 |
+
"grad_norm": 6.37501335144043,
|
| 389 |
+
"learning_rate": 1.5085714285714288e-05,
|
| 390 |
+
"loss": 0.3636,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.572246065808297,
|
| 395 |
+
"grad_norm": 89.96634674072266,
|
| 396 |
+
"learning_rate": 1.4926984126984128e-05,
|
| 397 |
+
"loss": 0.412,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.715307582260372,
|
| 402 |
+
"grad_norm": 85.16874694824219,
|
| 403 |
+
"learning_rate": 1.4768253968253969e-05,
|
| 404 |
+
"loss": 0.4309,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.858369098712446,
|
| 409 |
+
"grad_norm": 67.8319091796875,
|
| 410 |
+
"learning_rate": 1.4609523809523812e-05,
|
| 411 |
+
"loss": 0.4097,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 7.0,
|
| 416 |
+
"grad_norm": 0.12423601746559143,
|
| 417 |
+
"learning_rate": 1.4450793650793653e-05,
|
| 418 |
+
"loss": 0.3431,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 7.0,
|
| 423 |
+
"eval_accuracy": 0.8247978436657682,
|
| 424 |
+
"eval_loss": 0.8630949258804321,
|
| 425 |
+
"eval_macro_f1": 0.823499495513782,
|
| 426 |
+
"eval_runtime": 1.2358,
|
| 427 |
+
"eval_samples_per_second": 600.399,
|
| 428 |
+
"eval_steps_per_second": 38.031,
|
| 429 |
+
"eval_weighted_f1": 0.8238051937850761,
|
| 430 |
+
"step": 1225
|
| 431 |
+
}
|
| 432 |
+
],
|
| 433 |
+
"logging_steps": 25,
|
| 434 |
+
"max_steps": 3500,
|
| 435 |
+
"num_input_tokens_seen": 0,
|
| 436 |
+
"num_train_epochs": 20,
|
| 437 |
+
"save_steps": 500,
|
| 438 |
+
"stateful_callbacks": {
|
| 439 |
+
"EarlyStoppingCallback": {
|
| 440 |
+
"args": {
|
| 441 |
+
"early_stopping_patience": 3,
|
| 442 |
+
"early_stopping_threshold": 0.0
|
| 443 |
+
},
|
| 444 |
+
"attributes": {
|
| 445 |
+
"early_stopping_patience_counter": 0
|
| 446 |
+
}
|
| 447 |
+
},
|
| 448 |
+
"TrainerControl": {
|
| 449 |
+
"args": {
|
| 450 |
+
"should_epoch_stop": false,
|
| 451 |
+
"should_evaluate": false,
|
| 452 |
+
"should_log": false,
|
| 453 |
+
"should_save": true,
|
| 454 |
+
"should_training_stop": false
|
| 455 |
+
},
|
| 456 |
+
"attributes": {}
|
| 457 |
+
}
|
| 458 |
+
},
|
| 459 |
+
"total_flos": 0.0,
|
| 460 |
+
"train_batch_size": 8,
|
| 461 |
+
"trial_name": null,
|
| 462 |
+
"trial_params": null
|
| 463 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/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-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/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-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1750/trainer_state.json
ADDED
|
@@ -0,0 +1,643 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1225,
|
| 3 |
+
"best_metric": 0.823499495513782,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-05-s44-20260828T205632Z-496d6871/trainer/checkpoint-1225",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1750,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.1430615164520744,
|
| 14 |
+
"grad_norm": 10.956339836120605,
|
| 15 |
+
"learning_rate": 1.3714285714285717e-06,
|
| 16 |
+
"loss": 4.4118,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2861230329041488,
|
| 21 |
+
"grad_norm": 10.109342575073242,
|
| 22 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 23 |
+
"loss": 4.4152,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4291845493562232,
|
| 28 |
+
"grad_norm": 10.964703559875488,
|
| 29 |
+
"learning_rate": 4.228571428571429e-06,
|
| 30 |
+
"loss": 4.4096,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5722460658082976,
|
| 35 |
+
"grad_norm": 11.63598346710205,
|
| 36 |
+
"learning_rate": 5.6571428571428576e-06,
|
| 37 |
+
"loss": 4.4094,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7153075822603719,
|
| 42 |
+
"grad_norm": 14.79238510131836,
|
| 43 |
+
"learning_rate": 7.085714285714286e-06,
|
| 44 |
+
"loss": 4.4251,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8583690987124464,
|
| 49 |
+
"grad_norm": 15.961112976074219,
|
| 50 |
+
"learning_rate": 8.514285714285715e-06,
|
| 51 |
+
"loss": 4.3877,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"grad_norm": 8.027030944824219,
|
| 57 |
+
"learning_rate": 9.942857142857145e-06,
|
| 58 |
+
"loss": 4.3078,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.45417789757412397,
|
| 64 |
+
"eval_loss": 1.0636950731277466,
|
| 65 |
+
"eval_macro_f1": 0.34917420611286004,
|
| 66 |
+
"eval_runtime": 1.5377,
|
| 67 |
+
"eval_samples_per_second": 482.55,
|
| 68 |
+
"eval_steps_per_second": 30.566,
|
| 69 |
+
"eval_weighted_f1": 0.3559217202787933,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1430615164520743,
|
| 74 |
+
"grad_norm": 16.617094039916992,
|
| 75 |
+
"learning_rate": 1.1371428571428573e-05,
|
| 76 |
+
"loss": 4.1219,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2861230329041489,
|
| 81 |
+
"grad_norm": 30.883926391601562,
|
| 82 |
+
"learning_rate": 1.2800000000000001e-05,
|
| 83 |
+
"loss": 4.0633,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4291845493562232,
|
| 88 |
+
"grad_norm": 30.19638442993164,
|
| 89 |
+
"learning_rate": 1.422857142857143e-05,
|
| 90 |
+
"loss": 3.7612,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5722460658082977,
|
| 95 |
+
"grad_norm": 25.96612548828125,
|
| 96 |
+
"learning_rate": 1.5657142857142856e-05,
|
| 97 |
+
"loss": 3.1714,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.715307582260372,
|
| 102 |
+
"grad_norm": 42.82233428955078,
|
| 103 |
+
"learning_rate": 1.708571428571429e-05,
|
| 104 |
+
"loss": 2.8573,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.8583690987124464,
|
| 109 |
+
"grad_norm": 69.23664855957031,
|
| 110 |
+
"learning_rate": 1.8514285714285716e-05,
|
| 111 |
+
"loss": 2.524,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"grad_norm": 52.727821350097656,
|
| 117 |
+
"learning_rate": 1.9942857142857142e-05,
|
| 118 |
+
"loss": 2.4357,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7318059299191375,
|
| 124 |
+
"eval_loss": 0.6728650331497192,
|
| 125 |
+
"eval_macro_f1": 0.7215115431703327,
|
| 126 |
+
"eval_runtime": 1.3037,
|
| 127 |
+
"eval_samples_per_second": 569.151,
|
| 128 |
+
"eval_steps_per_second": 36.051,
|
| 129 |
+
"eval_weighted_f1": 0.7228918433060562,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.1430615164520743,
|
| 134 |
+
"grad_norm": 27.7490234375,
|
| 135 |
+
"learning_rate": 1.984761904761905e-05,
|
| 136 |
+
"loss": 1.8818,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2861230329041486,
|
| 141 |
+
"grad_norm": 41.11962890625,
|
| 142 |
+
"learning_rate": 1.968888888888889e-05,
|
| 143 |
+
"loss": 1.9635,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.429184549356223,
|
| 148 |
+
"grad_norm": 41.880027770996094,
|
| 149 |
+
"learning_rate": 1.953015873015873e-05,
|
| 150 |
+
"loss": 1.824,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5722460658082977,
|
| 155 |
+
"grad_norm": 35.391849517822266,
|
| 156 |
+
"learning_rate": 1.937142857142857e-05,
|
| 157 |
+
"loss": 1.975,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.715307582260372,
|
| 162 |
+
"grad_norm": 47.064674377441406,
|
| 163 |
+
"learning_rate": 1.9212698412698414e-05,
|
| 164 |
+
"loss": 1.8885,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8583690987124464,
|
| 169 |
+
"grad_norm": 44.9041748046875,
|
| 170 |
+
"learning_rate": 1.9053968253968257e-05,
|
| 171 |
+
"loss": 1.871,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"grad_norm": 24.269550323486328,
|
| 177 |
+
"learning_rate": 1.8895238095238096e-05,
|
| 178 |
+
"loss": 1.8795,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.7722371967654986,
|
| 184 |
+
"eval_loss": 0.6149473190307617,
|
| 185 |
+
"eval_macro_f1": 0.7639556933946593,
|
| 186 |
+
"eval_runtime": 1.2075,
|
| 187 |
+
"eval_samples_per_second": 614.511,
|
| 188 |
+
"eval_steps_per_second": 38.925,
|
| 189 |
+
"eval_weighted_f1": 0.7647290763983028,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.1430615164520743,
|
| 194 |
+
"grad_norm": 35.27069091796875,
|
| 195 |
+
"learning_rate": 1.8736507936507936e-05,
|
| 196 |
+
"loss": 1.5831,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.2861230329041486,
|
| 201 |
+
"grad_norm": 75.83830261230469,
|
| 202 |
+
"learning_rate": 1.857777777777778e-05,
|
| 203 |
+
"loss": 1.3505,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.429184549356223,
|
| 208 |
+
"grad_norm": 26.643152236938477,
|
| 209 |
+
"learning_rate": 1.8419047619047622e-05,
|
| 210 |
+
"loss": 1.2001,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5722460658082977,
|
| 215 |
+
"grad_norm": 80.0782699584961,
|
| 216 |
+
"learning_rate": 1.8260317460317462e-05,
|
| 217 |
+
"loss": 1.2431,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.715307582260372,
|
| 222 |
+
"grad_norm": 68.80171966552734,
|
| 223 |
+
"learning_rate": 1.81015873015873e-05,
|
| 224 |
+
"loss": 1.1344,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.8583690987124464,
|
| 229 |
+
"grad_norm": 44.56772232055664,
|
| 230 |
+
"learning_rate": 1.7942857142857144e-05,
|
| 231 |
+
"loss": 1.0834,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"grad_norm": 54.323429107666016,
|
| 237 |
+
"learning_rate": 1.7784126984126987e-05,
|
| 238 |
+
"loss": 1.2617,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.8018867924528302,
|
| 244 |
+
"eval_loss": 0.5251675248146057,
|
| 245 |
+
"eval_macro_f1": 0.7965134593890665,
|
| 246 |
+
"eval_runtime": 1.4242,
|
| 247 |
+
"eval_samples_per_second": 521.007,
|
| 248 |
+
"eval_steps_per_second": 33.002,
|
| 249 |
+
"eval_weighted_f1": 0.7968561401034333,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.143061516452074,
|
| 254 |
+
"grad_norm": 51.52719497680664,
|
| 255 |
+
"learning_rate": 1.7625396825396827e-05,
|
| 256 |
+
"loss": 0.7874,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.286123032904149,
|
| 261 |
+
"grad_norm": 40.51311492919922,
|
| 262 |
+
"learning_rate": 1.7466666666666667e-05,
|
| 263 |
+
"loss": 0.9152,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.429184549356223,
|
| 268 |
+
"grad_norm": 31.555131912231445,
|
| 269 |
+
"learning_rate": 1.730793650793651e-05,
|
| 270 |
+
"loss": 0.7222,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.572246065808297,
|
| 275 |
+
"grad_norm": 76.65121459960938,
|
| 276 |
+
"learning_rate": 1.7149206349206353e-05,
|
| 277 |
+
"loss": 0.8278,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.715307582260372,
|
| 282 |
+
"grad_norm": 60.05152893066406,
|
| 283 |
+
"learning_rate": 1.6990476190476192e-05,
|
| 284 |
+
"loss": 0.7635,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.858369098712446,
|
| 289 |
+
"grad_norm": 71.91492462158203,
|
| 290 |
+
"learning_rate": 1.6831746031746032e-05,
|
| 291 |
+
"loss": 0.9925,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"grad_norm": 40.52027893066406,
|
| 297 |
+
"learning_rate": 1.6673015873015875e-05,
|
| 298 |
+
"loss": 0.983,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.0,
|
| 303 |
+
"eval_accuracy": 0.807277628032345,
|
| 304 |
+
"eval_loss": 0.5738729238510132,
|
| 305 |
+
"eval_macro_f1": 0.8050125481376144,
|
| 306 |
+
"eval_runtime": 1.265,
|
| 307 |
+
"eval_samples_per_second": 586.573,
|
| 308 |
+
"eval_steps_per_second": 37.155,
|
| 309 |
+
"eval_weighted_f1": 0.8052675572176005,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.143061516452074,
|
| 314 |
+
"grad_norm": 83.3038101196289,
|
| 315 |
+
"learning_rate": 1.6514285714285714e-05,
|
| 316 |
+
"loss": 0.5309,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.286123032904149,
|
| 321 |
+
"grad_norm": 27.02752685546875,
|
| 322 |
+
"learning_rate": 1.6355555555555557e-05,
|
| 323 |
+
"loss": 0.5224,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.429184549356223,
|
| 328 |
+
"grad_norm": 13.689112663269043,
|
| 329 |
+
"learning_rate": 1.6196825396825397e-05,
|
| 330 |
+
"loss": 0.5628,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.572246065808297,
|
| 335 |
+
"grad_norm": 12.405098915100098,
|
| 336 |
+
"learning_rate": 1.603809523809524e-05,
|
| 337 |
+
"loss": 0.4872,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.715307582260372,
|
| 342 |
+
"grad_norm": 14.685806274414062,
|
| 343 |
+
"learning_rate": 1.587936507936508e-05,
|
| 344 |
+
"loss": 0.3895,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.858369098712446,
|
| 349 |
+
"grad_norm": 27.080974578857422,
|
| 350 |
+
"learning_rate": 1.5720634920634923e-05,
|
| 351 |
+
"loss": 0.5834,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"grad_norm": 20.062631607055664,
|
| 357 |
+
"learning_rate": 1.5561904761904762e-05,
|
| 358 |
+
"loss": 0.5135,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.0,
|
| 363 |
+
"eval_accuracy": 0.8113207547169812,
|
| 364 |
+
"eval_loss": 0.8136313557624817,
|
| 365 |
+
"eval_macro_f1": 0.8100761397749455,
|
| 366 |
+
"eval_runtime": 1.2443,
|
| 367 |
+
"eval_samples_per_second": 596.324,
|
| 368 |
+
"eval_steps_per_second": 37.773,
|
| 369 |
+
"eval_weighted_f1": 0.8103175567646799,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.143061516452074,
|
| 374 |
+
"grad_norm": 75.22364807128906,
|
| 375 |
+
"learning_rate": 1.5403174603174605e-05,
|
| 376 |
+
"loss": 0.3682,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.286123032904149,
|
| 381 |
+
"grad_norm": 80.3740463256836,
|
| 382 |
+
"learning_rate": 1.5244444444444447e-05,
|
| 383 |
+
"loss": 0.4421,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.429184549356223,
|
| 388 |
+
"grad_norm": 6.37501335144043,
|
| 389 |
+
"learning_rate": 1.5085714285714288e-05,
|
| 390 |
+
"loss": 0.3636,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.572246065808297,
|
| 395 |
+
"grad_norm": 89.96634674072266,
|
| 396 |
+
"learning_rate": 1.4926984126984128e-05,
|
| 397 |
+
"loss": 0.412,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.715307582260372,
|
| 402 |
+
"grad_norm": 85.16874694824219,
|
| 403 |
+
"learning_rate": 1.4768253968253969e-05,
|
| 404 |
+
"loss": 0.4309,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.858369098712446,
|
| 409 |
+
"grad_norm": 67.8319091796875,
|
| 410 |
+
"learning_rate": 1.4609523809523812e-05,
|
| 411 |
+
"loss": 0.4097,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 7.0,
|
| 416 |
+
"grad_norm": 0.12423601746559143,
|
| 417 |
+
"learning_rate": 1.4450793650793653e-05,
|
| 418 |
+
"loss": 0.3431,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 7.0,
|
| 423 |
+
"eval_accuracy": 0.8247978436657682,
|
| 424 |
+
"eval_loss": 0.8630949258804321,
|
| 425 |
+
"eval_macro_f1": 0.823499495513782,
|
| 426 |
+
"eval_runtime": 1.2358,
|
| 427 |
+
"eval_samples_per_second": 600.399,
|
| 428 |
+
"eval_steps_per_second": 38.031,
|
| 429 |
+
"eval_weighted_f1": 0.8238051937850761,
|
| 430 |
+
"step": 1225
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.143061516452074,
|
| 434 |
+
"grad_norm": 172.74609375,
|
| 435 |
+
"learning_rate": 1.4292063492063493e-05,
|
| 436 |
+
"loss": 0.4064,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.286123032904149,
|
| 441 |
+
"grad_norm": 59.42573547363281,
|
| 442 |
+
"learning_rate": 1.4133333333333334e-05,
|
| 443 |
+
"loss": 0.4991,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.429184549356223,
|
| 448 |
+
"grad_norm": 168.42855834960938,
|
| 449 |
+
"learning_rate": 1.3974603174603177e-05,
|
| 450 |
+
"loss": 0.3294,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.572246065808297,
|
| 455 |
+
"grad_norm": 16.392112731933594,
|
| 456 |
+
"learning_rate": 1.3815873015873018e-05,
|
| 457 |
+
"loss": 0.2872,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.715307582260372,
|
| 462 |
+
"grad_norm": 103.35064697265625,
|
| 463 |
+
"learning_rate": 1.3657142857142858e-05,
|
| 464 |
+
"loss": 0.2906,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.858369098712446,
|
| 469 |
+
"grad_norm": 0.17929917573928833,
|
| 470 |
+
"learning_rate": 1.34984126984127e-05,
|
| 471 |
+
"loss": 0.1945,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.0,
|
| 476 |
+
"grad_norm": 47.55339431762695,
|
| 477 |
+
"learning_rate": 1.333968253968254e-05,
|
| 478 |
+
"loss": 0.2856,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 8.0,
|
| 483 |
+
"eval_accuracy": 0.8207547169811321,
|
| 484 |
+
"eval_loss": 0.9973414540290833,
|
| 485 |
+
"eval_macro_f1": 0.8205517189024981,
|
| 486 |
+
"eval_runtime": 1.2132,
|
| 487 |
+
"eval_samples_per_second": 611.584,
|
| 488 |
+
"eval_steps_per_second": 38.739,
|
| 489 |
+
"eval_weighted_f1": 0.8208615316307343,
|
| 490 |
+
"step": 1400
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.143061516452075,
|
| 494 |
+
"grad_norm": 0.5621415972709656,
|
| 495 |
+
"learning_rate": 1.318095238095238e-05,
|
| 496 |
+
"loss": 0.3534,
|
| 497 |
+
"step": 1425
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.286123032904149,
|
| 501 |
+
"grad_norm": 0.4190160930156708,
|
| 502 |
+
"learning_rate": 1.3022222222222223e-05,
|
| 503 |
+
"loss": 0.1991,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.429184549356224,
|
| 508 |
+
"grad_norm": 150.7726287841797,
|
| 509 |
+
"learning_rate": 1.2863492063492065e-05,
|
| 510 |
+
"loss": 0.2662,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.572246065808297,
|
| 515 |
+
"grad_norm": 1.0449289083480835,
|
| 516 |
+
"learning_rate": 1.2704761904761906e-05,
|
| 517 |
+
"loss": 0.2201,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.715307582260372,
|
| 522 |
+
"grad_norm": 0.37539392709732056,
|
| 523 |
+
"learning_rate": 1.2546031746031746e-05,
|
| 524 |
+
"loss": 0.4938,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.858369098712446,
|
| 529 |
+
"grad_norm": 0.39994290471076965,
|
| 530 |
+
"learning_rate": 1.2387301587301589e-05,
|
| 531 |
+
"loss": 0.4489,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 9.0,
|
| 536 |
+
"grad_norm": 0.1733744591474533,
|
| 537 |
+
"learning_rate": 1.222857142857143e-05,
|
| 538 |
+
"loss": 0.2276,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 9.0,
|
| 543 |
+
"eval_accuracy": 0.8180592991913747,
|
| 544 |
+
"eval_loss": 1.0497444868087769,
|
| 545 |
+
"eval_macro_f1": 0.8161564262378972,
|
| 546 |
+
"eval_runtime": 1.2017,
|
| 547 |
+
"eval_samples_per_second": 617.453,
|
| 548 |
+
"eval_steps_per_second": 39.111,
|
| 549 |
+
"eval_weighted_f1": 0.8167106421250762,
|
| 550 |
+
"step": 1575
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.143061516452075,
|
| 554 |
+
"grad_norm": 0.10080797970294952,
|
| 555 |
+
"learning_rate": 1.2069841269841271e-05,
|
| 556 |
+
"loss": 0.2141,
|
| 557 |
+
"step": 1600
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.286123032904149,
|
| 561 |
+
"grad_norm": 100.0721206665039,
|
| 562 |
+
"learning_rate": 1.191111111111111e-05,
|
| 563 |
+
"loss": 0.1143,
|
| 564 |
+
"step": 1625
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.429184549356224,
|
| 568 |
+
"grad_norm": 0.10907715559005737,
|
| 569 |
+
"learning_rate": 1.1752380952380952e-05,
|
| 570 |
+
"loss": 0.0757,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.572246065808297,
|
| 575 |
+
"grad_norm": 1.453877329826355,
|
| 576 |
+
"learning_rate": 1.1593650793650795e-05,
|
| 577 |
+
"loss": 0.2001,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.715307582260372,
|
| 582 |
+
"grad_norm": 72.37555694580078,
|
| 583 |
+
"learning_rate": 1.1434920634920636e-05,
|
| 584 |
+
"loss": 0.1595,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.858369098712446,
|
| 589 |
+
"grad_norm": 95.98368072509766,
|
| 590 |
+
"learning_rate": 1.1276190476190476e-05,
|
| 591 |
+
"loss": 0.1795,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 10.0,
|
| 596 |
+
"grad_norm": 39.84410095214844,
|
| 597 |
+
"learning_rate": 1.1117460317460317e-05,
|
| 598 |
+
"loss": 0.1793,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 10.0,
|
| 603 |
+
"eval_accuracy": 0.8099730458221024,
|
| 604 |
+
"eval_loss": 1.1723982095718384,
|
| 605 |
+
"eval_macro_f1": 0.8100042904069665,
|
| 606 |
+
"eval_runtime": 1.2471,
|
| 607 |
+
"eval_samples_per_second": 594.963,
|
| 608 |
+
"eval_steps_per_second": 37.686,
|
| 609 |
+
"eval_weighted_f1": 0.8100532475833168,
|
| 610 |
+
"step": 1750
|
| 611 |
+
}
|
| 612 |
+
],
|
| 613 |
+
"logging_steps": 25,
|
| 614 |
+
"max_steps": 3500,
|
| 615 |
+
"num_input_tokens_seen": 0,
|
| 616 |
+
"num_train_epochs": 20,
|
| 617 |
+
"save_steps": 500,
|
| 618 |
+
"stateful_callbacks": {
|
| 619 |
+
"EarlyStoppingCallback": {
|
| 620 |
+
"args": {
|
| 621 |
+
"early_stopping_patience": 3,
|
| 622 |
+
"early_stopping_threshold": 0.0
|
| 623 |
+
},
|
| 624 |
+
"attributes": {
|
| 625 |
+
"early_stopping_patience_counter": 3
|
| 626 |
+
}
|
| 627 |
+
},
|
| 628 |
+
"TrainerControl": {
|
| 629 |
+
"args": {
|
| 630 |
+
"should_epoch_stop": false,
|
| 631 |
+
"should_evaluate": false,
|
| 632 |
+
"should_log": false,
|
| 633 |
+
"should_save": true,
|
| 634 |
+
"should_training_stop": true
|
| 635 |
+
},
|
| 636 |
+
"attributes": {}
|
| 637 |
+
}
|
| 638 |
+
},
|
| 639 |
+
"total_flos": 0.0,
|
| 640 |
+
"train_batch_size": 8,
|
| 641 |
+
"trial_name": null,
|
| 642 |
+
"trial_params": null
|
| 643 |
+
}
|
ood-vifactcheck-bamibert-topic-domain-06-s44-20260828T210529Z-496d6871/trainer/checkpoint-920/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-vifactcheck-bamibert-topic-domain-06-s44-20260828T210529Z-496d6871/trainer/checkpoint-920/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 |
+
}
|